 |
|
| If you can't view the Datasheet, Please click here to try to view without PDF Reader . |
|
|
|
| If you can't view the Datasheet, Please click here to try to view without PDF Reader . |
|
|
| Datasheet File OCR Text: |
| Features * High Performance, Low Power AVR(R)32 UC 32-Bit Microcontroller - Compact Single-cycle RISC Instruction Set Including DSP Instruction Set - Read-Modify-Write Instructions and Atomic Bit Manipulation - Performing up to 1.39 DMIPS / MHz Up to 83 DMIPS Running at 60 MHz from Flash Up to 46 DMIPS Running at 30 MHz from Flash - Memory Protection Unit Multi-hierarchy Bus System - High-Performance Data Transfers on Separate Buses for Increased Performance - 7 Peripheral DMA Channels Improves Speed for Peripheral Communication Internal High-Speed Flash - 512K Bytes, 256K Bytes, 128K Bytes, 64K Bytes Versions - Single Cycle Access up to 30 MHz - Prefetch Buffer Optimizing Instruction Execution at Maximum Speed - 4ms Page Programming Time and 8ms Full-Chip Erase Time - 100,000 Write Cycles, 15-year Data Retention Capability - Flash Security Locks and User Defined Configuration Area Internal High-Speed SRAM, Single-Cycle Access at Full Speed - 96K Bytes (512KB Flash), 32K Bytes (256KB and 128KB Flash), 16K Bytes (64KB Flash) Interrupt Controller - Autovectored Low Latency Interrupt Service with Programmable Priority System Functions - Power and Clock Manager Including Internal RC Clock and One 32KHz Oscillator - Two Multipurpose Oscillators and Two Phase-Lock-Loop (PLL) allowing Independant CPU Frequency from USB Frequency - Watchdog Timer, Real-Time Clock Timer Universal Serial Bus (USB) - Device 2.0 and Embedded Host Low Speed and Full Speed - Flexible End-Point Configuration and Management with Dedicated DMA Channels - On-chip Transceivers Including Pull-Ups - USB Wake Up from Sleep Functionality One Three-Channel 16-bit Timer/Counter (TC) - Three External Clock Inputs, PWM, Capture and Various Counting Capabilities One 7-Channel 20-bit Pulse Width Modulation Controller (PWM) Three Universal Synchronous/Asynchronous Receiver/Transmitters (USART) - Independant Baudrate Generator, Support for SPI, IrDA and ISO7816 interfaces - Support for Hardware Handshaking, RS485 Interfaces and Modem Line One Master/Slave Serial Peripheral Interfaces (SPI) with Chip Select Signals One Synchronous Serial Protocol Controller - Supports I2S and Generic Frame-Based Protocols One Master/Slave Two-Wire Interface (TWI), 400kbit/s I2C-compatible One 8-channel 10-bit Analog-To-Digital Converter, 384ks/s 16-bit Stereo Audio Bitstream DAC - Sample Rate Up to 50 KHz QTouch(R) Library Support - Capacitive Touch Buttons, Sliders, and Wheels - QTouch(R) and QMatrix(R) Acquisition * * 32-bit AVR(R) Microcontroller AT32UC3B0512 AT32UC3B0256 AT32UC3B0128 AT32UC3B064 AT32UC3B1512 AT32UC3B1256 AT32UC3B1128 AT32UC3B164 Preliminary * * * * * * * * * * * * * 32059I-06/2010 AT32UC3B * On-Chip Debug System (JTAG interface) - Nexus Class 2+, Runtime Control, Non-Intrusive Data and Program Trace * 64-pin TQFP/QFN (44 GPIO pins), 48-pin TQFP/QFN (28 GPIO pins) * 5V Input Tolerant I/Os, including 4 high-drive pins * Single 3.3V Power Supply or Dual 1.8V-3.3V Power Supply 2 32059I-06/2010 AT32UC3B 1. Description The AT32UC3B is a complete System-On-Chip microcontroller based on the AVR32 UC RISC processor running at frequencies up to 60 MHz. AVR32 UC is a high-performance 32-bit RISC microprocessor core, designed for cost-sensitive embedded applications, with particular emphasis on low power consumption, high code density and high performance. The processor implements a Memory Protection Unit (MPU) and a fast and flexible interrupt controller for supporting modern operating systems and real-time operating systems. Higher computation capability is achieved using a rich set of DSP instructions. The AT32UC3B incorporates on-chip Flash and SRAM memories for secure and fast access. The Peripheral Direct Memory Access controller enables data transfers between peripherals and memories without processor involvement. PDCA drastically reduces processing overhead when transferring continuous and large data streams between modules within the MCU. The Power Manager improves design flexibility and security: the on-chip Brown-Out Detector monitors the power supply, the CPU runs from the on-chip RC oscillator or from one of external oscillator sources, a Real-Time Clock and its associated timer keeps track of the time. The Timer/Counter includes three identical 16-bit timer/counter channels. Each channel can be independently programmed to perform frequency measurement, event counting, interval measurement, pulse generation, delay timing and pulse width modulation. The PWM modules provides seven independent channels with many configuration options including polarity, edge alignment and waveform non overlap control. One PWM channel can trigger ADC conversions for more accurate close loop control implementations. The AT32UC3B also features many communication interfaces for communication intensive applications. In addition to standard serial interfaces like UART, SPI or TWI, other interfaces like flexible Synchronous Serial Controller and USB are available. The Synchronous Serial Controller provides easy access to serial communication protocols and audio standards like I2S, UART or SPI. The Full-Speed USB 2.0 Device interface supports several USB Classes at the same time thanks to the rich End-Point configuration. The Embedded Host interface allows device like a USB Flash disk or a USB printer to be directly connected to the processor. Atmel offers the QTouch library for embedding capacitive touch buttons, sliders, and wheels functionality into AVR microcontrollers. The patented charge-transfer signal acquisition offers robust sensing and included fully debounced reporting of touch keys and includes Adjacent Key Suppression(R) (AKS(R)) technology for unambiguous detection of key events. The easy-to-use QTouch Suite toolchain allows you to explore, develop, and debug your own touch applications. AT32UC3B integrates a class 2+ Nexus 2.0 On-Chip Debug (OCD) System, with non-intrusive real-time trace, full-speed read/write memory access in addition to basic runtime control. The Nanotrace interface enables trace feature for JTAG-based debuggers. 3 32059I-06/2010 AT32UC3B 2. Overview 2.1 Blockdiagram Block diagram TCK MEMORY INTERFACE TDO TDI TMS Figure 2-1. JTAG INTERFACE MCKO MDO[5..0] MSEO[1..0] EVTI_N EVTO_N LOCAL BUS INTERFACE FAST GPIO NEXUS CLASS 2+ OCD UC CPU MEMORY PROTECTION UNIT INSTR INTERFACE DATA INTERFACE 16/32/96 KB SRAM USB INTERFACE DMA S M S FLASH CONTROLLER VBUS D+ DID VBOF M M HIGH SPEED BUS MATRIX M S 64/128/ 256/512 KB FLASH S M S CONFIGURATION REGISTERS BUS HSB PB HSB HSB-PB BRIDGE B GENERAL PURPOSE IOs HSB-PB BRIDGE A PB PERIPHERAL DMA CONTROLLER PA PB EXTINT[7..0] KPS[7..0] NMI EXTERNAL INTERRUPT CONTROLLER REAL TIME COUNTER PDC USART1 RXD TXD CLK RTS, CTS DSR, DTR, DCD, RI RXD TXD CLK RTS, CTS SCK MISO, MOSI NPCS[3..0] GENERAL PURPOSE IOs INTERRUPT CONTROLLER PA PB USART0 USART2 PDC WATCHDOG TIMER 115 kHz RCOSC XIN32 XOUT32 XIN0 XOUT0 XIN1 XOUT1 SERIAL PERIPHERAL INTERFACE SYNCHRONOUS SERIAL CONTROLLER PDC TX_CLOCK, TX_FRAME_SYNC TX_DATA RX_CLOCK, RX_FRAME_SYNC RX_DATA POWER MANAGER CLOCK GENERATOR CLOCK CONTROLLER SLEEP CONTROLLER RESET CONTROLLER 32 KHz OSC OSC0 OSC1 PLL0 PLL1 GCLK[3..0] RESET_N A[2..0] B[2..0] CLK[2..0] PDC TWO-WIRE INTERFACE ANALOG TO DIGITAL CONVERTER AUDIO BITSTREAM DAC PULSE WIDTH MODULATION CONTROLLER PDC SCL SDA PDC AD[7..0] ADVREF PDC DATA[1..0] DATAN[1..0] TIMER/COUNTER PWM[6..0] 4 32059I-06/2010 AT32UC3B 3. Configuration Summary The table below lists all AT32UC3B memory and package configurations: Table 3-1. Device AT32UC3B0512 AT32UC3B0256 AT32UC3B0128 AT32UC3B064 AT32UC3B1512 AT32UC3B1256 AT32UC3B1128 AT32UC3B164 Flash 512 Kbytes 256 Kbytes 128 Kbytes 64 Kbytes 512 Kbytes 256 Kbytes 128 Kbytes 64 Kbytes Memory and Package Configurations SSC 1 1 1 1 0 0 0 0 ADC 8 8 8 8 6 6 6 6 ABDAC 1 0 0 0 1 0 0 0 OSC 2 2 2 2 1 1 1 1 USB Configuration Mini-Host + Device Mini-Host + Device Mini-Host + Device Mini-Host + Device Device Device Device Device Package 64 lead TQFP/QFN 64 lead TQFP/QFN 64 lead TQFP/QFN 64 lead TQFP/QFN 48 lead QFN 48 lead TQFP/QFN 48 lead TQFP/QFN 48 lead TQFP/QFN SRAM 96 Kbytes 32 Kbytes 32 Kbytes 16 Kbytes 96 Kbytes 32 Kbytes 16 Kbytes 16 Kbytes 5 32059I-06/2010 AT32UC3B 4. Package and Pinout 4.1 Package The device pins are multiplexed with peripheral functions as described in the Peripheral Multiplexing on I/O Line section. Figure 4-1. TQFP64 / QFN64 Pinout VDDIO PA23 PA22 PA21 PA20 PB07 PA29 PA28 PA19 PA18 PB06 PA17 PA16 PA15 PA14 PA13 GND DP DM VBUS VDDPLL PB08 PB09 VDDCORE PB10 PB11 PA24 PA25 PA26 PA27 RESET_N VDDIO 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 PA31 PA30 PA08 PA07 PA06 PA05 PA04 PA03 VDDCORE PB01 PB00 PA02 PA01 PA00 TCK GND 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 VDDIO PA12 PA11 PA10 PA09 PB05 PB04 PB03 PB02 GND VDDCORE VDDIN VDDOUT VDDANA ADVREF GNDANA 6 32059I-06/2010 AT32UC3B Figure 4-2. TQFP48 / QFN48 Pinout VDDIO PA23 PA22 PA21 PA20 PA19 PA18 PA17 PA16 PA15 PA14 PA13 GND DP DM VBUS VDDPLL VDDCORE PA24 PA25 PA26 PA27 RESET_N VDDIO 37 38 39 40 41 42 43 44 45 46 47 48 12 11 10 9 8 7 6 5 4 3 2 1 PA08 PA07 PA06 PA05 PA04 PA03 VDDCORE PA02 PA01 PA00 TCK GND 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 VDDIO PA12 PA11 PA10 PA09 GND VDDCORE VDDIN VDDOUT VDDANA ADVREF GNDANA Note: On QFN packages, the exposed pad is not connected to anything. 4.2 4.2.1 Peripheral Multiplexing on I/O lines Multiplexed signals Each GPIO line can be assigned to one of 4 peripheral functions; A, B, C or D (D is only available for UC3Bx512 parts). The following table define how the I/O lines on the peripherals A, B,C or D are multiplexed by the GPIO. Table 4-1. TQFP48 /QFN48 3 4 5 7 8 9 10 GPIO Controller Function Multiplexing TQFP64 /QFN64 3 4 5 9 10 11 12 Function D PIN PA00 PA01 PA02 PA03 PA04 PA05 PA06 GPIO Pin GPIO 0 GPIO 1 GPIO 2 GPIO 3 GPIO 4 GPIO 5 GPIO 6 ADC - AD[0] ADC - AD[1] EIC - EXTINT[0] EIC - EXTINT[1] PM - GCLK[0] PM - GCLK[1] ADC - AD[2] ADC - AD[3] USBB - USB_ID USBB - USB_VBOF USART1 - DCD USART1 - DSR ABDAC - DATA[0] ABDAC - DATAN[0] ABDAC - DATA[1] ABDAC - DATAN[1] Function A Function B Function C (only for UC3Bx512) 7 32059I-06/2010 AT32UC3B Table 4-1. 11 12 20 21 22 23 25 26 27 28 29 30 31 32 33 34 35 43 44 45 46 GPIO Controller Function Multiplexing 13 14 28 29 30 31 33 34 35 36 37 39 40 44 45 46 47 59 60 61 62 41 42 15 16 6 7 24 25 26 27 38 43 54 PA07 PA08 PA09 PA10 PA11 PA12 PA13 PA14 PA15 PA16 PA17 PA18 PA19 PA20 PA21 PA22 PA23 PA24 PA25 PA26 PA27 PA28 PA29 PA30 PA31 PB00 PB01 PB02 PB03 PB04 PB05 PB06 PB07 PB08 GPIO 7 GPIO 8 GPIO 9 GPIO 10 GPIO 11 GPIO 12 GPIO 13 GPIO 14 GPIO 15 GPIO 16 GPIO 17 GPIO 18 GPIO 19 GPIO 20 GPIO 21 GPIO 22 GPIO 23 GPIO 24 GPIO 25 GPIO 26 GPIO 27 GPIO 28 GPIO 29 GPIO 30 GPIO 31 GPIO 32 GPIO 33 GPIO 34 GPIO 35 GPIO 36 GPIO 37 GPIO 38 GPIO 39 GPIO 40 PWM - PWM[0] PWM - PWM[1] TWI - SCL TWI - SDA USART0 - RTS USART0 - CTS EIC - NMI SPI0 - MOSI SPI0 - SCK SPI0 - NPCS[0] SPI0 - NPCS[1] USART0 - RXD USART0 - TXD USART1 - CLK PWM - PWM[2] PWM - PWM[6] USART1 - TXD USART1 - RXD SPI0 - MISO USBB - USB_ID USBB - USB_VBOF USART0 - CLK TC - CLK0 ADC - AD[6] ADC - AD[7] TC - A0 TC - B0 EIC - EXTINT[6] EIC - EXTINT[7] USART1 - CTS USART1 - RTS SSC - RX_CLOCK SSC - RX_DATA SSC RX_FRAME_SYNC ADC - AD[4] ADC - AD[5] SPI0 - NPCS[2] SPI0 - NPCS[3] TC - A2 TC - B2 PWM - PWM[2] PWM - PWM[3] PWM - PWM[4] TC - CLK1 TC - CLK2 PWM - PWM[5] PWM - PWM[6] TC - CLK0 TC - A1 TC - B1 SPI0 - NPCS[1] SPI0 - NPCS[0] PWM - PWM[3] USART2 - TXD USART2 - RXD PWM - PWM[4] TC - CLK1 EIC - SCAN[0] EIC - SCAN[1] EIC - SCAN[2] EIC - SCAN[3] TC - A1 TC - B1 SPI0 - NPCS[3] SPI0 - NPCS[2] USART1 - DCD USART1 - DSR USART1 - DTR USART1 - DTR USART1 - RI USART1 - CTS USART1 - RTS PWM - PWM[0] PWM - PWM[1] USART0 - CLK EIC - EXTINT[2] USART2 - CLK PWM - PWM[4] SPI0 - SCK SPI0 - MISO SPI0 - MOSI USART2 - RXD USART2 - TXD ADC - TRIGGER EIC - EXTINT[3] EIC - EXTINT[4] EIC - EXTINT[5] TC - A0 TC - B0 SPI0 - MISO SPI0 - MOSI PM - GCLK[2] PWM - PWM[6] USART2 - CTS USART2 - RTS USART1 - TXD USART1 - RXD TC - CLK2 PWM - PWM[5] EIC - SCAN[4] EIC - SCAN[5] EIC - SCAN[6] ABDAC - DATA[0] ABDAC - DATAN[0] ABDAC - DATA[1] ABDAC - DATA[1] ABDAC - DATAN[1] ABDAC - DATAN[0] USART1 - RXD SSC RX_FRAME_SYNC SSC - TX_CLOCK SSC - TX_DATA SSC TX_FRAME_SYNC ABDAC - DATA[0] PWM - PWM[0] PWM - PWM[1] SSC - RX_DATA USART1 - TXD SSC - RX_CLOCK PM - GCLK[2] SSC RX_FRAME_SYNC SSC - RX_CLOCK 8 32059I-06/2010 AT32UC3B Table 4-1. GPIO Controller Function Multiplexing 55 57 58 PB09 PB10 PB11 GPIO 41 GPIO 42 GPIO 43 SSC - TX_CLOCK SSC - TX_DATA SSC TX_FRAME_SYNC USART1 - RI TC - A2 TC - B2 EIC - SCAN[7] USART0 - RXD USART0 - TXD ABDAC - DATAN[1] 4.2.2 JTAG Port Connections If the JTAG is enabled, the JTAG will take control over a number of pins, irrespective of the I/O Controller configuration. Table 4-2. 64QFP/QFN 2 3 4 5 JTAG Pinout 48QFP/QFN 2 3 4 5 Pin name TCK PA00 PA01 PA02 JTAG pin TCK TDI TDO TMS 4.2.3 Nexus OCD AUX port connections If the OCD trace system is enabled, the trace system will take control over a number of pins, irrespectively of the PIO configuration. Two different OCD trace pin mappings are possible, depending on the configuration of the OCD AXS register. For details, see the AVR32 UCTechnical Reference Manual. Table 4-3. Pin EVTI_N MDO[5] MDO[4] MDO[3] MDO[2] MDO[1] MDO[0] EVTO_N MCKO MSEO[1] MSEO[0] Nexus OCD AUX port connections AXS=0 PB05 PB04 PB03 PB02 PB01 PB00 PA31 PA15 PA30 PB06 PB07 AXS=1 PA14 PA08 PA07 PA06 PA05 PA04 PA03 PA15 PA13 PA09 PA10 4.2.4 Oscillator Pinout The oscillators are not mapped to the normal A, B or C functions and their muxings are controlled by registers in the Power Manager (PM). Please refer to the power manager chapter for more information about this. 9 32059I-06/2010 AT32UC3B Table 4-4. Oscillator pinout QFP64 pin 39 41 22 31 30 40 42 23 31 Pad PA18 PA28 PA11 PA19 PA29 PA12 Oscillator pin XIN0 XIN1 XIN32 XOUT0 XOUT1 XOUT32 QFP48 pin 30 4.3 High Drive Current GPIO Ones of GPIOs can be used to drive twice current than other GPIO capability (see Electrical Characteristics section). Table 4-5. High Drive Current GPIO GPIO Name PA20 PA21 PA22 PA23 5. Signals Description The following table gives details on the signal name classified by peripheral. Table 5-1. Signal Name Signal Description List Function Power Type Active Level Comments VDDPLL PLL Power Supply Power Input Power Input Power Input Power Input Power Input 1.65V to 1.95 V VDDCORE Core Power Supply 1.65V to 1.95 V VDDIO I/O Power Supply 3.0V to 3.6V VDDANA Analog Power Supply 3.0V to 3.6V VDDIN Voltage Regulator Input Supply 3.0V to 3.6V 10 32059I-06/2010 AT32UC3B Table 5-1. Signal Name VDDOUT GNDANA GND Signal Description List (Continued) Function Voltage Regulator Output Analog Ground Ground Type Power Output Ground Ground Clocks, Oscillators, and PLL's Active Level Comments 1.65V to 1.95 V XIN0, XIN1, XIN32 XOUT0, XOUT1, XOUT32 Crystal 0, 1, 32 Input Crystal 0, 1, 32 Output JTAG Analog Analog TCK TDI TDO TMS Test Clock Test Data In Test Data Out Test Mode Select Input Input Output Input Auxiliary Port - AUX MCKO MDO0 - MDO5 MSEO0 - MSEO1 EVTI_N EVTO_N Trace Data Output Clock Trace Data Output Trace Frame Control Event In Event Out Output Output Output Output Output Power Manager - PM Low Low GCLK0 - GCLK2 RESET_N Generic Clock Pins Reset Pin Output Input External Interrupt Controller - EIC Low EXTINT0 - EXTINT7 KPS0 - KPS7 NMI External Interrupt Pins Keypad Scan Pins Non-Maskable Interrupt Pin Input Output Input Low General Purpose I/O pin- GPIOA, GPIOB PA0 - PA31 PB0 - PB11 Parallel I/O Controller GPIOA Parallel I/O Controller GPIOB I/O I/O 11 32059I-06/2010 AT32UC3B Table 5-1. Signal Name Signal Description List (Continued) Function Type Serial Peripheral Interface - SPI0 Active Level Comments MISO MOSI NPCS0 - NPCS3 SCK Master In Slave Out Master Out Slave In SPI Peripheral Chip Select Clock I/O I/O I/O Output Synchronous Serial Controller - SSC Low RX_CLOCK RX_DATA RX_FRAME_SYNC TX_CLOCK TX_DATA TX_FRAME_SYNC SSC Receive Clock SSC Receive Data SSC Receive Frame Sync SSC Transmit Clock SSC Transmit Data SSC Transmit Frame Sync I/O Input I/O I/O Output I/O Timer/Counter - TIMER A0 A1 A2 B0 B1 B2 CLK0 CLK1 CLK2 Channel 0 Line A Channel 1 Line A Channel 2 Line A Channel 0 Line B Channel 1 Line B Channel 2 Line B Channel 0 External Clock Input Channel 1 External Clock Input Channel 2 External Clock Input I/O I/O I/O I/O I/O I/O Input Input Input Two-wire Interface - TWI SCL SDA Serial Clock Serial Data I/O I/O Universal Synchronous Asynchronous Receiver Transmitter - USART0, USART1, USART2 CLK CTS Clock Clear To Send I/O Input 12 32059I-06/2010 AT32UC3B Table 5-1. Signal Name DCD DSR DTR RI RTS RXD TXD Signal Description List (Continued) Function Data Carrier Detect Data Set Ready Data Terminal Ready Ring Indicator Request To Send Receive Data Transmit Data Output Input Output Analog to Digital Converter - ADC Type Active Level Comments Only USART1 Only USART1 Only USART1 Only USART1 AD0 - AD7 Analog input pins Analog input Analog input 2.6 to 3.6V ADVREF Analog positive reference voltage input Audio Bitstream DAC - ABDAC DATA0 - DATA1 DATAN0 - DATAN1 D/A Data out D/A Data inverted out Output Output Pulse Width Modulator - PWM PWM0 - PWM6 PWM Output Pins Output Universal Serial Bus Device - USBB DDM DDP VBUS USBID USB_VBOF USB Device Port Data USB Device Port Data + USB VBUS Monitor and Embedded Host Negotiation ID Pin of the USB Bus USB VBUS On/off: bus power control port Analog Analog Analog Input Input output 5.1 JTAG pins TMS and TDI pins have pull-up resistors. TDO pin is an output, driven at up to VDDIO, and has no pull-up resistor. These 3 pins can be used as GPIO-pins. At reset state, these pins are in GPIO mode. TCK pin cannot be used as GPIO pin. JTAG interface is enabled when TCK pin is tied low. 13 32059I-06/2010 AT32UC3B 5.2 RESET_N pin The RESET_N pin is a schmitt input and integrates a permanent pull-up resistor to VDDIO. As the product integrates a power-on reset cell, the RESET_N pin can be left unconnected in case no reset from the system needs to be applied to the product. 5.3 TWI pins When these pins are used for TWI, the pins are open-drain outputs with slew-rate limitation and inputs with inputs with spike-filtering. When used as GPIO-pins or used for other peripherals, the pins have the same characteristics as GPIO pins. 5.4 GPIO pins All the I/O lines integrate a pull-up resistor. Programming of this pull-up resistor is performed independently for each I/O line through the GPIO Controllers. After reset, I/O lines default as inputs with pull-up resistors disabled, except when indicated otherwise in the column "Reset Value" of the GPIO Controller user interface table. 5.5 High drive pins The four pins PA20, PA21, PA22, PA23 have high drive output capabilities. 5.6 5.6.1 Power Considerations Power Supplies The AT32UC3B has several types of power supply pins: * * * * * VDDIO: Powers I/O lines. Voltage is 3.3V nominal. VDDANA: Powers the ADC Voltage is 3.3V nominal. VDDIN: Input voltage for the voltage regulator. Voltage is 3.3V nominal. VDDCORE: Powers the core, memories, and peripherals. Voltage is 1.8V nominal. VDDPLL: Powers the PLL. Voltage is 1.8V nominal. The ground pins GND are common to VDDCORE, VDDIO and VDDPLL. The ground pin for VDDANA is GNDANA. For QFN packages, the center pad must be left unconnected. Refer to "Electrical Characteristics" on page 615 for power consumption on the various supply pins. The main requirement for power supplies connection is to respect a star topology for all electrical connection. 14 32059I-06/2010 AT32UC3B Figure 5-1. Power Supply Single Power Supply Dual Power Supply 3.3V VDDANA 3.3V VDDANA VDDIO VDDIO ADVREF ADVREF VDDIN 1.8V Regulator VDDIN 1.8V Regulator VDDOUT 1.8 V VDDOUT VDDCORE VDDCORE VDDPLL VDDPLL 5.6.2 5.6.2.1 Voltage Regulator Single Power Supply The AT32UC3B embeds a voltage regulator that converts from 3.3V to 1.8V. The regulator takes its input voltage from VDDIN, and supplies the output voltage on VDDOUT that should be externally connected to the 1.8V domains. Adequate input supply decoupling is mandatory for VDDIN in order to improve startup stability and reduce source voltage drop. Two input decoupling capacitors must be placed close to the chip. Adequate output supply decoupling is mandatory for VDDOUT to reduce ripple and avoid oscillations. The best way to achieve this is to use two capacitors in parallel between VDDOUT and GND as close to the chip as possible Figure 5-2. Supply Decoupling 3.3V CIN2 CIN1 VDDIN 1.8V Regulator VDDOUT 1.8V COUT2 COUT1 15 32059I-06/2010 AT32UC3B Refer to Section 28.3 on page 618 for decoupling capacitors values and regulator characteristics. 5.6.2.2 Dual Power Supply In case of dual power supply, VDDIN and VDDOUT should be connected to ground to prevent from leakage current. To avoid over consumption during the power up sequence, VDDIO and VDDCORE voltage difference needs to stay in the range given Figure 5-3. Figure 5-3. 4 Extra consumption on VDDIO VDDIO versus VDDCORE during power up sequence 3.5 3 2.5 VDDIO (V) 2 1.5 1 0.5 0 0 0.2 0.4 0.6 0.8 1 VDDCORE (V) 1.2 1.4 1.6 1.8 2 Extra consumption on VDDCORE 5.6.3 Analog-to-Digital Converter (ADC) reference. The ADC reference (ADVREF) must be provided from an external source. Two decoupling capacitors must be used to insure proper decoupling. Figure 5-4. ADVREF Decoupling 3.3V C VREF2 ADVREF C VREF1 Refer to Section 28.4 on page 618 for decoupling capacitors values and electrical characteristics. In case ADC is not used, the ADVREF pin should be connected to GND to avoid extra consumption. 16 32059I-06/2010 AT32UC3B 6. Processor and Architecture Rev: 1.0.0.0 This chapter gives an overview of the AVR32UC CPU. AVR32UC is an implementation of the AVR32 architecture. A summary of the programming model, instruction set, and MPU is presented. For further details, see the AVR32 Architecture Manual and the AVR32UC Technical Reference Manual. 6.1 Features * 32-bit load/store AVR32A RISC architecture 15 general-purpose 32-bit registers 32-bit Stack Pointer, Program Counter and Link Register reside in register file Fully orthogonal instruction set Privileged and unprivileged modes enabling efficient and secure Operating Systems Innovative instruction set together with variable instruction length ensuring industry leading code density - DSP extention with saturating arithmetic, and a wide variety of multiply instructions * 3-stage pipeline allows one instruction per clock cycle for most instructions - Byte, halfword, word and double word memory access - Multiple interrupt priority levels * MPU allows for operating systems with memory protection - - - - - 6.2 AVR32 Architecture AVR32 is a high-performance 32-bit RISC microprocessor architecture, designed for cost-sensitive embedded applications, with particular emphasis on low power consumption and high code density. In addition, the instruction set architecture has been tuned to allow a variety of microarchitectures, enabling the AVR32 to be implemented as low-, mid-, or high-performance processors. AVR32 extends the AVR family into the world of 32- and 64-bit applications. Through a quantitative approach, a large set of industry recognized benchmarks has been compiled and analyzed to achieve the best code density in its class. In addition to lowering the memory requirements, a compact code size also contributes to the core's low power characteristics. The processor supports byte and halfword data types without penalty in code size and performance. Memory load and store operations are provided for byte, halfword, word, and double word data with automatic sign- or zero extension of halfword and byte data. The C-compiler is closely linked to the architecture and is able to exploit code optimization features, both for size and speed. In order to reduce code size to a minimum, some instructions have multiple addressing modes. As an example, instructions with immediates often have a compact format with a smaller immediate, and an extended format with a larger immediate. In this way, the compiler is able to use the format giving the smallest code size. Another feature of the instruction set is that frequently used instructions, like add, have a compact format with two operands as well as an extended format with three operands. The larger format increases performance, allowing an addition and a data move in the same instruction in a single cycle. Load and store instructions have several different formats in order to reduce code size and speed up execution. 17 32059I-06/2010 AT32UC3B The register file is organized as sixteen 32-bit registers and includes the Program Counter, the Link Register, and the Stack Pointer. In addition, register R12 is designed to hold return values from function calls and is used implicitly by some instructions. 6.3 The AVR32UC CPU The AVR32UC CPU targets low- and medium-performance applications, and provides an advanced OCD system, no caches, and a Memory Protection Unit (MPU). Java acceleration hardware is not implemented. AVR32UC provides three memory interfaces, one High Speed Bus master for instruction fetch, one High Speed Bus master for data access, and one High Speed Bus slave interface allowing other bus masters to access data RAMs internal to the CPU. Keeping data RAMs internal to the CPU allows fast access to the RAMs, reduces latency, and guarantees deterministic timing. Also, power consumption is reduced by not needing a full High Speed Bus access for memory accesses. A dedicated data RAM interface is provided for communicating with the internal data RAMs. A local bus interface is provided for connecting the CPU to device-specific high-speed systems, such as floating-point units and fast GPIO ports. This local bus has to be enabled by writing the LOCEN bit in the CPUCR system register. The local bus is able to transfer data between the CPU and the local bus slave in a single clock cycle. The local bus has a dedicated memory range allocated to it, and data transfers are performed using regular load and store instructions. Details on which devices that are mapped into the local bus space is given in the Memories chapter of this data sheet. Figure 6-1 on page 19 displays the contents of AVR32UC. 18 32059I-06/2010 AT32UC3B Figure 6-1. Interrupt controller interface Overview of the AVR32UC CPU Reset interface OCD interface OCD system Power/ Reset control AVR32UC CPU pipeline MPU Instruction memory controller High Speed Bus master High Speed Bus Data memory controller High Speed Bus slave High Speed Bus High Speed Bus master High Speed Bus 6.3.1 Pipeline Overview AVR32UC has three pipeline stages, Instruction Fetch (IF), Instruction Decode (ID), and Instruction Execute (EX). The EX stage is split into three parallel subsections, one arithmetic/logic (ALU) section, one multiply (MUL) section, and one load/store (LS) section. Instructions are issued and complete in order. Certain operations require several clock cycles to complete, and in this case, the instruction resides in the ID and EX stages for the required number of clock cycles. Since there is only three pipeline stages, no internal data forwarding is required, and no data dependencies can arise in the pipeline. Figure 6-2 on page 20 shows an overview of the AVR32UC pipeline stages. CPU Local Bus Data RAM interface CPU Local Bus master 19 32059I-06/2010 AT32UC3B Figure 6-2. The AVR32UC Pipeline MUL Multiply unit IF ID Regf ile Read A LU Regf ile w rite A LU unit Pref etch unit Decode unit Load-store unit LS 6.3.2 AVR32A Microarchitecture Compliance AVR32UC implements an AVR32A microarchitecture. The AVR32A microarchitecture is targeted at cost-sensitive, lower-end applications like smaller microcontrollers. This microarchitecture does not provide dedicated hardware registers for shadowing of register file registers in interrupt contexts. Additionally, it does not provide hardware registers for the return address registers and return status registers. Instead, all this information is stored on the system stack. This saves chip area at the expense of slower interrupt handling. Upon interrupt initiation, registers R8-R12 are automatically pushed to the system stack. These registers are pushed regardless of the priority level of the pending interrupt. The return address and status register are also automatically pushed to stack. The interrupt handler can therefore use R8-R12 freely. Upon interrupt completion, the old R8-R12 registers and status register are restored, and execution continues at the return address stored popped from stack. The stack is also used to store the status register and return address for exceptions and scall. Executing the rete or rets instruction at the completion of an exception or system call will pop this status register and continue execution at the popped return address. 6.3.3 Java Support AVR32UC does not provide Java hardware acceleration. 6.3.4 Memory Protection The MPU allows the user to check all memory accesses for privilege violations. If an access is attempted to an illegal memory address, the access is aborted and an exception is taken. The MPU in AVR32UC is specified in the AVR32UC Technical Reference manual. 6.3.5 Unaligned Reference Handling AVR32UC does not support unaligned accesses, except for doubleword accesses. AVR32UC is able to perform word-aligned st.d and ld.d. Any other unaligned memory access will cause an address exception. Doubleword-sized accesses with word-aligned pointers will automatically be performed as two word-sized accesses. 20 32059I-06/2010 AT32UC3B The following table shows the instructions with support for unaligned addresses. All other instructions require aligned addresses. Table 6-1. Instruction ld.d st.d Instructions with Unaligned Reference Support Supported alignment Word Word 6.3.6 Unimplemented Instructions The following instructions are unimplemented in AVR32UC, and will cause an Unimplemented Instruction Exception if executed: * All SIMD instructions * All coprocessor instructions if no coprocessors are present * retj, incjosp, popjc, pushjc * tlbr, tlbs, tlbw * cache 6.3.7 CPU and Architecture Revision Three major revisions of the AVR32UC CPU currently exist. The Architecture Revision field in the CONFIG0 system register identifies which architecture revision is implemented in a specific device. AVR32UC CPU revision 3 is fully backward-compatible with revisions 1 and 2, ie. code compiled for revision 1 or 2 is binary-compatible with revision 3 CPUs. 21 32059I-06/2010 AT32UC3B 6.4 6.4.1 Programming Model Register File Configuration The AVR32UC register file is shown below. Figure 6-3. Application Bit 31 Bit 0 Bit 31 The AVR32UC Register File INT0 Bit 31 Bit 0 Supervisor Bit 0 INT1 Bit 31 Bit 0 INT2 Bit 31 Bit 0 INT3 Bit 31 Bit 0 Exception Bit 31 Bit 0 NMI Bit 31 Bit 0 Secure Bit 31 Bit 0 PC LR SP_APP R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR PC LR SP_SEC R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR SS_STATUS SS_ADRF SS_ADRR SS_ADR0 SS_ADR1 SS_SP_SYS SS_SP_APP SS_RAR SS_RSR 6.4.2 Status Register Configuration The Status Register (SR) is split into two halfwords, one upper and one lower, see Figure 6-4 on page 22 and Figure 6-5 on page 23. The lower word contains the C, Z, N, V, and Q condition code flags and the R, T, and L bits, while the upper halfword contains information about the mode and state the processor executes in. Refer to the AVR32 Architecture Manual for details. Figure 6-4. Bit 31 The Status Register High Halfword Bit 16 - LC 1 0 - - DM D - M2 M1 M0 EM I3M I2M FE I1M I0M GM Bit name Initial value Global Interrupt Mask Interrupt Level 0 Mask Interrupt Level 1 Mask Interrupt Level 2 Mask Interrupt Level 3 Mask Exception Mask Mode Bit 0 Mode Bit 1 Mode Bit 2 Reserved Debug State Debug State Mask Reserved 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 22 32059I-06/2010 AT32UC3B Figure 6-5. Bit 15 The Status Register Low Halfword Bit 0 0 T 0 0 0 0 0 0 0 0 0 L 0 Q 0 V 0 N 0 Z 0 C 0 Bit name Initial value Carry Zero Sign Overflow Saturation Lock Reserved Scratch Reserved 6.4.3 6.4.3.1 Processor States Normal RISC State The AVR32 processor supports several different execution contexts as shown in Table 6-2 on page 23. Table 6-2. Priority 1 2 3 4 5 6 N/A N/A Overview of Execution Modes, their Priorities and Privilege Levels. Mode Non Maskable Interrupt Exception Interrupt 3 Interrupt 2 Interrupt 1 Interrupt 0 Supervisor Application Security Privileged Privileged Privileged Privileged Privileged Privileged Privileged Unprivileged Description Non Maskable high priority interrupt mode Execute exceptions General purpose interrupt mode General purpose interrupt mode General purpose interrupt mode General purpose interrupt mode Runs supervisor calls Normal program execution mode Mode changes can be made under software control, or can be caused by external interrupts or exception processing. A mode can be interrupted by a higher priority mode, but never by one with lower priority. Nested exceptions can be supported with a minimal software overhead. When running an operating system on the AVR32, user processes will typically execute in the application mode. The programs executed in this mode are restricted from executing certain instructions. Furthermore, most system registers together with the upper halfword of the status register cannot be accessed. Protected memory areas are also not available. All other operating modes are privileged and are collectively called System Modes. They have full access to all privileged and unprivileged resources. After a reset, the processor will be in supervisor mode. 6.4.3.2 Debug State The AVR32 can be set in a debug state, which allows implementation of software monitor routines that can read out and alter system information for use during application development. This implies that all system and application registers, including the status registers and program counters, are accessible in debug state. The privileged instructions are also available. 23 32059I-06/2010 AT32UC3B All interrupt levels are by default disabled when debug state is entered, but they can individually be switched on by the monitor routine by clearing the respective mask bit in the status register. Debug state can be entered as described in the AVR32UC Technical Reference Manual. Debug state is exited by the retd instruction. 6.4.4 System Registers The system registers are placed outside of the virtual memory space, and are only accessible using the privileged mfsr and mtsr instructions. The table below lists the system registers specified in the AVR32 architecture, some of which are unused in AVR32UC. The programmer is responsible for maintaining correct sequencing of any instructions following a mtsr instruction. For detail on the system registers, refer to the AVR32UC Technical Reference Manual. Table 6-3. Reg # 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 System Registers Address 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96 100 Name SR EVBA ACBA CPUCR ECR RSR_SUP RSR_INT0 RSR_INT1 RSR_INT2 RSR_INT3 RSR_EX RSR_NMI RSR_DBG RAR_SUP RAR_INT0 RAR_INT1 RAR_INT2 RAR_INT3 RAR_EX RAR_NMI RAR_DBG JECR JOSP JAVA_LV0 JAVA_LV1 JAVA_LV2 Function Status Register Exception Vector Base Address Application Call Base Address CPU Control Register Exception Cause Register Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Return Status Register for Debug mode Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Return Address Register for Debug mode Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC 24 32059I-06/2010 AT32UC3B Table 6-3. Reg # 26 27 28 29 30 31 32 33-63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 System Registers (Continued) Address 104 108 112 116 120 124 128 132-252 256 260 264 268 272 276 280 284 288 292 296 300 304 308 312 316 320 324 328 332 336 340 344 348 352 356 360 364 Name JAVA_LV3 JAVA_LV4 JAVA_LV5 JAVA_LV6 JAVA_LV7 JTBA JBCR Reserved CONFIG0 CONFIG1 COUNT COMPARE TLBEHI TLBELO PTBR TLBEAR MMUCR TLBARLO TLBARHI PCCNT PCNT0 PCNT1 PCCR BEAR MPUAR0 MPUAR1 MPUAR2 MPUAR3 MPUAR4 MPUAR5 MPUAR6 MPUAR7 MPUPSR0 MPUPSR1 MPUPSR2 MPUPSR3 Function Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Reserved for future use Configuration register 0 Configuration register 1 Cycle Counter register Compare register Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Unused in AVR32UC Bus Error Address Register MPU Address Register region 0 MPU Address Register region 1 MPU Address Register region 2 MPU Address Register region 3 MPU Address Register region 4 MPU Address Register region 5 MPU Address Register region 6 MPU Address Register region 7 MPU Privilege Select Register region 0 MPU Privilege Select Register region 1 MPU Privilege Select Register region 2 MPU Privilege Select Register region 3 25 32059I-06/2010 AT32UC3B Table 6-3. Reg # 92 93 94 95 96 97 98 99 100 101 102 103-191 192-255 System Registers (Continued) Address 368 372 376 380 384 388 392 396 400 404 408 448-764 768-1020 Name MPUPSR4 MPUPSR5 MPUPSR6 MPUPSR7 MPUCRA MPUCRB MPUBRA MPUBRB MPUAPRA MPUAPRB MPUCR Reserved IMPL Function MPU Privilege Select Register region 4 MPU Privilege Select Register region 5 MPU Privilege Select Register region 6 MPU Privilege Select Register region 7 Unused in this version of AVR32UC Unused in this version of AVR32UC Unused in this version of AVR32UC Unused in this version of AVR32UC MPU Access Permission Register A MPU Access Permission Register B MPU Control Register Reserved for future use IMPLEMENTATION DEFINED 6.5 Exceptions and Interrupts AVR32UC incorporates a powerful exception handling scheme. The different exception sources, like Illegal Op-code and external interrupt requests, have different priority levels, ensuring a welldefined behavior when multiple exceptions are received simultaneously. Additionally, pending exceptions of a higher priority class may preempt handling of ongoing exceptions of a lower priority class. When an event occurs, the execution of the instruction stream is halted, and execution control is passed to an event handler at an address specified in Table 6-4 on page 29. Most of the handlers are placed sequentially in the code space starting at the address specified by EVBA, with four bytes between each handler. This gives ample space for a jump instruction to be placed there, jumping to the event routine itself. A few critical handlers have larger spacing between them, allowing the entire event routine to be placed directly at the address specified by the EVBA-relative offset generated by hardware. All external interrupt sources have autovectored interrupt service routine (ISR) addresses. This allows the interrupt controller to directly specify the ISR address as an address relative to EVBA. The autovector offset has 14 address bits, giving an offset of maximum 16384 bytes. The target address of the event handler is calculated as (EVBA | event_handler_offset), not (EVBA + event_handler_offset), so EVBA and exception code segments must be set up appropriately. The same mechanisms are used to service all different types of events, including external interrupt requests, yielding a uniform event handling scheme. An interrupt controller does the priority handling of the external interrupts and provides the autovector offset to the CPU. 6.5.1 System Stack Issues Event handling in AVR32UC uses the system stack pointed to by the system stack pointer, SP_SYS, for pushing and popping R8-R12, LR, status register, and return address. Since event code may be timing-critical, SP_SYS should point to memory addresses in the IRAM section, since the timing of accesses to this memory section is both fast and deterministic. 26 32059I-06/2010 AT32UC3B The user must also make sure that the system stack is large enough so that any event is able to push the required registers to stack. If the system stack is full, and an event occurs, the system will enter an UNDEFINED state. 6.5.2 Exceptions and Interrupt Requests When an event other than scall or debug request is received by the core, the following actions are performed atomically: 1. The pending event will not be accepted if it is masked. The I3M, I2M, I1M, I0M, EM, and GM bits in the Status Register are used to mask different events. Not all events can be masked. A few critical events (NMI, Unrecoverable Exception, TLB Multiple Hit, and Bus Error) can not be masked. When an event is accepted, hardware automatically sets the mask bits corresponding to all sources with equal or lower priority. This inhibits acceptance of other events of the same or lower priority, except for the critical events listed above. Software may choose to clear some or all of these bits after saving the necessary state if other priority schemes are desired. It is the event source's responsability to ensure that their events are left pending until accepted by the CPU. 2. When a request is accepted, the Status Register and Program Counter of the current context is stored to the system stack. If the event is an INT0, INT1, INT2, or INT3, registers R8-R12 and LR are also automatically stored to stack. Storing the Status Register ensures that the core is returned to the previous execution mode when the current event handling is completed. When exceptions occur, both the EM and GM bits are set, and the application may manually enable nested exceptions if desired by clearing the appropriate bit. Each exception handler has a dedicated handler address, and this address uniquely identifies the exception source. 3. The Mode bits are set to reflect the priority of the accepted event, and the correct register file bank is selected. The address of the event handler, as shown in Table 6-4, is loaded into the Program Counter. The execution of the event handler routine then continues from the effective address calculated. The rete instruction signals the end of the event. When encountered, the Return Status Register and Return Address Register are popped from the system stack and restored to the Status Register and Program Counter. If the rete instruction returns from INT0, INT1, INT2, or INT3, registers R8-R12 and LR are also popped from the system stack. The restored Status Register contains information allowing the core to resume operation in the previous execution mode. This concludes the event handling. 6.5.3 Supervisor Calls The AVR32 instruction set provides a supervisor mode call instruction. The scall instruction is designed so that privileged routines can be called from any context. This facilitates sharing of code between different execution modes. The scall mechanism is designed so that a minimal execution cycle overhead is experienced when performing supervisor routine calls from timecritical event handlers. The scall instruction behaves differently depending on which mode it is called from. The behaviour is detailed in the instruction set reference. In order to allow the scall routine to return to the correct context, a return from supervisor call instruction, rets, is implemented. In the AVR32UC CPU, scall and rets uses the system stack to store the return address and the status register. 6.5.4 Debug Requests The AVR32 architecture defines a dedicated Debug mode. When a debug request is received by the core, Debug mode is entered. Entry into Debug mode can be masked by the DM bit in the 27 32059I-06/2010 AT32UC3B status register. Upon entry into Debug mode, hardware sets the SR[D] bit and jumps to the Debug Exception handler. By default, Debug mode executes in the exception context, but with dedicated Return Address Register and Return Status Register. These dedicated registers remove the need for storing this data to the system stack, thereby improving debuggability. The mode bits in the status register can freely be manipulated in Debug mode, to observe registers in all contexts, while retaining full privileges. Debug mode is exited by executing the retd instruction. This returns to the previous context. 6.5.5 Entry Points for Events Several different event handler entry points exists. In AVR32UC, the reset address is 0x8000_0000. This places the reset address in the boot flash memory area. TLB miss exceptions and scall have a dedicated space relative to EVBA where their event handler can be placed. This speeds up execution by removing the need for a jump instruction placed at the program address jumped to by the event hardware. All other exceptions have a dedicated event routine entry point located relative to EVBA. The handler routine address identifies the exception source directly. AVR32UC uses the ITLB and DTLB protection exceptions to signal a MPU protection violation. ITLB and DTLB miss exceptions are used to signal that an access address did not map to any of the entries in the MPU. TLB multiple hit exception indicates that an access address did map to multiple TLB entries, signalling an error. All external interrupt requests have entry points located at an offset relative to EVBA. This autovector offset is specified by an external Interrupt Controller. The programmer must make sure that none of the autovector offsets interfere with the placement of other code. The autovector offset has 14 address bits, giving an offset of maximum 16384 bytes. Special considerations should be made when loading EVBA with a pointer. Due to security considerations, the event handlers should be located in non-writeable flash memory, or optionally in a privileged memory protection region if an MPU is present. If several events occur on the same instruction, they are handled in a prioritized way. The priority ordering is presented in Table 6-4. If events occur on several instructions at different locations in the pipeline, the events on the oldest instruction are always handled before any events on any younger instruction, even if the younger instruction has events of higher priority than the oldest instruction. An instruction B is younger than an instruction A if it was sent down the pipeline later than A. The addresses and priority of simultaneous events are shown in Table 6-4. Some of the exceptions are unused in AVR32UC since it has no MMU, coprocessor interface, or floating-point unit. 28 32059I-06/2010 AT32UC3B Table 6-4. Priority 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Priority and Handler Addresses for Events Handler Address 0x8000_0000 Provided by OCD system EVBA+0x00 EVBA+0x04 EVBA+0x08 EVBA+0x0C EVBA+0x10 Autovectored Autovectored Autovectored Autovectored EVBA+0x14 EVBA+0x50 EVBA+0x18 EVBA+0x1C EVBA+0x20 EVBA+0x24 EVBA+0x28 EVBA+0x2C EVBA+0x30 EVBA+0x100 EVBA+0x34 EVBA+0x38 EVBA+0x60 EVBA+0x70 EVBA+0x3C EVBA+0x40 EVBA+0x44 Name Reset OCD Stop CPU Unrecoverable exception TLB multiple hit Bus error data fetch Bus error instruction fetch NMI Interrupt 3 request Interrupt 2 request Interrupt 1 request Interrupt 0 request Instruction Address ITLB Miss ITLB Protection Breakpoint Illegal Opcode Unimplemented instruction Privilege violation Floating-point Coprocessor absent Supervisor call Data Address (Read) Data Address (Write) DTLB Miss (Read) DTLB Miss (Write) DTLB Protection (Read) DTLB Protection (Write) DTLB Modified Event source External input OCD system Internal MPU Data bus Data bus External input External input External input External input External input CPU MPU MPU OCD system Instruction Instruction Instruction UNUSED Instruction Instruction CPU CPU MPU MPU MPU MPU UNUSED PC of offending instruction PC of offending instruction PC of offending instruction PC(Supervisor Call) +2 PC of offending instruction PC of offending instruction PC of offending instruction First non-completed instruction PC of offending instruction PC of offending instruction PC of offending instruction First non-completed instruction First non-completed instruction First non-completed instruction First non-completed instruction First non-completed instruction First non-completed instruction First non-completed instruction PC of offending instruction Stored Return Address Undefined First non-completed instruction PC of offending instruction 29 32059I-06/2010 AT32UC3B 6.6 Module Configuration All AT32UC3B parts do not implement the same CPU and Architecture Revision. Table 6-5. CPU and Architecture Revision Architecture Revision 2 1 1 1 Part Name AT32UC3Bx512 AT32UC3Bx256 AT32UC3Bx128 AT32UC3Bx64 30 32059I-06/2010 AT32UC3B 7. Memories 7.1 Embedded Memories * Internal High-Speed Flash 512KBytes (AT32UC3B0512, AT32UC3B1512) 256 KBytes (AT32UC3B0256, AT32UC3B1256) 128 KBytes (AT32UC3B0128, AT32UC3B1128) 64 KBytes (AT32UC3B064, AT32UC3B164) - 0 Wait State Access at up to 30 MHz in Worst Case Conditions - 1 Wait State Access at up to 60 MHz in Worst Case Conditions - Pipelined Flash Architecture, allowing burst reads from sequential Flash locations, hiding penalty of 1 wait state access - 100 000 Write Cycles, 15-year Data Retention Capability - 4 ms Page Programming Time, 8 ms Chip Erase Time - Sector Lock Capabilities, Bootloader Protection, Security Bit - 32 Fuses, Erased During Chip Erase - User Page For Data To Be Preserved During Chip Erase * Internal High-Speed SRAM, Single-cycle access at full speed - 96KBytes ((AT32UC3B0512, AT32UC3B1512) - 32KBytes (AT32UC3B0256, AT32UC3B0128, AT32UC3B1256 and AT32UC3B1128) - 16KBytes (AT32UC3B064 and AT32UC3B164) - - - - 7.2 Physical Memory Map The system bus is implemented as a bus matrix. All system bus addresses are fixed, and they are never remapped in any way, not even in boot. Note that AVR32 UC CPU uses unsegmented translation, as described in the AVR32UC Technical Architecture Manual. The 32-bit physical address space is mapped as follows: Table 7-1. AT32UC3B Physical Memory Map Embedded SRAM 0x0000_0000 AT32UC3B0512 AT32UC3B1512 AT32UC3B0256 AT32UC3B1256 96 Kbytes 32 Kbytes 32 Kbytes 16 Kbytes Embedded Flash 0x8000_0000 512 Kbytes 256 Kbytes 128 Kbytes 64 Kbytes HSB-PB Bridge A 0xFFFF_0000 64 Kbytes 64 Kbytes 64 Kbytes 64 Kbytes HSB-PB Bridge B 0xFFFE_0000 64 Kbytes 64 Kbytes 64 Kbytes 64 Kbytes Device Start Address USB Data 0xD000_0000 64 Kbytes 64 Kbytes 64 Kbytes 64 Kbytes Size AT32UC3B0128 AT32UC3B1128 AT32UC3B064 AT32UC3B164 31 32059I-06/2010 AT32UC3B 7.3 Peripheral Address Map Peripheral Address Mapping Address 0xFFFE0000 Table 7-2. Peripheral Name USB USB 2.0 Interface - USB 0xFFFE1000 HMATRIX 0xFFFE1400 HSB Matrix - HMATRIX HFLASHC 0xFFFF0000 Flash Controller - HFLASHC PDCA 0xFFFF0800 Peripheral DMA Controller - PDCA INTC 0xFFFF0C00 Interrupt controller - INTC PM 0xFFFF0D00 Power Manager - PM RTC 0xFFFF0D30 Real Time Counter - RTC WDT 0xFFFF0D80 Watchdog Timer - WDT EIM 0xFFFF1000 External Interrupt Controller - EIM GPIO 0xFFFF1400 General Purpose Input/Output Controller - GPIO Universal Synchronous/Asynchronous Receiver/Transmitter - USART0 Universal Synchronous/Asynchronous Receiver/Transmitter - USART1 Universal Synchronous/Asynchronous Receiver/Transmitter - USART2 Serial Peripheral Interface - SPI0 USART0 0xFFFF1800 USART1 0xFFFF1C00 USART2 0xFFFF2400 SPI0 0xFFFF2C00 TWI 0xFFFF3000 Two-wire Interface - TWI PWM 0xFFFF3400 Pulse Width Modulation Controller - PWM SSC 0xFFFF3800 Synchronous Serial Controller - SSC TC Timer/Counter - TC 32 32059I-06/2010 AT32UC3B Table 7-2. Peripheral Address Mapping 0xFFFF3C00 ADC 0xFFFF4000 Analog to Digital Converter - ADC ABDAC Audio Bitstream DAC - ABDAC 7.4 CPU Local Bus Mapping Some of the registers in the GPIO module are mapped onto the CPU local bus, in addition to being mapped on the Peripheral Bus. These registers can therefore be reached both by accesses on the Peripheral Bus, and by accesses on the local bus. Mapping these registers on the local bus allows cycle-deterministic toggling of GPIO pins since the CPU and GPIO are the only modules connected to this bus. Also, since the local bus runs at CPU speed, one write or read operation can be performed per clock cycle to the local busmapped GPIO registers. The following GPIO registers are mapped on the local bus: Table 7-3. Port 0 Local bus mapped GPIO registers Mode WRITE SET CLEAR TOGGLE Local Bus Address 0x4000_0040 0x4000_0044 0x4000_0048 0x4000_004C 0x4000_0050 0x4000_0054 0x4000_0058 0x4000_005C 0x4000_0060 0x4000_0140 0x4000_0144 0x4000_0148 0x4000_014C 0x4000_0150 0x4000_0154 0x4000_0158 0x4000_015C 0x4000_0160 Access Write-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Read-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Write-only Read-only Register Output Driver Enable Register (ODER) Output Value Register (OVR) WRITE SET CLEAR TOGGLE Pin Value Register (PVR) 1 Output Driver Enable Register (ODER) WRITE SET CLEAR TOGGLE Output Value Register (OVR) WRITE SET CLEAR TOGGLE Pin Value Register (PVR) - 33 32059I-06/2010 AT32UC3B 8. Boot Sequence This chapter summarizes the boot sequence of the AT32UC3B. The behaviour after power-up is controlled by the Power Manager. For specific details, refer to section Power Manager (PM). 8.1 Starting of clocks After power-up, the device will be held in a reset state by the Power-On Reset circuitry, until the power has stabilized throughout the device. Once the power has stabilized, the device will use the internal RC Oscillator as clock source. On system start-up, the PLLs are disabled. All clocks to all modules are running. No clocks have a divided frequency, all parts of the system recieves a clock with the same frequency as the internal RC Oscillator. 8.2 Fetching of initial instructions After reset has been released, the AVR32 UC CPU starts fetching instructions from the reset address, which is 0x8000_0000. This address points to the first address in the internal Flash. The code read from the internal Flash is free to configure the system to use for example the PLLs, to divide the frequency of the clock routed to some of the peripherals, and to gate the clocks to unused peripherals. 34 32059I-06/2010 AT32UC3B 9. Power Manager (PM) Rev: 2.3.0.2 9.1 Features * * * * * * * * * * * * * Controls integrated oscillators and PLLs Generates clocks and resets for digital logic Supports 2 crystal oscillators 0.4-20 MHz Supports 2 PLLs 80-240 MHz Supports 32 KHz ultra-low power oscillator Integrated low-power RC oscillator On-the fly frequency change of CPU, HSB, PBA, and PBB clocks Sleep modes allow simple disabling of logic clocks, PLLs, and oscillators Module-level clock gating through maskable peripheral clocks Wake-up from internal or external interrupts Generic clocks with wide frequency range provided Automatic identification of reset sources Controls brownout detector (BOD), RC oscillator, and bandgap voltage reference through control and calibration registers 9.2 Description The Power Manager (PM) controls the oscillators and PLLs, and generates the clocks and resets in the device. The PM controls two fast crystal oscillators, as well as two PLLs, which can multiply the clock from either oscillator to provide higher frequencies. Additionally, a low-power 32 KHz oscillator is used to generate the real-time counter clock for high accuracy real-time measurements. The PM also contains a low-power RC oscillator with fast start-up time, which can be used to clock the digital logic. The provided clocks are divided into synchronous and generic clocks. The synchronous clocks are used to clock the main digital logic in the device, namely the CPU, and the modules and peripherals connected to the HSB, PBA, and PBB buses. The generic clocks are asynchronous clocks, which can be tuned precisely within a wide frequency range, which makes them suitable for peripherals that require specific frequencies, such as timers and communication modules. The PM also contains advanced power-saving features, allowing the user to optimize the power consumption for an application. The synchronous clocks are divided into three clock domains, one for the CPU and HSB, one for modules on the PBA bus, and one for modules on the PBB bus.The three clocks can run at different speeds, so the user can save power by running peripherals at a relatively low clock, while maintaining a high CPU performance. Additionally, the clocks can be independently changed on-the-fly, without halting any peripherals. This enables the user to adjust the speed of the CPU and memories to the dynamic load of the application, without disturbing or re-configuring active peripherals. Each module also has a separate clock, enabling the user to switch off the clock for inactive modules, to save further power. Additionally, clocks and oscillators can be automatically switched off during idle periods by using the sleep instruction on the CPU. The system will return to normal on occurrence of interrupts. The Power Manager also contains a Reset Controller, which collects all possible reset sources, generates hard and soft resets, and allows the reset source to be identified by software. 35 32059I-06/2010 AT32UC3B 9.3 Block Diagram Figure 9-1. Power Manager block diagram Synchronous clocks CPU, HSB, PBA, PBB RCOSC Synchronous Clock Generator PLL0 Oscillator 0 Oscillator 1 PLL1 Generic Clock Generator G eneric clocks 32 KHz Oscillator O SC/PLL Control signals 32 KHz clock for RTC RC Oscillator Slow clock Oscillator and PLL Control Voltage Regulator Startup Counter Interrupts Sleep Controller Sleep instruction fuses Calibration Registers Brown-Out Detector Power-On Detector O ther reset sources External Reset Pad Reset Controller resets 36 32059I-06/2010 AT32UC3B 9.4 9.4.1 Product Dependencies I/O Lines The PM provides a number of generic clock outputs, which can be connected to output pins, multiplexed with GPIO lines. The programmer must first program the GPIO controller to assign these pins to their peripheral function. If the I/O pins of the PM are not used by the application, they can be used for other purposes by the GPIO controller. 9.4.2 Interrupt The PM interrupt line is connected to one of the internal sources of the interrupt controller. Using the PM interrupt requires the interrupt controller to be programmed first. 9.4.3 Clock implementation In AT32UC3B, the HSB shares the source clock with the CPU. This means that writing to the HSBDIV and HSBSEL bits in CKSEL has no effect. These bits will always read the same as CPUDIV and CPUSEL. 9.5 9.5.1 Functional Description Slow clock The slow clock is generated from an internal RC oscillator which is always running, except in Static mode. The slow clock can be used for the main clock in the device, as described in "Synchronous clocks" on page 39. The slow clock is also used for the Watchdog Timer and measuring various delays in the Power Manager. The RC oscillator has a 3 cycles startup time, and is always available when the CPU is running. The RC oscillator operates at approximately 115 kHz, and can be calibrated to a narrow range by the RCOSCCAL fuses. Software can also change RC oscillator calibration through the use of the RCCR register. Please see the Electrical Characteristics section for details. RC oscillator can also be used as the RTC clock when crystal accuracy is not required. 9.5.2 Oscillator 0 and 1 operation The two main oscillators are designed to be used with an external 450 kHz to 16 MHz crystal and two biasing capacitors, as shown in Figure . Oscillator 0 can be used for the main clock in the device, as described in "Synchronous clocks" on page 39. Both oscillators can be used as source for the generic clocks, as described in "Generic clocks" on page 43. The oscillators are disabled by default after reset. When the oscillators are disabled, the XIN and XOUT pins can be used as general purpose I/Os. When the oscillators are configured to use an external clock, the clock must be applied to the XIN pin while the XOUT pin can be used as a general purpose I/O. The oscillators can be enabled by writing to the OSCnEN bits in MCCTRL. Operation mode (external clock or crystal) is chosen by writing to the MODE field in OSCCTRLn. Oscillators are automatically switched off in certain sleep modes to reduce power consumption, as described in Section 9.5.7 on page 42. After a hard reset, or when waking up from a sleep mode that disabled the oscillators, the oscillators may need a certain amount of time to stabilize on the correct frequency. This start-up time can be set in the OSCCTRLn register. 37 32059I-06/2010 AT32UC3B The PM masks the oscillator outputs during the start-up time, to ensure that no unstable clocks propagate to the digital logic. The OSCnRDY bits in POSCSR are automatically set and cleared according to the status of the oscillators. A zero to one transition on these bits can also be configured to generate an interrupt, as described in "MODE: Oscillator Mode" on page 57. Figure 9-2. Oscillator connections C2 XO UT XIN C1 9.5.3 32 KHz oscillator operation The 32 KHz oscillator operates as described for Oscillator 0 and 1 above. The 32 KHz oscillator is used as source clock for the Real-Time Counter. The oscillator is disabled by default, but can be enabled by writing OSC32EN in OSCCTRL32. The oscillator is an ultra-low power design and remains enabled in all sleep modes except Static mode. While the 32 KHz oscillator is disabled, the XIN32 and XOUT32 pins are available as general purpose I/Os. When the oscillator is configured to work with an external clock (MODE field in OSCCTRL32 register), the external clock must be connected to XIN32 while the XOUT32 pin can be used as a general purpose I/O. The startup time of the 32 KHz oscillator can be set in the OSCCTRL32, after which OSC32RDY in POSCSR is set. An interrupt can be generated on a zero to one transition of OSC32RDY. As a crystal oscillator usually requires a very long startup time (up to 1 second), the 32 KHz oscillator will keep running across resets, except Power-On-Reset. 9.5.4 PLL operation The device contains two PLLs, PLL0 and PLL1. These are disabled by default, but can be enabled to provide high frequency source clocks for synchronous or generic clocks. The PLLs can take either Oscillator 0 or 1 as reference clock. The PLL output is divided by a multiplication factor, and the PLL compares the resulting clock to the reference clock. The PLL will adjust its output frequency until the two compared clocks are equal, thus locking the output frequency to a multiple of the reference clock frequency. The Voltage Controlled Oscillator inside the PLL can generate frequencies from 80 to 240 MHz. To make the PLL output frequencies under 80 MHz the OTP[1] bitfield could be set. This will divide the output of the PLL by two and bring the clock in range of the max frequency of the CPU. 38 32059I-06/2010 AT32UC3B When the PLL is switched on, or when changing the clock source or multiplication factor for the PLL, the PLL is unlocked and the output frequency is undefined. The PLL clock for the digital logic is automatically masked when the PLL is unlocked, to prevent connected digital logic from receiving a too high frequency and thus become unstable. Figure 9-3. PLL with control logic and filters PLLM UL P L L O P T [1 ] O u tp u t D iv id e r fvco 0 fP LL 1 /2 O sc0 c lo c k O sc1 c lo c k 0 1 M ask P L L c lo c k 1 In p u t D iv id e r P L L D IV Phase D e te c to r VCO Lock D e te c to r L o c k b it PLLO SC PLLO PT 9.5.4.1 Enabling the PLL PLLn is enabled by writing the PLLEN bit in the PLLn register. PLLOSC selects Oscillator 0 or 1 as clock source. The PLLMUL and PLLDIV bitfields must be written with the multiplication and division factors, respectively, creating the voltage controlled ocillator frequency fVCO and the PLL frequency fPLL : fVCO = (PLLMUL+1)/(PLLDIV) * fOSC if PLLDIV > 0. fVCO = 2*(PLLMUL+1) * fOSC if PLLDIV = 0. If PLLOPT[1] field is set to 0: fPLL = fVCO. If PLLOPT[1] field is set to 1: fPLL = fVCO / 2. The PLLn:PLLOPT field should be set to proper values according to the PLL operating frequency. The PLLOPT field can also be set to divide the output frequency of the PLLs by 2. The lock signal for each PLL is available as a LOCKn flag in POSCSR. An interrupt can be generated on a 0 to 1 transition of these bits. 9.5.5 Synchronous clocks The slow clock (default), Oscillator 0, or PLL0 provide the source for the main clock, which is the common root for the synchronous clocks for the CPU/HSB, PBA, and PBB modules. The main clock is divided by an 8-bit prescaler, and each of these four synchronous clocks can run from 39 32059I-06/2010 AT32UC3B any tapping of this prescaler, or the undivided main clock, as long as fCPU fPBA,B,. The synchronous clock source can be changed on-the fly, responding to varying load in the application. The clock domains can be shut down in sleep mode, as described in "Sleep modes" on page 42. Additionally, the clocks for each module in the four domains can be individually masked, to avoid power consumption in inactive modules. Figure 9-4. Synchronous clock generation Sleep instruction Sleep Controller 0 Slow clock Osc0 clock PLL0 clock Main clock Mask CPUMASK CPU clocks HSB clocks PBAclocks PBB clocks Prescaler 1 CPUDIV MCSEL CPUSEL 9.5.5.1 Selecting PLL or oscillator for the main clock The common main clock can be connected to the slow clock, Oscillator 0, or PLL0. By default, the main clock will be connected to the slow clock. The user can connect the main clock to Oscillator 0 or PLL0 by writing the MCSEL bitfield in the Main Clock Control Register (MCCTRL). This must only be done after that unit has been enabled, otherwise a deadlock will occur. Care should also be taken that the new frequency of the synchronous clocks does not exceed the maximum frequency for each clock domain. 9.5.5.2 Selecting synchronous clock division ratio The main clock feeds an 8-bit prescaler, which can be used to generate the synchronous clocks. By default, the synchronous clocks run on the undivided main clock. The user can select a prescaler division for the CPU clock by writing CKSEL:CPUDIV to 1 and CPUSEL to the prescaling value, resulting in a CPU clock frequency: fCPU = fmain / 2(CPUSEL+1) 40 32059I-06/2010 AT32UC3B Similarly, the clock for the PBA, and PBB can be divided by writing their respective bitfields. To ensure correct operation, frequencies must be selected so that fCPU fPBA,B. Also, frequencies must never exceed the specified maximum frequency for each clock domain. CKSEL can be written without halting or disabling peripheral modules. Writing CKSEL allows a new clock setting to be written to all synchronous clocks at the same time. It is possible to keep one or more clocks unchanged by writing the same value a before to the xxxDIV and xxxSEL bitfields. This way, it is possible to e.g. scale CPU and HSB speed according to the required performance, while keeping the PBA and PBB frequency constant. For modules connected to the HSB bus, the PB clock frequency must be set to the same frequency than the CPU clock. 9.5.5.3 Clock Ready flag There is a slight delay from CKSEL is written and the new clock setting becomes effective. During this interval, the Clock Ready (CKRDY) flag in ISR will read as 0. If IER:CKRDY is written to 1, the Power Manager interrupt can be triggered when the new clock setting is effective. CKSEL must not be re-written while CKRDY is 0, or the system may become unstable or hang. 9.5.6 Peripheral clock masking By default, the clock for all modules are enabled, regardless of which modules are actually being used. It is possible to disable the clock for a module in the CPU, HSB, PBA, or PBB clock domain by writing the corresponding bit in the Clock Mask register (CPU/HSB/PBA/PBB) to 0. When a module is not clocked, it will cease operation, and its registers cannot be read or written. The module can be re-enabled later by writing the corresponding mask bit to 1. A module may be connected to several clock domains, in which case it will have several mask bits. Table 9-6 contains a list of implemented maskable clocks. 9.5.6.1 Cautionary note Note that clocks should only be switched off if it is certain that the module will not be used. Switching off the clock for the internal RAM will cause a problem if the stack is mapped there. Switching off the clock to the Power Manager (PM), which contains the mask registers, or the corresponding PBx bridge, will make it impossible to write the mask registers again. In this case, they can only be re-enabled by a system reset. 9.5.6.2 Mask Ready flag Due to synchronization in the clock generator, there is a slight delay from a mask register is written until the new mask setting goes into effect. When clearing mask bits, this delay can usually be ignored. However, when setting mask bits, the registers in the corresponding module must not be written until the clock has actually be re-enabled. The status flag MSKRDY in ISR provides the required mask status information. When writing either mask register with any value, this bit is cleared. The bit is set when the clocks have been enabled and disabled according to the new mask setting. Optionally, the Power Manager interrupt can be enabled by writing the MSKRDY bit in IER. 41 32059I-06/2010 AT32UC3B 9.5.7 Sleep modes In normal operation, all clock domains are active, allowing software execution and peripheral operation. When the CPU is idle, it is possible to switch off the CPU clock and optionally other clock domains to save power. This is activated by the sleep instruction, which takes the sleep mode index number as argument. 9.5.7.1 Entering and exiting sleep modes The sleep instruction will halt the CPU and all modules belonging to the stopped clock domains. The modules will be halted regardless of the bit settings of the mask registers. Oscillators and PLLs can also be switched off to save power. Some of these modules have a relatively long start-up time, and are only switched off when very low power consumption is required. The CPU and affected modules are restarted when the sleep mode is exited. This occurs when an interrupt triggers. Note that even if an interrupt is enabled in sleep mode, it may not trigger if the source module is not clocked. 9.5.7.2 Supported sleep modes The following sleep modes are supported. These are detailed in Table 9-1. *Idle: The CPU is stopped, the rest of the chip is operating. Wake-up sources are any interrupt. *Frozen: The CPU and HSB modules are stopped, peripherals are operating. Wake-up sources are any interrupts from PB modules. *Standby: All synchronous clocks are stopped, but oscillators and PLLs are running, allowing quick wake-up to normal mode. Wake-up sources are RTC or external interrupt (EIC), external reset or any asynchronous interrupts from PB modules. *Stop: As Standby, but Oscillator 0 and 1, and the PLLs are stopped. 32 KHz (if enabled) and RC oscillators and RTC/WDT will still operate. Wake-up are the same as for Standby mode. *DeepStop: All synchronous clocks, Oscillator 0 and 1 and PLL 0 and 1 are stopped. 32 KHz oscillator can run if enabled. RC oscillator still operates. Bandgap voltage reference and BOD is turned off. Wake-up sources are RTC, external interrupt in asynchronous mode, external reset or any asynchronous interrupts from PB modules. *Static: All oscillators, including 32 KHz and RC oscillator are stopped. Bandgap voltage reference BOD detector is turned off. Wake-up sources are external interrupt (EIC) in asynchronous mode only, external reset pin or any asynchronous interrupts from PB modules. Table 9-1. Sleep modes PBA,B GCLK Run Run Stop Osc0,1 PLL0,1, SYSTIMER Run Run Run BOD & Bandgap On On On Voltage Regulator Full power Full power Full power Index 0 1 2 Sleep Mode Idle Frozen Standby CPU Stop Stop Stop HSB Run Stop Stop Osc32 Run Run Run RCOsc Run Run Run 42 32059I-06/2010 AT32UC3B Table 9-1. Sleep modes PBA,B GCLK Stop Stop Stop Osc0,1 PLL0,1, SYSTIMER Stop Stop Stop BOD & Bandgap On Off Off Voltage Regulator Low power Low power Low power Index 3 4 5 Sleep Mode Stop DeepStop Static CPU Stop Stop Stop HSB Stop Stop Stop Osc32 Run Run Stop RCOsc Run Run Stop The power level of the internal voltage regulator is also adjusted according to the sleep mode to reduce the internal regulator power consumption. 9.5.7.3 Precautions when entering sleep mode Modules communicating with external circuits should normally be disabled before entering a sleep mode that will stop the module operation. This prevents erratic behavior when entering or exiting sleep mode. Please refer to the relevant module documentation for recommended actions. Communication between the synchronous clock domains is disturbed when entering and exiting sleep modes. This means that bus transactions are not allowed between clock domains affected by the sleep mode. The system may hang if the bus clocks are stopped in the middle of a bus transaction. The CPU is automatically stopped in a safe state to ensure that all CPU bus operations are complete when the sleep mode goes into effect. Thus, when entering Idle mode, no further action is necessary. When entering a sleep mode (except Idle mode), all HSB masters must be stopped before entering the sleep mode. Also, if there is a chance that any PB write operations are incomplete, the CPU should perform a read operation from any register on the PB bus before executing the sleep instruction. This will stall the CPU while waiting for any pending PB operations to complete. 9.5.7.4 Wake Up The USB can be used to wake up the part from sleep modes through register AWEN of the Power Manager. 9.5.8 Generic clocks Timers, communication modules, and other modules connected to external circuitry may require specific clock frequencies to operate correctly. The Power Manager contains an implementation defined number of generic clocks that can provide a wide range of accurate clock frequencies. Each generic clock module runs from either Oscillator 0 or 1, or PLL0 or 1. The selected source can optionally be divided by any even integer up to 512. Each clock can be independently enabled and disabled, and is also automatically disabled along with peripheral clocks by the Sleep Controller. 43 32059I-06/2010 AT32UC3B Figure 9-5. Generic clock generation Sleep Controller 0 Osc0 clock Osc1 clock PLL0 clock PLL1 clock 0 Mask Divider 1 Generic Clock 1 PLLSEL OSCSEL DIV DIVEN CEN 9.5.8.1 Enabling a generic clock A generic clock is enabled by writing the CEN bit in GCCTRL to 1. Each generic clock can use either Oscillator 0 or 1 or PLL0 or 1 as source, as selected by the PLLSEL and OSCSEL bits. The source clock can optionally be divided by writing DIVEN to 1 and the division factor to DIV, resulting in the output frequency: fGCLK = fSRC / (2*(DIV+1)) 9.5.8.2 Disabling a generic clock The generic clock can be disabled by writing CEN to 0 or entering a sleep mode that disables the PB clocks. In either case, the generic clock will be switched off on the first falling edge after the disabling event, to ensure that no glitches occur. If CEN is written to 0, the bit will still read as 1 until the next falling edge occurs, and the clock is actually switched off. When writing CEN to 0, the other bits in GCCTRL should not be changed until CEN reads as 0, to avoid glitches on the generic clock. When the clock is disabled, both the prescaler and output are reset. 9.5.8.3 Changing clock frequency When changing generic clock frequency by writing GCCTRL, the clock should be switched off by the procedure above, before being re-enabled with the new clock source or division setting. This prevents glitches during the transition. 44 32059I-06/2010 AT32UC3B 9.5.8.4 Generic clock implementation In AT32UC3B, there are 5 generic clocks. These are allocated to different functions as shown in Table 9-2. Table 9-2. Generic clock allocation Function GCLK0 pin GCLK1 pin GCLK2 pin USBB ABDAC Clock number 0 1 2 3 4 9.5.9 Divided PB clocks The clock generator in the Power Manager provides divided PBA and PBB clocks for use by peripherals that require a prescaled PBx clock. This is described in the documentation for the relevant modules. The divided clocks are not directly maskable, but are stopped in sleep modes where the PBx clocks are stopped. 9.5.10 Debug operation During a debug session, the user may need to halt the system to inspect memory and CPU registers. The clocks normally keep running during this debug operation, but some peripherals may require the clocks to be stopped, e.g. to prevent timer overflow, which would cause the program to fail. For this reason, peripherals on the PBA and PBB buses may use "debug qualified" PBx clocks. This is described in the documentation for the relevant modules. The divided PBx clocks are always debug qualified clocks. Debug qualified PB clocks are stopped during debug operation. The debug system can optionally keep these clocks running during the debug operation. This is described in the documentation for the On-Chip Debug system. 9.5.11 Reset Controller The Reset Controller collects the various reset sources in the system and generates hard and soft resets for the digital logic. The device contains a Power-On Detector, which keeps the system reset until power is stable. This eliminates the need for external reset circuitry to guarantee stable operation when powering up the device. It is also possible to reset the device by asserting the RESET_N pin. This pin has an internal pullup, and does not need to be driven externally when negated. Table 9-4 lists these and other reset sources supported by the Reset Controller. 45 32059I-06/2010 AT32UC3B Figure 9-6. Reset Controller block diagram R C _R C AU S E RESET_N P o w e r-O n D e te c to r B ro w n o u t D e te c to r JTAG OCD W a tc h d o g R e s e t CPU, HSB, PBA, PBB R eset C o n tro lle r O C D , R T C /W D T C lo c k G e n e ra to In addition to the listed reset types, the JTAG can keep parts of the device statically reset through the JTAG Reset Register. See JTAG documentation for details. Table 9-3. Reset source Reset description Description Supply voltage below the power-on reset detector threshold voltage RESET_N pin asserted Supply voltage below the brownout reset detector threshold voltage Caused by an illegal CPU access to external memory while in Supervisor mode See watchdog timer documentation. See On-Chip Debug documentation Power-on Reset External Reset Brownout Reset CPU Error Watchdog Timer OCD When a Reset occurs, some parts of the chip are not necessarily reset, depending on the reset source. Only the Power On Reset (POR) will force a reset of the whole chip. 46 32059I-06/2010 AT32UC3B Table 9-4 lists parts of the device that are reset, depending on the reset source. Table 9-4. Effect of the different reset events Power-On Reset CPU/HSB/PBA/PBB (excluding Power Manager) 32 KHz oscillator RTC control register GPLP registers Watchdog control register Y Y Y Y Y Y Y Y Y Y Y Y Y External Reset Y N N N Y N N Y Y Y Y Y Y Watchdog Reset Y N N N N N N N N Y Y Y N BOD Reset Y N N N Y N N N N Y Y Y Y CPU Error Reset Y N N N Y N N N N Y Y Y Y OCD Reset Y N N N Y N N N N Y Y Y N Voltage Calibration register RC Oscillator Calibration register BOD control register Bandgap control register Clock control registers Osc0/Osc1 and control registers PLL0/PLL1 and control registers OCD system and OCD registers The cause of the last reset can be read from the RCAUSE register. This register contains one bit for each reset source, and can be read during the boot sequence of an application to determine the proper action to be taken. 9.5.11.1 Power-On Detector The Power-On Detector monitors the VDDCORE supply pin and generates a reset when the device is powered on. The reset is active until the supply voltage from the linear regulator is above the power-on threshold level. The reset will be re-activated if the voltage drops below the power-on threshold level. See Electrical Characteristics for parametric details. 9.5.11.2 Brown-Out Detector The Brown-Out Detector (BOD) monitors the VDDCORE supply pin and compares the supply voltage to the brown-out detection level, as set in BOD.LEVEL. The BOD is disabled by default, but can be enabled either by software or by flash fuses. The Brown-Out Detector can either generate an interrupt or a reset when the supply voltage is below the brown-out detection level. In any case, the BOD output is available in bit POSCR.BODET bit. Note 1 : Any change to the BOD.LEVEL field of the BOD register should be done with the BOD deactivated to avoid spurious reset or interrupt. Note 2 : If the BOD level is set to a value higher than VDDCORE and enabled by fuses, the part will be in constant reset. In order to leave reset state, an external voltage higher than the BOD level should be applied. Thus, it is possible to disable BOD. 47 32059I-06/2010 AT32UC3B See Electrical Characteristics for parametric details. 9.5.11.3 External Reset The external reset detector monitors the state of the RESET_N pin. By default, a low level on this pin will generate a reset. 9.5.12 Calibration registers The Power Manager controls the calibration of the RC oscillator, voltage regulator, bandgap voltage reference through several calibration registers. Those calibration registers are loaded after a Power On Reset with default values stored in factory-programmed flash fuses. Although it is not recommended to override default factory settings, it is still possible to override these default values by writing to those registers. To prevent unexpected writes due to software bugs, write access to these registers is protected by a "key". First, a write to the register must be made with the field "KEY" equal to 0x55 then a second write must be issued with the "KEY" field equal to 0xAA 48 32059I-06/2010 AT32UC3B 9.6 User Interface Table 9-5. Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0020 0x0024 0x0028 0x002C 0x0030 0x0040 0x0044 0x0048 0x004C 0x0050 0x0054 0x0060-0x070 0x00C0 0x00C4 0x00C8 0x00D0 0x0140 0x0144 0x0200 0x0204 PM Register Memory Map Register Register Name MCCTRL CKSEL CPUMASK HSBMASK PBAMASK PBBMASK PLL0 PLL1 OSCCTRL0 OSCCTRL1 OSCCTRL32 IER IDR IMR ISR ICR POSCSR GCCTRL RCCR BGCR VREGCR BOD RCAUSE AWEN GPLP0 GPLP1 Access Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Write-Only Write-Only Read-Only Read-Only Write-Only Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read-Only Read/Write Read/Write Read/Write Reset 0x00000000 0x00000000 0x00000003 0x0000007F 0x00007FFF 0x0000003F 0x00000000 0x00000000 0x00000000 0x00000000 0x00010000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 Factory settings Factory settings Factory settings BOD fuses in Flash Latest Reset Source 0x00000000 0x00000000 0x00000000 Main Clock Control Register Clock Select Register CPU Mask Register HSB Mask Register PBA Mask Register PBB Mask Register PLL0 Control Register PLL1 Control Register Oscillator 0 Control Register Oscillator 1 Control Register Oscillator 32 Control Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Interrupt Clear Register Power and Oscillators Status Register Generic Clock Control Register RC Oscillator Calibration Register Bandgap Calibration Register Linear Regulator Calibration Register BOD Level Register Reset Cause Register Asynchronous Wake Up Enable Register General Purpose Low-Power Register 0 General Purpose Low-Power Register 1 49 32059I-06/2010 AT32UC3B 9.6.1 Name: Access Type: Offset: Reset Value: Main Clock Control Register MCCTRL Read/Write 0x000 0x00000000 31 23 15 7 - 30 22 14 6 - 29 21 13 5 28 20 12 4 27 19 11 3 OSC1EN 26 18 10 2 OSC0EN 25 17 9 1 MCSEL 24 16 8 0 * OSC1EN: Oscillator 1 Enable 0: Oscillator 1 is disabled. 1: Oscillator 1 is enabled. * OSC0EN: Oscillator 0 Enable 0: Oscillator 0 is disabled. 1: Oscillator 0 is enabled. * MCSEL: Main Clock Select 0: The slow clock is the source for the main clock. 1: Oscillator 0 is the source for the main clock. 2: PLL0 is the source for the main clock. 3: Reserved. 50 32059I-06/2010 AT32UC3B 9.6.2 Name: Access Type: Offset: Reset Value: Clock Select Register CKSEL Read/Write 0x004 0x00000000 31 PBBDIV 23 PBADIV 15 HSBDIV 7 CPUDIV 30 22 14 6 - 29 21 13 5 - 28 20 12 4 - 27 19 11 3 - 26 25 PBBSEL 24 18 17 PBASEL 16 10 9 HSBSEL 8 2 1 CPUSEL 0 * PBBDIV, PBBSEL: PBB Division and Clock Select PBBDIV = 0: PBB clock equals main clock. PBBDIV = 1: PBB clock equals main clock divided by 2(PBBSEL+1). * PBADIV, PBASEL: PBA Division and Clock Select PBADIV = 0: PBA clock equals main clock. PBADIV = 1: PBA clock equals main clock divided by 2(PBASEL+1). * HSBDIV, HSBSEL: HSB Division and Clock Select For the AT32UC3B, HSBDIV always equals CPUDIV, and HSBSEL always equals CPUSEL, as the HSB clock is always equal to the CPU clock. * CPUDIV, CPUSEL: CPU Division and Clock Select CPUDIV = 0: CPU clock equals main clock. CPUDIV = 1: CPU clock equals main clock divided by 2(CPUSEL+1). Note that if xxxDIV is written to 0, xxxSEL should also be written to 0 to ensure correct operation. Also note that writing this register clears POSCSR:CKRDY. The register must not be re-written until CKRDY goes high. 51 32059I-06/2010 AT32UC3B 9.6.3 Name: Access Type: Offset: Reset Value: Clock Mask Register CPU/HSB/PBA/PBBMASK Read/Write 0x008, 0x00C, 0x010, 0x014 - 31 30 29 28 MASK[31:24] 27 26 25 24 23 22 21 20 MASK[23:16] 19 18 17 16 15 14 13 12 MASK[15:8] 11 10 9 8 7 6 5 4 MASK[7:0] 3 2 1 0 52 32059I-06/2010 AT32UC3B * MASK: Clock Mask If bit n is cleared, the clock for module n is stopped. If bit n is set, the clock for module n is enabled according to the current power mode. The number of implemented bits in each mask register, as well as which module clock is controlled by each bit, is shown in Table 9-6. Table 9-6. Bit 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Maskable module clocks in AT32UC3B. CPUMASK OCD(1) HSBMASK FLASHC PBA bridge PBB bridge USBB PDCA PBAMASK INTC GPIO PDCA PM/RTC/EIC ADC SPI TWI USART0 USART1 USART2 PWM SSC TC ABDAC PBBMASK HMATRIX USBB FLASHC - SYSTIMER (COMPARE/COUNT REGISTERS CLK) - 31: 17 - - - Note: 1. This bit must be one if the user wishes to debug the device with a JTAG debugger. 53 32059I-06/2010 AT32UC3B 9.6.4 Name: Access Type: Offset: Reset Value: PLL Control Register PLL0,1 Read/Write 0x020, 0x024 0x00000000 31 23 15 7 - 30 22 14 6 - 29 28 27 PLLCOUNT 26 25 24 21 13 5 - 20 12 4 19 18 PLLMUL 17 16 11 10 PLLDIV 9 8 3 PLLOPT 2 1 PLLOSC 0 PLLEN * PLLCOUNT: PLL Count Specifies the number of slow clock cycles before ISR:LOCKn will be set after PLLn has been written, or after PLLn has been automatically re-enabled after exiting a sleep mode. * PLLMUL: PLL Multiply Factor * PLLDIV: PLL Division Factor These fields determine the ratio of the ouput frequency of the internal VCO of the PLL (fVCO) to the source oscillator frequency: fVCO = (PLLMUL+1)/(PLLDIV) * fOSC if PLLDIV > 0. fVCO = 2 * (PLLMUL+1) * fOSC if PLLDIV = 0. If PLLOPT[1] bit is set to 0: fPLL = fVCO. If PLLOPT[1] bit is set to 1: fPLL = fVCO / 2. Note that the PLLMUL field cannot be equal to 0 or 1, or the behavior of the PLL will be undefined. PLLDIV gives also the input frequency of the PLL (fIN): if the PLLDIV field is set to 0: fIN = fOSC. if the PLLDIV field is greater than 0: fIN = fOSC / (2 * PLLDIV). * PLLOPT: PLL Option Select the operating range for the PLL. PLLOPT[0]: Select the VCO frequency range. PLLOPT[1]: Enable the extra output divider. PLLOPT[2]: Disable the Wide-Bandwidth mode (Wide-Bandwidth mode allows a faster startup time and out-of-lock time). 54 32059I-06/2010 AT32UC3B Table 9-7. PLLOPT Fields Description in AT32UC3B Description PLLOPT[0]: VCO frequency 0 1 PLLOPT[1]: Output divider 0 1 PLLOPT[2] 0 1 Wide Bandwidth Mode enabled Wide Bandwidth Mode disabled fPLL = fvco fPLL = fvco/2 160MHz 0: Oscillator 0 is the source for the PLL. 1: Oscillator 1 is the source for the PLL. * PLLEN: PLL Enable 0: PLL is disabled. 1: PLL is enabled. 55 32059I-06/2010 AT32UC3B 9.6.5 Name: Access Type: Offset: Reset Value: Oscillator 0/1 Control Register OSCCTRL0,1 Read/Write 0x028, 0x02C 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 9 STARTUP 8 7 - 6 - 5 - 4 - 3 - 2 1 MODE 0 * STARTUP: Oscillator Startup Time Select startup time for the oscillator. Table 9-8. STARTUP 0 1 2 3 4 5 6 7 Startup time for oscillators 0 and 1 Number of RC oscillator clock cycle 0 64 128 2048 4096 8192 16384 Reserved Approximative Equivalent time (RCOsc = 115 kHz) 0 560 us 1.1 ms 18 ms 36 ms 71 ms 142 ms Reserved * MODE: Oscillator Mode Choose between crystal, or external clock 0: External clock connected on XIN, XOUT can be used as an I/O (no crystal). 1 to 3: reserved . 4: Crystal is connected to XIN/XOUT - Oscillator is used with gain G0 ( XIN from 0.4 5: Crystal is connected to XIN/XOUT - Oscillator is used with gain G1 ( XIN from 0.9 6: Crystal is connected to XIN/XOUT - Oscillator is used with gain G2 ( XIN from 3.0 7: Crystal is connected to XIN/XOUT - Oscillator is used with gain G3 ( XIN from 8.0 MHz to 0.9 MHz ). MHz to 3.0 MHz ). MHz to 8.0 MHz ). Mhz). 56 32059I-06/2010 AT32UC3B 9.6.6 Name: Access Type: Offset: Reset Value: 32 KHz Oscillator Control Register OSCCTRL32 Read/Write 0x030 0x00010000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 17 STARTUP 16 15 - 14 - 13 - 12 - 11 - 10 9 MODE 8 7 - 6 - 5 - 4 - 3 - 2 - 1 - 0 OSC32EN Note: This register is only reset by Power-On Reset * STARTUP: Oscillator Startup Time Select startup time for 32 KHz oscillator. Table 9-9. STARTUP 0 1 2 3 4 5 6 7 Startup time for 32 KHz oscillator Number of RC oscillator clock cycle 0 128 8192 16384 65536 131072 262144 524288 Approximative Equivalent time (RCOsc = 115 kHz) 0 1.1 ms 72.3 ms 143 ms 570 ms 1.1 s 2.3 s 4.6 s * MODE: Oscillator Mode Choose between crystal, or external clock. 0: External clock connected on XIN32, XOUT32 can be used as a I/O (no crystal). 1: Crystal is connected to XIN32/XOUT32. 2 to 7: reserved . * OSC32EN: Enable the 32 KHz oscillator 0: 32 KHz Oscillator is disabled. 1: 32 KHz Oscillator is enabled. 57 32059I-06/2010 AT32UC3B 9.6.7 Name: Access Type: Offset: Reset Value: Interrupt Enable Register IER Write-only 0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 BODDET 15 - 14 - 13 - 12 - 11 - 10 - 9 OSC32RDY 8 OSC1RDY 7 OSC0RDY 6 MSKRDY 5 CKRDY 4 - 3 - 2 - 1 LOCK1 0 LOCK0 Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. 58 32059I-06/2010 AT32UC3B 9.6.8 Name: Access Type: Offset: Reset Value: Interrupt Disable Register IDR Write-only 0x044 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 BODDET 15 - 14 - 13 - 12 - 11 - 10 - 9 OSC32RDY 8 OSC1RDY 7 OSC0RDY 6 MSKRDY 5 CKRDY 4 - 3 - 2 - 1 LOCK1 0 LOCK0 Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. 59 32059I-06/2010 AT32UC3B 9.6.9 Name: Access Type: Offset: Reset Value: Interrupt Mask Register IMR Read-only 0x048 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 BODDET 15 - 14 - 13 - 12 - 11 - 10 - 9 OSC32RDY 8 OSC1RDY 7 OSC0RDY 6 MSKRDY 5 CKRDY 4 - 3 - 2 - 1 LOCK1 0 LOCK0 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is set when the corresponding bit in IER is written to one. 60 32059I-06/2010 AT32UC3B 9.6.10 Name: Access Type: Offset: Reset Value: Interrupt Status Register ISR Read-only 0x04C 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 BODDET 15 - 14 - 13 - 12 - 11 - 10 - 9 OSC32RDY 8 OSC1RDY 7 OSC0RDY 6 MSKRDY 5 CKRDY 4 - 3 - 2 - 1 LOCK1 0 LOCK0 * BODDET: Brown out detection Set to 1 when 0 to 1 transition on POSCSR:BODDET bit is detected: BOD has detected that power supply is going below BOD reference value. * OSC32RDY: 32 KHz oscillator Ready Set to 1 when 0 to 1 transition on the POSCSR:OSC32RDY bit is detected: The 32 KHz oscillator is stable and ready to be used as clock source. * OSC1RDY: Oscillator 1 Ready Set to 1 when 0 to 1 transition on the POSCSR:OSC1RDY bit is detected: Oscillator 1 is stable and ready to be used as clock source. * OSC0RDY: Oscillator 0 Ready Set to 1 when 0 to 1 transition on the POSCSR:OSC1RDY bit is detected: Oscillator 1 is stable and ready to be used as clock source. * MSKRDY: Mask Ready Set to 1 when 0 to 1 transition on the POSCSR:MSKRDY bit is detected: Clocks are now masked according to the (CPU/HSB/PBA/PBB)_MASK registers. * CKRDY: Clock Ready 0: The CKSEL register has been written, and the new clock setting is not yet effective. 1: The synchronous clocks have frequencies as indicated in the CKSEL register. Note: Writing ICR:CKRDY to 1 has no effect. * LOCK1: PLL1 locked Set to 1 when 0 to 1 transition on the POSCSR:LOCK1 bit is detected: PLL 1 is locked and ready to be selected as clock source. * LOCK0: PLL0 locked Set to 1 when 0 to 1 transition on the POSCSR:LOCK0 bit is detected: PLL 0 is locked and ready to be selected as clock source. 61 32059I-06/2010 AT32UC3B 9.6.11 Name: Access Type: Offset: Reset Value: Interrupt Clear Register ICR Write-only 0x050 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 BODDET 15 - 14 - 13 - 12 - 11 - 10 - 9 OSC32RDY 8 OSC1RDY 7 OSC0RDY 6 MSKRDY 5 CKRDY 4 - 3 - 2 - 1 LOCK1 0 LOCK0 Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in ISR. 62 32059I-06/2010 AT32UC3B 9.6.12 Name: Access Type: Offset: Reset Value: 31 - Power and Oscillators Status Register POSCSR Read-only 0x054 0x00000000 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 BODDET 15 - 14 - 13 - 12 - 11 - 10 - 9 OSC32RDY 8 OSC1RDY 7 OSC0RDY 6 MSKRDY 5 CKRDY 4 - 3 - 2 WAKE 1 LOCK1 0 LOCK0 * BODDET: Brown out detection 0: No BOD event. 1: BOD has detected that power supply is going below BOD reference value. OSC32RDY: 32 KHz oscillator Ready 0: The 32 KHz oscillator is not enabled or not ready. 1: The 32 KHz oscillator is stable and ready to be used as clock source. OSC1RDY: OSC1 ready 0: Oscillator 1 not enabled or not ready. 1: Oscillator 1 is stable and ready to be used as clock source. OSC0RDY: OSC0 ready 0: Oscillator 0 not enabled or not ready. 1: Oscillator 0 is stable and ready to be used as clock source. MSKRDY: Mask ready 0: Mask register has been changed, masking in progress. 1: Clock are masked according to xxx_MASK. CKRDY: 0: The CKSEL register has been written, and the new clock setting is not yet effective. 1: The synchronous clocks have frequencies as indicated in the CKSEL register. LOCK1: PLL 1 locked 0:PLL 1 is unlocked. 1:PLL 1 is locked, and ready to be selected as clock source. LOCK0: PLL 0 locked 0: PLL 0 is unlocked. 1: PLL 0 is locked, and ready to be selected as clock source. * * * * * * * 63 32059I-06/2010 AT32UC3B 9.6.13 Name: Access Type: Offset: Reset Value: Generic Clock Control Register GCCTRL Read/Write 0x060 - 0x070 0x00000000 31 23 15 30 22 14 29 21 13 28 20 12 DIV[7:0] 27 19 11 26 18 10 25 17 9 24 16 8 7 - 6 - 5 - 4 DIVEN 3 - 2 CEN 1 PLLSEL 0 OSCSEL * * * * * There is one GCCTRL register per generic clock in the design. DIV: Division Factor DIVEN: Divide Enable 0: The generic clock equals the undivided source clock. 1: The generic clock equals the source clock divided by 2*(DIV+1). CEN: Clock Enable 0: Clock is stopped. 1: Clock is running. PLLSEL: PLL Select 0: Oscillator is source for the generic clock. 1: PLL is source for the generic clock. OSCSEL: Oscillator Select 0: Oscillator (or PLL) 0 is source for the generic clock. 1: Oscillator (or PLL) 1 is source for the generic clock. 64 32059I-06/2010 AT32UC3B 9.6.14 Name: Access Type: Offset: Reset Value: RC Oscillator Calibration Register RCCR Read/Write 0x0C0 - 31 30 29 28 KEY 27 26 25 24 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 FCD 15 - 14 - 13 - 12 - 11 - 10 - 9 CALIB[9:8] 8 7 6 5 4 CALIB[7:0] 3 2 1 0 * KEY: Register Write protection This field must be written twice, first with key value 0x55, then 0xAA, for a write operation to have an effect. * CALIB: Calibration Value Calibration Value for the RC oscillator. * FCD: Flash Calibration Done Set to 1 when the CALIB field has been updated by the Flash fuses after power-on reset or Flash fuses update. 0: Allow subsequent overwriting of the CALIB value by Flash fuses. 1: The CALIB value will not be updated again by Flash fuses. 65 32059I-06/2010 AT32UC3B 9.6.15 Name: Access Type: Offset: Reset Value: Bandgap Calibration Register BGCR Read/Write 0x0C4 - 31 30 29 28 KEY 27 26 25 24 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 FCD 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 1 CALIB 0 * KEY: Register Write protection This field must be written twice, first with key value 0x55, then 0xAA, for a write operation to have an effect. * CALIB: Calibration value Calibration value for Bandgap. See Electrical Characteristics for voltage values. * FCD: Flash Calibration Done Set to 1 when the CALIB field has been updated by the Flash fuses after power-on reset or Flash fuses update. 0: Allow subsequent overwriting of the CALIB value by Flash fuses. 1: The CALIB value will not be updated again by Flash fuses. 66 32059I-06/2010 AT32UC3B 9.6.16 Name:: Register access: Offset: Reset Value: Voltage Regulator Calibration Register VREGCR Read/Write 0x0C8 - 31 30 29 28 KEY 27 26 25 24 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 FCD 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 1 CALIB 0 * KEY: Register Write protection This field must be written twice, first with key value 0x55, then 0xAA, for a write operation to have an effect. * CALIB: Calibration value Calibration value for Voltage Regulator. See Electrical Characteristics for voltage values. * FCD: Flash Calibration Done Set to 1 when the CALIB field has been updated by the Flash fuses after power-on reset or Flash fuses update. 0: Allow subsequent overwriting of the CALIB value by Flash fuses. 1: The CALIB value will not be updated again by Flash fuses. 67 32059I-06/2010 AT32UC3B 9.6.17 Name: Access Type: Offset: Reset Value: BOD Level Register BOD Read/Write 0x0D0 - 31 30 29 28 KEY 27 26 25 24 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 FCD 15 - 14 - 13 - 12 - 11 - 10 - 9 CTRL 8 7 - 6 HYST 5 4 3 LEVEL 2 1 0 * KEY: Register Write protection This field must be written twice, first with key value 0x55, then 0xAA, for a write operation to have an effect. * FCD: BOD Fuse calibration done Set to 1 when CTRL, HYST and LEVEL fields has been updated by the Flash fuses after power-on reset or Flash fuses update. 0: Allow subsequent overwriting of the value by Flash fuses. 1: The CTRL, HYST and LEVEL values will not be updated again by Flash fuses. * CTRL: BOD Control 0: BOD is off. 1: BOD is enabled and can reset the chip. 2: BOD is enabled and but cannot reset the chip. Only interrupt will be sent to interrupt controller, if enabled in the IMR register. 3: BOD is off. * HYST: BOD Hysteresis 0: No hysteresis 1: Hysteresis On * LEVEL: BOD Level This field sets the triggering threshold of the BOD. See Electrical Characteristics for actual voltage levels. Note that any change to the LEVEL field of the BOD register should be done with the BOD deactivated to avoid spurious reset or interrupt. 68 32059I-06/2010 AT32UC3B 9.6.18 Name: Access Type: Offset: Reset Value: 31 - Reset Cause Register RCAUSE Read-only 0x140 Latest Reset Source 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 OCDRST 7 CPUERR 6 SLEEP 5 - 4 JTAG 3 WDT 2 EXT 1 BOD 0 POR * OCDRST: OCD Reset The CPU was reset because the RES strobe in the OCD Development Control register has been written to one. * CPUERR: CPU Error The CPU was reset because it had detected an illegal access. * SLEEP: The CPU was reset because it went to SHUTDOWN or STATIC sleep mode. * JTAG: JTAG reset The CPU was reset by setting the bit RC_CPU in the JTAG reset register. * WDT: Watchdog Reset The CPU was reset because of a watchdog time-out. * EXT: External Reset Pin The CPU was reset due to the RESET pin being asserted. * BOD: Brown-out Reset The CPU was reset due to the supply voltage being lower than the brown-out threshold level. * POR Power-on Reset The CPU was reset due to the supply voltage being lower than the power-on threshold level. 69 32059I-06/2010 AT32UC3B 9.6.19 Name: Access Type: Offset: Reset Value: Asynchronous Wake Up Enable Register AWEN Read/Write 0x144 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 - 1 - 0 USB_WAKEN * USB_WAKEN : USB Wake Up Enable 0: The USB wake up is disabled. 1: The USB wake up is enabled. 70 32059I-06/2010 AT32UC3B 9.6.20 Name: Access Type: Offset: Reset Value: 31 General Purpose Low-power Register 0/1 GPLP0,1 Read/Write 0x200, 0x204 0x00000000 30 29 28 DATA[31:24] 27 26 25 24 23 22 21 20 DATA[23:16] 19 18 17 16 15 14 13 12 DATA[15:8] 11 10 9 8 7 6 5 4 DATA[7:0] 3 2 1 0 These registers are general purpose 32-bit registers that are reset only by power-on-reset. Any other reset will keep the content of these registers untouched. User software can use these registers to save context variables in a very low power mode. 71 32059I-06/2010 AT32UC3B 10. Real Time Counter (RTC) Rev: 2.3.1.1 10.1 Features * 32-bit real-time counter with 16-bit prescaler * Clocked from RC oscillator or 32KHz oscillator * Long delays * * * * - Max timeout 272years High resolution: Max count frequency 16KHz Extremely low power consumption Available in all sleep modes except Static Interrupt on wrap 10.2 Overview The Real Time Counter (RTC) enables periodic interrupts at long intervals, or accurate measurement of real-time sequences. The RTC is fed from a 16-bit prescaler, which is clocked from the system RC oscillator or the 32KHz crystal oscillator. Any tapping of the prescaler can be selected as clock source for the RTC, enabling both high resolution and long timeouts. The prescaler cannot be written directly, but can be cleared by the user. The RTC can generate an interrupt when the counter wraps around the value stored in the top register (TOP), producing accurate periodic interrupts. 10.3 Block Diagram Figure 10-1. Real Time Counter Block Diagram CTRL CLK32 CLK_32 RCSYS 1 TOP EN PSEL 16-bit Prescaler 0 32-bit counter TOPI IRQ VAL 10.4 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 72 32059I-06/2010 AT32UC3B 10.4.1 Power Management The RTC remains operating in all sleep modes except Static mode. Interrupts are not available in DeepStop mode. 10.4.2 Clocks The RTC can use the system RC oscillator as clock source. This oscillator is always enabled whenever this module is active. Please refer to the Electrical Characteristics chapter for the characteristic frequency of this oscillator (fRC). The RTC can also use the 32 KHz crystal oscillator as clock source. This oscillator must be enabled before use. Please refer to the Power Manager chapter for details. The clock for the RTC bus interface (CLK_RTC) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. It is recommended to disable the RTC before disabling the clock, to avoid freezing the RTC in an undefined state. 10.4.3 Interrupts The RTC interrupt request line is connected to the interrupt controller. Using the RTC interrupt requires the interrupt controller to be programmed first. 10.4.4 Debug Operation The RTC prescaler is frozen during debug operation, unless the OCD system keeps peripherals running in debug operation. 10.5 10.5.1 Functional Description RTC Operation Source clock The RTC is enabled by writing a one to the Enable bit in the Control Register (CTRL.EN). The 16-bit prescaler will then increment on the selected clock. The prescaler cannot be read or written, but it can be reset by writing a one to the Prescaler Clear bit in CTRL register (CTRL.PCLR). The 32KHz Oscillator Select bit in CTRL register (CTRL.CLK32) selects either the RC oscillator or the 32 KHz oscillator as clock source (defined as INPUT in the formula below) for the prescaler. The Prescale Select field in CTRL register (CTRL.PSEL) selects the prescaler tapping, selecting the source clock for the RTC: 10.5.1.1 f RTC = f INPUT 2 ( PSEL + 1 ) 10.5.1.2 Counter operation When enabled, the RTC will increment until it reaches TOP, and then wraps to 0x0. The status bit TOPI in Interrupt Status Register (ISR) is set to one when this occurs. From 0x0 the counter will count TOP+1 cycles of the source clock before it wraps back to 0x0. 73 32059I-06/2010 AT32UC3B The RTC count value can be read from or written to the Value register (VAL). Due to synchronization, continuous reading of the VAL register with the lowest prescaler setting will skip every other value. 10.5.1.3 RTC interrupt The RTC interrupt is enabled by writing a one to the Top Interrupt bit in the Interrupt Enable Register (IER.TOPI), and is disabled by writing a one to the Top Interrupt bit in the Interrupt Disable Register (IDR.TOPI). The Interrupt Mask Register (IMR) can be read to see whether or not the interrupt is enabled. If enabled, an interrupt will be generated if the TOPI bit in the Interrupt Status Register (ISR) is set. The TOPI bit in ISR can be cleared by writing a one to the TOPI bit in the Interrupt Clear Register (ICR.TOPI). The RTC interrupt can wake the CPU from all sleep modes except DeepStop and Static modes. 10.5.1.4 RTC wakeup The RTC can also wake up the CPU directly without triggering an interrupt when the ISR.TOPI bit is set. In this case, the CPU will continue executing from the instruction following the sleep instruction. This direct RTC wake-up is enabled by writing a one to the Wake Enable bit in the CTRL register (CTRL.WAKEN). When the CPU wakes from sleep, the CTRL.WAKEN bit must be written to zero to clear the internal wake signal to the sleep controller, otherwise a new sleep instruction will have no effect. The RTC wakeup is available in all sleep modes except Static mode. The RTC wakeup can be configured independently of the RTC interrupt. 10.5.1.5 Busy bit Due to the crossing of clock domains, the RTC uses a few clock cycles to propagate the values stored in CTRL, TOP, and VAL to the RTC. The RTC Busy bit in CTRL (CTRL.BUSY) indicates that a register write is still going on and all writes to TOP, CTRL, and VAL will be discarded until the CTRL.BUSY bit goes low again. 74 32059I-06/2010 AT32UC3B 10.6 User Interface RTC Register Memory Map Register Control Register Value Register Top Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Interrupt Clear Register Register Name CTRL VAL TOP IER IDR IMR ISR ICR Access Read/Write Read/Write Read/Write Write-only Write-only Read-only Read-only Write-only Reset 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 Table 10-1. Offset 0x00 0x04 0x08 0x10 0x14 0x18 0x1C 0x20 75 32059I-06/2010 AT32UC3B 10.6.1 Name: Access Type: Offset: Reset Value: Control Register CTRL Read/Write 0x00 0x00000000 31 23 15 7 - 30 22 14 6 - 29 21 13 5 - 28 20 12 4 BUSY 27 19 11 26 18 10 PSEL 25 17 9 24 16 CLKEN 8 3 CLK32 2 WAKEN 1 PCLR 0 EN * CLKEN: Clock Enable 1: The clock is enabled. 0: The clock is disabled. PSEL: Prescale Select Selects prescaler bit PSEL as source clock for the RTC. BUSY: RTC Busy This bit is set when the RTC is busy and will discard writes to TOP, VAL, and CTRL. This bit is cleared when the RTC accepts writes to TOP, VAL, and CTRL. CLK32: 32 KHz Oscillator Select 1: The RTC uses the 32 KHz oscillator as clock source. 0: The RTC uses the RC oscillator as clock source. WAKEN: Wakeup Enable 1: The RTC wakes up the CPU from sleep modes. 0: The RTC does not wake up the CPU from sleep modes. PCLR: Prescaler Clear Writing a one to this bit clears the prescaler. Writing a zero to this bit has no effect. This bit always reads as zero. EN: Enable 1: The RTC is enabled. 0: The RTC is disabled. * * * * * * 76 32059I-06/2010 AT32UC3B 10.6.2 Name: Access Type: Offset: Reset Value: Value Register VAL Read/Write 0x04 0x00000000 31 30 29 28 VAL[31:24] 27 26 25 24 23 22 21 20 VAL[23:16] 19 18 17 16 15 14 13 12 VAL[15:8] 11 10 9 8 7 6 5 4 VAL[7:0] 3 2 1 0 * VAL[31:0]: RTC Value This value is incremented on every rising edge of the source clock. 77 32059I-06/2010 AT32UC3B 10.6.3 Name: Access Type: Offset: Reset Value: Top Register TOP Read/Write 0x08 0x00000000 31 30 29 28 VAL[31:24] 27 26 25 24 23 22 21 20 VAL[23:16] 19 18 17 16 15 14 13 12 VAL[15:8] 11 10 9 8 7 6 5 4 VAL[7:0] 3 2 1 0 * VAL[31:0]: RTC Top Value VAL wraps at this value. 78 32059I-06/2010 AT32UC3B 10.6.4 Name: Access Type: Offset: Reset Value: Interrupt Enable Register IER Write-only 0x10 0x00000000 31 23 15 7 - 30 22 14 6 - 29 21 13 5 - 28 20 12 4 - 27 19 11 3 - 26 18 10 2 - 25 17 9 1 - 24 16 8 0 TOPI Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. 79 32059I-06/2010 AT32UC3B 10.6.5 Name: Access Type: Offset: Reset Value: Interrupt Disable Register IDR Write-only 0x14 0x00000000 31 23 15 7 - 30 22 14 6 - 29 21 13 5 - 28 20 12 4 - 27 19 11 3 - 26 18 10 2 - 25 17 9 1 - 24 16 8 0 TOPI Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. 80 32059I-06/2010 AT32UC3B 10.6.6 Name: Access Type: Offset: Reset Value: Interrupt Mask Register IMR Read-only 0x18 0x00000000 31 23 15 7 - 30 22 14 6 - 29 21 13 5 - 28 20 12 4 - 27 19 11 3 - 26 18 10 2 - 25 17 9 1 - 24 16 8 0 TOPI 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is set when the corresponding bit in IER is written to one. 81 32059I-06/2010 AT32UC3B 10.6.7 Name: Access Type: Offset: Reset Value: Interrupt Status Register ISR Read-only 0x1C 0x00000000 31 23 15 7 - 30 22 14 6 - 29 21 13 5 - 28 20 12 4 - 27 19 11 3 - 26 18 10 2 - 25 17 9 1 - 24 16 8 0 TOPI * TOPI: Top Interrupt This bit is set when VAL has wrapped at its top value. This bit is cleared when the corresponding bit in ICR is written to one. 82 32059I-06/2010 AT32UC3B 10.6.8 Name: Access Type: Offset: Reset Value: Interrupt Clear Register ICR Write-only 0x20 0x00000000 31 23 15 7 - 30 22 14 6 - 29 21 13 5 - 28 20 12 4 - 27 19 11 3 - 26 18 10 2 - 25 17 9 1 - 24 16 8 0 TOPI Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in SR and the corresponding interrupt request. 83 32059I-06/2010 AT32UC3B 11. Watchdog Timer (WDT) Rev: 2.3.1.1 11.1 Features * Watchdog timer counter with 32-bit prescaler * Clocked from the system RC oscillator (RCSYS) 11.2 Overview The Watchdog Timer (WDT) has a prescaler generating a time-out period. This prescaler is clocked from the RC oscillator. The watchdog timer must be periodically reset by software within the time-out period, otherwise, the device is reset and starts executing from the boot vector. This allows the device to recover from a condition that has caused the system to be unstable. 11.3 Block Diagram Figure 11-1. WDT Block Diagram CLR RCSYS 32-bit Prescaler Watchdog Detector Watchdog Reset EN CTRL 11.4 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 11.4.1 Power Management When the WDT is enabled, the WDT remains clocked in all sleep modes, and it is not possible to enter Static mode. 11.4.2 Clocks The WDT can use the system RC oscillator (RCSYS) as clock source. This oscillator is always enabled whenever these modules are active. Please refer to the Electrical Characteristics chapter for the characteristic frequency of this oscillator (fRC). 11.4.3 Debug Operation The WDT prescaler is frozen during debug operation, unless the On-Chip Debug (OCD) system keeps peripherals running in debug operation. 84 32059I-06/2010 AT32UC3B 11.5 Functional Description The WDT is enabled by writing a one to the Enable bit in the Control register (CTRL.EN). This also enables the system RC clock (CLK_RCSYS) for the prescaler. The Prescale Select field (PSEL) in the CTRL register selects the watchdog time-out period: TWDT = 2(PSEL+1) / fRC The next time-out period will begin as soon as the watchdog reset has occurred and count down during the reset sequence. Care must be taken when selecting the PSEL field value so that the time-out period is greater than the startup time of the chip, otherwise a watchdog reset can reset the chip before any code has been run. To avoid accidental disabling of the watchdog, the CTRL register must be written twice, first with the KEY field set to 0x55, then 0xAA without changing the other bits. Failure to do so will cause the write operation to be ignored, and the CTRL register value will not change. The Clear register (CLR) must be written with any value with regular intervals shorter than the watchdog time-out period. Otherwise, the device will receive a soft reset, and the code will start executing from the boot vector. When the WDT is enabled, it is not possible to enter Static mode. Attempting to do so will result in entering Shutdown mode, leaving the WDT operational. 11.6 User Interface WDT Register Memory Map Register Control Register Clear Register Register Name CTRL CLR Access Read/Write Write-only Reset 0x00000000 0x00000000 Table 11-1. Offset 0x00 0x04 85 32059I-06/2010 AT32UC3B 11.6.1 Name: Access Type: Offset: Reset Value: Control Register CTRL Read/Write 0x00 0x00000000 31 30 29 28 KEY 27 26 25 24 23 15 7 - 22 14 6 - 21 13 5 - 20 12 19 11 18 10 PSEL 17 9 16 8 4 - 3 - 2 - 1 - 0 EN * KEY: Write protection key This field must be written twice, first with key value 0x55, then 0xAA, for a write operation to be effective. This field always reads as zero. * PSEL: Prescale Select PSEL is used as watchdog timeout period. * EN: WDT Enable 1: WDT is enabled. 0: WDT is disabled. 86 32059I-06/2010 AT32UC3B 11.6.2 Name: Access Type: Offset: Reset Value: Clear Register CLR Write-only 0x04 0x00000000 31 30 29 28 CLR[31:24] 27 26 25 24 23 22 21 20 CLR[23:16] 19 18 17 16 15 14 13 12 CLR[15:8] 11 10 9 8 7 6 5 4 CLR[7:0] 3 2 1 0 * CLR: Writing periodically any value to this field when the WDT is enabled, within the watchdog time-out period, will prevent a watchdog reset. This field always reads as zero. 87 32059I-06/2010 AT32UC3B 12. Interrupt Controller (INTC) Rev: 1.0.1.5 12.1 Features * Autovectored low latency interrupt service with programmable priority - 4 priority levels for regular, maskable interrupts - One Non-Maskable Interrupt * Up to 64 groups of interrupts with up to 32 interrupt requests in each group 12.2 Overview The INTC collects interrupt requests from the peripherals, prioritizes them, and delivers an interrupt request and an autovector to the CPU. The AVR32 architecture supports 4 priority levels for regular, maskable interrupts, and a Non-Maskable Interrupt (NMI). The INTC supports up to 64 groups of interrupts. Each group can have up to 32 interrupt request lines, these lines are connected to the peripherals. Each group has an Interrupt Priority Register (IPR) and an Interrupt Request Register (IRR). The IPRs are used to assign a priority level and an autovector to each group, and the IRRs are used to identify the active interrupt request within each group. If a group has only one interrupt request line, an active interrupt group uniquely identifies the active interrupt request line, and the corresponding IRR is not needed. The INTC also provides one Interrupt Cause Register (ICR) per priority level. These registers identify the group that has a pending interrupt of the corresponding priority level. If several groups have a pending interrupt of the same level, the group with the lowest number takes priority. 12.3 Block Diagram Figure 12-1 gives an overview of the INTC. The grey boxes represent registers that can be accessed via the user interface. The interrupt requests from the peripherals (IREQn) and the NMI are input on the left side of the figure. Signals to and from the CPU are on the right side of the figure. 88 32059I-06/2010 AT32UC3B Figure 12-1. INTC Block Diagram Interrupt Controller NMIREQ CPU Masks SREG Masks I[3-0]M GM OR IRRn IREQ63 IREQ34 IREQ33 IREQ32 GrpReqN ValReqN IPRn . . . GrpReq1 OR IRR1 Request Masking ValReq1 IPR1 INT_level, offset . . . INT_level, offset . . . INTLEVEL Prioritizer AUTOVECTOR IREQ31 IREQ2 IREQ1 IREQ0 OR IRR0 IRR Registers GrpReq0 ValReq0 IPR0 INT_level, offset IPR Registers ICR Registers 12.4 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 12.4.1 Power Management If the CPU enters a sleep mode that disables CLK_SYNC, the INTC will stop functioning and resume operation after the system wakes up from sleep mode. 12.4.2 Clocks The clock for the INTC bus interface (CLK_INTC) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. The INTC sampling logic runs on a clock which is stopped in any of the sleep modes where the system RC oscillator is not running. This clock is referred to as CLK_SYNC. This clock is enabled at reset, and only turned off in sleep modes where the system RC oscillator is stopped. 12.4.3 Debug Operation When an external debugger forces the CPU into debug mode, the INTC continues normal operation. 12.5 Functional Description All of the incoming interrupt requests (IREQs) are sampled into the corresponding Interrupt Request Register (IRR). The IRRs must be accessed to identify which IREQ within a group that is active. If several IREQs within the same group are active, the interrupt service routine must prioritize between them. All of the input lines in each group are logically ORed together to form the GrpReqN lines, indicating if there is a pending interrupt in the corresponding group. The Request Masking hardware maps each of the GrpReq lines to a priority level from INT0 to INT3 by associating each group with the Interrupt Level (INTLEVEL) field in the corresponding 89 32059I-06/2010 AT32UC3B Interrupt Priority Register (IPR). The GrpReq inputs are then masked by the mask bits from the CPU status register. Any interrupt group that has a pending interrupt of a priority level that is not masked by the CPU status register, gets its corresponding ValReq line asserted. Masking of the interrupt requests is done based on five interrupt mask bits of the CPU status register, namely Interrupt Level 3 Mask (I3M) to Interrupt Level 0 Mask (I0M), and Global Interrupt Mask (GM). An interrupt request is masked if either the GM or the corresponding interrupt level mask bit is set. The Prioritizer hardware uses the ValReq lines and the INTLEVEL field in the IPRs to select the pending interrupt of the highest priority. If an NMI interrupt request is pending, it automatically gets the highest priority of any pending interrupt. If several interrupt groups of the highest pending interrupt level have pending interrupts, the interrupt group with the lowest number is selected. The INTLEVEL and handler autovector offset (AUTOVECTOR) of the selected interrupt are transmitted to the CPU for interrupt handling and context switching. The CPU does not need to know which interrupt is requesting handling, but only the level and the offset of the handler address. The IRR registers contain the interrupt request lines of the groups and can be read via user interface registers for checking which interrupts of the group are actually active. The delay through the INTC from the peripheral interrupt request is set until the interrupt request to the CPU is set is three cycles of CLK_SYNC. 12.5.1 Non-Maskable Interrupts A NMI request has priority over all other interrupt requests. NMI has a dedicated exception vector address defined by the AVR32 architecture, so AUTOVECTOR is undefined when INTLEVEL indicates that an NMI is pending. 12.5.2 CPU Response When the CPU receives an interrupt request it checks if any other exceptions are pending. If no exceptions of higher priority are pending, interrupt handling is initiated. When initiating interrupt handling, the corresponding interrupt mask bit is set automatically for this and lower levels in status register. E.g, if an interrupt of level 3 is approved for handling, the interrupt mask bits I3M, I2M, I1M, and I0M are set in status register. If an interrupt of level 1 is approved, the masking bits I1M and I0M are set in status register. The handler address is calculated by logical OR of the AUTOVECTOR to the CPU system register Exception Vector Base Address (EVBA). The CPU will then jump to the calculated address and start executing the interrupt handler. Setting the interrupt mask bits prevents the interrupts from the same and lower levels to be passed through the interrupt controller. Setting of the same level mask bit prevents also multiple requests of the same interrupt to happen. It is the responsibility of the handler software to clear the interrupt request that caused the interrupt before returning from the interrupt handler. If the conditions that caused the interrupt are not cleared, the interrupt request remains active. 12.5.3 Clearing an Interrupt Request Clearing of the interrupt request is done by writing to registers in the corresponding peripheral module, which then clears the corresponding NMIREQ/IREQ signal. The recommended way of clearing an interrupt request is a store operation to the controlling peripheral register, followed by a dummy load operation from the same register. This causes a 90 32059I-06/2010 AT32UC3B pipeline stall, which prevents the interrupt from accidentally re-triggering in case the handler is exited and the interrupt mask is cleared before the interrupt request is cleared. 91 32059I-06/2010 AT32UC3B 12.6 User Interface Table 12-1. Offset 0x000 0x004 ... 0x0FC 0x100 0x104 ... 0x1FC 0x200 0x204 0x208 0x20C INTC Register Memory Map Register Interrupt Priority Register 0 Interrupt Priority Register 1 ... Interrupt Priority Register 63 Interrupt Request Register 0 Interrupt Request Register 1 ... Interrupt Request Register 63 Interrupt Cause Register 3 Interrupt Cause Register 2 Interrupt Cause Register 1 Interrupt Cause Register 0 Register Name IPR0 IPR1 ... IPR63 IRR0 IRR1 ... IRR63 ICR3 ICR2 ICR1 ICR0 Access Read/Write Read/Write ... Read/Write Read-only Read-only ... Read-only Read-only Read-only Read-only Read-only Reset 0x00000000 0x00000000 ... 0x00000000 N/A N/A ... N/A N/A N/A N/A N/A 92 32059I-06/2010 AT32UC3B 12.6.1 Name: Access Type: Offset: Reset Value: Interrupt Priority Registers IPR0...IPR63 Read/Write 0x000 - 0x0FC 0x00000000 31 30 29 21 13 28 20 12 27 19 11 26 18 10 25 17 9 24 16 8 INTLEVEL[1:0] 23 15 7 22 14 6 AUTOVECTOR[13:8] 5 4 3 2 1 0 AUTOVECTOR[7:0] * INTLEVEL: Interrupt Level Indicates the EVBA-relative offset of the interrupt handler of the corresponding group: 00: INT0: Lowest priority 01: INT1 10: INT2 11: INT3: Highest priority * AUTOVECTOR: Autovector Address Handler offset is used to give the address of the interrupt handler. The least significant bit should be written to zero to give halfword alignment. 93 32059I-06/2010 AT32UC3B 12.6.2 Name: Access Type: Offset: Reset Value: Interrupt Request Registers IRR0...IRR63 Read-only 0x0FF - 0x1FC N/A 31 IRR[32*x+31] 23 IRR[32*x+23] 15 IRR[32*x+15] 7 IRR[32*x+7] 30 IRR[32*x+30] 22 IRR[32*x+22] 14 IRR[32*x+14] 6 IRR[32*x+6] 29 IRR[32*x+29] 21 IRR[32*x+21] 13 IRR[32*x+13] 5 IRR[32*x+5] 28 IRR[32*x+28] 20 IRR[32*x+20] 12 IRR[32*x+12] 4 IRR[32*x+4] 27 IRR[32*x+27] 19 IRR[32*x+19] 11 IRR[32*x+11] 3 IRR[32*x+3] 26 IRR[32*x+26] 18 IRR[32*x+18] 10 IRR[32*x+10] 2 IRR[32*x+2] 25 IRR[32*x+25] 17 IRR[32*x+17] 9 IRR[32*x+9] 1 IRR[32*x+1] 24 IRR[32*x+24] 16 IRR[32*x+16] 8 IRR[32*x+8] 0 IRR[32*x+0] * IRR: Interrupt Request line This bit is cleared when no interrupt request is pending on this input request line. This bit is set when an interrupt request is pending on this input request line. The are 64 IRRs, one for each group. Each IRR has 32 bits, one for each possible interrupt request, for a total of 2048 possible input lines. The IRRs are read by the software interrupt handler in order to determine which interrupt request is pending. The IRRs are sampled continuously, and are read-only. 94 32059I-06/2010 AT32UC3B 12.6.3 Name: Access Type: Offset: Reset Value: Interrupt Cause Registers ICR0...ICR3 Read-only 0x200 - 0x20C N/A 31 23 15 7 - 30 22 14 6 - 29 21 13 5 28 20 12 4 27 19 11 3 CAUSE 26 18 10 2 25 17 9 1 24 16 8 0 * CAUSE: Interrupt Group Causing Interrupt of Priority n ICRn identifies the group with the highest priority that has a pending interrupt of level n. This value is only defined when at least one interrupt of level n is pending. 95 32059I-06/2010 AT32UC3B 12.7 Interrupt Request Signal Map The various modules may output Interrupt request signals. These signals are routed to the Interrupt Controller (INTC), described in a later chapter. The Interrupt Controller supports up to 64 groups of interrupt requests. Each group can have up to 32 interrupt request signals. All interrupt signals in the same group share the same autovector address and priority level. Refer to the documentation for the individual submodules for a description of the semantics of the different interrupt requests. The interrupt request signals are connected to the INTC as follows. Table 12-2. Group 0 Interrupt Request Signal Map Line 0 0 1 2 3 4 Module AVR32 UC CPU with optional MPU and optional OCD External Interrupt Controller External Interrupt Controller External Interrupt Controller External Interrupt Controller External Interrupt Controller External Interrupt Controller External Interrupt Controller External Interrupt Controller Real Time Counter Power Manager General Purpose Input/Output Controller General Purpose Input/Output Controller General Purpose Input/Output Controller General Purpose Input/Output Controller General Purpose Input/Output Controller General Purpose Input/Output Controller Peripheral DMA Controller Peripheral DMA Controller Peripheral DMA Controller Peripheral DMA Controller Peripheral DMA Controller Peripheral DMA Controller Peripheral DMA Controller Flash Controller Universal Synchronous/Asynchronous Receiver/Transmitter Signal SYSREG COMPARE EIC 0 EIC 1 EIC 2 EIC 3 EIC 4 EIC 5 EIC 6 EIC 7 RTC PM GPIO 0 GPIO 1 GPIO 2 GPIO 3 GPIO 4 GPIO 5 PDCA 0 PDCA 1 PDCA 2 PDCA 3 PDCA 4 PDCA 5 PDCA 6 FLASHC USART0 1 5 6 7 8 9 0 1 2 2 3 4 5 0 1 2 3 3 4 5 6 4 5 0 0 96 32059I-06/2010 AT32UC3B Table 12-2. 6 7 9 11 12 13 Interrupt Request Signal Map 0 0 0 0 0 0 0 Universal Synchronous/Asynchronous Receiver/Transmitter Universal Synchronous/Asynchronous Receiver/Transmitter Serial Peripheral Interface Two-wire Interface Pulse Width Modulation Controller Synchronous Serial Controller Timer/Counter Timer/Counter Timer/Counter Analog to Digital Converter USB 2.0 Interface Audio Bitstream DAC USART1 USART2 SPI TWI PWM SSC TC0 TC1 TC2 ADC USBB ABDAC 14 1 2 15 17 18 0 0 0 97 32059I-06/2010 AT32UC3B 13. External Interrupt Controller (EIC) Rev: 2.3.1.0 13.1 Features * * * * * * * * Dedicated interrupt request for each interrupt Individually maskable interrupts Interrupt on rising or falling edge Interrupt on high or low level Asynchronous interrupts for sleep modes without clock Filtering of interrupt lines Maskable NMI interrupt Keypad scan support 13.2 Overview The External Interrupt Controller (EIC) allows pins to be configured as external interrupts. Each external interrupt has its own interrupt request and can be individually masked. Each external interrupt can generate an interrupt on rising or falling edge, or high or low level. Every interrupt input has a configurable filter to remove spikes from the interrupt source. Every interrupt pin can also be configured to be asynchronous in order to wake up the part from sleep modes where the CLK_SYNC clock has been disabled. A Non-Maskable Interrupt (NMI) is also supported. This has the same properties as the other external interrupts, but is connected to the NMI request of the CPU, enabling it to interrupt any other interrupt mode. The EIC can wake up the part from sleep modes without triggering an interrupt. In this mode, code execution starts from the instruction following the sleep instruction. The External Interrupt Controller has support for keypad scanning for keypads laid out in rows and columns. Columns are driven by a separate set of scanning outputs, while rows are sensed by the external interrupt lines. The pressed key will trigger an interrupt, which can be identified through the user registers of the module. 98 32059I-06/2010 AT32UC3B 13.3 Block Diagram Figure 13-1. EIC Block Diagram LEVEL MODE EDGE ASYNC EN D IS P o la r it y c o n tro l A s y n c h ro n u s d e te c to r LEVEL MODE EDGE IC R CTRL IE R ID R E n a b le E X T IN T n NMI CTRL F IL T E R IN T n IS R M ask IM R IR Q n F ilt e r E d g e /L e v e l D e te c to r CLK_SYNC C LK_R C SYS W ake d e te c t E IC _ W A K E P r e s c a le r PRESC S h if t e r P IN SCANm EN SCAN 13.4 I/O Lines Description Table 13-1. Pin Name NMI EXTINTn SCANm I/O Lines Description Pin Description Non-Maskable Interrupt External Interrupt Type Input Input Output Keypad scan pin m 13.5 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 13.5.1 I/O Lines The external interrupt pins (EXTINTn and NMI) are multiplexed with I/O lines. To generate an external interrupt from an external source the source pin must be configured as an input pins by the I/O Controller. It is also possible to trigger the interrupt by driving these pins from registers in the I/O Controller, or another peripheral output connected to the same pin. 99 32059I-06/2010 AT32UC3B 13.5.2 Power Management All interrupts are available in all sleep modes as long as the EIC module is powered. However, in sleep modes where CLK_SYNC is stopped, the interrupt must be configured to asynchronous mode. 13.5.3 Clocks The clock for the EIC bus interface (CLK_EIC) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. The filter and synchronous edge/level detector runs on a clock which is stopped in any of the sleep modes where the system RC oscillator is not running. This clock is referred to as CLK_SYNC. Refer to the Module Configuration section at the end of this chapter for details. The Keypad scan function operates on the system RC oscillator clock CLK_RCSYS. 13.5.4 Interrupts The external interrupt request lines are connected to the interrupt controller. Using the external interrupts requires the interrupt controller to be programmed first. Using the Non-Maskable Interrupt does not require the interrupt controller to be programmed. 13.5.5 Debug Operation The EIC is frozen during debug operation, unless the OCD system keeps peripherals running during debug operation. 13.6 13.6.1 Functional Description External Interrupts The external interrupts are not enabled by default, allowing the proper interrupt vectors to be set up by the CPU before the interrupts are enabled. Each external interrupt INTn can be configured to produce an interrupt on rising or falling edge, or high or low level. External interrupts are configured by the MODE, EDGE, and LEVEL registers. Each interrupt n has a bit INTn in each of these registers. Writing a zero to the INTn bit in the MODE register enables edge triggered interrupts, while writing a one to the bit enables level triggered interrupts. If INTn is configured as an edge triggered interrupt, writing a zero to the INTn bit in the EDGE register will cause the interrupt to be triggered on a falling edge on EXTINTn, while writing a one to the bit will cause the interrupt to be triggered on a rising edge on EXTINTn. If INTn is configured as a level triggered interrupt, writing a zero to the INTn bit in the LEVEL register will cause the interrupt to be triggered on a low level on EXTINTn, while writing a one to the bit will cause the interrupt to be triggered on a high level on EXTINTn. Each interrupt has a corresponding bit in each of the interrupt control and status registers. Writing a one to the INTn bit in the Interrupt Enable Register (IER) enables the external interrupt from pin EXTINTn to propagate from the EIC to the interrupt controller, while writing a one to INTn bit in the Interrupt Disable Register (IDR) disables this propagation. The Interrupt Mask Register (IMR) can be read to check which interrupts are enabled. When an interrupt triggers, the corresponding bit in the Interrupt Status Register (ISR) will be set. This bit remains set until a 100 32059I-06/2010 AT32UC3B one is written to the corresponding bit in the Interrupt Clear Register (ICR) or the interrupt is disabled. Writing a one to the INTn bit in the Enable Register (EN) enables the external interrupt on pin EXTINTn, while writing a one to INTn bit in the Disable Register (DIS) disables the external interrupt. The Control Register (CTRL) can be read to check which interrupts are enabled. If a bit in the CTRL register is set, but the corresponding bit in IMR is not set, an interrupt will not propagate to the interrupt controller. However, the corresponding bit in ISR will be set, and EIC_WAKE will be set. If the CTRL.INTn bit is zero, then the corresponding bit in ISR will always be zero. Disabling an external interrupt by writing to the DIS.INTn bit will clear the corresponding bit in ISR. 13.6.2 Synchronization and Filtering of External Interrupts In synchronous mode the pin value of the EXTINTn pin is synchronized to CLK_SYNC, so spikes shorter than one CLK_SYNC cycle are not guaranteed to produce an interrupt. The synchronization of the EXTINTn to CLK_SYNC will delay the propagation of the interrupt to the interrupt controller by two cycles of CLK_SYNC, see Figure 13-2 on page 101 and Figure 13-3 on page 102 for examples (FILTER off). It is also possible to apply a filter on EXTINTn by writing a one to INTn bit in the FILTER register. This filter is a majority voter, if the condition for an interrupt is true for more than one of the latest three cycles of CLK_SYNC the interrupt will be set. This will additionally delay the propagation of the interrupt to the interrupt controller by one or two cycles of CLK_SYNC, see Figure 13-2 on page 101 and Figure 13-3 on page 102 for examples (FILTER on). Figure 13-2. Timing Diagram, Synchronous Interrupts, High Level or Rising Edge CLK_SYNC EXTINTn/NMI ISR.INTn: FILTER off ISR.INTn: FILTER on 101 32059I-06/2010 AT32UC3B Figure 13-3. Timing Diagram, Synchronous Interrupts, Low Level or Falling Edge CLK_SYNC EXTINTn/NMI ISR.INTn: FILTER off ISR.INTn: FILTER on 13.6.3 Non-Maskable Interrupt The NMI supports the same features as the external interrupts, and is accessed through the same registers. The description in Section 13.6.1 should be followed, accessing the NMI bit instead of the INTn bits. The NMI is non-maskable within the CPU in the sense that it can interrupt any other execution mode. Still, as for the other external interrupts, the actual NMI input can be enabled and disabled by accessing the registers in the EIC. 13.6.4 Asynchronous Interrupts Each external interrupt can be made asynchronous by writing a one to INTn in the ASYNC register. This will route the interrupt signal through the asynchronous path of the module. All edge interrupts will be interpreted as level interrupts and the filter is disabled. If an interrupt is configured as edge triggered interrupt in asynchronous mode, a zero in EDGE.INTn will be interpreted as low level, and a one in EDGE.INTn will be interpreted as high level. EIC_WAKE will be set immediately after the source triggers the interrupt, while the corresponding bit in ISR and the interrupt to the interrupt controller will be set on the next rising edge of CLK_SYNC. Please refere to Figure 13-4 on page 103 for details. When CLK_SYNC is stopped only asynchronous interrupts remain active, and any short spike on this interrupt will wake up the device. EIC_WAKE will restart CLK_SYNC and ISR will be updated on the first rising edge of CLK_SYNC. 102 32059I-06/2010 AT32UC3B Figure 13-4. Timing Diagram, Asynchronous Interrupts CLK_SYNC CLK_SYNC EXTINTn/NMI EXTINTn/NMI ISR.INTn: rising EDGE or high LEVEL EIC_WAKE: rising EDGE or high LEVEL ISR.INTn: rising EDGE or high LEVEL EIC_WAKE: rising EDGE or high LEVEL 13.6.5 Wakeup The external interrupts can be used to wake up the part from sleep modes. The wakeup can be interpreted in two ways. If the corresponding bit in IMR is one, then the execution starts at the interrupt handler for this interrupt. If the bit in IMR is zero, then the execution starts from the next instruction after the sleep instruction. 103 32059I-06/2010 AT32UC3B 13.6.6 Keypad scan support The External Interrupt Controller also includes support for keypad scanning. The keypad scan feature is compatible with keypads organized as rows and columns, where a row is shorted against a column when a key is pressed. The rows should be connected to the external interrupt pins with pull-ups enabled in the I/O Controller. These external interrupts should be enabled as low level or falling edge interrupts. The columns should be connected to the available scan pins. The I/O Controller must be configured to let the required scan pins be controlled by the EIC. Unused external interrupt or scan pins can be left controlled by the I/O Controller or other peripherals. The Keypad Scan function is enabled by writing SCAN.EN to 1, which starts the keypad scan counter. The SCAN outputs are tri-stated, except SCAN[0], which is driven to zero. After 2(SCAN.PRESC+1) RC clock cycles this pattern is left shifted, so that SCAN[1] is driven to zero while the other outputs are tri-stated. This sequence repeats infinitely, wrapping from the most significant SCAN pin to SCAN[0]. When a key is pressed, the pulled-up row is driven to zero by the column, and an external interrupt triggers. The scanning stops, and the software can then identify the key pressed by the interrupt status register and the SCAN.PINS value. The scanning stops whenever there is an active interrupt request from the EIC to the CPU. When the CPU clears the interrupt flags, scanning resumes. 104 32059I-06/2010 AT32UC3B 13.7 User Interface EIC Register Memory Map Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Interrupt Clear Register Mode Register Edge Register Level Register Filter Register Test Register Asynchronous Register Scan Register Enable Register Disable Register Control Register Register Name IER IDR IMR ISR ICR MODE EDGE LEVEL FILTER TEST ASYNC SCAN EN DIS CTRL Access Write-only Write-only Read-only Read-only Write-only Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Write-only Write-only Read-only Reset 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 Table 13-2. Offset 0x000 0x004 0x008 0x00C 0x010 0x014 0x018 0x01C 0x020 0x024 0x028 0x2C 0x030 0x034 0x038 105 32059I-06/2010 AT32UC3B 13.7.1 Name: Access Type: Offset: Reset Value: Interrupt Enable Register IER Write-only 0x000 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n Writing a zero to this bit has no effect. Writing a one to this bit will set the corresponding bit in IMR. * NMI: Non-Maskable Interrupt Writing a zero to this bit has no effect. Writing a one to this bit will set the corresponding bit in IMR. 106 32059I-06/2010 AT32UC3B 13.7.2 Name: Access Type: Offset: Reset Value: Interrupt Disable Register IDR Write-only 0x004 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n Writing a zero to this bit has no effect. Writing a one to this bit will clear the corresponding bit in IMR. * NMI: Non-Maskable Interrupt Writing a zero to this bit has no effect. Writing a one to this bit will clear the corresponding bit in IMR. 107 32059I-06/2010 AT32UC3B 13.7.3 Name: Access Type: Offset: Reset Value: Interrupt Mask Register IMR Read-only 0x008 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. This bit is cleared when the corresponding bit in IDR is written to one. This bit is set when the corresponding bit in IER is written to one. * NMI: Non-Maskable Interrupt 0: The Non-Maskable Interrupt is disabled. 1: The Non-Maskable Interrupt is enabled. This bit is cleared when the corresponding bit in IDR is written to one. This bit is set when the corresponding bit in IER is written to one. 108 32059I-06/2010 AT32UC3B 13.7.4 Name: Access Type: Offset: Reset Value: Interrupt Status Register ISR Read-only 0x00C 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n 0: An interrupt event has not occurred 1: An interrupt event has occurred This bit is cleared by writing a one to the corresponding bit in ICR. * NMI: Non-Maskable Interrupt 0: An interrupt event has not occurred 1: An interrupt event has occurred This bit is cleared by writing a one to the corresponding bit in ICR. 109 32059I-06/2010 AT32UC3B 13.7.5 Name: Access Type: Offset: Reset Value: Interrupt Clear Register ICR Write-only 0x010 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n Writing a zero to this bit has no effect. Writing a one to this bit will clear the corresponding bit in ISR. * NMI: Non-Maskable Interrupt Writing a zero to this bit has no effect. Writing a one to this bit will clear the corresponding bit in ISR. 110 32059I-06/2010 AT32UC3B 13.7.6 Name: Access Type: Offset: Reset Value: Mode Register MODE Read/Write 0x014 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n 0: The external interrupt is edge triggered. 1: The external interrupt is level triggered. * NMI: Non-Maskable Interrupt 0: The Non-Maskable Interrupt is edge triggered. 1: The Non-Maskable Interrupt is level triggered. 111 32059I-06/2010 AT32UC3B 13.7.7 Name: Access Type: Offset: Reset Value: Edge Register EDGE Read/Write 0x018 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n 0: The external interrupt triggers on falling edge. 1: The external interrupt triggers on rising edge. * NMI: Non-Maskable Interrupt 0: The Non-Maskable Interrupt triggers on falling edge. 1: The Non-Maskable Interrupt triggers on rising edge. 112 32059I-06/2010 AT32UC3B 13.7.8 Name: Access Type: Offset: Reset Value: Level Register LEVEL Read/Write 0x01C 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n 0: The external interrupt triggers on low level. 1: The external interrupt triggers on high level. * NMI: Non-Maskable Interrupt 0: The Non-Maskable Interrupt triggers on low level. 1: The Non-Maskable Interrupt triggers on high level. 113 32059I-06/2010 AT32UC3B 13.7.9 Name: Access Type: Offset: Reset Value: Filter Register FILTER Read/Write 0x020 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n 0: The external interrupt is not filtered. 1: The external interrupt is filtered. * NMI: Non-Maskable Interrupt 0: The Non-Maskable Interrupt is not filtered. 1: The Non-Maskable Interrupt is filtered. 114 32059I-06/2010 AT32UC3B 13.7.10 Name: Access Type: Offset: Reset Value: Test Register TEST Read/Write 0x024 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * TESTEN: Test Enable 0: This bit disables external interrupt test mode. 1: This bit enables external interrupt test mode. * INTn: External Interrupt n If TESTEN is 1, the value written to this bit will be the value to the interrupt detector and the value on the pad will be ignored. * NMI: Non-Maskable Interrupt If TESTEN is 1, the value written to this bit will be the value to the interrupt detector and the value on the pad will be ignored. 115 32059I-06/2010 AT32UC3B 13.7.11 Name: Access Type: Offset: Reset Value: Asynchronous Register ASYNC Read/Write 0x028 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n 0: The external interrupt is synchronized to CLK_SYNC. 1: The external interrupt is asynchronous. * NMI: Non-Maskable Interrupt 0: The Non-Maskable Interrupt is synchronized to CLK_SYNC 1: The Non-Maskable Interrupt is asynchronous. 116 32059I-06/2010 AT32UC3B 13.7.12 Name: Access Type: Offset: Reset Value: Scan Register SCAN Read/Write 0x2C 0x0000000 31 - 30 - 29 - 28 - 27 - 26 25 PIN[2:0] 24 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 11 10 PRESC[4:0] 9 8 7 - 6 - 5 - 4 - 3 - 2 - 1 - 0 EN * EN 0: Keypad scanning is disabled 1: Keypad scanning is enabled * PRESC Prescale select for the keypad scan rate: Scan rate = 2(SCAN.PRESC+1) TRC The RC clock period can be found in the Electrical Characteristics section. * PIN The index of the currently active scan pin. Writing to this bitfield has no effect. 117 32059I-06/2010 AT32UC3B 13.7.13 Name: Access Type: Offset: Reset Value: Enable Register EN Write-only 0x030 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n Writing a zero to this bit has no effect. Writing a one to this bit will enable the corresponding external interrupt. * NMI: Non-Maskable Interrupt Writing a zero to this bit has no effect. Writing a one to this bit will enable the Non-Maskable Interrupt. 118 32059I-06/2010 AT32UC3B 13.7.14 Name: Access Type: Offset: Reset Value: Disable Register DIS Write-only 0x034 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n Writing a zero to this bit has no effect. Writing a one to this bit will disable the corresponding external interrupt. * NMI: Non-Maskable Interrupt Writing a zero to this bit has no effect. Writing a one to this bit will disable the Non-Maskable Interrupt. 119 32059I-06/2010 AT32UC3B 13.7.15 Name: Access Type: Offset: Reset Value: Control Register CTRL Read-only 0x038 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 NMI 7 INT7 6 INT6 5 INT5 4 INT4 3 INT3 2 INT2 1 INT1 0 INT0 * INTn: External Interrupt n 0: The corresponding external interrupt is disabled. 1: The corresponding external interrupt is enabled. * NMI: Non-Maskable Interrupt 0: The Non-Maskable Interrupt is disabled. 1: The Non-Maskable Interrupt is enabled. 120 32059I-06/2010 AT32UC3B 14. Flash Controller (FLASHC) Rev: 2.1.2.4 14.1 Features * Controls flash block with dual read ports allowing staggered reads. * Supports 0 and 1 wait state bus access. * Allows interleaved burst reads for systems with one wait state, outputting one 32-bit word per clock cycle. * 32-bit HSB interface for reads from flash array and writes to page buffer. * 32-bit PB interface for issuing commands to and configuration of the controller. * 16 lock bits, each protecting a region consisting of (total number of pages in the flash block / 16) * * * * * * * pages. Regions can be individually protected or unprotected. Additional protection of the Boot Loader pages. Supports reads and writes of general-purpose NVM bits. Supports reads and writes of additional NVM pages. Supports device protection through a security bit. Dedicated command for chip-erase, first erasing all on-chip volatile memories before erasing flash and clearing security bit. Interface to Power Manager for power-down of flash-blocks in sleep mode. 14.2 Overview The flash controller (FLASHC) interfaces a flash block with the 32-bit internal High-Speed Bus (HSB). Performance for uncached systems with high clock-frequency and one wait state is increased by placing words with sequential addresses in alternating flash subblocks. Having one read interface per subblock allows them to be read in parallel. While data from one flash subblock is being output on the bus, the sequential address is being read from the other flash subblock and will be ready in the next clock cycle. The controller also manages the programming, erasing, locking and unlocking sequences with dedicated commands. 14.3 14.3.1 Product dependencies Power Manager The FLASHC has two bus clocks connected: One High speed bus clock (CLK_FLASHC_HSB) and one Peripheral bus clock (CLK_FLASHC_PB). These clocks are generated by the Power manager. Bot h clocks are turned on by default, but the user has t o ensure that CLK_FLASHC_HSB is not turned off before reading the flash or writing the pagebuffer and that CLK_FLASHC_PB is not turned off before accessing the FLASHC configuration and control registers. 14.3.2 Interrupt Controller The FLASHC interrupt lines are connected to internal sources of the interrupt controller. Using FLASHC interrupts requires the interrupt controller to be programmed first. 121 32059I-06/2010 AT32UC3B 14.4 14.4.1 Functional description Bus interfaces The FLASHC has two bus interfaces, one HSB interface for reads from the flash array and writes to the page buffer, and one Peripheral Bus (PB) interface for writing commands and control to and reading status from the controller. 14.4.2 Memory organization To maximize performance for high clock-frequency systems, FLASHC interfaces to a flash block with two read ports. The flash block has several parameters, given by the design of the flash block. Refer to the "Memories" chapter for the device-specific values of the parameters. * p pages (FLASH_P) * w words in each page and in the page buffer (FLASH_W) * pw words in total (FLASH_PW) * f general-purpose fuse bits (FLASH_F) * 1 security fuse bit * 1 User Page 14.4.3 User page The User page is an additional page, outside the regular flash array, that can be used to store various data, like calibration data and serial numbers. This page is not erased by regular chip erase. The User page can only be written and erased by proprietary commands. Read accesses to the User page is performed just as any other read access to the flash. The address map of the User page is given in Figure 14-1. 14.4.4 Read operations The FLASHC provides two different read modes: * 0 wait state (0ws) for clock frequencies < (access time of the flash plus the bus delay) * 1 wait state (1ws) for clock frequencies < (access time of the flash plus the bus delay)/2 Higher clock frequencies that would require more wait states are not supported by the flash controller. The programmer can select the wait states required by writing to the FWS field in the Flash Control Register (FCR). It is the responsibility of the programmer to select a number of wait states compatible with the clock frequency and timing characteristics of the flash block. In 0ws mode, only one of the two flash read ports is accessed. The other flash read port is idle. In 1ws mode, both flash read ports are active. One read port reading the addressed word, and the other reading the next sequential word. If the clock frequency allows, the user should use 0ws mode, because this gives the lowest power consumption for low-frequency systems as only one flash read port is read. Using 1ws mode has a power/performance ratio approaching 0ws mode as the clock frequency approaches twice the max frequency of 0ws mode. Using two flash read ports use twice the power, but also give twice the performance. 122 32059I-06/2010 AT32UC3B The flash controller supports flash blocks with up to 2^21 word addresses, as displayed in Figure 14-1. Reading the memory space between address pw and 2^21-1 returns an undefined result. The User page is permanently mapped to word address 2^21. Table 14-1. Memory type Main array User User page addresses Start address, byte sized 0 2^23 = 8388608 Size pw words = 4pw bytes 128 words = 512 bytes Figure 14-1. Memory map for the Flash memories A ll a d d r e s s e s a r e w o r d a d d r e s s e s 2^21+128 2^21 U nused U ser page pw p w -1 0 F la s h w it h e x tra p a g e 14.4.5 Quick Page Read A dedicated command, Quick Page Read (QPR), is provided to read all words in an addressed page. All bits in all words in this page are AND'ed together, returning a 1-bit result. This result is placed in the Quick Page Read Result (QPRR) bit in Flash Status Register (FSR). The QPR command is useful to check that a page is in an erased state. The QPR instruction is much faster than performing the erased-page check using a regular software subroutine. 14.4.6 Write page buffer operations The internal memory area reserved for the embedded flash can also be written through a writeonly page buffer. The page buffer is addressed only by the address bits required to address w words (since the page buffer is word addressable) and thus wrap around within the internal memory area address space and appear to be repeated within it. When writing to the page buffer, the PAGEN field in the Flash Command register (FCMD) is updated with the page number corresponding to page address of the latest word written into the page buffer. Flash data array Unused 123 32059I-06/2010 AT32UC3B The page buffer is also used for writes to the User page. Write operations can be prevented by programming the Memory Protection Unit of the CPU. Writing 8-bit and 16-bit data to the page buffer is not allowed and may lead to unpredictable data corruption. Page buffer write operations are performed with 4 wait states. Writing to the page buffer can only change page buffer bits from one to zero, i.e. writing 0xaaaaaaaa to a page buffer location that has the value 0x00000000, will not change the page buffer value. The only way to change a bit from zero to one, is to reset the entire page buffer with the Clear Page Buffer command. The page buffer is not automatically reset after a page write. The programmer should do this manually by issuing the Clear Page Buffer flash command. This can be done after a page write, or before the page buffer is loaded with data to be stored to the flash page. Example: Writing a word into word address 130 of a flash with 128 words in the page buffer. PAGEN will be updated with the value 1, and the word will be written into word 2 in the page buffer. 14.4.7 Writing words to a page that is not completely erased This can be used for EEPROM emulation, i.e. writes with granularity of one word instead of an entire page. Only words that are in an completely erased state (0xFFFFFFFF) can be changed. The procedure is as follows: 1. Clear page buffer 2. Write to the page buffer the result of the logical bitwise AND operation between the contents of the flash page and the new data to write. Only words that were in an erased state can be changed from the original page. 3. Write Page. 14.5 Flash commands The FLASHC offers a command set to manage programming of the flash memory, locking and unlocking of regions, and full flash erasing. See chapter 14.8.2 for a complete list of commands. To run a command, the field FCMD.CMD has to be written with the command number. As soon as FCMD is written, the FRDY bit is automatically cleared. Once the current command is complete, the FRDY bit is automatically set. If an interrupt has been enabled by setting the bit FRDY in FCR, the interrupt line of the flash controller is activated. All flash commands except for Quick Page Read (QPR) will generate an interrupt request upon completion if FRDY is set. After a command has been written to FCMD, the programming algorithm should wait until the command has been executed before attempting to read instructions or data from the flash or writing to the page buffer, as the flash will be busy. The waiting can be performed either by polling the Flash Status Register (FSR) or by waiting for the flash ready interrupt. The command written to FCMD is initiated on the first clock cycle where the HSB bus interface in FLASHC is IDLE. The user must make sure that the access pattern to the FLASHC HSB interface contains an IDLE cycle so that the command is allowed to start. Make sure that no bus masters such as DMA controllers are performing endless burst transfers from the flash. Also, make sure that the CPU does not perform endless burst transfers from flash. This is done by letting the CPU enter sleep mode after writing to FCMD, or by polling FSR for command completion. This polling will result in an access pattern with IDLE HSB cycles. 124 32059I-06/2010 AT32UC3B All the commands are protected by the same keyword, which has to be written in the eight highest bits of FCMD. Writing FCMD with data that does not contain the correct key and/or with an invalid command has no effect on the flash memory; however, the PROGE bit is set in FSR. This bit is automatically cleared by a read access to FSR. Writing a command to FCMD while another command is being executed has no effect on the flash memory; however, the PROGE bit is set in FSR. This bit is automatically cleared by a read access to FSR. If the current command writes or erases a page in a locked region, or a page protected by the BOOTPROT fuses, the command has no effect on the flash memory; however, the LOCKE bit is set in FSR . This bit is automatically cleared by a read access to FSR. 14.5.1 Write/erase page operation Flash technology requires that an erase must be done before programming. The entire flash can be erased by an Erase All command. Alternatively, pages can be individually erased by the Erase Page command. The User page can be written and erased using the mechanisms described in this chapter. After programming, the page can be locked to prevent miscellaneous write or erase sequences. Locking is performed on a per-region basis, so locking a region locks all pages inside the region. Additional protection is provided for the lowermost address space of the flash. This address space is allocated for the Boot Loader, and is protected both by the lock bit(s) corresponding to this address space, and the BOOTPROT[2:0] fuses. Data to be written are stored in an internal buffer called page buffer. The page buffer contains w words. The page buffer wraps around within the internal memory area address space and appears to be repeated by the number of pages in it. Writing of 8-bit and 16-bit data to the page buffer is not allowed and may lead to unpredictable data corruption. Data must be written to the page buffer before the programming command is written to FCMD. The sequence is as follows: * Reset the page buffer with the Clear Page Buffer command. * Fill the page buffer with the desired contents, using only 32-bit access. * Programming starts as soon as the programming key and the programming command are written to the Flash Command Register. The FCMD.PAGEN field must contain the address of the page to write. PAGEN is automatically updated when writing to the page buffer, but can also be written to directly. The FRDY bit in FSR is automatically cleared when the page write operation starts. * When programming is completed, the bit FRDY in FSR is set. If an interrupt was enabled by setting the bit FRDY in FCR, the interrupt line of the flash controller is set. Two errors can be detected in FSR after a programming sequence: * Programming Error: A bad keyword and/or an invalid command have been written in FCMD. * Lock Error: The page to be programmed belongs to a locked region. A command must be executed to unlock the corresponding region before programming can start. 14.5.2 Erase All operation The entire memory is erased if the Erase All command (EA) is written to FCMD. Erase All erases all bits in the flash array. The User page is not erased. All flash memory locations, the generalpurpose fuse bits, and the security bit are erased (reset to 0xFF) after an Erase All. 125 32059I-06/2010 AT32UC3B The EA command also ensures that all volatile memories, such as register file and RAMs, are erased before the security bit is erased. Erase All operation is allowed only if no regions are locked, and the BOOTPROT fuses are programmed with a region size of 0. Thus, if at least one region is locked, the bit LOCKE in FSR is set and the command is cancelled. If the bit LOCKE has been written to 1 in FCR, the interrupt line rises. When the command is complete, the bit FRDY bit in FSR is set. If an interrupt has been enabled by setting the bit FRDY in FCR, the interrupt line of the flash controller is set. Two errors can be detected in FSR after issuing the command: * Programming Error: A bad keyword and/or an invalid command have been written in FCMD. * Lock Error: At least one lock region to be erased is protected, or BOOTPROT is different from 0. The erase command has been refused and no page has been erased. A Clear Lock Bit command must be executed previously to unlock the corresponding lock regions. 14.5.3 Region lock bits The flash block has p pages, and these pages are grouped into 16 lock regions, each region containing p/16 pages. Each region has a dedicated lock bit preventing writing and erasing pages in the region. After production, the device may have some regions locked. These locked regions are reserved for a boot or default application. Locked regions can be unlocked to be erased and then programmed with another application or other data. To lock or unlock a region, the commands Lock Region Containing Page (LP) and Unlock Region Containing Page (UP) are provided. Writing one of these commands, together with the number of the page whose region should be locked/unlocked, performs the desired operation. One error can be detected in FSR after issuing the command: * Programming Error: A bad keyword and/or an invalid command have been written in FCMD. The lock bits are implemented using the lowest 16 general-purpose fuse bits. This means that lock bits can also be set/cleared using the commands for writing/erasing general-purpose fuse bits, see chapter 14.6. The general-purpose bit being in an erased (1) state means that the region is unlocked. The lowermost pages in the Flash can additionally be protected by the BOOTPROT fuses, see Section 14.6. 14.6 General-purpose fuse bits Each flash block has a number of general-purpose fuse bits that the application programmer can use freely. The fuse bits can be written and erased using dedicated commands, and read 126 32059I-06/2010 AT32UC3B through a dedicated Peripheral Bus address. Some of the general-purpose fuse bits are reserved for special purposes, and should not be used for other functions.: Table 14-2. GeneralPurpose fuse number 15:0 General-purpose fuses with special functions Name LOCK Usage Region lock bits. External Privileged Fetch Lock. Used to prevent the CPU from fetching instructions from external memories when in privileged mode. This bit can only be changed when the security bit is cleared. The address range corresponding to external memories is device-specific, and not known to the flash controller. This fuse bit is simply routed out of the CPU or bus system, the flash controller does not treat this fuse in any special way, except that it can not be altered when the security bit is set. If the security bit is set, only an external JTAG Chip Erase can clear EPFL. No internal commands can alter EPFL if the security bit is set. When the fuse is erased (i.e. "1"), the CPU can execute instructions fetched from external memories. When the fuse is programmed (i.e. "0"), instructions can not be executed from external memories. Used to select one of eight different boot loader sizes. Pages included in the bootlegger area can not be erased or programmed except by a JTAG chip erase. BOOTPROT can only be changed when the security bit is cleared. If the security bit is set, only an external JTAG Chip Erase can clear BOOTPROT, and thereby allow the pages protected by BOOTPROT to be programmed. No internal commands can alter BOOTPROT or the pages protected by BOOTPROT if the security bit is set. 16 EPFL 19:17 BOOTPROT The BOOTPROT fuses protects the following address space for the Boot Loader: Table 14-3. BOOTPROT 7 6 5 4 3 2 1 0 Boot Loader area specified by BOOTPROT Pages protected by BOOTPROT None 0-1 0-3 0-7 0-15 0-31 0-63 0-127 Size of protected memory 0 1kByte 2kByte 4kByte 8kByte 16kByte 32kByte 64kByte 127 32059I-06/2010 AT32UC3B To erase or write a general-purpose fuse bit, the commands Write General-Purpose Fuse Bit (WGPB) and Erase General-Purpose Fuse Bit (EGPB) are provided. Writing one of these commands, together with the number of the fuse to write/erase, performs the desired operation. An entire General-Purpose Fuse byte can be written at a time by using the Program GP Fuse Byte (PGPFB) instruction. A PGPFB to GP fuse byte 2 is not allowed if the flash is locked by the security bit. The PFB command is issued with a parameter in the PAGEN field: * PAGEN[2:0] - byte to write * PAGEN[10:3] - Fuse value to write All General-Purpose fuses can be erased by the Erase All General-Purpose fuses (EAGP) command. An EAGP command is not allowed if the flash is locked by the security bit. Two errors can be detected in FSR after issuing these commands: * Programming Error: A bad keyword and/or an invalid command have been written in FCMD. * Lock Error: A write or erase of any of the special-function fuse bits in Table 14-3 was attempted while the flash is locked by the security bit. The lock bits are implemented using the lowest 16 general-purpose fuse bits. This means that the 16 lowest general-purpose fuse bits can also be written/erased using the commands for locking/unlocking regions, see Section 14.5.3. 14.7 Security bit The security bit allows the entire chip to be locked from external JTAG or other debug access for code security. The security bit can be written by a dedicated command, Set Security Bit (SSB). Once set, the only way to clear the security bit is through the JTAG Chip Erase command. Once the Security bit is set, the following Flash controller commands will be unavailable and return a lock error if attempted: * Write General-Purpose Fuse Bit (WGPB) to BOOTPROT or EPFL fuses * Erase General-Purpose Fuse Bit (EGPB) to BOOTPROT or EPFL fuses * Program General-Purpose Fuse Byte (PGPFB) of fuse byte 2 * Erase All General-Purpose Fuses (EAGPF) One error can be detected in FSR after issuing the command: * Programming Error: A bad keyword and/or an invalid command have been written in FCMD. 128 32059I-06/2010 AT32UC3B 14.8 User Interface FLASHC Register Memory Map Register Flash Control Register Flash Command Register Flash Status Register Flash General Purpose Fuse Register Hi Flash General Purpose Fuse Register Lo Name FCR FCMD FSR FGPFRHI FGPFRLO Access R/W R/W R/W R R Reset 0x00000000 0x00000000 0x00000000 (*) NA (*) NA (*) Table 14-4. Offset 0x0 0x4 0x8 0xc 0x10 (*) The value of the Lock bits is dependent of their programmed state. All other bits in FSR are 0. All bits in FGPFR and FCFR are dependent on the programmed state of the fuses they map to. Any bits in these registers not mapped to a fuse read 0. 129 32059I-06/2010 AT32UC3B 14.8.1 Flash Control Register FCR Read/Write 0x00 0x00000000 30 29 28 27 26 25 24 - Name: Access Type: Offset: Reset value: 31 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 FWS 5 - 4 - 3 PROGE 2 LOCKE 1 - 0 FRDY * FRDY: Flash Ready Interrupt Enable 0: Flash Ready does not generate an interrupt. 1: Flash Ready generates an interrupt. * LOCKE: Lock Error Interrupt Enable 0: Lock Error does not generate an interrupt. 1: Lock Error generates an interrupt. * PROGE: Programming Error Interrupt Enable 0: Programming Error does not generate an interrupt. 1: Programming Error generates an interrupt. * FWS: Flash Wait State 0: The flash is read with 0 wait states. 1: The flash is read with 1 wait state. 130 32059I-06/2010 AT32UC3B 14.8.2 Flash Command Register FCMD Read/Write 0x04 0x00000000 FCMD can not be written if the flash is in the process of performing a flash command. Doing so will cause the FCR write to be ignored, and the PROGE bit to be set. 31 30 29 28 KEY 27 26 25 24 Name: Access Type: Offset: Reset value: 23 22 21 20 PAGEN [15:8] 19 18 17 16 15 14 13 12 PAGEN [7:0] 11 10 9 8 7 - 6 - 5 4 3 CMD 2 1 0 * CMD: Command This field defines the flash command. Issuing any unused command will cause the Programming Error bit to be set, and the corresponding interrupt to be requested if FCR.PROGE is set. Table 14-5. Command No operation Write Page Erase Page Set of commands Value 0 1 2 3 4 5 6 7 8 9 10 11 12 Mnemonic NOP WP EP CPB LP UP EA WGPB EGPB SSB PGPFB EAGPF QPR Clear Page Buffer Lock region containing given Page Unlock region containing given Page Erase All Write General-Purpose Fuse Bit Erase General-Purpose Fuse Bit Set Security Bit Program GP Fuse Byte Erase All GPFuses Quick Page Read 131 32059I-06/2010 AT32UC3B Table 14-5. Command Write User Page Erase User Page Quick Page Read User Page Set of commands Value 13 14 15 Mnemonic WUP EUP QPRUP * PAGEN: Page number The PAGEN field is used to address a page or fuse bit for certain operations. In order to simplify programming, the PAGEN field is automatically updated every time the page buffer is written to. For every page buffer write, the PAGEN field is updated with the page number of the address being written to. Hardware automatically masks writes to the PAGEN field so that only bits representing valid page numbers can be written, all other bits in PAGEN are always 0. As an example, in a flash with 1024 pages (page 0 - page 1023), bits 15:10 will always be 0. Table 14-6. Command No operation Write Page Semantic of PAGEN field in different commands PAGEN description Not used The number of the page to write Not used Page number whose region should be locked Page number whose region should be unlocked Not used GPFUSE # GPFUSE # Not used WriteData[7:0], ByteAddress[2:0] Not used Page number Not used Not used Not used Clear Page Buffer Lock region containing given Page Unlock region containing given Page Erase All Write General-Purpose Fuse Bit Erase General-Purpose Fuse Bit Set Security Bit Program GP Fuse Byte Erase All GP Fuses Quick Page Read Write User Page Erase User Page Quick Page Read User Page * KEY: Write protection key This field should be written with the value 0xA5 to enable the command defined by the bits of the register. If the field is written with a different value, the write is not performed and no action is started. This field always reads as 0. 132 32059I-06/2010 AT32UC3B 14.8.3 Flash Status Register FSR Read/Write 0x08 0x00000000 30 LOCK14 29 LOCK13 28 LOCK12 27 LOCK11 26 LOCK10 25 LOCK9 24 LOCK8 Name: Access Type: Offset: Reset value: 31 LOCK15 23 LOCK7 22 LOCK6 21 LOCK5 20 LOCK4 19 LOCK3 18 LOCK2 17 LOCK1 16 LOCK0 15 14 FSZ 13 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 QPRR 4 SECURITY 3 PROGE 2 LOCKE 1 - 0 FRDY * FRDY: Flash Ready Status 0: The flash controller is busy and the application must wait before running a new command. 1: The flash controller is ready to run a new command. * LOCKE: Lock Error Status Automatically cleared when FSR is read. 0: No programming of at least one locked lock region has happened since the last read of FSR. 1: Programming of at least one locked lock region has happened since the last read of FSR. * PROGE: Programming Error Status Automatically cleared when FSR is read. 0: No invalid commands and no bad keywords were written in FCMD. 1: An invalid command and/or a bad keyword was/were written in FCMD. * SECURITY: Security Bit Status 0: The security bit is inactive. 1: The security bit is active. * QPRR: Quick Page Read Result 0: The result is zero, i.e. the page is not erased. 1: The result is one, i.e. the page is erased. 133 32059I-06/2010 AT32UC3B * FSZ: Flash Size The size of the flash. Not all device families will provide all flash sizes indicated in the table. Table 14-7. FSZ 0 1 2 3 4 5 6 7 Flash size Flash Size 32 Kbytes 64 Kbytes 128 Kbytes 256 Kbytes 384 Kbytes 512 Kbytes 768 Kbytes 1024 Kbytes * LOCKx: Lock Region x Lock Status 0: The corresponding lock region is not locked. 1: The corresponding lock region is locked. 134 32059I-06/2010 AT32UC3B 14.8.4 Flash General Purpose Fuse Register High FGPFRHI Read 0x0C N/A 30 GPF62 29 GPF61 28 GPF60 27 GPF59 26 GPF58 25 GPF57 24 GPF56 Name: Access Type: Offset: Reset value: 31 GPF63 23 GPF55 22 GPF54 21 GPF53 20 GPF52 19 GPF51 18 GPF50 17 GPF49 16 GPF48 15 GPF47 14 GPF46 13 GPF45 12 GPF44 11 GPF43 10 GPF42 9 GPF41 8 GPF40 7 GPF39 6 GPF38 5 GPF37 4 GPF36 3 GPF35 2 GPF34 1 GPF33 0 GPF32 This register is only used in systems with more than 32 GP fuses. * GPFxx: General Purpose Fuse xx 0: The fuse has a written/programmed state. 1: The fuse has an erased state. 135 32059I-06/2010 AT32UC3B 14.8.5 Flash General Purpose Fuse Register Low FGPFRLO Read 0x10 N/A 30 GPF30 29 GPF29 28 GPF28 27 GPF27 26 GPF26 25 GPF25 24 GPF24 Name: Access Type: Offset: Reset value: 31 GPF31 23 GPF23 22 GPF22 21 GPF21 20 GPF20 19 GPF19 18 GPF18 17 GPF17 16 GPF16 15 GPF15 14 GPF14 13 GPF13 12 GPF12 11 GPF11 10 GPF10 9 GPF09 8 GPF08 7 GPF07 6 GPF06 5 GPF05 4 GPF04 3 GPF03 2 GPF02 1 GPF01 0 GPF00 * GPFxx: General Purpose Fuse xx 0: The fuse has a written/programmed state. 1: The fuse has an erased state. 136 32059I-06/2010 AT32UC3B 14.9 Fuses Settings The flash block contains a number of general purpose fuses. Some of these fuses have defined meanings outside the flash controller and are described in this section. The general purpose fuses are erase by a JTAG chip erase. 14.9.1 31 GPF31 Flash General Purpose Fuse Register Low (FGPFRLO) 30 GPF30 29 GPF29 28 BODEN 27 26 BODHYST 25 24 BODLEVEL[5:4] 23 22 21 20 19 18 BOOTPROT 17 16 EPFL BODLEVEL[3:0] 15 14 13 12 LOCK[15:8] 11 10 9 8 7 6 5 4 LOCK[7:0] 3 2 1 0 BODEN: Brown Out Detector Enable BODEN 0x0 0x1 0x2 0x3 Description BOD disabled BOD enabled, BOD reset enabled BOD enabled, BOD reset disabled BOD disabled BODHYST: Brown Out Detector Hysteresis 0: The Brown out detector hysteresis is disabled 1: The Brown out detector hysteresis is enabled BODLEVEL: Brown Out Detector Trigger Level This controls the voltage trigger level for the Brown out detector. Refer to Electrical Characteristics section. If the BODLEVEL is set higher than VDDCORE and enabled byt fuses, the part will be in constant reset. To recover from this situation, apply an external voltage on VDDCORE that is higher than the BOD level and disable the BOD. LOCK, EPFL, BOOTPROT These are Flash controller fuses and are described in the FLASHC section. As no external memories can be connected to AT32UC3B the EPFL bit has no effect. 137 32059I-06/2010 AT32UC3B 14.9.2 Default Fuse Value The devices are shipped with the FGPFRLO register value: 0xFC07FFFF: * GPF31 fuse set to 1b. This fuse is used by the pre-programmed USB bootloader. * GPF30 fuse set to 1b. This fuse is used by the pre-programmed USB bootloader. * GPF29 fuse set to 1b. This fuse is used by the pre-programmed USB bootloader. * BODEN fuses set to 11b. BOD is disabled. * BODHYST fuse set to 1b. The BOD hysteresis is enabled. * BODLEVEL fuses set to 000000b. This is the minimum voltage trigger level for BOD. * BOOTPROT fuses set to 011b. The bootloader protected size is 8 Ko. * EPFL fuse set to 1b. External privileged fetch is not locked. * LOCK fuses set to 1111111111111111b. No region locked. See also the AT32UC3B Bootloader user guide document. After the JTAG chip erase command, the FGPFRLO register value is 0xFFFFFFFF. 14.10 Module configuration Table 14-8. Flash Memory Parameters Flash Size (FLASH_PW) 512 Kbytes 512 Kbytes 256 Kbytes 256 Kbytes 128 Kbytes 128 Kbytes 64 Kbytes 64 Kbytes Number of pages (FLASH_P) 1024 1024 512 512 256 256 128 128 Page size (FLASH_W) 128 words 128 words 128 words 128 words 128 words 128 words 128 words 128 words General Purpose Fuse bits (FLASH_L) 32 fuses 32 fuses 32 fuses 32 fuses 32 fuses 32 fuses 32 fuses 32 fuses Part Number AT32UC3B0512 AT32UC3B1512 AT32UC3B0256 AT32UC3B1256 AT32UC3B0128 AT32UC3B1128 AT32UC3B064 AT32UC3B164 138 32059I-06/2010 AT32UC3B 15. HSB Bus Matrix (HMATRIX) Rev: 2.3.0.2 15.1 Features * * * * * * * User Interface on peripheral bus Configurable Number of Masters (Up to sixteen) Configurable Number of Slaves (Up to sixteen) One Decoder for Each Master Three Different Memory Mappings for Each Master (Internal and External boot, Remap) One Remap Function for Each Master Programmable Arbitration for Each Slave - Round-Robin - Fixed Priority Programmable Default Master for Each Slave - No Default Master - Last Accessed Default Master - Fixed Default Master One Cycle Latency for the First Access of a Burst Zero Cycle Latency for Default Master One Special Function Register for Each Slave (Not dedicated) * * * * 15.2 Overview The Bus Matrix implements a multi-layer bus structure, that enables parallel access paths between multiple High Speed Bus (HSB) masters and slaves in a system, thus increasing the overall bandwidth. The Bus Matrix interconnects up to 16 HSB Masters to up to 16 HSB Slaves. The normal latency to connect a master to a slave is one cycle except for the default master of the accessed slave which is connected directly (zero cycle latency). The Bus Matrix provides 16 Special Function Registers (SFR) that allow the Bus Matrix to support application specific features. 15.3 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 15.3.1 Clocks The clock for the HMATRIX bus interface (CLK_HMATRIX) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. It is recommended to disable the HMATRIX before disabling the clock, to avoid freezing the HMATRIX in an undefined state. 15.4 Functional Description 15.4.1 Special Bus Granting Mechanism The Bus Matrix provides some speculative bus granting techniques in order to anticipate access requests from some masters. This mechanism reduces latency at first access of a burst or single transfer. This bus granting mechanism sets a different default master for every slave. 139 32059I-06/2010 AT32UC3B At the end of the current access, if no other request is pending, the slave remains connected to its associated default master. A slave can be associated with three kinds of default masters: no default master, last access master and fixed default master. 15.4.1.1 No Default Master At the end of the current access, if no other request is pending, the slave is disconnected from all masters. No Default Master suits low-power mode. 15.4.1.2 Last Access Master At the end of the current access, if no other request is pending, the slave remains connected to the last master that performed an access request. 15.4.1.3 Fixed Default Master At the end of the current access, if no other request is pending, the slave connects to its fixed default master. Unlike last access master, the fixed master does not change unless the user modifies it by a software action (field FIXED_DEFMSTR of the related SCFG). To change from one kind of default master to another, the Bus Matrix user interface provides the Slave Configuration Registers, one for each slave, that set a default master for each slave. The Slave Configuration Register contains two fields: DEFMSTR_TYPE and FIXED_DEFMSTR. The 2-bit DEFMSTR_TYPE field selects the default master type (no default, last access master, fixed default master), whereas the 4-bit FIXED_DEFMSTR field selects a fixed default master provided that DEFMSTR_TYPE is set to fixed default master. Please refer to the Bus Matrix user interface description. 15.4.2 Arbitration The Bus Matrix provides an arbitration mechanism that reduces latency when conflict cases occur, i.e. when two or more masters try to access the same slave at the same time. One arbiter per HSB slave is provided, thus arbitrating each slave differently. The Bus Matrix provides the user with the possibility of choosing between 2 arbitration types for each slave: 1. Round-Robin Arbitration (default) 2. Fixed Priority Arbitration This choice is made via the field ARBT of the Slave Configuration Registers (SCFG). Each algorithm may be complemented by selecting a default master configuration for each slave. When a re-arbitration must be done, specific conditions apply. See Section 15.4.2.1 "Arbitration Rules" on page 140. 15.4.2.1 Arbitration Rules Each arbiter has the ability to arbitrate between two or more different master requests. In order to avoid burst breaking and also to provide the maximum throughput for slave interfaces, arbitration may only take place during the following cycles: 1. Idle Cycles: When a slave is not connected to any master or is connected to a master which is not currently accessing it. 2. Single Cycles: When a slave is currently doing a single access. 140 32059I-06/2010 AT32UC3B 3. End of Burst Cycles: When the current cycle is the last cycle of a burst transfer. For defined length burst, predicted end of burst matches the size of the transfer but is managed differently for undefined length burst. 4. Slot Cycle Limit: When the slot cycle counter has reached the limit value indicating that the current master access is too long and must be broken. * Undefined Length Burst Arbitration In order to avoid long slave handling during undefined length bursts (INCR), the Bus Matrix provides specific logic in order to re-arbitrate before the end of the INCR transfer. A predicted end of burst is used as a defined length burst transfer and can be selected from among the following five possibilities: 1. Infinite: No predicted end of burst is generated and therefore INCR burst transfer will never be broken. 2. One beat bursts: Predicted end of burst is generated at each single transfer inside the INCP transfer. 3. Four beat bursts: Predicted end of burst is generated at the end of each four beat boundary inside INCR transfer. 4. Eight beat bursts: Predicted end of burst is generated at the end of each eight beat boundary inside INCR transfer. 5. Sixteen beat bursts: Predicted end of burst is generated at the end of each sixteen beat boundary inside INCR transfer. This selection can be done through the field ULBT of the Master Configuration Registers (MCFG). * Slot Cycle Limit Arbitration The Bus Matrix contains specific logic to break long accesses, such as very long bursts on a very slow slave (e.g., an external low speed memory). At the beginning of the burst access, a counter is loaded with the value previously written in the SLOT_CYCLE field of the related Slave Configuration Register (SCFG) and decreased at each clock cycle. When the counter reaches zero, the arbiter has the ability to re-arbitrate at the end of the current byte, half word or word transfer. 15.4.2.2 Round-Robin Arbitration This algorithm allows the Bus Matrix arbiters to dispatch the requests from different masters to the same slave in a round-robin manner. If two or more master requests arise at the same time, the master with the lowest number is first serviced, then the others are serviced in a round-robin manner. There are three round-robin algorithms implemented: 1. Round-Robin arbitration without default master 2. Round-Robin arbitration with last default master 3. Round-Robin arbitration with fixed default master * Round-Robin Arbitration without Default Master This is the main algorithm used by Bus Matrix arbiters. It allows the Bus Matrix to dispatch requests from different masters to the same slave in a pure round-robin manner. At the end of 141 32059I-06/2010 AT32UC3B the current access, if no other request is pending, the slave is disconnected from all masters. This configuration incurs one latency cycle for the first access of a burst. Arbitration without default master can be used for masters that perform significant bursts. * Round-Robin Arbitration with Last Default Master This is a biased round-robin algorithm used by Bus Matrix arbiters. It allows the Bus Matrix to remove the one latency cycle for the last master that accessed the slave. In fact, at the end of the current transfer, if no other master request is pending, the slave remains connected to the last master that performed the access. Other non privileged masters still get one latency cycle if they want to access the same slave. This technique can be used for masters that mainly perform single accesses. * Round-Robin Arbitration with Fixed Default Master This is another biased round-robin algorithm. It allows the Bus Matrix arbiters to remove the one latency cycle for the fixed default master per slave. At the end of the current access, the slave remains connected to its fixed default master. Every request attempted by this fixed default master will not cause any latency whereas other non privileged masters will still get one latency cycle. This technique can be used for masters that mainly perform single accesses. 15.4.2.3 Fixed Priority Arbitration This algorithm allows the Bus Matrix arbiters to dispatch the requests from different masters to the same slave by using the fixed priority defined by the user. If two or more master requests are active at the same time, the master with the highest priority number is serviced first. If two or more master requests with the same priority are active at the same time, the master with the highest number is serviced first. For each slave, the priority of each master may be defined through the Priority Registers for Slaves (PRAS and PRBS). 15.4.3 Slave and Master assignation The index number assigned to Bus Matrix slaves and masters are described in Memories chapter. 142 32059I-06/2010 AT32UC3B 15.5 User Interface HMATRIX Register Memory Map Register Master Configuration Register 0 Master Configuration Register 1 Master Configuration Register 2 Master Configuration Register 3 Master Configuration Register 4 Master Configuration Register 5 Master Configuration Register 6 Master Configuration Register 7 Master Configuration Register 8 Master Configuration Register 9 Master Configuration Register 10 Master Configuration Register 11 Master Configuration Register 12 Master Configuration Register 13 Master Configuration Register 14 Master Configuration Register 15 Slave Configuration Register 0 Slave Configuration Register 1 Slave Configuration Register 2 Slave Configuration Register 3 Slave Configuration Register 4 Slave Configuration Register 5 Slave Configuration Register 6 Slave Configuration Register 7 Slave Configuration Register 8 Slave Configuration Register 9 Slave Configuration Register 10 Slave Configuration Register 11 Slave Configuration Register 12 Slave Configuration Register 13 Slave Configuration Register 14 Slave Configuration Register 15 Priority Register A for Slave 0 Priority Register B for Slave 0 Priority Register A for Slave 1 Name MCFG0 MCFG1 MCFG2 MCFG3 MCFG4 MCFG5 MCFG6 MCFG7 MCFG8 MCFG9 MCFG10 MCFG11 MCFG12 MCFG13 MCFG14 MCFG15 SCFG0 SCFG1 SCFG2 SCFG3 SCFG4 SCFG5 SCFG6 SCFG7 SCFG8 SCFG9 SCFG10 SCFG11 SCFG12 SCFG13 SCFG14 SCFG15 PRAS0 PRBS0 PRAS1 Access Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Reset Value 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000002 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000010 0x00000000 0x00000000 0x00000000 Table 15-1. Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 0x0028 0x002C 0x0030 0x0034 0x0038 0x003C 0x0040 0x0044 0x0048 0x004C 0x0050 0x0054 0x0058 0x005C 0x0060 0x0064 0x0068 0x006C 0x0070 0x0074 0x0078 0x007C 0x0080 0x0084 0x0088 143 32059I-06/2010 AT32UC3B Table 15-1. Offset 0x008C 0x0090 0x0094 0x0098 0x009C 0x00A0 0x00A4 0x00A8 0x00AC 0x00B0 0x00B4 0x00B8 0x00BC 0x00C0 0x00C4 0x00C8 0x00CC 0x00D0 0x00D4 0x00D8 0x00DC 0x00E0 0x00E4 0x00E8 0x00EC 0x00F0 0x00F4 0x00F8 0x00FC 0x0110 0x0114 0x0118 0x011C 0x0120 0x0124 0x0128 HMATRIX Register Memory Map (Continued) Register Priority Register B for Slave 1 Priority Register A for Slave 2 Priority Register B for Slave 2 Priority Register A for Slave 3 Priority Register B for Slave 3 Priority Register A for Slave 4 Priority Register B for Slave 4 Priority Register A for Slave 5 Priority Register B for Slave 5 Priority Register A for Slave 6 Priority Register B for Slave 6 Priority Register A for Slave 7 Priority Register B for Slave 7 Priority Register A for Slave 8 Priority Register B for Slave 8 Priority Register A for Slave 9 Priority Register B for Slave 9 Priority Register A for Slave 10 Priority Register B for Slave 10 Priority Register A for Slave 11 Priority Register B for Slave 11 Priority Register A for Slave 12 Priority Register B for Slave 12 Priority Register A for Slave 13 Priority Register B for Slave 13 Priority Register A for Slave 14 Priority Register B for Slave 14 Priority Register A for Slave 15 Priority Register B for Slave 15 Special Function Register 0 Special Function Register 1 Special Function Register 2 Special Function Register 3 Special Function Register 4 Special Function Register 5 Special Function Register 6 Name PRBS1 PRAS2 PRBS2 PRAS3 PRBS3 PRAS4 PRBS4 PRAS5 PRBS5 PRAS6 PRBS6 PRAS7 PRBS7 PRAS8 PRBS8 PRAS9 PRBS9 PRAS10 PRBS10 PRAS11 PRBS11 PRAS12 PRBS12 PRAS13 PRBS13 PRAS14 PRBS14 PRAS15 PRBS15 SFR0 SFR1 SFR2 SFR3 SFR4 SFR5 SFR6 Access Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Reset Value 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 - - - - - - - 144 32059I-06/2010 AT32UC3B Table 15-1. Offset 0x012C 0x0130 0x0134 0x0138 0x013C 0x0140 0x0144 0x0148 0x014C HMATRIX Register Memory Map (Continued) Register Special Function Register 7 Special Function Register 8 Special Function Register 9 Special Function Register 10 Special Function Register 11 Special Function Register 12 Special Function Register 13 Special Function Register 14 Special Function Register 15 Name SFR7 SFR8 SFR9 SFR10 SFR11 SFR12 SFR13 SFR14 SFR15 Access Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Reset Value - - - - - - - - - 145 32059I-06/2010 AT32UC3B 15.5.1 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - Master Configuration Registers MCFG0...MCFG15 Read/Write 0x00 - 0x3C 0x00000002 30 - 22 - 14 - 6 - 29 - 21 - 13 - 5 - 28 - 20 - 12 - 4 - 27 - 19 - 11 - 3 - 26 - 18 - 10 - 2 25 - 17 - 9 - 1 ULBT 24 - 16 - 8 - 0 * ULBT: Undefined Length Burst Type 0: Infinite Length Burst No predicted end of burst is generated and therefore INCR bursts coming from this master cannot be broken. 1: Single Access The undefined length burst is treated as a succession of single accesses, allowing re-arbitration at each beat of the INCR burst. 2: Four Beat Burst The undefined length burst is split into a four-beat burst, allowing re-arbitration at each four-beat burst end. 3: Eight Beat Burst The undefined length burst is split into an eight-beat burst, allowing re-arbitration at each eight-beat burst end. 4: Sixteen Beat Burst The undefined length burst is split into a sixteen-beat burst, allowing re-arbitration at each sixteen-beat burst end. 146 32059I-06/2010 AT32UC3B 15.5.2 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 Slave Configuration Registers SCFG0...SCFG15 Read/Write 0x40 - 0x7C 0x00000010 30 - 22 - 14 - 6 13 - 5 29 - 21 28 - 20 27 - 19 26 - 18 25 - 17 24 ARBT 16 FIXED_DEFMSTR 12 - 4 SLOT_CYCLE 11 - 3 10 - 2 DEFMSTR_TYPE 9 - 1 8 - 0 * ARBT: Arbitration Type 0: Round-Robin Arbitration 1: Fixed Priority Arbitration * FIXED_DEFMSTR: Fixed Default Master This is the number of the Default Master for this slave. Only used if DEFMSTR_TYPE is 2. Specifying the number of a master which is not connected to the selected slave is equivalent to setting DEFMSTR_TYPE to 0. The size of this field depends on the number of masters. This size is log2(number of masters). * DEFMSTR_TYPE: Default Master Type 0: No Default Master At the end of the current slave access, if no other master request is pending, the slave is disconnected from all masters. This results in a one cycle latency for the first access of a burst transfer or for a single access. 1: Last Default Master At the end of the current slave access, if no other master request is pending, the slave stays connected to the last master having accessed it. This results in not having one cycle latency when the last master tries to access the slave again. 2: Fixed Default Master At the end of the current slave access, if no other master request is pending, the slave connects to the fixed master the number that has been written in the FIXED_DEFMSTR field. This results in not having one cycle latency when the fixed master tries to access the slave again. * SLOT_CYCLE: Maximum Number of Allowed Cycles for a Burst When the SLOT_CYCLE limit is reached for a burst, it may be broken by another master trying to access this slave. This limit has been placed to avoid locking a very slow slave when very long bursts are used. This limit must not be very small. Unreasonably small values break every burst and the Bus Matrix arbitrates without performing any data transfer. 16 cycles is a reasonable value for SLOT_CYCLE. 147 32059I-06/2010 AT32UC3B 15.5.3 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - Bus Matrix Priority Registers A For Slaves PRAS0...PRAS15 Read/Write 0x00000000 30 - 22 - 14 - 6 - 5 M1PR 13 M3PR 4 21 M5PR 12 29 M7PR 20 28 27 - 19 - 11 - 3 - 26 - 18 - 10 - 2 - 1 M0PR 9 M2PR 0 17 M4PR 8 25 M6PR 16 24 * MxPR: Master x Priority Fixed priority of Master x for accessing the selected slave. The higher the number, the higher the priority. 148 32059I-06/2010 AT32UC3B 15.5.4 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - Priority Registers B For Slaves PRBS0...PRBS15 Read/Write 0x00000000 30 - 22 - 14 - 6 - 5 M9PR 13 M11PR 4 21 M13PR 12 29 M15PR 20 28 27 - 19 - 11 - 3 - 26 - 18 - 10 - 2 - 1 M8PR 9 M10PR 0 17 M12PR 8 25 M14PR 16 24 * MxPR: Master x Priority Fixed priority of Master x for accessing the selected slave. The higher the number, the higher the priority. 149 32059I-06/2010 AT32UC3B 15.5.5 Name: Access Type: Offset: Reset Value: 31 Special Function Registers SFR0...SFR15 Read/Write 0x110 - 0x115 30 29 28 SFR 23 22 21 20 SFR 15 14 13 12 SFR 7 6 5 4 SFR 3 2 1 0 11 10 9 8 19 18 17 16 27 26 25 24 * SFR: Special Function Register Fields Those registers are not a HMATRIX specific register. The field of those will be defined where they are used. 150 32059I-06/2010 AT32UC3B 15.6 Bus Matrix Connections Accesses to unused areas returns an error result to the master requesting such an access. The bus matrix has the several masters and slaves. Each master has its own bus and its own decoder, thus allowing a different memory mapping per master. The master number in the table below can be used to index the HMATRIX control registers. For example, HMATRIX MCFG0 register is associated with the CPU Data master interface. Table 15-2. Master 0 Master 1 Master 2 Master 3 Master 4 High Speed Bus masters CPU Data CPU Instruction CPU SAB PDCA USBB DMA Each slave has its own arbiter, thus allowing a different arbitration per slave. The slave number in the table below can be used to index the HMATRIX control registers. For example, SCFG3 is associated with the Internal SRAM Slave Interface. Table 15-3. Slave 0 Slave 1 Slave 2 Slave 3 Slave 4 High Speed Bus slaves Internal Flash HSB-PB Bridge A HSB-PB Bridge B Internal SRAM USBB DPRAM 151 32059I-06/2010 AT32UC3B Figure 15-1. HMatrix Master / Slave Connections HMATRIX SLAVES USBB DPRAM 4 Internal SRAM 3 Internal Flash HSB-PB Bridge A 1 0 CPU Data 0 HMATRIX MASTERS CPU Instruction 1 CPU SAB 2 PDCA 3 USBB DMA 4 HSB-PB Bridge B 2 152 32059I-06/2010 AT32UC3B 16. Peripheral DMA Controller (PDCA) Rev: 1.0.2.1 16.1 Features * Multiple channels * Generates transfers to/from peripherals such as USART and SPI * Two address pointers/counters per channel allowing double buffering 16.2 Overview The Peripheral DMA Controller (PDCA) transfers data between on-chip peripheral modules such as USART, SPI and memories (those memories may be on- and off-chip memories). Using the PDCA avoids CPU intervention for data transfers, improving the performance of the microcontroller. The PDCA can transfer data from memory to a peripheral or from a peripheral to memory. The PDCA consists of multiple DMA channels. Each channel has: * A Peripheral Select Register * A 32-bit memory pointer * A 16-bit transfer counter * A 32-bit memory pointer reload value * A 16-bit transfer counter reload value The PDCA communicates with the peripheral modules over a set of handshake interfaces. The peripheral signals the PDCA when it is ready to receive or transmit data. The PDCA acknowledges the request when the transmission has started. When a transmit buffer is empty or a receive buffer is full, an optional interrupt request can be generated. 153 32059I-06/2010 AT32UC3B 16.3 Block Diagram Figure 16-1. PDCA Block Diagram Peripheral 0 HSB to PB Bridge HSB Peripheral Bus Peripheral 1 High Speed Bus Matrix HSB Peripheral 2 Peripheral DMA Controller (PDCA) Interrupt Controller IRQ Peripheral (n-1) Handshake Interfaces ... 16.4 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 16.4.1 Power Management If the CPU enters a sleep mode that disables the PDCA clocks, the PDCA will stop functioning and resume operation after the system wakes up from sleep mode. 16.4.2 Clocks The PDCA has two bus clocks connected: One High Speed Bus clock (CLK_PDCA_HSB) and one Peripheral Bus clock (CLK_PDCA_PB). These clocks are generated by the Power Manager. Both clocks are enabled at reset, and can be disabled by writing to the Power Manager. It is recommended to disable the PDCA before disabling the clocks, to avoid freezing the PDCA in an undefined state. 16.4.3 Interrupts The PDCA interrupt request lines are connected to the interrupt controller. Using the PDCA interrupts requires the interrupt controller to be programmed first. 154 32059I-06/2010 AT32UC3B 16.5 16.5.1 Functional Description Basic Operation The PDCA consists of multiple independent PDCA channels, each capable of handling DMA requests in parallel. Each PDCA channels contains a set of configuration registers which must be configured to start a DMA transfer. In this section the steps necessary to configure one PDCA channel is outlined. The peripheral to transfer data to or from must be configured correctly in the Peripheral Select Register (PSR). This is performed by writing the Peripheral Identity (PID) value for the corresponding peripheral to the PID field in the PSR register. The PID also encodes the transfer direction, i.e. memory to peripheral or peripheral to memory. See Section 16.5.5. The transfer size must be written to the Transfer Size field in the Mode Register (MR.SIZE). The size must match the data size produced or consumed by the selected peripheral. See Section 16.5.6. The memory address to transfer to or from, depending on the PSR, must be written to the Memory Address Register (MAR). For each transfer the memory address is increased by either a one, two or four, depending on the size set in MR. See Section 16.5.2. The number of data items to transfer is written to the TCR register. If the PDCA channel is enabled, a transfer will start immediately after writing a non-zero value to TCR or the reload version of TCR, TCRR. After each transfer the TCR value is decreased by one. Both MAR and TCR can be read while the PDCA channel is active to monitor the DMA progress. See Section 16.5.3. The channel must be enabled for a transfer to start. A channel is enable by writing a one to the EN bit in the Control Register (CR). 16.5.2 Memory Pointer Each channel has a 32-bit Memory Address Register (MAR). This register holds the memory address for the next transfer to be performed. The register is automatically updated after each transfer. The address will be increased by either one, two or four depending on the size of the DMA transfer (byte, halfword or word). The MAR can be read at any time during transfer. 16.5.3 Transfer Counter Each channel has a 16-bit Transfer Counter Register (TCR). This register must be programmed with the number of transfers to be performed. The TCR register should contain the number of data items to be transferred independently of the transfer size. The TCR can be read at any time during transfer to see the number of remaining transfers. 16.5.4 Reload Registers Both the MAR and the TCR have a reload register, respectively Memory Address Reload Register (MARR) and Transfer Counter Reload Register (TCRR). These registers provide the possibility for the PDCA to work on two memory buffers for each channel. When one buffer has completed, MAR and TCR will be reloaded with the values in MARR and TCRR. The reload logic is always enabled and will trigger if the TCR reaches zero while TCRR holds a non-zero value. After reload, the MARR and TCRR registers are cleared. If TCR is zero when writing to TCRR, the TCR and MAR are automatically updated with the value written in TCRR and MARR. 155 32059I-06/2010 AT32UC3B 16.5.5 Peripheral Selection The Peripheral Select Register (PSR) decides which peripheral should be connected to the PDCA channel. A peripheral is selected by writing the corresponding Peripheral Identity (PID) to the PID field in the PST register. Writing the PID will both select the direction of the transfer (memory to peripheral or peripheral to memory), which handshake interface to use, and the address of the peripheral holding register. Refer to the Peripheral Identity (PID) table in the Module Configuration section for the peripheral PID values. 16.5.6 Transfer Size The transfer size can be set individually for each channel to be either byte, halfword or word (8bit, 16-bit or 32-bit respectively). Transfer size is set by writing the desired value to the Transfer Size field in the Mode Register (MR.SIZE). When the PDCA moves data between peripherals and memory, data is automatically sized and aligned. When memory is accessed, the size specified in MR.SIZE and system alignment is used. When a peripheral register is accessed the data to be transferred is converted to a word where bit n in the data corresponds to bit n in the peripheral register. If the transfer size is byte or halfword, bits greater than 8 and16 respectively are set to zero. Refer to the Module Configuration section for information regarding what peripheral registers are used for the differen peripherals and then to the peripheral specific chapter for information about the size option available for the different registers. 16.5.7 Enabling and Disabling Each DMA channel is enabled by writing a one to the Transfer Enable bit in the Control Register (CR.TEN) and disabled by writing a one to the Transfer Disable bit (CR.TDIS). The current status can be read from the Status Register (SR). While the PDCA channel is enabled all DMA request will be handled as long the TCR and TCRR is not zero. 16.5.8 Interrupts Interrupts can be enabled by writing a one to the corresponding bit in the Interrupt Enable Register (IER) and disabled by writing a one to the corresponding bit in the Interrupt Disable Register (IDR). The Interrupt Mask Register (IMR) can be read to see whether an interrupt is enabled or not. The current status of an interrupt source can be read through the Interrupt Status Register (ISR). The PDCA has three interrupt sources: * Reload Counter Zero - The TCRR register is zero. * Transfer Finished - Both the TCR and TCRR registers are zero. * Transfer Error - An error has occurred in accessing memory. 16.5.9 Priority If more than one PDCA channel is requesting transfer at a given time, the PDCA channels are prioritized by their channel number. Channels with lower numbers have priority over channels with higher numbers, giving channel zero the highest priority. 156 32059I-06/2010 AT32UC3B 16.5.10 Error Handling If the Memory Address Register (MAR) is set to point to an invalid location in memory, an error will occur when the PDCA tries to perform a transfer. When an error occurs, the Transfer Error bit in the Interrupt Status Register (ISR.TERR) will be set and the DMA channel that caused the error will be stopped. In order to restart the channel, the user must program the Memory Address Register to a valid address and then write a one to the Error Clear bit in the Control Register (CR.ECLR). If the Transfer Error interrupt is enabled, an interrupt request will be generated when an transfer error occurs. 157 32059I-06/2010 AT32UC3B 16.6 16.6.1 User Interface Memory Map Overview PDCA Register Memory Map Contents DMA channel 0 configuration registers DMA channel 1 configuration registers ... DMA channel m configuration registers Table 16-1. Address Range 0x000 - 0x03F 0x040 - 0x07F ... (0x000 - 0x03F)+m*0x040 The channels are mapped as shown in Table 16-1. Each channel has a set of configuration registers, shown in Table 16-2, where n is the channel number. 16.6.2 Channel Memory Map PDCA Channel Configuration Registers Register Memory Address Register Peripheral Select Register Transfer Counter Register Memory Address Reload Register Transfer Counter Reload Register Control Register Mode Register Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Interrupt Status Register Note: Register Name MAR PSR TCR MARR TCRR CR MR SR IER IDR IMR ISR Access Read/Write Read/Write Read/Write Read/Write Read/Write Write-only Read/Write Read-only Write-only Write-only Read-only Read-only Reset 0x00000000 - (1) 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 Table 16-2. Offset 0x000 + n*0x040 0x004 + n*0x040 0x008 + n*0x040 0x00C + n*0x040 0x010 + n*0x040 0x014 + n*0x040 0x018 + n*0x040 0x01C + n*0x040 0x020 + n*0x040 0x024 + n*0x040 0x028 + n*0x040 0x02C + n*0x040 1. The reset values are device specific. Please refer to the Module Configuration section at the end of this chapter. 16.6.3 Note: 1. The reset values are device specific. Please refer to the Module Configuration section at the end of this chapter. 158 32059I-06/2010 AT32UC3B 16.6.4 Name: Access Type: Offset: Reset Value: Memory Address Register MAR Read/Write 0x000 + n*0x040 0x00000000 31 30 29 28 27 26 25 24 MADDR[31:24] 23 22 21 20 19 18 17 16 MADDR[23:16] 15 14 13 12 MADDR[15:8] 11 10 9 8 7 6 5 4 MADDR[7:0] 3 2 1 0 * MADDR: Memory Address Address of memory buffer. MADDR should be programmed to point to the start of the memory buffer when configuring the PDCA. During transfer, MADDR will point to the next memory location to be read/written. 159 32059I-06/2010 AT32UC3B 16.6.5 Name: Access Type: Offset: Reset Value: Peripheral Select Register PSR Read/Write 0x004 + n*0x040 - 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 6 5 4 PID 3 2 1 0 * PID: Peripheral Identifier The Peripheral Identifier selects which peripheral should be connected to the DMA channel. Writing a PID will select both which handshake interface to use, the direction of the transfer and also the address of the Receive/Transfer Holding Register for the peripheral. See the Module Configuration section of PDCA for details. The width of the PID field is device specific and dependent on the number of peripheral modules in the device. 160 32059I-06/2010 AT32UC3B 16.6.6 Name: Access Type: Offset: Reset Value: Transfer Counter Register TCR Read/Write 0x008 + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 TCV[15:8] 11 10 9 8 7 6 5 4 TCV[7:0] 3 2 1 0 * TCV: Transfer Counter Value Number of data items to be transferred by the PDCA. TCV must be programmed with the total number of transfers to be made. During transfer, TCV contains the number of remaining transfers to be done. 161 32059I-06/2010 AT32UC3B 16.6.7 Name: Access Type: Offset: Reset Value: Memory Address Reload Register MARR Read/Write 0x00C + n*0x040 0x00000000 31 30 29 28 MARV[31:24] 27 26 25 24 23 22 21 20 MARV[23:16] 19 18 17 16 15 14 13 12 MARV[15:8] 11 10 9 8 7 6 5 4 MARV[7:0] 3 2 1 0 * MARV: Memory Address Reload Value Reload Value for the MAR register. This value will be loaded into MAR when TCR reaches zero if the TCRR register has a nonzero value. 162 32059I-06/2010 AT32UC3B 16.6.8 Name: Access Type: Offset: Reset Value: Transfer Counter Reload Register TCRR Read/Write 0x010 + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 TCRV[15:8] 11 10 9 8 7 6 5 4 TCRV[7:0] 3 2 1 0 * TCRV: Transfer Counter Reload Value Reload value for the TCR register. When TCR reaches zero, it will be reloaded with TCRV if TCRV has a positive value. If TCRV is zero, no more transfers will be performed for the channel. When TCR is reloaded, the TCRR register is cleared. 163 32059I-06/2010 AT32UC3B 16.6.9 Name: Access Type: Offset: Reset Value: Control Register CR Write-only 0x014 + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 ECLR 7 - 6 - 5 - 4 - 3 - 2 - 1 TDIS 0 TEN * ECLR: Transfer Error Clear Writing a zero to this bit has no effect. Writing a one to this bit will clear the Transfer Error bit in the Status Register (SR.TERR). Clearing the SR.TERR bit will allow the channel to transmit data. The memory address must first be set to point to a valid location. * TDIS: Transfer Disable Writing a zero to this bit has no effect. Writing a one to this bit will disable transfer for the DMA channel. * TEN: Transfer Enable Writing a zero to this bit has no effect. Writing a one to this bit will enable transfer for the DMA channel. 164 32059I-06/2010 AT32UC3B 16.6.10 Name: Access Type: Offset: Reset Value: Mode Register MR Read/Write 0x018 + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 - 1 SIZE 0 * SIZE: Size of Transfer Table 16-3. SIZE 0 1 2 3 Size of Transfer Size of Transfer Byte Halfword Word Reserved 165 32059I-06/2010 AT32UC3B 16.6.11 Name: Access Type: Offset: Reset Value: Status Register SR Read-only 0x01C + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 - 1 - 0 TEN * TEN: Transfer Enabled This bit is cleared when the TDIS bit in CR is written to one. This bit is set when the TEN bit in CR is written to one. 0: Transfer is disabled for the DMA channel. 1: Transfer is enabled for the DMA channel. 166 32059I-06/2010 AT32UC3B 16.6.12 Name: Access Type: Offset: Reset Value: Interrupt Enable Register IER Write-only 0x020 + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 TERR 1 TRC 0 RCZ Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. 167 32059I-06/2010 AT32UC3B 16.6.13 Name: Access Type: Offset: Reset Value: Interrupt Disable Register IDR Write-only 0x024 + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 TERR 1 TRC 0 RCZ Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. 168 32059I-06/2010 AT32UC3B 16.6.14 Name: Access Type: Offset: Reset Value: Interrupt Mask Register IMR Read-only 0x028 + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 TERR 1 TRC 0 RCZ 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is set when the corresponding bit in IER is written to one. 169 32059I-06/2010 AT32UC3B 16.6.15 Name: Access Type: Offset: Reset Value: Interrupt Status Register ISR Read-only 0x02C + n*0x040 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 TERR 1 TRC 0 RCZ * TERR: Transfer Error This bit is cleared when no transfer errors have occurred since the last write to CR.ECLR. This bit is set when one or more transfer errors has occurred since reset or the last write to CR.ECLR. * TRC: Transfer Complete This bit is cleared when the TCR and/or the TCRR holds a non-zero value. This bit is set when both the TCR and the TCRR are zero. * RCZ: Reload Counter Zero This bit is cleared when the TCRR holds a non-zero value. This bit is set when TCRR is zero. * 170 32059I-06/2010 AT32UC3B 16.7 Module Configuration The specific configuration for the PDCA instance is listed in the following tables. Table 16-4. Features Number of channels PDCA Configuration PDCA 7 Table 16-5. Register PSRn Register Reset Values Reset Value n 16.7.1 DMA Handshake Signals The following table defines the valid settings for the Peripheral Identifier (PID) in the PDCA Peripheral Select Register (PSR).). Table 16-6. PID Value 0 1 2 3 4 5 6 7 8 9 10 11 12 13 PDCA Handshake Signals Peripheral module & direction ADC SSC - RX USART0 - RX USART1 - RX USART2 - RX TWI - RX SPI0 - RX SSC - TX USART0 - TX USART1 - TX USART2 - TX TWI - TX SPI0 - TX ABDAC - TX 171 32059I-06/2010 AT32UC3B 17. General-Purpose Input/Output Controller (GPIO) Rev: 1.1.0.4 17.1 Features * * * * * Each I/O line of the GPIO features: Configurable pin-change, rising-edge or falling-edge interrupt on any I/O line A glitch filter providing rejection of pulses shorter than one clock cycle Input visibility and output control Multiplexing of up to four peripheral functions per I/O line Programmable internal pull-up resistor 17.2 Overview The General Purpose Input/Output Controller manages the I/O pins of the microcontroller. Each I/O line may be dedicated as a general-purpose I/O or be assigned to a function of an embedded peripheral. This assures effective optimization of the pins of a product. 17.3 Block Diagram Figure 17-1. GPIO Block Diagram PB Configuration Interface Interrupt Controller GPIO Interrupt Request PIN General Purpose Input/Output - GPIO Power Manager CLK_GPIO PIN PIN PIN PIN MCU I/O Pins Embedded Peripheral Pin Control Signals 17.4 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 172 32059I-06/2010 AT32UC3B 17.4.1 Module Configuration Most of the features of the GPIO are configurable for each product. The user must refer to the Package and Pinout chapter for these settings. Product specific settings includes: * Number of I/O pins. * Functions implemented on each pin * Peripheral function(s) multiplexed on each I/O pin * Reset value of registers 17.4.2 Clocks The clock for the GPIO bus interface (CLK_GPIO) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. The CLK_GPIO must be enabled in order to access the configuration registers of the GPIO or to use the GPIO interrupts. After configuring the GPIO, the CLK_GPIO can be disabled if interrupts are not used. 17.4.3 Interrupts The GPIO interrupt lines are connected to the interrupt controller. Using the GPIO interrupt requires the interrupt controller to be configured first. 17.5 Functional Description The GPIO controls the I/O lines of the microcontroller. The control logic associated with each pin is represented in the figure below: 173 32059I-06/2010 AT32UC3B Figure 17-2. Overview of the GPIO Pad Connections ODER PUER 1 Periph. A output enable Periph. B output enable 0 Periph. C output enable Periph. D output enable PMR1 GPER PMR0 Periph. A output data Periph. B output data Periph. C output data Periph. D output data 1 0 PAD OVR Periph. A input data Periph. B input data Periph. C input data Periph. D input data IER 0 Edge Detector Glitch Filter 1 0 Interrupt Request 1 PVR IMR1 GFER IMR0 17.5.1 17.5.1.1 Basic Operation I/O Line or peripheral function selection When a pin is multiplexed with one or more peripheral functions, the selection is controlled with the GPIO Enable Register (GPER). If a bit in GPER is written to one, the corresponding pin is controlled by the GPIO. If a bit is written to zero, the corresponding pin is controlled by a peripheral function. 17.5.1.2 Peripheral selection The GPIO provides multiplexing of up to four peripheral functions on a single pin. The selection is performed by accessing Peripheral Mux Register 0 (PMR0) and Peripheral Mux Register 1 (PMR1). 17.5.1.3 Output control When the I/O line is assigned to a peripheral function, i.e. the corresponding bit in GPER is written to zero, the drive of the I/O line is controlled by the peripheral. The peripheral, depending on the value in PMR0 and PMR1, determines whether the pin is driven or not. When the I/O line is controlled by the GPIO, the value of the Output Driver Enable Register (ODER) determines if the pin is driven or not. When a bit in this register is written to one, the cor- 174 32059I-06/2010 AT32UC3B responding I/O line is driven by the GPIO. When the bit is written to zero, the GPIO does not drive the line. The level driven on an I/O line can be determined by writing to the Output Value Register (OVR). 17.5.1.4 Inputs The level on each I/O line can be read through the Pin Value Register (PVR). This register indicates the level of the I/O lines regardless of whether the lines are driven by the GPIO or by an external component. Note that due to power saving measures, the PVR register can only be read when GPER is written to one for the corresponding pin or if interrupt is enabled for the pin. 17.5.1.5 Output line timings The figure below shows the timing of the I/O line when writing a one and a zero to OVR. The same timing applies when performing a `set' or `clear' access, i.e., writing a one to the Output Value Set Register (OVRS) or the Output Value Clear Register (OVRC). The timing of PVR is also shown. Figure 17-3. Output Line Timings CLK_GPIO Write OVR to 1 Write OVR to 0 OVR / I/O Line PVR PB Access PB Access 17.5.2 17.5.2.1 Advanced Operation Pull-up resistor control Each I/O line is designed with an embedded pull-up resistor. The pull-up resistor can be enabled or disabled by writing a one or a zero to the corresponding bit in the Pull-up Enable Register (PUER). Control of the pull-up resistor is possible whether an I/O line is controlled by a peripheral or the GPIO. 17.5.2.2 Input glitch filter Optional input glitch filters can be enabled on each I/O line. When the glitch filter is enabled, a glitch with duration of less than 1 clock cycle is automatically rejected, while a pulse with duration of 2 clock cycles or more is accepted. For pulse durations between 1 clock cycle and 2 clock cycles, the pulse may or may not be taken into account, depending on the precise timing of its occurrence. Thus for a pulse to be guaranteed visible it must exceed 2 clock cycles, whereas for a glitch to be reliably filtered out, its duration must not exceed 1 clock cycle. The filter introduces 2 clock cycles of latency. The glitch filters are controlled by the Glitch Filter Enable Register (GFER). When a bit is written to one in GFER, the glitch filter on the corresponding pin is enabled. The glitch filter affects only interrupt inputs. Inputs to peripherals or the value read through PVR are not affected by the glitch filters. 175 32059I-06/2010 AT32UC3B 17.5.3 Interrupts The GPIO can be configured to generate an interrupt when it detects an input change on an I/O line. The module can be configured to signal an interrupt whenever a pin changes value or only to trigger on rising edges or falling edges. Interrupts are enabled on a pin by writing a one to the corresponding bit in the Interrupt Enable Register (IER). The interrupt mode is set by writing to the Interrupt Mode Register 0 (IMR0) and the Interrupt Mode Register 1(IMR1). Interrupts can be enabled on a pin, regardless of the configuration of the I/O line, i.e. whether it is controlled by the GPIO or assigned to a peripheral function. In every port there are four interrupt lines connected to the interrupt controller. Groups of eight interrupts in the port are ORed together to form an interrupt line. When an interrupt event is detected on an I/O line, and the corresponding bit in IER is written to one, the GPIO interrupt request line is asserted. A number of interrupt signals are ORed-wired together to generate a single interrupt signal to the interrupt controller. The Interrupt Flag Register (IFR) can by read to determine which pin(s) caused the interrupt. The interrupt bit must be cleared by writing a one to the Interrupt Flag Clear Register (IFRC). GPIO interrupts can only be triggered when the CLK_GPIO is enabled. 17.5.4 Interrupt Timings The figure below shows the timing for rising edge (or pin-change) interrupts when the glitch filter is disabled. For the pulse to be registered, it must be sampled at the rising edge of the clock. In this example, this is not the case for the first pulse. The second pulse is however sampled on a rising edge and will trigger an interrupt request. Figure 17-4. Interrupt Timing With Glitch Filter Disabled clock Pin Level GPIO_IFR The figure below shows the timing for rising edge (or pin-change) interrupts when the glitch filter is enabled. For the pulse to be registered, it must be sampled on two subsequent rising edges. In the example, the first pulse is rejected while the second pulse is accepted and causes an interrupt request. Figure 17-5. Interrupt Timing With Glitch Filter Enabled clock Pin Level GPIO_IFR 176 32059I-06/2010 AT32UC3B 17.6 User Interface The GPIO controls all the I/O pins on the AVR32 microcontroller. The pins are managed as 32bit ports that are configurable through a PB interface. Each port has a set of configuration registers. The overall memory map of the GPIO is shown below. The number of pins and hence the number of ports are product specific. Figure 17-6. Overall Mermory Map 0x0000 Port 0 Configuration Registers 0x0100 Port 1 Configuration Registers 0x0200 Port 2 Configuration Registers 0x0300 Port 3 Configuration Registers 0x0400 Port 4 Configuration Registers In the GPIO Controller Function Multiplexingtable in the Package and Pinout chapter, each GPIO line has a unique number. Note that the PA, PB, PC and PX ports do not directly correspond to the GPIO ports. To find the corresponding port and pin the following formula can be used: GPIO port = floor((GPIO number) / 32), example: floor((36)/32) = 1 GPIO pin = GPIO number mod 32, example: 36 mod 32 = 4 The table below shows the configuration registers for one port. Addresses shown are relative to the port address offset. The specific address of a configuration register is found by adding the 177 32059I-06/2010 AT32UC3B register offset and the port offset to the GPIO start address. One bit in each of the configuration registers corresponds to an I/O pin. Table 17-1. Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x40 0x44 0x48 0x4C 0x50 0x54 0x58 0x5c 0x60 0x70 0x74 0x78 0x7C 0x90 0x94 0x98 0x9C 0xA0 0xA4 0xA8 0xAC 0xB0 GPIO Register Memory Map Register GPIO Enable Register GPIO Enable Register GPIO Enable Register GPIO Enable Register Peripheral Mux Register 0 Peripheral Mux Register 0 Peripheral Mux Register 0 Peripheral Mux Register 0 Peripheral Mux Register 1 Peripheral Mux Register 1 Peripheral Mux Register 1 Peripheral Mux Register 1 Function Read/Write Set Clear Toggle Read/Write Set Clear Toggle Read/Write Set Clear Toggle Read/Write Set Clear Toggle Read/Write Set Clear Toggle Read Read/Write Set Clear Toggle Read/Write Set Clear Toggle Read/Write Set Clear Toggle Read/Write Name GPER GPERS GPERC GPERT PMR0 PMR0S PMR0C PMR0T PMR1 PMR1S PMR1C PMR1T ODER ODERS ODERC ODERT OVR OVRS OVRC OVRT PVR PUER PUERS PUERC PUERT IER IERS IERC IERT IMR0 IMR0S IMR0C IMR0T IMR1 Access Read/Write Write-Only Write-Only Write-Only Read/Write Write-Only Write-Only Write-Only Read/Write Write-Only Write-Only Write-Only Read/Write Write-Only Write-Only Write-Only Read/Write Write-Only Write-Only Write-Only Read-Only Read/Write Write-Only Write-Only Write-Only Read/Write Write-Only Write-Only Write-Only Read/Write Write-Only Write-Only Write-Only Read/Write (1) (1) (1) (2) (1) (1) (1) (1) (1) Reset value (1) Output Driver Enable Register Output Driver Enable Register Output Driver Enable Register Output Driver Enable Register Output Value Register Output Value Register Output Value Register Output Value Register Pin Value Register Pull-up Enable Register Pull-up Enable Register Pull-up Enable Register Pull-up Enable Register Interrupt Enable Register Interrupt Enable Register Interrupt Enable Register Interrupt Enable Register Interrupt Mode Register 0 Interrupt Mode Register 0 Interrupt Mode Register 0 Interrupt Mode Register 0 Interrupt Mode Register 1 178 32059I-06/2010 AT32UC3B Table 17-1. Offset 0xB4 0xB8 0xBC 0xC0 0xC4 0xC8 0xCC 0xD0 0xD4 0xD8 0xDC GPIO Register Memory Map Register Interrupt Mode Register 1 Interrupt Mode Register 1 Interrupt Mode Register 1 Glitch Filter Enable Register Glitch Filter Enable Register Glitch Filter Enable Register Glitch Filter Enable Register Interrupt Flag Register Interrupt Flag Register Interrupt Flag Register Interrupt Flag Register Function Set Clear Toggle Read/Write Set Clear Toggle Read Clear Name IMR1S IMR1C IMR1T GFER GFERS GFERC GFERT IFR IFRC Access Write-Only Write-Only Write-Only Read/Write Write-Only Write-Only Write-Only Read-Only Write-Only (1) (1) Reset value 1) The reset value for these registers are device specific. Please refer to the Module Configuration section at the end of this chapter. 2) The reset value is undefined depending on the pin states. 17.6.1 Access Types Each configuration register can be accessed in four different ways. The first address location can be used to write the register directly. This address can also be used to read the register value. The following addresses facilitate three different types of write access to the register. Performing a "set" access, all bits written to one will be set. Bits written to zero will be unchanged by the operation. Performing a "clear" access, all bits written to one will be cleared. Bits written to zero will be unchanged by the operation. Finally, a toggle access will toggle the value of all bits written to one. Again all bits written to zero remain unchanged. Note that for some registers (e.g. IFR), not all access methods are permitted. Note that for ports with less than 32 bits, the corresponding control registers will have unused bits. This is also the case for features that are not implemented for a specific pin. Writing to an unused bit will have no effect. Reading unused bits will always return 0. 179 32059I-06/2010 AT32UC3B 17.6.2 Name: Access Type: Offset: Reset Value: 31 Enable Register GPER Read, Write, Set, Clear, Toggle 0x00, 0x04, 0x08, 0x0C 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-P31: Pin Enable 0: A peripheral function controls the corresponding pin. 1: The GPIO controls the corresponding pin. 180 32059I-06/2010 AT32UC3B 17.6.3 Name: Access Type: Offset: Reset Value: 31 Peripheral Mux Register 0 PMR0 Read, Write, Set, Clear, Toggle 0x10, 0x14, 0x18, 0x1C 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Peripheral Multiplexer Select bit 0 181 32059I-06/2010 AT32UC3B 17.6.4 Name: Access Type: Offset: Reset Value: 31 Peripheral Mux Register 1 PMR1 Read, Write, Set, Clear, Toggle 0x20, 0x24, 0x28, 0x2C 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Peripheral Multiplexer Select bit 1 {PMR1, PMR0} 00 01 10 11 Selected Peripheral Function A B C D 182 32059I-06/2010 AT32UC3B 17.6.5 Name: Access Type: Offset: Reset Value: 31 Output Driver Enable Register ODER Read, Write, Set, Clear, Toggle 0x40, 0x44, 0x48, 0x4C 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Output Driver Enable 0: The output driver is disabled for the corresponding pin. 1: The output driver is enabled for the corresponding pin. 183 32059I-06/2010 AT32UC3B 17.6.6 Name: Access Type: Offset: Reset Value: 31 Output Value Register OVR Read, Write, Set, Clear, Toggle 0x50, 0x54, 0x58, 0x5C 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Output Value 0: The value to be driven on the I/O line is 0. 1: The value to be driven on the I/O line is 1. 184 32059I-06/2010 AT32UC3B 17.6.7 Name: Access Type: Offset: Reset Value: 31 Pin Value Register PVR Read 0x60, 0x64, 0x68, 0x6C 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Pin Value 0: The I/O line is at level `0'. 1: The I/O line is at level `1'. Note that the level of a pin can only be read when GPER is set or interrupt is enabled for the pin. 185 32059I-06/2010 AT32UC3B 17.6.8 Name: Access Type: Offset: Reset Value: 31 Pull-up Enable Register PUER Read, Write, Set, Clear, Toggle 0x70, 0x74, 0x78, 0x7C 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Pull-up Enable 0: The internal pull-up resistor is disabled for the corresponding pin. 1: The internal pull-up resistor is enabled for the corresponding pin. 186 32059I-06/2010 AT32UC3B 17.6.9 Name: Access Type: Offset: Reset Value: 31 Interrupt Enable Register IER Read, Write, Set, Clear, Toggle 0x90, 0x94, 0x98, 0x9C 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Interrupt Enable 0: Interrupt is disabled for the corresponding pin. 1: Interrupt is enabled for the corresponding pin. 187 32059I-06/2010 AT32UC3B 17.6.10 Name: Access Type: Offset: Reset Value: 31 Interrupt Mode Register 0 IMR0 Read, Write, Set, Clear, Toggle 0xA0, 0xA4, 0xA8, 0xAC 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Interrupt Mode Bit 0 188 32059I-06/2010 AT32UC3B 17.6.11 Name: Access Type: Offset: Reset Value: 31 Interrupt Mode Register 1 IMR1 Read, Write, Set, Clear, Toggle 0xB0, 0xB4, 0xB8, 0xBC 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Interrupt Mode Bit 1 {IMR1, IMR0} 00 01 10 11 Interrupt Mode Pin Change Rising Edge Falling Edge Reserved 189 32059I-06/2010 AT32UC3B 17.6.12 Name: Access Type: Offset: Reset Value: 31 Glitch Filter Enable Register GFER Read, Write, Set, Clear, Toggle 0xC0, 0xC4, 0xC8, 0xCC 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Glitch Filter Enable 0: Glitch filter is disabled for the corresponding pin. 1: Glitch filter is enabled for the corresponding pin. NOTE! The value of this register should only be changed when IER is `0'. Updating this GFER while interrupt on the corresponding pin is enabled can cause an unintentional interrupt to be triggered. 190 32059I-06/2010 AT32UC3B 17.6.13 Name: Access Type: Offset: Reset Value: 31 Interrupt Flag Register IFR Read, Clear 0xD0, 0xD8 30 29 28 27 26 25 24 P31 23 P30 22 P29 21 P28 20 P27 19 P26 18 P25 17 P24 16 P23 15 P22 14 P21 13 P20 12 P19 11 P18 10 P17 9 P16 8 P15 7 P14 6 P13 5 P12 4 P11 3 P10 2 P9 1 P8 0 P7 P6 P5 P4 P3 P2 P1 P0 * P0-31: Interrupt Flag 1: An interrupt condition has been detected on the corresponding pin. 0: No interrupt condition has beedn detected on the corresponding pin since reset or the last time it was cleared. The number of interrupt request lines is dependant on the number of I/O pins on the MCU. Refer to the product specific data for details. Note also that a bit in the Interrupt Flag register is only valid if the corresponding bit in IER is set. 191 32059I-06/2010 AT32UC3B 17.7 17.7.1 Programming Examples 8-bit LED-Chaser // Set R0 to GPIO base address mov orh R0, LO(AVR32_GPIO_ADDRESS) R0, HI(AVR32_GPIO_ADDRESS) // Enable GPIO control of pin 0-8 mov st.w R1, 0xFF R0[AVR32_GPIO_GPERS], R1 // Set initial value of port mov st.w R2, 0x01 R0[AVR32_GPIO_OVRS], R2 // Set up toggle value. Two pins are toggled // in each round. The bit that is currently set, // and the next bit to be set. mov orh R2, 0x0303 R2, 0x0303 loop: // Only change 8 LSB mov and st.w rol rcall rjmp R3, 0x00FF R3, R2 R0[AVR32_GPIO_OVRT], R3 R2 delay loop It is assumed in this example that a subroutine "delay" exists that returns after a given time. 17.7.2 Configuration of USART pins The example below shows how to configure a peripheral module to control I/O pins. It assumed in this example that the USART receive pin (RXD) is connected to PC16 and that the USART transmit pin (TXD) is connected to PC17. For both pins, the USART is peripheral B. In this example, the state of the GPIO registers is assumed to be unknown. The two USART pins are therefore first set to be controlled by the GPIO with output drivers disabled. The pins can then be assured to be tri-stated while changing the Peripheral Mux Registers. // Set up pointer to GPIO, PORTC mov orh R0, LO(AVR32_GPIO_ADDRESS + PORTC_OFFSET) R0, HI(AVR32_GPIO_ADDRESS + PORTC_OFFSET) // Disable output drivers 192 32059I-06/2010 AT32UC3B mov orh st.w R1, 0x0000 R1, 0x0003 R0[AVR32_GPIO_ODERC], R1 // Make the GPIO control the pins st.w R0[AVR32_GPIO_GPERS], R1 // Select peripheral B on PC16-PC17 st.w st.w R0[AVR32_GPIO_PMR0S], R1 R0[AVR32_GPIO_PMR1C], R1 // Enable peripheral control st.w R0[AVR32_GPIO_GPERC], R1 193 32059I-06/2010 AT32UC3B 17.8 Module Configuration The specific configuration for each GPIO instance is listed in the following tables. The module bus clocks listed here are connected to the system bus clocks. Refer to the Power Manager chapter for details. Table 17-2. Feature Number of GPIO ports Number of peripheral functions Module Configuration GPIO 2 4 Table 17-3. Module Name GPIO Module Clock Name Clock Name CLK_GPIO The reset values for all GPIO registers are zero, with the following exceptions: Table 17-4. Port 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 Register Reset Values Register GPER PMR0 PMR1 ODER OVR PUER IER IMR0 IMR1 GFER IFR GPER PMR0 PMR1 ODER OVR PUER IER IMR0 IMR1 Reset Value TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD TBD 194 32059I-06/2010 AT32UC3B Table 17-4. Port 1 1 1 Register Reset Values Register GFER IFR GPER Reset Value TBD TBD TBD 195 32059I-06/2010 AT32UC3B 18. Serial Peripheral Interface (SPI) Rev. 1.9.9.2 18.1 Features * Supports Communication with Serial External Devices - Four Chip Selects with External Decoder Support Allow Communication with Up to 15 Peripherals - Serial Memories, such as DataFlash and 3-wire EEPROMs - Serial Peripherals, such as ADCs, DACs, LCD Controllers, CAN Controllers and Sensors - External Co-processors * Master or Slave Serial Peripheral Bus Interface - 8 - to 16-bit Programmable Data Length Per Chip Select - Programmable Phase and Polarity Per Chip Select - Programmable Transfer Delays Between Consecutive Transfers and Between Clock and Data Per Chip Select - Programmable Delay Between Consecutive Transfers - Selectable Mode Fault Detection * Connection to PDCA Channel Capabilities Optimizes Data Transfers - One Channel for the Receiver, One Channel for the Transmitter - Next Buffer Support 18.2 Overview The Serial Peripheral Interface (SPI) circuit is a synchronous serial data link that provides communication with external devices in Master or Slave Mode. It also enables communication between processors if an external processor is connected to the system. The Serial Peripheral Interface is essentially a shift register that serially transmits data bits to other SPIs. During a data transfer, one SPI system acts as the "master"' which controls the data flow, while the other devices act as "slaves'' which have data shifted into and out by the master. Different CPUs can take turn being masters (Multiple Master Protocol opposite to Single Master Protocol where one CPU is always the master while all of the others are always slaves) and one master may simultaneously shift data into multiple slaves. However, only one slave may drive its output to write data back to the master at any given time. A slave device is selected when the master asserts its NSS signal. If multiple slave devices exist, the master generates a separate slave select signal for each slave (NPCS). The SPI system consists of two data lines and two control lines: * Master Out Slave In (MOSI): This data line supplies the output data from the master shifted into the input(s) of the slave(s). * Master In Slave Out (MISO): This data line supplies the output data from a slave to the input of the master. There may be no more than one slave transmitting data during any particular transfer. * Serial Clock (SPCK): This control line is driven by the master and regulates the flow of the data bits. The master may transmit data at a variety of baud rates; the SPCK line cycles once for each bit that is transmitted. * Slave Select (NSS): This control line allows slaves to be turned on and off by hardware. 196 32059I-06/2010 AT32UC3B 18.3 Block Diagram Figure 18-1. Block Diagram PDCA Peripheral Bus SPCK MISO CLK_SPI Spi Interface GPIO MOSI NPCS0/NSS NPCS1 CLK_SPI 32 Interrupt Control NPCS2 NPCS3 SPI Interrupt DIV 18.4 Application Block Diagram Figure 18-2. Application Block Diagram: Single Master/Multiple Slave Implementation SPCK MISO MOSI Spi Master NPCS0 NPCS1 NPCS2 NPCS3 NC SPCK MISO MOSI NSS Slave 1 SPCK MISO MOSI NSS Slave 0 SPCK MISO MOSI NSS Slave 2 197 32059I-06/2010 AT32UC3B 18.5 Signal Description Table 18-1. Type Pin Name MISO MOSI SPCK NPCS1-NPCS3 NPCS0/NSS Pin Description Master In Slave Out Master Out Slave In Serial Clock Peripheral Chip Selects Peripheral Chip Select/Slave Select Master Input Output Output Output Output Slave Output Input Input Unused Input 18.6 18.6.1 Product Dependencies I/O Lines The pins used for interfacing the compliant external devices may be multiplexed with GPIO lines. The programmer must first program the GPIO controller to assign the SPI pins to their peripheral functions.To use the local loopback function the SPI pins must be controlled by the SPI. 18.6.2 Power Management The SPI may be clocked through the Power Manager, Before using the SPI, the programmer must ensure that the SPI clock is enabled in the Power Manager. In the SPI description, CLK_SPI is the clock of the peripheral bus to which the SPI is connected. 18.6.3 Interrupt The SPI interface has an interrupt line connected to the Interrupt Controller (INTC). Handling the SPI interrupt requires programming the INTC before configuring the SPI. 18.7 18.7.1 Functional Description Modes of Operation The SPI operates in Master Mode or in Slave Mode. Operation in Master Mode is programmed by writing at 1 the MSTR bit in the Mode Register. The pins NPCS0 to NPCS3 are all configured as outputs, the SPCK pin is driven, the MISO line is wired on the receiver input and the MOSI line driven as an output by the transmitter. If the MSTR bit is written at 0, the SPI operates in Slave Mode. The MISO line is driven by the transmitter output, the MOSI line is wired on the receiver input, the SPCK pin is driven by the transmitter to synchronize the receiver. The NPCS0 pin becomes an input, and is used as a Slave Select signal (NSS). The pins NPCS1 to NPCS3 are not driven and can be used for other purposes. The data transfers are identically programmable for both modes of operations. The baud rate generator is activated only in Master Mode. 18.7.2 Data Transfer Four combinations of polarity and phase are available for data transfers. The clock polarity is programmed with the CPOL bit in the Chip Select Register. The clock phase is programmed with 198 32059I-06/2010 AT32UC3B the NCPHA bit. These two parameters determine the edges of the clock signal on which data is driven and sampled. Each of the two parameters has two possible states, resulting in four possible combinations that are incompatible with one another. Thus, a master/slave pair must use the same parameter pair values to communicate. If multiple slaves are used and fixed in different configurations, the master must reconfigure itself each time it needs to communicate with a different slave. Table 18-2 shows the four modes and corresponding parameter settings. Table 18-2. SPI modes SPI Mode 0 1 2 3 CPOL 0 0 1 1 NCPHA 1 0 1 0 Figure 18-3 on page 199 and Figure 18-4 on page 199 show examples of data transfers. Figure 18-3. SPI Transfer Format (NCPHA = 1, 8 bits per transfer) SPCK cycle (for reference) 1 2 3 4 5 6 7 8 SPCK (CPOL = 0) SPCK (CPOL = 1) MOSI (from master) MSB 6 5 4 3 2 1 LSB MISO (from slave) MSB 6 5 4 3 2 1 LSB *** NSS (to slave) *** Not Defined, but normaly MSB of previous character received Figure 18-4. SPI Transfer Format (NCPHA = 0, 8 bits per transfer) SPCK cycle (for reference) 1 2 3 4 5 6 7 8 SPCK (CPOL = 0) SPCK (CPOL = 1) MOSI (from master) MSB 6 5 4 3 2 1 LSB MISO (from slave) *** MSB 6 5 4 3 2 1 LSB NSS (to slave) 199 32059I-06/2010 AT32UC3B 18.7.3 Master Mode Operations When configured in Master Mode, the SPI uses the internal programmable baud rate generator as clock source. It fully controls the data transfers to and from the slave(s) connected to the SPI bus. The SPI drives the chip select line to the slave and the serial clock signal (SPCK). The SPI features two holding registers, the Transmit Data Register and the Receive Data Register, and a single Shift Register. The holding registers maintain the data flow at a constant rate. After enabling the SPI, a data transfer begins when the processor writes to the TDR (Transmit Data Register). The written data is immediately transferred in the Shift Register and transfer on the SPI bus starts. While the data in the Shift Register is shifted on the MOSI line, the MISO line is sampled and shifted in the Shift Register. Transmission cannot occur without reception. Before writing the TDR, the PCS field must be set in order to select a slave. If new data is written in TDR during the transfer, it stays in it until the current transfer is completed. Then, the received data is transferred from the Shift Register to RDR, the data in TDR is loaded in the Shift Register and a new transfer starts. The transfer of a data written in TDR in the Shift Register is indicated by the TDRE bit (Transmit Data Register Empty) in the Status Register (SR). When new data is written in TDR, this bit is cleared. The TDRE bit is used to trigger the Transmit Peripheral DMA Controller channel. The end of transfer is indicated by the TXEMPTY flag in the SR register. If a transfer delay (DLYBCT) is greater than 0 for the last transfer, TXEMPTY is set after the completion of said delay. The CLK_SPI can be switched off at this time. The transfer of received data from the Shift Register in RDR is indicated by the RDRF bit (Receive Data Register Full) in the Status Register (SR). When the received data is read, the RDRF bit is cleared. If the RDR (Receive Data Register) has not been read before new data is received, the Overrun Error bit (OVRES) in SR is set. When this bit is set the SPI will continue to update RDR when data is received, overwriting the previously received data. The user has to read the status register to clear the OVRES bit. Figure 18-5 on page 201 shows a block diagram of the SPI when operating in Master Mode. Figure 18-6 on page 202 shows a flow chart describing how transfers are handled. 200 32059I-06/2010 AT32UC3B 18.7.3.1 Master Mode Block Diagram Figure 18-5. Master Mode Block Diagram CSR0..3 SCBR CLK_SPI Baud Rate Generator SPCK SPI Clock RDR CSR0..3 BITS NCPHA CPOL MISO LSB Shift Register MSB MOSI RD RDRF OVRES TDR TD TDRE RDR PCS CSR0..3 PS MR PCS 0 TDR PCS 1 Current Peripheral PCSDEC CSAAT NPCS3 NPCS2 NPCS1 NPCS0 MSTR MODF NPCS0 MODFDIS 201 32059I-06/2010 AT32UC3B 18.7.3.2 Master Mode Flow Diagram Figure 18-6. Master Mode Flow Diagram 202 32059I-06/2010 AT32UC3B 18.7.3.3 Clock Generation The SPI Baud rate clock is generated by dividing the CLK_SPI by a value between 1 and 255. This allows a maximum operating baud rate at up to CLK_SPI and a minimum operating baud rate of CLK_SPI divided by 255. Programming the SCBR field at 0 is forbidden. Triggering a transfer while SCBR is at 0 can lead to unpredictable results. At reset, SCBR is 0 and the user has to program it at a valid value before performing the first transfer. The divisor can be defined independently for each chip select, as it has to be programmed in the SCBR field of the Chip Select Registers. This allows the SPI to automatically adapt the baud rate for each interfaced peripheral without reprogramming. 18.7.3.4 Transfer Delays Figure 18-7 on page 203 shows a chip select transfer change and consecutive transfers on the same chip select. Three delays can be programmed to modify the transfer waveforms: * The delay between chip selects, programmable only once for all the chip selects by writing the DLYBCS field in the Mode Register. Allows insertion of a delay between release of one chip select and before assertion of a new one. * The delay before SPCK, independently programmable for each chip select by writing the field DLYBS. Allows the start of SPCK to be delayed after the chip select has been asserted. * The delay between consecutive transfers, independently programmable for each chip select by writing the DLYBCT field. Allows insertion of a delay between two transfers occurring on the same chip select These delays allow the SPI to be adapted to the interfaced peripherals and their speed and bus release time. Figure 18-7. Programmable Delays Chip Select 1 Chip Select 2 SPCK DLYBCS DLYBS DLYBCT DLYBCT 203 32059I-06/2010 AT32UC3B 18.7.3.5 Peripheral Selection The serial peripherals are selected through the assertion of the NPCS0 to NPCS3 signals. By default, all the NPCS signals are high before and after each transfer. The peripheral selection can be performed in two different ways: * Fixed Peripheral Select: SPI exchanges data with only one peripheral * Variable Peripheral Select: Data can be exchanged with more than one peripheral Fixed Peripheral Select is activated by writing the PS bit to zero in MR (Mode Register). In this case, the current peripheral is defined by the PCS field in MR and the PCS field in the TDR has no effect. Variable Peripheral Select is activated by setting PS bit to one. The PCS field in TDR is used to select the current peripheral. This means that the peripheral selection can be defined for each new data. The Fixed Peripheral Selection allows buffer transfers with a single peripheral. Using the Peripheral DMA Controller is an optimal means, as the size of the data transfer between the memory and the SPI is either 8 bits or 16 bits. However, changing the peripheral selection requires the Mode Register to be reprogrammed. The Variable Peripheral Selection allows buffer transfers with multiple peripherals without reprogramming the Mode Register. Data written in TDR is 32 bits wide and defines the real data to be transmitted and the peripheral it is destined to. Using the Peripheral DMA Controller in this mode requires 32-bit wide buffers, with the data in the LSBs and the PCS and LASTXFER fields in the MSBs, however the SPI still controls the number of bits (8 to16) to be transferred through MISO and MOSI lines with the chip select configuration registers. This is not the optimal means in term of memory size for the buffers, but it provides a very effective means to exchange data with several peripherals without any intervention of the processor. 18.7.3.6 Peripheral Chip Select Decoding The user can program the SPI to operate with up to 15 peripherals by decoding the four Chip Select lines, NPCS0 to NPCS3 with an external logic. This can be enabled by writing the PCSDEC bit at 1 in the Mode Register (MR). When operating without decoding, the SPI makes sure that in any case only one chip select line is activated, i.e. driven low at a time. If two bits are defined low in a PCS field, only the lowest numbered chip select is driven low. When operating with decoding, the SPI directly outputs the value defined by the PCS field of either the Mode Register or the Transmit Data Register (depending on PS). As the SPI sets a default value of 0xF on the chip select lines (i.e. all chip select lines at 1) when not processing any transfer, only 15 peripherals can be decoded. The SPI has only four Chip Select Registers, not 15. As a result, when decoding is activated, each chip select defines the characteristics of up to four peripherals. As an example, CRS0 defines the characteristics of the externally decoded peripherals 0 to 3, corresponding to the PCS values 0x0 to 0x3. Thus, the user has to make sure to connect compatible peripherals on the decoded chip select lines 0 to 3, 4 to 7, 8 to 11 and 12 to 14. 204 32059I-06/2010 AT32UC3B 18.7.3.7 Peripheral Deselection When operating normally, as soon as the transfer of the last data written in TDR is completed, the NPCS lines all rise. This might lead to runtime error if the processor is too long in responding to an interrupt, and thus might lead to difficulties for interfacing with some serial peripherals requiring the chip select line to remain active during a full set of transfers. To facilitate interfacing with such devices, the Chip Select Register can be programmed with the CSAAT bit (Chip Select Active After Transfer) at 1. This allows the chip select lines to remain in their current state (low = active) until transfer to another peripheral is required. Figure 18-8 on page 205 shows different peripheral deselection cases and the effect of the CSAAT bits. Figure 18-8. Peripheral Deselection CSAAT = 0 CSAAT = 1 TDRE DLYBCT A DLYBCS PCS = A A A DLYBCT A DLYBCS PCS = A A NPCS[0..3] Write SPI_TDR TDRE DLYBCT A DLYBCS PCS=A A A DLYBCT A DLYBCS PCS = A A NPCS[0..3] Write SPI_TDR TDRE NPCS[0..3] DLYBCT A DLYBCS PCS = B B A DLYBCT B DLYBCS PCS = B Write SPI_TDR 18.7.3.8 Mode Fault Detection A mode fault is detected when the SPI is programmed in Master Mode and a low level is driven by an external master on the NPCS0/NSS signal. NPCS0, MOSI, MISO and SPCK must be configured in open drain through the GPIO controller, so that external pull up resistors are needed to guarantee high level. 205 32059I-06/2010 AT32UC3B When a mode fault is detected, the MODF bit in the SR is set until the SR is read and the SPI is automatically disabled until re-enabled by writing the SPIEN bit in the CR (Control Register) at 1. By default, the Mode Fault detection circuitry is enabled. The user can disable Mode Fault detection by setting the MODFDIS bit in the SPI Mode Register (MR). 18.7.4 SPI Slave Mode When operating in Slave Mode, the SPI processes data bits on the clock provided on the SPI clock pin (SPCK). The SPI waits for NSS to go active before receiving the serial clock from an external master. When NSS falls, the clock is validated on the serializer, which processes the number of bits defined by the BITS field of the Chip Select Register 0 (CSR0). These bits are processed following a phase and a polarity defined respectively by the NCPHA and CPOL bits of the CSR0. Note that BITS, CPOL and NCPHA of the other Chip Select Registers have no effect when the SPI is programmed in Slave Mode. The bits are shifted out on the MISO line and sampled on the MOSI line. When all the bits are processed, the received data is transferred in the Receive Data Register and the RDRF bit rises. If the RDR (Receive Data Register) has not been read before new data is received, the Overrun Error bit (OVRES) in SR is set. As long as this flag is set, data is loaded in RDR. The user has to read the status register to clear the OVRES bit. When a transfer starts, the data shifted out is the data present in the Shift Register. If no data has been written in the Transmit Data Register (TDR), the last data received is transferred. If no data has been received since the last reset, all bits are transmitted low, as the Shift Register resets at 0. When a first data is written in TDR, it is transferred immediately in the Shift Register and the TDRE bit rises. If new data is written, it remains in TDR until a transfer occurs, i.e. NSS falls and there is a valid clock on the SPCK pin. When the transfer occurs, the last data written in TDR is transferred in the Shift Register and the TDRE bit rises. This enables frequent updates of critical variables with single transfers. Then, a new data is loaded in the Shift Register from the Transmit Data Register. In case no character is ready to be transmitted, i.e. no character has been written in TDR since the last load from TDR to the Shift Register, the Shift Register is not modified and the last received character is retransmitted. Figure 18-9 on page 207 shows a block diagram of the SPI when operating in Slave Mode. 206 32059I-06/2010 AT32UC3B Figure 18-9. Slave Mode Functional Block Diagram SPCK NSS SPIEN SPIENS SPIDIS SPI_CSR0 BITS NCPHA CPOL MOSI LSB SPI_RDR RD RDRF OVRES SPI Clock Shift Register MSB MISO SPI_TDR FLOAD TD TDRE 207 32059I-06/2010 AT32UC3B 18.8 User Interface SPI Register Memory Map Register Control Register Mode Register Receive Data Register Transmit Data Register Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Reserved Chip Select Register 0 Chip Select Register 1 Chip Select Register 2 Chip Select Register 3 Reserved Version Register Reserved CSR0 CSR1 CSR2 CSR3 - VERSION - Read/write Read/write Read/write Read/write - Read-only - 0x00000000 0x00000000 0x00000000 0x00000000 - 0x- (1) - Name CR MR RDR TDR SR IER IDR IMR Access Write-only Read/write Read-only Write-only Read-only Write-only Write-only Read-only Reset 0x00000000 0x00000000 0x00000000 0x00000000 0x000000F0 0x00000000 0x00000000 0x00000000 Table 18-3. Offset 0x00 0x04 0x08 0x0C 0x10 0x14 0x18 0x1C 0x20 - 0x2C 0x30 0x34 0x38 0x3C 0x004C - 0x00F8 0x00FC 0x004C - 0x00FC Note: 1. Values in the Version Register vary with the version of the IP block implementation. 208 32059I-06/2010 AT32UC3B 18.8.1 Name: Access Type: Offset: Reset Value: 31 Control Register CR Write-only 0x00 0x00000000 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 LASTXFER 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 - 1 0 SWRST - - - - - SPIDIS SPIEN * LASTXFER: Last Transfer 0: No effect. 1: The current NPCS will be deasserted after the character written in TD has been transferred. When CSAAT is set, this allows to close the communication with the current serial peripheral by raising the corresponding NPCS line as soon as TD transfer has completed. * SWRST: SPI Software Reset 0: No effect. 1: Reset the SPI. A software-triggered hardware reset of the SPI interface is performed. The SPI is in slave mode after software reset. Peripheral DMA Controller channels are not affected by software reset. * SPIDIS: SPI Disable 0: No effect. 1: Disables the SPI. As soon as SPIDIS is set, SPI finishes its tranfer. All pins are set in input mode and no data is received or transmitted. If a transfer is in progress, the transfer is finished before the SPI is disabled. If both SPIEN and SPIDIS are equal to one when the control register is written, the SPI is disabled. * SPIEN: SPI Enable 0: No effect. 1: Enables the SPI to transfer and receive data. 209 32059I-06/2010 AT32UC3B 18.8.2 Name: Access Type: Offset: Reset Value: 31 Mode Register MR Read/Write 0x04 0x00000000 30 29 28 27 26 25 24 DLYBCS 23 22 21 20 19 18 17 16 - 15 - 14 - 13 - 12 11 10 PCS 9 8 - 7 - 6 - 5 - 4 - 3 - 2 - 1 - 0 LLB - - MODFDIS - PCSDEC PS MSTR * DLYBCS: Delay Between Chip Selects This field defines the delay from NPCS inactive to the activation of another NPCS. The DLYBCS time guarantees nonoverlapping chip selects and solves bus contentions in case of peripherals having long data float times. If DLYBCS is less than or equal to six, six CLK_SPI periods will be inserted by default. Otherwise, the following equation determines the delay: DLYBCS Delay Between Chip Selects = ---------------------CLKSPI * PCS: Peripheral Chip Select This field is only used if Fixed Peripheral Select is active (PS = 0). If PCSDEC = 0: PCS = xxx0NPCS[3:0] = 1110 PCS = xx01NPCS[3:0] = 1101 PCS = x011NPCS[3:0] = 1011 PCS = 0111NPCS[3:0] = 0111 PCS = 1111forbidden (no peripheral is selected) (x = don't care) If PCSDEC = 1: NPCS[3:0] output signals = PCS. * LLB: Local Loopback Enable 0: Local loopback path disabled. 1: Local loopback path enabled ( LLB controls the local loopback on the data serializer for testing in Master Mode only. (MISO is internally connected on MOSI.) * MODFDIS: Mode Fault Detection 0: Mode fault detection is enabled. 1: Mode fault detection is disabled. * PCSDEC: Chip Select Decode 0: The chip selects are directly connected to a peripheral device. 1: The four chip select lines are connected to a 4- to 16-bit decoder. When PCSDEC equals one, up to 15 Chip Select signals can be generated with the four lines using an external 4- to 16-bit decoder. The Chip Select Registers define the characteristics of the 15 chip selects according to the following rules: CSR0 defines peripheral chip select signals 0 to 3. CSR1 defines peripheral chip select signals 4 to 7. 210 32059I-06/2010 AT32UC3B CSR2 defines peripheral chip select signals 8 to 11. CSR3 defines peripheral chip select signals 12 to 14. * PS: Peripheral Select 0: Fixed Peripheral Select. 1: Variable Peripheral Select. * MSTR: Master/Slave Mode 0: SPI is in Slave mode. 1: SPI is in Master mode. 211 32059I-06/2010 AT32UC3B 18.8.3 Name: Access Type: Offset: Reset Value: 31 Receive Data Register RDR Read-only 0x08 0x00000000 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 RD 7 6 5 4 3 2 1 0 RD * RD: Receive Data Data received by the SPI Interface is stored in this register right-justified. Unused bits read zero. 212 32059I-06/2010 AT32UC3B 18.8.4 Name: Access Type: Offset: Reset Value: 31 Transmit Data Register TDR Write-only 0x0C 0x00000000 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 LASTXFER 16 - 15 - 14 - 13 - 12 11 10 PCS 9 8 TD 7 6 5 4 3 2 1 0 TD * LASTXFER: Last Transfer 0: No effect. 1: The current NPCS will be deasserted after the character written in TD has been transferred. When CSAAT is set, this allows to close the communication with the current serial peripheral by raising the corresponding NPCS line as soon as TD transfer has completed. This field is only used if Variable Peripheral Select is active (PS = 1). * PCS: Peripheral Chip Select This field is only used if Variable Peripheral Select is active (PS = 1). If PCSDEC = 0: PCS = xxx0NPCS[3:0] = 1110 PCS = xx01NPCS[3:0] = 1101 PCS = x011NPCS[3:0] = 1011 PCS = 0111NPCS[3:0] = 0111 PCS = 1111forbidden (no peripheral is selected) (x = don't care) If PCSDEC = 1: NPCS[3:0] output signals = PCS * TD: Transmit Data Data to be transmitted by the SPI Interface is stored in this register. Information to be transmitted must be written to the transmit data register in a right-justified format. 213 32059I-06/2010 AT32UC3B 18.8.5 Name: Access Type: Offset: Reset Value: 31 Status Register SR Read-only 0x10 0x00000000 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 SPIENS 8 - 7 - 6 - 5 - 4 - 3 2 TXEMPTY 1 NSSR 0 - - - - OVRES MODF TDRE RDRF * SPIENS: SPI Enable Status 0: SPI is disabled. 1: SPI is enabled. * TXEMPTY: Transmission Registers Empty 0: As soon as data is written in TDR. 1: TDR and internal shifter are empty. If a transfer delay has been defined, TXEMPTY is set after the completion of such delay. * NSSR: NSS Rising 0: No rising edge detected on NSS pin since last read. 1: A rising edge occurred on NSS pin since last read. * OVRES: Overrun Error Status 0: No overrun has been detected since the last read of SR. 1: An overrun has occurred since the last read of SR. An overrun occurs when RDR is loaded at least twice from the serializer since the last read of the RDR. * MODF: Mode Fault Error 0: No Mode Fault has been detected since the last read of SR. 1: A Mode Fault occurred since the last read of the SR. * TDRE: Transmit Data Register Empty 0: Data has been written to TDR and not yet transferred to the serializer. 1: The last data written in the Transmit Data Register has been transferred to the serializer. TDRE equals zero when the SPI is disabled or at reset. The SPI enable command sets this bit to one. * RDRF: Receive Data Register Full 0: No data has been received since the last read of RDR 1: Data has been received and the received data has been transferred from the serializer to RDR since the last read of RDR. 214 32059I-06/2010 AT32UC3B 18.8.6 Name: Access Type: Offset: Reset Value: 31 Interrupt Enable Register IER Write-only 0x14 0x00000000 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 2 TXEMPTY 1 NSSR 0 - - - - OVRES MODF TDRE RDRF Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. 215 32059I-06/2010 AT32UC3B 18.8.7 Name: Access Type: Offset: Reset Value: 31 Interrupt Disable Register IDR Write-only 0x18 0x00000000 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 2 TXEMPTY 1 NSSR 0 - - - - OVRES MODF TDRE RDRF Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. 216 32059I-06/2010 AT32UC3B 18.8.8 Name: Access Type: Offset: Reset Value: 31 Interrupt Mask Register IMR Read-only 0x1C 0x00000000 30 29 28 27 26 25 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 2 TXEMPTY 1 NSSR 0 - - - - OVRES MODF TDRE RDRF 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is set when the corresponding bit in IER is written to one. 217 32059I-06/2010 AT32UC3B 18.8.9 Name: Access Type: Offset: Reset Value: 31 Chip Select Register n CSRn Read/Write 0x30 +0x04*n 0x00000000 30 29 28 27 26 25 24 DLYBCT 23 22 21 20 19 18 17 16 DLYBS 15 14 13 12 11 10 9 8 SCBR 7 6 5 4 3 2 1 0 BITS CSAAT CSNAAT NCPHA CPOL * DLYBCT: Delay Between Consecutive Transfers This field defines the delay between two consecutive transfers with the same peripheral without removing the chip select. The delay is always inserted after each transfer and before removing the chip select if needed. When DLYBCT equals zero, no delay between consecutive transfers is inserted and the clock keeps its duty cycle over the character transfers. Otherwise, the following equation determines the delay: Delay Between Consecutive Transfers = 32 x DLYBCT ----------------------------------CLKSPI * DLYBS: Delay Before SPCK This field defines the delay from NPCS valid to the first valid SPCK transition. When DLYBS equals zero, the NPCS valid to SPCK transition is 1/2 the SPCK clock period. Otherwise, the following equations determine the delay: I DLYBS Delay Before SPCK = -------------------CLKSPI * ISCBR: Serial Clock Baud Rate In Master Mode, the SPI Interface uses a modulus counter to derive the SPCK baud rate from the CLK_SPI. The Baud rate is selected by writing a value from 1 to 255 in the SCBR field. The following equations determine the SPCK baud rate: CLKSPI SPCK Baudrate = --------------------SCBR Writing 0 to the SCBR field is forbidden. Triggering a transfer while SCBR is 0 can lead to unpredictable results. At reset, SCBR is 0 and the user has to write it at a valid value before performing the first transfer. IIf a clock divider (SCBRn) is set to 1 and the other SCBR differ from 1, access on CSn is correct but no correct access will be possible on others CS. 218 32059I-06/2010 AT32UC3B * BITS: Bits Per Transfer The BITS field determines the number of data bits transferred. Reserved values should not be used. BITS 0000 0001 0010 0011 0100 0101 0110 0111 1000 1001 1010 1011 1100 1101 1110 1111 Bits Per Transfer 8 9 10 11 12 13 14 15 16 Reserved Reserved Reserved Reserved Reserved Reserved Reserved * CSAAT: Chip Select Active After Transfer 0: The Peripheral Chip Select Line rises as soon as the last transfer is achieved. 1: The Peripheral Chip Select does not rise after the last transfer is achieved. It remains active until a new transfer is requested on a different chip select. * CSNAAT: Chip Select Not Active After Transfer 0 = The Peripheral Chip Select Line rises as soon as the last transfer is acheived 1 = The Peripheral Chip Select Line rises after every transfer CSNAAT can be used to force the Peripheral Chip Select Line to go inactive after every transfer. This allows successful interfacing to SPI slave devices that require this behavior. * NCPHA: Clock Phase 0: Data is changed on the leading edge of SPCK and captured on the following edge of SPCK. 1: Data is captured on the leading edge of SPCK and changed on the following edge of SPCK. NCPHA determines which edge of SPCK causes data to change and which edge causes data to be captured. NCPHA is used with CPOL to produce the required clock/data relationship between master and slave devices. * CPOL: Clock Polarity 0: The inactive state value of SPCK is logic level zero. 1: The inactive state value of SPCK is logic level one. CPOL is used to determine the inactive state value of the serial clock (SPCK). It is used with NCPHA to produce the required clock/data relationship between master and slave devices. 219 32059I-06/2010 AT32UC3B 19. Two-Wire Interface (TWI) 2.1.1.0 19.1 Features Compatible with Atmel Two-wire Interface Serial Memory and IC Compatible Devices(1) One, Two or Three Bytes for Slave Address Sequential Read-write Operations Master, Multi-master and Slave Mode Operation Bit Rate: Up to 400 Kbits General Call Supported in Slave mode Connection to Peripheral DMA Controller Channel Capabilities Optimizes Data Transfers in Master Mode Only - One Channel for the Receiver, One Channel for the Transmitter - Next Buffer Support Note: 1. See Table 19-1 below for details on compatibility with IC Standard. * * * * * * * 19.2 Overview The Atmel Two-wire Interface (TWI) interconnects components on a unique two-wire bus, made up of one clock line and one data line with speeds of up to 400 Kbits per second, based on a byte-oriented transfer format. It can be used with any Atmel Two-wire Interface bus Serial EEPROM and IC compatible device such as Real Time Clock (RTC), Dot Matrix/Graphic LCD Controllers and Temperature Sensor, to name but a few. The TWI is programmable as a master or a slave with sequential or single-byte access. Multiple master capability is supported. Arbitration of the bus is performed internally and puts the TWI in slave mode automatically if the bus arbitration is lost. A configurable baud rate generator permits the output data rate to be adapted to a wide range of core clock frequencies. Below, Table 19-1 lists the compatibility level of the Atmel Two-wire Interface in Master Mode and a full I2C compatible device. Table 19-1. I2C Standard Standard Mode Speed (100 KHz) Fast Mode Speed (400 KHz) 7 or 10 bits Slave Addressing START BYTE (1) Atmel TWI compatibility with I2C Standard Atmel TWI Supported Supported Supported Not Supported Supported Supported Not Supported Supported Repeated Start (Sr) Condition ACK and NACK Management Slope control and input filtering (Fast mode) Clock stretching Note: 1. START + b000000001 + Ack + Sr 220 32059I-06/2010 AT32UC3B 19.3 List of Abbreviations Table 19-2. Abbreviation TWI A NA P S Sr SADR ADR R W Abbreviations Description Two-wire Interface Acknowledge Non Acknowledge Stop Start Repeated Start Slave Address Any address except SADR Read Write 19.4 Block Diagram Figure 19-1. Block Diagram Peripheral Bus Bridge TWCK PIO Two-wire Interface TWD PM MCK TWI Interrupt INTC 221 32059I-06/2010 AT32UC3B 19.5 Application Block Diagram Figure 19-2. Application Block Diagram VDD Rp TWD TWCK Rp Host with TWI Interface Atmel TWI Serial EEPROM Slave 1 IC RTC Slave 2 IC LCD Controller Slave 3 IC Temp. Sensor Slave 4 Rp: Pull up value as given by the IC Standard 19.6 I/O Lines Description I/O Lines Description Pin Description Two-wire Serial Data Two-wire Serial Clock Type Input/Output Input/Output Table 19-3. Pin Name TWD TWCK 19.7 19.7.1 Product Dependencies I/O Lines Both TWD and TWCK are bidirectional lines, connected to a positive supply voltage via a current source or pull-up resistor (see Figure 19-2 on page 222). When the bus is free, both lines are high. The output stages of devices connected to the bus must have an open-drain or open-collector to perform the wired-AND function. TWD and TWCK pins may be multiplexed with GPIO lines. To enable the TWI, the programmer must perform the following steps: * Program the GPIO controller to: - Dedicate TWD and TWCK as peripheral lines. - Define TWD and TWCK as open-drain. 19.7.2 Power Management The TWI clock is generated by the Power Manager (PM). Before using the TWI, the programmer must ensure that the TWI clock is enabled in the PM. In the TWI description, Master Clock (MCK) is the clock of the peripheral bus to which the TWI is connected. 222 32059I-06/2010 AT32UC3B 19.7.3 Interrupt The TWI interface has an interrupt line connected to the Interrupt Controller (INTC). In order to handle interrupts, the INTC must be programmed before configuring the TWI. 19.8 19.8.1 Functional Description Transfer Format The data put on the TWD line must be 8 bits long. Data is transferred MSB first; each byte must be followed by an acknowledgement. The number of bytes per transfer is unlimited (see Figure 19-4). Each transfer begins with a START condition and terminates with a STOP condition (see Figure 19-3). * A high-to-low transition on the TWD line while TWCK is high defines the START condition. * A low-to-high transition on the TWD line while TWCK is high defines a STOP condition. Figure 19-3. START and STOP Conditions TWD TWCK Start Stop Figure 19-4. Transfer Format TWD TWCK Start Address R/W Ack Data Ack Data Ack Stop 19.9 Modes of Operation The TWI has six modes of operations: * Master transmitter mode * Master receiver mode * Multi-master transmitter mode * Multi-master receiver mode * Slave transmitter mode * Slave receiver mode These modes are described in the following chapters. 223 32059I-06/2010 AT32UC3B 19.10 Master Mode 19.10.1 Definition The Master is the device which starts a transfer, generates a clock and stops it. 19.10.2 Application Block Diagram Figure 19-5. Master Mode Typical Application Block Diagram VDD Rp TWD TWCK Rp Host with TWI Interface Atmel TWI Serial EEPROM Slave 1 IC RTC Slave 2 IC LCD Controller Slave 3 IC Temp. Sensor Slave 4 Rp: Pull up value as given by the IC Standard 19.10.3 Programming Master Mode The following registers have to be programmed before entering Master mode: 1. DADR (+ IADRSZ + IADR if a 10 bit device is addressed): The device address is used to access slave devices in read or write mode. 2. CKDIV + CHDIV + CLDIV: Clock Waveform. 3. SVDIS: Disable the slave mode. 4. MSEN: Enable the master mode. 19.10.4 Master Transmitter Mode After the master initiates a Start condition when writing into the Transmit Holding Register, THR, it sends a 7-bit slave address, configured in the Master Mode register (DADR in MMR), to notify the slave device. The bit following the slave address indicates the transfer direction, 0 in this case (MREAD = 0 in MMR). The TWI transfers require the slave to acknowledge each received byte. During the acknowledge clock pulse (9th pulse), the master releases the data line (HIGH), enabling the slave to pull it down in order to generate the acknowledge. The master polls the data line during this clock pulse and sets the NACK in the status register if the slave does not acknowledge the byte. As with the other status bits, an interrupt can be generated if enabled in the interrupt enable register (IER). If the slave acknowledges the byte, the data written in the THR, is then shifted in the internal shifter and transferred. When an acknowledge is detected, the TXRDY bit is set until a new write in the THR. When no more data is written into the THR, the master generates a stop condition to end the transfer. The end of the complete transfer is marked by the TXCOMP bit set to one. See Figure 19-6, Figure 19-7, and Figure 19-8 on page 225. TXRDY is used as Transmit Ready for the Peripheral DMA Controller transmit channel. 224 32059I-06/2010 AT32UC3B Figure 19-6. Master Write with One Data Byte TWD S DADR W A DATA A P TXCOMP TXRDY Write THR (DATA) STOP sent automaticaly (ACK received and TXRDY = 1) Figure 19-7. Master Write with Multiple Data Byte TWD S DADR W A DATA n A DATA n+5 A DATA n+x A P TXCOMP TXRDY Write THR (Data n) Write THR (Data n+1) Write THR (Data n+x) Last data sent STOP sent automaticaly (ACK received and TXRDY = 1) Figure 19-8. Master Write with One Byte Internal Address and Multiple Data Bytes TWD S DADR W A IADR(7:0) A DATA n A DATA n+5 A DATA n+x A P TXCOMP TXRDY Write THR (Data n) Write THR (Data n+1) Write THR (Data n+x) STOP sent automaticaly Last data sent (ACK received and TXRDY = 1) 19.10.5 Master Receiver Mode The read sequence begins by setting the START bit. After the start condition has been sent, the master sends a 7-bit slave address to notify the slave device. The bit following the slave address indicates the transfer direction, 1 in this case (MREAD = 1 in MMR). During the acknowledge clock pulse (9th pulse), the master releases the data line (HIGH), enabling the slave to pull it down in order to generate the acknowledge. The master polls the data line during this clock pulse and sets the NACK bit in the status register if the slave does not acknowledge the byte. If an acknowledge is received, the master is then ready to receive data from the slave. After data has been received, the master sends an acknowledge condition to notify the slave that the data has been received except for the last data, after the stop condition. See Figure 19-9. When the 225 32059I-06/2010 AT32UC3B RXRDY bit is set in the status register, a character has been received in the receive-holding register (RHR). The RXRDY bit is reset when reading the RHR. When a single data byte read is performed, with or without internal address (IADR), the START and STOP bits must be set at the same time. See Figure 19-9. When a multiple data byte read is performed, with or without IADR, the STOP bit must be set after the next-to-last data received. See Figure 19-10. For Internal Address usage see "Internal Address" on page 226. Figure 19-9. Master Read with One Data Byte TWD S DADR R A DATA N P TXCOMP Write START & STOP Bit RXRDY Read RHR Figure 19-10. Master Read with Multiple Data Bytes TWD S DADR R A DATA n A DATA (n+1) A DATA (n+m)-1 A DATA (n+m) N P TXCOMP Write START Bit RXRDY Read RHR DATA n Read RHR DATA (n+1) Read RHR DATA (n+m)-1 Read RHR DATA (n+m) Write STOP Bit after next-to-last data read RXRDY is used as Receive Ready for the Peripheral DMA Controller receive channel. 19.10.6 Internal Address The TWI interface can perform various transfer formats: Transfers with 7-bit slave address devices and 10-bit slave address devices. 19.10.6.1 7-bit Slave Addressing When Addressing 7-bit slave devices, the internal address bytes are used to perform random address (read or write) accesses to reach one or more data bytes, within a memory page location in a serial memory, for example. When performing read operations with an internal address, the TWI performs a write operation to set the internal address into the slave device, and then switch to Master Receiver mode. Note that the second start condition (after sending the IADR) is sometimes called "repeated start" (Sr) in I2C fully-compatible devices. See Figure 19-12. See Figure 19-11 and Figure 19-13 for Master Write operation with internal address. The three internal address bytes are configurable through the Master Mode register (MMR). 226 32059I-06/2010 AT32UC3B If the slave device supports only a 7-bit address, i.e. no internal address, IADRSZ must be set to 0. n the figures below the following abbreviations are used:I *S *Sr *P *W *R *A *N *DADR *IADR Start Repeated Start Stop Write Read Acknowledge Not Acknowledge Device Address Internal Address Figure 19-11. Master Write with One, Two or Three Bytes Internal Address and One Data Byte Three bytes internal address TWD S DADR W A IADR(23:16) A IADR(15:8) A IADR(7:0) A DATA A P Two bytes internal address TWD S DADR W A IADR(15:8) A IADR(7:0) A DATA A P One byte internal address TWD S DADR W A IADR(7:0) A DATA A P Figure 19-12. Master Read with One, Two or Three Bytes Internal Address and One Data Byte Three bytes internal address TWD S DADR W A IADR(23:16) A IADR(15:8) A IADR(7:0) A Sr DADR R A DATA Two bytes internal address TWD S DADR W A IADR(15:8) A IADR(7:0) A Sr DADR R A DATA N P N P One byte internal address TWD S DADR W A IADR(7:0) A Sr DADR R A DATA N P 227 32059I-06/2010 AT32UC3B 19.10.6.2 10-bit Slave Addressing For a slave address higher than 7 bits, the user must configure the address size (IADRSZ) and set the other slave address bits in the internal address register (IADR). The two remaining Internal address bytes, IADR[15:8] and IADR[23:16] can be used the same as in 7-bit Slave Addressing. Example: Address a 10-bit device: (10-bit device address is b1 b2 b3 b4 b5 b6 b7 b8 b9 b10) 1. Program IADRSZ = 1, 2. Program DADR with 1 1 1 1 0 b1 b2 (b1 is the MSB of the 10-bit address, b2, etc.) 3. Program IADR with b3 b4 b5 b6 b7 b8 b9 b10 (b10 is the LSB of the 10-bit address) Figure 19-13 below shows a byte write to an Atmel AT24LC512 EEPROM. This demonstrates the use of internal addresses to access the device. Figure 19-13. Internal Address Usage S T A R T W R I T E Device Address 0 M S B FIRST WORD ADDRESS SECOND WORD ADDRESS DATA S T O P LRA S/C BW K M S B A C K LA SC BK A C K 19.11 Using the Peripheral DMA Controller The use of the Peripheral DMA Controller significantly reduces the CPU load. To assure correct implementation, respect the following programming sequences: 19.11.1 Data Transmit with the Peripheral DMA Controller 1. Initialize the Peripheral DMA Controller TX channel (memory pointers, size, etc.). 2. Configure the master mode (DADR, CKDIV, etc.). 3. Start the transfer by setting the Peripheral DMA Controller TXEN bit. 4. Wait for the Peripheral DMA Controller end TX flag. 5. Disable the Peripheral DMA Controller by setting the Peripheral DMA Controller TXDIS bit. 19.11.2 Data Receive with the Peripheral DMA Controller 1. Initialize the Peripheral DMA Controller TX channel (memory pointers, size, etc.). 2. Configure the master mode (DADR, CKDIV, etc.). 3. Start the transfer by setting the Peripheral DMA Controller RXEN bit. 4. Wait for the Peripheral DMA Controller end RX flag. 228 32059I-06/2010 AT32UC3B 5. Disable the Peripheral DMA Controller by setting the Peripheral DMA Controller RXDIS bit. 229 32059I-06/2010 AT32UC3B 19.11.3 Read-write Flowcharts The following flowcharts shown in Figure 19-14 to Figure 19-19 on page 235 give examples for read and write operations. A polling or interrupt method can be used to check the status bits. The interrupt method requires that the interrupt enable register (IER) be configured first. Figure 19-14. TWI Write Operation with Single Data Byte without Internal Address. BEGIN Set TWI clock (CLDIV, CHDIV, CKDIV) in TWI_CWGR (Needed only once) Set the Control register: - Master enable TWI_CR = MSEN + SVDIS Set the Master Mode register: - Device slave address (DADR) - Transfer direction bit Write ==> bit MREAD = 0 Load Transmit register TWI_THR = Data to send Read Status register No TXRDY = 1? Yes Read Status register No TXCOMP = 1? Yes Transfer finished 230 32059I-06/2010 AT32UC3B Figure 19-15. TWI Write Operation with Single Data Byte and Internal Address BEGIN Set TWI clock (CLDIV, CHDIV, CKDIV) in TWI_CWGR (Needed only once) Set the Control register: - Master enable TWI_CR = MSEN + SVDIS Set the Master Mode register: - Device slave address (DADR) - Internal address size (IADRSZ) - Transfer direction bit Write ==> bit MREAD = 0 Set the internal address TWI_IADR = address Load transmit register TWI_THR = Data to send Read Status register No TXRDY = 1? Yes Read Status register TXCOMP = 1? No Yes Transfer finished 231 32059I-06/2010 AT32UC3B Figure 19-16. TWI Write Operation with Multiple Data Bytes with or without Internal Address BEGIN Set TWI clock (CLDIV, CHDIV, CKDIV) in TWI_CWGR (Needed only once) Set the Control register: - Master enable TWI_CR = MSEN + SVDIS Set the Master Mode register: - Device slave address - Internal address size (if IADR used) - Transfer direction bit Write ==> bit MREAD = 0 No Internal address size = 0? Set the internal address TWI_IADR = address Yes Load Transmit register TWI_THR = Data to send Read Status register TWI_THR = data to send TXRDY = 1? Yes Data to send? Yes No Read Status register Yes No TXCOMP = 1? END 232 32059I-06/2010 AT32UC3B Figure 19-17. TWI Read Operation with Single Data Byte without Internal Address BEGIN Set TWI clock (CLDIV, CHDIV, CKDIV) in TWI_CWGR (Needed only once) Set the Control register: - Master enable TWI_CR = MSEN + SVDIS Set the Master Mode register: - Device slave address - Transfer direction bit Read ==> bit MREAD = 1 Start the transfer TWI_CR = START | STOP Read status register RXRDY = 1? Yes Read Receive Holding Register No Read Status register No TXCOMP = 1? Yes END 233 32059I-06/2010 AT32UC3B Figure 19-18. TWI Read Operation with Single Data Byte and Internal Address BEGIN Set TWI clock (CLDIV, CHDIV, CKDIV) in TWI_CWGR (Needed only once) Set the Control register: - Master enable TWI_CR = MSEN + SVDIS Set the Master Mode register: - Device slave address - Internal address size (IADRSZ) - Transfer direction bit Read ==> bit MREAD = 1 Set the internal address TWI_IADR = address Start the transfer TWI_CR = START | STOP Read Status register No RXRDY = 1? Yes Read Receive Holding register Read Status register No TXCOMP = 1? Yes END 234 32059I-06/2010 AT32UC3B Figure 19-19. TWI Read Operation with Multiple Data Bytes with or without Internal Address BEGIN Set TWI clock (CLDIV, CHDIV, CKDIV) in TWI_CWGR (Needed only once) Set the Control register: - Master enable TWI_CR = MSEN + SVDIS Set the Master Mode register: - Device slave address - Internal address size (if IADR used) - Transfer direction bit Read ==> bit MREAD = 1 Internal address size = 0? Set the internal address TWI_IADR = address Yes Start the transfer TWI_CR = START Read Status register RXRDY = 1? Yes Read Receive Holding register (TWI_RHR) No No Last data to read but one? Yes Stop the transfer TWI_CR = STOP Read Status register No RXRDY = 1? Yes Read Receive Holding register (TWI_RHR) Read status register TXCOMP = 1? Yes END No 235 32059I-06/2010 AT32UC3B 19.12 Multi-master Mode 19.12.1 Definition More than one master may handle the bus at the same time without data corruption by using arbitration. Arbitration starts as soon as two or more masters place information on the bus at the same time, and stops (arbitration is lost) for the master that intends to send a logical one while the other master sends a logical zero. As soon as arbitration is lost by a master, it stops sending data and listens to the bus in order to detect a stop. When the stop is detected, the master who has lost arbitration may put its data on the bus by respecting arbitration. Arbitration is illustrated in Figure 19-21 on page 237. 19.12.2 Different Multi-master Modes Two multi-master modes may be distinguished: 1. TWI is considered as a Master only and will never be addressed. 2. TWI may be either a Master or a Slave and may be addressed. Note: Arbitration is supported in both Multi-master modes. 19.12.2.1 TWI as Master Only In this mode, TWI is considered as a Master only (MSEN is always at one) and must be driven like a Master with the ARBLST (ARBitration Lost) flag in addition. If arbitration is lost (ARBLST = 1), the programmer must reinitiate the data transfer. If the user starts a transfer (ex.: DADR + START + W + Write in THR) and if the bus is busy, the TWI automatically waits for a STOP condition on the bus to initiate the transfer (see Figure 1920 on page 237). Note: The state of the bus (busy or free) is not indicated in the user interface. 19.12.2.2 TWI as Master or Slave The automatic reversal from Master to Slave is not supported in case of a lost arbitration. Then, in the case where TWI may be either a Master or a Slave, the programmer must manage the pseudo Multi-master mode described in the steps below. 1. Program TWI in Slave mode (SADR + MSDIS + SVEN) and perform Slave Access (if TWI is addressed). 2. If TWI has to be set in Master mode, wait until TXCOMP flag is at 1. 3. Program Master mode (DADR + SVDIS + MSEN) and start the transfer (ex: START + Write in THR). 4. As soon as the Master mode is enabled, TWI scans the bus in order to detect if it is busy or free. When the bus is considered as free, TWI initiates the transfer. 5. As soon as the transfer is initiated and until a STOP condition is sent, the arbitration becomes relevant and the user must monitor the ARBLST flag. 6. If the arbitration is lost (ARBLST is set to 1), the user must program the TWI in Slave mode in the case where the Master that won the arbitration wanted to access the TWI. 236 32059I-06/2010 AT32UC3B 7. If TWI has to be set in Slave mode, wait until TXCOMP flag is at 1 and then program the Slave mode. Note: In the case where the arbitration is lost and TWI is addressed, TWI will not acknowledge even if it is programmed in Slave mode as soon as ARBLST is set to 1. Then, the Master must repeat SADR. Figure 19-20. Programmer Sends Data While the Bus is Busy TWCK STOP sent by the master TWD DATA sent by a master Bus is busy Bus is free TWI DATA transfer Transfer is kept START sent by the TWI DATA sent by the TWI A transfer is programmed (DADR + W + START + Write THR) Bus is considered as free Transfer is initiated Figure 19-21. Arbitration Cases TWCK TWD TWCK Data from a Master Data from TWI TWD S S S 1 1 1 0 0 11 0 1 Arbitration is lost TWI stops sending data P S S 1 1 1 0 1 Arbitration is lost The master stops sending data 0 01 0 01 1 1 Data from the TWI 00 11 Data from the master P S ARBLST Bus is busy Bus is free TWI DATA transfer A transfer is programmed (DADR + W + START + Write THR) Transfer is stopped Transfer is kept Transfer is programmed again (DADR + W + START + Write THR) Bus is considered as free Transfer is initiated The flowchart shown in Figure 19-22 on page 238 gives an example of read and write operations in Multi-master mode. 237 32059I-06/2010 AT32UC3B Figure 19-22. Multi-master Flowchart START Programm the SLAVE mode: SADR + MSDIS + SVEN Read Status Register SVACC = 1 ? Yes GACC = 1 ? SVREAD = 0 ? EOSACC = 1 ? Yes TXCOMP = 1 ? Yes Yes TXRDY= 1 ? Yes Write in TWI_THR RXRDY= 0 ? Yes Read TWI_RHR GENERAL CALL TREATMENT Need to perform a master access ? Yes Decoding of the programming sequence Prog seq OK ? Change SADR Program the Master mode DADR + SVDIS + MSEN + CLK + R / W Read Status Register Yes ARBLST = 1 ? Yes Yes RXRDY= 0 ? MREAD = 1 ? TXRDY= 0 ? Yes Read TWI_RHR Yes Data to read? Data to send ? Yes Write in TWI_THR Stop transfer Read Status Register Yes TXCOMP = 0 ? 238 32059I-06/2010 AT32UC3B 19.13 Slave Mode 19.13.1 Definition The Slave Mode is defined as a mode where the device receives the clock and the address from another device called the master. In this mode, the device never initiates and never completes the transmission (START, REPEATED_START and STOP conditions are always provided by the master). 19.13.2 Application Block Diagram Figure 19-23. Slave Mode Typical Application Block Diagram VDD R TWD TWCK R Master Host with TWI Interface Host with TWI Interface Slave 1 Host with TWI Interface Slave 2 LCD Controller Slave 3 19.13.3 Programming Slave Mode The following fields must be programmed before entering Slave mode: 1. SADR (SMR): The slave device address is used in order to be accessed by master devices in read or write mode. 2. MSDIS (CR): Disable the master mode. 3. SVEN (CR): Enable the slave mode. As the device receives the clock, values written in CWGR are not taken into account. 19.13.4 Receiving Data After a Start or Repeated Start condition is detected and if the address sent by the Master matches with the Slave address programmed in the SADR (Slave ADdress) field, SVACC (Slave ACCess) flag is set and SVREAD (Slave READ) indicates the direction of the transfer. SVACC remains high until a STOP condition or a repeated START is detected. When such a condition is detected, EOSACC (End Of Slave ACCess) flag is set. 19.13.4.1 Read Sequence In the case of a Read sequence (SVREAD is high), TWI transfers data written in the THR (TWI Transmit Holding Register) until a STOP condition or a REPEATED_START + an address different from SADR is detected. Note that at the end of the read sequence TXCOMP (Transmission Complete) flag is set and SVACC reset. 239 32059I-06/2010 AT32UC3B As soon as data is written in the THR, TXRDY (Transmit Holding Register Ready) flag is reset, and it is set when the shift register is empty and the sent data acknowledged or not. If the data is not acknowledged, the NACK flag is set. Note that a STOP or a repeated START always follows a NACK. See Figure 19-24 on page 241. 19.13.4.2 Write Sequence In the case of a Write sequence (SVREAD is low), the RXRDY (Receive Holding Register Ready) flag is set as soon as a character has been received in the RHR (TWI Receive Holding Register). RXRDY is reset when reading the RHR. TWI continues receiving data until a STOP condition or a REPEATED_START + an address different from SADR is detected. Note that at the end of the write sequence TXCOMP flag is set and SVACC reset. See Figure 19-25 on page 241. 19.13.4.3 Clock Synchronization Sequence In the case where THR or RHR is not written/read in time, TWI performs a clock synchronization. Clock stretching information is given by the SCLWS (Clock Wait state) bit. See Figure 19-27 on page 243 and Figure 19-28 on page 244. 19.13.4.4 General Call In the case where a GENERAL CALL is performed, GACC (General Call ACCess) flag is set. After GACC is set, it is up to the programmer to interpret the meaning of the GENERAL CALL and to decode the new address programming sequence. See Figure 19-26 on page 242. 19.13.4.5 Peripheral DMA Controller As it is impossible to know the exact number of data to receive/send, the use of Peripheral DMA Controller is NOT recommended in SLAVE mode. 19.13.5 19.13.5.1 Data Transfer Read Operation The read mode is defined as a data requirement from the master. After a START or a REPEATED START condition is detected, the decoding of the address starts. If the slave address (SADR) is decoded, SVACC is set and SVREAD indicates the direction of the transfer. Until a STOP or REPEATED START condition is detected, TWI continues sending data loaded in the THR register. If a STOP condition or a REPEATED START + an address different from SADR is detected, SVACC is reset. Figure 19-24 on page 241 describes the write operation. 240 32059I-06/2010 AT32UC3B Figure 19-24. Read Access Ordered by a MASTER SADR does not match, TWI answers with a NACK SADR matches, TWI answers with an ACK ACK/NACK from the Master A DATA NA S/Sr TWD TXRDY NACK SVACC SVREAD EOSVACC S ADR R NA DATA NA P/S/Sr SADR R A DATA A Write THR Read RHR SVREAD has to be taken into account only while SVACC is active Notes: 1. When SVACC is low, the state of SVREAD becomes irrelevant. 2. TXRDY is reset when data has been transmitted from THR to the shift register and set when this data has been acknowledged or non acknowledged. 19.13.5.2 Write Operation The write mode is defined as a data transmission from the master. After a START or a REPEATED START, the decoding of the address starts. If the slave address is decoded, SVACC is set and SVREAD indicates the direction of the transfer (SVREAD is low in this case). Until a STOP or REPEATED START condition is detected, TWI stores the received data in the RHR register. If a STOP condition or a REPEATED START + an address different from SADR is detected, SVACC is reset. Figure 19-25 on page 241 describes the Write operation. Figure 19-25. Write Access Ordered by a Master SADR does not match, TWI answers with a NACK SADR matches, TWI answers with an ACK Read RHR TWD RXRDY SVACC SVREAD EOSVACC S ADR W NA DATA NA P/S/Sr SADR W A DATA A A DATA NA S/Sr SVREAD has to be taken into account only while SVACC is active Notes: 1. When SVACC is low, the state of SVREAD becomes irrelevant. 2. RXRDY is set when data has been transmitted from the shift register to the RHR and reset when this data is read. 241 32059I-06/2010 AT32UC3B 19.13.5.3 General Call The general call is performed in order to change the address of the slave. If a GENERAL CALL is detected, GACC is set. After the detection of General Call, it is up to the programmer to decode the commands which come afterwards. In case of a WRITE command, the programmer has to decode the programming sequence and program a new SADR if the programming sequence matches. Figure 19-26 on page 242 describes the General Call access. Figure 19-26. Master Performs a General Call 0000000 + W RESET command = 00000110X WRITE command = 00000100X A TXD S GENERAL CALL A Reset or write DADD A DATA1 A DATA2 A New SADR P New SADR Programming sequence GCACC Reset after read SVACC Note: 1. This method allows the user to create an own programming sequence by choosing the programming bytes and the number of them. The programming sequence has to be provided to the master. 242 32059I-06/2010 AT32UC3B 19.13.6 Clock Synchronization In both read and write modes, it may happen that THR/RHR buffer is not filled /emptied before the emission/reception of a new character. In this case, to avoid sending/receiving undesired data, a clock stretching mechanism is implemented. 19.13.6.1 Clock Synchronization in Read Mode The clock is tied low if the shift register is empty and if a STOP or REPEATED START condition was not detected. It is tied low until the shift register is loaded. Figure 19-27 on page 243 describes the clock synchronization in Read mode. Figure 19-27. Clock Synchronization in Read Mode TWI_THR DATA0 1 DATA1 DATA2 S SADR R A DATA0 A DATA1 A XXXXXXX 2 DATA2 NA S TWCK Write THR CLOCK is tied low by the TWI as long as THR is empty SCLWS TXRDY SVACC SVREAD TXCOMP As soon as a START is detected TWI_THR is transmitted to the shift register 1 2 The data is memorized in TWI_THR until a new value is written Ack or Nack from the master The clock is stretched after the ACK, the state of TWD is undefined during clock stretching Notes: 1. TXRDY is reset when data has been written in the TH to the shift register and set when this data has been acknowledged or non acknowledged. 2. At the end of the read sequence, TXCOMP is set after a STOP or after a REPEATED_START + an address different from SADR. 3. SCLWS is automatically set when the clock synchronization mechanism is started. 243 32059I-06/2010 AT32UC3B 19.13.6.2 Clock Synchronization in Write Mode The clock is tied low if the shift register and the RHR is full. If a STOP or REPEATED_START condition was not detected, it is tied low until RHR is read. Figure 19-28 on page 244 describes the clock synchronization in Read mode. Figure 19-28. Clock Synchronization in Write Mode TWCK CLOCK is tied low by the TWI as long as RHR is full TWD S SADR W A DATA0 A DATA1 A DATA2 NA S ADR TWI_RHR SCLWS DATA0 is not read in the RHR DATA1 DATA2 SCL is stretched on the last bit of DATA1 RXRDY Rd DATA0 SVACC SVREAD TXCOMP As soon as a START is detected Rd DATA1 Rd DATA2 Notes: 1. At the end of the read sequence, TXCOMP is set after a STOP or after a REPEATED_START + an address different from SADR. 2. SCLWS is automatically set when the clock synchronization mechanism is started and automatically reset when the mechanism is finished. 244 32059I-06/2010 AT32UC3B 19.13.7 19.13.7.1 Reversal after a Repeated Start Reversal of Read to Write The master initiates the communication by a read command and finishes it by a write command. Figure 19-29 on page 245 describes the repeated start + reversal from Read to Write mode. Figure 19-29. Repeated Start + Reversal from Read to Write Mode TWI_THR DATA0 DATA1 TWD S SADR R A DATA0 A DATA1 NA Sr SADR W A DATA2 A DATA3 A DATA3 P TWI_RHR SVACC SVREAD TXRDY RXRDY EOSACC TXCOMP As soon as a START is detected DATA2 Cleared after read Note: 1. TXCOMP is only set at the end of the transmission because after the repeated start, SADR is detected again. 19.13.7.2 Reversal of Write to Read The master initiates the communication by a write command and finishes it by a read command.Figure 19-30 on page 245 describes the repeated start + reversal from Write to Read mode. Figure 19-30. Repeated Start + Reversal from Write to Read Mode TWI_THR DATA2 DATA3 TWD TWI_RHR SVACC SVREAD TXRDY RXRDY EOSACC TXCOMP S SADR W A DATA0 A DATA1 A Sr SADR R A DATA2 A DATA3 NA P DATA0 DATA1 Read TWI_RHR As soon as a START is detected Cleared after read 245 32059I-06/2010 AT32UC3B Notes: 1. In this case, if THR has not been written at the end of the read command, the clock is automatically stretched before the ACK. 2. TXCOMP is only set at the end of the transmission because after the repeated start, SADR is detected again. 19.13.8 Read Write Flowcharts The flowchart shown in Figure 19-31 on page 246 gives an example of read and write operations in Slave mode. A polling or interrupt method can be used to check the status bits. The interrupt method requires that the interrupt enable register (IER) be configured first. Figure 19-31. Read Write Flowchart in Slave Mode Set the SLAVE mode: SADR + MSDIS + SVEN Read Status Register SVACC = 1 ? GACC = 1 ? SVREAD = 0 ? EOSACC = 1 ? TXRDY= 1 ? Write in TWI_THR TXCOMP = 1 ? RXRDY= 0 ? END Read TWI_RHR GENERAL CALL TREATMENT Decoding of the programming sequence Prog seq OK ? Change SADR 246 32059I-06/2010 AT32UC3B 19.14 User Interface Table 19-4. Offset 0x00 0x04 0x08 0x0C 0x10 0x20 0x24 0x28 0x2C 0x30 0x34 TWI User Interface Register Control Register Master Mode Register Slave Mode Register Internal Address Register Clock Waveform Generator Register Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Receive Holding Register Transmit Holding Register Register Name CR MMR SMR IADR CWGR SR IER IDR IMR RHR THR Access Write-only Read/Write Read/Write Read/Write Read/Write Read-only Write-only Write-only Read-only Read-only Write-only Reset N/A 0x00000000 0x00000000 0x00000000 0x00000000 0x0000F009 N/A N/A 0x00000000 0x00000000 0x00000000 247 32059I-06/2010 AT32UC3B 19.14.1 Name: Access: Offset: Control Register CR Write-only 0x00 Reset Value: 0x00000000 31 - 23 - 15 - 7 SWRST 30 - 22 - 14 - 6 - 29 - 21 - 13 - 5 SVDIS 28 - 20 - 12 - 4 SVEN 27 - 19 - 11 - 3 MSDIS 26 - 18 - 10 - 2 MSEN 25 - 17 - 9 - 1 STOP 24 - 16 - 8 - 0 START * SWRST: Software Reset 0 = No effect. 1 = Equivalent to a system reset. SVDIS: TWI Slave Mode Disabled 0 = No effect. 1 = The slave mode is disabled. The shifter and holding characters (if it contains data) are transmitted in case of read operation. In write operation, the character being transferred must be completely received before disabling. SVEN: TWI Slave Mode Enabled 0 = No effect. 1 = If SVDIS = 0, the slave mode is enabled. Switching from Master to Slave mode is only permitted when TXCOMP = 1. MSDIS: TWI Master Mode Disabled 0 = No effect. 1 = The master mode is disabled, all pending data is transmitted. The shifter and holding characters (if it contains data) are transmitted in case of write operation. In read operation, the character being transferred must be completely received before disabling. MSEN: TWI Master Mode Enabled 0 = No effect. 1 = If MSDIS = 0, the master mode is enabled. Switching from Slave to Master mode is only permitted when TXCOMP = 1. STOP: Send a STOP Condition 0 = No effect. 1 = STOP Condition is sent just after completing the current byte transmission in master read mode. - In single data byte master read, the START and STOP must both be set. - In multiple data bytes master read, the STOP must be set after the last data received but one. - In master read mode, if a NACK bit is received, the STOP is automatically performed. - In multiple data write operation, when both THR and shift register are empty, a STOP condition is automatically sent. * * * * * * START: Send a START Condition 0 = No effect. 248 32059I-06/2010 AT32UC3B 1 = A frame beginning with a START bit is transmitted according to the features defined in the mode register. This action is necessary when the TWI peripheral wants to read data from a slave. When configured in Master Mode with a write operation, a frame is sent as soon as the user writes a character in the Transmit Holding Register (THR). 249 32059I-06/2010 AT32UC3B 19.14.2 Name: Access: Offset: Master Mode Register MMR Read-write 0x04 Reset Value: 0x00000000 31 - 23 - 15 - 7 - 30 - 22 29 - 21 28 - 20 27 - 19 DADR 11 - 3 - 26 - 18 25 - 17 24 - 16 14 - 6 - 13 - 5 - 12 MREAD 4 - 10 - 2 - 9 IADRSZ 1 - 8 0 - * DADR: Device Address The device address is used to access slave devices in read or write mode. Those bits are only used in Master mode. * MREAD: Master Read Direction 0 = Master write direction. 1 = Master read direction. * IADRSZ: Internal Device Address Size IADRSZ[9:8] 0 0 1 1 0 1 0 1 Description No internal device address One-byte internal device address Two-byte internal device address Three-byte internal device address 250 32059I-06/2010 AT32UC3B 19.14.3 Name: Access: Offset: Slave Mode Register SMR Read-write 0x08 Reset Value: 0x00000000 31 - 23 - 15 - 7 - 30 - 22 29 - 21 28 - 20 27 - 19 SADR 11 - 3 - 26 - 18 25 - 17 24 - 16 14 - 6 - 13 - 5 - 12 - 4 - 10 - 2 - 9 8 1 - 0 - * SADR: Slave Address The slave device address is used in Slave mode in order to be accessed by master devices in read or write mode. SADR must be programmed before enabling the Slave mode or after a general call. Writes at other times have no effect. 251 32059I-06/2010 AT32UC3B 19.14.4 Name: Access: Offset: Internal Address Register IADR Read-write 0x0C Reset Value: 0x00000000 31 - 23 30 - 22 29 - 21 28 - 20 IADR 15 14 13 12 IADR 7 6 5 4 IADR 3 2 1 0 11 10 9 8 27 - 19 26 - 18 25 - 17 24 - 16 * IADR: Internal Address 0, 1, 2 or 3 bytes depending on IADRSZ. 252 32059I-06/2010 AT32UC3B 19.14.5 Name: Access: Offset: Clock Waveform Generator Register CWGR Read-write 0x10 Reset Value: 0x00000000 31 - 23 30 - 22 29 - 21 28 - 20 27 - 19 26 - 18 25 - 17 CKDIV 9 24 - 16 15 14 13 12 CHDIV 11 10 8 7 6 5 4 CLDIV 3 2 1 0 CWGR is only used in Master mode. * CKDIV: Clock Divider The CKDIV is used to increase both SCL high and low periods. * CHDIV: Clock High Divider The SCL high period is defined as follows: T high = ( ( CHDIV x 2 CKDIV ) + 4 ) x T MCK * CLDIV: Clock Low Divider The SCL low period is defined as follows: T low = ( ( CLDIV x 2 CKDIV ) + 4 ) x T MCK 253 32059I-06/2010 AT32UC3B 19.14.6 Name: Access: Offset: Status Register SR Read-only 0x20 Reset Value: 0x0000F009 31 - 23 - 15 TXBUFE 7 - 30 - 22 - 14 RXBUFF 6 OVRE 29 - 21 - 13 ENDTX 5 GACC 28 - 20 - 12 ENDRX 4 SVACC 27 - 19 - 11 EOSACC 3 SVREAD 26 - 18 - 10 SCLWS 2 TXRDY 25 - 17 - 9 ARBLST 1 RXRDY 24 - 16 - 8 NACK 0 TXCOMP * TXBUFE: TX Buffer Empty This bit is only used in Master mode. 0 = TCR or TNCR have a value other than 0. 1 = Both TCR and TNCR have a value of 0. RXBUFF: RX Buffer Full This bit is only used in Master mode. 0 = RCR or RNCR have a value other than 0. 1 = Both RCR and RNCR have a value of 0. ENDTX: End of TX buffer This bit is only used in Master mode. 0 = The Transmit Counter Register has not reached 0 since the last write in TCR or TNCR. 1 = The Transmit Counter Register has reached 0 since the last write in TCR or TNCR. ENDRX: End of RX buffer This bit is only used in Master mode. 0 = The Receive Counter Register has not reached 0 since the last write in RCR or RNCR. 1 = The Receive Counter Register has reached 0 since the last write in RCR or RNCR. EOSACC: End Of Slave Access (clear on read) This bit is only used in Slave mode. 0 = A slave access is being performing. 1 = The Slave Access is finished. End Of Slave Access is automatically set as soon as SVACC is reset. EOSACC behavior can be seen in Figure 19-29 on page 245 and Figure 19-30 on page 245 SCLWS: Clock Wait State (automatically set / reset) This bit is only used in Slave mode. 0 = The clock is not stretched. 1 = The clock is stretched. THR / RHR buffer is not filled / emptied before the emission / reception of a new character. SCLWS behavior can be seen in Figure 19-27 on page 243 and Figure 19-28 on page 244. ARBLST: Arbitration Lost (clear on read) This bit is only used in Master mode. 0 = Arbitration won. 1 = Arbitration lost. Another master of the TWI bus has won the multi-master arbitration. TXCOMP is set at the same time. NACK: Not Acknowledged (clear on read) * * * * * * * 254 32059I-06/2010 AT32UC3B NACK used in Master mode: 0 = Each data byte has been correctly received by the far-end side TWI slave component. 1 = A data byte has not been acknowledged by the slave component. Set at the same time as TXCOMP. NACK used in Slave Read mode: 0 = Each data byte has been correctly received by the Master. 1 = In read mode, a data byte has not been acknowledged by the Master. When NACK is set the programmer must not fill THR even if TXRDY is set, because it means that the Master will stop the data transfer or re initiate it. Note that in Slave Write mode all data are acknowledged by the TWI. * OVRE: Overrun Error (clear on read) This bit is only used in Master mode. 0 = RHR has not been loaded while RXRDY was set 1 = RHR has been loaded while RXRDY was set. Reset by read in SR when TXCOMP is set. GACC: General Call Access (clear on read) This bit is only used in Slave mode. 0 = No General Call has been detected. 1 = A General Call has been detected. After the detection of General Call, the programmer decoded the commands that follow and the programming sequence. GACC behavior can be seen in Figure 19-26 on page 242. SVACC: Slave Access (automatically set / reset) This bit is only used in Slave mode. 0 = TWI is not addressed. SVACC is automatically cleared after a NACK or a STOP condition is detected. 1 = Indicates that the address decoding sequence has matched (A Master has sent SADR). SVACC remains high until a NACK or a STOP condition is detected. SVACC behavior can be seen in Figure 19-24 on page 241, Figure 19-25 on page 241, Figure 19-29 on page 245 and Figure 19-30 on page 245. SVREAD: Slave Read (automatically set / reset) This bit is only used in Slave mode. When SVACC is low (no Slave access has been detected) SVREAD is irrelevant. 0 = Indicates that a write access is performed by a Master. 1 = Indicates that a read access is performed by a Master. SVREAD behavior can be seen in Figure 19-24 on page 241, Figure 19-25 on page 241, Figure 19-29 on page 245 and Figure 19-30 on page 245. TXRDY: Transmit Holding Register Ready (automatically set / reset) TXRDY used in Master mode: 0 = The transmit holding register has not been transferred into shift register. Set to 0 when writing into THR register. 1 = As soon as a data byte is transferred from THR to internal shifter or if a NACK error is detected, TXRDY is set at the same time as TXCOMP and NACK. TXRDY is also set when MSEN is set (enable TWI). TXRDY behavior in Master mode can be seen in Figure 19-8 on page 225. TXRDY used in Slave mode: 0 = As soon as data is written in the THR, until this data has been transmitted and acknowledged (ACK or NACK). 1 = It indicates that the THR is empty and that data has been transmitted and acknowledged. If TXRDY is high and if a NACK has been detected, the transmission will be stopped. Thus when TRDY = NACK = 1, the programmer must not fill THR to avoid losing it. TXRDY behavior in Slave mode can be seen in Figure 19-24 on page 241, Figure 19-27 on page 243, Figure 19-29 on page 245 and Figure 19-30 on page 245. RXRDY: Receive Holding Register Ready (automatically set / reset) 0 = No character has been received since the last RHR read operation. 1 = A byte has been received in the RHR since the last read. RXRDY behavior in Master mode can be seen in Figure 19-10 on page 226. RXRDY behavior in Slave mode can be seen in Figure 19-25 on page 241, Figure 19-28 on page 244, Figure 19-29 on page 245 and Figure 19-30 on page 245. * * * * * 255 32059I-06/2010 AT32UC3B * TXCOMP: Transmission Completed (automatically set / reset) TXCOMP used in Master mode: 0 = During the length of the current frame. 1 = When both holding and shifter registers are empty and STOP condition has been sent. TXCOMP behavior in Master mode can be seen in Figure 19-8 on page 225 and in Figure 19-10 on page 226. TXCOMP used in Slave mode: 0 = As soon as a Start is detected. 1 = After a Stop or a Repeated Start + an address different from SADR is detected. TXCOMP behavior in Slave mode can be seen in Figure 19-27 on page 243, Figure 19-28 on page 244, Figure 19-29 on page 245 and Figure 19-30 on page 245. 256 32059I-06/2010 AT32UC3B 19.14.7 Name: Access: Offset: Interrupt Enable Register IER Write-only 0x24 Reset Value: 0x00000000 31 - 23 - 15 TXBUFE 7 - 30 - 22 - 14 RXBUFF 6 OVRE 29 - 21 - 13 ENDTX 5 GACC 28 - 20 - 12 ENDRX 4 SVACC 27 - 19 - 11 EOSACC 3 - 26 - 18 - 10 SCL_WS 2 TXRDY 25 - 17 - 9 ARBLST 1 RXRDY 24 - 16 - 8 NACK 0 TXCOMP Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. 257 32059I-06/2010 AT32UC3B 19.14.8 Name: Access: Offset: Interrupt Disable Register IDR Write-only 0x28 Reset Value: 0x00000000 31 - 23 - 15 TXBUFE 7 - 30 - 22 - 14 RXBUFF 6 OVRE 29 - 21 - 13 ENDTX 5 GACC 28 - 20 - 12 ENDRX 4 SVACC 27 - 19 - 11 EOSACC 3 - 26 - 18 - 10 SCL_WS 2 TXRDY 25 - 17 - 9 ARBLST 1 RXRDY 24 - 16 - 8 NACK 0 TXCOMP Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. 258 32059I-06/2010 AT32UC3B 19.14.9 Name: Access: Offset: Interrupt Mask Register IMR Read-only 0x2C Reset Value: 0x00000000 31 - 23 - 15 TXBUFE 7 - 30 - 22 - 14 RXBUFF 6 OVRE 29 - 21 - 13 ENDTX 5 GACC 28 - 20 - 12 ENDRX 4 SVACC 27 - 19 - 11 EOSACC 3 - 26 - 18 - 10 SCL_WS 2 TXRDY 25 - 17 - 9 ARBLST 1 RXRDY 24 - 16 - 8 NACK 0 TXCOMP 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is set when the corresponding bit in IER is written to one. 259 32059I-06/2010 AT32UC3B 19.14.10 Receive Holding Register Name: Access: Offset: RHR Read-only 0x30 Reset Value: 0x00000000 31 - 23 - 15 - 7 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 RXDATA 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 - 1 24 - 16 - 8 - 0 * RXDATA: Master or Slave Receive Holding Data 260 32059I-06/2010 AT32UC3B 19.14.11 Transmit Holding Register Name: Access: Offset: THR Read-write 0x34 Reset Value: 0x00000000 31 - 23 - 15 - 7 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 TXDATA 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 - 1 24 - 16 - 8 - 0 * TXDATA: Master or Slave Transmit Holding Data 261 32059I-06/2010 AT32UC3B 20. Synchronous Serial Controller (SSC) Rev: 3.1.0.2 20.1 Features * * * * * Provides serial synchronous communication links used in audio and telecom applications Independent receiver and transmitter, common clock divider Interfaced with two Peripheral DMA Controller channels to reduce processor overhead Configurable frame sync and data length Receiver and transmitter can be configured to start automatically or on detection of different events on the frame sync signal * Receiver and transmitter include a data signal, a clock signal and a frame synchronization signal 20.2 Overview The Synchronous Serial Controller (SSC) provides a synchronous communication link with external devices. It supports many serial synchronous communication protocols generally used in audio and telecom applications such as I2S, Short Frame Sync, Long Frame Sync, etc. The SSC consists of a receiver, a transmitter, and a common clock divider. Both the receiver and the transmitter interface with three signals: * the TX_DATA/RX_DATA signal for data * the TX_CLOCK/RX_CLOCK signal for the clock * the TX_FRAME_SYNC/RX_FRAME_SYNC signal for the frame synchronization The transfers can be programmed to start automatically or on different events detected on the Frame Sync signal. The SSC's high-level of programmability and its two dedicated Peripheral DMA Controller channels of up to 32 bits permit a continuous high bit rate data transfer without processor intervention. Featuring connection to two Peripheral DMA Controller channels, the SSC permits interfacing with low processor overhead to the following: * CODEC's in master or slave mode * DAC through dedicated serial interface, particularly I2S * Magnetic card reader 262 32059I-06/2010 AT32UC3B 20.3 Block Diagram Figure 20-1. SSC Block Diagram High Speed Bus Peripheral Bus Bridge Peripheral DMA Controller Peripheral Bus TX_FRAME_SYNC TX_CLOCK Power CLK_SSC Manager TX_DATA SSC Interface I/O Controller RX_FRAME_SYNC RX_CLOCK Interrupt Control RX_DATA SSC Interrupt 20.4 Application Block Diagram Figure 20-2. SSC Application Block Diagram Power Management SSC Serial AUDIO Codec Time Slot Frame Management Management Line Interface Interrupt Management Test Management OS or RTOS Driver 263 32059I-06/2010 AT32UC3B 20.5 I/O Lines Description Table 20-1. Pin Name I/O Lines Description Pin Description Receiver Frame Synchro Receiver Clock Receiver Data Transmitter Frame Synchro Transmitter Clock Transmitter Data Type Input/Output Input/Output Input Input/Output Input/Output Output RX_FRAME_SYNC RX_CLOCK RX_DATA TX_FRAME_SYNC TX_CLOCK TX_DATA 20.6 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 20.6.1 I/O Lines The pins used for interfacing the compliant external devices may be multiplexed with I/O lines. Before using the SSC receiver, the I/O Controller must be configured to dedicate the SSC receiver I/O lines to the SSC peripheral mode. Before using the SSC transmitter, the I/O Controller must be configured to dedicate the SSC transmitter I/O lines to the SSC peripheral mode. 20.6.2 Clocks The clock for the SSC bus interface (CLK_SSC) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. It is recommended to disable the SSC before disabling the clock, to avoid freezing the SSC in an undefined state. 20.6.3 Interrupts The SSC interrupt request line is connected to the interrupt controller. Using the SSC interrupt requires the interrupt controller to be programmed first. 20.7 Functional Description This chapter contains the functional description of the following: SSC functional block, clock management, data framing format, start, transmitter, receiver, and frame sync. The receiver and the transmitter operate separately. However, they can work synchronously by programming the receiver to use the transmit clock and/or to start a data transfer when transmission starts. Alternatively, this can be done by programming the transmitter to use the receive clock and/or to start a data transfer when reception starts. The transmitter and the receiver can be programmed to operate with the clock signals provided on either the TX_CLOCK or RX_CLOCK pins. This allows the SSC to support many slave-mode data transfers. The maximum clock speed allowed on the TX_CLOCK and RX_CLOCK pins is CLK_SSC divided by two. 264 32059I-06/2010 AT32UC3B Figure 20-3. SSC Functional Block Diagram Transmitter Clock Output Controller TX_CLOCK TX_CLOCK Input CLK_SSC Clock Divider RX clock TX_FRAME_SYNC RX_FRAME_SYNC Transmit Clock TX clock Controller Frame Sync Controller TX_FRAME_SYNC Start Selector TX_DMA Transmit Shift Register Transmit Holding Register Transmit Sync Holding Register TX_DATA Peripheral Bus User Interface Load Shift Receiver Clock Output Controller RX_CLOCK RX_CLOCK Input TX clock TX_FRAME_SYNC RX_FRAME_SYNC Receive Clock RX clock Controller Frame Sync Controller RX_FRAME_SYNC Start Selector RX_DMA Receive Shift Register Receive Holding Register Receive Sync Holding Register RX_DATA DMA Interrupt Control Load Shift Interrupt Controller 20.7.1 Clock Management The transmitter clock can be generated by: * an external clock received on the TX_CLOCK pin * the receiver clock * the internal clock divider The receiver clock can be generated by: * an external clock received on the RX_CLOCK pin * the transmitter clock * the internal clock divider Furthermore, the transmitter block can generate an external clock on the TX_CLOCK pin, and the receiver block can generate an external clock on the RX_CLOCK pin. This allows the SSC to support many Master and Slave Mode data transfers. 265 32059I-06/2010 AT32UC3B 20.7.1.1 Clock divider Figure 20-4. Divided Clock Block Diagram Clock Divider CMR CLK_SSC Divided Clock /2 12-bit Counter The peripheral clock divider is determined by the 12-bit Clock Divider field (its maximal value is 4095) in the Clock Mode Register (CMR.DIV), allowing a peripheral clock division by up to 8190. The divided clock is provided to both the receiver and transmitter. When this field is written to zero, the clock divider is not used and remains inactive. When CMR.DIV is written to a value equal to or greater than one, the divided clock has a frequency of CLK_SSC divided by two times CMR.DIV. Each level of the divided clock has a duration of the peripheral clock multiplied by CMR.DIV. This ensures a 50% duty cycle for the divided clock regardless of whether the CMR.DIV value is even or odd. Figure 20-5. Divided Clock Generation CLK_SSC Divided Clock DIV = 1 Divided Clock Frequency = CLK_SSC/2 CLK_SSC Divided Clock DIV = 3 Divided Clock Frequency = CLK_SSC/6 Table 20-2. Maximum CLK_SSC / 2 Range of Clock Divider Minimum CLK_SSC / 8190 20.7.1.2 Transmitter clock management The transmitter clock is generated from the receiver clock, the divider clock, or an external clock scanned on the TX_CLOCK pin. The transmitter clock is selected by writing to the Transmit 266 32059I-06/2010 AT32UC3B Clock Selection field in the Transmit Clock Mode Register (TCMR.CKS). The transmit clock can be inverted independently by writing a one to the Transmit Clock Inversion bit in TCMR (TCMR.CKI). The transmitter can also drive the TX_CLOCK pin continuously or be limited to the actual data transfer, depending on the Transmit Clock Output Mode Selection field in the TCMR register (TCMR.CKO). The TCMR.CKI bit has no effect on the clock outputs. Writing 0b10 to the TCMR.CKS field to select TX_CLOCK pin and 0b001 to the TCMR.CKO field to select Continuous Transmit Clock can lead to unpredictable results. Figure 20-6. Transmitter Clock Management TX_CLOCK Tri-state Controller Clock Output MUX Receiver Clock Divider Clock CKO Data Transfer CKS INV MUX Tri-state Controller Transmitter Clock CKI CKG 20.7.1.3 Receiver clock management The receiver clock is generated from the transmitter clock, the divider clock, or an external clock scanned on the RX_CLOCK pin. The receive clock is selected by writing to the Receive Clock Selection field in the Receive Clock Mode Register (RCMR.CKS). The receive clock can be inverted independently by writing a one to the Receive Clock Inversion bit in RCMR (RCMR.CKI). The receiver can also drive the RX_CLOCK pin continuously or be limited to the actual data transfer, depending on the Receive Clock Output Mode Selection field in the RCMR register (RCMR.CKO). The RCMR.CKI bit has no effect on the clock outputs. Writing 0b10 to the RCMR.CKS field to select RX_CLOCK pin and 0b001 to the RCMR.CKO field to select Continuous Receive Clock can lead to unpredictable results. 267 32059I-06/2010 AT32UC3B Figure 20-7. Receiver Clock Management RX_CLOCK MUX Transmitter Clock Tri-state Controller Clock Output Divider Clock CKO Data Transfer CKS INV MUX Tri-state Controller Receiver Clock CKI CKG 20.7.1.4 Serial clock ratio considerations The transmitter and the receiver can be programmed to operate with the clock signals provided on either the TX_CLOCK or RX_CLOCK pins. This allows the SSC to support many slave-mode data transfers. In this case, the maximum clock speed allowed on the RX_CLOCK pin is: - CLK_SSC divided by two if RX_FRAME_SYNC is input. - CLK_SSC divided by three if RX_FRAME_SYNC is output. In addition, the maximum clock speed allowed on the TX_CLOCK pin is: - CLK_SSC divided by six if TX_FRAME_SYNC is input. - CLK_SSC divided by two if TX_FRAME_SYNC is output. 20.7.2 Transmitter Operations A transmitted frame is triggered by a start event and can be followed by synchronization data before data transmission. The start event is configured by writing to the TCMR register. See Section 20.7.4. The frame synchronization is configured by writing to the Transmit Frame Mode Register (TFMR). See Section 20.7.5. To transmit data, the transmitter uses a shift register clocked by the transmitter clock signal and the start mode selected in the TCMR register. Data is written by the user to the Transmit Holding Register (THR) then transferred to the shift register according to the data format selected. When both the THR and the transmit shift registers are empty, the Transmit Empty bit is set in the Status Register (SR.TXEMPTY). When the THR register is transferred in the transmit shift register, the Transmit Ready bit is set in the SR register (SR.TXREADY) and additional data can be loaded in the THR register. 268 32059I-06/2010 AT32UC3B Figure 20-8. Transmitter Block Diagram CR.TXEN SR.TXEN CR.TXDIS TFMR.DATDEF TCMR.STTDLY TFMR.FSDEN TFMR.DATNB TX_DATA 1 TX_FRAME_SYNC RX_FRAME_SYNC Transmitter Clock Start Selector TFMR.MSBF 0 Transmit Shift Register TFMR.FSDEN TCMR.STTDLY TFMR.DATLEN THR 0 1 TSHR TFMR.FSLEN 20.7.3 Receiver Operations A received frame is triggered by a start event and can be followed by synchronization data before data transmission. The start event is configured by writing to the RCMR register. See Section 20.7.4. The frame synchronization is configured by writing to the Receive Frame Mode Register (RFMR). See Section 20.7.5. The receiver uses a shift register clocked by the receiver clock signal and the start mode selected in the RCMR register. The data is transferred from the shift register depending on the data format selected. When the receiver shift register is full, the SSC transfers this data in the Receive Holding Register (RHR), the Receive Ready bit is set in the SR register (SR.RXREADY) and the data can be read in the RHR register. If another transfer occurs before a read of the RHR register, the Receive Overrun bit is set in the SR register (SR.OVRUN) and the receiver shift register is transferred to the RHR register. 269 32059I-06/2010 AT32UC3B Figure 20-9. Receiver Block Diagram RX_CLO CK MUX T ra n sm itte r C lo ck T ri-sta te C o n tro lle r C lo ck O u tp u t D ivid e r C lo ck CKO D a ta T ra n sfe r CKS IN V MUX T ri-sta te C o n tro lle r R e ce ive r C lo ck CKI CKG 20.7.4 Start The transmitter and receiver can both be programmed to start their operations when an event occurs, respectively in the Transmit Start Selection field of the TCMR register (TCMR.START) and in the Receive Start Selection field of the RCMR register (RCMR.START). Under the following conditions the start event is independently programmable: * Continuous: in this case, the transmission starts as soon as a word is written to the THR register and the reception starts as soon as the receiver is enabled * Synchronously with the transmitter/receiver * On detection of a falling/rising edge on TX_FRAME_SYNC/RX_FRAME_SYNC * On detection of a low/high level on TX_FRAME_SYNC/RX_FRAME_SYNC * On detection of a level change or an edge on TX_FRAME_SYNC/RX_FRAME_SYNC A start can be programmed in the same manner on either side of the Transmit/Receive Clock Mode Register (TCMR/RCMR). Thus, the start could be on TX_FRAME_SYNC (transmit) or RX_FRAME_SYNC (receive). Moreover, the receiver can start when data is detected in the bit stream with the compare functions. See Section 20.7.6 for more details on receive compare modes. Detection on TX_FRAME_SYNC input/output is done by the Transmit Frame Sync Output Selection field in the TFMR register (TFMR.FSOS). Similarly, detection on RX_FRAME_SYNC input/output is done by the Receive Frame Output Sync Selection field in the RFMR register (RFMR.FSOS). 270 32059I-06/2010 AT32UC3B Figure 20-10. Transmit Start Mode TX_CLOCK (Input) TX_FRAME_SYNC (Input) TX_DATA (Output) Start= Low Level on TX_FRAME_SYNC TX_DATA (Output) Start= Falling Edge on TX_FRAME_SYNC TX_DATA (Output) Start= High Level on TX_FRAME_SYNC TX_DATA (Output) Start= Rising Edge on TX_FRAME_SYNC TX_DATA (Output) Start= Level Change on TX_FRAME_SYNC TX_DATA (Output) Start= Any Edge on TX_FRAME_SYNC X B0 B1 STTDLY X B0 B1 STTDLY B0 B1 STTDLY B0 B1 STTDLY X X X B0 B1 B0 B1 STTDLY X B0 B1 B0 B1 STTDLY Figure 20-11. Receive Pulse/Edge Start Modes RX_CLOCK RX_FRAME_SYNC (Input) RX_DATA (Input) Start = Low Level on RX_FRAME_SYNC RX_DATA (Input) Start = Falling Edge on RX_FRAME_SYNC RX_DATA (Input) Start = High Level on RX_FRAME_SYNC RX_DATA (Input) Start = Rising Edge on RX_FRAME_SYNC RX_DATA (Input) Start = Level Change on RX_FRAME_SYNC RX_DATA (Input) Start = Any Edge on RX_FRAME_SYNC X X X X STTDLY B0 B1 X STTDLY B0 B1 STTDLY B0 B1 STTDLY B0 B1 X B0 B1 STTDLY B0 B1 B0 B1 STTDLY 271 32059I-06/2010 AT32UC3B 20.7.5 Frame Sync The transmitter and receiver frame synchro pins, TX_FRAME_SYNC and RX_FRAME_SYNC, can be programmed to generate different kinds of frame synchronization signals. The RFMR.FSOS and TFMR.FSOS fields are used to select the required waveform. * Programmable low or high levels during data transfer are supported. * Programmable high levels before the start of data transfers or toggling are also supported. If a pulse waveform is selected, in reception, the Receive Frame Sync Length High Part and the Receive Frame Sync Length fields in the RFMR register (RFMR.FSLENHI and RFMR.FSLEN) define the length of the pulse, from 1 bit time up to 256 bit time. Reception Pulse Length = ((16 x FSLENHI ) + FSLEN + 1) receive clock periods Similarly, in transmission, the Transmit Frame Sync Length High Part and the Transmit Frame Sync Length fields in the TFMR register (TFMR.FSLENHI and TFMR.FSLEN) define the length of the pulse, from 1 bit up to 256 bit time. Transmission Pulse Length = ((16 x FSLENHI ) + FSLEN + 1) transmit clock periods The periodicity of the RX_FRAME_SYNC and TX_FRAME_SYNC pulse outputs can be configured respectively through the Receive Period Divider Selection field in the RCMR register (RCMR.PERIOD) and the Transmit Period Divider Selection field in the TCMR register (TCMR.PERIOD). 20.7.5.1 Frame sync data Frame Sync Data transmits or receives a specific tag during the Frame Sync signal. During the Frame Sync signal, the receiver can sample the RX_DATA line and store the data in the Receive Sync Holding Register (RSHR) and the transmitter can transfer the Transmit Sync Holding Register (TSHR) in the shifter register. The data length to be sampled in reception during the Frame Sync signal shall be written to the RFMR.FSLENHI and RFMR.FSLEN fields. The data length to be shifted out in transmission during the Frame Sync signal shall be written to the TFMR.FSLENHI and TFMR.FSLEN fields. Concerning the Receive Frame Sync Data operation, if the Frame Sync Length is equal to or lower than the delay between the start event and the actual data reception, the data sampling operation is performed in the RSHR through the receive shift register. The Transmit Frame Sync operation is performed by the transmitter only if the Frame Sync Data Enable bit in TFMR register (TFMR.FSDEN) is written to one. If the Frame Sync length is equal to or lower than the delay between the start event and the actual data transmission, the normal transmission has priority and the data contained in the TSHR is transferred in the transmit register, then shifted out. 20.7.5.2 Frame sync edge detection The Frame Sync Edge detection is configured by writing to the Frame Sync Edge Detection bit in the RFMR/TFMR registers (RFMR.FSEDGE and TFMR.FSEDGE). This sets the Receive Sync 272 32059I-06/2010 AT32UC3B and Transmit Sync bits in the SR register (SR.RXSYN and SR.TXSYN) on frame synchro edge detection (signals RX_FRAME_SYNC/TX_FRAME_SYNC). 20.7.6 Receive Compare Modes Figure 20-12. Receive Compare Modes RX_CLOCK RX_DATA (Input) CMP0 CMP1 CMP2 CMP3 Start Ignored B0 B1 B2 {FSLENHI,FSLEN} Up to 256 Bits (4 in This Example) STTDLY DATLEN 20.7.6.1 Compare functions Compare 0 can be one start event of the receiver. In this case, the receiver compares at each new sample the last {RFMR.FSLENHI, RFMR.FSLEN} bits received to the {RFMR.FSLENHI, RFMR.FSLEN} lower bits of the data contained in the Receive Compare 0 Register (RC0R). When this start event is selected, the user can program the receiver to start a new data transfer either by writing a new Compare 0, or by receiving continuously until Compare 1 occurs. This selection is done with the Receive Stop Selection bit in the RCMR register (RCMR.STOP). 20.7.7 Data Framing Format The data framing format of both the transmitter and the receiver are programmable through the TFMR, TCMR, RFMR, and RCMR registers. In either case, the user can independently select: * the event that starts the data transfer (RCMR.START and TCMR.START) * the delay in number of bit periods between the start event and the first data bit (RCMR.STTDLY and TCMR.STTDLY) * the length of the data (RFMR.DATLEN and TFMR.DATLEN) * the number of data to be transferred for each start event (RFMR.DATNB and TFMR.DATLEN) * the length of synchronization transferred for each start event (RFMR.FSLENHI, RFMR.FSLEN, TFMR.FSLENHI, and TFMR.FSLEN) * the bit sense: most or lowest significant bit first (RFMR.MSBF and TFMR.MSBF) Additionally, the transmitter can be used to transfer synchronization and select the level driven on the TX_DATA pin while not in data transfer operation. This is done respectively by writing to the Frame Sync Data Enable and the Data Default Value bits in the TFMR register (TFMR.FSDEN and TFMR.DATDEF). Table 20-3. Transmitter TCMR TCMR TCMR TFMR Data Framing Format Registers Receiver RCMR RCMR RCMR RFMR Bit/Field PERIOD START STTDLY DATNB Up to 255 Up to 16 Length Up to 512 Frame size Start selection Size of transmit start delay Number of words transmitted in frame Comment 273 32059I-06/2010 AT32UC3B Table 20-3. Transmitter TFMR TFMR TFMR TFMR TFMR Data Framing Format Registers Receiver RFMR RFMR RFMR Bit/Field DATLEN {FSLENHI,FSLEN} MSBF FSDEN DATDEF Length Up to 32 Up to 256 Comment Size of word Size of Synchro data register Most significant bit first Enable send TSHR Data default value ended Figure 20-13. Transmit and Receive Frame Format in Edge/Pulse Start Modes Start PERIOD TX_FRAME_SYNC / (1) RX_FRAME_SYNC FSLEN TX_DATA (If FSDEN = 1) TX_DATA (If FSDEN = 0) Sync Data From TSHR Default From DATDEF Data From THR Data From THR Data To RHR DATLEN Data From THR Data From THR Data To RHR DATLEN Default From DATDEF Default From DATDEF Ignored Sync Data Sync Data Start Default From DATDEF RX_DATA Sync Data To RSHR Ignored STTDLY DATNB Note: Example of input on falling edge of TX_FRAME_SYNC/RX_FRAME_SYNC. Figure 20-14. Transmit Frame Format in Continuous Mode Start TX_DATA Data From THR DATLEN Data From THR DATLEN Default Start: 1. TXEMPTY set to one 2. Write into the THR Note: STTDLY is written to zero. In this example, THR is loaded twice. FSDEN value has no effect on the transmission. SyncData cannot be output in continuous mode. 274 32059I-06/2010 AT32UC3B Figure 20-15. Receive Frame Format in Continuous Mode Start = Enable Receiver RX_DATA Data To RHR DATLEN Data To RHR DATLEN Note: STTDLY is written to zero. 20.7.8 Loop Mode The receiver can be programmed to receive transmissions from the transmitter. This is done by writing a one to the Loop Mode bit in RFMR register (RFMR.LOOP). In this case, RX_DATA is connected to TX_DATA, RX_FRAME_SYNC is connected to TX_FRAME_SYNC and RX_CLOCK is connected to TX_CLOCK. 20.7.9 Interrupt Most bits in the SR register have a corresponding bit in interrupt management registers. The SSC can be programmed to generate an interrupt when it detects an event. The interrupt is controlled by writing to the Interrupt Enable Register (IER) and Interrupt Disable Register (IDR). These registers enable and disable, respectively, the corresponding interrupt by setting and clearing the corresponding bit in the Interrupt Mask Register (IMR), which controls the generation of interrupts by asserting the SSC interrupt line connected to the interrupt controller. Figure 20-16. Interrupt Block Diagram IM R IE R Set ID R C le a r T ra n s m itte r TXRDY TXEM PTY TXSYNC In te rru p t C o n tro l S S C In te rru p t R e c e iv e r RXRDY OVRUN RXSYNC 275 32059I-06/2010 AT32UC3B 20.8 SSC Application Examples The SSC can support several serial communication modes used in audio or high speed serial links. Some standard applications are shown in the following figures. All serial link applications supported by the SSC are not listed here. Figure 20-17. Audio Application Block Diagram Clock SCK TX_CLOCK Word Select WS TX_FRAME_SYNC TX_DATA SSC RX_DATA RX_FRAME_SYNC RX_CLOCK Clock SCK Word Select WS Data SD I2S RECEIVER Data SD MSB Left Channel LSB MSB Right Channel Figure 20-18. Codec Application Block Diagram Serial Data Clock (SCLK) TX_CLOCK Frame sync (FSYNC) TX_FRAME_SYNC TX_DATA SSC RX_DATA RX_FRAME_SYNC RX_CLOCK Serial Data In Serial Data Out CODEC Serial Data Clock (SCLK) Frame sync (FSYNC) First Time Slot Dstart Serial Data Out Dend Serial Data In 276 32059I-06/2010 AT32UC3B Figure 20-19. Time Slot Application Block Diagram SCLK TX_CLOCK FSYNC TX_FRAME_SYNC TX_DATA SSC RX_DATA RX_FRAME_SYNC RX_CLOCK Data in Data Out CODEC First Time Slot CODEC Second Time Slot Serial Data Clock (SCLK) Frame sync (FSYNC) Serial Data Out Serial Data In First Time Slot Dstart Second Time Slot Dend 277 32059I-06/2010 AT32UC3B 20.9 User Interface SSC Register Memory Map Register Control Register Clock Mode Register Receive Clock Mode Register Receive Frame Mode Register Transmit Clock Mode Register Transmit Frame Mode Register Receive Holding Register Transmit Holding Register Receive Synchronization Holding Register Transmit Synchronization Holding Register Receive Compare 0 Register Receive Compare 1 Register Status Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Register Name CR CMR RCMR RFMR TCMR TFMR RHR THR RSHR TSHR RC0R RC1R SR IER IDR IMR Access Write-only Read/Write Read/Write Read/Write Read/Write Read/Write Read-only Write-only Read-only Read/Write Read/Write Read/Write Read-only Write-only Write-only Read-only Reset 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x000000CC 0x00000000 0x00000000 0x00000000 Table 20-4. Offset 0x00 0x04 0x10 0x14 0x18 0x1C 0x20 0x24 0x30 0x34 0x38 0x3C 0x40 0x44 0x48 0x4C 278 32059I-06/2010 AT32UC3B 20.9.1 Name: Access Type: Offset: Reset value: Control Register CR Write-only 0x00 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 SWRST 14 - 13 - 12 - 11 - 10 - 9 TXDIS 8 TXEN 7 - 6 - 5 - 4 - 3 - 2 - 1 RXDIS 0 RXEN * SWRST: Software Reset 1: Writing a one to this bit will perform a software reset. This software reset has priority on any other bit in CR. 0: Writing a zero to this bit has no effect. * TXDIS: Transmit Disable 1: Writing a one to this bit will disable the transmission. If a character is currently being transmitted, the disable occurs at the end of the current character transmission. 0: Writing a zero to this bit has no effect. * TXEN: Transmit Enable 1: Writing a one to this bit will enable the transmission if the TXDIS bit is not written to one. 0: Writing a zero to this bit has no effect. * RXDIS: Receive Disable 1: Writing a one to this bit will disable the reception. If a character is currently being received, the disable occurs at the end of current character reception. 0: Writing a zero to this bit has no effect. * RXEN: Receive Enable 1: Writing a one to this bit will enables the reception if the RXDIS bit is not written to one. 0: Writing a zero to this bit has no effect. 279 32059I-06/2010 AT32UC3B 20.9.2 Name: Access Type: Offset: Reset value: Clock Mode Register CMR Read/Write 0x04 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 10 DIV[11:8] 9 8 7 6 5 4 DIV[7:0] 3 2 1 0 * DIV[11:0]: Clock Divider The divided clock equals the CLK_SSC divided by two times DIV. The maximum bit rate is CLK_SSC/2. The minimum bit rate is CLK_SSC/(2 x 4095) = CLK_SSC/8190. The clock divider is not active when DIV equals zero. Divided Clock = CLK_SSC ( DIV x 2) 280 32059I-06/2010 AT32UC3B 20.9.3 Name: Access Type: Offset: Reset value: Receive Clock Mode Register RCMR Read/Write 0x10 0x00000000 31 30 29 28 PERIOD 27 26 25 24 23 22 21 20 STTDLY 19 18 17 16 15 - 14 - 13 - 12 STOP 11 10 START 9 8 7 CKG 6 5 CKI 4 3 CKO 2 1 CKS 0 * PERIOD: Receive Period Divider Selection This field selects the divider to apply to the selected receive clock in order to generate a periodic Frame Sync Signal. If equal to zero, no signal is generated. If not equal to zero, a signal is generated each 2 x (PERIOD+1) receive clock periods. * STTDLY: Receive Start Delay If STTDLY is not zero, a delay of STTDLY clock cycles is inserted between the start event and the actual start of reception. When the receiver is programmed to start synchronously with the transmitter, the delay is also applied. Note: It is very important that STTDLY be written carefully. If STTDLY must be written, it should be done in relation to Receive Sync Data reception. * STOP: Receive Stop Selection 1: After starting a receive with a Compare 0, the receiver operates in a continuous mode until a Compare 1 is detected. 0: After completion of a data transfer when starting with a Compare 0, the receiver stops the data transfer and waits for a new Compare 0. 281 32059I-06/2010 AT32UC3B * START: Receive Start Selection START 0 1 2 3 4 5 6 7 8 Others Receive Start Continuous, as soon as the receiver is enabled, and immediately after the end of transfer of the previous data. Transmit start Detection of a low level on RX_FRAME_SYNC signal Detection of a high level on RX_FRAME_SYNC signal Detection of a falling edge on RX_FRAME_SYNC signal Detection of a rising edge on RX_FRAME_SYNC signal Detection of any level change on RX_FRAME_SYNC signal Detection of any edge on RX_FRAME_SYNC signal Compare 0 Reserved * CKG: Receive Clock Gating Selection CKG 0 1 2 3 Receive Clock Gating None, continuous clock Receive Clock enabled only if RX_FRAME_SYNC is low Receive Clock enabled only if RX_FRAME_SYNC is high Reserved * CKI: Receive Clock Inversion CKI affects only the receive clock and not the output clock signal. 1: The data inputs (Data and Frame Sync signals) are sampled on receive clock rising edge. The Frame Sync signal output is shifted out on receive clock falling edge. 0: The data inputs (Data and Frame Sync signals) are sampled on receive clock falling edge. The Frame Sync signal output is shifted out on receive clock rising edge. * CKO: Receive Clock Output Mode Selection CKO 0 1 2 Others Receive Clock Output Mode None Continuous receive clock Receive clock only during data transfers Reserved RX_CLOCK pin Input-only Output Output * CKS: Receive Clock Selection CKS 0 1 2 3 Selected Receive Clock Divided clock TX_CLOCK clock signal RX_CLOCK pin Reserved 282 32059I-06/2010 AT32UC3B 20.9.4 Name: Access Type: Offset: Reset value: Receive Frame Mode Register RFMR Read/Write 0x14 0x00000000 31 30 FSLENHI 29 28 27 - 26 - 25 - 24 FSEDGE 23 - 22 21 FSOS 20 19 18 FSLEN 17 16 15 - 14 - 13 - 12 - 11 10 DATNB 9 8 7 MSBF 6 - 5 LOOP 4 3 2 DATLEN 1 0 * FSLENHI: Receive Frame Sync Length High Part The four MSB of the FSLEN field. * FSEDGE: Receive Frame Sync Edge Detection Determines which edge on Frame Sync will generate the SR.RXSYN interrupt. FSEDGE 0 1 Frame Sync Edge Detection Positive edge detection Negative edge detection * FSOS: Receive Frame Sync Output Selection FSOS 0 1 2 3 4 5 Others Selected Receive Frame Sync Signal None Negative Pulse Positive Pulse Driven Low during data transfer Driven High during data transfer Toggling at each start of data transfer Reserved RX_FRAME_SYNC Pin Input-only Output Output Output Output Output Undefined * FSLEN: Receive Frame Sync Length This field defines the length of the Receive Frame Sync signal and the number of bits sampled and stored in the RSHR register. When this mode is selected by the RCMR.START field, it also determines the length of the sampled data to be compared to the Compare 0 or Compare 1 register. Note: The four most significant bits for this field are located in the FSLENHI field. 283 32059I-06/2010 AT32UC3B The pulse length is equal to ({FSLENHI,FSLEN} + 1) receive clock periods. Thus, if {FSLENHI,FSLEN} is zero, the Receive Frame Sync signal is generated during one receive clock period. DATNB: Data Number per Frame This field defines the number of data words to be received after each transfer start, which is equal to (DATNB + 1). MSBF: Most Significant Bit First 1: The most significant bit of the data register is sampled first in the bit stream. 0: The lowest significant bit of the data register is sampled first in the bit stream. LOOP: Loop Mode 1: RX_DATA is driven by TX_DATA, RX_FRAME_SYNC is driven by TX_FRAME_SYNC and TX_CLOCK drives RX_CLOCK. 0: Normal operating mode. DATLEN: Data Length The bit stream contains (DATLEN + 1) data bits. This field also defines the transfer size performed by the Peripheral DMA Controller assigned to the receiver. * * * * DATLEN 0 1-7 8-15 Others Transfer Size Forbidden value Data transfer are in bytes Data transfer are in halfwords Data transfer are in words 284 32059I-06/2010 AT32UC3B 20.9.5 Name: Access Type: Offset: Reset value: Transmit Clock Mode Register TCMR Read/Write 0x18 0x00000000 31 30 29 28 PERIOD 27 26 25 24 23 22 21 20 STTDLY 19 18 17 16 15 - 14 - 13 - 12 - 11 10 START 9 8 7 CKG 6 5 CKI 4 3 CKO 2 1 CKS 0 * PERIOD: Transmit Period Divider Selection This field selects the divider to apply to the selected transmit clock in order to generate a periodic Frame Sync Signal. If equal to zero, no signal is generated. If not equal to zero, a signal is generated each 2 x (PERIOD+1) transmit clock periods. * STTDLY: Transmit Start Delay If STTDLY is not zero, a delay of STTDLY clock cycles is inserted between the start event and the actual start of transmission. When the transmitter is programmed to start synchronously with the receiver, the delay is also applied. Note: STTDLY must be written carefully, in relation to Transmit Sync Data transmission. * START: Transmit Start Selection START 0 1 2 3 4 5 6 7 Others Transmit Start Continuous, as soon as a word is written to the THR Register (if Transmit is enabled), and immediately after the end of transfer of the previous data. Receive start Detection of a low level on TX_FRAME_SYNC signal Detection of a high level on TX_FRAME_SYNC signal Detection of a falling edge on TX_FRAME_SYNC signal Detection of a rising edge on TX_FRAME_SYNC signal Detection of any level change on TX_FRAME_SYNC signal Detection of any edge on TX_FRAME_SYNC signal Reserved 285 32059I-06/2010 AT32UC3B * CKG: Transmit Clock Gating Selection CKG 0 1 2 3 Transmit Clock Gating None, continuous clock Transmit Clock enabled only if TX_FRAME_SYNC is low Transmit Clock enabled only if TX_FRAME_SYNC is high Reserved * CKI: Transmit Clock Inversion CKI affects only the Transmit Clock and not the output clock signal. 1: The data outputs (Data and Frame Sync signals) are shifted out on transmit clock rising edge. The Frame sync signal input is sampled on transmit clock falling edge. 0: The data outputs (Data and Frame Sync signals) are shifted out on transmit clock falling edge. The Frame sync signal input is sampled on transmit clock rising edge. * CKO: Transmit Clock Output Mode Selection CKO 0 1 2 Others Transmit Clock Output Mode None Continuous transmit clock Transmit clock only during data transfers Reserved TX_CLOCK pin Input-only Output Output * CKS: Transmit Clock Selection CKS 0 1 2 3 Selected Transmit Clock Divided Clock RX_CLOCK clock signal TX_CLOCK Pin Reserved 286 32059I-06/2010 AT32UC3B 20.9.6 Name: Access Type: Offset: Reset value: Transmit Frame Mode Register TFMR Read/Write 0x1C 0x00000000 31 30 FSLENHI 29 28 27 - 26 - 25 - 24 FSEDGE 23 FSDEN 22 21 FSOS 20 19 18 FSLEN 17 16 15 - 14 - 13 - 12 - 11 10 DATNB 9 8 7 MSBF 6 - 5 DATDEF 4 3 2 DATLEN 1 0 * FSLENHI: Transmit Frame Sync Length High Part The four MSB of the FSLEN field. * FSEDGE: Transmit Frame Sync Edge Detection Determines which edge on Frame Sync will generate the SR.TXSYN interrupt. FSEDGE 0 1 Frame Sync Edge Detection Positive Edge Detection Negative Edge Detection * FSDEN: Transmit Frame Sync Data Enable 1: TSHR value is shifted out during the transmission of the Transmit Frame Sync signal. 0: The TX_DATA line is driven with the default value during the Transmit Frame Sync signal. * FSOS: Transmit Frame Sync Output Selection FSOS 0 1 2 3 4 5 Others Selected Transmit Frame Sync Signal None Negative Pulse Positive Pulse Driven Low during data transfer Driven High during data transfer Toggling at each start of data transfer Reserved TX_FRAME_SYNC Pin Input-only Output Output Output Output Output Undefined * FSLEN: Transmit Frame Sync Length This field defines the length of the Transmit Frame Sync signal and the number of bits shifted out from the TSHR register if TFMR.FSDEN is equal to one. 287 32059I-06/2010 AT32UC3B Note: The four most significant bits for this field are located in the FSLENHI field. The pulse length is equal to ({FSLENHI,FSLEN} + 1) transmit clock periods, i.e., the pulse length can range from 1 to 256 transmit clock periods. If {FSLENHI,FSLEN} is zero, the Transmit Frame Sync signal is generated during one transmit clock period. DATNB: Data Number per Frame This field defines the number of data words to be transferred after each transfer start, which is equal to (DATNB + 1). MSBF: Most Significant Bit First 1: The most significant bit of the data register is shifted out first in the bit stream. 0: The lowest significant bit of the data register is shifted out first in the bit stream. DATDEF: Data Default Value This bit defines the level driven on the TX_DATA pin while out of transmission. Note that if the pin is defined as multi-drive by the I/O Controller, the pin is enabled only if the TX_DATA output is one. 1: The level driven on the TX_DATA pin while out of transmission is one. 0: The level driven on the TX_DATA pin while out of transmission is zero. DATLEN: Data Length The bit stream contains (DATLEN + 1) data bits. This field also defines the transfer size performed by the Peripheral DMA Controller assigned to the transmitter. * * * * DATLEN 0 1-7 8-15 Others Transfer Size Forbidden value (1-bit data length is not supported) Data transfer are in bytes Data transfer are in halfwords Data transfer are in words 288 32059I-06/2010 AT32UC3B 20.9.7 Name: Access Type: Offset: Reset value: Receive Holding Register RHR Read-only 0x20 0x00000000 31 30 29 28 RDAT[31:24] 27 26 25 24 23 22 21 20 RDAT[23:16] 19 18 17 16 15 14 13 12 RDAT[15:8] 11 10 9 8 7 6 5 4 RDAT[7:0] 3 2 1 0 * RDAT: Receive Data Right aligned regardless of the number of data bits defined by the RFMR.DATLEN field. 289 32059I-06/2010 AT32UC3B 20.9.8 Name: Access Type: Offset: Reset value: Transmit Holding Register THR Write-only 0x24 0x00000000 31 30 29 28 TDAT[31:24] 27 26 25 24 23 22 21 20 TDAT[23:16] 19 18 17 16 15 14 13 12 TDAT[15:8] 11 10 9 8 7 6 5 4 TDAT[7:0] 3 2 1 0 * TDAT: Transmit Data Right aligned regardless of the number of data bits defined by the TFMR.DATLEN field. 290 32059I-06/2010 AT32UC3B 20.9.9 Name: Access Type: Offset: Reset value: Receive Synchronization Holding Register RSHR Read-only 0x30 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 RSDAT[15:8] 11 10 9 8 7 6 5 4 RSDAT[7:0] 3 2 1 0 * RSDAT: Receive Synchronization Data 291 32059I-06/2010 AT32UC3B 20.9.10 Name: Access Type: Offset: Reset value: Transmit Synchronization Holding Register TSHR Read/Write 0x34 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 TSDAT[15:8] 11 10 9 8 7 6 5 4 TSDAT[7:0] 3 2 1 0 * TSDAT: Transmit Synchronization Data 292 32059I-06/2010 AT32UC3B 20.9.11 Name: Access Type: Offset: Reset value: Receive Compare 0 Register RC0R Read/Write 0x38 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 CP0[15:8] 11 10 9 8 7 6 5 4 CP0[7:0] 3 2 1 0 * CP0: Receive Compare Data 0 293 32059I-06/2010 AT32UC3B 20.9.12 Name: Access Type: Offset: Reset value: Receive Compare 1 Register RC1R Read/Write 0x3C 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 CP1[[15:8] 11 10 9 8 7 6 5 4 CP1[7:0] 3 2 1 0 * CP1: Receive Compare Data 1 294 32059I-06/2010 AT32UC3B 20.9.13 Name: Access Type: Offset: Reset value: Status Register SR Read-only 0x40 0x000000CC 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 RXEN 16 TXEN 15 - 14 - 13 - 12 - 11 RXSYN 10 TXSYN 9 CP1 8 CP0 7 - 6 - 5 OVRUN 4 RXRDY 3 - 2 - 1 TXEMPTY 0 TXRDY * RXEN: Receive Enable This bit is set when the CR.RXEN bit is written to one. This bit is cleared when no data are being processed and the CR.RXDIS bit has been written to one. * TXEN: Transmit Enable This bit is set when the CR.TXEN bit is written to one. This bit is cleared when no data are being processed and the CR.TXDIS bit has been written to one. * RXSYN: Receive Sync This bit is set when a Receive Sync has occurred. This bit is cleared when the SR register is read. * TXSYN: Transmit Sync This bit is set when a Transmit Sync has occurred. This bit is cleared when the SR register is read. * CP1: Compare 1 This bit is set when compare 1 has occurred. This bit is cleared when the SR register is read. * CP0: Compare 0 This bit is set when compare 0 has occurred. This bit is cleared when the SR register is read. * OVRUN: Receive Overrun This bit is set when data has been loaded in the RHR register while previous data has not yet been read. This bit is cleared when the SR register is read. * RXRDY: Receive Ready This bit is set when data has been received and loaded in the RHR register. This bit is cleared when the RHR register is empty. * TXEMPTY: Transmit Empty This bit is set when the last data written in the THR register has been loaded in the TSR register and last data loaded in the TSR register has been transmitted. 295 32059I-06/2010 AT32UC3B This bit is cleared when data remains in the THR register or is currently transmitted from the TSR register. * TXRDY: Transmit Ready This bit is set when the THR register is empty. This bit is cleared when data has been loaded in the THR register and is waiting to be loaded in the TSR register. 296 32059I-06/2010 AT32UC3B 20.9.14 Name: Access Type: Offset: Reset value: Interrupt Enable Register IER Write-only 0x44 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 RXSYN 10 TXSYN 9 CP1 8 CP0 7 - 6 - 5 OVRUN 4 RXRDY 3 - 2 - 1 TXEMPTY 0 TXRDY Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. 297 32059I-06/2010 AT32UC3B 20.9.15 Name: Access Type: Offset: Reset value: Interrupt Disable Register IDR Write-only 0x48 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 RXSYN 10 TXSYN 9 CP1 8 CP0 7 - 6 - 5 OVRUN 4 RXRDY 3 - 2 - 1 TXEMPTY 0 TXRDY Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. 298 32059I-06/2010 AT32UC3B 20.9.16 Name: Access Type: Offset: Reset value: Interrupt Mask Register IMR Read-only 0x4C 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 RXSYN 10 TXSYN 9 CP1 8 CP0 7 - 6 - 5 OVRUN 4 RXRDY 3 - 2 - 1 TXEMPTY 0 TXRDY 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is set when the corresponding bit in IER is written to one. 299 32059I-06/2010 AT32UC3B 21. Universal Synchronous Asynchronous Receiver Transmitter (USART) Rev.4.0.0.5 21.1 Features * Programmable Baud Rate Generator * 5- to 9-bit Full-duplex Synchronous or Asynchronous Serial Communications - 1, 1.5 or 2 Stop Bits in Asynchronous Mode or 1 or 2 Stop Bits in Synchronous Mode - Parity Generation and Error Detection - Framing Error Detection, Overrun Error Detection - MSB- or LSB-first - Optional Break Generation and Detection - By 8 or by 16 Over-sampling Receiver Frequency - Optional Hardware Handshaking RTS-CTS - Optional Modem Signal Management DTR-DSR-DCD-RI - Receiver Time-out and Transmitter Timeguard - Optional Multidrop Mode with Address Generation and Detection RS485 with Driver Control Signal ISO7816, T = 0 or T = 1 Protocols for Interfacing with Smart Cards - NACK Handling, Error Counter with Repetition and Iteration Limit IrDA Modulation and Demodulation - Communication at up to 115.2 Kbps SPI Mode - Master or Slave - Serial Clock Programmable Phase and Polarity - SPI Serial Clock (CLK) Frequency up to Internal Clock Frequency CLK_USART/4 Test Modes - Remote Loopback, Local Loopback, Automatic Echo Supports Connection of Two Peripheral DMA Controller Channels (PDCA) - Offers Buffer Transfer without Processor Intervention * * * * * * 21.2 Overview The Universal Synchronous Asynchronous Receiver Transceiver (USART) provides one full duplex universal synchronous asynchronous serial link. Data frame format is widely programmable (data length, parity, number of stop bits) to support a maximum of standards. The receiver implements parity error, framing error and overrun error detection. The receiver time-out enables handling variable-length frames and the transmitter timeguard facilitates communications with slow remote devices. Multidrop communications are also supported through address bit handling in reception and transmission. The USART features three test modes: remote loopback, local loopback and automatic echo. The USART supports specific operating modes providing interfaces on RS485 and SPI buses, with ISO7816 T = 0 or T = 1 smart card slots, infrared transceivers and connection to modem ports. The hardware handshaking feature enables an out-of-band flow control by automatic management of the pins RTS and CTS. The USART supports the connection to the Peripheral DMA Controller, which enables data transfers to the transmitter and from the receiver. The Peripheral DMA Controller provides chained buffer management without any intervention of the processor. 300 32059I-06/2010 AT32UC3B 21.3 Block Diagram Figure 21-1. USART Block Diagram Peripheral DMA Controller Channel Channel USART I/O Controller RXD Receiver RTS INTC USART Interrupt TXD Transmitter CTS DTR CLK_USART CLK_USART/DIV Modem Signals Control DSR DCD RI BaudRate Generator CLK Power Manager DIV User Interface Peripheral bus Table 21-1. PIN RXD TXD RTS CTS SPI Operating Mode USART RXD TXD RTS CTS SPI Slave MOSI MISO - CS SPI Master MISO MOSI CS - 301 32059I-06/2010 AT32UC3B 21.4 I/O Lines Description I/O Lines Description Description Serial Clock Transmit Serial Data or Master Out Slave In (MOSI) in SPI Master Mode or Master In Slave Out (MISO) in SPI Slave Mode Receive Serial Data or Master In Slave Out (MISO) in SPI Master Mode or Master Out Slave In (MOSI) in SPI Slave Mode Ring Indicator Data Set Ready Data Carrier Detect Data Terminal Ready Clear to Send or Slave Select (NSS) in SPI Slave Mode Request to Send or Slave Select (NSS) in SPI Master Mode Type I/O Output Active Level Table 21-2. Name CLK TXD RXD RI DSR DCD DTR CTS RTS Input Input Input Input Output Input Output Low Low Low Low Low Low 302 32059I-06/2010 AT32UC3B 21.5 21.5.1 Product Dependencies I/O Lines The pins used for interfacing the USART may be multiplexed with the I/O Controller lines. The programmer must first program the I/O Controller to assign the desired USART pins to their peripheral function. If I/O lines of the USART are not used by the application, they can be used for other purposes by the I/O Controller. To prevent the TXD line from falling when the USART is disabled, the use of an internal pull up is mandatory. If the hardware handshaking feature or Modem mode is used, the internal pull up on TXD must also be enabled. All the pins of the modems may or may not be implemented on the USART. On USARTs not equipped with the corresponding pins, the associated control bits and statuses have no effect on the behavior of the USART. 21.5.2 Clocks The clock for the USART bus interface (CLK_USART) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. It is recommended to disable the USART before disabling the clock, to avoid freezing the USART in an undefined state. 21.5.3 Interrupts The USART interrupt request line is connected to the interrupt controller. Using the USART interrupt requires the interrupt controller to be programmed first. 303 32059I-06/2010 AT32UC3B 21.6 Functional Description The USART is capable of managing several types of serial synchronous or asynchronous communications. It supports the following communication modes: * 5- to 9-bit full-duplex asynchronous serial communication - MSB- or LSB-first - 1, 1.5 or 2 stop bits - Parity even, odd, marked, space or none - By 8 or by 16 over-sampling receiver frequency - Optional hardware handshaking - Optional modem signals management - Optional break management - Optional multidrop serial communication * High-speed 5- to 9-bit full-duplex synchronous serial communication - MSB- or LSB-first - 1 or 2 stop bits - Parity even, odd, marked, space or none - By 8 or by 16 over-sampling frequency - Optional hardware handshaking - Optional modem signals management - Optional break management - Optional multidrop serial communication * RS485 with driver control signal * ISO7816, T0 or T1 protocols for interfacing with smart cards - NACK handling, error counter with repetition and iteration limit * InfraRed IrDA Modulation and Demodulation * SPI Mode - Master or Slave - Serial Clock Programmable Phase and Polarity - SPI Serial Clock (CLK) Frequency up to Internal Clock Frequency CLK_USART/4 * Test modes - Remote loopback, local loopback, automatic echo 21.6.1 Baud Rate Generator The Baud Rate Generator provides the bit period clock named the Baud Rate Clock to both the receiver and the transmitter. The Baud Rate Generator clock source can be selected by setting the USCLKS field in the Mode Register (MR) between: * CLK_USART * a division of CLK_USART, the divider being product dependent, but generally set to 8 * the external clock, available on the CLK pin 304 32059I-06/2010 AT32UC3B The Baud Rate Generator is based upon a 16-bit divider, which is programmed with the CD field of the Baud Rate Generator Register (BRGR). If CD is programmed at 0, the Baud Rate Generator does not generate any clock. If CD is programmed at 1, the divider is bypassed and becomes inactive. If the external CLK clock is selected, the duration of the low and high levels of the signal provided on the CLK pin must be longer than a CLK_USART period. The frequency of the signal provided on CLK must be at least 4.5 times lower than CLK_USART. Figure 21-2. Baud Rate Generator USCLKS CLK_USART CLK_USART/DIV CLK Reserved CD CD CLK 0 1 2 3 0 16-bit Counter >1 1 0 1 1 SYNC USCLKS= 3 Sampling Clock OVER 0 Sampling Divider 0 BaudRate Clock FIDI SYNC 21.6.1.1 Baud Rate in Asynchronous Mode If the USART is programmed to operate in asynchronous mode, the selected clock is first divided by CD, which is field programmed in the Baud Rate Generator Register (BRGR). The resulting clock is provided to the receiver as a sampling clock and then divided by 16 or 8, depending on the programming of the OVER bit in MR. If OVER is set to 1, the receiver sampling is 8 times higher than the baud rate clock. If OVER is cleared, the sampling is performed at 16 times the baud rate clock. The following formula performs the calculation of the Baud Rate. SelectedClock Baudrate = -------------------------------------------( 8 ( 2 - Over )CD ) This gives a maximum baud rate of CLK_USART divided by 8, assuming that CLK_USART is the highest possible clock and that OVER is programmed at 1. 305 32059I-06/2010 AT32UC3B 21.6.1.2 Baud Rate Calculation Example Table 21-3 shows calculations of CD to obtain a baud rate at 38400 bauds for different source clock frequencies. This table also shows the actual resulting baud rate and the error. Table 21-3. Baud Rate Example (OVER = 0) Expected Baud Rate Bit/s 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 38 400 6.00 8.00 8.14 12.00 13.02 19.53 20.00 23.30 24.00 30.00 39.06 40.00 40.69 52.08 53.33 53.71 65.10 81.38 97.66 6 8 8 12 13 20 20 23 24 30 39 40 40 52 53 54 65 81 98 Calculation Result CD Actual Baud Rate Bit/s 38 400.00 38 400.00 39 062.50 38 400.00 38 461.54 37 500.00 38 400.00 38 908.10 38 400.00 38 400.00 38 461.54 38 400.00 38 109.76 38 461.54 38 641.51 38 194.44 38 461.54 38 580.25 38 265.31 0.00% 0.00% 1.70% 0.00% 0.16% 2.40% 0.00% 1.31% 0.00% 0.00% 0.16% 0.00% 0.76% 0.16% 0.63% 0.54% 0.16% 0.47% 0.35% Error Source Clock MHz 3 686 400 4 915 200 5 000 000 7 372 800 8 000 000 12 000 000 12 288 000 14 318 180 14 745 600 18 432 000 24 000 000 24 576 000 25 000 000 32 000 000 32 768 000 33 000 000 40 000 000 50 000 000 60 000 000 The baud rate is calculated with the following formula: BaudRate = ( CLKUSART ) ( CD x 16 ) The baud rate error is calculated with the following formula. It is not recommended to work with an error higher than 5%. ExpectedBaudRate Error = 1 - -------------------------------------------------- ActualBaudRate 21.6.1.3 Fractional Baud Rate in Asynchronous Mode The Baud Rate generator previously defined is subject to the following limitation: the output frequency changes by only integer multiples of the reference frequency. An approach to this problem is to integrate a fractional N clock generator that has a high resolution. The generator architecture is modified to obtain Baud Rate changes by a fraction of the reference source clock. This fractional part is programmed with the FP field in the Baud Rate Generator Register 306 32059I-06/2010 AT32UC3B (BRGR). If FP is not 0, the fractional part is activated. The resolution is one eighth of the clock divider. This feature is only available when using USART normal mode. The fractional Baud Rate is calculated using the following formula: SelectedClock Baudrate = --------------------------------------------------------------- 8 ( 2 - Over ) CD + FP ----- 8 The modified architecture is presented below: Figure 21-3. Fractional Baud Rate Generator FP USCLKS CLK_USART CLK_USART/DIV CLK Reserved CD Modulus Control FP CD CLK FIDI >1 1 0 0 1 1 SYNC USCLKS = 3 Sampling Clock 0 OVER Sampling Divider SYNC 0 1 2 3 16-bit Counter glitch-free logic 0 BaudRate Clock 21.6.1.4 Baud Rate in Synchronous Mode or SPI Mode If the USART is programmed to operate in synchronous mode, the selected clock is simply divided by the field CD in BRGR. BaudRate = SelectedClock ------------------------------------CD In synchronous mode, if the external clock is selected (USCLKS = 3), the clock is provided directly by the signal on the USART CLK pin. No division is active. The value written in BRGR has no effect. The external clock frequency must be at least 4.5 times lower than the system clock. When either the external clock CLK or the internal clock divided (CLK_USART/DIV) is selected, the value programmed in CD must be even if the user has to ensure a 50:50 mark/space ratio on the CLK pin. If the internal clock CLK_USART is selected, the Baud Rate Generator ensures a 50:50 duty cycle on the CLK pin, even if the value programmed in CD is odd. 307 32059I-06/2010 AT32UC3B 21.6.1.5 Baud Rate in ISO 7816 Mode The ISO7816 specification defines the bit rate with the following formula: Di B = ----- x f Fi where: * B is the bit rate * Di is the bit-rate adjustment factor * Fi is the clock frequency division factor * f is the ISO7816 clock frequency (Hz) Di is a binary value encoded on a 4-bit field, named DI, as represented in Table 21-4. Table 21-4. DI field Di (decimal) Binary and Decimal Values for Di 0001 1 0010 2 0011 4 0100 8 0101 16 0110 32 1000 12 1001 20 Fi is a binary value encoded on a 4-bit field, named FI, as represented in Table 21-5. Table 21-5. FI field Fi (decimal Binary and Decimal Values for Fi 0000 372 0001 372 0010 558 0011 744 0100 1116 0101 1488 0110 1860 1001 512 1010 768 1011 1024 1100 1536 1101 2048 Table 21-6 shows the resulting Fi/Di Ratio, which is the ratio between the ISO7816 clock and the baud rate clock. Table 21-6. Fi/Di 1 2 4 8 16 32 12 20 Possible Values for the Fi/Di Ratio 372 372 186 93 46.5 23.25 11.62 31 18.6 558 558 279 139.5 69.75 34.87 17.43 46.5 27.9 774 744 372 186 93 46.5 23.25 62 37.2 1116 1116 558 279 139.5 69.75 34.87 93 55.8 1488 1488 744 372 186 93 46.5 124 74.4 1806 1860 930 465 232.5 116.2 58.13 155 93 512 512 256 128 64 32 16 42.66 25.6 768 768 384 192 96 48 24 64 38.4 1024 1024 512 256 128 64 32 85.33 51.2 1536 1536 768 384 192 96 48 128 76.8 2048 2048 1024 512 256 128 64 170.6 102.4 If the USART is configured in ISO7816 Mode, the clock selected by the USCLKS field in the Mode Register (MR) is first divided by the value programmed in the field CD in the Baud Rate Generator Register (BRGR). The resulting clock can be provided to the CLK pin to feed the smart card clock inputs. This means that the CLKO bit can be set in MR. This clock is then divided by the value programmed in the FI_DI_RATIO field in the FI_DI_Ratio register (FIDI). This is performed by the Sampling Divider, which performs a division by up to 2047 in ISO7816 Mode. The non-integer values of the Fi/Di Ratio are not supported and the user must program the FI_DI_RATIO field to a value as close as possible to the expected value. The FI_DI_RATIO field resets to the value 0x174 (372 in decimal) and is the most common divider between the ISO7816 clock and the bit rate (Fi = 372, Di = 1). 308 32059I-06/2010 AT32UC3B Figure 21-4 shows the relation between the Elementary Time Unit, corresponding to a bit time, and the ISO 7816 clock. Figure 21-4. Elementary Time Unit (ETU) FI_DI_RATIO ISO7816 Clock Cycles ISO7816 Clock on CLK ISO7816 I/O Line on TXD 1 ETU 21.6.2 Receiver and Transmitter Control After reset, the receiver is disabled. The user must enable the receiver by setting the RXEN bit in the Control Register (CR). However, the receiver registers can be programmed before the receiver clock is enabled. After reset, the transmitter is disabled. The user must enable it by setting the TXEN bit in the Control Register (CR). However, the transmitter registers can be programmed before being enabled. The Receiver and the Transmitter can be enabled together or independently. At any time, the software can perform a reset on the receiver or the transmitter of the USART by setting the corresponding bit, RSTRX and RSTTX respectively, in the Control Register (CR). The software resets clear the status flag and reset internal state machines but the user interface configuration registers hold the value configured prior to software reset. Regardless of what the receiver or the transmitter is performing, the communication is immediately stopped. The user can also independently disable the receiver or the transmitter by setting RXDIS and TXDIS respectively in CR. If the receiver is disabled during a character reception, the USART waits until the end of reception of the current character, then the reception is stopped. If the transmitter is disabled while it is operating, the USART waits the end of transmission of both the current character and character being stored in the Transmit Holding Register (THR). If a timeguard is programmed, it is handled normally. 21.6.3 21.6.3.1 Synchronous and Asynchronous Modes Transmitter Operations The transmitter performs the same in both synchronous and asynchronous operating modes (SYNC = 0 or SYNC = 1). One start bit, up to 9 data bits, one optional parity bit and up to two stop bits are successively shifted out on the TXD pin at each falling edge of the programmed serial clock. The number of data bits is selected by the CHRL field and the MODE 9 bit in the Mode Register (MR). Nine bits are selected by setting the MODE 9 bit regardless of the CHRL field. The parity bit is set according to the PAR field in MR. The even, odd, space, marked or none parity bit can be configured. The MSBF field in MR configures which data bit is sent first. If written at 1, the most significant bit is sent first. At 0, the less significant bit is sent first. The number of stop bits is selected by the NBSTOP field in MR. The 1.5 stop bit is supported in asynchronous mode only. 309 32059I-06/2010 AT32UC3B Figure 21-5. Character Transmit Example: 8-bit, Parity Enabled One Stop Baud Rate Clock TXD Start Bit D0 D1 D2 D3 D4 D5 D6 D7 Parity Bit Stop Bit The characters are sent by writing in the Transmit Holding Register (THR). The transmitter reports two status bits in the Channel Status Register (CSR): TXRDY (Transmitter Ready), which indicates that THR is empty and TXEMPTY, which indicates that all the characters written in THR have been processed. When the current character processing is completed, the last character written in THR is transferred into the Shift Register of the transmitter and THR becomes empty, thus TXRDY rises. Both TXRDY and TXEMPTY bits are low when the transmitter is disabled. Writing a character in THR while TXRDY is low has no effect and the written character is lost. Figure 21-6. Transmitter Status Baud Rate Clock TXD Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Start D0 Bit Bit Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit Write THR TXRDY TXEMPTY 21.6.3.2 Manchester Encoder When the Manchester encoder is in use, characters transmitted through the USART are encoded based on biphase Manchester II format. To enable this mode, set the MAN field in the MR register to 1. Depending on polarity configuration, a logic level (zero or one), is transmitted as a coded signal one-to-zero or zero-to-one. Thus, a transition always occurs at the midpoint of each bit time. It consumes more bandwidth than the original NRZ signal (2x) but the receiver has more error control since the expected input must show a change at the center of a bit cell. An example of Manchester encoded sequence is: the byte 0xB1 or 10110001 encodes to 10 01 10 10 01 01 01 10, assuming the default polarity of the encoder. Figure 21-7 illustrates this coding scheme. 310 32059I-06/2010 AT32UC3B Figure 21-7. NRZ to Manchester Encoding NRZ encoded data Manchester encoded data 1 0 1 1 0 0 0 1 Txd The Manchester encoded character can also be encapsulated by adding both a configurable preamble and a start frame delimiter pattern. Depending on the configuration, the preamble is a training sequence, composed of a pre-defined pattern with a programmable length from 1 to 15 bit times. If the preamble length is set to 0, the preamble waveform is not generated prior to any character. The preamble pattern is chosen among the following sequences: ALL_ONE, ALL_ZERO, ONE_ZERO or ZERO_ONE, writing the field TX_PP in the MAN register, the field TX_PL is used to configure the preamble length. Figure 21-8 illustrates and defines the valid patterns. To improve flexibility, the encoding scheme can be configured using the TX_MPOL field in the MAN register. If the TX_MPOL field is set to zero (default), a logic zero is encoded with a zero-to-one transition and a logic one is encoded with a one-to-zero transition. If the TX_MPOL field is set to one, a logic one is encoded with a one-to-zero transition and a logic zero is encoded with a zero-to-one transition. Figure 21-8. Preamble Patterns, Default Polarity Assumed Manchester encoded data Txd SFD DATA 8 bit width "ALL_ONE" Preamble Manchester encoded data Txd SFD DATA 8 bit width "ALL_ZERO" Preamble Manchester encoded data Txd SFD DATA 8 bit width "ZERO_ONE" Preamble Manchester encoded data Txd SFD DATA 8 bit width "ONE_ZERO" Preamble A start frame delimiter is to be configured using the ONEBIT field in the MR register. It consists of a user-defined pattern that indicates the beginning of a valid data. Figure 21-9 illustrates these patterns. If the start frame delimiter, also known as start bit, is one bit, (ONEBIT at 1), a logic zero is Manchester encoded and indicates that a new character is being sent serially on the line. If the start frame delimiter is a synchronization pattern also referred to as sync (ONEBIT at 0), a sequence of 3 bit times is sent serially on the line to indicate the start of a new character. 311 32059I-06/2010 AT32UC3B The sync waveform is in itself an invalid Manchester waveform as the transition occurs at the middle of the second bit time. Two distinct sync patterns are used: the command sync and the data sync. The command sync has a logic one level for one and a half bit times, then a transition to logic zero for the second one and a half bit times. If the MODSYNC field in the MR register is set to 1, the next character is a command. If it is set to 0, the next character is a data. When direct memory access is used, the MODSYNC field can be immediately updated with a modified character located in memory. To enable this mode, VAR_SYNC field in MR register must be set to 1. In this case, the MODSYNC field in MR is bypassed and the sync configuration is held in the TXSYNH in the THR register. The USART character format is modified and includes sync information. Figure 21-9. Start Frame Delimiter Preamble Length is set to 0 SFD Manchester encoded data Txd DATA One bit start frame delimiter SFD Manchester encoded data Txd DATA SFD Manchester encoded data Txd Command Sync start frame delimiter DATA Data Sync start frame delimiter Drift Compensation Drift compensation is available only in 16X oversampling mode. An hardware recovery system allows a larger clock drift. To enable the hardware system, the bit in the MAN register must be set. If the RXD edge is one 16X clock cycle from the expected edge, this is considered as normal jitter and no corrective actions is taken. If the RXD event is between 4 and 2 clock cycles before the expected edge, then the current period is shortened by one clock cycle. If the RXD event is between 2 and 3 clock cycles after the expected edge, then the current period is lengthened by one clock cycle. These intervals are considered to be drift and so corrective actions are automatically taken. 312 32059I-06/2010 AT32UC3B Figure 21-10. Bit Resynchronization Oversampling 16x Clock RXD Sampling point Expected edge Synchro. Error Synchro. Jump Tolerance Sync Jump Synchro. Error 21.6.3.3 Asynchronous Receiver If the USART is programmed in asynchronous operating mode (SYNC = 0), the receiver oversamples the RXD input line. The oversampling is either 16 or 8 times the Baud Rate clock, depending on the OVER bit in the Mode Register (MR). The receiver samples the RXD line. If the line is sampled during one half of a bit time at 0, a start bit is detected and data, parity and stop bits are successively sampled on the bit rate clock. If the oversampling is 16, (OVER at 0), a start is detected at the eighth sample at 0. Then, data bits, parity bit and stop bit are sampled on each 16 sampling clock cycle. If the oversampling is 8 (OVER at 1), a start bit is detected at the fourth sample at 0. Then, data bits, parity bit and stop bit are sampled on each 8 sampling clock cycle. The number of data bits, first bit sent and parity mode are selected by the same fields and bits as the transmitter, i.e. respectively CHRL, MODE9, MSBF and PAR. For the synchronization mechanism only, the number of stop bits has no effect on the receiver as it considers only one stop bit, regardless of the field NBSTOP, so that resynchronization between the receiver and the transmitter can occur. Moreover, as soon as the stop bit is sampled, the receiver starts looking for a new start bit so that resynchronization can also be accomplished when the transmitter is operating with one stop bit. Figure 21-11 and Figure 21-12 illustrate start detection and character reception when USART operates in asynchronous mode. 313 32059I-06/2010 AT32UC3B Figure 21-11. Asynchronous Start Detection Baud Rate Clock Sampling Clock (x16) RXD Sampling 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 D0 Sampling Start Detection RXD Sampling 1 2 3 4 5 6 01 Start Rejection 7 2 3 4 Figure 21-12. Asynchronous Character Reception Example: 8-bit, Parity Enabled Baud Rate Clock RXD Start Detection 16 16 16 16 16 16 16 16 16 16 samples samples samples samples samples samples samples samples samples samples D0 D1 D2 D3 D4 D5 D6 D7 Parity Bit Stop Bit 21.6.3.4 Manchester Decoder When the MAN field in MR register is set to 1, the Manchester decoder is enabled. The decoder performs both preamble and start frame delimiter detection. One input line is dedicated to Manchester encoded input data. An optional preamble sequence can be defined, its length is user-defined and totally independent of the emitter side. Use RX_PL in MAN register to configure the length of the preamble sequence. If the length is set to 0, no preamble is detected and the function is disabled. In addition, the polarity of the input stream is programmable with RX_MPOL field in MAN register. Depending on the desired application the preamble pattern matching is to be defined via the RX_PP field in MAN. See Figure 21-8 for available preamble patterns. Unlike preamble, the start frame delimiter is shared between Manchester Encoder and Decoder. So, if ONEBIT field is set to 1, only a zero encoded Manchester can be detected as a valid start frame delimiter. If ONEBIT is set to 0, only a sync pattern is detected as a valid start frame delimiter. Decoder operates by detecting transition on incoming stream. If RXD is sampled during one quarter of a bit time at zero, a start bit is detected. See Figure 21-13. The sample pulse rejection mechanism applies. 314 32059I-06/2010 AT32UC3B Figure 21-13. Asynchronous Start Bit Detection Sampling Clock (16 x) Manchester encoded data Txd Start Detection 1 2 3 4 The receiver is activated and starts Preamble and Frame Delimiter detection, sampling the data at one quarter and then three quarters. If a valid preamble pattern or start frame delimiter is detected, the receiver continues decoding with the same synchronization. If the stream does not match a valid pattern or a valid start frame delimiter, the receiver re-synchronizes on the next valid edge.The minimum time threshold to estimate the bit value is three quarters of a bit time. If a valid preamble (if used) followed with a valid start frame delimiter is detected, the incoming stream is decoded into NRZ data and passed to USART for processing. Figure 21-14 illustrates Manchester pattern mismatch. When incoming data stream is passed to the USART, the receiver is also able to detect Manchester code violation. A code violation is a lack of transition in the middle of a bit cell. In this case, MANE flag in CSR register is raised. It is cleared by writing the Control Register (CR) with the RSTSTA bit at 1. See Figure 21-15 for an example of Manchester error detection during data phase. Figure 21-14. Preamble Pattern Mismatch Preamble Mismatch Manchester coding error Preamble Mismatch invalid pattern Manchester encoded data Txd SFD DATA Preamble Length is set to 8 Figure 21-15. Manchester Error Flag Preamble Length is set to 4 Elementary character bit time SFD Manchester encoded data Txd Entering USART character area sampling points Preamble subpacket and Start Frame Delimiter were successfully decoded Manchester Coding Error detected 315 32059I-06/2010 AT32UC3B When the start frame delimiter is a sync pattern (ONEBIT field at 0), both command and data delimiter are supported. If a valid sync is detected, the received character is written as RXCHR field in the RHR register and the RXSYNH is updated. RXCHR is set to 1 when the received character is a command, and it is set to 0 if the received character is a data. This mechanism alleviates and simplifies the direct memory access as the character contains its own sync field in the same register. As the decoder is setup to be used in unipolar mode, the first bit of the frame has to be a zero-toone transition. 21.6.3.5 Radio Interface: Manchester Encoded USART Application This section describes low data rate RF transmission systems and their integration with a Manchester encoded USART. These systems are based on transmitter and receiver ICs that support ASK and FSK modulation schemes. The goal is to perform full duplex radio transmission of characters using two different frequency carriers. See the configuration in Figure 21-16. Figure 21-16. Manchester Encoded Characters RF Transmission Fup frequency Carrier ASK/FSK Upstream Receiver Upstream Emitter LNA VCO RF filter Demod Serial Configuration Interface control Fdown frequency Carrier bi-dir line ASK/FSK downstream transmitter Manchester decoder USART Receiver Downstream Receiver Manchester encoder PA RF filter Mod VCO USART Emitter control The USART module is configured as a Manchester encoder/decoder. Looking at the downstream communication channel, Manchester encoded characters are serially sent to the RF emitter. This may also include a user defined preamble and a start frame delimiter. Mostly, preamble is used in the RF receiver to distinguish between a valid data from a transmitter and signals due to noise. The Manchester stream is then modulated. See Figure 21-17 for an example of ASK modulation scheme. When a logic one is sent to the ASK modulator, the power amplifier, referred to as PA, is enabled and transmits an RF signal at downstream frequency. When a logic zero is transmitted, the RF signal is turned off. If the FSK modulator is activated, two different frequencies are used to transmit data. When a logic 1 is sent, the modulator outputs an RF signal at frequency F0 and switches to F1 if the data sent is a 0. See Figure 21-18. From the receiver side, another carrier frequency is used. The RF receiver performs a bit check operation examining demodulated data stream. If a valid pattern is detected, the receiver 316 32059I-06/2010 AT32UC3B switches to receiving mode. The demodulated stream is sent to the Manchester decoder. Because of bit checking inside RF IC, the data transferred to the microcontroller is reduced by a user-defined number of bits. The Manchester preamble length is to be defined in accordance with the RF IC configuration. Figure 21-17. ASK Modulator Output 1 NRZ stream Manchester encoded data default polarity unipolar output ASK Modulator Output Uptstream Frequency F0 0 0 1 Txd Figure 21-18. FSK Modulator Output 1 NRZ stream Manchester encoded data default polarity unipolar output FSK Modulator Output Uptstream Frequencies [F0, F0+offset] 0 0 1 Txd 21.6.3.6 Synchronous Receiver In synchronous mode (SYNC = 1), the receiver samples the RXD signal on each rising edge of the Baud Rate Clock. If a low level is detected, it is considered as a start. All data bits, the parity bit and the stop bits are sampled and the receiver waits for the next start bit. Synchronous mode operations provide a high speed transfer capability. Configuration fields and bits are the same as in asynchronous mode. Figure 21-19 illustrates a character reception in synchronous mode. 317 32059I-06/2010 AT32UC3B Figure 21-19. Synchronous Mode Character Reception Example: 8-bit, Parity Enabled 1 Stop Baud Rate Clock RXD Sampling Start D0 D1 D2 D3 D4 D5 D6 D7 Parity Bit Stop Bit 21.6.3.7 Receiver Operations When a character reception is completed, it is transferred to the Receive Holding Register (RHR) and the RXRDY bit in the Status Register (CSR) rises. If a character is completed while the RXRDY is set, the OVRE (Overrun Error) bit is set. The last character is transferred into RHR and overwrites the previous one. The OVRE bit is cleared by writing the Control Register (CR) with the RSTSTA (Reset Status) bit at 1. Figure 21-20. Receiver Status Baud Rate Clock RXD Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Start D0 Bit Bit Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit RSTSTA = 1 Write CR Read RHR RXRDY OVRE 318 32059I-06/2010 AT32UC3B 21.6.3.8 Parity The USART supports five parity modes selected by programming the PAR field in the Mode Register (MR). The PAR field also enables the Multidrop mode, see "Multidrop Mode" on page 320. Even and odd parity bit generation and error detection are supported. If even parity is selected, the parity generator of the transmitter drives the parity bit at 0 if a number of 1s in the character data bit is even, and at 1 if the number of 1s is odd. Accordingly, the receiver parity checker counts the number of received 1s and reports a parity error if the sampled parity bit does not correspond. If odd parity is selected, the parity generator of the transmitter drives the parity bit at 1 if a number of 1s in the character data bit is even, and at 0 if the number of 1s is odd. Accordingly, the receiver parity checker counts the number of received 1s and reports a parity error if the sampled parity bit does not correspond. If the mark parity is used, the parity generator of the transmitter drives the parity bit at 1 for all characters. The receiver parity checker reports an error if the parity bit is sampled at 0. If the space parity is used, the parity generator of the transmitter drives the parity bit at 0 for all characters. The receiver parity checker reports an error if the parity bit is sampled at 1. If parity is disabled, the transmitter does not generate any parity bit and the receiver does not report any parity error. Table 21-7 shows an example of the parity bit for the character 0x41 (character ASCII "A") depending on the configuration of the USART. Because there are two bits at 1, 1 bit is added when a parity is odd, or 0 is added when a parity is even. Table 21-7. Character A A A A A Parity Bit Examples Hexa 0x41 0x41 0x41 0x41 0x41 Binary 0100 0001 0100 0001 0100 0001 0100 0001 0100 0001 Parity Bit 1 0 1 0 None Parity Mode Odd Even Mark Space None When the receiver detects a parity error, it sets the PARE (Parity Error) bit in the Channel Status Register (CSR). The PARE bit can be cleared by writing the Control Register (CR) with the RSTSTA bit at 1. Figure 21-21 illustrates the parity bit status setting and clearing. 319 32059I-06/2010 AT32UC3B Figure 21-21. Parity Error Baud Rate Clock RXD Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Bad Stop Parity Bit Bit RSTSTA = 1 Write CR PARE RXRDY 21.6.3.9 Multidrop Mode If the PAR field in the Mode Register (MR) is programmed to the value 0x6 or 0x07, the USART runs in Multidrop Mode. This mode differentiates the data characters and the address characters. Data is transmitted with the parity bit at 0 and addresses are transmitted with the parity bit at 1. If the USART is configured in multidrop mode, the receiver sets the PARE parity error bit when the parity bit is high and the transmitter is able to send a character with the parity bit high when the Control Register is written with the SENDA bit at 1. To handle parity error, the PARE bit is cleared when the Control Register is written with the bit RSTSTA at 1. The transmitter sends an address byte (parity bit set) when SENDA is written to CR. In this case, the next byte written to THR is transmitted as an address. Any character written in THR without having written the command SENDA is transmitted normally with the parity at 0. 21.6.3.10 Transmitter Timeguard The timeguard feature enables the USART interface with slow remote devices. The timeguard function enables the transmitter to insert an idle state on the TXD line between two characters. This idle state actually acts as a long stop bit. The duration of the idle state is programmed in the TG field of the Transmitter Timeguard Register (TTGR). When this field is programmed at zero no timeguard is generated. Otherwise, the transmitter holds a high level on TXD after each transmitted byte during the number of bit periods programmed in TG in addition to the number of stop bits. As illustrated in Figure 21-22, the behavior of TXRDY and TXEMPTY status bits is modified by the programming of a timeguard. TXRDY rises only when the start bit of the next character is sent, and thus remains at 0 during the timeguard transmission if a character has been written in THR. TXEMPTY remains low until the timeguard transmission is completed as the timeguard is part of the current character being transmitted. 320 32059I-06/2010 AT32UC3B Figure 21-22. Timeguard Operations TG = 4 Baud Rate Clock TXD Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit TG = 4 Write THR TXRDY TXEMPTY Table 21-8 indicates the maximum length of a timeguard period that the transmitter can handle in relation to the function of the Baud Rate. Table 21-8. Maximum Timeguard Length Depending on Baud Rate Bit time s 833 104 69.4 52.1 34.7 29.9 17.9 17.4 8.7 Timeguard ms 212.50 26.56 17.71 13.28 8.85 7.63 4.55 4.43 2.21 Baud Rate Bit/sec 1 200 9 600 14400 19200 28800 33400 56000 57600 115200 21.6.3.11 Receiver Time-out The Receiver Time-out provides support in handling variable-length frames. This feature detects an idle condition on the RXD line. When a time-out is detected, the bit TIMEOUT in the Channel Status Register (CSR) rises and can generate an interrupt, thus indicating to the driver an end of frame. The time-out delay period (during which the receiver waits for a new character) is programmed in the TO field of the Receiver Time-out Register (RTOR). If the TO field is programmed at 0, the Receiver Time-out is disabled and no time-out is detected. The TIMEOUT bit in CSR remains at 0. Otherwise, the receiver loads a counter with the value programmed in TO. This counter is decremented at each bit period and reloaded each time a new character is received. If the counter reaches 0, the TIMEOUT bit in the Status Register rises. Then, the user can either: * Stop the counter clock until a new character is received. This is performed by writing the Control Register (CR) with the STTTO (Start Time-out) bit at 1. In this case, the idle state on RXD before a new character is received will not provide a time-out. This prevents having to 321 32059I-06/2010 AT32UC3B handle an interrupt before a character is received and allows waiting for the next idle state on RXD after a frame is received. * Obtain an interrupt while no character is received. This is performed by writing CR with the RETTO (Reload and Start Time-out) bit at 1. If RETTO is performed, the counter starts counting down immediately from the value TO. This enables generation of a periodic interrupt so that a user time-out can be handled, for example when no key is pressed on a keyboard. If STTTO is performed, the counter clock is stopped until a first character is received. The idle state on RXD before the start of the frame does not provide a time-out. This prevents having to obtain a periodic interrupt and enables a wait of the end of frame when the idle state on RXD is detected. If RETTO is performed, the counter starts counting down immediately from the value TO. This enables generation of a periodic interrupt so that a user time-out can be handled, for example when no key is pressed on a keyboard. Figure 21-23 shows the block diagram of the Receiver Time-out feature. Figure 21-23. Receiver Time-out Block Diagram Baud Rate Clock TO 1 STTTO D Q Clock 16-bit Time-out Counter Load 16-bit Value = TIMEOUT Character Received RETTO Clear 0 Table 21-9 gives the maximum time-out period for some standard baud rates. Table 21-9. Maximum Time-out Period Bit Time s 1 667 833 417 208 104 69 52 35 30 Time-out ms 109 225 54 613 27 306 13 653 6 827 4 551 3 413 2 276 1 962 Baud Rate bit/sec 600 1 200 2 400 4 800 9 600 14400 19200 28800 33400 322 32059I-06/2010 AT32UC3B Table 21-9. Maximum Time-out Period (Continued) Bit Time 18 17 5 Time-out 1 170 1 138 328 Baud Rate 56000 57600 200000 21.6.3.12 Framing Error The receiver is capable of detecting framing errors. A framing error happens when the stop bit of a received character is detected at level 0. This can occur if the receiver and the transmitter are fully desynchronized. A framing error is reported on the FRAME bit of the Channel Status Register (CSR). The FRAME bit is asserted in the middle of the stop bit as soon as the framing error is detected. It is cleared by writing the Control Register (CR) with the RSTSTA bit at 1. Figure 21-24. Framing Error Status Baud Rate Clock RXD Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit RSTSTA = 1 Write CR FRAME RXRDY 21.6.3.13 Transmit Break The user can request the transmitter to generate a break condition on the TXD line. A break condition drives the TXD line low during at least one complete character. It appears the same as a 0x00 character sent with the parity and the stop bits at 0. However, the transmitter holds the TXD line at least during one character until the user requests the break condition to be removed. A break is transmitted by writing the Control Register (CR) with the STTBRK bit at 1. This can be performed at any time, either while the transmitter is empty (no character in either the Shift Register or in THR) or when a character is being transmitted. If a break is requested while a character is being shifted out, the character is first completed before the TXD line is held low. Once STTBRK command is requested further STTBRK commands are ignored until the end of the break is completed. The break condition is removed by writing CR with the STPBRK bit at 1. If the STPBRK is requested before the end of the minimum break duration (one character, including start, data, parity and stop bits), the transmitter ensures that the break condition completes. 323 32059I-06/2010 AT32UC3B The transmitter considers the break as though it is a character, i.e. the STTBRK and STPBRK commands are taken into account only if the TXRDY bit in CSR is at 1 and the start of the break condition clears the TXRDY and TXEMPTY bits as if a character is processed. Writing CR with the both STTBRK and STPBRK bits at 1 can lead to an unpredictable result. All STPBRK commands requested without a previous STTBRK command are ignored. A byte written into the Transmit Holding Register while a break is pending, but not started, is ignored. After the break condition, the transmitter returns the TXD line to 1 for a minimum of 12 bit times. Thus, the transmitter ensures that the remote receiver detects correctly the end of break and the start of the next character. If the timeguard is programmed with a value higher than 12, the TXD line is held high for the timeguard period. After holding the TXD line for this period, the transmitter resumes normal operations. Figure 21-25 illustrates the effect of both the Start Break (STTBRK) and Stop Break (STPBRK) commands on the TXD line. Figure 21-25. Break Transmission Baud Rate Clock TXD Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit Break Transmission STPBRK = 1 End of Break STTBRK = 1 Write CR TXRDY TXEMPTY 21.6.3.14 Receive Break The receiver detects a break condition when all data, parity and stop bits are low. This corresponds to detecting a framing error with data at 0x00, but FRAME remains low. When the low stop bit is detected, the receiver asserts the RXBRK bit in CSR. This bit may be cleared by writing the Control Register (CR) with the bit RSTSTA at 1. An end of receive break is detected by a high level for at least 2/16 of a bit period in asynchronous operating mode or one sample at high level in synchronous operating mode. The end of break detection also asserts the RXBRK bit. 21.6.3.15 Hardware Handshaking The USART features a hardware handshaking out-of-band flow control. The RTS and CTS pins are used to connect with the remote device, as shown in Figure 21-26. 324 32059I-06/2010 AT32UC3B Figure 21-26. Connection with a Remote Device for Hardware Handshaking USART TXD RXD CTS RTS Remote Device RXD TXD RTS CTS Setting the USART to operate with hardware handshaking is performed by writing the MODE field in the Mode Register (MR) to the value 0x2. The USART behavior when hardware handshaking is enabled is the same as the behavior in standard synchronous or asynchronous mode, except that the receiver drives the RTS pin as described below and the level on the CTS pin modifies the behavior of the transmitter as described below. Using this mode requires using the Peripheral DMA Controller channel for reception. The transmitter can handle hardware handshaking in any case. Figure 21-27 shows how the receiver operates if hardware handshaking is enabled. The RTS pin is driven high if the receiver is disabled and if the status RXBUFF (Receive Buffer Full) coming from the Peripheral DMA Controller channel is high. Normally, the remote device does not start transmitting while its CTS pin (driven by RTS) is high. As soon as the Receiver is enabled, the RTS falls, indicating to the remote device that it can start transmitting. Defining a new buffer to the Peripheral DMA Controller clears the status bit RXBUFF and, as a result, asserts the pin RTS low. Figure 21-27. Receiver Behavior when Operating with Hardware Handshaking RXD RXEN = 1 Write CR RTS RXBUFF RXDIS = 1 Figure 21-28 shows how the transmitter operates if hardware handshaking is enabled. The CTS pin disables the transmitter. If a character is being processing, the transmitter is disabled only after the completion of the current character and transmission of the next character happens as soon as the pin CTS falls. Figure 21-28. Transmitter Behavior when Operating with Hardware Handshaking CTS TXD 325 32059I-06/2010 AT32UC3B 21.6.4 ISO7816 Mode The USART features an ISO7816-compatible operating mode. This mode permits interfacing with smart cards and Security Access Modules (SAM) communicating through an ISO7816 link. Both T = 0 and T = 1 protocols defined by the ISO7816 specification are supported. Setting the USART in ISO7816 mode is performed by writing the MODE field in the Mode Register (MR) to the value 0x4 for protocol T = 0 and to the value 0x6 for protocol T = 1. 21.6.4.1 ISO7816 Mode Overview The ISO7816 is a half duplex communication on only one bidirectional line. The baud rate is determined by a division of the clock provided to the remote device (see "Baud Rate Generator" on page 304). The USART connects to a smart card as shown in Figure 21-29. The TXD line becomes bidirectional and the Baud Rate Generator feeds the ISO7816 clock on the CLK pin. As the TXD pin becomes bidirectional, its output remains driven by the output of the transmitter but only when the transmitter is active while its input is directed to the input of the receiver. The USART is considered as the master of the communication as it generates the clock. Figure 21-29. Connection of a Smart Card to the USART USART CLK TXD CLK I/O Smart Card When operating in ISO7816, either in T = 0 or T = 1 modes, the character format is fixed. The configuration is 8 data bits, even parity and 1 or 2 stop bits, regardless of the values programmed in the CHRL, MODE9, PAR and CHMODE fields. MSBF can be used to transmit LSB or MSB first. Parity Bit (PAR) can be used to transmit in normal or inverse mode. Refer to "Mode Register" on page 343 and "PAR: Parity Type" on page 344. The USART cannot operate concurrently in both receiver and transmitter modes as the communication is unidirectional at a time. It has to be configured according to the required mode by enabling or disabling either the receiver or the transmitter as desired. Enabling both the receiver and the transmitter at the same time in ISO7816 mode may lead to unpredictable results. The ISO7816 specification defines an inverse transmission format. Data bits of the character must be transmitted on the I/O line at their negative value. The USART does not support this format and the user has to perform an exclusive OR on the data before writing it in the Transmit Holding Register (THR) or after reading it in the Receive Holding Register (RHR). 21.6.4.2 Protocol T = 0 In T = 0 protocol, a character is made up of one start bit, eight data bits, one parity bit and one guard time, which lasts two bit times. The transmitter shifts out the bits and does not drive the I/O line during the guard time. If no parity error is detected, the I/O line remains at 1 during the guard time and the transmitter can continue with the transmission of the next character, as shown in Figure 21-30. 326 32059I-06/2010 AT32UC3B If a parity error is detected by the receiver, it drives the I/O line at 0 during the guard time, as shown in Figure 21-31. This error bit is also named NACK, for Non Acknowledge. In this case, the character lasts 1 bit time more, as the guard time length is the same and is added to the error bit time which lasts 1 bit time. When the USART is the receiver and it detects an error, it does not load the erroneous character in the Receive Holding Register (RHR). It appropriately sets the PARE bit in the Status Register (SR) so that the software can handle the error. Figure 21-30. T = 0 Protocol without Parity Error Baud Rate Clock RXD Start Bit D0 D1 D2 D3 D4 D5 D6 D7 Parity Guard Guard Next Bit Time 1 Time 2 Start Bit Figure 21-31. T = 0 Protocol with Parity Error Baud Rate Clock I/O Start Bit D0 D1 D2 D3 D4 D5 D6 D7 Error Parity Guard Bit Time 1 Guard Start Time 2 Bit D0 D1 Repetition 21.6.4.3 Receive Error Counter The USART receiver also records the total number of errors. This can be read in the Number of Error (NER) register. The NB_ERRORS field can record up to 255 errors. Reading NER automatically clears the NB_ERRORS field. 21.6.4.4 Receive NACK Inhibit The USART can also be configured to inhibit an error. This can be achieved by setting the INACK bit in the Mode Register (MR). If INACK is at 1, no error signal is driven on the I/O line even if a parity bit is detected. Moreover, if INACK is set, the erroneous received character is stored in the Receive Holding Register, as if no error occurred. However, the RXRDY bit does raise. 21.6.4.5 Transmit Character Repetition When the USART is transmitting a character and gets a NACK, it can automatically repeat the character before moving on to the next one. Repetition is enabled by writing the MAX_ITERATION field in the Mode Register (MR) at a value higher than 0. Each character can be transmitted up to eight times; the first transmission plus seven repetitions. If MAX_ITERATION does not equal zero, the USART repeats the character as many times as the value loaded in MAX_ITERATION. 327 32059I-06/2010 AT32UC3B When the USART repetition number reaches MAX_ITERATION, the ITERATION bit is set in the Channel Status Register (CSR). If the repetition of the character is acknowledged by the receiver, the repetitions are stopped and the iteration counter is cleared. The ITERATION bit in CSR can be cleared by writing the Control Register with the RSIT bit at 1. 21.6.4.6 Disable Successive Receive NACK The receiver can limit the number of successive NACKs sent back to the remote transmitter. This is programmed by setting the bit DSNACK in the Mode Register (MR). The maximum number of NACK transmitted is programmed in the MAX_ITERATION field. As soon as MAX_ITERATION is reached, the character is considered as correct, an acknowledge is sent on the line and the ITERATION bit in the Channel Status Register is set. 21.6.4.7 Protocol T = 1 When operating in ISO7816 protocol T = 1, the transmission is similar to an asynchronous format with only one stop bit. The parity is generated when transmitting and checked when receiving. Parity error detection sets the PARE bit in the Channel Status Register (CSR). 21.6.5 IrDA Mode The USART features an IrDA mode supplying half-duplex point-to-point wireless communication. It embeds the modulator and demodulator which allows a glueless connection to the infrared transceivers, as shown in Figure 21-32. The modulator and demodulator are compliant with the IrDA specification version 1.1 and support data transfer speeds ranging from 2.4 Kb/s to 115.2 Kb/s. The USART IrDA mode is enabled by setting the MODE field in the Mode Register (MR) to the value 0x8. The IrDA Filter Register (IFR) allows configuring the demodulator filter. The USART transmitter and receiver operate in a normal asynchronous mode and all parameters are accessible (except those fixed by IrDA specification : one start bit , 8 data bits and one stop bit). Note that the modulator and the demodulator are activated. Figure 21-32. Connection to IrDA Transceivers USART Receiver Demodulator RXD RX TX Transmitter Modulator TXD IrDA Transceivers The receiver and the transmitter must be enabled or disabled according to the direction of the transmission to be managed. To receive IrDA signals, the following needs to be done: * Disable TX and Enable RX 328 32059I-06/2010 AT32UC3B * Configure the TXD pin as I/O and set it as an output at 0 (to avoid LED emission). Disable the internal pull-up (better for power consumption). * Receive data 21.6.5.1 IrDA Modulation For baud rates up to and including 115.2 Kbits/sec, the RZI modulation scheme is used. "0" is represented by a light pulse of 3/16th of a bit time. Some examples of signal pulse duration are shown in Table 21-10. Table 21-10. IrDA Pulse Duration Baud Rate 2.4 Kb/s 9.6 Kb/s 19.2 Kb/s 38.4 Kb/s 57.6 Kb/s 115.2 Kb/s Pulse Duration (3/16) 78.13 s 19.53 s 9.77 s 4.88 s 3.26 s 1.63 s Figure 21-33 shows an example of character transmission. Figure 21-33. IrDA Modulation Start Bit Transmitter Output 0 1 0 1 Data Bits 0 0 1 1 0 Stop Bit 1 TXD Bit Period 3 16 Bit Period 21.6.5.2 IrDA Baud Rate Table 21-11 gives some examples of CD values, baud rate error and pulse duration. Note that the requirement on the maximum acceptable error of 1.87% must be met. Table 21-11. IrDA Baud Rate Error Peripheral Clock 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 Baud Rate 115 200 115 200 115 200 115 200 57 600 57 600 57 600 CD 2 11 18 22 4 22 36 Baud Rate Error 0.00% 1.38% 1.25% 1.38% 0.00% 1.38% 1.25% Pulse Time 1.63 1.63 1.63 1.63 3.26 3.26 3.26 329 32059I-06/2010 AT32UC3B Table 21-11. IrDA Baud Rate Error (Continued) Peripheral Clock 40 000 000 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 40 000 000 3 686 400 20 000 000 32 768 000 Baud Rate 57 600 38 400 38 400 38 400 38 400 19 200 19 200 19 200 19 200 9 600 9 600 9 600 9 600 2 400 2 400 2 400 CD 43 6 33 53 65 12 65 107 130 24 130 213 260 96 521 853 Baud Rate Error 0.93% 0.00% 1.38% 0.63% 0.16% 0.00% 0.16% 0.31% 0.16% 0.00% 0.16% 0.16% 0.16% 0.00% 0.03% 0.04% Pulse Time 3.26 4.88 4.88 4.88 4.88 9.77 9.77 9.77 9.77 19.53 19.53 19.53 19.53 78.13 78.13 78.13 21.6.5.3 IrDA Demodulator The demodulator is based on the IrDA Receive filter comprised of an 8-bit down counter which is loaded with the value programmed in IFR. When a falling edge is detected on the RXD pin, the Filter Counter starts counting down at the CLK_USART speed. If a rising edge is detected on the RXD pin, the counter stops and is reloaded with IFR. If no rising edge is detected when the counter reaches 0, the input of the receiver is driven low during one bit time. Figure 21-34 illustrates the operations of the IrDA demodulator. Figure 21-34. IrDA Demodulator Operations CLK_USART RXD Counter Value Receiver Input Pulse Accepted 6 5 4 3 2 6 6 5 4 3 2 1 0 Pulse Rejected Driven Low During 16 Baud Rate Clock Cycles As the IrDA mode uses the same logic as the ISO7816, note that the FI_DI_RATIO field in FIDI must be set to a value higher than 0 in order to assure IrDA communications operate correctly. 330 32059I-06/2010 AT32UC3B 21.6.6 RS485 Mode The USART features the RS485 mode to enable line driver control. While operating in RS485 mode, the USART behaves as though in asynchronous or synchronous mode and configuration of all the parameters is possible. The difference is that the RTS pin is driven high when the transmitter is operating. The behavior of the RTS pin is controlled by the TXEMPTY bit. A typical connection of the USART to a RS485 bus is shown in Figure 21-35. Figure 21-35. Typical Connection to a RS485 Bus USART RXD TXD RTS Differential Bus The USART is set in RS485 mode by programming the MODE field in the Mode Register (MR) to the value 0x1. The RTS pin is at a level inverse to the TXEMPTY bit. Significantly, the RTS pin remains high when a timeguard is programmed so that the line can remain driven after the last character completion. Figure 21-36 gives an example of the RTS waveform during a character transmission when the timeguard is enabled. Figure 21-36. Example of RTS Drive with Timeguard TG = 4 Baud Rate Clock TXD Start D0 Bit D1 D2 D3 D4 D5 D6 D7 Parity Stop Bit Bit Write THR TXRDY TXEMPTY RTS 331 32059I-06/2010 AT32UC3B 21.6.7 Modem Mode The USART features modem mode, which enables control of the signals: DTR (Data Terminal Ready), DSR (Data Set Ready), RTS (Request to Send), CTS (Clear to Send), DCD (Data Carrier Detect) and RI (Ring Indicator). While operating in modem mode, the USART behaves as a DTE (Data Terminal Equipment) as it drives DTR and RTS and can detect level change on DSR, DCD, CTS and RI. Setting the USART in modem mode is performed by writing the MODE field in the Mode Register (MR) to the value 0x3. While operating in modem mode the USART behaves as though in asynchronous mode and all the parameter configurations are available. Table 21-12 gives the correspondence of the USART signals with modem connection standards. Table 21-12. Circuit References USART Pin TXD RTS DTR RXD CTS DSR DCD RI V24 2 4 20 3 5 6 8 22 CCITT 103 105 108.2 104 106 107 109 125 Direction From terminal to modem From terminal to modem From terminal to modem From modem to terminal From terminal to modem From terminal to modem From terminal to modem From terminal to modem The control of the DTR output pin is performed by writing the Control Register (CR) with the DTRDIS and DTREN bits respectively at 1. The disable command forces the corresponding pin to its inactive level, i.e. high. The enable command forces the corresponding pin to its active level, i.e. low. RTS output pin is automatically controlled in this mode The level changes are detected on the RI, DSR, DCD and CTS pins. If an input change is detected, the RIIC, DSRIC, DCDIC and CTSIC bits in the Channel Status Register (CSR) are set respectively and can trigger an interrupt. The status is automatically cleared when CSR is read. Furthermore, the CTS automatically disables the transmitter when it is detected at its inactive state. If a character is being transmitted when the CTS rises, the character transmission is completed before the transmitter is actually disabled. 332 32059I-06/2010 AT32UC3B 21.6.8 SPI Mode The Serial Peripheral Interface (SPI) Mode is a synchronous serial data link that provides communication with external devices in Master or Slave Mode. It also enables communication between processors if an external processor is connected to the system. The Serial Peripheral Interface is essentially a shift register that serially transmits data bits to other SPIs. During a data transfer, one SPI system acts as the "master" which controls the data flow, while the other devices act as "slaves'' which have data shifted into and out by the master. Different CPUs can take turns being masters and one master may simultaneously shift data into multiple slaves. (Multiple Master Protocol is the opposite of Single Master Protocol, where one CPU is always the master while all of the others are always slaves.) However, only one slave may drive its output to write data back to the master at any given time. A slave device is selected when its NSS signal is asserted by the master. The USART in SPI Master mode can address only one SPI Slave because it can generate only one NSS signal. The SPI system consists of two data lines and two control lines: * Master Out Slave In (MOSI): This data line supplies the output data from the master shifted into the input of the slave. * Master In Slave Out (MISO): This data line supplies the output data from a slave to the input of the master. * Serial Clock (CLK): This control line is driven by the master and regulates the flow of the data bits. The master may transmit data at a variety of baud rates. The CLK line cycles once for each bit that is transmitted. * Slave Select (NSS): This control line allows the master to select or deselect the slave. 21.6.8.1 Modes of Operation The USART can operate in Master Mode or in Slave Mode. Operation in SPI Master Mode is programmed by writing at 0xE the MODE field in the Mode Register. In this case the SPI lines must be connected as described below: * the MOSI line is driven by the output pin TXD * the MISO line drives the input pin RXD * the CLK line is driven by the output pin CLK * the NSS line is driven by the output pin RTS Operation in SPI Slave Mode is programmed by writing at 0xF the MODE field in the Mode Register. In this case the SPI lines must be connected as described below: * the MOSI line drives the input pin RXD * the MISO line is driven by the output pin TXD * the CLK line drives the input pin CLK * the NSS line drives the input pin CTS In order to avoid unpredicted behavior, any change of the SPI Mode must be followed by a software reset of the transmitter and of the receiver (except the initial configuration after a hardware reset). 333 32059I-06/2010 AT32UC3B 21.6.8.2 Baud Rate In SPI Mode, the baudrate generator operates in the same way as in USART synchronous mode: See Section "21.6.1.4" on page 307. However, there are some restrictions: In SPI Master Mode: * the external clock CLK must not be selected (USCLKS ... 0x3), and the bit CLKO must be set to "1" in the Mode Register (MR), in order to generate correctly the serial clock on the CLK pin. * to obtain correct behavior of the receiver and the transmitter, the value programmed in CD of must be superior or equal to 4. * if the internal clock divided (CLK_USART/DIV) is selected, the value programmed in CD must be even to ensure a 50:50 mark/space ratio on the CLK pin, this value can be odd if the internal clock is selected (CLK_USART). In SPI Slave Mode: * the external clock (CLK) selection is forced regardless of the value of the USCLKS field in the Mode Register (MR). Likewise, the value written in BRGR has no effect, because the clock is provided directly by the signal on the USART CLK pin. * to obtain correct behavior of the receiver and the transmitter, the external clock (CLK) frequency must be at least 4 times lower than the system clock. 334 32059I-06/2010 AT32UC3B 21.6.8.3 Data Transfer Up to 9 data bits are successively shifted out on the TXD pin at each rising or falling edge (depending of CPOL and CPHA) of the programmed serial clock. There is no Start bit, no Parity bit and no Stop bit. The number of data bits is selected by the CHRL field and the MODE 9 bit in the Mode Register (MR). The 9 bits are selected by setting the MODE 9 bit regardless of the CHRL field. The MSB data bit is always sent first in SPI Mode (Master or Slave). Four combinations of polarity and phase are available for data transfers. The clock polarity is programmed with the CPOL bit in the Mode Register. The clock phase is programmed with the CPHA bit. These two parameters determine the edges of the clock signal upon which data is driven and sampled. Each of the two parameters has two possible states, resulting in four possible combinations that are incompatible with one another. Thus, a master/slave pair must use the same parameter pair values to communicate. If multiple slaves are used and fixed in different configurations, the master must reconfigure itself each time it needs to communicate with a different slave. Table 21-13. SPI Bus Protocol Mode SPI Bus Protocol Mode 0 1 2 3 CPOL 0 0 1 1 CPHA 1 0 1 0 335 32059I-06/2010 AT32UC3B Figure 21-37. SPI Transfer Format (CPHA=1, 8 bits per transfer) CLK cycle (for reference) CLK (CPOL= 0) 1 2 3 4 5 6 7 8 CLK (CPOL= 1) MOSI SPI Master ->TXD SPI Slave ->RXD MSB 6 5 4 3 2 1 LSB MISO SPI Master ->RXD SPI Slave ->TXD MSB 6 5 4 3 2 1 LSB NSS SPI Master ->RTS SPI Slave ->CTS Figure 21-38. SPI Transfer Format (CPHA=0, 8 bits per transfer) CLK cycle (for reference) CLK (CPOL= 0) 1 2 3 4 5 6 7 8 CLK (CPOL= 1) MOSI SPI Master -> TXD SPI Slave -> RXD MSB 6 5 4 3 2 1 LSB MISO SPI Master -> RXD SPI Slave -> TXD MSB 6 5 4 3 2 1 LSB NSS SPI Master -> RTS SPI Slave -> CTS 336 32059I-06/2010 AT32UC3B 21.6.8.4 Receiver and Transmitter Control See Section "21.6.2" on page 309. 21.6.8.5 Character Transmission The characters are sent by writing in the Transmit Holding Register (THR). The transmitter reports two status bits in the Channel Status Register (CSR): TXRDY (Transmitter Ready), which indicates that THR is empty and TXEMPTY, which indicates that all the characters written in THR have been processed. When the current character processing is completed, the last character written in THR is transferred into the Shift Register of the transmitter and THR becomes empty, thus TXRDY rises. Both TXRDY and TXEMPTY bits are low when the transmitter is disabled. Writing a character in THR while TXRDY is low has no effect and the written character is lost. If the USART is in SPI Slave Mode and if a character must be sent while the Transmit Holding Register (THR) is empty, the UNRE (Underrun Error) bit is set. The TXD transmission line stays at high level during all this time. The UNRE bit is cleared by writing the Control Register (CR) with the RSTSTA (Reset Status) bit at 1. In SPI Master Mode, the slave select line (NSS) is asserted at low level 1 Tbit before the transmission of the MSB bit and released at high level 1 Tbit after the transmission of the LSB bit. So, the slave select line (NSS) is always released between each character transmission and a minimum delay of 3 Tbits always inserted. However, in order to address slave devices supporting the CSAAT mode (Chip Select Active After Transfer), the slave select line (NSS) can be forced at low level by writing the Control Register (CR) with the RTSEN bit at 1. The slave select line (NSS) can be released at high level only by writing the Control Register (CR) with the RTSDIS bit at 1 (for example, when all data have been transferred to the slave device). In SPI Slave Mode, the transmitter does not require a falling edge of the slave select line (NSS) to initiate a character transmission but only a low level. However, this low level must be present on the slave select line (NSS) at least 1 Tbit before the first serial clock cycle corresponding to the MSB bit. 21.6.8.6 Character Reception When a character reception is completed, it is transferred to the Receive Holding Register (RHR) and the RXRDY bit in the Status Register (CSR) rises. If a character is completed while RXRDY is set, the OVRE (Overrun Error) bit is set. The last character is transferred into RHR and overwrites the previous one. The OVRE bit is cleared by writing the Control Register (CR) with the RSTSTA (Reset Status) bit at 1. To ensure correct behavior of the receiver in SPI Slave Mode, the master device sending the frame must ensure a minimum delay of 1 Tbit between each character transmission. The receiver does not require a falling edge of the slave select line (NSS) to initiate a character reception but only a low level. However, this low level must be present on the slave select line (NSS) at least 1 Tbit before the first serial clock cycle corresponding to the MSB bit. 21.6.8.7 Receiver Timeout Because the receiver baudrate clock is active only during data transfers in SPI Mode, a receiver timeout is impossible in this mode, whatever the Time-out value is (field TO) in the Time-out Register (RTOR). 337 32059I-06/2010 AT32UC3B 21.6.9 Test Modes The USART can be programmed to operate in three different test modes. The internal loopback capability allows on-board diagnostics. In the loopback mode the USART interface pins are disconnected or not and reconfigured for loopback internally or externally. 21.6.9.1 Normal Mode Normal mode connects the RXD pin on the receiver input and the transmitter output on the TXD pin. Figure 21-39. Normal Mode Configuration RXD Receiver TXD Transmitter 21.6.9.2 Automatic Echo Mode Automatic echo mode allows bit-by-bit retransmission. When a bit is received on the RXD pin, it is sent to the TXD pin, as shown in Figure 21-40. Programming the transmitter has no effect on the TXD pin. The RXD pin is still connected to the receiver input, thus the receiver remains active. Figure 21-40. Automatic Echo Mode Configuration RXD Receiver TXD Transmitter 21.6.9.3 Local Loopback Mode Local loopback mode connects the output of the transmitter directly to the input of the receiver, as shown in Figure 21-41. The TXD and RXD pins are not used. The RXD pin has no effect on the receiver and the TXD pin is continuously driven high, as in idle state. Figure 21-41. Local Loopback Mode Configuration RXD Receiver Transmitter 1 TXD 338 32059I-06/2010 AT32UC3B 21.6.9.4 Remote Loopback Mode Remote loopback mode directly connects the RXD pin to the TXD pin, as shown in Figure 21-42. The transmitter and the receiver are disabled and have no effect. This mode allows bit-by-bit retransmission. Figure 21-42. Remote Loopback Mode Configuration Receiver 1 RXD TXD Transmitter 339 32059I-06/2010 AT32UC3B 21.7 User Interface Table 21-14. USART Register Memory Map Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0024 0x0028 0x0040 0x0044 0x004C 0x0050 0x00FC Register Control Register Mode Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Channel Status Register Receiver Holding Register Transmitter Holding Register Baud Rate Generator Register Receiver Time-out Register Transmitter Timeguard Register FI DI Ratio Register Number of Errors Register IrDA Filter Register Manchester Encoder Decoder Register Version Register Note: Name CR MR IER IDR IMR CSR RHR THR BRGR RTOR TTGR FIDI NER IFR MAN VERSION Access Write-only Read-write Write-only Write-only Read-only Read-only Read-only Write-only Read-write Read-write Read-write Read-write Read-only Read-write Read-write Read-only Reset - 0x00000000 - - 0x00000000 0x00000000 0x00000000 - 0x00000000 0x00000000 0x00000000 0x00000174 0x00000000 0x00000000 0x30011004 0x-(1) 1. Values in the Version Register vary with the version of the IP block implementation. 340 32059I-06/2010 AT32UC3B 21.7.1 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 RETTO 7 TXDIS Control Register CR Write-only 0x0 30 - 22 - 14 RSTNACK 6 TXEN 29 - 21 - 13 RSTIT 5 RXDIS 28 - 20 - 12 SENDA 4 RXEN 27 - 19 RTSDIS/RCS 11 STTTO 3 RSTTX 26 - 18 RTSEN/FCS 10 STPBRK 2 RSTRX 25 - 17 DTRDIS 9 STTBRK 1 - 24 - 16 DTREN 8 RSTSTA 0 - * RTSDIS/RCS: Request to Send Disable/Release SPI Chip Select If USART does not operate in SPI Master Mode (MODE ... 0xE): 0: No effect. 1: Drives the pin RTS to 1. If USART operates in SPI Master Mode (MODE = 0xE): RCS = 0: No effect. RCS = 1: Releases the Slave Select Line NSS (RTS pin). RTSEN/FCS: Request to Send Enable/Force SPI Chip Select If USART does not operate in SPI Master Mode (MODE ... 0xE): 0: No effect. 1: Drives the pin RTS to 0. If USART operates in SPI Master Mode (MODE = 0xE): FCS = 0: No effect. FCS = 1: Forces the Slave Select Line NSS (RTS pin) to 0, even if USART is no transmitting, in order to address SPI slave devices supporting the CSAAT Mode (Chip Select Active After Transfer). DTRDIS: Data Terminal Ready Disable 0: No effect. 1: Drives the pin DTR to 1. DTREN: Data Terminal Ready Enable 0: No effect. 1: Drives the pin DTR at 0. RETTO: Rearm Time-out 0: No effect 1: Restart Time-out RSTNACK: Reset Non Acknowledge 0: No effect 1: Resets NACK in CSR. RSTIT: Reset Iterations 0: No effect. * * * * * * 341 32059I-06/2010 AT32UC3B 1: Resets ITERATION in CSR. No effect if the ISO7816 is not enabled. * SENDA: Send Address 0: No effect. 1: In Multidrop Mode only, the next character written to the THR is sent with the address bit set. STTTO: Start Time-out 0: No effect. 1: Starts waiting for a character before clocking the time-out counter. Resets the status bit TIMEOUT in CSR. STPBRK: Stop Break 0: No effect. 1: Stops transmission of the break after a minimum of one character length and transmits a high level during 12-bit periods. No effect if no break is being transmitted. STTBRK: Start Break 0: No effect. 1: Starts transmission of a break after the characters present in THR and the Transmit Shift Register have been transmitted. No effect if a break is already being transmitted. RSTSTA: Reset Status Bits 0: No effect. 1: Resets the status bits PARE, FRAME, OVRE, MANERR, and RXBRK in CSR. TXDIS: Transmitter Disable 0: No effect. 1: Disables the transmitter. TXEN: Transmitter Enable 0: No effect. 1: Enables the transmitter if TXDIS is 0. RXDIS: Receiver Disable 0: No effect. 1: Disables the receiver. RXEN: Receiver Enable 0: No effect. 1: Enables the receiver, if RXDIS is 0. RSTTX: Reset Transmitter 0: No effect. 1: Resets the transmitter. RSTRX: Reset Receiver 0: No effect. 1: Resets the receiver. * * * * * * * * * * 342 32059I-06/2010 AT32UC3B 21.7.2 Name: Access Type: Offset: Reset Value: 31 ONEBIT 23 - 15 CHMODE 7 CHRL 6 5 USCLKS Mode Register MR Read-write 0x4 30 MODSYNC 22 VAR_SYNC 14 29 MAN 21 DSNACK 13 NBSTOP 4 3 28 FILTER 20 INACK 12 27 - 19 OVER 11 26 25 MAX_ITERATION 17 MODE9 9 24 18 CLKO 10 PAR 2 MODE 16 MSBF/CPOL 8 SYNC/CPHA 0 1 This register can only be written if the WPEN bit is cleared in the Write Protect Mode Register(if exists). * ONEBIT: Start Frame Delimiter Selector 0: Start Frame delimiter is COMMAND or DATA SYNC. 1: Start Frame delimiter is One Bit. MODSYNC: Manchester Synchronization Mode 0:The Manchester Start bit is a 0 to 1 transition 1: The Manchester Start bit is a 1 to 0 transition. MAN: Manchester Encoder/Decoder Enable 0: Manchester Encoder/Decoder are disabled. 1: Manchester Encoder/Decoder are enabled. FILTER: Infrared Receive Line Filter 0: The USART does not filter the receive line. 1: The USART filters the receive line using a three-sample filter (1/16-bit clock) (2 over 3 majority). MAX_ITERATION Defines the maximum number of iterations in mode ISO7816, protocol T= 0. VAR_SYNC: Variable Synchronization of Command/Data Sync Start Frame Delimiter 0: User defined configuration of command or data sync field depending on SYNC value. 1: The sync field is updated when a character is written into THR register. DSNACK: Disable Successive NACK 0: NACK is sent on the ISO line as soon as a parity error occurs in the received character (unless INACK is set). 1: Successive parity errors are counted up to the value specified in the MAX_ITERATION field. These parity errors generate a NACK on the ISO line. As soon as this value is reached, no additional NACK is sent on the ISO line. The flag ITERATION is asserted. INACK: Inhibit Non Acknowledge 0: The NACK is generated. 1: The NACK is not generated. OVER: Oversampling Mode 0: 16x Oversampling. 1: 8x Oversampling. * * * * * * * * 343 32059I-06/2010 AT32UC3B * CLKO: Clock Output Select 0: The USART does not drive the CLK pin. 1: The USART drives the CLK pin if USCLKS does not select the external clock CLK. * MODE9: 9-bit Character Length 0: CHRL defines character length. 1: 9-bit character length. * MSBF/CPOL: Bit Order or SPI Clock Polarity If USART does not operate in SPI Mode (MODE ... 0xE and 0xF): MSBF = 0: Least Significant Bit is sent/received first. MSBF = 1: Most Significant Bit is sent/received first. If USART operates in SPI Mode (Slave or Master, MODE = 0xE or 0xF): CPOL = 0: The inactive state value of SPCK is logic level zero. CPOL = 1: The inactive state value of SPCK is logic level one. CPOL is used to determine the inactive state value of the serial clock (SPCK). It is used with CPHA to produce the required clock/data relationship between master and slave devices. * CHMODE: Channel Mode Table 21-15. CHMODE 0 0 1 1 0 1 0 1 Mode Description Normal Mode Automatic Echo. Receiver input is connected to the TXD pin. Local Loopback. Transmitter output is connected to the Receiver Input. Remote Loopback. RXD pin is internally connected to the TXD pin. * NBSTOP: Number of Stop Bits Table 21-16. NBSTOP 0 0 1 1 0 1 0 1 Asynchronous (SYNC = 0) 1 stop bit 1.5 stop bits 2 stop bits Reserved Synchronous (SYNC = 1) 1 stop bit Reserved 2 stop bits Reserved * PAR: Parity Type Table 21-17. PAR 0 0 0 0 1 1 0 0 1 1 0 1 0 1 0 1 x x Parity Type Even parity Odd parity Parity forced to 0 (Space) Parity forced to 1 (Mark) No parity Multidrop mode 344 32059I-06/2010 AT32UC3B * SYNC/CPHA: Synchronous Mode Select or SPI Clock Phase If USART does not operate in SPI Mode (MODE is ... 0xE and 0xF): SYNC = 0: USART operates in Asynchronous Mode. SYNC = 1: USART operates in Synchronous Mode. If USART operates in SPI Mode (MODE = 0xE or 0xF): CPHA = 0: Data is changed on the leading edge of SPCK and captured on the following edge of SPCK. CPHA = 1: Data is captured on the leading edge of SPCK and changed on the following edge of SPCK. CPHA determines which edge of SPCK causes data to change and which edge causes data to be captured. CPHA is used with CPOL to produce the required clock/data relationship between master and slave devices. * CHRL: Character Length. Table 21-18. CHRL 0 0 1 1 0 1 0 1 Character Length 5 bits 6 bits 7 bits 8 bits * USCLKS: Clock Selection Table 21-19. USCLKS 0 0 1 1 0 1 0 1 Note: Selected Clock CLK_USART CLK_USART/DIV(1) Reserved CLK 1. The value of DIV is device dependent. Please refer to the Module Configuration section at the end of this chapter. * MODE Table 21-20. MODE 0 0 0 0 0 0 1 1 1 0 0 0 0 1 1 0 1 1 Others 0 0 1 1 0 1 0 1 1 0 1 0 1 0 0 0 0 1 Mode of the USART Normal RS485 Hardware Handshaking Modem IS07816 Protocol: T = 0 IS07816 Protocol: T = 1 IrDA SPI Master SPI Slave Reserved 345 32059I-06/2010 AT32UC3B 21.7.3 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 PARE Interrupt Enable Register IER Write-only 0x8 30 - 22 - 14 - 6 FRAME 29 - 21 - 13 NACK 5 OVRE 28 - 20 MANE 12 RXBUFF 4 - 27 - 19 CTSIC 11 - 3 - 26 - 18 DCDIC 10 ITER/UNRE 2 RXBRK 25 - 17 DSRIC 9 TXEMPTY 1 TXRDY 24 MANEA 16 RIIC 8 TIMEOUT 0 RXRDY Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. For backward compatibility the MANE bit has been duplicated to the MANEA bit position. Writing either one or the other has the same effect. 346 32059I-06/2010 AT32UC3B 21.7.4 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 PARE Interrupt Disable Register IDR Write-only 0xC 30 - 22 - 14 - 6 FRAME 29 - 21 - 13 NACK 5 OVRE 28 - 20 MANE 12 RXBUFF 4 - 27 - 19 CTSIC 11 - 3 - 26 - 18 DCDIC 10 ITER/UNRE 2 RXBRK 25 - 17 DSRIC 9 TXEMPTY 1 TXRDY 24 MANEA 16 RIIC 8 TIMEOUT 0 RXRDY Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. For backward compatibility the MANE bit has been duplicated to the MANEA bit position. Writing either one or the other has the same effect. 347 32059I-06/2010 AT32UC3B 21.7.5 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 PARE Interrupt Mask Register IMR Read-only 0x10 30 - 22 - 14 - 6 FRAME 29 - 21 - 13 NACK 5 OVRE 28 - 20 MANE 12 RXBUFF 4 - 27 - 19 CTSIC 11 - 3 - 26 - 18 DCDIC 10 ITER/UNRE 2 RXBRK 25 - 17 DSRIC 9 TXEMPTY 1 TXRDY 24 MANEA 16 RIIC 8 TIMEOUT 0 RXRDY 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is set when the corresponding bit in IER is written to one. For backward compatibility the MANE bit has been duplicated to the MANEA bit position. Reading either one or the other has the same effect. 348 32059I-06/2010 AT32UC3B 21.7.6 Name: Access Type: Offset: Reset Value: 31 - 23 CTS 15 - 7 PARE Channel Status Register CSR Read-only 0x14 30 - 22 DCD 14 - 6 FRAME 29 - 21 DSR 13 NACK 5 OVRE 28 - 20 RI 12 RXBUFF 4 - 27 - 19 CTSIC 11 - 3 - 26 - 18 DCDIC 10 ITER/UNRE 2 RXBRK 25 - 17 DSRIC 9 TXEMPTY 1 TXRDY 24 MANERR 16 RIIC 8 TIMEOUT 0 RXRDY * MANERR: Manchester Error 0: No Manchester error has been detected since the last RSTSTA. 1: At least one Manchester error has been detected since the last RSTSTA. * CTS: Image of CTS Input 0: CTS is at 0. 1: CTS is at 1. * DCD: Image of DCD Input 0: DCD is at 0. 1: DCD is at 1. * DSR: Image of DSR Input 0: DSR is at 0 1: DSR is at 1. 349 32059I-06/2010 AT32UC3B * RI: Image of RI Input 0: RI is at 0. 1: RI is at 1. CTSIC: Clear to Send Input Change Flag 0: No input change has been detected on the CTS pin since the last read of CSR. 1: At least one input change has been detected on the CTS pin since the last read of CSR. DCDIC: Data Carrier Detect Input Change Flag 0: No input change has been detected on the DCD pin since the last read of CSR. 1: At least one input change has been detected on the DCD pin since the last read of CSR. DSRIC: Data Set Ready Input Change Flag 0: No input change has been detected on the DSR pin since the last read of CSR. 1: At least one input change has been detected on the DSR pin since the last read of CSR. RIIC: Ring Indicator Input Change Flag 0: No input change has been detected on the RI pin since the last read of CSR. 1: At least one input change has been detected on the RI pin since the last read of CSR. NACK: Non Acknowledge 0: No Non Acknowledge has not been detected since the last RSTNACK. 1: At least one Non Acknowledge has been detected since the last RSTNACK. RXBUFF: Reception Buffer Full 0: The signal Buffer Full from the Receive Peripheral DMA Controller channel is inactive. 1: The signal Buffer Full from the Receive Peripheral DMA Controller channel is active. ITER/UNRE: Max number of Repetitions Reached or SPI Underrun Error If USART does not operate in SPI Slave Mode (MODE ... 0xF): ITER = 0: Maximum number of repetitions has not been reached since the last RSTSTA. ITER = 1: Maximum number of repetitions has been reached since the last RSTSTA. If USART operates in SPI Slave Mode (MODE = 0xF): UNRE = 0: No SPI underrun error has occurred since the last RSTSTA. UNRE = 1: At least one SPI underrun error has occurred since the last RSTSTA. TXEMPTY: Transmitter Empty 0: There are characters in either THR or the Transmit Shift Register, or the transmitter is disabled. 1: There are no characters in THR, nor in the Transmit Shift Register. TIMEOUT: Receiver Time-out 0: There has not been a time-out since the last Start Time-out command (STTTO in CR) or the Time-out Register is 0. 1: There has been a time-out since the last Start Time-out command (STTTO in CR). PARE: Parity Error 0: No parity error has been detected since the last RSTSTA. 1: At least one parity error has been detected since the last RSTSTA. FRAME: Framing Error 0: No stop bit has been detected low since the last RSTSTA. 1: At least one stop bit has been detected low since the last RSTSTA. OVRE: Overrun Error 0: No overrun error has occurred since the last RSTSTA. 1: At least one overrun error has occurred since the last RSTSTA. RXBRK: Break Received/End of Break 0: No Break received or End of Break detected since the last RSTSTA. 1: Break Received or End of Break detected since the last RSTSTA. TXRDY: Transmitter Ready 0: A character is in the THR waiting to be transferred to the Transmit Shift Register, or an STTBRK command has been requested, or the transmitter is disabled. As soon as the transmitter is enabled, TXRDY becomes 1. 1: There is no character in the THR. * * * * * * * * * * * * * * 350 32059I-06/2010 AT32UC3B * RXRDY: Receiver Ready 0: No complete character has been received since the last read of RHR or the receiver is disabled. If characters were being received when the receiver was disabled, RXRDY changes to 1 when the receiver is enabled. 1: At least one complete character has been received and RHR has not yet been read. 351 32059I-06/2010 AT32UC3B 21.7.7 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 RXSYNH 7 Receive Holding Register RHR Read-only 0x18 0x00000000 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 RXCHR 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 - 1 24 - 16 - 8 RXCHR 0 * RXSYNH: Received Sync 0: Last Character received is a Data. 1: Last Character received is a Command. * RXCHR: Received Character Last character received if RXRDY is set. 352 32059I-06/2010 AT32UC3B 21.7.8 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 TXSYNH 7 USART Transmit Holding Register THR Write-only 0x1C 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 TXCHR 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 - 1 24 - 16 - 8 TXCHR 0 * TXSYNH: Sync Field to be transmitted 0: The next character sent is encoded as a data. Start Frame Delimiter is DATA SYNC. 1: The next character sent is encoded as a command. Start Frame Delimiter is COMMAND SYNC. * TXCHR: Character to be Transmitted Next character to be transmitted after the current character if TXRDY is not set. 353 32059I-06/2010 AT32UC3B 21.7.9 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 Baud Rate Generator Register BRGR Read-write 0x20 0x00000000 30 - 22 - 14 29 - 21 - 13 28 - 20 - 12 CD 7 6 5 4 CD 3 2 1 0 27 - 19 - 11 26 - 18 25 - 17 FP 9 24 - 16 10 8 * FP: Fractional Part 0: Fractional divider is disabled. 1 - 7: Baudrate resolution, defined by FP x 1/8. * CD: Clock Divider Table 21-21. MODE ISO7816 SYNC = 1 or MODE = SPI (Master or Slave) OVER = 1 Baud Rate Clock Disabled Baud Rate = Selected Clock/16/CD Baud Rate = Selected Clock/8/CD Baud Rate = Selected Clock /CD Baud Rate = Selected Clock/CD/FI_DI_RATIO MODE = ISO7816 SYNC = 0 CD 0 1 to 65535 OVER = 0 354 32059I-06/2010 AT32UC3B 21.7.10 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 Receiver Time-out Register RTOR Read-write 0x24 0x00000000 30 - 22 - 14 29 - 21 - 13 28 - 20 - 12 TO 27 - 19 - 11 26 - 18 - 10 25 - 17 - 9 24 - 16 - 8 7 6 5 4 TO 3 2 1 0 * TO: Time-out Value 0: The Receiver Time-out is disabled. 1 - 65535: The Receiver Time-out is enabled and the Time-out delay is TO x Bit Period. Note that the size of the TO counter can change depending of implementation. See the Module Configuration section. 355 32059I-06/2010 AT32UC3B 21.7.11 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 Transmitter Timeguard Register TTGR Read-write 0x28 0x00000000 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 TG 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 - 1 24 - 16 - 8 - 0 * TG: Timeguard Value 0: The Transmitter Timeguard is disabled. 1 - 255: The Transmitter timeguard is enabled and the timeguard delay is TG x Bit Period. 356 32059I-06/2010 AT32UC3B 21.7.12 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 FI DI RATIO Register FIDI Read-write 0x40 0x00000174 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 FI_DI_RATIO 27 - 19 - 11 - 3 26 - 18 - 10 25 - 17 - 9 FI_DI_RATIO 1 24 - 16 - 8 2 0 This register can only be written if the WPEN bit is cleared in the Write Protect Mode Register. * FI_DI_RATIO: FI Over DI Ratio Value 0: If ISO7816 mode is selected, the Baud Rate Generator generates no signal. 1 - 2047: If ISO7816 mode is selected, the Baud Rate is the clock provided on CLK divided by FI_DI_RATIO. 357 32059I-06/2010 AT32UC3B 21.7.13 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 Number of Errors Register NER Read-only 0x44 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 NB_ERRORS 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 - 1 24 - 16 - 8 - 0 * NB_ERRORS: Number of Errors Total number of errors that occurred during an ISO7816 transfer. This register automatically clears when read. 358 32059I-06/2010 AT32UC3B 21.7.14 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 IrDA FILTER Register IFR Read-write 0x4C 0x00000000 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 IRDA_FILTER 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 - 1 24 - 16 - 8 - 0 This register can only be written if the WPEN bit is cleared in the Write Protect Mode Register(if exists). IRDA_FILTER: IrDA Filter Sets the filter of the IrDA demodulator. 359 32059I-06/2010 AT32UC3B 21.7.15 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - Manchester Configuration Register MAN Read-write 0x50 0x30011004 30 DRIFT 22 - 14 - 6 - 29 1 21 - 13 - 5 - 28 RX_MPOL 20 - 12 TX_MPOL 4 - 27 - 19 26 - 18 RX_PL 11 - 3 10 - 2 TX_PL 9 TX_PP 1 0 8 25 RX_PP 17 16 24 This register can only be written if the WPEN bit is cleared in the Write Protect Mode Register(if exists). * DRIFT: Drift compensation 0: The USART can not recover from an important clock drift 1: The USART can recover from clock drift. The 16X clock mode must be enabled. * RX_MPOL: Receiver Manchester Polarity 0: Logic Zero is coded as a zero-to-one transition, Logic One is coded as a one-to-zero transition. 1: Logic Zero is coded as a one-to-zero transition, Logic One is coded as a zero-to-one transition. * RX_PP: Receiver Preamble Pattern detected Table 21-22. RX_PP 0 0 1 1 0 1 0 1 Preamble Pattern default polarity assumed (RX_MPOL field not set) ALL_ONE ALL_ZERO ZERO_ONE ONE_ZERO * RX_PL: Receiver Preamble Length 0: The receiver preamble pattern detection is disabled 1 - 15: The detected preamble length is RX_PL x Bit Period * TX_MPOL: Transmitter Manchester Polarity 0: Logic Zero is coded as a zero-to-one transition, Logic One is coded as a one-to-zero transition. 1: Logic Zero is coded as a one-to-zero transition, Logic One is coded as a zero-to-one transition. 360 32059I-06/2010 AT32UC3B * TX_PP: Transmitter Preamble Pattern Table 21-23. TX_PP 0 0 1 1 0 1 0 1 Preamble Pattern default polarity assumed (TX_MPOL field not set) ALL_ONE ALL_ZERO ZERO_ONE ONE_ZERO * TX_PL: Transmitter Preamble Length 0: The Transmitter Preamble pattern generation is disabled 1 - 15: The Preamble Length is TX_PL x Bit Period 361 32059I-06/2010 AT32UC3B 362 32059I-06/2010 AT32UC3B 21.7.16 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 Version Register VERSION Read-only 0xFC 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 VERSION 27 - 19 26 - 18 VARIANT 11 10 VERSION 3 2 1 0 9 8 25 - 17 24 - 16 * VARIANT Reserved. No functionality associated. * VERSION Version of the module. No functionality associated. 363 32059I-06/2010 AT32UC3B 21.8 Module Configuration The specific configuration for each USART instance is listed in the following tables.The module bus clocks listed here are connected to the system bus clocks according to the table in the System Bus Clock Connections section. Table 21-24. Module Configuration Feature SPI Logic RS485 Logic Manchester Logic Modem Logic IRDA Logic Fractional Baudrate ISO7816 DIV Receiver Time-out Counter Size USART0 Implemented Not Implemented Not Implemented Not Implemented Not Implemented Implemented Not Implemented 8 8-bits USART1 Implemented Not Implemented Implemented Implemented Implemented Implemented Implemented 8 17-bits USART2 Implemented Not Implemented Not Implemented Not Implemented Not Implemented Implemented Not Implemented 8 8-bits Table 21-25. Module Clock Name Module name USART0 USART1 USART2 Clock name CLK_USART0 CLK_USART1 CLK_USART2 364 32059I-06/2010 AT32UC3B 22. USB On-The-Go Interface (USBB) Rev: 3.1.0.1.6 22.1 Features * * * * * * * * Compatible with the USB 2.0 specification Supports Full (12Mbit/s) and Low (1.5 Mbit/s) speed Device and Embedded Host seven pipes/endpoints 960 of Embedded Dual-Port RAM (DPRAM) for Pipes/Endpoints Up to 2 memory banks per Pipe/Endpoint (Not for Control Pipe/Endpoint) Flexible Pipe/Endpoint configuration and management with dedicated DMA channels On-Chip transceivers including Pull-Ups/Pull-downs On-Chip pad including VBUS analog comparator 22.2 Overview The Universal Serial Bus (USB) MCU device complies with the Universal Serial Bus (USB) 2.0 specification, but it does NOT feature Hi-Speed USB (480 Mbit/s). Each pipe/endpoint can be configured in one of several transfer types. It can be associated with one or more banks of a dual-port RAM (DPRAM) used to store the current data payload. If several banks are used ("ping-pong" mode), then one DPRAM bank is read or written by the CPU or the DMA while the other is read or written by the USBB core. This feature is mandatory for isochronous pipes/endpoints. Table 22-1 on page 365 describes the hardware configuration of the USB MCU device. Table 22-1. Description of USB Pipes/Endpoints Mnemonic PEP0 PEP1 PEP2 PEP3 PEP4 PEP5 PEP6 Max. Size 64 bytes 64 bytes 64 bytes 64 bytes 64 bytes 256 bytes 256 bytes Max. Nb. Banks 1 2 2 2 2 2 2 DMA N Y Y Y Y Y Y Type Control Isochronous/Bulk/Interrupt/Control Isochronous/Bulk/Interrupt/Control Isochronous/Bulk/Interrupt/Control Isochronous/Bulk/Interrupt/Control Isochronous/Bulk/Interrupt/Control Isochronous/Bulk/Interrupt/Control Pipe/Endpoint 0 1 2 3 4 5 6 The theoretical maximal pipe/endpoint configuration (1600) exceeds the real DPRAM size (960). The user needs to be aware of this when configuring pipes/endpoints. To fully use the 960 of DPRAM, the user could for example use the configuration described in Table 22-2 on page 365. Table 22-2. Example of Configuration of Pipes/Endpoints Using the Whole DPRAM Mnemonic PEP0 PEP1 PEP2 PEP3 Size 64 bytes 64 bytes 64 bytes 64 bytes Nb. Banks 1 1 1 2 Pipe/Endpoint 0 1 2 3 365 32059I-06/2010 AT32UC3B Table 22-2. Example of Configuration of Pipes/Endpoints Using the Whole DPRAM Mnemonic PEP4 PEP5 PEP6 Size 64 bytes 256 bytes 256 bytes Nb. Banks 2 1 1 Pipe/Endpoint 4 5 6 22.3 Block Diagram The USBB provides a hardware device to interface a USB link to a data flow stored in a dual-port RAM (DPRAM). The USBB requires a 48MHz 0.25% reference clock, which is the USB generic clock generated from one of the power manager oscillators, optionally through one of the power manager PLLs. The 48MHz clock is used to generate a 12MHz full-speed (or 1.5 MHz low-speed) bit clock from the received USB differential data and to transmit data according to full- or low-speed USB device tolerance. Clock recovery is achieved by a digital phase-locked loop (a DPLL, not represented), which complies with the USB jitter specifications. 366 32059I-06/2010 AT32UC3B Figure 22-1. USBB Block Diagram USB 32 bits DPRAM Local HSB Slave Interface Slave HSB MUX HSB Master HSB0 DMA HSB1 PB PEP Allocation VBUS DUSB 2.0 Core D+ I/O Controller User Interface Interrupt Controller USB Interrupts USB_ID USB_VBOF Power Manager USB GCLK @ 48 MHz System Clock USB Clock Domain Domain 367 32059I-06/2010 AT32UC3B 22.4 Application Block Diagram Depending on the USB operating mode (device-only, reduced-host modes) and the power source (bus-powered or self-powered), there are different typical hardware implementations. 22.4.1 22.4.1.1 Device Mode Bus-Powered device Figure 22-2. Bus-Powered Device Application Block Diagram VDD 3.3 V Regulator USB USB Connector VBUS DD+ ID GND USB_VBOF VBUS DD+ USB_ID 39 1% 39 1% 22.4.1.2 Self-Powered device Figure 22-3. Self-Powered Device Application Block Diagram USB USB Connector VBUS DD+ ID GND USB_VBOF VBUS DD+ USB_ID 39 1% 39 1% 368 32059I-06/2010 AT32UC3B 22.4.2 Host Mode Figure 22-4. Host Application Block Diagram VDD 5 V DC/DC Generator USB USB Connector VBUS DD+ ID GND USB_VBOF VBUS DD+ USB_ID 39 1% 39 1% 369 32059I-06/2010 AT32UC3B 22.5 I/O Lines Description I/O Lines Description Pin Description USB VBus On/Off: Bus Power Control Port VBus: Bus Power Measurement Port Data -: Differential Data Line - Port Data +: Differential Data Line + Port USB Identification: Mini Connector Identification Port Type Output Input Input/Output Input/Output Input Low: Mini-A plug High Z: Mini-B plug Active Level VBUSPO Table 22-3. PIn Name USB_VBOF USB_VBUS DD+ USB_ID 370 32059I-06/2010 AT32UC3B 22.6 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 22.6.1 I/O Lines The USB_VBOF and USB_ID pins are multiplexed with I/O Controller lines and may also be multiplexed with lines of other peripherals. In order to use them with the USB, the user must first configure the I/O Controller to assign them to their USB peripheral functions. If USB_ID is used, the I/O Controller must be configured to enable the internal pull-up resistor of its pin. If USB_VBOF or USB_ID is not used by the application, the corresponding pin can be used for other purposes by the I/O Controller or by other peripherals. 22.6.2 Clocks The clock for the USBB bus interface (CLK_USBB) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. It is recommended to disable the USBB before disabling the clock, to avoid freezing the USBB in an undefined state. The 48 MHz USB clock is generated by a dedicated generic clock from the Power Manager. Before using the USB, the user must ensure that the USB generic clock (GCLK_USBB) is enabled at 48MHz in the Power Manager. 22.6.3 Interrupts The USBB interrupt request line is connected to the interrupt controller. Using the USBB interrupt requires the interrupt controller to be programmed first. 371 32059I-06/2010 AT32UC3B 22.7 22.7.1 22.7.1.1 Functional Description USB General Operation Introduction After a hardware reset, the USBB is disabled. When enabled, the USBB runs either in device mode or in host mode according to the ID detection. If the USB_ID pin is not connected to ground, the USB_ID Pin State bit in the General Status register (USBSTA.ID) is set (the internal pull-up resistor of the USB_ID pin must be enabled by the I/O Controller) and device mode is engaged. The USBSTA.ID bit is cleared when a low level has been detected on the USB_ID pin. Host mode is then engaged. 22.7.1.2 Power-On and reset Figure 22-5 on page 372 describes the USBB main states. Figure 22-5. General States Macro off: USBE = 0 Clock stopped: FRZCLK = 1 USBE = 0 Reset HW RESET USBE = 1 ID = 1 USBE = 0 USBE = 1 ID = 0 USBE = 0 Device Host After a hardware reset, the USBB is in the Reset state. In this state: * The macro is disabled. The USBB Enable bit in the General Control register (USBCON.USBE) is zero. * The macro clock is stopped in order to minimize power consumption. The Freeze USB Clock bit in USBCON (USBON.FRZCLK) is set. * The pad is in suspend mode. * The internal states and registers of the device and host modes are reset. * The DPRAM is not cleared and is accessible. * The USBSTA.ID bit and the VBus Level bit in the UBSTA (UBSTA.VBUS) reflect the states of the USB_ID and USB_VBUS input pins. * The OTG Pad Enable (OTGPADE) bit, the VBus Polarity (VBUSPO) bit, the FRZCLK bit, the USBE bit, the USB_ID Pin Enable (UIDE) bit, the USBB Mode (UIMOD) bit in USBCON, and the Low-Speed Mode Force bit in the Device General Control (UDCON.LS) register can be written by software, so that the user can program pads and speed before enabling the macro, but their value is only taken into account once the macro is enabled and unfrozen. 372 32059I-06/2010 AT32UC3B After writing a one to USBCON.USBE, the USBB enters the Device or the Host mode (according to the ID detection) in idle state. The USBB can be disabled at any time by writing a zero to USBCON.USBE. In fact, writing a zero to USBCON.USBE acts as a hardware reset, except that the OTGPADE, VBUSPO, FRZCLK, UIDE, UIMOD and, LS bits are not reset. 22.7.1.3 Interrupts One interrupt vector is assigned to the USB interface. Figure 22-6 on page 374 shows the structure of the USB interrupt system. 373 32059I-06/2010 AT32UC3B Figure 22-6. Interrupt System USBSTA.IDTI USBCON.IDTE USBSTA.VBUSTI USBCON.VBUSTE USBSTA.SRPI USBCON.SRPE USBSTA.VBERRI USBCON.VBERRE USBSTA.BCERRI USBCON.BCERRE USBSTA.ROLEEXI USBCON.ROLEEXE USBSTA.HNPERRI USBCON.HNPERRE USBSTA.STOI USBCON.STOE USB General Interrupt UESTAX.TXINI UECONX.TXINE UESTAX.RXOUTI UECONX.RXOUTE UESTAX.RXSTPI UECONX.RXSTPE UESTAX.UNDERFI UECONX.UNDERFE UESTAX.NAKOUTI UECONX.NAKOUTE UESTAX.NAKINI UECONX.NAKINE UESTAX.OVERFI UECONX.OVERFE UESTAX.STALLEDI UECONX.STALLEDE UESTAX.CRCERRI UECONX.CRCERRE UESTAX.SHORTPACKET UECONX.SHORTPACKETE UESTAX.NBUSYBK UECONX.NBUSYBKE UDDMAX_STATUS.EOT_STA UDDMAX_CONTROL.EOT_IRQ_EN UDDMAX_STATUS.EOCH_BUFF_STA UDDMAX_CONTROL.EOBUFF_IRQ_EN UDDMAX_STATUS.DESC_LD_STA UDDMAX_CONTROL.DESC_LD_IRQ_EN UPSTAX.RXINI UPCONX.RXINE UPSTAX.TXOUTI UPCONX.TXOUTE UPSTAX.TXSTPI UPCONX.TXSTPE UPSTAX.UNDERFI UPCONX.UNDERFIE UPSTAX.PERRI UPCONX.PERRE UPSTAX.NAKEDI UPCONX.NAKEDE UPSTAX.OVERFI UPCONX.OVERFIE UPSTAX.RXSTALLDI UPCONX.RXSTALLDE UPSTAX.CRCERRI UPCONX.CRCERRE UPSTAX.SHORTPACKETI UPCONX.SHORTPACKETIE UPSTAX.NBUSYBK UPCONX.NBUSYBKE UHDMAX_STATUS.EOT_STA UHDMAX_CONTROL.EOT_IRQ_EN UHDMAX_STATUS.EOCH_BUFF_STA UHDMAX_CONTROL.EOBUFF_IRQ_EN UHDMAX_STATUS.DESC_LD_STA UHDMAX_CONTROL.DESC_LD_IRQ_EN USB Host DMA Channel X Interrupt UHINT.DMAXINT UHINTE.DMAXINTE UHINT.HWUPI UHINTE.HWUPIE UHINT.PXINT UHINTE.PXINTE USB Host Pipe X Interrupt UHINT.RSMEDI UHINTE.RSMEDIE UHINT.RXRSMI UHINTE.RXRSMIE UHINT.HSOFI UHINTE.HSOFIE USB Host Interrupt UHINT.RSTI UHINTE.RSTIE UHINT.DDISCI UHINTE.DDISCIE UHINT.DCONNI UHINTE.DCONNIE USB Device DMA Channel X Interrupt UDINT.DMAXINT UDINTE.DMAXINTE UDINT.UPRSM UDINTE.UPRSME UDINT.EPXINT UDINTE.EPXINTE USB Device Endpoint X Interrupt UDINT.EORST UDINTE.EORSTE UDINT.WAKEUP UDINTE.WAKEUPE UDINT.EORSM UDINTE.EORSME USB Device Interrupt USB Interrupt UDINT.SOF UDINTE.SOFE UDINT.SUSP UDINTE.SUSPE Asynchronous interrupt source See Section 22.7.2.17 and Section 22.7.3.13 for further details about device and host interrupts. There are two kinds of general interrupts: processing, i.e. their generation is part of the normal processing, and exception, i.e. errors (not related to CPU exceptions). The processing general interrupts are: 374 32059I-06/2010 AT32UC3B * The ID Transition Interrupt (IDTI) * The VBus Transition Interrupt (VBUSTI) * The Role Exchange Interrupt (ROLEEXI) The exception general interrupts are: * The VBus Error Interrupt (VBERRI) * The B-Connection Error Interrupt (BCERRI) * The Suspend Time-Out Interrupt (STOI) 22.7.1.4 MCU Power modes *Run mode In this mode, all MCU clocks can run, including the USB clock. *Idle mode In this mode, the CPU is halted, i.e. the CPU clock is stopped. The Idle mode is entered whatever the state of the USBB. The MCU wakes up on any USB interrupt. *Frozen mode Same as the Idle mode, except that the HSB module is stopped, so the USB DMA, which is an HSB master, can not be used. Moreover, the USB DMA must be stopped before entering this sleep mode in order to avoid erratic behavior. The MCU wakes up on any USB interrupt. *Standby, Stop, DeepStop and Static modes Same as the Frozen mode, except that the USB generic clock and other clocks are stopped, so the USB macro is frozen. Only the asynchronous USB interrupt sources can wake up the MCU in these modes. The Power Manager (PM) may have to be configured to enable asynchronous wake up from USB. The USB module must be frozen by writing a one to the FRZCLK bit. *USB clock frozen In the run, idle and frozen MCU modes, the USBB can be frozen when the USB line is in the suspend mode, by writing a one to the FRZCLK bit, what reduces power consumption. In deeper MCU power modes (from StandBy mode), the USBC must be frozen. In this case, it is still possible to access the following elements, but only in Run mode: * The OTGPADE, VBUSPO, FRZCLK, USBE, UIDE, UIMOD and LS bits in the USBCON register * The DPRAM (through the USB Pipe/Endpoint n FIFO Data (USBFIFOnDATA) registers, but not through USB bus transfers which are frozen) Moreover, when FRZCLK is written to one, only the asynchronous interrupt sources may trigger the USB interrupt: * The ID Transition Interrupt (IDTI) * The VBus Transition Interrupt (VBUSTI) * The Wake-up Interrupt (WAKEUP) 375 32059I-06/2010 AT32UC3B * The Host Wake-up Interrupt (HWUPI) *USB Suspend mode In peripheral mode, the Suspend Interrupt bit in the Device Global Interrupt register (UDINT.SUSP)indicates that the USB line is in the suspend mode. In this case, the USB Data UTMI transceiver is automatically set in suspend mode to reduce the consumption. 22.7.1.5 Speed control *Device mode When the USBB interface is in device mode, the speed selection (full-/low-speed) depends on which of D+ and D- is pulled up. The LS bit allows to connect an internal pull-up resistor either on D+ (full-speed mode) or on D- (low-speed mode). The LS bit shall be written before attaching the device, what can be done by clearing the DETACH bit in UDCON. Figure 22-7. Speed Selection in Device Mode VBUS RPU UDCON.DETACH UDCON.LS D+ D- *Host mode When the USB interface is in host mode, internal pull-down resistors are connected on both D+ and D- and the interface detects the speed of the connected device, which is reflected by the Speed Status (SPEED) field in USBSTA. 22.7.1.6 DPRAM management Pipes and endpoints can only be allocated in ascending order (from the pipe/endpoint 0 to the last pipe/endpoint to be allocated). The user shall therefore configure them in the same order. The allocation of a pipe/endpoint n starts when the Endpoint Memory Allocate bit in the Endpoint n Configuration register (UECFGn.ALLOC) is written to one. Then, the hardware allocates a memory area in the DPRAM and inserts it between the n-1 and n+1 pipes/endpoints. The n+1 pipe/endpoint memory window slides up and its data is lost. Note that the following pipe/endpoint memory windows (from n+2) do not slide. Disabling a pipe, by writing a zero to the Pipe n Enable bit in the Pipe Enable/Reset register (UPRST.PENn), or disabling an endpoint, by writing a zero to the Endpoint n Enable bit in the Endpoint Enable/Reset register (UERST.EPENn), resets neither the UECFGn.ALLOC bit nor its 376 32059I-06/2010 AT32UC3B configuration (the Pipe Banks (PBK) field, the Pipe Size (PSIZE) field, the Pipe Token (PTOKEN) field, the Pipe Type (PTYPE) field, the Pipe Endpoint Number (PEPNUM) field, and the Pipe Interrupt Request Frequency (INTFRQ) field in the Pipe n Configuration (UPCFGn) register/the Endpoint Banks (EPBK) field, the Endpoint Size (EPSIZE) field, the Endpoint Direction (EPDIR) field, and the Endpoint Type (EPTYPE) field in UECFGn). To free its memory, the user shall write a zero to the UECFGn.ALLOC bit. The n+1 pipe/endpoint memory window then slides down and its data is lost. Note that the following pipe/endpoint memory windows (from n+2) does not slide. Figure 22-8 on page 377 illustrates the allocation and reorganization of the DPRAM in a typical example. Figure 22-8. Allocation and Reorganization of the DPRAM Free Memory Free Memory Free Memory Free Memory PEP5 PEP4 PEP3 PEP2 PEP5 PEP4 PEP3 (ALLOC stays at 1) PEP2 PEP5 PEP4 Lost Memory PEP4 PEP2 PEP5 PEP4 PEP3 (larger size) Conflict PEP2 PEP1 PEP0 PEP1 PEP0 PEP1 PEP0 PEP1 PEP0 U(P/E)RST.(E)PENn = 1 U(P/E)CFGn.ALLOC = 1 U(P/E)RST.(E)PEN3 = 0 U(P/E)CFG3.ALLOC = 0 U(P/E)RST.(E)PEN3 = 1 U(P/E)CFG3.ALLOC = 1 Pipes/Endpoints 0..5 Activated Pipe/Endpoint 3 Disabled Pipe/Endpoint 3 Memory Freed Pipe/Endpoint 3 Activated 1. The pipes/endpoints 0 to 5 are enabled, configured and allocated in ascending order. Each pipe/endpoint then owns a memory area in the DPRAM. 2. The pipe/endpoint 3 is disabled, but its memory is kept allocated by the controller. 3. In order to free its memory, its ALLOC bit is written to zero. The pipe/endpoint 4 memory window slides down, but the pipe/endpoint 5 does not move. 4. If the user chooses to reconfigure the pipe/endpoint 3 with a larger size, the controller allocates a memory area after the pipe/endpoint 2 memory area and automatically slides up the pipe/endpoint 4 memory window. The pipe/endpoint 5 does not move and a memory conflict appears as the memory windows of the pipes/endpoints 4 and 5 overlap. The data of these pipes/endpoints is potentially lost. Note that: * There is no way the data of the pipe/endpoint 0 can be lost (except if it is de-allocated) as memory allocation and de-allocation may affect only higher pipes/endpoints. * Deactivating then reactivating a same pipe/endpoint with the same configuration only modifies temporarily the controller DPRAM pointer and size for this pipe/endpoint, but nothing changes in the DPRAM, so higher endpoints seem to not have been moved and their data is preserved 377 32059I-06/2010 AT32UC3B as far as nothing has been written or received into them while changing the allocation state of the first pipe/endpoint. * When the user write a one to the ALLOC bit, the Configuration OK Status bit in the Endpoint n Status register (UESTAn.CFGOK) is set only if the configured size and number of banks are correct compared to their maximal allowed values for the endpoint and to the maximal FIFO size (i.e. the DPRAM size), so the value of CFGOK does not consider memory allocation conflicts. 22.7.1.7 Pad Suspend Figure 22-9 on page 378 shows the pad behavior. Figure 22-9. Pad Behavior Idle USBE = 1 & DETACH = 0 & Suspend USBE = 0 | DETACH = 1 | Suspend Active * In the Idle state, the pad is put in low power consumption mode, i.e., the differential receiver of the USB pad is off, and internal pull-down with strong value(15K) are set in both DP/DM to avoid floating lines. * In the Active state, the pad is working. Figure 22-10 on page 378 illustrates the pad events leading to a PAD state change. Figure 22-10. Pad Events SUSP Suspend detected Cleared on wake-up WAKEUP Wake-up detected Cleared by software to acknowledge the interrupt PAD State Active Idle Active 378 32059I-06/2010 AT32UC3B The SUSP bit is set and the Wake-Up Interrupt (WAKEUP) bit in UDINT is cleared when a USB "Suspend" state has been detected on the USB bus. This event automatically puts the USB pad in the Idle state. The detection of a non-idle event sets WAKEUP, clears SUSP and wakes up the USB pad. Moreover, the pad goes to the Idle state if the macro is disabled or if the DETACH bit is written to one. It returns to the Active state when USBE is written to one and DETACH is written to zero. 22.7.1.8 Plug-In detection The USB connection is detected from the USB_VBUS pad. Figure 22-11 on page 379 shows the architecture of the plug-in detector. Figure 22-11. Plug-In Detection Input Block Diagram VDD VBus_pulsing RPU Session_valid Logic USB_VBUS RPD Va_Vbus_valid VBUS USBSTA VBUSTI USBSTA VBus_discharge GND Pad Logic The control logic of the USB_VBUS pad outputs two signals: * The Session_valid signal is high when the voltage on the USB_VBUS pad is higher than or equal to 1.4V. * The Va_Vbus_valid signal is high when the voltage on the USB_VBUS pad is higher than or equal to 4.4V. In device mode, the USBSTA.VBUS bit follows the Session_valid comparator output: * It is set when the voltage on the USB_VBUS pad is higher than or equal to 1.4V. * It is cleared when the voltage on the VBUS pad is lower than 1.4V. In host mode, the USBSTA.VBUS bit follows an hysteresis based on Session_valid and Va_Vbus_valid: * It is set when the voltage on the USB_VBUS pad is higher than or equal to 4.4V. * It is cleared when the voltage on the USB_VBUS pad is lower than 1.4V. The VBus Transition interrupt (VBUSTI) bit in USBSTA is set on each transition of the USBSTA.VBUS bit. The USBSTA.VBUS bit is effective whether the USBB is enabled or not. 22.7.1.9 ID detection Figure 22-12 on page 380 shows how the ID transitions are detected. 379 32059I-06/2010 AT32UC3B Figure 22-12. ID Detection Input Block Diagram VDD RPU 1 USB_ID 0 ID USBSTA IDTI USBSTA UIMOD USBCON UIDE USBCON I/O Controller The USB mode (device or host) can be either detected from the USB_ID pin or software selected by writing to the UIMOD bit, according to the UIDE bit. This allows the USB_ID pin to be used as a general purpose I/O pin even when the USB interface is enabled. By default, the USB_ID pin is selected (UIDE is written to one) and the USBB is in device mode (UBSTA.ID is one), what corresponds to the case where no Mini-A plug is connected, i.e. no plug or a Mini-B plug is connected and the USB_ID pin is kept high by the internal pull-up resistor from the I/O Controller (which must be enabled if USB_ID is used). The ID Transition Interrupt (IDTI) bit in USBSTA is set on each transition of the ID bit, i.e. when a Mini-A plug (host mode) is connected or disconnected. This does not occur when a Mini-B plug (device mode) is connected or disconnected. The USBSTA.ID bit is effective whether the USBB is enabled or not. 380 32059I-06/2010 AT32UC3B 22.7.2 22.7.2.1 USB Device Operation Introduction In device mode, the USBB supports full- and low-speed data transfers. In addition to the default control endpoint, six endpoints are provided, which can be configured with the types isochronous, bulk or interrupt, as described in Table 22-1 on page 365. The device mode starts in the Idle state, so the pad consumption is reduced to the minimum. 22.7.2.2 Power-On and reset Figure 22-13 on page 381 describes the USBB device mode main states. Figure 22-13. Device Mode States USBE = 0 | ID = 0 USBE = 0 | ID = 0 Reset HW RESET USBE = 1 & ID = 1 Idle After a hardware reset, the USBB device mode is in the Reset state. In this state: * The macro clock is stopped in order to minimize power consumption (FRZCLK is written to one). * The internal registers of the device mode are reset. * The endpoint banks are de-allocated. * Neither D+ nor D- is pulled up (DETACH is written to one). D+ or D- will be pulled up according to the selected speed as soon as the DETACH bit is written to zero and VBus is present. See "Device mode" for further details. When the USBB is enabled (USBE is written to one) in device mode (ID is one), its device mode state goes to the Idle state with minimal power consumption. This does not require the USB clock to be activated. The USBB device mode can be disabled and reset at any time by disabling the USBB (by writing a zero to USBE) or when host mode is engaged (ID is zero). 22.7.2.3 USB reset The USB bus reset is managed by hardware. It is initiated by a connected host. When a USB reset is detected on the USB line, the following operations are performed by the controller: 381 32059I-06/2010 AT32UC3B * All the endpoints are disabled, except the default control endpoint. * The default control endpoint is reset (see Section 22.7.2.4 for more details). * The data toggle sequence of the default control endpoint is cleared. * At the end of the reset process, the End of Reset (EORST) bit in UDINT interrupt is set. 22.7.2.4 Endpoint reset An endpoint can be reset at any time by writing a one to the Endpoint n Reset (EPRSTn) bit in the UERST register. This is recommended before using an endpoint upon hardware reset or when a USB bus reset has been received. This resets: * The internal state machine of this endpoint. * The receive and transmit bank FIFO counters. * All the registers of this endpoint (UECFGn, UESTAn, the Endpoint n Control (UECONn) register), except its configuration (ALLOC, EPBK, EPSIZE, EPDIR, EPTYPE) and the Data Toggle Sequence (DTSEQ) field of the UESTAn register. Note that the interrupt sources located in the UESTAn register are not cleared when a USB bus reset has been received. The endpoint configuration remains active and the endpoint is still enabled. The endpoint reset may be associated with a clear of the data toggle sequence as an answer to the CLEAR_FEATURE USB request. This can be achieved by writing a one to the Reset Data Toggle Set bit in the Endpoint n Control Set register (UECONnSET.RSTDTS).(This will set the Reset Data Toggle (RSTD) bit in UECONn). In the end, the user has to write a zero to the EPRSTn bit to complete the reset operation and to start using the FIFO. 22.7.2.5 Endpoint activation The endpoint is maintained inactive and reset (see Section 22.7.2.4 for more details) as long as it is disabled (EPENn is written to zero). DTSEQ is also reset. The algorithm represented on Figure 22-14 on page 383 must be followed in order to activate an endpoint. 382 32059I-06/2010 AT32UC3B Figure 22-14. Endpoint Activation Algorithm Endpoint Activation Enable the endpoint. Configure the endpoint: - type - direction - size - number of banks Allocate the configured DPRAM banks. Test if the endpoint configuration is correct. No EPENn = 1 UECFGn EPTYPE EPDIR EPSIZE EPBK ALLOC CFGOK == 1? Yes Endpoint Activated ERROR As long as the endpoint is not correctly configured (CFGOK is zero), the controller does not acknowledge the packets sent by the host to this endpoint. The CFGOK bit is set only if the configured size and number of banks are correct compared to their maximal allowed values for the endpoint (see Table 22-1 on page 365) and to the maximal FIFO size (i.e. the DPRAM size). See Section 22.7.1.6 for more details about DPRAM management. 22.7.2.6 Address setup The USB device address is set up according to the USB protocol. * After all kinds of resets, the USB device address is 0. * The host starts a SETUP transaction with a SET_ADDRESS(addr) request. * The user write this address to the USB Address (UADD) field in UDCON, and write a zero to the Address Enable (ADDEN) bit in UDCON, so the actual address is still 0. * The user sends a zero-length IN packet from the control endpoint. * The user enables the recorded USB device address by writing a one to ADDEN. Once the USB device address is configured, the controller filters the packets to only accept those targeting the address stored in UADD. UADD and ADDEN shall not be written all at once. UADD and ADDEN are cleared: * On a hardware reset. * When the USBB is disabled (USBE written to zero). * When a USB reset is detected. When UADD or ADDEN is cleared, the default device address 0 is used. 383 32059I-06/2010 AT32UC3B 22.7.2.7 Suspend and wake-up When an idle USB bus state has been detected for 3 ms, the controller set the Suspend (SUSP) interrupt bit in UDINT. The user may then write a one to the FRZCLK bit to reduce power consumption. The MCU can also enter the Idle or Frozen sleep mode to lower again power consumption. To recover from the Suspend mode, the user shall wait for the Wake-Up (WAKEUP) interrupt bit, which is set when a non-idle event is detected, then write a zero to FRZCLK. As the WAKEUP interrupt bit in UDINT is set when a non-idle event is detected, it can occur whether the controller is in the Suspend mode or not. The SUSP and WAKEUP interrupts are thus independent of each other except that one bit is cleared when the other is set. 22.7.2.8 Detach The reset value of the DETACH bit is one. It is possible to initiate a device re-enumeration simply by writing a one then a zero to DETACH. DETACH acts on the pull-up connections of the D+ and D- pads. See "Device mode" for further details. 22.7.2.9 Remote wake-up The Remote Wake-Up request (also known as Upstream Resume) is the only one the device may send on its own initiative, but the device should have beforehand been allowed to by a DEVICE_REMOTE_WAKEUP request from the host. * First, the USBB must have detected a "Suspend" state on the bus, i.e. the Remote Wake-Up request can only be sent after a SUSP interrupt has been set. * The user may then write a one to the Remote Wake-Up (RMWKUP) bit in UDCON to send an upstream resume to the host for a remote wake-up. This will automatically be done by the controller after 5ms of inactivity on the USB bus. * When the controller sends the upstream resume, the Upstream Resume (UPRSM) interrupt is set and SUSP is cleared. * RMWKUP is cleared at the end of the upstream resume. * If the controller detects a valid "End of Resume" signal from the host, the End of Resume (EORSM) interrupt is set. 22.7.2.10 STALL request For each endpoint, the STALL management is performed using: * The STALL Request (STALLRQ) bit in UECONn to initiate a STALL request. * The STALLed Interrupt (STALLEDI) bit in UESTAn is set when a STALL handshake has been sent. To answer the next request with a STALL handshake, STALLRQ has to be set by writing a one to the STALL Request Set (STALLRQS) bit. All following requests will be discarded (RXOUTI, etc. will not be set) and handshaked with a STALL until the STALLRQ bit is cleared, what is done when a new SETUP packet is received (for control endpoints) or when the STALL Request Clear (STALLRQC) bit is written to one. Each time a STALL handshake is sent, the STALLEDI bit is set by the USBB and the EPnINT interrupt is set. 384 32059I-06/2010 AT32UC3B *Special considerations for control endpoints If a SETUP packet is received into a control endpoint for which a STALL is requested, the Received SETUP Interrupt (RXSTPI) bit in UESTAn is set and STALLRQ and STALLEDI are cleared. The SETUP has to be ACKed. This management simplifies the enumeration process management. If a command is not supported or contains an error, the user requests a STALL and can return to the main task, waiting for the next SETUP request. *STALL handshake and retry mechanism The retry mechanism has priority over the STALL handshake. A STALL handshake is sent if the STALLRQ bit is set and if there is no retry required. 22.7.2.11 Management of control endpoints *Overview A SETUP request is always ACKed. When a new SETUP packet is received, the RXSTPI is set, but not the Received OUT Data Interrupt (RXOUTI) bit. The FIFO Control (FIFOCON) bit in UECONn and the Read/Write Allowed (RWALL) bit in UESTAn are irrelevant for control endpoints. The user shall therefore never use them on these endpoints. When read, their value are always zero. Control endpoints are managed using: * The RXSTPI bit which is set when a new SETUP packet is received and which shall be cleared by firmware to acknowledge the packet and to free the bank. * The RXOUTI bit which is set when a new OUT packet is received and which shall be cleared by firmware to acknowledge the packet and to free the bank. * The Transmitted IN Data Interrupt (TXINI) bit which is set when the current bank is ready to accept a new IN packet and which shall be cleared by firmware to send the packet. *Control write Figure 22-15 on page 386 shows a control write transaction. During the status stage, the controller will not necessarily send a NAK on the first IN token: * If the user knows the exact number of descriptor bytes that must be read, it can then anticipate the status stage and send a zero-length packet after the next IN token. * Or it can read the bytes and wait for the NAKed IN Interrupt (NAKINI) which tells that all the bytes have been sent by the host and that the transaction is now in the status stage. 385 32059I-06/2010 AT32UC3B Figure 22-15. Control Write SETUP USB Bus RXSTPI RXOUTI TXINI SETUP HW SW DATA OUT OUT IN NAK STATUS IN HW SW HW SW SW *Control read Figure 22-16 on page 386 shows a control read transaction. The USBB has to manage the simultaneous write requests from the CPU and the USB host. Figure 22-16. Control Read SETUP USB Bus RXSTPI RXOUTI TXINI Wr Enable HOST Wr Enable CPU SW HW DATA IN SW STATUS IN OUT NAK OUT SETUP HW HW SW SW A NAK handshake is always generated on the first status stage command. When the controller detects the status stage, all the data written by the CPU are lost and clearing TXINI has no effect. The user checks if the transmission or the reception is complete. The OUT retry is always ACKed. This reception sets RXOUTI and TXINI. Handle this with the following software algorithm: set TXINI wait for RXOUTI OR TXINI if RXOUTI, then clear bit and return if TXINI, then continue Once the OUT status stage has been received, the USBB waits for a SETUP request. The SETUP request has priority over any other request and has to be ACKed. This means that any other bit should be cleared and the FIFO reset when a SETUP is received. The user has to take care of the fact that the byte counter is reset when a zero-length OUT packet is received. 386 32059I-06/2010 AT32UC3B 22.7.2.12 Management of IN endpoints *Overview IN packets are sent by the USB device controller upon IN requests from the host. All the data can be written which acknowledges or not the bank when it is full. The endpoint must be configured first. The TXINI bit is set at the same time as FIFOCON when the current bank is free. This triggers an EPnINT interrupt if the Transmitted IN Data Interrupt Enable (TXINE) bit in UECONn is one. TXINI shall be cleared by software (by writing a one to the Transmitted IN Data Interrupt Enable Clear bit in the Endpoint n Control Clear register (UECONnCLR.TXINIC)) to acknowledge the interrupt, what has no effect on the endpoint FIFO. The user then writes into the FIFO and write a one to the FIFO Control Clear (FIFOCONC) bit in UECONnCLR to clear the FIFOCON bit. This allows the USBB to send the data. If the IN endpoint is composed of multiple banks, this also switches to the next bank. The TXINI and FIFOCON bits are updated in accordance with the status of the next bank. TXINI shall always be cleared before clearing FIFOCON. The RWALL bit is set when the current bank is not full, i.e. the software can write further data into the FIFO. Figure 22-17. Example of an IN Endpoint with 1 Data Bank NAK IN DATA (bank 0) ACK IN HW TXINI SW SW FIFOCON write data to CPU BANK 0 SW write data to CPU BANK 0 SW Figure 22-18. Example of an IN Endpoint with 2 Data Banks IN DATA (bank 0) ACK IN DATA (bank 1) ACK HW TXINI SW SW SW FIFOCON write data to CPU BANK 0 SW write data to CPU BANK 1 SW write data to CPU BANK0 387 32059I-06/2010 AT32UC3B *Detailed description The data is written, following the next flow: * When the bank is empty, TXINI and FIFOCON are set, what triggers an EPnINT interrupt if TXINE is one. * The user acknowledges the interrupt by clearing TXINI. * The user writes the data into the current bank by using the USB Pipe/Endpoint nFIFO Data (USBFIFOnDATA) register, until all the data frame is written or the bank is full (in which case RWALL is cleared and the Byte Count (BYCT) field in UESTAn reaches the endpoint size). * The user allows the controller to send the bank and switches to the next bank (if any) by clearing FIFOCON. If the endpoint uses several banks, the current one can be written while the previous one is being read by the host. Then, when the user clears FIFOCON, the following bank may already be free and TXINI is set immediately. An "Abort" stage can be produced when a zero-length OUT packet is received during an IN stage of a control or isochronous IN transaction. The Kill IN Bank (KILLBK) bit in UECONn is used to kill the last written bank. The best way to manage this abort is to apply the algorithm represented on Figure 22-19 on page 388. Figure 22-19. Abort Algorithm Endpoint Abort TXINEC = 1 Disable the TXINI interrupt. NBUSYBK == 0? Yes EPRSTn = 1 No Abort is based on the fact that no bank is busy, i.e., that nothing has to be sent KILLBKS = 1 Kill the last written bank. Yes KILLBK == 1? No Wait for the end of the procedure Abort Done 22.7.2.13 Management of OUT endpoints *Overview OUT packets are sent by the host. All the data can be read which acknowledges or not the bank when it is empty. The endpoint must be configured first. 388 32059I-06/2010 AT32UC3B The RXOUTI bit is set at the same time as FIFOCON when the current bank is full. This triggers an EPnINT interrupt if the Received OUT Data Interrupt Enable (RXOUTE) bit in UECONn is one. RXOUTI shall be cleared by software (by writing a one to the Received OUT Data Interrupt Clear (RXOUTIC) bit) to acknowledge the interrupt, what has no effect on the endpoint FIFO. The user then reads from the FIFO and clears the FIFOCON bit to free the bank. If the OUT endpoint is composed of multiple banks, this also switches to the next bank. The RXOUTI and FIFOCON bits are updated in accordance with the status of the next bank. RXOUTI shall always be cleared before clearing FIFOCON. The RWALL bit is set when the current bank is not empty, i.e. the software can read further data from the FIFO. Figure 22-20. Example of an OUT Endpoint with one Data Bank OUT DATA (bank 0) ACK NAK OUT DATA (bank 0) ACK HW RXOUTI SW HW SW FIFOCON read data from CPU BANK 0 SW read data from CPU BANK 0 Figure 22-21. Example of an OUT Endpoint with two Data Banks OUT DATA (bank 0) ACK OUT DATA (bank 1) ACK HW RXOUTI SW HW SW FIFOCON read data from CPU BANK 0 SW read data from CPU BANK 1 *Detailed description The data is read, following the next flow: * When the bank is full, RXOUTI and FIFOCON are set, what triggers an EPnINT interrupt if RXOUTE is one. * The user acknowledges the interrupt by writing a one to RXOUTIC in order to clear RXOUTI. 389 32059I-06/2010 AT32UC3B * The user can read the byte count of the current bank from BYCT to know how many bytes to read, rather than polling RWALL. * The user reads the data from the current bank by using the USBFIFOnDATA register, until all the expected data frame is read or the bank is empty (in which case RWALL is cleared and BYCT reaches zero). * The user frees the bank and switches to the next bank (if any) by clearing FIFOCON. If the endpoint uses several banks, the current one can be read while the following one is being written by the host. Then, when the user clears FIFOCON, the following bank may already be ready and RXOUTI is set immediately. 22.7.2.14 Underflow This error exists only for isochronous IN/OUT endpoints. It set the Underflow Interrupt (UNDERFI) bit in UESTAn, what triggers an EPnINT interrupt if the Underflow Interrupt Enable (UNDERFE) bit is one. An underflow can occur during IN stage if the host attempts to read from an empty bank. A zerolength packet is then automatically sent by the USBB. An underflow can not occur during OUT stage on a CPU action, since the user may read only if the bank is not empty (RXOUTI is one or RWALL is one). An underflow can also occur during OUT stage if the host sends a packet while the bank is already full. Typically, the CPU is not fast enough. The packet is lost. An underflow can not occur during IN stage on a CPU action, since the user may write only if the bank is not full (TXINI is one or RWALL is one). 22.7.2.15 Overflow This error exists for all endpoint types. It set the Overflow interrupt (OVERFI) bit in UESTAn, what triggers an EPnINT interrupt if the Overflow Interrupt Enable (OVERFE) bit is one. An overflow can occur during OUT stage if the host attempts to write into a bank that is too small for the packet. The packet is acknowledged and the RXOUTI bit is set as if no overflow had occurred. The bank is filled with all the first bytes of the packet that fit in. An overflow can not occur during IN stage on a CPU action, since the user may write only if the bank is not full (TXINI is one or RWALL is one). 22.7.2.16 CRC error This error exists only for isochronous OUT endpoints. It set the CRC Error Interrupt (CRCERRI) bit in UESTAn, what triggers an EPnINT interrupt if the CRC Error Interrupt Enable (CRCERRE) bit is one. A CRC error can occur during OUT stage if the USBB detects a corrupted received packet. The OUT packet is stored in the bank as if no CRC error had occurred (RXOUTI is set). 22.7.2.17 Interrupts See the structure of the USB device interrupt system on Figure 22-6 on page 374. There are two kinds of device interrupts: processing, i.e. their generation is part of the normal processing, and exception, i.e. errors (not related to CPU exceptions). 390 32059I-06/2010 AT32UC3B *Global interrupts The processing device global interrupts are: * The Suspend (SUSP) interrupt * The Start of Frame (SOF) interrupt with no frame number CRC error (the Frame Number CRC Error (FNCERR) bit in the Device Frame Number (UDFNUM) register is zero) * The End of Reset (EORST) interrupt * The Wake-Up (WAKEUP) interrupt * The End of Resume (EORSM) interrupt * The Upstream Resume (UPRSM) interrupt * The Endpoint n (EPnINT) interrupt * The DMA Channel n (DMAnINT) interrupt The exception device global interrupts are: * The Start of Frame (SOF) interrupt with a frame number CRC error (FNCERR is one) *Endpoint interrupts The processing device endpoint interrupts are: * The Transmitted IN Data Interrupt (TXINI) * The Received OUT Data Interrupt (RXOUTI) * The Received SETUP Interrupt (RXSTPI) * The Short Packet (SHORTPACKET) interrupt * The Number of Busy Banks (NBUSYBK) interrupt The exception device endpoint interrupts are: * The Underflow Interrupt (UNDERFI) * The NAKed OUT Interrupt (NAKOUTI) * The NAKed IN Interrupt (NAKINI) * The Overflow Interrupt (OVERFI) * The STALLed Interrupt (STALLEDI) * The CRC Error Interrupt (CRCERRI) *DMA interrupts The processing device DMA interrupts are: * The End of USB Transfer Status (EOTSTA) interrupt * The End of Channel Buffer Status (EOCHBUFFSTA) interrupt * The Descriptor Loaded Status (DESCLDSTA) interrupt There is no exception device DMA interrupt. 391 32059I-06/2010 AT32UC3B 22.7.3 22.7.3.1 USB Host Operation Description of pipes For the USBB in host mode, the term "pipe" is used instead of "endpoint" (used in device mode). A host pipe corresponds to a device endpoint, as described by the Figure 22-22 on page 392 from the USB specification. Figure 22-22. USB Communication Flow In host mode, the USBB associates a pipe to a device endpoint, considering the device configuration descriptors. 22.7.3.2 Power-On and reset Figure 22-23 on page 392 describes the USBB host mode main states. Figure 22-23. Host Mode States Macro off Clock stopped Idle Device Connection Device Disconnection Device Disconnection Ready SOFE = 0 SOFE = 1 Suspend After a hardware reset, the USBB host mode is in the Reset state. When the USBB is enabled (USBE is one) in host mode (ID is zero), its host mode state goes to the Idle state. In this state, the controller waits for device connection with minimal power con- 392 32059I-06/2010 AT32UC3B sumption. The USB pad should be in the Idle state. Once a device is connected, the macro enters the Ready state, what does not require the USB clock to be activated. The controller enters the Suspend state when the USB bus is in a "Suspend" state, i.e., when the host mode does not generate the "Start of Frame (SOF)". In this state, the USB consumption is minimal. The host mode exits the Suspend state when starting to generate the SOF over the USB line. 22.7.3.3 Device detection A device is detected by the USBB host mode when D+ or D- is no longer tied low, i.e., when the device D+ or D- pull-up resistor is connected. To enable this detection, the host controller has to provide the VBus power supply to the device by setting the VBUSRQ bit (by writing a one to the VBUSRQS bit). The device disconnection is detected by the host controller when both D+ and D- are pulled down. 22.7.3.4 USB reset The USBB sends a USB bus reset when the user write a one to the Send USB Reset bit in the Host General Control register (UHCON.RESET). The USB Reset Sent Interrupt bit in the Host Global Interrupt register (UHINT.RSTI) is set when the USB reset has been sent. In this case, all the pipes are disabled and de-allocated. If the bus was previously in a "Suspend" state (the Start of Frame Generation Enable (SOFE) bit in UHCON is zero), the USBB automatically switches it to the "Resume" state, the Host WakeUp Interrupt (HWUPI) bit in UHINT is set and the SOFE bit is set in order to generate SOFs immediately after the USB reset. 22.7.3.5 Pipe reset A pipe can be reset at any time by writing a one to the Pipe n Reset (PRSTn) bit in the UPRST register. This is recommended before using a pipe upon hardware reset or when a USB bus reset has been sent. This resets: * The internal state machine of this pipe * The receive and transmit bank FIFO counters * All the registers of this pipe (UPCFGn, UPSTAn, UPCONn), except its configuration (ALLOC, PBK, PSIZE, PTOKEN, PTYPE, PEPNUM, INTFRQ in UPCFGn) and its Data Toggle Sequence field in the Pipe n Status register (UPSTAn.DTSEQ). The pipe configuration remains active and the pipe is still enabled. The pipe reset may be associated with a clear of the data toggle sequence. This can be achieved by setting the Reset Data Toggle bit in the Pipe n Control register (UPCONn.RSTDT) (by writing a one to the Reset Data Toggle Set bit in the Pipe n Control Set register (UPCONnSET.RSTDTS)). In the end, the user has to write a zero to the PRSTn bit to complete the reset operation and to start using the FIFO. 22.7.3.6 Pipe activation The pipe is maintained inactive and reset (see Section 22.7.3.5 for more details) as long as it is disabled (PENn is zero). The Data Toggle Sequence field (DTSEQ) is also reset. 393 32059I-06/2010 AT32UC3B The algorithm represented on Figure 22-24 on page 394 must be followed in order to activate a pipe. Figure 22-24. Pipe Activation Algorithm Pipe Activation PENn = 1 UPCFGn INTFRQ PEPNUM PTYPE PTOKEN PSIZE PBK ALLOC Enable the pipe. Configure the pipe: - interrupt request frequency - endpoint number - type - size - number of banks Allocate the configured DPRAM banks. Test if the pipe configuration is correct. CFGOK == 1? Yes Pipe Activated No ERROR As long as the pipe is not correctly configured (UPSTAn.CFGOK is zero), the controller can not send packets to the device through this pipe. The UPSTAn.CFGOK bit is set only if the configured size and number of banks are correct compared to their maximal allowed values for the pipe (see Table 22-1 on page 365) and to the maximal FIFO size (i.e. the DPRAM size). See Section 22.7.1.6 for more details about DPRAM management. Once the pipe is correctly configured (UPSTAn.CFGOK is zero), only the PTOKEN and INTFRQ fields can be written by software. INTFRQ is meaningless for non-interrupt pipes. When starting an enumeration, the user gets the device descriptor by sending a GET_DESCRIPTOR USB request. This descriptor contains the maximal packet size of the device default control endpoint (bMaxPacketSize0) and the user re-configures the size of the default control pipe with this size parameter. 22.7.3.7 Address setup Once the device has answered the first host requests with the default device address 0, the host assigns a new address to the device. The host controller has to send an USB reset to the device and to send a SET_ADDRESS(addr) SETUP request with the new address to be used by the device. Once this SETUP transaction is over, the user writes the new address into the USB Host Address for Pipe n field in the USB Host Device Address register (UHADDR.UHADDRPn). All following requests, on all pipes, will be performed using this new address. When the host controller sends an USB reset, the UHADDRPn field is reset by hardware and the following host requests will be performed using the default device address 0. 394 32059I-06/2010 AT32UC3B 22.7.3.8 Remote wake-up The controller host mode enters the Suspend state when the UHCON.SOFE bit is written to zero. No more "Start of Frame" is sent on the USB bus and the USB device enters the Suspend state 3ms later. The device awakes the host by sending an Upstream Resume (Remote Wake-Up feature). When the host controller detects a non-idle state on the USB bus, it set the Host Wake-Up interrupt (HWUPI) bit in UHINT. If the non-idle bus state corresponds to an Upstream Resume (K state), the Upstream Resume Received Interrupt (RXRSMI) bit in UHINT is set. The user has to generate a Downstream Resume within 1ms and for at least 20ms by writing a one to the Send USB Resume (RESUME) bit in UHCON. It is mandatory to write a one to UHCON.SOFE before writing a one to UHCON.RESUME to enter the Ready state, else UHCON.RESUME will have no effect. 22.7.3.9 Management of control pipes A control transaction is composed of three stages: * SETUP * Data (IN or OUT) * Status (OUT or IN) The user has to change the pipe token according to each stage. For the control pipe, and only for it, each token is assigned a specific initial data toggle sequence: * SETUP: Data0 * IN: Data1 * OUT: Data1 22.7.3.10 Management of IN pipes IN packets are sent by the USB device controller upon IN requests from the host. All the data can be read which acknowledges or not the bank when it is empty. The pipe must be configured first. When the host requires data from the device, the user has to select beforehand the IN request mode with the IN Request Mode bit in the Pipe n IN Request register (UPINRQn.INMODE): * When INMODE is written to zero, the USBB will perform (INRQ + 1) IN requests before freezing the pipe. * When INMODE is written to one, the USBB will perform IN requests endlessly when the pipe is not frozen by the user. The generation of IN requests starts when the pipe is unfrozen (the Pipe Freeze (PFREEZE) field in UPCONn is zero). The Received IN Data Interrupt (RXINI) bit in UPSTAn is set at the same time as the FIFO Control (FIFOCON) bit in UPCONn when the current bank is full. This triggers a PnINT interrupt if the Received IN Data Interrupt Enable (RXINE) bit in UPCONn is one. RXINI shall be cleared by software (by writing a one to the Received IN Data Interrupt Clear bit in the Pipe n Control Clear register(UPCONnCLR.RXINIC)) to acknowledge the interrupt, what has no effect on the pipe FIFO. 395 32059I-06/2010 AT32UC3B The user then reads from the FIFO and clears the FIFOCON bit (by writing a one to the FIFO Control Clear (FIFOCONC) bit in UPCONnCLR) to free the bank. If the IN pipe is composed of multiple banks, this also switches to the next bank. The RXINI and FIFOCON bits are updated in accordance with the status of the next bank. RXINI shall always be cleared before clearing FIFOCON. The Read/Write Allowed (RWALL) bit in UPSTAn is set when the current bank is not empty, i.e., the software can read further data from the FIFO. Figure 22-25. Example of an IN Pipe with 1 Data Bank IN DATA (bank 0) ACK IN DATA (bank 0) ACK HW RXINI SW HW SW FIFOCON read data from CPU BANK 0 SW read data from CPU BANK 0 Figure 22-26. Example of an IN Pipe with 2 Data Banks IN DATA (bank 0) ACK IN DATA (bank 1) ACK HW RXINI SW HW SW FIFOCON read data from CPU BANK 0 SW read data from CPU BANK 1 22.7.3.11 Management of OUT pipes OUT packets are sent by the host. All the data can be written which acknowledges or not the bank when it is full. The pipe must be configured and unfrozen first. The Transmitted OUT Data Interrupt (TXOUTI) bit in UPSTAn is set at the same time as FIFOCON when the current bank is free. This triggers a PnINT interrupt if the Transmitted OUT Data Interrupt Enable (TXOUTE) bit in UPCONn is one. TXOUTI shall be cleared by software (by writing a one to the Transmitted OUT Data Interrupt Clear (TXOUTIC) bit in UPCONnCLR) to acknowledge the interrupt, what has no effect on the pipe FIFO. 396 32059I-06/2010 AT32UC3B The user then writes into the FIFO and clears the FIFOCON bit to allow the USBB to send the data. If the OUT pipe is composed of multiple banks, this also switches to the next bank. The TXOUTI and FIFOCON bits are updated in accordance with the status of the next bank. TXOUTI shall always be cleared before clearing FIFOCON. The UPSTAn.RWALL bit is set when the current bank is not full, i.e., the software can write further data into the FIFO. Note that if the user decides to switch to the Suspend state (by writing a zero to the UHCON.SOFE bit) while a bank is ready to be sent, the USBB automatically exits this state and the bank is sent. Figure 22-27. Example of an OUT Pipe with one Data Bank OUT DATA (bank 0) ACK OUT HW TXOUTI SW SW FIFOCON write data to CPU BANK 0 SW write data to CPU BANK 0 SW Figure 22-28. Example of an OUT Pipe with two Data Banks and no Bank Switching Delay OUT DATA (bank 0) ACK OUT DATA (bank 1) ACK HW TXOUTI SW SW SW FIFOCON write data to CPU SW BANK 0 write data to CPU BANK 1 SW write data to CPU BANK0 397 32059I-06/2010 AT32UC3B Figure 22-29. Example of an OUT Pipe with two Data Banks and a Bank Switching Delay OUT DATA (bank 0) ACK OUT DATA (bank 1) ACK HW TXOUTI SW SW SW FIFOCON write data to CPU BANK 0 SW write data to CPU BANK 1 SW write data to CPU BANK0 22.7.3.12 CRC error This error exists only for isochronous IN pipes. It set the CRC Error Interrupt (CRCERRI) bit, what triggers a PnINT interrupt if then the CRC Error Interrupt Enable (CRCERRE) bit in UPCONn is one. A CRC error can occur during IN stage if the USBB detects a corrupted received packet. The IN packet is stored in the bank as if no CRC error had occurred (RXINI is set). 22.7.3.13 Interrupts See the structure of the USB host interrupt system on Figure 22-6 on page 374. There are two kinds of host interrupts: processing, i.e. their generation is part of the normal processing, and exception, i.e. errors (not related to CPU exceptions). *Global interrupts The processing host global interrupts are: * The Device Connection Interrupt (DCONNI) * The Device Disconnection Interrupt (DDISCI) * The USB Reset Sent Interrupt (RSTI) * The Downstream Resume Sent Interrupt (RSMEDI) * The Upstream Resume Received Interrupt (RXRSMI) * The Host Start of Frame Interrupt (HSOFI) * The Host Wake-Up Interrupt (HWUPI) * The Pipe n Interrupt (PnINT) * The DMA Channel n Interrupt (DMAnINT) There is no exception host global interrupt. *Pipe interrupts The processing host pipe interrupts are: * The Received IN Data Interrupt (RXINI) 398 32059I-06/2010 AT32UC3B * The Transmitted OUT Data Interrupt (TXOUTI) * The Transmitted SETUP Interrupt (TXSTPI) * The Short Packet Interrupt (SHORTPACKETI) * The Number of Busy Banks (NBUSYBK) interrupt The exception host pipe interrupts are: * The Underflow Interrupt (UNDERFI) * The Pipe Error Interrupt (PERRI) * The NAKed Interrupt (NAKEDI) * The Overflow Interrupt (OVERFI) * The Received STALLed Interrupt (RXSTALLDI) * The CRC Error Interrupt (CRCERRI) *DMA interrupts The processing host DMA interrupts are: * The End of USB Transfer Status (EOTSTA) interrupt * The End of Channel Buffer Status (EOCHBUFFSTA) interrupt * The Descriptor Loaded Status (DESCLDSTA) interrupt There is no exception host DMA interrupt. 399 32059I-06/2010 AT32UC3B 22.7.4 22.7.4.1 USB DMA Operation Introduction USB packets of any length may be transferred when required by the USBB. These transfers always feature sequential addressing. These two characteristics mean that in case of high USBB throughput, both HSB ports will benefit from "incrementing burst of unspecified length" since the average access latency of HSB slaves can then be reduced. The DMA uses word "incrementing burst of unspecified length" of up to 256 beats for both data transfers and channel descriptor loading. A burst may last on the HSB busses for the duration of a whole USB packet transfer, unless otherwise broken by the HSB arbitration or the HSB 1kbyte boundary crossing. Packet data HSB bursts may be locked on a DMA buffer basis for drastic overall HSB bus bandwidth performance boost with paged memories. This is because these memories row (or bank) changes, which are very clock-cycle consuming, will then likely not occur or occur once instead of dozens of times during a single big USB packet DMA transfer in case other HSB masters address the memory. This means up to 128 words single cycle unbroken HSB bursts for bulk pipes/endpoints and 256 words single cycle unbroken bursts for isochronous pipes/endpoints. This maximal burst length is then controlled by the lowest programmed USB pipe/endpoint size (PSIZE/EPSIZE) and the Channel Byte Length (CHBYTELENGTH) field in the Device DMA Channel n Control (UDDMAnCONTROL) register. The USBB average throughput may be up to nearly 12 Mbit/s. Its average access latency decreases as burst length increases due to the zero wait-state side effect of unchanged pipe/endpoint. Word access allows reducing the HSB bandwidth required for the USB by four compared to native byte access. If at least 0 wait-state word burst capability is also provided by the other DMA HSB bus slaves, each of both DMA HSB busses need less than 1.1% bandwidth allocation for full USB bandwidth usage at 33MHz, and less than 0.6% at 66MHz. 400 32059I-06/2010 AT32UC3B Figure 22-30. Example of DMA Chained List Transfer Descriptor USB DMA Channel X Registers (Current Transfer Descriptor) Next Descriptor Address Next Descriptor Address HSB Address Control HSB Address Control Status Transfer Descriptor Next Descriptor Address HSB Address Control Transfer Descriptor Next Descriptor Address HSB Address Control NULL Memory Area Data Buffer 1 Data Buffer 2 Data Buffer 3 22.7.4.2 DMA Channel descriptor The DMA channel transfer descriptor is loaded from the memory. Be careful with the alignment of this buffer. The structure of the DMA channel transfer descriptor is defined by three parameters as described below: * Offset 0: - The address must be aligned: 0xXXXX0 - DMA Channel n Next Descriptor Address Register: DMAnNXTDESCADDR * Offset 4: - The address must be aligned: 0xXXXX4 - DMA Channel n HSB Address Register: DMAnADDR * Offset 8: - The address must be aligned: 0xXXXX8 - DMA Channel n Control Register: DMAnCONTROL 22.7.4.3 Programming a chanel: Each DMA transfer is unidirectionnal. Direction depends on the type of the associated endpoint (IN or OUT). Three registers, the UDDMAnNEXTDESC, the UDDMAnADDR and UDDMAnCONTROL need to be programmed to set up wether single or multiple transfer is used. The following example refers to OUT endpoint. For IN endpoint, the programming is symmetric. 401 32059I-06/2010 AT32UC3B *Single-block transfer programming example for OUT transfer : The following sequence may be used: * Configure the targerted endpoint (source) as OUT type, and set the automatic bank switching for this endpoint in the UECFGn register to handle multiple OUT packet. * Write the starting destination address in the UDDMAnADDR register. * There is no need to program the UDDMAnNEXTDESC register. * Program the channel byte length in the UDDMAnCONTROL register. * Program the UDDMAnCONTROL according to Row 2 as shown in Figure 22-6 on page 451 to set up a single block transfer. The UDDMAnSTATUS.CHEN bit is set indicating that the dma channel is enable. As soon as an OUT packet is stored inside the endpoint, the UDDMAnSTATUS.CHACTIVE bit is set to one, indicating that the DMA channel is transfering data from the endpoint to the destination address until the endpoint is empty or the channel byte length is reached. Once the endpoint is empty, the UDDMAnSTATUS.CHACTIVE bit is cleared. Once the DMA channel is completed (i.e : the channel byte length is reached), after one or multiple processed OUT packet, the UDDMAnCONTROL.CHEN bit is cleared. As a consequence, the UDDMAnSTATUS.CHEN bit is also cleared, and the UDDMAnSTATUS.EOCHBUFFSTA bit is set indicating a end of dma channel. If the UDDMAnCONTROL.DMAENDEN bit was set, the last endpoint bank will be properly released even if there are some residual datas inside, i.e: OUT packet truncation at the end of DMA buffer when the dma channel byte lenght is not an integral multiple of the endpoint size. *Programming example for single-block dma transfer with automatic closure for OUT transfer : The idea is to automatically close the DMA transfer at the end of the OUT transaction (received short packet). The following sequence may be used: * Configure the targerted endpoint (source) as OUT type, and set the automatic bank switching for this endpoint in the UECFGn register to handle multiple OUT packet. * Write the starting destination address in the UDDMAnADDR register. * There is no need to program the UDDMAnNEXTDESC register. * Program the channel byte length in the UDDMAnCONTROL register. * Set the BUFFCLOSEINEN bit in the UDDMAnCONTROL register. * Program the UDDMAnCONTROL according to Row 2 as shown in Figure 22-6 on page 451 to set up a single block transfer. As soon as an OUT packet is stored inside the endpoint, the UDDMAnSTATUS.CHACTIVE bit is set to one, indicating that the DMA channel is transfering data from the endpoint to the destination address until the endpoint is empty. Once the endpoint is empty, the UDDMAnSTATUS.CHACTIVE bit is cleared. After one or multiple processed OUT packet, the DMA channel is completed after sourcing a short packet. Then, the UDDMAnCONTROL.CHEN bit is cleared. As a consequence, after a few cycles latency, the UDDMAnSTATUS.CHEN bit is also cleared, and the UDDMAnSTATUS.EOTSTA bit is set indicating that the DMA was closed by a end of USB transaction. 402 32059I-06/2010 AT32UC3B *Programming example for multi-block dma transfer : run and link at end of buffer The idea is to run first a single block transfer followed automatically by a linked list of DMA. The following sequence may be used: * Configure the targerted endpoint (source) as OUT type, and set the automatic bank switching for this endpoint in the UECFGn register to handle multiple OUT packet. * Set up the chain of linked list of descripor in memory. Each descriptor is composed of 3 items : channel next descriptor address, channel destination address and channel control. The last descriptor should be programmed according to row 2 as shown in Figure 22-6 on page 451. * Write the starting destination address in the UDDMAnADDR register. * Program the UDDMAnNEXTDESC register. * Program the channel byte length in the UDDMAnCONTROL register. * Optionnaly set the BUFFCLOSEINEN bit in the UDDMAnCONTROL register. * Program the UDDMAnCONTROL according to Row 4 as shown in Figure 22-6 on page 451. The UDDMAnSTATUS.CHEN bit is set indicating that the dma channel is enable. As soon as an OUT packet is stored inside the endpoint, the UDDMAnSTATUS.CHACTIVE bit is set to one, indicating that the DMA channel is transfering data from the endpoint to the destination address until the endpoint is empty or the channel byte length is reached. Once the endpoint is empty, the UDDMAnSTATUS.CHACTIVE bit is cleared. Once the first DMA channel is completed (i.e : the channel byte length is reached), after one or multiple processed OUT packet, the UDDMAnCONTROL.CHEN bit is cleared. As a consequence, the UDDMAnSTATUS.CHEN bit is also cleared, and the UDDMAnSTATUS.EOCHBUFFSTA bit is set indicating a end of dma channel. If the UDDMAnCONTROL.DMAENDEN bit was set, the last endpoint bank will be properly released even if there are some residual datas inside, i.e: OUT packet truncation at the end of DMA buffer when the dma channel byte lenght is not an integral multiple of the endpoint size. Note that the UDDMAnCONTROL.LDNXTCH bit remains to one indicating that a linked descriptor will be loaded. Once the new descriptor is loaded from the UDDMAnNEXTDESC memory address, the UDDMAnSTATUS.DESCLDSTA bit is set, and the UDDMAnCONTROL register is updated from the memory. As a consequence, the UDDMAnSTATUS.CHEN bit is set, and the UDDMAnSTATUS.CHACTIVE is set as soon as the endpoint is ready to be sourced by the DMA (received OUT data packet). This sequence is repeated until a last linked descriptor is processed. The last descriptor is detected according to row 2 as shown in Figure 22-6 on page 451. At the end of the last descriptor, the UDDMAnCONTROL.CHEN bit is cleared. As a consequence, after a few cycles latency, the UDDMAnSTATUS.CHEN bit is also cleared. *Programming example for multi-block dma transfer : load next descriptor now The idea is to directly run first a linked list of DMA. The following sequence may be used: The following sequence may be used: * Configure the targerted endpoint (source) as OUT type, and set the automatic bank switching for this endpoint in the UECFGn register to handle multiple OUT packet. 403 32059I-06/2010 AT32UC3B * Set up the chain of linked list of descripor in memory. Each descriptor is composed of 3 items : channel next descriptor address, channel destination address and channel control. The last descriptor should be programmed according to row 2 as shown in Figure 22-6 on page 451. * Program the UDDMAnNEXTDESC register. * Program the UDDMAnCONTROL according to Row 3 as shown in Figure 22-6 on page 451. The UDDMAnSTATUS.CHEN bit is 0 and the UDDMAnSTATUS.LDNXTCHDESCEN is set indicating that the DMA channel is pending until the endpoint is ready (received OUT packet). As soon as an OUT packet is stored inside the endpoint, the UDDMAnSTATUS.CHACTIVE bit is set to one. Then after a few cycle latency, the new descriptor is loaded from the memory and the UDDMAnSTATUS.DESCLDSTA is set. At the end of this DMA (for instance when the channel byte length has reached 0), the UDDMAnCONTROL.CHEN bit is cleared, and then the UDDMAnSTATUS.CHEN bit is also cleared. If the UDDMAnCONTROL.LDNXTCH value is one, a new descriptor is loaded. This sequence is repeated until a last linked descriptor is processed. The last descriptor is detected according to row 2 as shown in Figure 22-6 on page 451. At the end of the last descriptor, the UDDMAnCONTROL.CHEN bit is cleared. As a consequence, after a few cycles latency, the UDDMAnSTATUS.CHEN bit is also cleared. 404 32059I-06/2010 AT32UC3B 22.8 User Interface USBB Register Memory Map Register Device General Control Register Device Global Interrupt Register Device Global Interrupt Clear Register Device Global Interrupt Set Register Device Global Interrupt Enable Register Device Global Interrupt Enable Clear Register Device Global Interrupt Enable Set Register Endpoint Enable/Reset Register Device Frame Number Register Endpoint 0 Configuration Register Endpoint 1 Configuration Register Endpoint 2 Configuration Register Endpoint 3 Configuration Register Endpoint 4 Configuration Register Endpoint 5 Configuration Register Endpoint 6 Configuration Register Endpoint 0 Status Register Endpoint 1 Status Register Endpoint 2 Status Register Endpoint 3 Status Register Endpoint 4 Status Register Endpoint 5 Status Register Endpoint 6 Status Register Endpoint 0 Status Clear Register Endpoint 1 Status Clear Register Endpoint 2 Status Clear Register Endpoint 3 Status Clear Register Endpoint 4 Status Clear Register Endpoint 5 Status Clear Register Endpoint 6 Status Clear Register Endpoint 7 Status Clear Register Endpoint 0 Status Set Register Endpoint 1 Status Set Register Endpoint 2 Status Set Register Endpoint 3 Status Set Register Name UDCON UDINT UDINTCLR UDINTSET UDINTE UDINTECLR UDINTESET UERST UDFNUM UECFG0 UECFG1 UECFG2 UECFG3 UECFG4 UECFG5 UECFG6 UESTA0 UESTA1 UESTA2 UESTA3 UESTA4 UESTA5 UESTA6 UESTA0CLR UESTA1CLR UESTA2CLR UESTA3CLR UESTA4CLR UESTA5CLR UESTA6CLR UESTA7CLR UESTA0SET UESTA1SET UESTA2SET UESTA3SET Access Read/Write Read-Only Write-Only Write-Only Read-Only Write-Only Write-Only Read/Write Read-Only Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Reset Value 0x00000100 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00002000 0x00002000 0x00002000 0x00002000 0x00002000 0x00002000 0x00002000 0x00000100 0x00000100 0x00000100 0x00000100 0x00000100 0x00000100 0x00000100 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 Table 22-4. Offset 0x0000 0x0004 0x0008 0x000C 0x0010 0x0014 0x0018 0x001C 0x0020 0x0100 0x0104 0x0108 0x010C 0x0110 0x0114 0x0118 0x0130 0x0134 0x0138 0x013C 0x0140 0x0144 0x0148 0x0160 0x0164 0x0168 0x016C 0x0170 0x0174 0x0178 0x017C 0x0190 0x0194 0x0198 0x019C 405 32059I-06/2010 AT32UC3B Table 22-4. Offset 0x01A0 0x01A4 0x01A8 0x01AC 0x01C0 0x01C4 0x01C8 0x01CC 0x01D0 0x01D4 0x01D8 0x01DC 0x01F0 0x01F4 0x01F8 0x01FC 0x0200 0x0204 0x0208 0x020C 0x0220 0x0224 0x0228 0x022C 0x0230 0x0234 0x0238 0x023C 0x0310 0x0314 0x0318 0x031C 0x0320 USBB Register Memory Map Register Endpoint 4 Status Set Register Endpoint 5 Status Set Register Endpoint 6 Status Set Register Endpoint 7 Status Set Register Endpoint 0 Control Register Endpoint 1 Control Register Endpoint 2 Control Register Endpoint 3 Control Register Endpoint 4 Control Register Endpoint 5 Control Register Endpoint 6 Control Register Endpoint 7 Control Register Endpoint 0 Control Set Register Endpoint 1 Control Set Register Endpoint 2 Control Set Register Endpoint 3 Control Set Register Endpoint 4 Control Set Register Endpoint 5 Control Set Register Endpoint 6 Control Set Register Endpoint 7 Control Set Register Endpoint 0 Control Clear Register Endpoint 1 Control Clear Register Endpoint 2 Control Clear Register Endpoint 3 Control Clear Register Endpoint 4 Control Clear Register Endpoint 5 Control Clear Register Endpoint 6 Control Clear Register Endpoint 7 Control Clear Register Device DMA Channel 1 Next Descriptor Address Register Device DMA Channel 1 HSB Address Register Device DMA Channel 1 Control Register Device DMA Channel 1 Status Register Device DMA Channel 2 Next Descriptor Address Register Name UESTA4SET UESTA5SET UESTA6SET UESTA7SET UECON0 UECON1 UECON2 UECON3 UECON4 UECON5 UECON6 UECON7 UECON0SET UECON1SET UECON2SET UECON3SET UECON4SET UECON5SET UECON6SET UECON7SET UECON0CLR UECON1CLR UECON2CLR UECON3CLR UECON4CLR UECON5CLR UECON6CLR UECON7CLR UDDMA1 NEXTDESC UDDMA1 ADDR UDDMA1 CONTROL UDDMA1 STATUS UDDMA2 NEXTDESC Access Write-Only Write-Only Write-Only Write-Only Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Read/Write Read/Write Read/Write Read/Write Read/Write Reset Value 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 406 32059I-06/2010 AT32UC3B Table 22-4. Offset 0x0324 0x0328 0x032C 0x0330 0x0334 0x0338 0x033C 0x0340 0x0344 0x0348 0x034C 0x0350 0x0354 0x0358 0x035C 0x0360 0x0364 0x0368 0x036C 0x0400 0x0404 0x0408 0x040C 0x0410 USBB Register Memory Map Register Device DMA Channel 2 HSB Address Register Device DMA Channel 2 Control Register Device DMA Channel 2 Status Register Device DMA Channel 3 Next Descriptor Address Register Device DMA Channel 3 HSB Address Register Device DMA Channel 3 Control Register Device DMA Channel 3 Status Register Device DMA Channel 4 Next Descriptor Address Register Device DMA Channel 4 HSB Address Register Device DMA Channel 4 Control Register Device DMA Channel 4 Status Register Device DMA Channel 5 Next Descriptor Address Register Device DMA Channel 5 HSB Address Register Device DMA Channel 5 Control Register Device DMA Channel 5 Status Register Device DMA Channel 6 Next Descriptor Address Register Device DMA Channel 6 HSB Address Register Device DMA Channel 6 Control Register Device DMA Channel 6 Status Register Host General Control Register Host Global Interrupt Register Host Global Interrupt Clear Register Host Global Interrupt Set Register Host Global Interrupt Enable Register Name UDDMA2 ADDR UDDMA2 CONTROL UDDMA2 STATUS UDDMA3 NEXTDESC UDDMA3 ADDR UDDMA3 CONTROL UDDMA3 STATUS UDDMA4 NEXTDESC UDDMA4 ADDR UDDMA4 CONTROL UDDMA4 STATUS UDDMA5 NEXTDESC UDDMA5 ADDR UDDMA5 CONTROL UDDMA5 STATUS UDDMA6 NEXTDESC UDDMA6 ADDR UDDMA6 CONTROL UDDMA6 STATUS UHCON UHINT UHINTCLR UHINTSET UHINTE Access Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read-Only Write-Only Write-Only Read-Only Reset Value 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 407 32059I-06/2010 AT32UC3B Table 22-4. Offset 0x0414 0x0418 0x0041C 0x0420 0x0424 0x0428 0x0500 0x0504 0x0508 0x050C 0x0510 0x0514 0x0518 0x0530 0x0534 0x0538 0x053C 0x0540 0x0544 0x0548 0x0560 0x0564 0x0568 0x056C 0x0570 0x0574 0x0578 0x0590 0x0594 0x0598 0x059C 0x05A0 0x05A4 0x05A8 0x05C0 0x05C4 USBB Register Memory Map Register Host Global Interrupt Enable Clear Register Host Global Interrupt Enable Set Register Pipe Enable/Reset Register Host Frame Number Register Host Address 1 Register Host Address 2 Register Pipe 0 Configuration Register Pipe 1 Configuration Register Pipe 2 Configuration Register Pipe 3 Configuration Register Pipe 4 Configuration Register Pipe 5 Configuration Register Pipe 6 Configuration Register Pipe 0 Status Register Pipe 1 Status Register Pipe 2 Status Register Pipe 3 Status Register Pipe 4 Status Register Pipe 5 Status Register Pipe 6 Status Register Pipe 0 Status Clear Register Pipe 1 Status Clear Register Pipe 2 Status Clear Register Pipe 3 Status Clear Register Pipe 4 Status Clear Register Pipe 5 Status Clear Register Pipe 6 Status Clear Register Pipe 0 Status Set Register Pipe 1 Status Set Register Pipe 2 Status Set Register Pipe 3 Status Set Register Pipe 4 Status Set Register Pipe 5 Status Set Register Pipe 6 Status Set Register Pipe 0 Control Register Pipe 1 Control Register Name UHINTECLR UHINTESET UPRST UHFNUM UHADDR1 UHADDR2 UPCFG0 UPCFG1 UPCFG2 UPCFG3 UPCFG4 UPCFG5 UPCFG6 UPSTA0 UPSTA1 UPSTA2 UPSTA3 UPSTA4 UPSTA5 UPSTA6 UPSTA0CLR UPSTA1CLR UPSTA2CLR UPSTA3CLR UPSTA4CLR UPSTA5CLR UPSTA6CLR UPSTA0SET UPSTA1SET UPSTA2SET UPSTA3SET UPSTA4SET UPSTA5SET UPSTA6SET UPCON0 UPCON1 Access Write-Only Write-Only Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Read-Only Read-Only Reset Value 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 408 32059I-06/2010 AT32UC3B Table 22-4. Offset 0x05C8 0x05CC 0x05D0 0x05D4 0x05D8 0x05DC 0x05F0 0x05F4 0x05F8 0x05FC 0x0600 0x0604 0x0608 0x0620 0x0624 0x0628 0x062C 0x0630 0x0634 0x0638 0x0650 0x0654 0x0658 0x065C 0x0660 0x0664 0x0668 0x0680 0x0684 0x0688 0x068C 0x0690 0x0694 0x0698 0x0710 USBB Register Memory Map Register Pipe 2 Control Register Pipe 3 Control Register Pipe 4 Control Register Pipe 5 Control Register Pipe 6 Control Register Pipe 7 Control Register Pipe 0 Control Set Register Pipe 1 Control Set Register Pipe 2 Control Set Register Pipe 3 Control Set Register Pipe 4 Control Set Register Pipe 5 Control Set Register Pipe 6 Control Set Register Pipe 0 Control Clear Register Pipe 1 Control Clear Register Pipe 2 Control Clear Register Pipe 3 Control Clear Register Pipe 4 Control Clear Register Pipe 5 Control Clear Register Pipe 6 Control Clear Register Pipe 0 IN Request Register Pipe 1 IN Request Register Pipe 2 IN Request Register Pipe 3 IN Request Register Pipe 4 IN Request Register Pipe 5 IN Request Register Pipe 6 IN Request Register Pipe 0 Error Register Pipe 1 Error Register Pipe 2 Error Register Pipe 3 Error Register Pipe 4 Error Register Pipe 5 Error Register Pipe 6 Error Register Host DMA Channel 1 Next Descriptor Address Register Name UPCON2 UPCON3 UPCON4 UPCON5 UPCON6 UPCON7 UPCON0SET UPCON1SET UPCON2SET UPCON3SET UPCON4SET UPCON5SET UPCON6SET UPCON0CLR UPCON1CLR UPCON2CLR UPCON3CLR UPCON4CLR UPCON5CLR UPCON6CLR UPINRQ0 UPINRQ1 UPINRQ2 UPINRQ3 UPINRQ4 UPINRQ5 UPINRQ6 UPERR0 UPERR1 UPERR2 UPERR3 UPERR4 UPERR5 UPERR6 UHDMA1 NEXTDESC Access Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Write-Only Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Reset Value 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 409 32059I-06/2010 AT32UC3B Table 22-4. Offset 0x0714 0x0718 0x071C 0x0720 0x0724 0x0728 0x072C 0x0730 0x0734 0x0738 0x073C 0x0740 0x0744 0x0748 0x074C 0x0750 0x0754 0x0758 0x075C 0x0760 0x0764 0x0768 USBB Register Memory Map Register Host DMA Channel 1 HSB Address Register Host DMA Channel 1 Control Register Host DMA Channel 1 Status Register Host DMA Channel 2 Next Descriptor Address Register Host DMA Channel 2 HSB Address Register Host DMA Channel 2 Control Register Host DMA Channel 2 Status Register Host DMA Channel 3 Next Descriptor Address Register Host DMA Channel 3 HSB Address Register Host DMA Channel 3 Control Register Host DMA Channel 3Status Register Host DMA Channel 4 Next Descriptor Address Register Host DMA Channel 4 HSB Address Register Host DMA Channel 4 Control Register Host DMA Channel 4 Status Register Host DMA Channel 5 Next Descriptor Address Register Host DMA Channel 5 HSB Address Register Host DMA Channel 5 Control Register Host DMA Channel 5 Status Register Host DMA Channel 6 Next Descriptor Address Register Host DMA Channel 6 HSB Address Register Host DMA Channel 6 Control Register Name UHDMA1 ADDR UHDMA1 CONTROL UHDMA1 STATUS UHDMA2 NEXTDESC UHDMA2 ADDR UHDMA2 CONTROL UHDMA2 STATUS UHDMA3 NEXTDESC UHDMA3 ADDR UHDMA3 CONTROL UHDMA3 STATUS UHDMA4 NEXTDESC UHDMA4 ADDR UHDMA4 CONTROL UHDMA4 STATUS UHDMA5 NEXTDESC UHDMA5 ADDR UHDMA5 CONTROL UHDMA5 STATUS UHDMA6 NEXTDESC UHDMA6 ADDR UHDMA6 CONTROL Access Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Reset Value 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 410 32059I-06/2010 AT32UC3B Table 22-4. Offset 0x076C 0x0800 0x0804 0x0808 0x080C 0x0818 0x081C 0x0820 0x0824 0x0828 0x082C USBB Register Memory Map Register Host DMA Channel 6 Status Register General Control Register General Status Register General Status Clear Register General Status Set Register IP Version Register IP Features Register IP PB Address Size Register IP Name Register 1 IP Name Register 2 USB Finite State Machine Status Register Name UHDMA6 STATUS USBCON USBSTA USBSTACLR USBSTASET UVERS UFEATURES UADDRSIZE UNAME1 UNAME2 USBFSM Access Read/Write Read/Write Read-Only Write-Only Write-Only Read-Only Read-Only Read-Only Read-Only Read-Only Read-Only Reset Value 0x00000000 0x03004000 0x00000400 0x00000000 0x00000000 -(1) -(1) -(1) -(1) -(1) 0x00000009 Table 22-5. Offset 0x00000 0x0FFFC 0x10000 0x1FFFC 0x20000 0x2FFFC 0x30000 0x3FFFC 0x40000 0x4FFFC 0x50000 0x5FFFC 0x60000 0x6FFFC Note: USB HSB Memory Map Register Pipe/Endpoint 0 FIFO Data Register Pipe/Endpoint 1 FIFO Data Register Pipe/Endpoint 2 FIFO Data Register Pipe/Endpoint 3 FIFO Data Register Pipe/Endpoint 4 FIFO Data Register Pipe/Endpoint 5 FIFO Data Register Pipe/Endpoint 6 FIFO Data Register Name USB FIFO0DATA USB FIFO1DATA USB FIFO2DATA USB FIFO3DATA USB FIFO4DATA USB FIFO5DATA USB FIFO6DATA Access Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Read/Write Reset Value Undefined Undefined Undefined Undefined Undefined Undefined Undefined 1. The reset values are device specific. Please refer to the Module Configuration section at the end of this chapter. 411 32059I-06/2010 AT32UC3B 22.8.1 22.8.1.1 USB General Registers General Control Register USBCON Read/Write 0x0800 0x03004000 Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 UIMOD 24 UIDE 23 - 22 UNLOCK 21 TIMPAGE 20 19 - 18 - 17 TIMVALUE 16 15 USBE 14 FRZCLK 13 VBUSPO 12 OTGPADE 11 10 9 8 VBUSHWC 7 STOE 6 5 ROLEEXE 4 BCERRE 3 VBERRE 2 1 VBUSTE 0 IDTE * UIMOD: USBB Mode This bit has no effect when UIDE is one (USB_ID input pin activated). 0: The module is in USB host mode. 1: The module is in USB device mode. This bit can be written even if USBE is zero or FRZCLK is one. Disabling the USBB (by writing a zero to the USBE bit) does not reset this bit. UIDE: USB_ID Pin Enable 0: The USB mode (device/host) is selected from the UIMOD bit. 1: The USB mode (device/host) is selected from the USB_ID input pin. This bit can be written even if USBE is zero or FRZCLK is one. Disabling the USBB (by writing a zero to the USBE bit) does not reset this bit. UNLOCK: Timer Access Unlock 1: The TIMPAGE and TIMVALUE fields are unlocked. 0: The TIMPAGE and TIMVALUE fields are locked. The TIMPAGE and TIMVALUE fields can always be read, whatever the value of UNLOCK. TIMPAGE: Timer Page This field contains the page value to access a special timer register. TIMVALUE: Timer Value This field selects the timer value that is written to the special time register selected by TIMPAGE. See Section 22.7.1.8 for details. USBE: USBB Enable Writing a zero to this bit will reset the USBB, disable the USB transceiver and, disable the USBB clock inputs. Unless explicitly stated, all registers then will become read-only and will be reset. 1: The USBB is enabled. 0: The USBB is disabled. This bit can be written even if FRZCLK is one. FRZCLK: Freeze USB Clock 1: The clock input are disabled (the resume detection is still active).This reduces power consumption. Unless explicitly stated, all * * * * * * 412 32059I-06/2010 AT32UC3B registers then become read-only. 0: The clock inputs are enabled. This bit can be written even if USBE is zero. Disabling the USBB (by writing a zero to the USBE bit) does not reset this bit, but this freezes the clock inputs whatever its value. VBUSPO: VBus Polarity 1: The USB_VBOF output signal is inverted (active low). 0: The USB_VBOF output signal is in its default mode (active high). To be generic. May be useful to control an external VBus power module. This bit can be written even if USBE is zero or FRZCLK is one. Disabling the USBB (by writing a zero to the USBE bit) does not reset this bit. OTGPADE: OTG Pad Enable 1: The OTG pad is enabled. 0: The OTG pad is disabled. This bit can be written even if USBE is zero or FRZCLK is one. Disabling the USBB (by writing a zero to the USBE bit) does not reset this bit. VBUSHWC: VBus Hardware Control 1: The hardware control over the USB_VBOF output pin is disabled. 0: The hardware control over the USB_VBOF output pin is enabled. The USBB resets the USB_VBOF output pin when a VBUS problem occurs. STOE: Suspend Time-Out Interrupt Enable 1: The Suspend Time-Out Interrupt (STOI) is enabled. 0: The Suspend Time-Out Interrupt (STOI) is disabled. ROLEEXE: Role Exchange Interrupt Enable 1: The Role Exchange Interrupt (ROLEEXI) is enabled. 0: The Role Exchange Interrupt (ROLEEXI) is disabled. BCERRE: B-Connection Error Interrupt Enable 1: The B-Connection Error Interrupt (BCERRI) is enabled. 0: The B-Connection Error Interrupt (BCERRI) is disabled. VBERRE: VBus Error Interrupt Enable 1: The VBus Error Interrupt (VBERRI) is enabled. 0: The VBus Error Interrupt (VBERRI) is disabled. VBUSTE: VBus Transition Interrupt Enable 1: The VBus Transition Interrupt (VBUSTI) is enabled. 0: The VBus Transition Interrupt (VBUSTI) is disabled. IDTE: ID Transition Interrupt Enable 1: The ID Transition interrupt (IDTI) is enabled. 0: The ID Transition interrupt (IDTI) is disabled. * * * * * * * * * 413 32059I-06/2010 AT32UC3B 22.8.1.2 General Status Register USBSTA Read-Only 0x0804 0x00000400 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 SPEED 12 11 VBUS 10 ID 9 VBUSRQ 8 - 7 STOI 6 5 ROLEEXI 4 BCERRI 3 VBERRI 2 1 VBUSTI 0 IDTI * SPEED: Speed Status This field is set according to the controller speed mode. This field shall only be used in device mode. SPEED 0 1 X 0 0 1 Speed Status Full-Speed mode Low-Speed mode Reserved * VBUS: VBus Level This bit is set when the VBus line level is high, even if USBE is zero. This bit is cleared when the VBus line level is low, even if USBE is zero. This bit can be used in device mode to monitor the USB bus connection state of the application. * ID: USB_ID Pin State This bit is cleared when the USB_ID level is low, even if USBE is zero. This bit is set when the USB_ID level is high, event if USBE is zero. * VBUSRQ: VBus Request This bit is set when the USBSTASET.VBUSRQS bit is written to one. This bit is cleared when the USBSTACLR.VBUSRQC bit is written to one or when a VBus error occurs and VBUSHWC is zero. 1: The USB_VBOF output pin is driven high to enable the VBUS power supply generation. 0: The USB_VBOF output pin is driven low to disable the VBUS power supply generation. This bit shall only be used in host mode. 414 32059I-06/2010 AT32UC3B * STOI: Suspend Time-Out Interrupt This bit is set when a time-out error (more than 200ms) has been detected after a suspend. This triggers a USB interrupt if STOE is one. This bit is cleared when the UBSTACLR.STOIC bit is written to one. This bit shall only be used in host mode. ROLEEXI: Role Exchange Interrupt This bit is set when the USBB has successfully switched its mode because of an negotiation (host to device or device to host). This triggers a USB interrupt if ROLEEXE is one. This bit is cleared when the UBSTACLR.ROLEEXIC bit is written to one. BCERRI: B-Connection Error Interrupt This bit is set when an error occurs during the B-connection. This triggers a USB interrupt if BCERRE is one. This bit is cleared when the UBSTACLR.BCERRIC bit is written to one. This bit shall only be used in host mode. VBERRI: VBus Error Interrupt This bit is set when a VBus drop has been detected. This triggers a USB interrupt if VBERRE is one. This bit is cleared when the UBSTACLR.VBERRIC bit is written to one. This bit shall only be used in host mode. If a VBus problem occurs, then the VBERRI interrupt is generated even if the USBB does not go to an error state because of VBUSHWC is one. VBUSTI: VBus Transition Interrupt This bit is set when a transition (high to low, low to high) has been detected on the USB_VBUS pad. This triggers an USB interrupt if VBUSTE is one. This bit is cleared when the UBSTACLR.VBUSTIC bit is written to one. This interrupt is generated even if the clock is frozen by the FRZCLK bit. IDTI: ID Transition Interrupt This bit is set when a transition (high to low, low to high) has been detected on the USB_ID input pin. This triggers an USB interrupt if IDTE is one. This bit is cleared when the UBSTACLR.IDTIC bit is written to one. This interrupt is generated even if the clock is frozen by the FRZCLK bit. * * * * * 415 32059I-06/2010 AT32UC3B 22.8.1.3 General Status Clear Register USBSTACLR Write-Only 0x0808 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 VBUSRQC 8 - 7 STOIC 6 5 ROLEEXIC 4 BCERRIC 3 VBERRIC 2 1 VBUSTIC 0 IDTIC Writing a one to a bit in this register will clear the corresponding bit in UBSTA. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 416 32059I-06/2010 AT32UC3B 22.8.1.4 General Status Set Register USBSTASET Write-Only 0x080C 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 VBUSRQS 8 - 7 STOIS 6 5 ROLEEXIS 4 BCERRIS 3 VBERRIS 2 1 VBUSTIS 0 IDTIS Writing a one to a bit in this register will set the corresponding bit in UBSTA, what may be useful for test or debug purposes. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 417 32059I-06/2010 AT32UC3B 22.8.1.5 Version Register UVERS Read-Only 0x0818 - Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 18 VARIANT 17 16 15 - 14 - 13 - 12 - 11 10 9 8 VERSION[11:8] 7 6 5 4 VERSION[7:0] 3 2 1 0 * VARIANT: Variant Number Reserved. No functionality associated. * VERSION: Version Number Version number of the module. No functionality associated. 418 32059I-06/2010 AT32UC3B 22.8.1.6 Features Register UFEATURES Read-Only 0x081C - Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 BYTEWRITE DPRAM 14 13 FIFOMAXSIZE 12 11 10 9 8 DMAFIFOWORDDEPTH 7 DMABUFFE RSIZE 6 5 DMACHANNELNBR 4 3 2 EPTNBRMAX 1 0 * BYTEWRITEDPRAM: DPRAM Byte-Write Capability 1: The DPRAM is natively byte-write capable. 0: The DPRAM byte write lanes have shadow logic implemented in the USBB IP interface. * FIFOMAXSIZE: Maximal FIFO Size This field indicates the maximal FIFO size, i.e., the DPRAM size: FIFOMAXSIZE 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 Maximal FIFO Size < 256 bytes < 512 bytes < 1024 bytes < 2048 bytes < 4096 bytes < 8192 bytes < 16384 bytes >= 16384 bytes 419 32059I-06/2010 AT32UC3B * DMAFIFOWORDDEPTH: DMA FIFO Depth in Words This field indicates the DMA FIFO depth controller in words: DMAFIFOWORDDEPTH 0 0 0 0 0 0 0 0 1 0 1 0 DMA FIFO Depth in Words 16 1 2 ... 1 1 1 1 15 * DMABUFFERSIZE: DMA Buffer Size 1: The DMA buffer size is 24bits. 0: The DMA buffer size is 16bits. * DMACHANNELNBR: Number of DMA Channels This field indicates the number of hardware-implemented DMA channels: DMACHANNELNBR 0 0 0 0 0 1 0 1 0 Number of DMA Channels Reserved 1 2 ... 1 1 1 7 * EPTNBRMAX: Maximal Number of Pipes/Endpoints This field indicates the number of hardware-implemented pipes/endpoints: EPTNBRMAX 0 0 0 0 0 0 0 0 1 0 1 0 Maximal Number of Pipes/Endpoints 16 1 2 ... 1 1 1 1 15 420 32059I-06/2010 AT32UC3B 22.8.1.7 Address Size Register UADDRSIZE Read-Only 0x0820 - Register Name: Access Type: Offset: Read Value: 31 30 29 28 27 26 25 24 UADDRSIZE[31:24] 23 22 21 20 19 18 17 16 UADDRSIZE[23:16] 15 14 13 12 11 10 9 8 UADDRSIZE[15:8] 7 6 5 4 3 2 1 0 UADDRSIZE[7:0] * UADDRSIZE: IP PB Address Size This field indicates the size of the PB address space reserved for the USBB IP interface. 421 32059I-06/2010 AT32UC3B 22.8.1.8 Name Register 1 UNAME1 Read-Only 0x0824 - Register Name: Access Type: Offset: Read Value: 31 30 29 28 27 26 25 24 UNAME1[31:24] 23 22 21 20 19 18 17 16 UNAME1[23:16] 15 14 13 12 11 10 9 8 UNAME1[15:8] 7 6 5 4 UNAME1[7:0] 3 2 1 0 * UNAME1: IP Name Part One This field indicates the first part of the ASCII-encoded name of the USBB IP. 422 32059I-06/2010 AT32UC3B 22.8.1.9 Name Register 2 UNAME2 Read-Only 0x0828 Register Name: Access Type: Offset: Read Value: 31 30 29 28 27 26 25 24 UNAME2[31:24] 23 22 21 20 19 18 17 16 UNAME2[23:16] 15 14 13 12 11 10 9 8 UNAME2[15:8] 7 6 5 4 UNAME2[7:0] 3 2 1 0 * UNAME2: IP Name Part Two This field indicates the second part of the ASCII-encoded name of the USBB IP. 423 32059I-06/2010 AT32UC3B 22.8.1.10 Finite State Machine Status Register USBFSM Read-Only 0x082C 0x00000009 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 2 DRDSTATE 1 0 * DRDSTATE This field indicates the state of the USBB. DRDSTATE 0 1 2 3 4 5 6 7 8 9 10 11 12 Description a_idle state: this is the start state for A-devices (when the ID pin is 0) a_wait_vrise: In this state, the A-device waits for the voltage on VBus to rise above the Adevice VBus Valid threshold (4.4 V). a_wait_bcon: In this state, the A-device waits for the B-device to signal a connection. a_host: In this state, the A-device that operates in Host mode is operational. a_suspend: The A-device operating as a host is in the suspend mode. a_peripheral: The A-device operates as a peripheral. a_wait_vfall: In this state, the A-device waits for the voltage on VBus to drop below the Adevice Session Valid threshold (1.4 V). a_vbus_err: In this state, the A-device waits for recovery of the over-current condition that caused it to enter this state. a_wait_discharge: In this state, the A-device waits for the data usb line to discharge (100 us). b_idle: this is the start state for B-device (when the ID pin is 1). b_peripheral: In this state, the B-device acts as the peripheral. b_wait_begin_hnp: In this state, the B-device is in suspend mode and waits until 3 ms before initiating the HNP protocol if requested. b_wait_discharge: In this state, the B-device waits for the data usb line to discharge (100 us) before becoming Host. 424 32059I-06/2010 AT32UC3B DRDSTATE 13 14 15 Description b_wait_acon: In this state, the B-device waits for the A-device to signal a connect before becoming B-Host. b_host: In this state, the B-device acts as the Host. b_srp_init: In this state, the B-device attempts to start a session using the SRP protocol. 425 32059I-06/2010 AT32UC3B 22.8.2 22.8.2.1 USB Device Registers Device General Control Register UDCON Read/Write 0x0000 0x00000100 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 LS 11 - 10 - 9 RMWKUP 8 DETACH 7 ADDEN 6 5 4 3 UADD 2 1 0 * LS: Low-Speed Mode Force 1: The low-speed mode is active. 0: The full-speed mode is active. This bit can be written even if USBE is zero or FRZCLK is one. Disabling the USBB (by writing a zero to the USBE bit) does not reset this bit. RMWKUP: Remote Wake-Up Writing a one to this bit will send an upstream resume to the host for a remote wake-up. Writing a zero to this bit has no effect. This bit is cleared when the USBB receive a USB reset or once the upstream resume has been sent. DETACH: Detach Writing a one to this bit will physically detach the device (disconnect internal pull-up resistor from D+ and D-). Writing a zero to this bit will reconnect the device. ADDEN: Address Enable Writing a one to this bit will activate the UADD field (USB address). Writing a zero to this bit has no effect. This bit is cleared when a USB reset is received. UADD: USB Address This field contains the device address. This field is cleared when a USB reset is received. * * * * 426 32059I-06/2010 AT32UC3B 22.8.2.2 Device Global Interrupt Register UDINT Read-Only 0x0004 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 DMA6INT 29 DMA5INT 28 DMA4INT 27 DMA3INT 26 DMA2INT 25 DMA1INT 24 - 23 - 22 - 21 - 20 - 19 - 18 EP6INT 17 EP5INT 16 EP4INT 15 EP3INT 14 EP2INT 13 EP1INT 12 EP0INT 11 - 10 - 9 - 8 - 7 - 6 UPRSM 5 EORSM 4 WAKEUP 3 EORST 2 SOF 1 - 0 SUSP * DMAnINT: DMA Channel n Interrupt * This bit is set when an interrupt is triggered by the DMA channel n. This triggers a USB interrupt if DMAnINTE is one. This bit is cleared when the UDDMAnSTATUS interrupt source is cleared. EPnINT: Endpoint n Interrupt This bit is set when an interrupt is triggered by the endpoint n (UESTAn, UECONn). This triggers a USB interrupt if EPnINTE is one. This bit is cleared when the interrupt source is serviced. UPRSM: Upstream Resume Interrupt This bit is set when the USBB sends a resume signal called "Upstream Resume". This triggers a USB interrupt if UPRSME is one. This bit is cleared when the UDINTCLR.UPRSMC bit is written to one to acknowledge the interrupt (USB clock inputs must be enabled before). EORSM: End of Resume Interrupt This bit is set when the USBB detects a valid "End of Resume" signal initiated by the host. This triggers a USB interrupt if EORSME is one. This bit is cleared when the UDINTCLR.EORSMC bit is written to one to acknowledge the interrupt. WAKEUP: Wake-Up Interrupt This bit is set when the USBB is reactivated by a filtered non-idle signal from the lines (not by an upstream resume). This triggers an interrupt if WAKEUPE is one. This bit is cleared when the UDINTCLR.WAKEUPC bit is written to one to acknowledge the interrupt (USB clock inputs must be enabled before). This bit is cleared when the Suspend (SUSP) interrupt bit is set. This interrupt is generated even if the clock is frozen by the FRZCLK bit. EORST: End of Reset Interrupt This bit is set when a USB "End of Reset" has been detected. This triggers a USB interrupt if EORSTE is one. This bit is cleared when the UDINTCLR.EORSTC bit is written to one to acknowledge the interrupt. SOF: Start of Frame Interrupt This bit is set when a USB "Start of Frame" PID (SOF) has been detected (every 1 ms). This triggers a USB interrupt if SOFE is one. The FNUM field is updated. This bit is cleared when the UDINTCLR.SOFC bit is written to one to acknowledge the interrupt. * * * * * 427 32059I-06/2010 AT32UC3B * SUSP: Suspend Interrupt This bit is set when a USB "Suspend" idle bus state has been detected for 3 frame periods (J state for 3 ms). This triggers a USB interrupt if SUSPE is one. This bit is cleared when the UDINTCLR.SUSPC bit is written to one to acknowledge the interrupt. This bit is cleared when the Wake-Up (WAKEUP) interrupt bit is set. 428 32059I-06/2010 AT32UC3B 22.8.2.3 Device Global Interrupt Clear Register UDINTCLR Write-Only 0x0008 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 UPRSMC 5 EORSMC 4 WAKEUPC 3 EORSTC 2 SOFC 1 - 0 SUSPC Writing a one to a bit in this register will clear the corresponding bit in UDINT. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 429 32059I-06/2010 AT32UC3B 22.8.2.4 Device Global Interrupt Set Register UDINTSET Write-Only 0x000C 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 DMA6INTS 29 DMA5INTS 28 DMA4INTS 27 DMA3INTS 26 DMA2INTS 25 DMA1INTS 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 UPRSMS 5 EORSMS 4 WAKEUPS 3 EORSTS 2 SOFS 1 - 0 SUSPS Writing a one to a bit in this register will set the corresponding bit in UDINT, what may be useful for test or debug purposes. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 430 32059I-06/2010 AT32UC3B 22.8.2.5 Device Global Interrupt Enable Register UDINTE Read-Only 0x0010 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 DMA6INTE 29 DMA5INTE 28 DMA4INTE 27 DMA3INTE 26 DMA2INTE 25 DMA1INTE 24 - 23 - 22 - 21 - 20 - 19 - 18 EP6INTE 17 EP5INTE 16 EP4INTE 15 EP3INTE 14 EP2INTE 13 EP1INTE 12 EP0INTE 11 - 10 - 9 - 8 - 7 - 6 UPRSME 5 EORSME 4 WAKEUPE 3 EORSTE 2 SOFE 1 - 0 SUSPE 1: The corresponding interrupt is enabled. 0: The corresponding interrupt is disabled. A bit in this register is set when the corresponding bit in UDINTESET is written to one. A bit in this register is cleared when the corresponding bit in UDINTECLR is written to one. 431 32059I-06/2010 AT32UC3B 22.8.2.6 Device Global Interrupt Enable Clear Register UDINTECLR Write-Only 0x0014 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 DMA6INTEC 29 DMA5INTEC 28 DMA4INTEC 27 DMA3INTEC 26 DMA2INTEC 25 DMA1INTEC 24 - 23 - 22 - 21 - 20 - 19 - 18 EP6INTEC 17 EP5INTEC 16 EP4INTEC 15 EP3INTEC 14 EP2INTEC 13 EP1INTEC 12 EP0INTEC 11 - 10 - 9 - 8 - 7 - 6 UPRSMEC 5 EORSMEC 4 WAKEUPEC 3 EORSTEC 2 SOFEC 1 - 0 SUSPEC Writing a one to a bit in this register will clear the corresponding bit in UDINTE. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 432 32059I-06/2010 AT32UC3B 22.8.2.7 Device Global Interrupt Enable Set Register UDINTESET Write-Only 0x0018 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 DMA6INTES 29 DMA5INTES 28 DMA4INTES 27 DMA3INTES 26 DMA2INTES 25 DMA1INTES 24 - 23 - 22 - 21 - 20 - 19 - 18 EP6INTES 17 EP5INTES 16 EP4INTES 15 EP3INTES 14 EP2INTES 13 EP1INTES 12 EP0INTES 11 - 10 - 9 - 8 - 7 - 6 UPRSMES 5 EORSMES 4 WAKEUPES 3 EORSTES 2 SOFES 1 - 0 SUSPES Writing a one to a bit in this register will set the corresponding bit in UDINTE. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 433 32059I-06/2010 AT32UC3B 22.8.2.8 Endpoint Enable/Reset Register UERST Read/Write 0x001C 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 EPRST6 21 EPRST5 20 EPRST4 19 EPRST3 18 EPRST2 17 EPRST1 16 EPRST0 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 EPEN6 5 EPEN5 4 EPEN4 3 EPEN3 2 EPEN2 1 EPEN1 0 EPEN0 * EPRSTn: Endpoint n Reset Writing a one to this bit will reset the endpoint n FIFO prior to any other operation, upon hardware reset or when a USB bus reset has been received. This resets the endpoint n registers (UECFGn, UESTAn, UECONn) but not the endpoint configuration (ALLOC, EPBK, EPSIZE, EPDIR, EPTYPE). All the endpoint mechanism (FIFO counter, reception, transmission, etc.) is reset apart from the Data Toggle Sequence field (DTSEQ) which can be cleared by setting the RSTDT bit (by writing a one to the RSTDTS bit). The endpoint configuration remains active and the endpoint is still enabled. Writing a zero to this bit will complete the reset operation and start using the FIFO. This bit is cleared upon receiving a USB reset. * EPENn: Endpoint n Enable 1: The endpoint n is enabled. 0: The endpoint n is disabled, what forces the endpoint n state to inactive (no answer to USB requests) and resets the endpoint n registers (UECFGn, UESTAn, UECONn) but not the endpoint configuration (ALLOC, EPBK, EPSIZE, EPDIR, EPTYPE). 434 32059I-06/2010 AT32UC3B 22.8.2.9 Device Frame Number Register UDFNUM Read-Only 0x0020 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 FNCERR 14 - 13 12 11 FNUM[10:5] 10 9 8 7 6 5 FNUM[4:0] 4 3 2 - 1 - 0 - * FNCERR: Frame Number CRC Error This bit is set when a corrupted frame number is received. This bit and the SOF interrupt bit are updated at the same time. This bit is cleared upon receiving a USB reset. * FNUM: Frame Number This field contains the 11-bit frame number information. It is provided in the last received SOF packet. This field is cleared upon receiving a USB reset. FNUM is updated even if a corrupted SOF is received. 435 32059I-06/2010 AT32UC3B 22.8.2.10 Endpoint n Configuration Register UECFGn, n in [0..6] Read/Write 0x0100 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 EPTYPE 11 10 - 9 AUTOSW 8 EPDIR 7 - 6 5 EPSIZE 4 3 EPBK 2 1 ALLOC 0 - * EPTYPE: Endpoint Type This field shall be written to select the endpoint type: EPTYPE 0 0 1 1 0 1 0 1 Endpoint Type Control Isochronous Bulk Interrupt This field is cleared upon receiving a USB reset. 436 32059I-06/2010 AT32UC3B * AUTOSW: Automatic Switch This bit is cleared upon receiving a USB reset. 1: The automatic bank switching is enabled. 0: The automatic bank switching is disabled. * EPDIR: Endpoint Direction This bit is cleared upon receiving a USB reset. 1: The endpoint direction is IN (nor for control endpoints). 0: The endpoint direction is OUT. * EPSIZE: Endpoint Size This field shall be written to select the size of each endpoint bank: EPSIZE 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 Endpoint Size 8 bytes 16 bytes 32 bytes 64 bytes 128 bytes 256 bytes 512 bytes 1024 bytes This field is cleared upon receiving a USB reset (except for the endpoint 0). * EPBK: Endpoint Banks This field shall be written to select the number of banks for the endpoint: EPBK 0 0 1 1 0 1 0 1 Endpoint Banks 1 (single-bank endpoint) 2 (double-bank endpoint) 3 (triple-bank endpoint) Reserved For control endpoints, a single-bank endpoint (0b00) shall be selected. This field is cleared upon receiving a USB reset (except for the endpoint 0). * ALLOC: Endpoint Memory Allocate Writing a one to this bit will allocate the endpoint memory. The user should check the CFGOK bit to know whether the allocation of this endpoint is correct. Writing a zero to this bit will free the endpoint memory. This bit is cleared upon receiving a USB reset (except for the endpoint 0). 437 32059I-06/2010 AT32UC3B 22.8.2.11 Endpoint n Status Register UESTAn, n in [0..6] Read-Only 0x0100 0x0130 + (n * 0x04) 0x00000100 Register Name: Access Type: Offset: Reset Value: 31 - 30 29 28 27 BYCT 26 25 24 23 22 BYCT 21 20 19 - 18 CFGOK 17 CTRLDIR 16 RWALL 15 CURRBK 14 13 NBUSYBK 12 11 - 10 - 9 DTSEQ 8 7 SHORT PACKET 6 STALLEDI/ CRCERRI 5 OVERFI 4 NAKINI 3 NAKOUTI 2 RXSTPI/ UNDERFI 1 RXOUTI 0 TXINI * BYCT: Byte Count This field is set with the byte count of the FIFO. For IN endpoints, incremented after each byte written by the software into the endpoint and decremented after each byte sent to the host. For OUT endpoints, incremented after each byte received from the host and decremented after each byte read by the software from the endpoint. This field may be updated one clock cycle after the RWALL bit changes, so the user should not poll this field as an interrupt bit. * CFGOK: Configuration OK Status This bit is updated when the ALLOC bit is written to one. This bit is set if the endpoint n number of banks (EPBK) and size (EPSIZE) are correct compared to the maximal allowed number of banks and size for this endpoint and to the maximal FIFO size (i.e. the DPRAM size). If this bit is cleared, the user shall rewrite correct values to the EPBK and EPSIZE fields in the UECFGn register. * CTRLDIR: Control Direction This bit is set after a SETUP packet to indicate that the following packet is an IN packet. This bit is cleared after a SETUP packet to indicate that the following packet is an OUT packet. Writing a zero or a one to this bit has no effect. * RWALL: Read/Write Allowed This bit is set for IN endpoints when the current bank is not full, i.e., the user can write further data into the FIFO. This bit is set for OUT endpoints when the current bank is not empty, i.e., the user can read further data from the FIFO. This bit is never set if STALLRQ is one or in case of error. This bit is cleared otherwise. This bit shall not be used for control endpoints. 438 32059I-06/2010 AT32UC3B * CURRBK: Current Bank This bit is set for non-control endpoints, to indicate the current bank: CURRBK 0 0 1 1 0 1 0 1 Current Bank Bank0 Bank1 Bank2 Reserved This field may be updated one clock cycle after the RWALL bit changes, so the user should not poll this field as an interrupt bit. * NBUSYBK: Number of Busy Banks This field is set to indicate the number of busy banks: NBUSYBK 0 0 1 1 0 1 0 1 Number of Busy Banks 0 (all banks free) 1 2 3 For IN endpoints, it indicates the number of banks filled by the user and ready for IN transfer. When all banks are free, this triggers an EPnINT interrupt if NBUSYBKE is one. For OUT endpoints, it indicates the number of banks filled by OUT transactions from the host. When all banks are busy, this triggers an EPnINT interrupt if NBUSYBKE is one. When the FIFOCON bit is cleared (by writing a one to the FIFOCONC bit) to validate a new bank, this field is updated two or three clock cycles later to calculate the address of the next bank. An EPnINT interrupt is triggered if: - for IN endpoint, NBUSYBKE is one and all the banks are free. - for OUT endpoint, NBUSYBKE is one and all the banks are busy. * DTSEQ: Data Toggle Sequence This field is set to indicate the PID of the current bank: DTSEQ 0 0 1 0 1 X Data Toggle Sequence Data0 Data1 Reserved For IN transfers, it indicates the data toggle sequence that will be used for the next packet to be sent. This is not relative to the current bank. For OUT transfers, this value indicates the last data toggle sequence received on the current bank. By default DTSEQ is 0b01, as if the last data toggle sequence was Data1, so the next sent or expected data toggle sequence should be Data0. * SHORTPACKET: Short Packet Interrupt This bit is set for non-control OUT endpoints, when a short packet has been received. This bit is set for non-control IN endpoints, a short packet is transmitted upon ending a DMA transfer, thus signaling an end of isochronous frame or a bulk or interrupt end of transfer, this only if the End of DMA Buffer Output Enable (DMAENDEN) bit and the Automatic Switch (AUTOSW) bit are written to one. This triggers an EPnINT interrupt if SHORTPACKETE is one. This bit is cleared when the SHORTPACKETC bit is written to one. This will acknowledge the interrupt. * STALLEDI: STALLed Interrupt This bit is set to signal that a STALL handshake has been sent. To do that, the software has to set the STALLRQ bit (by writing a 439 32059I-06/2010 AT32UC3B one to the STALLRQS bit). This triggers an EPnINT interrupt if STALLEDE is one. This bit is cleared when the STALLEDIC bit is written to one. This will acknowledge the interrupt. CRCERRI: CRC Error Interrupt This bit is set to signal that a CRC error has been detected in an isochronous OUT endpoint. The OUT packet is stored in the bank as if no CRC error had occurred. This triggers an EPnINT interrupt if CRCERRE is one. This bit is cleared when the CRCERRIC bit is written to one. This will acknowledge the interrupt. OVERFI: Overflow Interrupt This bit is set when an overflow error occurs. This triggers an EPnINT interrupt if OVERFE is one. For all endpoint types, an overflow can occur during OUT stage if the host attempts to write into a bank that is too small for the packet. The packet is acknowledged and the RXOUTI bit is set as if no overflow had occurred. The bank is filled with all the first bytes of the packet that fit in. This bit is cleared when the OVERFIC bit is written to one. This will acknowledge the interrupt. NAKINI: NAKed IN Interrupt This bit is set when a NAK handshake has been sent in response to an IN request from the host. This triggers an EPnINT interrupt if NAKINE is one. This bit is cleared when the NAKINIC bit is written to one. This will acknowledge the interrupt. NAKOUTI: NAKed OUT Interrupt This bit is set when a NAK handshake has been sent in response to an OUT request from the host. This triggers an EPnINT interrupt if NAKOUTE is one. This bit is cleared when the NAKOUTIC bit is written to one. This will acknowledge the interrupt. UNDERFI: Underflow Interrupt This bit is set, for isochronous IN/OUT endpoints, when an underflow error occurs. This triggers an EPnINT interrupt if UNDERFE is one. An underflow can occur during IN stage if the host attempts to read from an empty bank. A zero-length packet is then automatically sent by the USBB. An underflow can also occur during OUT stage if the host sends a packet while the bank is already full. Typically, the CPU is not fast enough. The packet is lost. Shall be cleared by writing a one to the UNDERFIC bit. This will acknowledge the interrupt. This bit is inactive (cleared) for bulk and interrupt IN/OUT endpoints and it means RXSTPI for control endpoints. * * * * * 440 32059I-06/2010 AT32UC3B * RXSTPI: Received SETUP Interrupt This bit is set, for control endpoints, to signal that the current bank contains a new valid SETUP packet. This triggers an EPnINT interrupt if RXSTPE is one. Shall be cleared by writing a one to the RXSTPIC bit. This will acknowledge the interrupt and free the bank. This bit is inactive (cleared) for bulk and interrupt IN/OUT endpoints and it means UNDERFI for isochronous IN/OUT endpoints. * RXOUTI: Received OUT Data Interrupt This bit is set, for control endpoints, when the current bank contains a bulk OUT packet (data or status stage). This triggers an EPnINT interrupt if RXOUTE is one. Shall be cleared for control end points, by writing a one to the RXOUTIC bit. This will acknowledge the interrupt and free the bank. This bit is set for isochronous, bulk and, interrupt OUT endpoints, at the same time as FIFOCON when the current bank is full. This triggers an EPnINT interrupt if RXOUTE is one. Shall be cleared for isochronous, bulk and, interrupt OUT endpoints, by writing a one to the RXOUTIC bit. This will acknowledge the interrupt, what has no effect on the endpoint FIFO. The user then reads from the FIFO and clears the FIFOCON bit to free the bank. If the OUT endpoint is composed of multiple banks, this also switches to the next bank. The RXOUTI and FIFOCON bits are set/cleared in accordance with the status of the next bank. RXOUTI shall always be cleared before clearing FIFOCON. This bit is inactive (cleared) for isochronous, bulk and interrupt IN endpoints. * TXINI: Transmitted IN Data Interrupt This bit is set for control endpoints, when the current bank is ready to accept a new IN packet. This triggers an EPnINT interrupt if TXINE is one. This bit is cleared when the TXINIC bit is written to one. This will acknowledge the interrupt and send the packet. This bit is set for isochronous, bulk and interrupt IN endpoints, at the same time as FIFOCON when the current bank is free. This triggers an EPnINT interrupt if TXINE is one. This bit is cleared when the TXINIC bit is written to one. This will acknowledge the interrupt, what has no effect on the endpoint FIFO. The user then writes into the FIFO and clears the FIFOCON bit to allow the USBB to send the data. If the IN endpoint is composed of multiple banks, this also switches to the next bank. The TXINI and FIFOCON bits are set/cleared in accordance with the status of the next bank. TXINI shall always be cleared before clearing FIFOCON. This bit is inactive (cleared) for isochronous, bulk and interrupt OUT endpoints. 441 32059I-06/2010 AT32UC3B 22.8.2.12 Endpoint n Status Clear Register UESTAnCLR, n in [0..6] Write-Only 0x0160 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 SHORT PACKETC 6 STALLEDIC/ CRCERRIC 5 OVERFIC 4 NAKINIC 3 NAKOUTIC 2 RXSTPIC/ UNDERFIC 1 RXOUTIC 0 TXINIC Writing a one to a bit in this register will clear the corresponding bit in UESTA. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 442 32059I-06/2010 AT32UC3B 22.8.2.13 Endpoint n Status Set Register UESTAnSET, n in [0..6] Write-Only 0x0190 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 NBUSYBKS 11 - 10 - 9 8 - 7 SHORT PACKETS 6 STALLEDIS/ CRCERRIS 5 OVERFIS 4 NAKINIS 3 NAKOUTIS 2 RXSTPIS/ UNDERFIS 1 RXOUTIS 0 TXINIS Writing a one to a bit in this register will set the corresponding bit in UESTA, what may be useful for test or debug purposes. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 443 32059I-06/2010 AT32UC3B 22.8.2.14 Endpoint n Control Register UECONn, n in [0..6] Read-Only 0x01C0 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 STALLRQ 18 RSTDT 17 - 16 EPDISHDMA 15 - 14 FIFOCON 13 KILLBK 12 NBUSYBKE 11 - 10 - 9 - 8 - 7 SHORT PACKETE 6 STALLEDE/ CRCERRE 5 OVERFE 4 NAKINE 3 NAKOUTE 2 RXSTPE/ UNDERFE 1 RXOUTE 0 TXINE * STALLRQ: STALL Request This bit is set when the STALLRQS bit is written to one. This will request to send a STALL handshake to the host. This bit is cleared when a new SETUP packet is received or when the STALLRQC bit is written to zero. * RSTDT: Reset Data Toggle This bit is set when the RSTDTS bit is written to one. This will clear the data toggle sequence, i.e., set to Data0 the data toggle sequence of the next sent (IN endpoints) or received (OUT endpoints) packet. This bit is cleared instantaneously. The user does not have to wait for this bit to be cleared. * EPDISHDMA: Endpoint Interrupts Disable HDMA Request Enable This bit is set when the EPDISHDMAS is written to one. This will pause the on-going DMA channel n transfer on any Endpoint n interrupt (EPnINT), whatever the state of the Endpoint n Interrupt Enable bit (EPnINTE). The user then has to acknowledge or to disable the interrupt source (e.g. RXOUTI) or to clear the EPDISHDMA bit (by writing a one to the EPDISHDMAC bit) in order to complete the DMA transfer. In ping-pong mode, if the interrupt is associated to a new system-bank packet (e.g. Bank1) and the current DMA transfer is running on the previous packet (Bank0), then the previous-packet DMA transfer completes normally, but the new-packet DMA transfer will not start (not requested). If the interrupt is not associated to a new system-bank packet (NAKINI, NAKOUTI, etc.), then the request cancellation may occur at any time and may immediately pause the current DMA transfer. This may be used for example to identify erroneous packets, to prevent them from being transferred into a buffer, to complete a DMA transfer by software after reception of a short packet, etc. * FIFOCON: FIFO Control For control endpoints: The FIFOCON and RWALL bits are irrelevant. The software shall therefore never use them on these endpoints. When read, their value is always 0. For IN endpoints: This bit is set when the current bank is free, at the same time as TXINI. This bit is cleared (by writing a one to the FIFOCONC bit) to send the FIFO data and to switch to the next bank. For OUT endpoints: This bit is set when the current bank is full, at the same time as RXOUTI. 444 32059I-06/2010 AT32UC3B This bit is cleared (by writing a one to the FIFOCONC bit) to free the current bank and to switch to the next bank. * KILLBK: Kill IN Bank This bit is set when the KILLBKS bit is written to one. This will kill the last written bank. This bit is cleared when the bank is killed. Caution: The bank is really cleared when the "kill packet" procedure is accepted by the USBB core. This bit is automatically cleared after the end of the procedure: The bank is really killed: NBUSYBK is decremented. The bank is not cleared but sent (IN transfer): NBUSYBK is decremented. The bank is not cleared because it was empty. The user shall wait for this bit to be cleared before trying to kill another packet. This kill request is refused if at the same time an IN token is coming and the last bank is the current one being sent on the USB line. If at least 2 banks are ready to be sent, there is no problem to kill a packet even if an IN token is coming. Indeed, in this case, the current bank is sent (IN transfer) while the last bank is killed. NBUSYBKE: Number of Busy Banks Interrupt Enable This bit is set when the NBUSYBKES bit is written to one. This will enable the Number of Busy Banks interrupt (NBUSYBK). This bit is cleared when the NBUSYBKEC bit is written to zero. This will disable the Number of Busy Banks interrupt (NBUSYBK). SHORTPACKETE: Short Packet Interrupt Enable This bit is set when the SHORTPACKETES bit is written to one. This will enable the Short Packet interrupt (SHORTPACKET). This bit is cleared when the SHORTPACKETEC bit is written to one. This will disable the Short Packet interrupt (SHORTPACKET). STALLEDE: STALLed Interrupt Enable This bit is set when the STALLEDES bit is written to one. This will enable the STALLed interrupt (STALLEDI). This bit is cleared when the STALLEDEC bit is written to one. This will disable the STALLed interrupt (STALLEDI). CRCERRE: CRC Error Interrupt Enable This bit is set when the CRCERRES bit is written to one. This will enable the CRC Error interrupt (CRCERRI). This bit is cleared when the CRCERREC bit is written to one. This will disable the CRC Error interrupt (CRCERRI). OVERFE: Overflow Interrupt Enable This bit is set when the OVERFES bit is written to one. This will enable the Overflow interrupt (OVERFI). This bit is cleared when the OVERFEC bit is written to one. This will disable the Overflow interrupt (OVERFI). NAKINE: NAKed IN Interrupt Enable This bit is set when the NAKINES bit is written to one. This will enable the NAKed IN interrupt (NAKINI). This bit is cleared when the NAKINEC bit is written to one. This will disable the NAKed IN interrupt (NAKINI). NAKOUTE: NAKed OUT Interrupt Enable This bit is set when the NAKOUTES bit is written to one. This will enable the NAKed OUT interrupt (NAKOUTI). This bit is cleared when the NAKOUTEC bit is written to one. This will disable the NAKed OUT interrupt (NAKOUTI). RXSTPE: Received SETUP Interrupt Enable This bit is set when the RXSTPES bit is written to one. This will enable the Received SETUP interrupt (RXSTPI). This bit is cleared when the RXSTPEC bit is written to one. This will disable the Received SETUP interrupt (RXSTPI). UNDERFE: Underflow Interrupt Enable This bit is set when the UNDERFES bit is written to one. This will enable the Underflow interrupt (UNDERFI). This bit is cleared when the UNDERFEC bit is written to one. This will disable the Underflow interrupt (UNDERFI). RXOUTE: Received OUT Data Interrupt Enable This bit is set when the RXOUTES bit is written to one. This will enable the Received OUT Data interrupt (RXOUT). This bit is cleared when the RXOUTEC bit is written to one. This will disable the Received OUT Data interrupt (RXOUT). TXINE: Transmitted IN Data Interrupt Enable This bit is set when the TXINES bit is written to one. This will enable the Transmitted IN Data interrupt (TXINI). This bit is cleared when the TXINEC bit is written to one. This will disable the Transmitted IN Data interrupt (TXINI). * * * * * * * * * * * 445 32059I-06/2010 AT32UC3B 22.8.2.15 Endpoint n Control Clear Register UECONnCLR, n in [0..6] Write-Only 0x0220 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 STALLRQC 18 - 17 - 16 EPDISHDMAC 15 - 14 FIFOCONC 13 - 12 NBUSYBKEC 11 - 10 - 9 - 8 - 7 SHORT PACKETEC 6 STALLEDEC/ CRCERREC 5 OVERFEC 4 NAKINEC 3 NAKOUTEC 2 RXSTPEC/ UNDERFEC 1 RXOUTEC 0 TXINEC Writing a one to a bit in this register will clear the corresponding bit in UECONn. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 446 32059I-06/2010 AT32UC3B 22.8.2.16 Endpoint n Control Set Register UECONnSET, n in [0..6] Write-Only 0x01F0 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 STALLRQS 18 RSTDTS 17 - 16 EPDISHDMAS 15 - 14 - 13 KILLBKS 12 NBUSYBKES 11 - 10 - 9 - 8 - 7 SHORT PACKETES 6 STALLEDES/ CRCERRES 5 OVERFES 4 NAKINES 3 NAKOUTES 2 RXSTPES/ UNDERFES 1 RXOUTES 0 TXINES Writing a one to a bit in this register will set the corresponding bit in UECONn. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 447 32059I-06/2010 AT32UC3B 22.8.2.17 Device DMA Channel n Next Descriptor Address Register UDDMAnNEXTDESC, n in [1..6] Read/Write 0x0310 + (n - 1) * 0x10 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 27 26 25 24 NXTDESCADDR[31:24] 23 22 21 20 19 18 17 16 NXTDESCADDR[23:16] 15 14 13 12 11 10 9 8 NXTDESCADDR[15:8] 7 6 5 4 3 - 2 - 1 - 0 - NXTDESCADDR[7:4] * NXTDESCADDR: Next Descriptor Address This field contains the bits 31:4 of the 16-byte aligned address of the next channel descriptor to be processed. This field is written either or by descriptor loading. 448 32059I-06/2010 AT32UC3B 22.8.2.18 Device DMA Channel n HSB Address Register UDDMAnADDR, n in [1..6] Read/Write 0x0314 + (n - 1) * 0x10 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 27 26 25 24 HSBADDR[31:24] 23 22 21 20 19 18 17 16 HSBADDR[23:16] 15 14 13 12 11 10 9 8 HSBADDR[15:8] 7 6 5 4 3 2 1 0 HSBADDR[7:0] * HSBADDR: HSB Address This field determines the HSB bus current address of a channel transfer. The address written to the HSB address bus is HSBADDR rounded down to the nearest word-aligned address, i.e., HSBADDR[1:0] is considered as 0b00 since only word accesses are performed. Channel HSB start and end addresses may be aligned on any byte boundary. The user may write this field only when the Channel Enabled bit (CHEN) of the UDDMAnSTATUS register is cleared. This field is updated at the end of the address phase of the current access to the HSB bus. It is incremented of the HSB access byte-width. The HSB access width is 4 bytes, or less at packet start or end if the start or end address is not aligned on a word boundary. The packet start address is either the channel start address or the next channel address to be accessed in the channel buffer. The packet end address is either the channel end address or the latest channel address accessed in the channel buffer. The channel start address is written or loaded from the descriptor, whereas the channel end address is either determined by the end of buffer or the end of USB transfer if the Buffer Close Input Enable bit (BUFFCLOSEINEN) is set. 449 32059I-06/2010 AT32UC3B 22.8.2.19 Device DMA Channel n Control Register UDDMAnCONTROL, n in [1..6] Read/Write 0x0318 + (n - 1) * 0x10 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 27 26 25 24 CHBYTELENGTH[15:8] 23 22 21 20 19 18 17 16 CHBYTELENGTH[7:0] 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 BURSTLOCKEN 6 DESCLDIRQEN 5 EOBUFFIRQEN 4 EOTIRQEN 3 DMAENDEN 2 BUFFCLOSE INEN 1 LDNXTCH DESCEN 0 CHEN * CHBYTELENGTH: Channel Byte Length This field determines the total number of bytes to be transferred for this buffer. The maximum channel transfer size 64kB is reached when this field is zero (default value). If the transfer size is unknown, the transfer end is controlled by the peripheral and this field should be written to zero. This field can be written or descriptor loading only after the UDDMAnSTATUS.CHEN bit has been cleared, otherwise this field is ignored. BURSTLOCKEN: Burst Lock Enable 1: The USB data burst is locked for maximum optimization of HSB busses bandwidth usage and maximization of fly-by duration. 0: The DMA never locks the HSB access. DESCLDIRQEN: Descriptor Loaded Interrupt Enable 1: The Descriptor Loaded interrupt is enabled.This interrupt is generated when a Descriptor has been loaded from the system bus. 0: The Descriptor Loaded interrupt is disabled. EOBUFFIRQEN: End of Buffer Interrupt Enable 1: The end of buffer interrupt is enabled.This interrupt is generated when the channel byte count reaches zero. 0: The end of buffer interrupt is disabled. EOTIRQEN: End of USB Transfer Interrupt Enable 1: The end of usb OUT data transfer interrupt is enabled. This interrupt is generated only if the BUFFCLOSEINEN bit is set. 0: The end of usb OUT data transfer interrupt is disabled. DMAENDEN: End of DMA Buffer Output Enable Writing a one to this bit will properly complete the usb transfer at the end of the dma transfer. For IN endpoint, it means that a short packet (or a Zero Length Packet) will be sent to the USB line to properly closed the usb transfer at the end of the dma transfer. For OUT endpoint, it means that all the banks will be properly released. (NBUSYBK=0) at the end of the dma transfer. * * * * * 450 32059I-06/2010 AT32UC3B * BUFFCLOSEINEN: Buffer Close Input Enable For Bulk and Interrupt endpoint, writing a one to this bit will automatically close the current DMA transfer at the end of the USB OUT data transfer (received short packet). For Full-speed Isochronous, it does not make sense, so BUFFCLOSEINEN should be left to zero. Writing a zero to this bit to disable this feature. * LDNXTCHDESCEN: Load Next Channel Descriptor Enable 1: the channel controller loads the next descriptor after the end of the current transfer, i.e. when the UDDMAnSTATUS.CHEN bit is reset. 0: no channel register is loaded after the end of the channel transfer. If the CHEN bit is written to zero, the next descriptor is immediately loaded upon transfer request (endpoint is free for IN endpoint, or endpoint is full for OUT endpoint). Table 22-6. LDNXTCHDES CEN 0 0 1 1 DMA Channel Control Command Summary CHEN 0 1 0 1 Current Bank stop now Run and stop at end of buffer Load next descriptor now Run and link at end of buffer * CHEN: Channel Enable Writing this bit to zero will disabled the DMA channel and no transfer will occur upon request. If the LDNXTCHDESCEN bit is written to zero, the channel is frozen and the channel registers may then be read and/or written reliably as soon as both UDDMAnSTATUS.CHEN and CHACTIVE bits are zero. Writing this bit to one will set the UDDMAnSTATUS.CHEN bit and enable DMA channel data transfer. Then any pending request will start the transfer. This may be used to start or resume any requested transfer. This bit is cleared when the channel source bus is disabled at end of buffer. If the LDNXTCHDESCEN bit has been cleared by descriptor loading, the user will have to write to one the corresponding CHEN bit to start the described transfer, if needed. If a channel request is currently serviced when this bit is zero, the DMA FIFO buffer is drained until it is empty, then the UDDMAnSTATUS.CHEN bit is cleared. If the LDNXTCHDESCEN bit is set or after this bit clearing, then the currently loaded descriptor is skipped (no data transfer occurs) and the next descriptor is immediately loaded. 451 32059I-06/2010 AT32UC3B 22.8.2.20 Device DMA Channel n Status Register UDDMAnSTATUS, n in [1..6] Read/Write 0x031C + (n - 1) * 0x10 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 27 26 25 24 CHBYTECNT[15:8] 23 22 21 20 19 18 17 16 CHBYTECNT[7:0] 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 DESCLD STA 5 EOCHBUFF STA 4 EOTSTA 3 - 2 - 1 CHACTIVE 0 CHEN * CHBYTECNT: Channel Byte Count This field contains the current number of bytes still to be transferred for this buffer. This field is decremented at each dma access. This field is reliable (stable) only if the CHEN bit is zero. DESCLDSTA: Descriptor Loaded Status This bit is set when a Descriptor has been loaded from the HSB bus. This bit is cleared when read by the user. EOCHBUFFSTA: End of Channel Buffer Status This bit is set when the Channel Byte Count counts down to zero. This bit is automatically cleared when read by software. EOTSTA: End of USB Transfer Status This bit is set when the completion of the usb data transfer has closed the dma transfer. It is valid only if UDDMAnCONTROL.BUFFCLOSEINEN is one. This bit is automatically cleared when read by software. CHACTIVE: Channel Active 0: the DMA channel is no longer trying to source the packet data. 1: the DMA channel is currently trying to source packet data, i.e. selected as the highest-priority requesting channel. When a packet transfer cannot be completed due to an EOCHBUFFSTA, this bit stays set during the next channel descriptor load (if any) and potentially until USB packet transfer completion, if allowed by the new descriptor. When programming a DMA by descriptor (Load next descriptor now), the CHACTIVE bit is set only once the DMA is running (the endpoint is free for IN transaction, the endpoint is full for OUT transaction). CHEN: Channel Enabled This bit is set (after one cycle latency) when the L.CHEN is written to one or when the descriptor is loaded. This bit is cleared when any transfer is ended either due to an elapsed byte count or a USB device initiated transfer end. 0: the DMA channel no longer transfers data, and may load the next descriptor if the UDDMAnCONTROL.LDNXTCHDESCEN bit is zero. 1: the DMA channel is currently enabled and transfers data upon request. If a channel request is currently serviced when the UDDMAnCONTROL.CHEN bit is written to zero, the DMA FIFO buffer is drained until it is empty, then this status bit is cleared. * * * * * 452 32059I-06/2010 AT32UC3B 22.8.3 22.8.3.1 USB Host Registers Host General Control Register UHCON Read/Write 0x0400 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 RESUME 9 RESET 8 SOFE 7 - 6 - 5 - 4 - 3 - 2 - 1 - 0 - * RESUME: Send USB Resume Writing a one to this bit will generate a USB Resume on the USB bus. This bit is cleared when the USB Resume has been sent or when a USB reset is requested. Writing a zero to this bit has no effect. This bit should be written to one only when the start of frame generation is enable. (SOFE bit is one). * RESET: Send USB Reset Writing a one to this bit will generate a USB Reset on the USB bus. This bit is cleared when the USB Reset has been sent. It may be useful to write a zero to this bit when a device disconnection is detected (UHINT.DDISCI is one) whereas a USB Reset is being sent. * SOFE: Start of Frame Generation Enable Writing a one to this bit will generate SOF on the USB bus in full speed mode and keep alive in low speed mode. Writing a zero to this bit will disable the SOF generation and to leave the USB bus in idle state. This bit is set when a USB reset is requested or an upstream resume interrupt is detected (UHINT.TXRSMI). 453 32059I-06/2010 AT32UC3B 22.8.3.2 Host Global Interrupt Register UHINT Read-Only 0x0404 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 DMA6INT 29 DMA5INT 28 DMA4INT 27 DMA3INT 26 DMA2INT 25 DMA1INT 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 15 - 14 P6INT 13 P5INT 12 P4INT 11 P3INT 10 P2INT 9 P1INT 8 P0INT 7 - 6 HWUPI 5 HSOFI 4 RXRSMI 3 RSMEDI 2 RSTI 1 DDISCI 0 DCONNI * DMAnINT: DMA Channel n Interrupt This bit is set when an interrupt is triggered by the DMA channel n. This triggers a USB interrupt if the corresponding DMAnINTE is one (UHINTE register). This bit is cleared when the UHDMAnSTATUS interrupt source is cleared. PnINT: Pipe n Interrupt This bit is set when an interrupt is triggered by the endpoint n (UPSTAn). This triggers a USB interrupt if the corresponding pipe interrupt enable bit is one (UHINTE register). This bit is cleared when the interrupt source is served. HWUPI: Host Wake-Up Interrupt This bit is set when the host controller is in the suspend mode (SOFE is zero) and an upstream resume from the peripheral is detected. This bit is set when the host controller is in the suspend mode (SOFE is zero) and a peripheral disconnection is detected. This bit is set when the host controller is in the Idle state (USBSTA.VBUSRQ is zero, no VBus is generated). This interrupt is generated even if the clock is frozen by the FRZCLK bit. HSOFI: Host Start of Frame Interrupt This bit is set when a SOF is issued by the Host controller. This triggers a USB interrupt when HSOFE is one. When using the host controller in low speed mode, this bit is also set when a keep-alive is sent. This bit is cleared when the HSOFIC bit is written to one. RXRSMI: Upstream Resume Received Interrupt This bit is set when an Upstream Resume has been received from the Device. This bit is cleared when the RXRSMIC is written to one. RSMEDI: Downstream Resume Sent Interrupt This bit set when a Downstream Resume has been sent to the Device. This bit is cleared when the RSMEDIC bit is written to one. RSTI: USB Reset Sent Interrupt This bit is set when a USB Reset has been sent to the device. This bit is cleared when the RSTIC bit is written to one. DDISCI: Device Disconnection Interrupt This bit is set when the device has been removed from the USB bus. This bit is cleared when the DDISCIC bit is written to one. * * * * * * * 454 32059I-06/2010 AT32UC3B * DCONNI: Device Connection Interrupt This bit is set when a new device has been connected to the USB bus. This bit is cleared when the DCONNIC bit is written to one. 455 32059I-06/2010 AT32UC3B 22.8.3.3 Host Global Interrupt Clear Register UHINTCLR Write-Only 0x0408 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 HWUPIC 5 HSOFIC 4 RXRSMIC 3 RSMEDIC 2 RSTIC 1 DDISCIC 0 DCONNIC Writing a one to a bit in this register will clear the corresponding bit in UHINT. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 456 32059I-06/2010 AT32UC3B 22.8.3.4 Host Global Interrupt Set Register UHINTSET Write-Only 0x040C 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 DMA6INTS 29 DMA5INTS 28 DMA4INTS 27 DMA3INTS 26 DMA2INTS 25 DMA1INTS 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 HWUPIS 5 HSOFIS 4 RXRSMIS 3 RSMEDIS 2 RSTIS 1 DDISCIS 0 DCONNIS Writing a one to a bit in this register will set the corresponding bit in UHINT, what may be useful for test or debug purposes. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 457 32059I-06/2010 AT32UC3B 22.8.3.5 Host Global Interrupt Enable Register UHINTE Read-Only 0x0410 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 DMA6INTE 29 DMA5INTE 28 DMA4INTE 27 DMA3INTE 26 DMA2INTE 25 DMA1INTE 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 P6INTE 13 P5INTE 12 P4INTE 11 P3INTE 10 P2INTE 9 P1INTE 8 P0INTE 7 - 6 HWUPIE 5 HSOFIE 4 RXRSMIE 3 RSMEDIE 2 RSTIE 1 DDISCIE 0 DCONNIE * DMAnINTE: DMA Channel n Interrupt Enable * * * * * * * * This bit is set when the DMAnINTES bit is written to one. This will enable the DMA Channel n Interrupt (DMAnINT). This bit is cleared when the DMAnINTEC bit is written to one. This will disable the DMA Channel n Interrupt (DMAnINT). PnINTE: Pipe n Interrupt Enable This bit is set when the PnINTES bit is written to one. This will enable the Pipe n Interrupt (PnINT). This bit is cleared when the PnINTEC bit is written to one. This will disable the Pipe n Interrupt (PnINT). HWUPIE: Host Wake-Up Interrupt Enable This bit is set when the HWUPIES bit is written to one. This will enable the Host Wake-up Interrupt (HWUPI). This bit is cleared when the HWUPIEC bit is written to one. This will disable the Host Wake-up Interrupt (HWUPI). HSOFIE: Host Start of Frame Interrupt Enable This bit is set when the HSOFIES bit is written to one. This will enable the Host Start of Frame interrupt (HSOFI). This bit is cleared when the HSOFIEC bit is written to one. This will disable the Host Start of Frame interrupt (HSOFI). RXRSMIE: Upstream Resume Received Interrupt Enable This bit is set when the RXRSMIES bit is written to one. This will enable the Upstream Resume Received interrupt (RXRSMI). This bit is cleared when the RXRSMIEC bit is written to one. This will disable the Downstream Resume interrupt (RXRSMI). RSMEDIE: Downstream Resume Sent Interrupt Enable This bit is set when the RSMEDIES bit is written to one. This will enable the Downstream Resume interrupt (RSMEDI). This bit is cleared when the RSMEDIEC bit is written to one. This will disable the Downstream Resume interrupt (RSMEDI). RSTIE: USB Reset Sent Interrupt Enable This bit is set when the RSTIES bit is written to one. This will enable the USB Reset Sent interrupt (RSTI). This bit is cleared when the RSTIEC bit is written to one. This will disable the USB Reset Sent interrupt (RSTI). DDISCIE: Device Disconnection Interrupt Enable This bit is set when the DDISCIES bit is written to one. This will enable the Device Disconnection interrupt (DDISCI). This bit is cleared when the DDISCIEC bit is written to one. This will disable the Device Disconnection interrupt (DDISCI). DCONNIE: Device Connection Interrupt Enable This bit is set when the DCONNIES bit is written to one. This will enable the Device Connection interrupt (DCONNI). This bit is cleared when the DCONNIEC bit is written to one. This will disable the Device Connection interrupt (DCONNI). 458 32059I-06/2010 AT32UC3B 22.8.3.6 Host Global Interrupt Enable Clear Register UHINTECLR Write-Only 0x0414 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 DMA6INTEC 29 DMA5INTEC 28 DMA4INTEC 27 DMA3INTEC 26 DMA2INTEC 25 DMA1INTEC 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 P6INTEC 13 P5INTEC 12 P4INTEC 11 P3INTEC 10 P2INTEC 9 P1INTEC 8 P0INTEC 7 - 6 HWUPIEC 5 HSOFIEC 4 RXRSMIEC 3 RSMEDIEC 2 RSTIEC 1 DDISCIEC 0 DCONNIEC Writing a one to a bit in this register will clear the corresponding bit in UHINTE. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 459 32059I-06/2010 AT32UC3B 22.8.3.7 Host Global Interrupt Enable Set Register UHINTESET Write-Only 0x0418 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 DMA6INTES 29 DMA5INTES 28 DMA4INTES 27 DMA3INTES 26 DMA2INTES 25 DMA1INTES 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 P6INTES 13 P5INTES 12 P4INTES 11 P3INTES 10 P2INTES 9 P1INTES 8 P0INTES 7 - 6 HWUPIES 5 HSOFIES 4 RXRSMIES 3 RSMEDIES 2 RSTIES 1 DDISCIES 0 DCONNIES Writing a one to a bit in this register will set the corresponding bit in UHINT. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 460 32059I-06/2010 AT32UC3B 22.8.3.8 Host Frame Number Register UHFNUM Read/Write 0x0420 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 22 21 20 FLENHIGH 19 18 17 16 15 - 14 - 13 12 11 FNUM[10:5] 10 9 8 7 6 5 FNUM[4:0] 4 3 2 - 1 - 0 - * FLENHIGH: Frame Length This field contains the 8 high-order bits of the 14-bits internal frame counter (frame counter at 12MHz, counter length is 12000 to ensure a SOF generation every 1 ms). * FNUM: Frame Number This field contains the current SOF number. This field can be written. 461 32059I-06/2010 AT32UC3B 22.8.3.9 USB Host Frame Number Register (UHADDR1) UHADDR1 Read/Write 0x0424 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 29 28 27 UHADDRP3 26 25 24 23 - 22 21 20 19 UHADDRP2 18 17 16 15 - 14 13 12 11 UHADDRP1 10 9 8 7 - 6 5 4 3 UHADDRP0 2 1 0 * UHADDRP3: USB Host Address This field contains the address of the Pipe3 of the USB Device. This field is cleared when a USB reset is requested. * UHADDRP2: USB Host Address This field contains the address of the Pipe2 of the USB Device. This field is cleared when a USB reset is requested. * UHADDRP1: USB Host Address This field contains the address of the Pipe1 of the USB Device. This field is cleared when a USB reset is requested. * UHADDRP0: USB Host Address This field contains the address of the Pipe0 of the USB Device. This field is cleared when a USB reset is requested. 462 32059I-06/2010 AT32UC3B 22.8.3.10 Host Frame Number Register UHADDR2 Read/Write 0x0428 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 21 20 19 UHADDRP6 18 17 16 15 - 14 13 12 11 UHADDRP5 10 9 8 7 - 6 5 4 3 UHADDRP4 2 1 0 * UHADDRP6: USB Host Address This field contains the address of the Pipe6 of the USB Device. This field is cleared when a USB reset is requested. * UHADDRP5: USB Host Address This field contains the address of the Pipe5 of the USB Device. This field is cleared when a USB reset is requested. * UHADDRP4: USB Host Address This field contains the address of the Pipe4 of the USB Device. This field is cleared when a USB reset is requested. 463 32059I-06/2010 AT32UC3B 22.8.3.11 Pipe Enable/Reset Register UPRST Read/Write 0x0041C 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 PRST6 21 PRST5 20 PRST4 19 PRST3 18 PRST2 17 PRST1 16 PRST0 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 PEN6 5 PEN5 4 PEN4 3 PEN3 2 PEN2 1 PEN1 0 PEN0 * PRSTn: Pipe n Reset Writing a one to this bit will reset the Pipe n FIFO. This resets the endpoint n registers (UPCFGn, UPSTAn, UPCONn) but not the endpoint configuration (ALLOC, PBK, PSIZE, PTOKEN, PTYPE, PEPNUM, INTFRQ). All the endpoint mechanism (FIFO counter, reception, transmission, etc.) is reset apart from the Data Toggle management. The endpoint configuration remains active and the endpoint is still enabled. Writing a zero to this bit will complete the reset operation and allow to start using the FIFO. * PENn: Pipe n Enable Writing a one to this bit will enable the Pipe n. Writing a zero to this bit will disable the Pipe n, what forces the Pipe n state to inactive and resets the pipe n registers (UPCFGn, UPSTAn, UPCONn) but not the pipe configuration (ALLOC, PBK, PSIZE). 464 32059I-06/2010 AT32UC3B 22.8.3.12 Pipe n Configuration Register UPCFGn, n in [0..6] Read/Write 0x0500 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 INTFRQ 27 26 25 24 23 - 22 - 21 - 20 - 19 18 PEPNUM 17 16 15 - 14 - 13 PTYPE 12 11 - 10 AUTOSW 9 PTOKEN 8 7 - 6 5 PSIZE 4 3 PBK 2 1 ALLOC 0 - * INTFRQ: Pipe Interrupt Request Frequency This field contains the maximum value in millisecond of the polling period for an Interrupt Pipe. This value has no effect for a non-Interrupt Pipe. This field is cleared upon sending a USB reset. * PEPNUM: Pipe Endpoint Number This field contains the number of the endpoint targeted by the pipe. This value is from 0 to 15. This field is cleared upon sending a USB reset. * PTYPE: Pipe Type This field contains the pipe type. PTYPE 0 0 1 1 0 1 0 1 Pipe Type Control Isochronous Bulk Interrupt This field is cleared upon sending a USB reset. * AUTOSW: Automatic Switch This bit is cleared upon sending a USB reset. 1: The automatic bank switching is enabled. 0: The automatic bank switching is disabled. 465 32059I-06/2010 AT32UC3B * PTOKEN: Pipe Token This field contains the endpoint token. PTOKEN 00 01 10 11 Endpoint Direction SETUP IN OUT reserved * PSIZE: Pipe Size This field contains the size of each pipe bank. PSIZE 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 Endpoint Size 8 bytes 16 bytes 32 bytes 64 bytes 128 bytes 256 bytes 512 bytes 1024 bytes This field is cleared upon sending a USB reset. * PBK: Pipe Banks This field contains the number of banks for the pipe. PBK 0 0 1 1 0 1 0 1 Endpoint Banks 1 (single-bank pipe) 2 (double-bank pipe) 3 (triple-bank pipe) Reserved For control endpoints, a single-bank pipe (0b00) should be selected. This field is cleared upon sending a USB reset. * ALLOC: Pipe Memory Allocate Writing a one to this bit will allocate the pipe memory. Writing a zero to this bit will free the pipe memory. This bit is cleared when a USB Reset is requested. Refer to the DPRAM Management chapter for more details. 466 32059I-06/2010 AT32UC3B 22.8.3.13 Pipe n Status Register UPSTAn, n in [0..6] Read-Only 0x0530 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 29 28 27 PBYCT[10:4] 26 25 24 23 22 PBYCT[3:0] 21 20 19 - 18 CFGOK 17 - 16 RWALL 15 CURRBK 14 13 NBUSYBK 12 11 - 10 - 9 DTSEQ 8 7 SHORT PACKETI 6 RXSTALLDI/ CRCERRI 5 OVERFI 4 NAKEDI 3 PERRI 2 TXSTPI/ UNDERFI 1 TXOUTI 0 RXINI * PBYCT: Pipe Byte Count This field contains the byte count of the FIFO. For OUT pipe, incremented after each byte written by the user into the pipe and decremented after each byte sent to the peripheral. For IN pipe, incremented after each byte received from the peripheral and decremented after each byte read by the user from the pipe. This field may be updated 1 clock cycle after the RWALL bit changes, so the user should not poll this field as an interrupt bit. * CFGOK: Configuration OK Status This bit is set/cleared when the UPCFGn.ALLOC bit is set. This bit is set if the pipe n number of banks (UPCFGn.PBK) and size (UPCFGn.PSIZE) are correct compared to the maximal allowed number of banks and size for this pipe and to the maximal FIFO size (i.e., the DPRAM size). If this bit is cleared, the user should rewrite correct values ot the PBK and PSIZE field in the UPCFGn register. * RWALL: Read/Write Allowed For OUT pipe, this bit is set when the current bank is not full, i.e., the software can write further data into the FIFO. For IN pipe, this bit is set when the current bank is not empty, i.e., the software can read further data from the FIFO. This bit is cleared otherwise. This bit is also cleared when the RXSTALL or the PERR bit is one. * CURRBK: Current Bank For non-control pipe, this field indicates the number of the current bank. CURRBK 0 0 1 1 0 1 0 1 Current Bank Bank0 Bank1 Bank2 Reserved 467 32059I-06/2010 AT32UC3B This field may be updated 1 clock cycle after the RWALL bit changes, so the user shall not poll this field as an interrupt bit. * NBUSYBK: Number of Busy Banks This field indicates the number of busy bank. For OUT pipe, this field indicates the number of busy bank(s), filled by the user, ready for OUT transfer. When all banks are busy, this triggers an PnINT interrupt if UPCONn.NBUSYBKE is one. For IN pipe, this field indicates the number of busy bank(s) filled by IN transaction from the Device. When all banks are free, this triggers an PnINT interrupt if UPCONn.NBUSYBKE is one. NBUSYBK 0 0 1 1 0 1 0 1 Number of busy bank All banks are free. 1 busy bank 2 busy banks reserved * DTSEQ: Data Toggle Sequence This field indicates the data PID of the current bank. DTSEQ 0 0 1 1 0 1 0 1 Data toggle sequence Data0 Data1 reserved reserved * * * * * * * For OUT pipe, this field indicates the data toggle of the next packet that will be sent. For IN pipe, this field indicates the data toggle of the received packet stored in the current bank. SHORTPACKETI: Short Packet Interrupt This bit is set when a short packet is received by the host controller (packet length inferior to the PSIZE programmed field). This bit is cleared when the SHORTPACKETIC bit is written to one. RXSTALLDI: Received STALLed Interrupt This bit is set, for all endpoints but isochronous, when a STALL handshake has been received on the current bank of the pipe. The Pipe is automatically frozen. This triggers an interrupt if the RXSTALLE bit is one. This bit is cleared when the RXSTALLDIC bit is written to one. CRCERRI: CRC Error Interrupt This bit is set, for isochronous endpoint, when a CRC error occurs on the current bank of the Pipe. This triggers an interrupt if the TXSTPE bit is one. This bit is cleared when the CRCERRIC bit is written to one. OVERFI: Overflow Interrupt This bit is set when the current pipe has received more data than the maximum length of the current pipe. An interrupt is triggered if the OVERFIE bit is one. This bit is cleared when the OVERFIC bit is written to one. NAKEDI: NAKed Interrupt This bit is set when a NAK has been received on the current bank of the pipe. This triggers an interrupt if the NAKEDE bit is one. This bit is cleared when the NAKEDIC bit written to one. PERRI: Pipe Error Interrupt This bit is set when an error occurs on the current bank of the pipe. This triggers an interrupt if the PERRE bit is set. Refers to the UPERRn register to determine the source of the error. This bit is cleared when the error source bit is cleared. TXSTPI: Transmitted SETUP Interrupt This bit is set, for Control endpoints, when the current SETUP bank is free and can be filled. This triggers an interrupt if the TXSTPE bit is one. This bit is cleared when the TXSTPIC bit is written to one. 468 32059I-06/2010 AT32UC3B * UNDERFI: Underflow Interrupt This bit is set, for isochronous and Interrupt IN/OUT pipe, when an error flow occurs. This triggers an interrupt if the UNDERFIE bit is one. This bit is set, for Isochronous or interrupt OUT pipe, when a transaction underflow occurs in the current pipe. (the pipe can't send the OUT data packet in time because the current bank is not ready). A zero-length-packet (ZLP) will be sent instead of. This bit is set, for Isochronous or interrupt IN pipe, when a transaction flow error occurs in the current pipe. i.e, the current bank of the pipe is not free whereas a new IN USB packet is received. This packet is not stored in the bank. For Interrupt pipe, the overflowed packet is ACKed to respect the USB standard. This bit is cleared when the UNDERFIEC bit is written to one. * TXOUTI: Transmitted OUT Data Interrupt This bit is set when the current OUT bank is free and can be filled. This triggers an interrupt if the TXOUTE bit is one. This bit is cleared when the TXOUTIC bit is written to one. * RXINI: Received IN Data Interrupt This bit is set when a new USB message is stored in the current bank of the pipe. This triggers an interrupt if the RXINE bit is one. This bit is cleared when the RXINIC bit is written to one. 469 32059I-06/2010 AT32UC3B 22.8.3.14 Pipe n Status Clear Register UPSTAnCLR, n in [0..6] Write-Only 0x0560 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 SHORT PACKETIC 6 RXSTALLDI C/ CRCERRIC 5 OVERFIC 4 NAKEDIC 3 - 2 TXSTPIC/ UNDERFIC 1 TXOUTIC 0 RXINIC Writing a one to a bit in this register will clear the corresponding bit in UPSTAn. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 470 32059I-06/2010 AT32UC3B 22.8.3.15 Pipe n Status Set Register UPSTAnSET, n in [0..6] Write-Only 0x0590 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 NBUSYBKS 11 - 10 - 9 - 8 - 7 SHORT PACKETIS 6 RXSTALLDIS/ 5 OVERFIS 4 NAKEDIS 3 PERRIS 2 TXSTPIS/ UNDERFIS 1 TXOUTIS 0 RXINIS CRCERRIS Writing a one to a bit in this register will set the corresponding bit in UPSTAn, what may be useful for test or debug purposes. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 471 32059I-06/2010 AT32UC3B 22.8.3.16 Pipe n Control Register UPCONn, n in [0..6] Read-Only 0x05C0 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 RSTDT 17 PFREEZE 16 PDISHDMA 15 - 14 FIFOCON 13 - 12 NBUSYBKE 11 - 10 - 9 - 8 - 7 SHORT PACKETIE 6 RXSTALLDE/ CRCERRE 5 OVERFIE 4 NAKEDE 3 PERRE 2 TXSTPE/ UNDERFIE 1 TXOUTE 0 RXINE * RSTDT: Reset Data Toggle * This bit is set when the RSTDTS bit is written to one. This will reset the Data Toggle to its initial value for the current Pipe. This bit is cleared when proceed. PFREEZE: Pipe Freeze This bit is set when the PFREEZES bit is written to one or when the pipe is not configured or when a STALL handshake has been received on this Pipe or when an error occurs on the Pipe (PERR is one) or when (INRQ+1) In requests have been processed or when after a Pipe reset (UPRST.PRSTn rising) or a Pipe Enable (UPRST.PEN rising). This will Freeze the Pipe requests generation. This bit is cleared when the PFREEZEC bit is written to one. This will enable the Pipe request generation. PDISHDMA: Pipe Interrupts Disable HDMA Request Enable See the UECONn.EPDISHDMA bit description. FIFOCON: FIFO Control For OUT and SETUP Pipe: This bit is set when the current bank is free, at the same time than TXOUTI or TXSTPI. This bit is cleared when the FIFOCONC bit is written to one. This will send the FIFO data and switch the bank. For IN Pipe: This bit is set when a new IN message is stored in the current bank, at the same time than RXINI. This bit is cleared when the FIFOCONC bit is written to one. This will free the current bank and switch to the next bank. NBUSYBKE: Number of Busy Banks Interrupt Enable This bit is set when the NBUSYBKES bit is written to one.This will enable the Transmitted IN Data interrupt (NBUSYBKE). This bit is cleared when the NBUSYBKEC bit is written to one. This will disable the Transmitted IN Data interrupt (NBUSYBKE). SHORTPACKETIE: Short Packet Interrupt Enable This bit is set when the SHORTPACKETES bit is written to one. This will enable the Transmitted IN Data IT (SHORTPACKETIE). This bit is cleared when the SHORTPACKETEC bit is written to one. This will disable the Transmitted IN Data IT (SHORTPACKETE). * * * * 472 32059I-06/2010 AT32UC3B * RXSTALLDE: Received STALLed Interrupt Enable This bit is set when the RXSTALLDES bit is written to one. This will enable the Transmitted IN Data interrupt (RXSTALLDE). This bit is cleared when the RXSTALLDEC bit is written to one. This will disable the Transmitted IN Data interrupt (RXSTALLDE). CRCERRE: CRC Error Interrupt Enable This bit is set when the CRCERRES bit is written to one. This will enable the Transmitted IN Data interrupt (CRCERRE). This bit is cleared when the CRCERREC bit is written to one. This will disable the Transmitted IN Data interrupt (CRCERRE). OVERFIE: Overflow Interrupt Enable This bit is set when the OVERFIES bit is written to one. This will enable the Transmitted IN Data interrupt (OVERFIE). This bit is cleared when the OVERFIEC bit is written to one. This will disable the Transmitted IN Data interrupt (OVERFIE). NAKEDE: NAKed Interrupt Enable This bit is set when the NAKEDES bit is written to one. This will enable the Transmitted IN Data interrupt (NAKEDE). This bit is cleared when the NAKEDEC bit is written to one. This will disable the Transmitted IN Data interrupt (NAKEDE). PERRE: Pipe Error Interrupt Enable This bit is set when the PERRES bit is written to one. This will enable the Transmitted IN Data interrupt (PERRE). This bit is cleared when the PERREC bit is written to one. This will disable the Transmitted IN Data interrupt (PERRE). TXSTPE: Transmitted SETUP Interrupt Enable This bit is set when the TXSTPES bit is written to one. This will enable the Transmitted IN Data interrupt (TXSTPE). This bit is cleared when the TXSTPEC bit is written to one. This will disable the Transmitted IN Data interrupt (TXSTPE). UNDERFIE: Underflow Interrupt Enable This bit is set when the UNDERFIES bit is written to one. This will enable the Transmitted IN Data interrupt (UNDERFIE). This bit is cleared when the UNDERFIEC bit is written to one. This will disable the Transmitted IN Data interrupt (UNDERFIE). TXOUTE: Transmitted OUT Data Interrupt Enable This bit is set when the TXOUTES bit is written to one. This will enable the Transmitted IN Data interrupt (TXOUTE). This bit is cleared when the TXOUTEC bit is written to one. This will disable the Transmitted IN Data interrupt (TXOUTE). RXINE: Received IN Data Interrupt Enable This bit is set when the RXINES bit is written to one. This will enable the Transmitted IN Data interrupt (RXINE). This bit is cleared when the RXINEC bit is written to one. This will disable the Transmitted IN Data interrupt (RXINE). * * * * * * * * 473 32059I-06/2010 AT32UC3B 22.8.3.17 Pipe n Control Clear Register UPCONnCLR, n in [0..6] Write-Only 0x0620 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 PFREEZEC 16 PDISHDMAC 15 - 14 FIFOCONC 13 - 12 NBUSYBKEC 11 - 10 - 9 - 8 - 7 SHORT PACKETIEC 6 RXSTALLDEC/ 5 OVERFIEC 4 NAKEDEC 3 PERREC 2 TXSTPEC/ UNDERFIEC 1 TXOUTEC 0 RXINEC CRCERREC Writing a one to a bit in this register will clear the corresponding bit in UPCONn. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 474 32059I-06/2010 AT32UC3B 22.8.3.18 Pipe n Control Set Register UPCONnSET, n in [0..6] Write-Only 0x05F0 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Read Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 RSTDTS 17 PFREEZES 16 PDISHDMAS 15 - 14 - 13 - 12 NBUSYBKES 11 - 10 - 9 - 8 - 7 SHORT PACKETIES 6 RXSTALLDES/ 5 OVERFIES 4 NAKEDES 3 PERRES 2 TXSTPES/ UNDERFIES 1 TXOUTES 0 RXINES CRCERRES Writing a one to a bit in this register will set the corresponding bit in UPCONn. Writing a zero to a bit in this register has no effect. This bit always reads as zero. 475 32059I-06/2010 AT32UC3B 22.8.3.19 Pipe n IN Request Register UPINRQn, n in [0..6] Read/Write 0x0650 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 INMODE 7 6 5 4 INRQ 3 2 1 0 * INMODE: IN Request Mode Writing a one to this bit will allow the USBB to perform infinite IN requests when the Pipe is not frozen. Writing a zero to this bit will perform a pre-defined number of IN requests. This number is the INRQ field. * INRQ: IN Request Number before Freeze This field contains the number of IN transactions before the USBB freezes the pipe. The USBB will perform (INRQ+1) IN requests before to freeze the Pipe. This counter is automatically decreased by 1 each time a IN request has been successfully performed. This register has no effect when the INMODE bit is one (infinite IN requests generation till the pipe is not frozen). 476 32059I-06/2010 AT32UC3B 22.8.3.20 Pipe n Error Register UPERRn, n in [0..6] Read/Write 0x0680 + (n * 0x04) 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 COUNTER 5 4 CRC16 3 TIMEOUT 2 PID 1 DATAPID 0 DATATGL * COUNTER: Error Counter This field is incremented each time an error occurs (CRC16, TIMEOUT, PID, DATAPID or DATATGL). This field is cleared when receiving a good usb packet without any error. When this field reaches 3 (i.e., 3 consecutive errors), this pipe is automatically frozen (UPCONn.PFREEZE is set). Writing 0b00 to this field will clear the counter. CRC16: CRC16 Error This bit is set when a CRC16 error has been detected. Writing a zero to this bit will clear the bit. Writing a one to this bit has no effect. TIMEOUT: Time-Out Error This bit is set when a Time-Out error has been detected. Writing a zero to this bit will clear the bit. Writing a one to this bit has no effect. PID: PID Error This bit is set when a PID error has been detected. Writing a zero to this bit will clear the bit. Writing a one to this bit has no effect. DATAPID: Data PID Error This bit is set when a Data PID error has been detected. Writing a zero to this bit will clear the bit. Writing a one to this bit has no effect. DATATGL: Data Toggle Error This bit is set when a Data Toggle error has been detected. Writing a zero to this bit will clear the bit. Writing a one to this bit has no effect. * * * * * 477 32059I-06/2010 AT32UC3B 22.8.3.21 Host DMA Channel n Next Descriptor Address Register UHDMAnNEXTDESC, n in [1..6] Read/Write 0x0710 + (n - 1) * 0x10 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 27 26 25 24 NXTDESCADDR[31:24] 23 22 21 20 19 18 17 16 NXTDESCADDR[23:16] 15 14 13 12 11 10 9 8 NXTDESCADDR[15:8] 7 6 5 4 3 - 2 - 1 - 0 - NXTDESCADDR[7:4] Same as Section 22.8.2.17. 478 32059I-06/2010 AT32UC3B 22.8.3.22 Host DMA Channel n HSB Address Register UHDMAnADDR, n in [1..6] Read/Write 0x0714 + (n - 1) * 0x10 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 27 26 25 24 HSBADDR[31:24] 23 22 21 20 19 18 17 16 HSBADDR[23:16] 15 14 13 12 11 10 9 8 HSBADDR[15:8] 7 6 5 4 3 2 1 0 HSBADDR[7:0] Same as Section 22.8.2.18. 479 32059I-06/2010 AT32UC3B 22.8.3.23 USB Host DMA Channel n Control Register UHDMAnCONTROL, n in [1..6] Read/Write 0x0718 + (n - 1) * 0x10 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 27 26 25 24 CHBYTELENGTH[15:8] 23 22 21 20 19 18 17 16 CHBYTELENGTH[7:0] 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 BURSTLOC KEN 6 DESCLD IRQEN 5 EOBUFF IRQEN 4 EOTIRQEN 3 DMAENDEN 2 BUFFCLOSE INEN 1 LDNXTCHD ESCEN 0 CHEN Same as Section 22.8.2.19. (just replace the IN endpoint term by OUT endpoint, and vice-versa) 480 32059I-06/2010 AT32UC3B 22.8.3.24 USB Host DMA Channel n Status Register UHDMAnSTATUS, n in [1..6] Read/Write 0x071C + (n - 1) * 0x10 0x00000000 Register Name: Access Type: Offset: Reset Value: 31 30 29 28 27 26 25 24 CHBYTECNT[15:8] 23 22 21 20 19 18 17 16 CHBYTECNT[7:0] 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 DESCLD STA 5 EOCHBUFFS TA 4 EOTSTA 3 - 2 - 1 CHACTIVE 0 CHEN Same as Section 22.8.2.20. 481 32059I-06/2010 AT32UC3B 22.8.4 USB Pipe/Endpoint n FIFO Data Register (USBFIFOnDATA) The application has access to the physical DPRAM reserved for the Endpoint/Pipe through a 64KB logical address space. The application can access a 64KB buffer linearly or fixedly as the DPRAM address increment is fully handled by hardware. Byte, half-word and word access are supported. Data should be access in a big-endian way. Disabling the USBB (by writing a zero to the USBE bit) does not reset the DPRAM. 482 32059I-06/2010 AT32UC3B 23. Timer/Counter (TC) Rev: 2.2.2.1 23.1 Features * Three 16-bit Timer Counter channels * A wide range of functions including: - Frequency measurement - Event counting - Interval measurement - Pulse generation - Delay timing - Pulse width modulation - Up/down capabilities * Each channel is user-configurable and contains: - Three external clock inputs - Five internal clock inputs - Two multi-purpose input/output signals * Internal interrupt signal * Two global registers that act on all three TC channels 23.2 Overview The Timer Counter (TC) includes three identical 16-bit Timer Counter channels. Each channel can be independently programmed to perform a wide range of functions including frequency measurement, event counting, interval measurement, pulse generation, delay timing, and pulse width modulation. Each channel has three external clock inputs, five internal clock inputs, and two multi-purpose input/output signals which can be configured by the user. Each channel drives an internal interrupt signal which can be programmed to generate processor interrupts. The TC block has two global registers which act upon all three TC channels. The Block Control Register (BCR) allows the three channels to be started simultaneously with the same instruction. The Block Mode Register (BMR) defines the external clock inputs for each channel, allowing them to be chained. 483 32059I-06/2010 AT32UC3B 23.3 Block Diagram Figure 23-1. TC Block Diagram TIMER_CLOCK1 TCLK0 TIMER_CLOCK2 TIMER_CLOCK3 TCLK1 TIMER_CLOCK4 TIMER_CLOCK5 TCLK2 TIOA1 TIOA2 XC0 XC1 XC2 TC0XC0S SYNC Timer/Counter Channel 0 TIOA TIOB I/O Contr oller CLK0 CLK1 CLK2 A0 B0 TIOA0 TIOB0 INT0 TCLK0 TCLK1 TIOA0 TIOA2 TCLK2 XC0 XC1 XC2 TC1XC1S SYNC INT1 Timer/Counter Channel 1 TIOA TIOB TIOA1 TIOB1 A1 B1 TCLK0 TCLK1 TCLK2 TIOA0 TIOA1 XC0 XC1 XC2 TC2XC2S Timer/Counter Channel 2 TIOA TIOB TIOA2 TIOB2 A2 B2 SYNC INT2 Timer Count er Interrupt Controller 23.4 I/O Lines Description Table 23-1. Pin Name CLK0-CLK2 A0-A2 B0-B2 I/O Lines Description Description External Clock Input I/O Line A I/O Line B Type Input Input/Output Input/Output 23.5 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 23.5.1 I/O Lines The pins used for interfacing the compliant external devices may be multiplexed with I/O lines. The user must first program the I/O Controller to assign the TC pins to their peripheral functions. 484 32059I-06/2010 AT32UC3B 23.5.2 Power Management If the CPU enters a sleep mode that disables clocks used by the TC, the TC will stop functioning and resume operation after the system wakes up from sleep mode. 23.5.3 Clocks The clock for the TC bus interface (CLK_TC) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. It is recommended to disable the TC before disabling the clock, to avoid freezing the TC in an undefined state. 23.5.4 Interrupts The TC interrupt request line is connected to the interrupt controller. Using the TC interrupt requires the interrupt controller to be programmed first. 23.5.5 Debug Operation The Timer Counter clocks are frozen during debug operation, unless the OCD system keeps peripherals running in debug operation. 23.6 23.6.1 Functional Description TC Description The three channels of the Timer Counter are independent and identical in operation. The registers for channel programming are listed in Figure 23-3 on page 500. 23.6.1.1 Channel I/O Signals As described in Figure 23-1 on page 484, each Channel has the following I/O signals. Table 23-2. Channel I/O Signals Description Signal Name XC0, XC1, XC2 TIOA Description External Clock Inputs Capture mode: Timer Counter Input Waveform mode: Timer Counter Output Capture mode: Timer Counter Input Waveform mode: Timer Counter Input/Output Interrupt Signal Output Synchronization Input Signal Block/Channel Channel Signal TIOB INT SYNC 23.6.1.2 16-bit counter Each channel is organized around a 16-bit counter. The value of the counter is incremented at each positive edge of the selected clock. When the counter has reached the value 0xFFFF and passes to 0x0000, an overflow occurs and the Counter Overflow Status bit in the Channel n Status Register (SRn.COVFS) is set. The current value of the counter is accessible in real time by reading the Channel n Counter Value Register (CVn). The counter can be reset by a trigger. In this case, the counter value passes to 0x0000 on the next valid edge of the selected clock. 485 32059I-06/2010 AT32UC3B 23.6.1.3 Clock selection At block level, input clock signals of each channel can either be connected to the external inputs TCLK0, TCLK1 or TCLK2, or be connected to the configurable I/O signals A0, A1 or A2 for chaining by writing to the BMR register. See Figure 23-2 on page 486. Each channel can independently select an internal or external clock source for its counter: * Internal clock signals: TIMER_CLOCK1, TIMER_CLOCK2, TIMER_CLOCK3, TIMER_CLOCK4, TIMER_CLOCK5. See the Module Configuration Chapter for details about the connection of these clock sources. * External clock signals: XC0, XC1 or XC2. See the Module Configuration Chapter for details about the connection of these clock sources. This selection is made by the Clock Selection field in the Channel n Mode Register (CMRn.TCCLKS). The selected clock can be inverted with the Clock Invert bit in CMRn (CMRn.CLKI). This allows counting on the opposite edges of the clock. The burst function allows the clock to be validated when an external signal is high. The Burst Signal Selection field in the CMRn register (CMRn.BURST) defines this signal. Note: In all cases, if an external clock is used, the duration of each of its levels must be longer than the CLK_TC period. The external clock frequency must be at least 2.5 times lower than the CLK_TC. Figure 23-2. Clock Selection TCCLKS TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 XC0 XC1 XC2 CLKI Selected Clock BURST 1 23.6.1.4 Clock control The clock of each counter can be controlled in two different ways: it can be enabled/disabled and started/stopped. See Figure 23-3 on page 487. 486 32059I-06/2010 AT32UC3B * The clock can be enabled or disabled by the user by writing to the Counter Clock Enable/Disable Command bits in the Channel n Clock Control Register (CCRn.CLKEN and CCRn.CLKDIS). In Capture mode it can be disabled by an RB load event if the Counter Clock Disable with RB Loading bit in CMRn is written to one (CMRn.LDBDIS). In Waveform mode, it can be disabled by an RC Compare event if the Counter Clock Disable with RC Compare bit in CMRn is written to one (CMRn.CPCDIS). When disabled, the start or the stop actions have no effect: only a CLKEN command in CCRn can re-enable the clock. When the clock is enabled, the Clock Enabling Status bit is set in SRn (SRn.CLKSTA). * The clock can also be started or stopped: a trigger (software, synchro, external or compare) always starts the clock. In Capture mode the clock can be stopped by an RB load event if the Counter Clock Stopped with RB Loading bit in CMRn is written to one (CMRn.LDBSTOP). In Waveform mode it can be stopped by an RC compare event if the Counter Clock Stopped with RC Compare bit in CMRn is written to one (CMRn.CPCSTOP). The start and the stop commands have effect only if the clock is enabled. Figure 23-3. Clock Control Selected Clock Trigger CLKSTA CLKEN CLKDIS Q Q S R S R Counter Clock Stop Event Disable Event 23.6.1.5 TC operating modes Each channel can independently operate in two different modes: * Capture mode provides measurement on signals. * Waveform mode provides wave generation. The TC operating mode selection is done by writing to the Wave bit in the CCRn register (CCRn.WAVE). In Capture mode, TIOA and TIOB are configured as inputs. In Waveform mode, TIOA is always configured to be an output and TIOB is an output if it is not selected to be the external trigger. 487 32059I-06/2010 AT32UC3B 23.6.1.6 Trigger A trigger resets the counter and starts the counter clock. Three types of triggers are common to both modes, and a fourth external trigger is available to each mode. The following triggers are common to both modes: * Software Trigger: each channel has a software trigger, available by writing a one to the Software Trigger Command bit in CCRn (CCRn.SWTRG). * SYNC: each channel has a synchronization signal SYNC. When asserted, this signal has the same effect as a software trigger. The SYNC signals of all channels are asserted simultaneously by writing a one to the Synchro Command bit in the BCR register (BCR.SYNC). * Compare RC Trigger: RC is implemented in each channel and can provide a trigger when the counter value matches the RC value if the RC Compare Trigger Enable bit in CMRn (CMRn.CPCTRG) is written to one. The channel can also be configured to have an external trigger. In Capture mode, the external trigger signal can be selected between TIOA and TIOB. In Waveform mode, an external event can be programmed to be one of the following signals: TIOB, XC0, XC1, or XC2. This external event can then be programmed to perform a trigger by writing a one to the External Event Trigger Enable bit in CMRn (CMRn.ENETRG). If an external trigger is used, the duration of the pulses must be longer than the CLK_TC period in order to be detected. Regardless of the trigger used, it will be taken into account at the following active edge of the selected clock. This means that the counter value can be read differently from zero just after a trigger, especially when a low frequency signal is selected as the clock. 23.6.2 Capture Operating Mode This mode is entered by writing a zero to the CMRn.WAVE bit. Capture mode allows the TC channel to perform measurements such as pulse timing, frequency, period, duty cycle and phase on TIOA and TIOB signals which are considered as inputs. Figure 23-4 on page 490 shows the configuration of the TC channel when programmed in Capture mode. 23.6.2.1 Capture registers A and B Registers A and B (RA and RB) are used as capture registers. This means that they can be loaded with the counter value when a programmable event occurs on the signal TIOA. The RA Loading Selection field in CMRn (CMRn.LDRA) defines the TIOA edge for the loading of the RA register, and the RB Loading Selection field in CMRn (CMRn.LDRB) defines the TIOA edge for the loading of the RB register. RA is loaded only if it has not been loaded since the last trigger or if RB has been loaded since the last loading of RA. RB is loaded only if RA has been loaded since the last trigger or the last loading of RB. Loading RA or RB before the read of the last value loaded sets the Load Overrun Status bit in SRn (SRn.LOVRS). In this case, the old value is overwritten. 488 32059I-06/2010 AT32UC3B 23.6.2.2 Trigger conditions In addition to the SYNC signal, the software trigger and the RC compare trigger, an external trigger can be defined. The TIOA or TIOB External Trigger Selection bit in CMRn (CMRn.ABETRG) selects TIOA or TIOB input signal as an external trigger. The External Trigger Edge Selection bit in CMRn (CMRn.ETREDG) defines the edge (rising, falling or both) detected to generate an external trigger. If CMRn.ETRGEDG is zero (none), the external trigger is disabled. 489 32059I-06/2010 32059I-06/2010 TCCLKS CLKI CLKSTA CLKEN CLKDIS TIMER_CLOCK1 Q Q S R LDBSTOP BURST LDBDIS Register C Capture Register A SWTRG CLK OVF RESET Trig ABETRG ETRGEDG Edge Detector LD A R LDRB SR Edge Detector If RA is Loaded IMR CPCTRG 16-bit Counter Capture Register B S R Figure 23-4. Capture Mode TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 XC0 XC1 XC2 1 Compare RC = SYNC MTIOB TIOB CPCS COVFS LDRBS LDRAS LOVRS ETRGS MTIOA If RA is not Loaded or RB is Loaded Edge Detector TIOA Timer/Counter Channel INT AT32UC3B 490 AT32UC3B 23.6.3 Waveform Operating Mode Waveform operating mode is entered by writing a one to the CMRn.WAVE bit. In Waveform operating mode the TC channel generates one or two PWM signals with the same frequency and independently programmable duty cycles, or generates different types of oneshot or repetitive pulses. In this mode, TIOA is configured as an output and TIOB is defined as an output if it is not used as an external event. Figure 23-5 on page 492 shows the configuration of the TC channel when programmed in Waveform operating mode. 23.6.3.1 Waveform selection Depending on the Waveform Selection field in CMRn (CMRn.WAVSEL), the behavior of CVn varies. With any selection, RA, RB and RC can all be used as compare registers. RA Compare is used to control the TIOA output, RB Compare is used to control the TIOB output (if correctly configured) and RC Compare is used to control TIOA and/or TIOB outputs. 491 32059I-06/2010 O utput Contr oller EN G ETR SR TIOB Edge Detector BSWTRG IMR Timer/Counter Channel AT32UC3B 492 INT O utput Cont r oller 32059I-06/2010 TCCLKS CLKSTA CLKDIS CLKI Q Q R CPCSTOP AEEVT S R CPCDIS S ACPA CLKEN ACPC MTIOA TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 XC0 XC1 Figure 23-5. Waveform Mode TIOA XC2 BURST Register A WAVSEL 1 16-bit Counter CLK Register B Register C ASWTRG Compare RA = Compare RB = Compare RC = SWTRG RESET OVF BCPC T rig BCPB WAVSEL E VT E BEEVT E VTEDG E CPCS CPBS CPAS ETRGS COVFS MTIOB SYNC TIOB AT32UC3B 23.6.3.2 WAVSEL = 0 When CMRn.WAVSEL is zero, the value of CVn is incremented from 0 to 0xFFFF. Once 0xFFFF has been reached, the value of CVn is reset. Incrementation of CVn starts again and the cycle continues. See Figure 23-6 on page 493. An external event trigger or a software trigger can reset the value of CVn. It is important to note that the trigger may occur at any time. See Figure 23-7 on page 494. RC Compare cannot be programmed to generate a trigger in this configuration. At the same time, RC Compare can stop the counter clock (CMRn.CPCSTOP = 1) and/or disable the counter clock (CMRn.CPCDIS = 1). Figure 23-6. WAVSEL= 0 Without Trigger Counter Value 0xFFFF RC Counter cleared by compare match with 0xFFFF RB RA Waveform Examples TIOB Time TIOA 493 32059I-06/2010 AT32UC3B Figure 23-7. WAVSEL= 0 With Trigger Counter Value 0xFFFF RC RB Counter cleared by trigger Counter cleared by compare match with 0xFFFF RA Waveform Examples TIOB Time TIOA 23.6.3.3 WAVSEL = 2 When CMRn.WAVSEL is two, the value of CVn is incremented from zero to the value of RC, then automatically reset on a RC Compare. Once the value of CVn has been reset, it is then incremented and so on. See Figure 23-8 on page 495. It is important to note that CVn can be reset at any time by an external event or a software trigger if both are programmed correctly. See Figure 23-9 on page 495. In addition, RC Compare can stop the counter clock (CMRn.CPCSTOP) and/or disable the counter clock (CMRn.CPCDIS = 1). 494 32059I-06/2010 AT32UC3B Figure 23-8. WAVSEL = 2 Without Trigger Counter Value 0xFFFF RC Counter cleared by compare match with RC RB RA Waveform Examples TIOB Time TIOA Figure 23-9. WAVSEL = 2 With Trigger Counter Value 0xFFFF Counter cleared by compare match with RC RC RB Counter cleared by trigger RA Waveform Examples TIOB Time TIOA 23.6.3.4 WAVSEL = 1 When CMRn.WAVSEL is one, the value of CVn is incremented from 0 to 0xFFFF. Once 0xFFFF is reached, the value of CVn is decremented to 0, then re-incremented to 0xFFFF and so on. See Figure 23-10 on page 496. 495 32059I-06/2010 AT32UC3B A trigger such as an external event or a software trigger can modify CVn at any time. If a trigger occurs while CVn is incrementing, CVn then decrements. If a trigger is received while CVn is decrementing, CVn then increments. See Figure 23-11 on page 496. RC Compare cannot be programmed to generate a trigger in this configuration. At the same time, RC Compare can stop the counter clock (CMRn.CPCSTOP = 1) and/or disable the counter clock (CMRn.CPCDIS = 1). Figure 23-10. WAVSEL = 1 Without Trigger Counter Value 0xFFFF Counter decremented by compare match with 0xFFFF RC RB RA Waveform Examples TIOB Time TIOA Figure 23-11. WAVSEL = 1 With Trigger Counter Value 0xFFFF Counter decremented by trigger RC RB Counter incremented by trigger RA Counter decremented by compare match with 0xFFFF Waveform Examples TIOB Time TIOA 496 32059I-06/2010 AT32UC3B 23.6.3.5 WAVSEL = 3 When CMRn.WAVSEL is three, the value of CVn is incremented from zero to RC. Once RC is reached, the value of CVn is decremented to zero, then re-incremented to RC and so on. See Figure 23-12 on page 497. A trigger such as an external event or a software trigger can modify CVn at any time. If a trigger occurs while CVn is incrementing, CVn then decrements. If a trigger is received while CVn is decrementing, CVn then increments. See Figure 23-13 on page 498. RC Compare can stop the counter clock (CMRn.CPCSTOP = 1) and/or disable the counter clock (CMRn.CPCDIS = 1). Figure 23-12. WAVSEL = 3 Without Trigger Counter Value 0xFFFF Counter cleared by compare match with RC RC RB RA Waveform Examples TIOB Time TIOA 497 32059I-06/2010 AT32UC3B Figure 23-13. WAVSEL = 3 With Trigger Counter Value 0xFFFF Counter decremented by compare match with RC RC Counter decremented by trigger RB Counter incremented by trigger RA Waveform Examples TIOB Time TIOA 23.6.3.6 External event/trigger conditions An external event can be programmed to be detected on one of the clock sources (XC0, XC1, XC2) or TIOB. The external event selected can then be used as a trigger. The External Event Selection field in CMRn (CMRn.EEVT) selects the external trigger. The External Event Edge Selection field in CMRn (CMRn.EEVTEDG) defines the trigger edge for each of the possible external triggers (rising, falling or both). If CMRn.EEVTEDG is written to zero, no external event is defined. If TIOB is defined as an external event signal (CMRn.EEVT = 0), TIOB is no longer used as an output and the compare register B is not used to generate waveforms and subsequently no IRQs. In this case the TC channel can only generate a waveform on TIOA. When an external event is defined, it can be used as a trigger by writing a one to the CMRn.ENETRG bit. As in Capture mode, the SYNC signal and the software trigger are also available as triggers. RC Compare can also be used as a trigger depending on the CMRn.WAVSEL field. 23.6.3.7 Output controller The output controller defines the output level changes on TIOA and TIOB following an event. TIOB control is used only if TIOB is defined as output (not as an external event). The following events control TIOA and TIOB: * software trigger * external event * RC compare RA compare controls TIOA and RB compare controls TIOB. Each of these events can be programmed to set, clear or toggle the output as defined in the following fields in CMRn: 498 32059I-06/2010 AT32UC3B * RC Compare Effect on TIOB (CMRn.BCPC) * RB Compare Effect on TIOB (CMRn.BCPB) * RC Compare Effect on TIOA (CMRn.ACPC) * RA Compare Effect on TIOA (CMRn.ACPA) 499 32059I-06/2010 AT32UC3B 23.7 User Interface TC Register Memory Map Register Channel 0 Control Register Channel 0 Mode Register Channel 0 Counter Value Channel 0 Register A Channel 0 Register B Channel 0 Register C Channel 0 Status Register Interrupt Enable Register Channel 0 Interrupt Disable Register Channel 0 Interrupt Mask Register Channel 1 Control Register Channel 1 Mode Register Channel 1 Counter Value Channel 1 Register A Channel 1 Register B Channel 1 Register C Channel 1 Status Register Channel 1 Interrupt Enable Register Channel 1 Interrupt Disable Register Channel 1 Interrupt Mask Register Channel 2 Control Register Channel 2 Mode Register Channel 2 Counter Value Channel 2 Register A Channel 2 Register B Channel 2 Register C Channel 2 Status Register Channel 2 Interrupt Enable Register Channel 2 Interrupt Disable Register Channel 2 Interrupt Mask Register Block Control Register Block Mode Register Notes: Register Name CCR0 CMR0 CV0 RA0 RB0 RC0 SR0 IER0 IDR0 IMR0 CCR1 CMR1 CV1 RA1 RB1 RC1 SR1 IER1 IDR1 IMR1 CCR2 CMR2 CV2 RA2 RB2 RC2 SR2 IER2 IDR2 IMR2 BCR BMR Access Write-only Read/Write Read-only Read/Write(1) Read/Write(1) Read/Write Read-only Write-only Write-only Read-only Write-only Read/Write Read-only Read/Write Read/Write (1) (1) Table 23-3. Offset 0x00 0x04 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x40 0x44 0x50 0x54 0x58 0x5C 0x60 0x64 0x68 0x6C 0x80 0x84 0x90 0x94 0x98 0x9C 0xA0 0xA4 0xA8 0xAC 0xC0 0xC4 Reset 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 00x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 Read/Write Read-only Write-only Write-only Read-only Write-only Read/Write Read-only Read/Write(1) Read/Write(1) Read/Write Read-only Write-only Write-only Read-only Write-only Read/Write 1. Read-only if CMRn.WAVE is zero 500 32059I-06/2010 AT32UC3B 23.7.1 Name: Access Type: Offset: Reset Value: Channel Control Register CCR Write-only 0x00 + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 SWTRG 1 CLKDIS 0 CLKEN * SWTRG: Software Trigger Command 1: Writing a one to this bit will perform a software trigger: the counter is reset and the clock is started. 0: Writing a zero to this bit has no effect. * CLKDIS: Counter Clock Disable Command 1: Writing a one to this bit will disable the clock. 0: Writing a zero to this bit has no effect. * CLKEN: Counter Clock Enable Command 1: Writing a one to this bit will enable the clock if CLKDIS is not one. 0: Writing a zero to this bit has no effect. 501 32059I-06/2010 AT32UC3B 23.7.2 Name: Access Type: Offset: Reset Value: Channel Mode Register: Capture Mode CMR Read/Write 0x04 + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 LDRB 18 17 LDRA 16 15 WAVE 14 CPCTRG 13 - 12 - 11 - 10 ABETRG 9 ETRGEDG 8 7 LDBDIS 6 LDBSTOP 5 BURST 4 3 CLKI 2 1 TCCLKS 0 * LDRB: RB Loading Selection LDRB 0 1 2 3 Edge none rising edge of TIOA falling edge of TIOA each edge of TIOA * LDRA: RA Loading Selection LDRA 0 1 2 3 Edge none rising edge of TIOA falling edge of TIOA each edge of TIOA * WAVE 1: Capture mode is disabled (Waveform mode is enabled). 0: Capture mode is enabled. * CPCTRG: RC Compare Trigger Enable 1: RC Compare resets the counter and starts the counter clock. 0: RC Compare has no effect on the counter and its clock. * ABETRG: TIOA or TIOB External Trigger Selection 1: TIOA is used as an external trigger. 502 32059I-06/2010 AT32UC3B 0: TIOB is used as an external trigger. * ETRGEDG: External Trigger Edge Selection ETRGEDG 0 1 2 3 Edge none rising edge falling edge each edge * LDBDIS: Counter Clock Disable with RB Loading 1: Counter clock is disabled when RB loading occurs. 0: Counter clock is not disabled when RB loading occurs. * LDBSTOP: Counter Clock Stopped with RB Loading 1: Counter clock is stopped when RB loading occurs. 0: Counter clock is not stopped when RB loading occurs. * BURST: Burst Signal Selection BURST 0 1 2 3 Burst Signal Selection The clock is not gated by an external signal XC0 is ANDed with the selected clock XC1 is ANDed with the selected clock XC2 is ANDed with the selected clock * CLKI: Clock Invert 1: The counter is incremented on falling edge of the clock. 0: The counter is incremented on rising edge of the clock. * TCCLKS: Clock Selection TCCLKS 0 1 2 3 4 5 6 7 Clock Selected TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 XC0 XC1 XC2 503 32059I-06/2010 AT32UC3B 23.7.3 Name: Access Type: Offset: Reset Value: Channel Mode Register: Waveform Mode CMR Read/Write 0x04 + n * 0x40 0x00000000 31 BSWTRG 30 29 BEEVT 28 27 BCPC 26 25 BCPB 24 23 ASWTRG 22 21 AEEVT 20 19 ACPC 18 17 ACPA 16 15 WAVE 14 WAVSEL 13 12 ENETRG 11 EEVT 10 9 EEVTEDG 8 7 CPCDIS 6 CPCSTOP 5 BURST 4 3 CLKI 2 1 TCCLKS 0 * BSWTRG: Software Trigger Effect on TIOB BSWTRG 0 1 2 3 Effect none set clear toggle * BEEVT: External Event Effect on TIOB BEEVT 0 1 2 3 Effect none set clear toggle 504 32059I-06/2010 AT32UC3B * BCPC: RC Compare Effect on TIOB BCPC 0 1 2 3 Effect none set clear toggle * BCPB: RB Compare Effect on TIOB BCPB 0 1 2 3 Effect none set clear toggle * ASWTRG: Software Trigger Effect on TIOA ASWTRG 0 1 2 3 Effect none set clear toggle * AEEVT: External Event Effect on TIOA AEEVT 0 1 2 3 Effect none set clear toggle * ACPC: RC Compare Effect on TIOA ACPC 0 1 2 3 Effect none set clear toggle 505 32059I-06/2010 AT32UC3B * ACPA: RA Compare Effect on TIOA ACPA 0 1 2 3 Effect none set clear toggle * WAVE 1: Waveform mode is enabled. 0: Waveform mode is disabled (Capture mode is enabled). * WAVSEL: Waveform Selection WAVSEL 0 1 2 3 Effect UP mode without automatic trigger on RC Compare UPDOWN mode without automatic trigger on RC Compare UP mode with automatic trigger on RC Compare UPDOWN mode with automatic trigger on RC Compare * ENETRG: External Event Trigger Enable 1: The external event resets the counter and starts the counter clock. 0: The external event has no effect on the counter and its clock. In this case, the selected external event only controls the TIOA output. * EEVT: External Event Selection EEVT 0 1 2 3 Note: Signal selected as external event TIOB XC0 XC1 XC2 TIOB Direction input(1) output output output 1. If TIOB is chosen as the external event signal, it is configured as an input and no longer generates waveforms and subsequently no IRQs. * EEVTEDG: External Event Edge Selection EEVTEDG 0 1 2 3 Edge none rising edge falling edge each edge * CPCDIS: Counter Clock Disable with RC Compare 1: Counter clock is disabled when counter reaches RC. 0: Counter clock is not disabled when counter reaches RC. * CPCSTOP: Counter Clock Stopped with RC Compare 1: Counter clock is stopped when counter reaches RC. 506 32059I-06/2010 AT32UC3B 0: Counter clock is not stopped when counter reaches RC. * BURST: Burst Signal Selection BURST 0 1 2 3 Burst Signal Selection The clock is not gated by an external signal. XC0 is ANDed with the selected clock. XC1 is ANDed with the selected clock. XC2 is ANDed with the selected clock. * CLKI: Clock Invert 1: Counter is incremented on falling edge of the clock. 0: Counter is incremented on rising edge of the clock. * TCCLKS: Clock Selection TCCLKS 0 1 2 3 4 5 6 7 Clock Selected TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 XC0 XC1 XC2 507 32059I-06/2010 AT32UC3B 23.7.4 Name: Access Type: Offset: Reset Value: Channel Counter Value Register CV Read-only 0x10 + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 CV[15:8] 11 10 9 8 7 6 5 4 CV[7:0] 3 2 1 0 * CV: Counter Value CV contains the counter value in real time. 508 32059I-06/2010 AT32UC3B 23.7.5 Name: Access Type: Offset: Reset Value: Channel Register A RA Read-only if CMRn.WAVE = 0, Read/Write if CMRn.WAVE = 1 0x14 + n * 0X40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 RA[15:8] 11 10 9 8 7 6 5 4 RA[7:0] 3 2 1 0 * RA: Register A RA contains the Register A value in real time. 509 32059I-06/2010 AT32UC3B 23.7.6 Name: Access Type: Offset: Reset Value: Channel Register B RB Read-only if CMRn.WAVE = 0, Read/Write if CMRn.WAVE = 1 0x18 + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 RB[15:8] 11 10 9 8 7 6 5 4 RB[7:0] 3 2 1 0 * RB: Register B RB contains the Register B value in real time. 510 32059I-06/2010 AT32UC3B 23.7.7 Name: Access Type: Offset: Reset Value: Channel Register C RC Read/Write 0x1C + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 14 13 12 RC[15:8] 11 10 9 8 7 6 5 4 RC[7:0] 3 2 1 0 * RC: Register C RC contains the Register C value in real time. 511 32059I-06/2010 AT32UC3B 23.7.8 Name: Access Type: Offset: Reset Value: Channel Status Register SR Read-only 0x20 + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 MTIOB 17 MTIOA 16 CLKSTA 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 ETRGS 6 LDRBS 5 LDRAS 4 CPCS 3 CPBS 2 CPAS 1 LOVRS 0 COVFS Note: Reading the Status Register will also clear the interrupt bit for the corresponding interrupts. * MTIOB: TIOB Mirror 1: TIOB is high. If CMRn.WAVE is zero, this means that TIOB pin is high. If CMRn.WAVE is one, this means that TIOB is driven high. 0: TIOB is low. If CMRn.WAVE is zero, this means that TIOB pin is low. If CMRn.WAVE is one, this means that TIOB is driven low. * MTIOA: TIOA Mirror 1: TIOA is high. If CMRn.WAVE is zero, this means that TIOA pin is high. If CMRn.WAVE is one, this means that TIOA is driven high. 0: TIOA is low. If CMRn.WAVE is zero, this means that TIOA pin is low. If CMRn.WAVE is one, this means that TIOA is driven low. * CLKSTA: Clock Enabling Status 1: This bit is set when the clock is enabled. 0: This bit is cleared when the clock is disabled. * ETRGS: External Trigger Status 1: This bit is set when an external trigger has occurred. 0: This bit is cleared when the SR register is read. * LDRBS: RB Loading Status 1: This bit is set when an RB Load has occurred and CMRn.WAVE is zero. 0: This bit is cleared when the SR register is read. * LDRAS: RA Loading Status 1: This bit is set when an RA Load has occurred and CMRn.WAVE is zero. 0: This bit is cleared when the SR register is read. * CPCS: RC Compare Status 1: This bit is set when an RC Compare has occurred. 0: This bit is cleared when the SR register is read. 512 32059I-06/2010 AT32UC3B * CPBS: RB Compare Status 1: This bit is set when an RB Compare has occurred and CMRn.WAVE is one. 0: This bit is cleared when the SR register is read. * CPAS: RA Compare Status 1: This bit is set when an RA Compare has occurred and CMRn.WAVE is one. 0: This bit is cleared when the SR register is read. * LOVRS: Load Overrun Status 1: This bit is set when RA or RB have been loaded at least twice without any read of the corresponding register and CMRn.WAVE is zero. 0: This bit is cleared when the SR register is read. * COVFS: Counter Overflow Status 1: This bit is set when a counter overflow has occurred. 0: This bit is cleared when the SR register is read. 513 32059I-06/2010 AT32UC3B 23.7.9 Name: Access Type: Offset: Reset Value: Channel Interrupt Enable Register IER Write-only 0x24 + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 ETRGS 6 LDRBS 5 LDRAS 4 CPCS 3 CPBS 2 CPAS 1 LOVRS 0 COVFS Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. 514 32059I-06/2010 AT32UC3B 23.7.10 Name: Access Type: Offset: Reset Value: Channel Interrupt Disable Register IDR Write-only 0x28 + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 ETRGS 6 LDRBS 5 LDRAS 4 CPCS 3 CPBS 2 CPAS 1 LOVRS 0 COVFS Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. 515 32059I-06/2010 AT32UC3B 23.7.11 Name: Access Type: Offset: Reset Value: Channel Interrupt Mask Register IMR Read-only 0x2C + n * 0x40 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 ETRGS 6 LDRBS 5 LDRAS 4 CPCS 3 CPBS 2 CPAS 1 LOVRS 0 COVFS 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is set when the corresponding bit in IER is written to one. 516 32059I-06/2010 AT32UC3B 23.7.12 Name: Access Type: Offset: Reset Value: Block Control Register BCR Write-only 0xC0 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 - 1 - 0 SYNC * SYNC: Synchro Command 1: Writing a one to this bit asserts the SYNC signal which generates a software trigger simultaneously for each of the channels. 0: Writing a zero to this bit has no effect. 517 32059I-06/2010 AT32UC3B 23.7.13 Name: Access Type: Offset: Reset Value: Block Mode Register BMR Read/Write 0xC4 0x00000000 31 - 30 - 29 - 28 - 27 - 26 - 25 - 24 - 23 - 22 - 21 - 20 - 19 - 18 - 17 - 16 - 15 - 14 - 13 - 12 - 11 - 10 - 9 - 8 - 7 - 6 - 5 TC2XC2S 4 3 TC1XC1S 2 1 TC0XC0S 0 * TC2XC2S: External Clock Signal 2 Selection TC2XC2S 0 1 2 3 Signal Connected to XC2 TCLK2 none TIOA0 TIOA1 * TC1XC1S: External Clock Signal 1 Selection TC1XC1S 0 1 2 3 Signal Connected to XC1 TCLK1 none TIOA0 TIOA2 * TC0XC0S: External Clock Signal 0 Selection TC0XC0S 0 Signal Connected to XC0 TCLK0 518 32059I-06/2010 AT32UC3B 1 2 3 none TIOA1 TIOA2 519 32059I-06/2010 AT32UC3B 23.8 Module Configuration The specific configuration for each TC instance is listed in the following tables.The module bus clocks listed here are connected to the system bus clocks according to the table in the Power Manager section. Table 23-4. Module name TC0 Module Clock Name Clock name CLK_TC0 23.8.1 Clock Connections Each Timer/Counter channel can independently select an internal or external clock source for its counter: Table 23-5. Source Internal Timer/Counter clock connections Name TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 Connection 32 KHz Oscillator PBA Clock / 2 PBA Clock / 8 PBA Clock / 32 PBA Clock / 128 520 32059I-06/2010 AT32UC3B 24. Pulse Width Modulation Controller (PWM) Rev: 1.3.0.1 24.1 Features * 7 Channels * One 20-bit Counter Per Channel * Common Clock Generator Providing Thirteen Different Clocks - A Modulo n Counter Providing Eleven Clocks - Two Independent Linear Dividers Working on Modulo n Counter Outputs * Independent Channels - Independent Enable Disable Command for Each Channel - Independent Clock Selection for Each Channel - Independent Period and Duty Cycle for Each Channel - Double Buffering of Period or Duty Cycle for Each Channel - Programmable Selection of The Output Waveform Polarity for Each Channel - Programmable Center or Left Aligned Output Waveform for Each Channel 24.2 Description The PWM macrocell controls several channels independently. Each channel controls one square output waveform. Characteristics of the output waveform such as period, duty-cycle and polarity are configurable through the user interface. Each channel selects and uses one of the clocks provided by the clock generator. The clock generator provides several clocks resulting from the division of the PWM macrocell master clock. All PWM macrocell accesses are made through registers mapped on the peripheral bus. Channels can be synchronized, to generate non overlapped waveforms. All channels integrate a double buffering system in order to prevent an unexpected output waveform while modifying the period or the duty-cycle. 521 32059I-06/2010 AT32UC3B 24.3 Block Diagram Figure 24-1. Pulse Width Modulation Controller Block Diagram PWM Controller PWMx Channel Period Update Duty Cycle Comparator PWMx PWMx Clock Selector Counter PIO PWM0 Channel Period Update Duty Cycle Comparator PWM0 PWM0 Clock Selector Power Manager MCK Counter Interrupt Controller Clock Generator PB Interface Interrupt Generator Peripheral Bus 24.4 I/O Lines Description Each channel outputs one waveform on one external I/O line. Table 24-1. Name PWMx I/O Line Description Description PWM Waveform Output for channel x Type Output 522 32059I-06/2010 AT32UC3B 24.5 24.5.1 Product Dependencies I/O Lines The pins used for interfacing the PWM may be multiplexed with I/O controller lines. The programmer must first program the I/O controller to assign the desired PWM pins to their peripheral function. If I/O lines of the PWM are not used by the application, they can be used for other purposes by the I/O controller. Not all PWM outputs may be enabled. If an application requires only four channels, then only four I/O lines will be assigned to PWM outputs. 24.5.2 Debug operation The PWM clock is running during debug operation. 24.5.3 Power Management The PWM clock is generated by the Power Manager. Before using the PWM, the programmer must ensure that the PWM clock is enabled in the Power Manager. However, if the application does not require PWM operations, the PWM clock can be stopped when not needed and be restarted later. In this case, the PWM will resume its operations where it left off. In the PWM description, Master Clock (MCK) is the clock of the peripheral bus to which the PWM is connected. 24.5.4 Interrupt Sources The PWM interrupt line is connected to the interrupt controller. Using the PWM interrupt requires the interrupt controller to be programmed first. 523 32059I-06/2010 AT32UC3B 24.6 Functional Description The PWM macrocell is primarily composed of a clock generator module and 7 channels. - Clocked by the system clock, MCK, the clock generator module provides 13 clocks. - Each channel can independently choose one of the clock generator outputs. - Each channel generates an output waveform with attributes that can be defined independently for each channel through the user interface registers. 24.6.1 PWM Clock Generator Figure 24-2. Functional View of the Clock Generator Block Diagram MCK modulo n counter MCK MCK/2 MCK/4 MCK/8 MCK/16 MCK/32 MCK/64 MCK/128 MCK/256 MCK/512 MCK/1024 Divider A clkA PREA DIVA PWM_MR Divider B clkB PREB DIVB PWM_MR Caution: Before using the PWM macrocell, the programmer must ensure that the PWM clock in the Power Manager is enabled. The PWM macrocell master clock, MCK, is divided in the clock generator module to provide different clocks available for all channels. Each channel can independently select one of the divided clocks. 524 32059I-06/2010 AT32UC3B The clock generator is divided in three blocks: - a modulo n counter which provides 11 clocks: FMCK, FMCK/2, FMCK/4, FMCK/8, FMCK/16, FMCK/32, FMCK/64, FMCK/128, FMCK/256, FMCK/512, FMCK/1024 - two linear dividers (1, 1/2, 1/3, ... 1/255) that provide two separate clocks: clkA and clkB Each linear divider can independently divide one of the clocks of the modulo n counter. The selection of the clock to be divided is made according to the PREA (PREB) field of the PWM Mode register (MR). The resulting clock clkA (clkB) is the clock selected divided by DIVA (DIVB) field value in the PWM Mode register (MR). After a reset of the PWM controller, DIVA (DIVB) and PREA (PREB) in the PWM Mode register are set to 0. This implies that after reset clkA (clkB) are turned off. At reset, all clocks provided by the modulo n counter are turned off except clock "clk". This situation is also true when the PWM master clock is turned off through the Power Management Controller. 24.6.2 24.6.2.1 PWM Channel Block Diagram Figure 24-3. Functional View of the Channel Block Diagram inputs from clock generator Channel Clock Selector Internal Counter Comparator PWMx output waveform inputs from Peripheral Bus Each of the 7 channels is composed of three blocks: * A clock selector which selects one of the clocks provided by the clock generator described in Section 24.6.1 "PWM Clock Generator" on page 524. * An internal counter clocked by the output of the clock selector. This internal counter is incremented or decremented according to the channel configuration and comparators events. The size of the internal counter is 20 bits. * A comparator used to generate events according to the internal counter value. It also computes the PWMx output waveform according to the configuration. 24.6.2.2 Waveform Properties The different properties of output waveforms are: * the internal clock selection. The internal channel counter is clocked by one of the clocks provided by the clock generator described in the previous section. This channel parameter is defined in the CPRE field of the CMRx register. This field is reset at 0. * the waveform period. This channel parameter is defined in the CPRD field of the CPRDx register. 525 32059I-06/2010 AT32UC3B - If the waveform is left aligned, then the output waveform period depends on the counter source clock and can be calculated: By using the Master Clock (MCK) divided by an X given prescaler value (with X being 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, or 1024), the resulting period formula will be: ( X x CPRD ) -----------------------------MCK By using a Master Clock divided by one of both DIVA or DIVB divider, the formula becomes, respectively: ( CRPD x DIVA ) ----------------------------------------- or ( CRPD x DIVAB ) --------------------------------------------MCK MCK If the waveform is center aligned then the output waveform period depends on the counter source clock and can be calculated: By using the Master Clock (MCK) divided by an X given prescaler value (with X being 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, or 1024). The resulting period formula will be: ( 2 x X x CPRD ) ---------------------------------------MCK By using a Master Clock divided by one of both DIVA or DIVB divider, the formula becomes, respectively: ( 2 x CPRD x DIVA ) --------------------------------------------------- or ( 2 x CPRD x DIVB ) --------------------------------------------------MCK MCK * the waveform duty cycle. This channel parameter is defined in the CDTY field of the CDTYx register. If the waveform is left aligned then: duty cycle = ( period - 1 fchannel_x_clock x CDTY ) ------------------------------------------------------------------------------------------------------period If the waveform is center aligned, then: ( ( period 2 ) - 1 fchannel_x_clock x CDTY ) ) duty cycle = ---------------------------------------------------------------------------------------------------------------------( period 2 ) * the waveform polarity. At the beginning of the period, the signal can be at high or low level. This property is defined in the CPOL field of the CMRx register. By default the signal starts by a low level. * the waveform alignment. The output waveform can be left or center aligned. Center aligned waveforms can be used to generate non overlapped waveforms. This property is defined in the CALG field of the CMRx register. The default mode is left aligned. 526 32059I-06/2010 AT32UC3B Figure 24-4. Non Overlapped Center Aligned Waveforms No overlap PWM0 PWM1 Period Note: 1. See Figure 24-5 on page 528 for a detailed description of center aligned waveforms. When center aligned, the internal channel counter increases up to CPRD and.decreases down to 0. This ends the period. When left aligned, the internal channel counter increases up to CPRD and is reset. This ends the period. Thus, for the same CPRD value, the period for a center aligned channel is twice the period for a left aligned channel. Waveforms are fixed at 0 when: * CDTY = CPRD and CPOL = 0 * CDTY = 0 and CPOL = 1 Waveforms are fixed at 1 (once the channel is enabled) when: * CDTY = 0 and CPOL = 0 * CDTY = CPRD and CPOL = 1 The waveform polarity must be set before enabling the channel. This immediately affects the channel output level. Changes on channel polarity are not taken into account while the channel is enabled. 527 32059I-06/2010 AT32UC3B Figure 24-5. Waveform Properties PWM_MCKx CHIDx(PWM_SR) CHIDx(PWM_ENA) CHIDx(PWM_DIS) Center Aligned CALG(PWM_CMRx) = 1 PWM_CCNTx CPRD(PWM_CPRDx) CDTY(PWM_CDTYx) Period Output Waveform PWMx CPOL(PWM_CMRx) = 0 Output Waveform PWMx CPOL(PWM_CMRx) = 1 CHIDx(PWM_ISR) PWM_CCNTx CPRD(PWM_CPRDx) CDTY(PWM_CDTYx) Left Aligned CALG(PWM_CMRx) = 0 Period Output Waveform PWMx CPOL(PWM_CMRx) = 0 Output Waveform PWMx CPOL(PWM_CMRx) = 1 CHIDx(PWM_ISR) 528 32059I-06/2010 AT32UC3B 24.6.3 24.6.3.1 PWM Controller Operations Initialization Before enabling the output channel, this channel must have been configured by the software application: * Configuration of the clock generator if DIVA and DIVB are required * Selection of the clock for each channel (CPRE field in the CMRx register) * Configuration of the waveform alignment for each channel (CALG field in the CMRx register) * Configuration of the period for each channel (CPRD in the CPRDx register). Writing in CPRDx Register is possible while the channel is disabled. After validation of the channel, the user must use CUPDx Register to update CPRDx as explained below. * Configuration of the duty cycle for each channel (CDTY in the CDTYx register). Writing in CDTYx Register is possible while the channel is disabled. After validation of the channel, the user must use CUPDx Register to update CDTYx as explained below. * Configuration of the output waveform polarity for each channel (CPOL in the CMRx register) * Enable Interrupts (Writing CHIDx in the IER register) * Enable the PWM channel (Writing CHIDx in the ENA register) It is possible to synchronize different channels by enabling them at the same time by means of writing simultaneously several CHIDx bits in the ENA register. In such a situation, all channels may have the same clock selector configuration and the same period specified. 24.6.3.2 Source Clock Selection Criteria The large number of source clocks can make selection difficult. The relationship between the value in the Period Register (CPRDx) and the Duty Cycle Register (CDTYx) can help the user in choosing. The event number written in the Period Register gives the PWM accuracy. The Duty Cycle quantum cannot be lower than 1/CPRDx value. The higher the value of CPRDx, the greater the PWM accuracy. For example, if the user sets 15 (in decimal) in CPRDx, the user is able to set a value between 1 up to 14 in CDTYx Register. The resulting duty cycle quantum cannot be lower than 1/15 of the PWM period. 24.6.3.3 Changing the Duty Cycle or the Period It is possible to modulate the output waveform duty cycle or period. To prevent unexpected output waveform, the user must use the update register (PWM_CUPDx) to change waveform parameters while the channel is still enabled. The user can write a new period value or duty cycle value in the update register (CUPDx). This register holds the new value until the end of the current cycle and updates the value for the next cycle. Depending on the CPD field in the CMRx register, CUPDx either updates CPRDx or CDTYx. Note that even if the update register is used, the period must not be smaller than the duty cycle. 529 32059I-06/2010 AT32UC3B Figure 24-6. Synchronized Period or Duty Cycle Update User's Writing PWM_CUPDx Value 1 0 PWM_CMRx. CPD PWM_CPRDx PWM_CDTYx End of Cycle To prevent overwriting the CUPDx by software, the user can use status events in order to synchronize his software. Two methods are possible. In both, the user must enable the dedicated interrupt in IER at PWM Controller level. The first method (polling method) consists of reading the relevant status bit in ISR Register according to the enabled channel(s). See Figure 24-7. The second method uses an Interrupt Service Routine associated with the PWM channel. Note: Reading the ISR register automatically clears CHIDx flags. Figure 24-7. Polling Method PWM_ISR Read Acknowledgement and clear previous register state Writing in CPD field Update of the Period or Duty Cycle CHIDx = 1 YES Writing in PWM_CUPDx The last write has been taken into account Note: Polarity and alignment can be modified only when the channel is disabled. 530 32059I-06/2010 AT32UC3B 24.6.3.4 Interrupts Depending on the interrupt mask in the IMR register, an interrupt is generated at the end of the corresponding channel period. The interrupt remains active until a read operation in the ISR register occurs. A channel interrupt is enabled by setting the corresponding bit in the IER register. A channel interrupt is disabled by setting the corresponding bit in the IDR register. 531 32059I-06/2010 AT32UC3B 24.7 User Interface PWM Controller Memory Map Register PWM Mode Register PWM Enable Register PWM Disable Register PWM Status Register PWM Interrupt Enable Register PWM Interrupt Disable Register PWM Interrupt Mask Register PWM Interrupt Status Register Channel 0 Mode Register Channel 0 Duty Cycle Register Channel 0 Period Register Channel 0 Counter Register Channel 0 Update Register Channel 1 Mode Register Channel 1 Duty Cycle Register Channel 1 Period Register Channel 1 Counter Register Channel 1 Update Register Name MR ENA DIS SR IER IDR IMR ISR CMR0 CDTY0 CPRD0 CCNT0 CUPD0 CMR1 CDTY1 CPRD1 CCNT1 CUPD1 Access Read/Write Write-only Write-only Read-only Write-only Write-only Read-only Read-only Read/Write Read/Write Read/Write Read-only Write-only Read/Write Read/Write Read/Write Read-only Write-only Peripheral Reset Value 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 - Table 24-2. Offset 0x000 0x004 0x008 0x00C 0x010 0x014 0x018 0x01C 0x200 0x204 0x208 0x20C 0x210 0x220 0x224 0x228 0x22C 0x230 532 32059I-06/2010 AT32UC3B 24.7.1 Name: Access Type: Offset: Reset Value: 31 - 23 30 - 22 Mode Register MR Read/Write 0x000 0x00000000 29 - 21 28 - 20 DIVB 15 - 7 14 - 6 13 - 5 12 - 4 DIVA 11 10 PREA 3 2 1 0 9 8 27 26 PREB 19 18 17 16 25 24 * DIVA, DIVB: CLKA, CLKB Divide Factor DIVA, DIVB 0 1 2-255 CLKA, CLKB CLKA, CLKB clock is turned off CLKA, CLKB clock is clock selected by PREA, PREB CLKA, CLKB clock is clock selected by PREA, PREB divided by DIVA, DIVB factor. * PREA, PREB PREA, PREB 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 1 1 1 1 0 0 0 Other 0 0 1 1 0 0 1 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 MCK. MCK/2 MCK/4 MCK/8 MCK/16 MCK/32 MCK/64 MCK/128 MCK/256 MCK/512 MCK/1024 Reserved Divider Input Clock 533 32059I-06/2010 AT32UC3B 24.7.2 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - 30 - 22 - 14 - 6 CHID6 Enable Register ENA Write-only 0x004 29 - 21 - 13 - 5 CHID5 28 - 20 - 12 - 4 CHID4 27 - 19 - 11 - 3 CHID3 26 - 18 - 10 - 2 CHID2 25 - 17 - 9 - 1 CHID1 24 - 16 - 8 - 0 CHID0 * CHIDx: Channel ID 0 = No effect. 1 = Enable PWM output for channel x. 534 32059I-06/2010 AT32UC3B 24.7.3 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - 30 - 22 - 14 - 6 CHID6 Disable Register DIS Write-only 0x008 29 - 21 - 13 - 5 CHID5 28 - 20 - 12 - 4 CHID4 27 - 19 - 11 - 3 CHID3 26 - 18 - 10 - 2 CHID2 25 - 17 - 9 - 1 CHID1 24 - 16 - 8 - 0 CHID0 * CHIDx: Channel ID 0 = No effect. 1 = Disable PWM output for channel x. 535 32059I-06/2010 AT32UC3B 24.7.4 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - 30 - 22 - 14 - 6 CHID6 Status Register SR Read-only 0x00C 0x00000000 29 - 21 - 13 - 5 CHID5 28 - 20 - 12 - 4 CHID4 27 - 19 - 11 - 3 CHID3 26 - 18 - 10 - 2 CHID2 25 - 17 - 9 - 1 CHID1 24 - 16 - 8 - 0 CHID0 * CHIDx: Channel ID 0 = PWM output for channel x is disabled. 1 = PWM output for channel x is enabled. 536 32059I-06/2010 AT32UC3B 24.7.5 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - 30 - 22 - 14 - 6 CHID6 PWM Interrupt Enable Register IER Write-only 0x010 29 - 21 - 13 - 5 CHID5 28 - 20 - 12 - 4 CHID4 27 - 19 - 11 - 3 CHID3 26 - 18 - 10 - 2 CHID2 25 - 17 - 9 - 1 CHID1 24 - 16 - 8 - 0 CHID0 * CHIDx: Channel ID. 0 = No effect. 1 = Enable interrupt for PWM channel x. 537 32059I-06/2010 AT32UC3B 24.7.6 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - 30 - 22 - 14 - 6 CHID6 Interrupt Disable Register IDR Write-only 0x014 29 - 21 - 13 - 5 CHID5 28 - 20 - 12 - 4 CHID4 27 - 19 - 11 - 3 CHID3 26 - 18 - 10 - 2 CHID2 25 - 17 - 9 - 1 CHID1 24 - 16 - 8 - 0 CHID0 * CHIDx: Channel ID. 0 = No effect. 1 = Disable interrupt for PWM channel x. 538 32059I-06/2010 AT32UC3B 24.7.7 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - 30 - 22 - 14 - 6 CHID6 Interrupt Mask Register IMR Read-only 0x018 0x00000000 29 - 21 - 13 - 5 CHID5 28 - 20 - 12 - 4 CHID4 27 - 19 - 11 - 3 CHID3 26 - 18 - 10 - 2 CHID2 25 - 17 - 9 - 1 CHID1 24 - 16 - 8 - 0 CHID0 * CHIDx: Channel ID. 0 = Interrupt for PWM channel x is disabled. 1 = Interrupt for PWM channel x is enabled. 539 32059I-06/2010 AT32UC3B 24.7.8 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - 30 - 22 - 14 - 6 CHID6 Interrupt Status Register ISR Read-only 0x01C 0x00000000 29 - 21 - 13 - 5 CHID5 28 - 20 - 12 - 4 CHID4 27 - 19 - 11 - 3 CHID3 26 - 18 - 10 - 2 CHID2 25 - 17 - 9 - 1 CHID1 24 - 16 - 8 - 0 CHID0 * CHIDx: Channel ID 0 = No new channel period since the last read of the ISR register. 1 = At least one new channel period since the last read of the ISR register. Note: Reading ISR automatically clears CHIDx flags. 540 32059I-06/2010 AT32UC3B 24.7.9 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - 30 - 22 - 14 - 6 - Channel Mode Register CMRx Read/Write 0x200 0x00000000 29 - 21 - 13 - 5 - 28 - 20 - 12 - 4 - 27 - 19 - 11 - 3 26 - 18 - 10 CPD 2 CPRE 25 - 17 - 9 CPOL 1 24 - 16 - 8 CALG 0 * CPRE: Channel Pre-scaler CPRE 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 Other 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 MCK MCK/2 MCK/4 MCK/8 MCK/16 MCK/32 MCK/64 MCK/128 MCK/256 MCK/512 MCK/1024 CLKA CLKB Reserved Channel Pre-scaler * CALG: Channel Alignment 0 = The period is left aligned. 1 = The period is center aligned. * CPOL: Channel Polarity 0 = The output waveform starts at a low level. 541 32059I-06/2010 AT32UC3B 1 = The output waveform starts at a high level. * CPD: Channel Update Period 0 = Writing to the CUPDx will modify the duty cycle at the next period start event. 1 = Writing to the CUPDx will modify the period at the next period start event. 542 32059I-06/2010 AT32UC3B 24.7.10 Name: Access Type: Offset: Reset Value: 31 30 Channel Duty Cycle Register CDTYx Read/Write 0x204 0x00000000 29 28 CDTY 27 26 25 24 23 22 21 20 CDTY 19 18 17 16 15 14 13 12 CDTY 11 10 9 8 7 6 5 4 CDTY 3 2 1 0 Only the first 20 bits (internal channel counter size) are significant. * CDTY: Channel Duty Cycle Defines the waveform duty cycle. This value must be defined between 0 and CPRD (CPRx). 543 32059I-06/2010 AT32UC3B 24.7.11 Name: Access Type: Offset: Reset Value: 31 30 Channel Period Register CPRDx Read/Write 0x208 0x00000000 29 28 CPRD 27 26 25 24 23 22 21 20 CPRD 19 18 17 16 15 14 13 12 CPRD 11 10 9 8 7 6 5 4 CPRD 3 2 1 0 Only the first 20 bits (internal channel counter size) are significant. * CPRD: Channel Period If the waveform is left-aligned, then the output waveform period depends on the counter source clock and can be calculated: - By using the Master Clock (MCK) divided by an X given prescaler value (with X being 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, or 1024). The resulting period formula will be: ( X x CPRD ) -----------------------------MCK - By using a Master Clock divided by one of both DIVA or DIVB divider, the formula becomes, respectively: ( CRPD x DIVA ) ----------------------------------------- or ( CRPD x DIVAB ) --------------------------------------------MCK MCK If the waveform is center-aligned, then the output waveform period depends on the counter source clock and can be calculated: - By using the Master Clock (MCK) divided by an X given prescaler value (with X being 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, or 1024). The resulting period formula will be: ( 2 x X x CPRD ) ---------------------------------------MCK - By using a Master Clock divided by one of both DIVA or DIVB divider, the formula becomes, respectively: ( 2 x CPRD x DIVA ) ( 2 x CPRD x DIVB ) --------------------------------------------------- or --------------------------------------------------MCK MCK 544 32059I-06/2010 AT32UC3B 24.7.12 Name: Access Type: Offset: Reset Value: 31 30 Channel Counter Register CCNTx Read-only 0x20C 0x00000000 29 28 CNT 27 26 25 24 23 22 21 20 CNT 19 18 17 16 15 14 13 12 CNT 11 10 9 8 7 6 5 4 CNT 3 2 1 0 * CNT: Channel Counter Register Internal counter value. This register is reset when: * the channel is enabled (writing CHIDx in the ENA register). * the counter reaches CPRD value defined in the CPRDx register if the waveform is left aligned. 545 32059I-06/2010 AT32UC3B 24.7.13 Name: Access Type: Offset: Reset Value: 31 30 PWM Channel Update Register CUPDx Write-only 0x210 0x00000000 29 28 CUPD 27 26 25 24 23 22 21 20 CUPD 19 18 17 16 15 14 13 12 CUPD 11 10 9 8 7 6 5 4 CUPD 3 2 1 0 This register acts as a double buffer for the period or the duty cycle. This prevents an unexpected waveform when modifying the waveform period or duty-cycle. Only the first 20 bits (internal channel counter size) are significant. CPD (CMRx Register) 0 1 The duty-cycle (CDTY in the CDTYx register) is updated with the CUPD value at the beginning of the next period. The period (CPRD in the CPRDx register) is updated with the CUPD value at the beginning of the next period. 546 32059I-06/2010 AT32UC3B 25. Analog-to-Digital Converter (ADC) Rev: 2.0.0.1 25.1 Features * Integrated multiplexer offering up to eight independent analog inputs * Individual enable and disable of each channel * Hardware or software trigger - External trigger pin - Timer counter outputs (corresponding TIOA trigger) * Peripheral DMA Controller support * Possibility of ADC timings configuration * Sleep mode and conversion sequencer - Automatic wakeup on trigger and back to sleep mode after conversions of all enabled channels 25.2 Overview The Analog-to-Digital Converter (ADC) is based on a Successive Approximation Register (SAR) 10-bit ADC. It also integrates an 8-to-1 analog multiplexer, making possible the analog-to-digital conversions of 8 analog lines. The conversions extend from 0V to ADVREF. The ADC supports an 8-bit or 10-bit resolution mode, and conversion results are reported in a common register for all channels, as well as in a channel-dedicated register. Software trigger, external trigger on rising edge of the TRIGGER pin, or internal triggers from timer counter output(s) are configurable. The ADC also integrates a sleep mode and a conversion sequencer and connects with a Peripheral DMA Controller channel. These features reduce both power consumption and processor intervention. Finally, the user can configure ADC timings, such as startup time and sample & hold time. 547 32059I-06/2010 AT32UC3B 25.3 Block Diagram Timer Counter Channels Figure 25-1. ADC Block Diagram ADC Trigger Selection TRIGGER Control Logic ADC Interrupt Interrupt Controller VDDANA ADVREF Peripheral DMA Controller User Interface High Speed Bus (HSB) ADDedicated Analog Inputs AD- Peripheral Bridge AD- Analog Inputs Multiplexed With I/O lines ADADI/O Controller Successive Approximation Register Analog-to-Digital Converter Peripheral Bus (PB) AD- GND 25.4 I/O Lines Description ADC Pins Description Description Analog power supply Reference voltage Analog input channels External trigger Table 25-1. Pin Name VDDANA ADVREF AD[0] - AD[7] TRIGGER 25.5 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 25.5.1 I/O Lines The TRIGGER pin may be shared with other peripheral functions through the I/O Controller. 548 32059I-06/2010 AT32UC3B 25.5.2 Power Management In sleep mode, the ADC clock is automatically stopped after each conversion. As the logic is small and the ADC cell can be put into sleep mode, the Power Manager has no effect on the ADC behavior. 25.5.3 Clocks The clock for the ADC bus interface (CLK_ADC) is generated by the Power Manager. This clock is enabled at reset, and can be disabled in the Power Manager. It is recommended to disable the ADC before disabling the clock, to avoid freezing the ADC in an undefined state. The CLK_ADC clock frequency must be in line with the ADC characteritics. Refer to Electrical Characteristics section for details. 25.5.4 Interrupts The ADC interrupt request line is connected to the interrupt controller. Using the ADC interrupt requires the interrupt controller to be programmed first. 25.5.5 Analog Inputs The analog input pins can be multiplexed with I/O lines. In this case, the assignment of the ADC input is automatically done as soon as the corresponding I/O is configured through the I/O contoller. By default, after reset, the I/O line is configured as a logic input. 25.5.6 Timer Triggers Timer Counters may or may not be used as hardware triggers depending on user requirements. Thus, some or all of the timer counters may be non-connected. 25.6 25.6.1 Functional Description Analog-to-digital Conversion The ADC uses the ADC Clock to perform conversions. Converting a single analog value to a 10bit digital data requires sample and hold clock cycles as defined in the Sample and Hold Time field of the Mode Register (MR.SHTIM) and 10 ADC Clock cycles. The ADC Clock frequency is selected in the Prescaler Rate Selection field of the MR register (MR.PRESCAL). The ADC Clock range is between CLK_ADC/2, if the PRESCAL field is 0, and CLK_ADC/128, if the PRESCAL field is 63 (0x3F). The PRESCAL field must be written in order to provide an ADC Clock frequency according to the parameters given in the Electrical Characteristics chapter. 25.6.2 Conversion Reference The conversion is performed on a full range between 0V and the reference voltage pin ADVREF. Analog input values between these voltages are converted to digital values based on a linear conversion. 25.6.3 Conversion Resolution The ADC supports 8-bit or 10-bit resolutions. The 8-bit selection is performed by writing a one to the Resolution bit in the MR register (MR.LOWRES). By default, after a reset, the resolution is the highest and the Converted Data field in the Channel Data Registers (CDRn.DATA) is fully used. By writing a one to the LOWRES bit, the ADC switches in the lowest resolution and the conversion results can be read in the eight lowest significant bits of the Channel Data Registers 549 32059I-06/2010 AT32UC3B (CDRn). The two highest bits of the DATA field in the corresponding CDRn register will be read as zero. The two highest bits of the Last Data Converted field in the Last Converted Data Register (LCDR.LDATA) will be read as zero too. Moreover, when a Peripheral DMA channel is connected to the ADC, a 10-bit resolution sets the transfer request size to 16-bit. Writing a one to the LOWRES bit automatically switches to 8-bit data transfers. In this case, the destination buffers are optimized. 25.6.4 Conversion Results When a conversion is completed, the resulting 10-bit digital value is stored in the CDR register of the current channel and in the LCDR register. Channels are enabled by writing a one to the Channel n Enable bit (CHn) in the CHER register. The corresponding channel End of Conversion bit in the Status Register (SR.EOCn) and the Data Ready bit in the SR register (SR.DRDY) are set. In the case of a connected Peripheral DMA channel, DRDY rising triggers a data transfer request. In any case, either EOC or DRDY can trigger an interrupt. Reading one of the CDRn registers clears the corresponding EOC bit. Reading LCDR clears the DRDY bit and the EOC bit corresponding to the last converted channel. Figure 25-2. EOCn and DRDY Flag Behavior Write CR With START=1 Read CDRn Write CR With START=1 Read LCDR CHn(CHSR) EOCn(SR) Conversion Time Conversion Time DRDY(SR) 550 32059I-06/2010 AT32UC3B If the CDR register is not read before further incoming data is converted, the corresponding Overrun Error bit in the SR register (SR.OVREn) is set. In the same way, new data converted when DRDY is high sets the General Overrun Error bit in the SR register (SR.GOVRE). The OVREn and GOVRE bits are automatically cleared when the SR register is read. Figure 25-3. GOVRE and OVREn Flag Behavior Read SR TRIGGER CH0(CHSR) CH1(CHSR) LCDR CRD0 CRD1 Undefined Data Undefined Data Undefined Data Conversion Data A Data A Data B Data C Data C Data B EOC0(SR) Conversion Read CDR0 EOC1(SR) Conversion Read CDR1 GOVRE(SR) DRDY(ASR) OVRE0(SR) Warning: If the corresponding channel is disabled during a conversion or if it is disabled and then reenabled during a conversion, its associated data and its corresponding EOC and OVRE flags in SR are unpredictable. 551 32059I-06/2010 AT32UC3B 25.6.5 Conversion Triggers Conversions of the active analog channels are started with a software or a hardware trigger. The software trigger is provided by writing a one to the START bit in the Control Register (CR.START). The hardware trigger can be one of the TIOA outputs of the Timer Counter channels, or the external trigger input of the ADC (TRIGGER). The hardware trigger is selected with the Trigger Selection field in the Mode Register (MR.TRIGSEL). The selected hardware trigger is enabled by writing a one to the Trigger Enable bit in the Mode Register (MR.TRGEN). If a hardware trigger is selected, the start of a conversion is detected at each rising edge of the selected signal. If one of the TIOA outputs is selected, the corresponding Timer Counter channel must be programmed in Waveform Mode. Only one start command is necessary to initiate a conversion sequence on all the channels. The ADC hardware logic automatically performs the conversions on the active channels, then waits for a new request. The Channel Enable (CHER) and Channel Disable (CHDR) Registers enable the analog channels to be enabled or disabled independently. If the ADC is used with a Peripheral DMA Controller, only the transfers of converted data from enabled channels are performed and the resulting data buffers should be interpreted accordingly. Warning: Enabling hardware triggers does not disable the software trigger functionality. Thus, if a hardware trigger is selected, the start of a conversion can be initiated either by the hardware or the software trigger. 25.6.6 Sleep Mode and Conversion Sequencer The ADC Sleep Mode maximizes power saving by automatically deactivating the ADC when it is not being used for conversions. Sleep Mode is selected by writing a one to the Sleep Mode bit in the Mode Register (MR.SLEEP). The SLEEP mode is automatically managed by a conversion sequencer, which can automatically process the conversions of all channels at lowest power consumption. When a start conversion request occurs, the ADC is automatically activated. As the analog cell requires a start-up time, the logic waits during this time and starts the conversion on the enabled channels. When all conversions are complete, the ADC is deactivated until the next trigger. Triggers occurring during the sequence are not taken into account. The conversion sequencer allows automatic processing with minimum processor intervention and optimized power consumption. Conversion sequences can be performed periodically using a Timer/Counter output. The periodic acquisition of several samples can be processed automatically without any intervention of the processor thanks to the Peripheral DMA Controller. Note: The reference voltage pins always remain connected in normal mode as in sleep mode. 552 32059I-06/2010 AT32UC3B 25.6.7 ADC Timings Each ADC has its own minimal startup time that is defined through the Start Up Time field in the Mode Register (MR.STARTUP). This startup time is given in the Electrical Characteristics chapter. In the same way, a minimal sample and hold time is necessary for the ADC to guarantee the best converted final value between two channels selection. This time has to be defined through the Sample and Hold Time field in the Mode Register (MR.SHTIM). This time depends on the input impedance of the analog input, but also on the output impedance of the driver providing the signal to the analog input, as there is no input buffer amplifier. 25.6.8 Conversion Performances For performance and electrical characteristics of the ADC, see the Electrical Characteristics chapter. 553 32059I-06/2010 AT32UC3B 25.7 User Interface ADC Register Memory Map Register Control Register Mode Register Channel Enable Register Channel Disable Register Channel Status Register Status Register Last Converted Data Register Interrupt Enable Register Interrupt Disable Register Interrupt Mask Register Channel Data Register 0 ...(if implemented) Channel Data Register 7(if implemented) Version Register Name CR MR CHER CHDR CHSR SR LCDR IER IDR IMR CDR0 ... CDR7 VERSION Access Write-only Read/Write Write-only Write-only Read-only Read-only Read-only Write-only Write-only Read-only Read-only ... Read-only Read-only Reset State 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x000C0000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 ... 0x00000000 - (1) Table 25-2. Offset 0x00 0x04 0x10 0x14 0x18 0x1C 0x20 0x24 0x28 0x2C 0x30 ... 0x4C 0xFC Note: 1. The reset values are device specific. Please refer to the Module Configuration section at the end of this chapter. 554 32059I-06/2010 AT32UC3B 25.7.1 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 - Control Register CR Write-only 0x00 0x00000000 30 - 22 - 14 - 6 - 29 - 21 - 13 - 5 - 28 - 20 - 12 - 4 - 27 - 19 - 11 - 3 - 26 - 18 - 10 - 2 - 25 - 17 - 9 - 1 START 24 - 16 - 8 - 0 SWRST * START: Start Conversion Writing a one to this bit will begin an analog-to-digital conversion. Writing a zero to this bit has no effect. This bit always reads zero. * SWRST: Software Reset Writing a one to this bit will reset the ADC. Writing a zero to this bit has no effect. This bit always reads zero. 555 32059I-06/2010 AT32UC3B 25.7.2 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 Mode Register MR Read/Write 0x04 0x00000000 30 - 22 29 - 21 28 - 20 27 26 SHTIM 19 STARTUP 11 PRESCAL 7 - 6 - 5 SLEEP 4 LOWRES 3 2 TRGSEL 1 0 TRGEN 18 17 16 25 24 14 13 12 10 9 8 * SHTIM: Sample & Hold Time Sample & Hold Time = (SHTIM+3) / ADCClock * STARTUP: Start Up Time Startup Time = (STARTUP+1) * 8 / ADCClock * PRESCAL: Prescaler Rate Selection ADCClock = CLK_ADC / ( (PRESCAL+1) * 2 ) * SLEEP: Sleep Mode 1: Sleep Mode is selected. 0: Normal Mode is selected. * LOWRES: Resolution 1: 8-bit resolution is selected. 0: 10-bit resolution is selected. * TRGSEL: Trigger Selection TRGSEL 0 0 0 0 1 1 1 0 0 1 1 0 0 1 0 1 0 1 0 1 0 Selected TRGSEL Internal Trigger 0, depending of chip integration Internal Trigger 1, depending of chip integration Internal Trigger 2, depending of chip integration Internal Trigger 3, depending of chip integration Internal Trigger 4, depending of chip integration Internal Trigger 5, depending of chip integration External trigger * TRGEN: Trigger Enable 1: The hardware trigger selected by the TRGSEL field is enabled. 0: The hardware triggers are disabled. Starting a conversion is only possible by software. 556 32059I-06/2010 AT32UC3B 25.7.3 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 CH7 Channel Enable Register CHER Write-only 0x10 0x00000000 30 - 22 - 14 - 6 CH6 29 - 21 - 13 - 5 CH5 28 - 20 - 12 - 4 CH4 27 - 19 - 11 - 3 CH3 26 - 18 - 10 - 2 CH2 25 - 17 - 9 - 1 CH1 24 - 16 - 8 - 0 CH0 * CHn: Channel n Enable Writing a one to these bits will set the corresponding bit in CHSR. Writing a zero to these bits has no effect. These bits always read a zero. 557 32059I-06/2010 AT32UC3B 25.7.4 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 CH7 Channel Disable Register CHDR Write-only 0x14 0x00000000 30 - 22 - 14 - 6 CH6 29 - 21 - 13 - 5 CH5 28 - 20 - 12 - 4 CH4 27 - 19 - 11 - 3 CH3 26 - 18 - 10 - 2 CH2 25 - 17 - 9 - 1 CH1 24 - 16 - 8 - 0 CH0 * CHn: Channel n Disable Writing a one to these bits will clear the corresponding bit in CHSR. Writing a zero to these bits has no effect. These bits always read a zero. Warning: If the corresponding channel is disabled during a conversion or if it is disabled then reenabled during a conversion, its associated data and its corresponding EOC and OVRE flags in SR are unpredictable. 558 32059I-06/2010 AT32UC3B 25.7.5 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 CH7 Channel Status Register CHSR Read-only 0x18 0x00000000 30 - 22 - 14 - 6 CH6 29 - 21 - 13 - 5 CH5 28 - 20 - 12 - 4 CH4 27 - 19 - 11 - 3 CH3 26 - 18 - 10 - 2 CH2 25 - 17 - 9 - 1 CH1 24 - 16 - 8 - 0 CH0 * CHn: Channel n Status These bits are set when the corresponding bits in CHER is written to one. These bits are cleared when the corresponding bits in CHDR is written to one. 1: The corresponding channel is enabled. 0: The corresponding channel is disabled. 559 32059I-06/2010 AT32UC3B 25.7.6 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 OVRE7 7 EOC7 Status Register SR Read-only 0x1C 0x000C0000 30 - 22 - 14 OVRE6 6 EOC6 29 - 21 - 13 OVRE5 5 EOC5 28 - 20 - 12 OVRE4 4 EOC4 27 - 19 RXBUFF 11 OVRE3 3 EOC3 26 - 18 ENDRX 10 OVRE2 2 EOC2 25 - 17 GOVRE 9 OVRE1 1 EOC1 24 - 16 DRDY 8 OVRE0 0 EOC0 * RXBUFF: RX Buffer Full This bit is set when the Buffer Full signal from the Peripheral DMA is active. This bit is cleared when the Buffer Full signal from the Receive Peripheral DMA is inactive. * ENDRX: End of RX Buffer This bit is set when the End Receive signal from the Peripheral DMA is active. This bit is cleared when the End Receive signal from the Peripheral DMA is inactive. * GOVRE: General Overrun Error This bit is set when a General Overrun Error has occurred. This bit is cleared when the SR register is read. 1: At least one General Overrun Error has occurred since the last read of the SR register. 0: No General Overrun Error occurred since the last read of the SR register. * DRDY: Data Ready This bit is set when a data has been converted and is available in the LCDR register. This bit is cleared when the LCDR register is read. 0: No data has been converted since the last read of the LCDR register. 1: At least one data has been converted and is available in the LCDR register. * OVREn: Overrun Error n These bits are set when an overrun error on the corresponding channel has occurred (if implemented). These bits are cleared when the SR register is read. 0: No overrun error on the corresponding channel (if implemented) since the last read of SR. 1: There has been an overrun error on the corresponding channel (if implemented) since the last read of SR. * EOCn: End of Conversion n These bits are set when the corresponding conversion is complete. These bits are cleared when the corresponding CDR or LCDR registers are read. 0: Corresponding analog channel (if implemented) is disabled, or the conversion is not finished. 1: Corresponding analog channel (if implemented) is enabled and conversion is complete. 560 32059I-06/2010 AT32UC3B 25.7.7 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 Last Converted Data Register LCDR Read-only 0x20 0x00000000 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 LDATA[7:0] 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 LDATA[9:8] 1 0 24 - 16 - 8 * LDATA: Last Data Converted The analog-to-digital conversion data is placed into this register at the end of a conversion and remains until a new conversion is completed. 561 32059I-06/2010 AT32UC3B 25.7.8 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 OVRE7 7 EOC7 Interrupt Enable Register IER Write-only 0x24 0x00000000 30 - 22 - 14 OVRE6 6 EOC6 29 - 21 - 13 OVRE5 5 EOC5 28 - 20 - 12 OVRE4 4 EOC4 27 - 19 RXBUFF 11 OVRE3 3 EOC3 26 - 18 ENDRX 10 OVRE2 2 EOC2 25 - 17 GOVRE 9 OVRE1 1 EOC1 24 - 16 DRDY 8 OVRE0 0 EOC0 Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will set the corresponding bit in IMR. 562 32059I-06/2010 AT32UC3B 25.7.9 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 OVRE7 7 EOC7 Interrupt Disable Register IDR Write-only 0x28 0x00000000 30 - 22 - 14 OVRE6 6 EOC6 29 - 21 - 13 OVRE5 5 EOC5 28 - 20 - 12 OVRE4 4 EOC4 27 - 19 RXBUFF 11 OVRE3 3 EOC3 26 - 18 ENDRX 10 OVRE2 2 EOC2 25 - 17 GOVRE 9 OVRE1 1 EOC1 24 - 16 DRDY 8 OVRE0 0 EOC0 Writing a zero to a bit in this register has no effect. Writing a one to a bit in this register will clear the corresponding bit in IMR. 563 32059I-06/2010 AT32UC3B 25.7.10 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 OVRE7 7 EOC7 Interrupt Mask Register IMR Read-only 0x2C 0x00000000 30 - 22 - 14 OVRE6 6 EOC6 29 - 21 - 13 OVRE5 5 EOC5 28 - 20 - 12 OVRE4 4 EOC4 27 - 19 RXBUFF 11 OVRE3 3 EOC3 26 - 18 ENDRX 10 OVRE2 2 EOC2 25 - 17 GOVRE 9 OVRE1 1 EOC1 24 - 16 DRDY 8 OVRE0 0 EOC0 0: The corresponding interrupt is disabled. 1: The corresponding interrupt is enabled. A bit in this register is cleared when the corresponding bit in IDR is written to one. A bit in this register is cleared when the corresponding bit in IER is written to one. 564 32059I-06/2010 AT32UC3B 25.7.11 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 Channel Data Register CDRx Read-only 0x2C-0x4C 0x00000000 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 DATA[7:0] 27 - 19 - 11 - 3 26 - 18 - 10 - 2 25 - 17 - 9 DATA[9:8] 1 0 24 - 16 - 8 * DATA: Converted Data The analog-to-digital conversion data is placed into this register at the end of a conversion and remains until a new conversion is completed. The Convert Data Register (CDR) is only loaded if the corresponding analog channel is enabled. 565 32059I-06/2010 AT32UC3B 25.7.12 Name: Access Type: Offset: Reset Value: 31 - 23 - 15 - 7 Version Register VERSION Read-only 0xFC - 30 - 22 - 14 - 6 29 - 21 - 13 - 5 28 - 20 - 12 - 4 VERSION[7:0] 27 - 19 26 - 18 VARIANT 11 10 9 VERSION[11:8] 2 1 8 25 - 17 24 - 16 3 0 * VARIANT: Variant Number Reserved. No functionality associated. * VERSION: Version Number Version number of the module. No functionality associated. 566 32059I-06/2010 AT32UC3B 25.8 Module Configuration The specific configuration for the ADC instance is listed in the following tables. Table 25-3. Feature Module configuration ADC 8 TIOA Ouput A of the Timer Counter Channel 0 TIOB Ouput B of the Timer Counter Channel 0 TIOA Ouput A of the Timer Counter Channel 1 TIOB Ouput B of the Timer Counter Channel 1 TIOA Ouput A of the Timer Counter Channel 2 TIOB Ouput B of the Timer Counter Channel 2 Number of Channels Internal Trigger 0 Internal Trigger 1 Internal Trigger 2 Internal Trigger 3 Internal Trigger 4 Internal Trigger 5 Table 25-4. Module name ADC Module Clock Name Clock name CLK_ADC Table 25-5. Module name VERSION Register Reset Values Reset Value 0x00000200 567 32059I-06/2010 AT32UC3B 26. Audio Bitstream DAC (ABDAC) Rev: 1.0.1.1 26.1 Features * Digital Stereo DAC * Oversampled D/A conversion architecture - Oversampling ratio fixed 128x - FIR equalization filter - Digital interpolation filter: Comb4 - 3rd Order Sigma-Delta D/A converters * Digital bitstream outputs * Parallel interface * Connected to DMA Controller for background transfer without CPU intervention 26.2 Overview The Audio Bitstream DAC converts a 16-bit sample value to a digital bitstream with an average value proportional to the sample value. Two channels are supported, making the Audio Bitstream DAC particularly suitable for stereo audio. Each channel has a pair of complementary digital outputs, DATAn and DATANn, which can be connected to an external high input impedance amplifier. The output DATAn and DATANn should be as ideal as possible before filtering, to achieve the best SNR and THD quality. The outputs can be connected to a class D amplifier output stage to drive a speaker directly, or it can be low pass filtered and connected to a high input impedance amplifier. A simple 1st order low pass filter that filters all the frequencies above 50kHz should be adequate when applying the signal to a speaker or a bandlimited amplifier, as the speaker or amplifier will act as a filter and remove high frequency components from the signal. In some cases high frequency components might be folded down into the audible range, and in that case a higher order filter is required. For performance measurements on digital equipment a minimum of 4th order low pass filter should be used. This is to prevent aliasing in the measurements. For the best performance when not using a class D amplifier approach, the two outputs DATAn and DATANn, should be applied to a differential stage amplifier, as this will increase the SNR and THD. 568 32059I-06/2010 AT32UC3B 26.3 Block Diagram Figure 26-1. ABDAC Block Diagram Audio Bitstream DAC PM GCLK_ABDAC Clock Generator sample_clk bit_clk CHANNEL0[15:0] Equalization FIR COMB (INT=128) COMB (INT=128) Sigma-Delta DA-MOD Sigma-Delta DA-MOD DATA0 User Interface CHANNEL1[15:0] Equalization FIR DATA1 26.4 I/O Lines Description I/O Lines Description Pin Description Output from Audio Bitstream DAC Channel 0 Output from Audio Bitstream DAC Channel 1 Inverted output from Audio Bitstream DAC Channel 0 Inverted output from Audio Bitstream DAC Channel 1 Type Output Output Output Output Table 26-1. Pin Name DATA0 DATA1 DATAN0 DATAN1 26.5 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 26.5.1 I/O Lines The output pins used for the output bitstream from the Audio Bitstream DAC may be multiplexed with IO lines. Before using the Audio Bitstream DAC, the I/O Controller must be configured in order for the Audio Bitstream DAC I/O lines to be in Audio Bitstream DAC peripheral mode. 569 32059I-06/2010 AT32UC3B 26.5.2 Clocks The CLK_ABDAC to the Audio Bitstream DAC is generated by the Power Manager (PM). Before using the Audio Bitstream DAC, the user must ensure that the Audio Bitstream DAC clock is enabled in the Power Manager. The ABDAC needs a separate clock for the D/A conversion operation. This clock, GCLK_ABDAC should be set up in the Generic Clock register in the Power Manager and its frequency must be as follow: f GCLK = 256 x f S where fs is the samping rate of the data stream to convert. For fs= 48 kHz this means that the GCLK_ABDAC clock must have a frequency of 12.288MHz. The two clocks, CLK_ABDAC and GCLK_ABDAC, must be in phase with each other. 26.5.3 Interrupts The ABDAC interrupt request line is connected to the interrupt controller. Using the ABDAC interrupt requires the interrupt controller to be programmed first. 26.6 26.6.1 Functional Description How to Initialize the Module In order to use the Audio Bitstream DAC the product dependencies given in Section 26.5 on page 569 must be resolved. Particular attention should be given to the configuration of clocks and I/O lines in order to ensure correct operation of the Audio Bitstream DAC. The Audio Bitstream DAC is enabled by writing a one to the enable bit in the Audio Bitstream DAC Control Register (CR.EN). The Transmit Ready Interrupt Status bit in the Interrupt Status Register (ISR.TXREADY) will be set whenever the ABDAC is ready to receive a new sample. A new sample value should be written to SDR before 256 ABDAC clock cycles, or an underrun will occur, as indicated by the Underrun Interrupt Status bit in ISR (ISR.UNDERRUN). ISR is cleared when read, or when writing one to the corresponding bits in the Interrupt Clear Register (ICR). 26.6.2 Data Format The input data format is two's complement. Two 16-bit sample values for channel 0 and 1 can be written to the least and most significant halfword of the Sample Data Register (SDR), respectively. An input value of 0x7FFF will result in an output voltage of approximately: 3838V OUT ( 0x7FFF ) --------- VDDIO = --------- 3, 3 0, 98V 128 128 An Input value of 0x8000 will result in an output value of approximately: 9090V OUT ( 0x8000 ) --------- VDDIO = --------- 3, 3 2, 32V 128 128 570 32059I-06/2010 AT32UC3B If one want to get coherence between the sign of the input data and the output voltage one can use the DATAN signal or invert the sign of the input data by software. 26.6.3 Data Swapping When the SWAP bit in the ABDAC Control Register (CR.SWAP) is written to one, writing to the Sample Data Register (SDR) will cause the values written to the CHANNEL0 and CHANNEL1 fields to be swapped. 26.6.4 Peripheral DMA Controller The Audio Bitstream DAC is connected to the Peripheral DMA Controller. The Peripheral DMA Controller can be programmed to automatically transfer samples to the Audio Bitstream DAC Sample Data Register (SDR) when the Audio Bitstream DAC is ready for new samples. In this case only the CR.EN bit needs to be set in the Audio Bitstream DAC module. This enables the Audio Bitstream DAC to operate without any CPU intervention such as polling the Interrupt Status Register (ISR) or using interrupts. See the Peripheral DMA Controller documentation for details on how to setup Peripheral DMA transfers. 26.6.5 Construction The Audio Bitstream DAC is constructed of two 3rd order Sigma-Delta D/A converter with an oversampling ratio of 128. The samples are upsampled with a 4th order Sinc interpolation filter (Comb4) before being applied to the Sigma-Delta Modulator. In order to compensate for the pass band frequency response of the interpolation filter and flatten the overall frequency response, the input to the interpolation filter is first filtered with a simple 3-tap FIR filter.The total frequency response of the Equalization FIR filter and the interpolation filter is given in Figure 262 on page 572. The digital output bitstreams from the Sigma-Delta Modulators should be lowpass filtered to remove high frequency noise inserted by the modulation process. 26.6.6 Equalization Filter The equalization filter is a simple 3-tap FIR filter. The purpose of this filter is to compensate for the pass band frequency response of the sinc interpolation filter. The equalization filter makes the pass band response more flat and moves the -3dB corner a little higher. 26.6.7 Interpolation Filter The interpolation filter interpolates from fs to 128fs. This filter is a 4thorder Cascaded IntegratorComb filter, and the basic building blocks of this filter is a comb part and an integrator part. 26.6.8 Sigma-Delta Modulator This part is a 3rdorder Sigma-Delta Modulator consisting of three differentiators (delta blocks), three integrators (sigma blocks) and a one bit quantizer. The purpose of the integrators is to shape the noise, so that the noise is reduced in the band of interest and increased at the higher frequencies, where it can be filtered. 571 32059I-06/2010 AT32UC3B 26.6.9 Frequency Response Figure 26-2. Frequency Response, EQ-FIR+COMB4 10 0 [dB ] A m p li t u d e -1 0 -2 0 -3 0 -4 0 -5 0 -6 0 0 1 2 3 4 5 F re q u e n c y 6 [F s ] 7 8 9 x 10 10 4 572 32059I-06/2010 AT32UC3B 26.7 User Interface ABDAC Register Memory Map Register Sample Data Register Control Register Interrupt Mask Register Interrupt Enable Register Interrupt Disable Register Interrupt Clear Register Interrupt Status Register Register Name SDR CR IMR IER IDR ICR ISR Access Read/Write Read/Write Read-only Write-only Write-only Write-only Read-only Reset 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 0x00000000 Table 26-2. Offset 0x00 0x08 0x0C 0x10 0x14 0x18 0x1C 573 32059I-06/2010 AT32UC3B 26.7.1 Name: Access Type: Offset: Reset Value: 31 Sample Data Register SDR Read/Write 0x00 0x00000000 30 29 28 27 CHANNEL1[15:8] 20 19 CHANNEL1[7:0] 12 11 CHANNEL0[15:8] 4 3 CHANNEL0[7:0] 26 25 24 23 22 21 18 17 16 15 14 13 10 9 8 7 6 5 2 1 0 * CHANNEL1: Sample Data for Channel 1 signed 16-bit Sample Data for channel 1. * CHANNEL0: Signed 16-bit Sample Data for Channel 0 signed 16-bit Sample Data for channel 0. 574 32059I-06/2010 AT32UC3B 26.7.2 Name: Access Type: Offset: Reset Value: 31 EN 23 15 7 - Control Register CR Read/Write 0x08 0x00000000 30 SWAP 22 14 6 29 21 13 5 28 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 - * EN: Enable Audio Bitstream DAC 1: The module is enabled. 0: The module is disabled. * SWAP: Swap Channels 1: The swap of CHANNEL0 and CHANNEL1 samples is enabled. 0: The swap of CHANNEL0 and CHANNEL1 samples is disabled. 575 32059I-06/2010 AT32UC3B 26.7.3 Name: Access Type: Offset: Reset Value: 31 23 15 7 - Interrupt Mask Register IMR Read-only 0x0C 0x00000000 30 22 14 6 29 TXREADY 21 13 5 28 UNDERRUN 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 - 1: The corresponding interrupt is enabled. 0: The corresponding interrupt is disabled. A bit in this register is set when the corresponding bit in IER is written to one. A bit in this register is cleared when the corresponding bit in IDR is written to one. 576 32059I-06/2010 AT32UC3B 26.7.4 Name: Access Type: Offset: Reset Value: 31 23 15 7 - Interrupt Enable Register IER Write-only 0x10 0x00000000 30 22 14 6 29 TXREADY 21 13 5 28 UNDERRUN 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 - Writing a one to a bit in this register will set the corresponding bit in IMR. Writing a zero to a bit in this register has no effect. 577 32059I-06/2010 AT32UC3B 26.7.5 Name: Access Type: Offset: Reset Value: 31 23 15 7 - Interrupt Disable Register IDR Write-only 0x14 0x00000000 30 22 14 6 29 TXREADY 21 13 5 28 UNDERRUN 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 - Writing a one to a bit in this register will clear the corresponding bit in IMR. Writing a zero to a bit in this register has no effect. 578 32059I-06/2010 AT32UC3B 26.7.6 Name: Access Type: Offset: Reset Value: 31 23 15 7 - Interrupt Clear Register ICR Write-only 0x18 0x00000000 30 22 14 6 29 TXREADY 21 13 5 28 UNDERRUN 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 - Writing a one to a bit in this register will clear the corresponding bit in ISR and the corresponding interrupt request. Writing a zero to a bit in this register has no effect. 579 32059I-06/2010 AT32UC3B 26.7.7 Name: Access Type: Offset: Reset Value: 31 23 15 7 - Interrupt Status Register ISR Read-only 0x1C 0x00000000 30 22 14 6 29 TXREADY 21 13 5 28 UNDERRUN 20 12 4 27 19 11 3 26 18 10 2 25 17 9 1 24 16 8 0 - * TXREADY: TX Ready Interrupt Status This bit is set when the Audio Bitstream DAC is ready to receive a new data in SDR. This bit is cleared when the Audio Bitstream DAC is not ready to receive a new data in SDR. * UNDERRUN: Underrun Interrupt Status This bit is set when at least one Audio Bitstream DAC Underrun has occurred since the last time this bit was cleared (by reset or by writing in ICR). This bit is cleared when no Audio Bitstream DAC Underrun has occurred since the last time this bit was cleared (by reset or by writing in ICR). 580 32059I-06/2010 AT32UC3B 27. Programming and Debugging 27.1 Overview General description of programming and debug features, block diagram and introduction of main concepts. 27.2 Service Access Bus The AVR32 architecture offers a common interface for access to On-Chip Debug, programming, and test functions. These are mapped on a common bus called the Service Access Bus (SAB), which is linked to the JTAG port through a bus master module, which also handles synchronization between the debugger and SAB clocks. When accessing the SAB through the debugger there are no limitations on debugger frequency compared to chip frequency, although there must be an active system clock in order for the SAB accesses to complete. If the system clock is switched off in sleep mode, activity on the debugger will restart the system clock automatically, without waking the device from sleep. Debuggers may optimize the transfer rate by adjusting the frequency in relation to the system clock. This ratio can be measured with debug protocol specific instructions. The Service Access Bus uses 36 address bits to address memory or registers in any of the slaves on the bus. The bus supports sized accesses of bytes (8 bits), halfwords (16 bits), or words (32 bits). All accesses must be aligned to the size of the access, i.e. halfword accesses must have the lowest address bit cleared, and word accesses must have the two lowest address bits cleared. 27.2.1 SAB address map The Service Access Bus (SAB) gives the user access to the internal address space and other features through a 36 bits address space. The 4 MSBs identify the slave number, while the 32 LSBs are decoded within the slave's address space. The SAB slaves are shown in Table 27-1 on page 581. Table 27-1. Slave SAB Slaves, addresses and descriptions. Address [35:32] 0x0 0x1 0x4 0x5 0x6 Other Description Intentionally unallocated OCD registers HSB memory space, as seen by the CPU Alternative mapping for HSB space, for compatibility with other 32-bit AVR devices. Memory Service Unit registers Unused Unallocated OCD HSB HSB Memory Service Unit Reserved 27.2.2 SAB security restrictions The Service Access bus can be restricted by internal security measures. A short description of the security measures are found in the table below. 581 32059I-06/2010 AT32UC3B 27.2.2.1 Security measure and control location A security measure is a mechanism to either block or allow SAB access to a certain address or address range. A security measure is enabled or disabled by one or several control signals. This is called the control location for the security measure. These security measures can be used to prevent an end user from reading out the code programmed in the flash, for instance. Table 27-2. SAB Security measures. Control Location FLASHC security bit set FLASHC UPROT + security bit set Description Programming and debugging not possible, very restricted access. Restricts all access except parts of the flash and the flash controller for programming user code. Debugging is not possible unless an OS running from the secure part of the flash supports it. Security measure Security bit User code programming Below follows a more in depth description of what locations are accessible when the security measures are active. Table 27-3. Name Security bit SAB restrictions Address start 0x100000110 0x580800000 Address end 0x100000118 0x581000000 Access Read/Write Read Blocked OCD DCCPU, OCD DCEMU, OCD DCSR User page Other accesses Table 27-4. Name User code programming SAB restrictions Address start 0x100000110 0x580800000 0x5FFFE0000 0x580000000 + BOOTPROT size Address end 0x100000118 0x581000000 0x5FFFE0400 Access Read/Write Read Read/Write OCD DCCPU, OCD DCEMU, OCD DCSR User page FLASHC PB interface FLASH pages outside BOOTPROT Other accesses 0x580000000 + Flash size - Read/Write Blocked 582 32059I-06/2010 AT32UC3B 27.3 On-Chip Debug (OCD) Rev: 1.4.3.1 27.3.1 Features * * * * * * * * Debug interface in compliance with IEEE-ISTO 5001-2003 (Nexus 2.0) Class 2+ JTAG access to all on-chip debug functions Advanced program, data, ownership, and watchpoint trace supported NanoTrace JTAG-based trace access Auxiliary port for high-speed trace information Hardware support for 6 program and 2 data breakpoints Unlimited number of software breakpoints supported Automatic CRC check of memory regions 27.3.2 Overview Debugging on the AT32UC3B is facilitated by a powerful On-Chip Debug (OCD) system. The user accesses this through an external debug tool which connects to the JTAG port and the Auxiliary (AUX) port. The AUX port is primarily used for trace functions, and a JTAG-based debugger is sufficient for basic debugging. The debug system is based on the Nexus 2.0 standard, class 2+, which includes: * Basic run-time control * Program breakpoints * Data breakpoints * Program trace * Ownership trace * Data trace In addition to the mandatory Nexus debug features, the AT32UC3B implements several useful OCD features, such as: * Debug Communication Channel between CPU and JTAG * Run-time PC monitoring * CRC checking * NanoTrace * Software Quality Assurance (SQA) support The OCD features are controlled by OCD registers, which can be accessed by JTAG when the NEXUS_ACCESS JTAG instruction is loaded. The CPU can also access OCD registers directly using mtdr/mfdr instructions in any privileged mode. The OCD registers are implemented based on the recommendations in the Nexus 2.0 standard, and are detailed in the AVR32UC Technical Reference Manual. 583 32059I-06/2010 AT32UC3B 27.3.3 Block Diagram Figure 27-1. On-Chip Debug Block Diagram JTAG JTAG AUX On-Chip Debug Service Access Bus Memory Service Unit Transmit Queue Watchpoints Debug PC Debug Instruction Breakpoints Program Trace Data Trace Ownership Trace CPU Internal SRAM HSB Bus Matrix Memories and peripherals 27.3.4 JTAG-based Debug Features A debugger can control all OCD features by writing OCD registers over the JTAG interface. Many of these do not depend on output on the AUX port, allowing a JTAG-based debugger to be used. A JTAG-based debugger should connect to the device through a standard 10-pin IDC connector as described in the AVR32UC Technical Reference Manual. 584 32059I-06/2010 AT32UC3B Figure 27-2. JTAG-based Debugger PC JTAG-based debug tool 10-pin IDC JTAG AVR32 27.3.4.1 Debug Communication Channel The Debug Communication Channel (DCC) consists of a pair OCD registers with associated handshake logic, accessible to both CPU and JTAG. The registers can be used to exchange data between the CPU and the JTAG master, both runtime as well as in debug mode. 27.3.4.2 breakpoints One of the most fundamental debug features is the ability to halt the CPU, to examine registers and the state of the system. This is accomplished by breakpoints, of which many types are available: * Unconditional breakpoints are set by writing OCD registers by JTAG, halting the CPU immediately. * Program breakpoints halt the CPU when a specific address in the program is executed. * Data breakpoints halt the CPU when a specific memory address is read or written, allowing variables to be watched. * Software breakpoints halt the CPU when the breakpoint instruction is executed. When a breakpoint triggers, the CPU enters debug mode, and the D bit in the Status Register is set. This is a privileged mode with dedicated return address and return status registers. All privileged instructions are permitted. Debug mode can be entered as either OCD mode, running instructions from JTAG, or monitor mode, running instructions from program memory. 585 32059I-06/2010 AT32UC3B 27.3.4.3 OCD mode When a breakpoint triggers, the CPU enters OCD mode, and instructions are fetched from the Debug Instruction OCD register. Each time this register is written by JTAG, the instruction is executed, allowing the JTAG to execute CPU instructions directly. The JTAG master can e.g. read out the register file by issuing mtdr instructions to the CPU, writing each register to the Debug Communication Channel OCD registers. 27.3.4.4 monitor mode Since the OCD registers are directly accessible by the CPU, it is possible to build a softwarebased debugger that runs on the CPU itself. Setting the Monitor Mode bit in the Development Control register causes the CPU to enter monitor mode instead of OCD mode when a breakpoint triggers. Monitor mode is similar to OCD mode, except that instructions are fetched from the debug exception vector in regular program memory, instead of issued by JTAG. 27.3.4.5 program counter monitoring Normally, the CPU would need to be halted for a JTAG-based debugger to examine the current PC value. However, the AT32UC3B provides a Debug Program Counter OCD register, where the debugger can continuously read the current PC without affecting the CPU. This allows the debugger to generate a simple statistic of the time spent in various areas of the code, easing code optimization. 27.3.5 Memory Service Unit The Memory Service Unit (MSU) is a block dedicated to test and debug functionality. It is controlled through a dedicated set of registers addressed through the MEMORY_SERVICE JTAG command. 27.3.5.1 Cyclic Redundancy Check (CRC) The MSU can be used to automatically calculate the CRC of a block of data in memory. The OCD will then read out each word in the specified memory block and report the CRC32-value in an OCD register. 27.3.5.2 NanoTrace The MSU additionally supports NanoTrace. This is an AVR32-specific feature, in which trace data is output to memory instead of the AUX port. This allows the trace data to be extracted by JTAG MEMORY_ACCESS, enabling trace features for JTAG-based debuggers. The user must write MSU registers to configure the address and size of the memory block to be used for NanoTrace. The NanoTrace buffer can be anywhere in the physical address range, including internal and external RAM, through an EBI, if present. This area may not be used by the application running on the CPU. 27.3.6 AUX-based Debug Features Utilizing the Auxiliary (AUX) port gives access to a wide range of advanced debug features. Of prime importance are the trace features, which allow an external debugger to receive continuous information on the program execution in the CPU. Additionally, Event In and Event Out pins allow external events to be correlated with the program flow. The AUX port contains a number of pins, as shown in Table 27-5 on page 587. These are multiplexed with I/O Controller lines, and must explicitly be enabled by writing OCD registers before 586 32059I-06/2010 AT32UC3B the debug session starts. The AUX port is mapped to two different locations, selectable by OCD Registers, minimizing the chance that the AUX port will need to be shared with an application. Debug tools utilizing the AUX port should connect to the device through a Nexus-compliant Mictor-38 connector, as described in the AVR32UC Technical Reference manual. This connector includes the JTAG signals and the RESET_N pin, giving full access to the programming and debug features in the device. Table 27-5. Signal MCKO MDO[5:0] MSEO[1:0] EVTI_N EVTO_N Auxiliary Port Signals Direction Output Output Output Input Output Description Trace data output clock Trace data output Trace frame control Event In Event Out Figure 27-3. AUX+JTAG based Debugger PC T ra c e b u ffe r A U X +JTA G d e b u g to o l M ic t o r 3 8 AUX h ig h s p e e d JTA G AVR 32 27.3.6.1 trace operation Trace features are enabled by writing OCD registers by JTAG. The OCD extracts the trace information from the CPU, compresses this information and formats it into variable-length messages 587 32059I-06/2010 AT32UC3B according to the Nexus standard. The messages are buffered in a 16-frame transmit queue, and are output on the AUX port one frame at a time. The trace features can be configured to be very selective, to reduce the bandwidth on the AUX port. In case the transmit queue overflows, error messages are produced to indicate loss of data. The transmit queue module can optionally be configured to halt the CPU when an overflow occurs, to prevent the loss of messages, at the expense of longer run-time for the program. 27.3.6.2 program trace Program trace allows the debugger to continuously monitor the program execution in the CPU. Program trace messages are generated for every branch in the program, and contains compressed information, which allows the debugger to correlate the message with the source code to identify the branch instruction and target address. 27.3.6.3 data trace Data trace outputs a message every time a specific location is read or written. The message contains information about the type (read/write) and size of the access, as well as the address and data of the accessed location. The AT32UC3B contains two data trace channels, each of which are controlled by a pair of OCD registers which determine the range of addresses (or single address) which should produce data trace messages. 27.3.6.4 ownership trace Program and data trace operate on virtual addresses. In cases where an operating system runs several processes in overlapping virtual memory segments, the Ownership Trace feature can be used to identify the process switch. When the O/S activates a process, it will write the process ID number to an OCD register, which produces an Ownership Trace Message, allowing the debugger to switch context for the subsequent program and data trace messages. As the use of this feature depends on the software running on the CPU, it can also be used to extract other types of information from the system. 27.3.6.5 watchpoint messages The breakpoint modules normally used to generate program and data breakpoints can also be used to generate Watchpoint messages, allowing a debugger to monitor program and data events without halting the CPU. Watchpoints can be enabled independently of breakpoints, so a breakpoint module can optionally halt the CPU when the trigger condition occurs. Data trace modules can also be configured to produce watchpoint messages instead of regular data trace messages. 27.3.6.6 Event In and Event Out pins The AUX port also contains an Event In pin (EVTI_N) and an Event Out pin (EVTO_N). EVTI_N can be used to trigger a breakpoint when an external event occurs. It can also be used to trigger specific program and data trace synchronization messages, allowing an external event to be correlated to the program flow. When the CPU enters debug mode, a Debug Status message is transmitted on the trace port. All trace messages can be timestamped when they are received by the debug tool. However, due to the latency of the transmit queue buffering, the timestamp will not be 100% accurate. To improve this, EVTO_N can toggle every time a message is inserted into the transmit queue, allowing trace messages to be timestamped precisely. EVTO_N can also toggle when a break- 588 32059I-06/2010 AT32UC3B point module triggers, or when the CPU enters debug mode, for any reason. This can be used to measure precisely when the respective internal event occurs. 27.3.6.7 Software Quality Analysis (SQA) Software Quality Analysis (SQA) deals with two important issues regarding embedded software development. Code coverage involves identifying untested parts of the embedded code, to improve test procedures and thus the quality of the released software. Performance analysis allows the developer to precisely quantify the time spent in various parts of the code, allowing bottlenecks to be identified and optimized. Program trace must be used to accomplish these tasks without instrumenting (altering) the code to be examined. However, traditional program trace cannot reconstruct the current PC value without correlating the trace information with the source code, which cannot be done on-the-fly. This limits program trace to a relatively short time segment, determined by the size of the trace buffer in the debug tool. The OCD system in AT32UC3B extends program trace with SQA capabilities, allowing the debug tool to reconstruct the PC value on-the-fly. Code coverage and performance analysis can thus be reported for an unlimited execution sequence. 589 32059I-06/2010 AT32UC3B 27.4 JTAG and Boundary-Scan (JTAG) Rev: 2.0.1.4 27.4.1 Features * IEEE1149.1 compliant JTAG Interface * Boundary-Scan Chain for board-level testing * Direct memory access and programming capabilities through JTAG Interface 27.4.2 Overview The JTAG Interface offers a four pin programming and debug solution, including boundary-scan support for board-level testing. Figure 27-4 on page 591 shows how the JTAG is connected in an 32-bit AVR device. The TAP Controller is a state machine controlled by the TCK and TMS signals. The TAP Controller selects either the JTAG Instruction Register or one of several Data Registers as the scan chain (shift register) between the TDI-input and TDO-output. The Instruction Register holds JTAG instructions controlling the behavior of a Data Register. The Device Identification Register, Bypass Register, and the boundary-scan chain are the Data Registers used for board-level testing. The Reset Register can be used to keep the device reset during test or programming. The Service Access Bus (SAB) interface contains address and data registers for the Service Access Bus, which gives access to On-Chip Debug, programming, and other functions in the device. The SAB offers several modes of access to the address and data registers, as described in Section 27.4.11. Section 27.5 lists the supported JTAG instructions, with references to the description in this document. 590 32059I-06/2010 AT32UC3B 27.4.3 Block Diagram Figure 27-4. JTAG and Boundary-Scan Access 32-bit AVR device JTAG JTAG master TMS TCK TDO TDI TAP Controller Boundary scan enable JTAG Pins TCK TMS TDI TDO Instruction register scan enable Data register scan enable Instruction Register TMS TCK TDO TDI JTAG data registers 2nd JTAG device Device Identification Register By-pass Register Pins and analog blocks Boundary Scan Chain Internal I/O lines Reset Register Part specific registers ... Service Access Bus interface SAB 27.4.4 I/O Lines Description I/O Line Description Pin Description Test Clock Input. Fully asynchronous to system clock frequency. Test Mode Select, sampled on rising TCK. Test Data In, sampled on rising TCK. Test Data Out, driven on falling TCK. Type Input Input Input Output Active Level Table 27-6. Pin Name TCK TMS TDI TDO 27.4.5 Product Dependencies In order to use this module, other parts of the system must be configured correctly, as described below. 591 32059I-06/2010 AT32UC3B 27.4.5.1 I/O Lines The TMS, TDI, and TDO pins are multiplexed with I/O lines. When the JTAG is used the associated pins must be enabled. To enable the JTAG pins, refer to Section 27.4.7. While using the multiplexed JTAG lines all normal peripheral activity on these lines is disabled. The user must make sure that no external peripheral is blocking the JTAG lines while debugging. 27.4.5.2 Power Management When an instruction that accesses the SAB is loaded in the instruction register, before entering a sleep mode, the system clocks are not switched off to allow debugging in sleep modes. This can lead to a program behaving differently when debugging. 27.4.5.3 Clocks The JTAG Interface uses the external TCK pin as clock source. This clock must be provided by the JTAG master. Instructions that use the SAB bus requires the internal main clock to be running. 27.4.6 JTAG Interface The JTAG Interface is accessed through the dedicated JTAG pins shown in Table 27-6 on page 591. The TMS control line navigates the TAP controller, as shown in Figure 27-5 on page 593. The TAP controller manages the serial access to the JTAG Instruction and Data registers. Data is scanned into the selected instruction or data register on TDI, and out of the register on TDO, in the Shift-IR and Shift-DR states, respectively. The LSB is shifted in and out first. TDO is highZ in other states than Shift-IR and Shift-DR. The device implements a 5-bit Instruction Register (IR). A number of public JTAG instructions defined by the JTAG standard are supported, as described in Section 27.5.2, as well as a number of 32-bit AVR-specific private JTAG instructions described in Section 27.5.3. Each instruction selects a specific data register for the Shift-DR path, as described for each instruction. 592 32059I-06/2010 AT32UC3B Figure 27-5. TAP Controller State Diagram 1 Test-LogicReset 0 1 1 Select-DR Scan 0 Capture-DR 0 Shift-DR 1 Exit1-DR 0 Pause-DR 1 0 Exit2-DR 1 1 Update-DR 0 1 0 1 0 0 1 Select-IR Scan 0 Capture-IR 0 Shift-IR 1 Exit1-IR 0 Pause-IR 1 Exit2-IR 1 Update-IR 0 0 1 0 0 Run-Test/ Idle 1 1 593 32059I-06/2010 AT32UC3B 27.4.7 How to Initialize the Module To enable the TMS, TDI and TDO pins one clock pulse should be applied on TCK. Independent of the initial state of the TAP Controller, the Test-Logic-Reset state can always be entered by holding TMS high for 5 TCK clock periods. This sequence should always be applied at the start of a JTAG session and after enabling the JTAG pins to bring the TAP Controller into a defined state before applying JTAG commands. Applying a 0 on TMS for 1 TCK period brings the TAP Controller to the Run-Test/Idle state, which is the starting point for JTAG operations. 27.4.8 How to disable the module To disable the TMS, TDI, and TDO pins the RESET_N pin must be pulled low. 27.4.9 Typical Sequence Assuming Run-Test/Idle is the present state, a typical scenario for using the JTAG Interface follows. 27.4.9.1 Scanning in JTAG Instruction At the TMS input, apply the sequence 1, 1, 0, 0 at the rising edges of TCK to enter the Shift Instruction Register (Shift-IR) state. While in this state, shift the 5 bits of the JTAG instructions into the JTAG instruction register from the TDI input at the rising edge of TCK. During shifting, the JTAG outputs status bits on TDO, refer to Section 27.5 for a description of these. The TMS input must be held low during input of the 4 LSBs in order to remain in the Shift-IR state. The JTAG Instruction selects a particular Data Register as path between TDI and TDO and controls the circuitry surrounding the selected Data Register. Apply the TMS sequence 1, 1, 0 to re-enter the Run-Test/Idle state. The instruction is latched onto the parallel output from the shift register path in the Update-IR state. The Exit-IR, Pause-IR, and Exit2-IR states are only used for navigating the state machine. Figure 27-6. Scanning in JTAG Instruction TCK TAP State TMS TDI TDO 27.4.9.2 Scanning in/out Data At the TMS input, apply the sequence 1, 0, 0 at the rising edges of TCK to enter the Shift Data Register (Shift-DR) state. While in this state, upload the selected Data Register (selected by the present JTAG instruction in the JTAG Instruction Register) from the TDI input at the rising edge of TCK. In order to remain in the Shift-DR state, the TMS input must be held low. While the Data TLR RTI SelDR SelIR CapIR ShIR Ex1IR UpdIR RTI Instruction ImplDefined 594 32059I-06/2010 AT32UC3B Register is shifted in from the TDI pin, the parallel inputs to the Data Register captured in the Capture-DR state is shifted out on the TDO pin. Apply the TMS sequence 1, 1, 0 to re-enter the Run-Test/Idle state. If the selected Data Register has a latched parallel-output, the latching takes place in the Update-DR state. The Exit-DR, Pause-DR, and Exit2-DR states are only used for navigating the state machine. As shown in the state diagram, the Run-Test/Idle state need not be entered between selecting JTAG instruction and using Data Registers. 27.4.10 Boundary-Scan The boundary-scan chain has the capability of driving and observing the logic levels on the digital I/O pins, as well as the boundary between digital and analog logic for analog circuitry having off-chip connections. At system level, all ICs having JTAG capabilities are connected serially by the TDI/TDO signals to form a long shift register. An external controller sets up the devices to drive values at their output pins, and observe the input values received from other devices. The controller compares the received data with the expected result. In this way, boundary-scan provides a mechanism for testing interconnections and integrity of components on Printed Circuits Boards by using the 4 TAP signals only. The four IEEE 1149.1 defined mandatory JTAG instructions IDCODE, BYPASS, SAMPLE/PRELOAD, and EXTEST can be used for testing the Printed Circuit Board. Initial scanning of the data register path will show the ID-code of the device, since IDCODE is the default JTAG instruction. It may be desirable to have the 32-bit AVR device in reset during test mode. If not reset, inputs to the device may be determined by the scan operations, and the internal software may be in an undetermined state when exiting the test mode. If needed, the BYPASS instruction can be issued to make the shortest possible scan chain through the device. The device can be set in the reset state either by pulling the external RESETn pin low, or issuing the AVR_RESET instruction with appropriate setting of the Reset Data Register. The EXTEST instruction is used for sampling external pins and loading output pins with data. The data from the output latch will be driven out on the pins as soon as the EXTEST instruction is loaded into the JTAG IR-register. Therefore, the SAMPLE/PRELOAD should also be used for setting initial values to the scan ring, to avoid damaging the board when issuing the EXTEST instruction for the first time. SAMPLE/PRELOAD can also be used for taking a snapshot of the external pins during normal operation of the part. When using the JTAG Interface for boundary-scan, the JTAG TCK clock is independent of the internal chip clock. The internal chip clock is not required to run during boundary-scan operations. NOTE: For pins connected to 5V lines care should be taken to not drive the pins to a logic one using boundary-scan, as this will create a current flowing from the 3,3V driver to the 5V pull-up on the line. Optionally a series resistor can be added between the line and the pin to reduce the current. Details about the boundary-scan chain can be found in the BSDL file for the device. This can be found on the Atmel website. 27.4.11 Service Access Bus The AVR32 architecture offers a common interface for access to On-Chip Debug, programming, and test functions. These are mapped on a common bus called the Service Access Bus (SAB), 595 32059I-06/2010 AT32UC3B which is linked to the JTAG through a bus master module, which also handles synchronization between the TCK and SAB clocks. For more information about the SAB and a list of SAB slaves see the Service Access Bus chapter. 27.4.11.1 SAB Address Mode The MEMORY_SIZED_ACCESS instruction allows a sized read or write to any 36-bit address on the bus. MEMORY_WORD_ACCESS is a shorthand instruction for 32-bit accesses to any 36-bit address, while the NEXUS_ACCESS instruction is a Nexus-compliant shorthand instruction for accessing the 32-bit OCD registers in the 7-bit address space reserved for these. These instructions require two passes through the Shift-DR TAP state: one for the address and control information, and one for data. 27.4.11.2 Block Transfer To increase the transfer rate, consecutive memory accesses can be accomplished by the MEMORY_BLOCK_ACCESS instruction, which only requires a single pass through Shift-DR for data transfer only. The address is automatically incremented according to the size of the last SAB transfer. 27.4.11.3 Canceling a SAB Access It is possible to abort an ongoing SAB access by the CANCEL_ACCESS instruction, to avoid hanging the bus due to an extremely slow slave. 27.4.11.4 Busy Reporting As the time taken to perform an access may vary depending on system activity and current chip frequency, all the SAB access JTAG instructions can return a busy indicator. This indicates whether a delay needs to be inserted, or an operation needs to be repeated in order to be successful. If a new access is requested while the SAB is busy, the request is ignored. The SAB becomes busy when: * Entering Update-DR in the address phase of any read operation, e.g., after scanning in a NEXUS_ACCESS address with the read bit set. * Entering Update-DR in the data phase of any write operation, e.g., after scanning in data for a NEXUS_ACCESS write. * Entering Update-DR during a MEMORY_BLOCK_ACCESS. * Entering Update-DR after scanning in a counter value for SYNC. * Entering Update-IR after scanning in a MEMORY_BLOCK_ACCESS if the previous access was a read and data was scanned after scanning the address. The SAB becomes ready again when: * A read or write operation completes. * A SYNC countdown completed. * A operation is cancelled by the CANCEL_ACCESS instruction. What to do if the busy bit is set: * During Shift-IR: The new instruction is selected, but the previous operation has not yet completed and will continue (unless the new instruction is CANCEL_ACCESS). You may 596 32059I-06/2010 AT32UC3B continue shifting the same instruction until the busy bit clears, or start shifting data. If shifting data, you must be prepared that the data shift may also report busy. * During Shift-DR of an address: The new address is ignored. The SAB stays in address mode, so no data must be shifted. Repeat the address until the busy bit clears. * During Shift-DR of read data: The read data is invalid. The SAB stays in data mode. Repeat scanning until the busy bit clears. * During Shift-DR of write data: The write data is ignored. The SAB stays in data mode. Repeat scanning until the busy bit clears. 27.4.11.5 Error Reporting The Service Access Bus may not be able to complete all accesses as requested. This may be because the address is invalid, the addressed area is read-only or cannot handle byte/halfword accesses, or because the chip is set in a protected mode where only limited accesses are allowed. The error bit is updated when an access completes, and is cleared when a new access starts. What to do if the error bit is set: * During Shift-IR: The new instruction is selected. The last operation performed using the old instruction did not complete successfully. * During Shift-DR of an address: The previous operation failed. The new address is accepted. If the read bit is set, a read operation is started. * During Shift-DR of read data: The read operation failed, and the read data is invalid. * During Shift-DR of write data: The previous write operation failed. The new data is accepted and a write operation started. This should only occur during block writes or stream writes. No error can occur between scanning a write address and the following write data. * While polling with CANCEL_ACCESS: The previous access was cancelled. It may or may not have actually completed. * After power-up: The error bit is set after power up, but there has been no previous SAB instruction so this error can be discarded. 27.4.11.6 Protected Reporting A protected status may be reported during Shift-IR or Shift-DR. This indicates that the security bit in the Flash Controller is set and that the chip is locked for access, according to Section 27.5.1. The protected state is reported when: * The Flash Controller is under reset. This can be due to the AVR_RESET command or the RESET_N line. * The Flash Controller has not read the security bit from the flash yet (This will take a a few ms). Happens after the Flash Controller reset has been released. * The security bit in the Flash Controller is set. What to do if the protected bit is set: * Release all active AVR_RESET domains, if any. * Release the RESET_N line. * Wait a few ms for the security bit to clear. It can be set temporarily due to a reset. 597 32059I-06/2010 AT32UC3B * Perform a CHIP_ERASE to clear the security bit. NOTE: This will erase all the contents of the non-volatile memory. 27.5 JTAG Instruction Summary The implemented JTAG instructions in the 32-bit AVR are shown in the table below. Table 27-7. Instruction OPCODE 0x01 0x02 0x03 0x04 0x06 0x0C 0x0F 0x10 0x11 0x12 0x13 0x14 0x15 0x17 0x1C 0x1F Others JTAG Instruction Summary Instruction IDCODE SAMPLE_PRELOAD EXTEST INTEST CLAMP AVR_RESET CHIP_ERASE NEXUS_ACCESS MEMORY_WORD_ACCESS MEMORY_BLOCK_ACCESS CANCEL_ACCESS MEMORY_SERVICE MEMORY_SIZED_ACCESS SYNC HALT BYPASS N/A Description Select the 32-bit Device Identification register as data register. Take a snapshot of external pin values without affecting system operation. Select boundary-scan chain as data register for testing circuitry external to the device. Select boundary-scan chain for internal testing of the device. Bypass device through Bypass register, while driving outputs from boundaryscan register. Apply or remove a static reset to the device Erase the device Select the SAB Address and Data registers as data register for the TAP. The registers are accessed in Nexus mode. Select the SAB Address and Data registers as data register for the TAP. Select the SAB Data register as data register for the TAP. The address is auto-incremented. Cancel an ongoing Nexus or Memory access. Select the SAB Address and Data registers as data register for the TAP. The registers are accessed in Memory Service mode. Select the SAB Address and Data registers as data register for the TAP. Synchronization counter Halt the CPU for safe programming. Bypass this device through the bypass register. Acts as BYPASS 27.5.1 Security Restrictions When the security fuse in the Flash is programmed, the following JTAG instructions are restricted: * NEXUS_ACCESS * MEMORY_WORD_ACCESS * MEMORY_BLOCK_ACCESS * MEMORY_SIZED_ACCESS For description of what memory locations remain accessible, please refer to the SAB address map. Full access to these instructions is re-enabled when the security fuse is erased by the CHIP_ERASE JTAG instruction. 598 32059I-06/2010 AT32UC3B Note that the security bit will read as programmed and block these instructions also if the Flash Controller is statically reset. Other security mechanisms can also restrict these functions. If such mechanisms are present they are listed in the SAB address map section. 27.5.1.1 Notation Table 27-9 on page 599 shows bit patterns to be shifted in a format like "peb01". Each character corresponds to one bit, and eight bits are grouped together for readability. The least significantbit is always shifted first, and the most significant bit shifted last. The symbols used are shown in Table 27-8. Table 27-8. Symbol 0 1 a b d e Symbol Description Description Constant low value - always reads as zero. Constant high value - always reads as one. An address bit - always scanned with the least significant bit first A busy bit. Reads as one if the SAB was busy, or zero if it was not. See Section 27.4.11.4 for details on how the busy reporting works. A data bit - always scanned with the least significant bit first. An error bit. Reads as one if an error occurred, or zero if not. See Section 27.4.11.5 for details on how the error reporting works. The chip protected bit. Some devices may be set in a protected state where access to chip internals are severely restricted. See the documentation for the specific device for details. On devices without this possibility, this bit always reads as zero. A direction bit. Set to one to request a read, set to zero to request a write. A size bit. The size encoding is described where used. A don't care bit. Any value can be shifted in, and output data should be ignored. p r s x In many cases, it is not required to shift all bits through the data register. Bit patterns are shown using the full width of the shift register, but the suggested or required bits are emphasized using bold text. I.e. given the pattern "aaaaaaar xxxxxxxx xxxxxxxx xxxxxxxx xx", the shift register is 34 bits, but the test or debug unit may choose to shift only 8 bits "aaaaaaar". The following describes how to interpret the fields in the instruction description tables: Table 27-9. Instruction Instruction Description Description Shows the bit pattern to shift into IR in the Shift-IR state in order to select this instruction. The pattern is show both in binary and in hexadecimal form for convenience. Example: 10000 (0x10) Shows the bit pattern shifted out of IR in the Shift-IR state when this instruction is active. Example: peb01 IR input value IR output value 599 32059I-06/2010 AT32UC3B Table 27-9. Instruction DR Size Instruction Description (Continued) Description Shows the number of bits in the data register chain when this instruction is active. Example: 34 bits Shows which bit pattern to shift into the data register in the Shift-DR state when this instruction is active. Multiple such lines may exist, e.g., to distinguish between reads and writes. Example: aaaaaaar xxxxxxxx xxxxxxxx xxxxxxxx xx Shows the bit pattern shifted out of the data register in the Shift-DR state when this instruction is active. Multiple such lines may exist, e.g., to distinguish between reads and writes. Example: xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb DR input value DR output value 27.5.2 Public JTAG Instructions The JTAG standard defines a number of public JTAG instructions. These instructions are described in the sections below. 27.5.2.1 IDCODE This instruction selects the 32 bit Device Identification register (DID) as Data Register. The DID register consists of a version number, a device number, and the manufacturer code chosen by JEDEC. This is the default instruction after a JTAG reset. Details about the DID register can be found in the module configuration section at the end of this chapter. Starting in Run-Test/Idle, the Device Identification register is accessed in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 6. In Capture-DR: The IDCODE value is latched into the shift register. 7. In Shift-DR: The IDCODE scan chain is shifted by the TCK input. 8. Return to Run-Test/Idle. Table 27-10. IDCODE Details Instructions IR input value IR output value DR Size DR input value DR output value Details 00001 (0x01) p0001 32 xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx Device Identification Register 27.5.2.2 SAMPLE_PRELOAD This instruction takes a snap-shot of the input/output pins without affecting the system operation, and pre-loading the scan chain without updating the DR-latch. The boundary-scan chain is selected as Data Register. 600 32059I-06/2010 AT32UC3B Starting in Run-Test/Idle, the Device Identification register is accessed in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 6. In Capture-DR: The Data on the external pins are sampled into the boundary-scan chain. 7. In Shift-DR: The boundary-scan chain is shifted by the TCK input. 8. Return to Run-Test/Idle. Table 27-11. SAMPLE_PRELOAD Details Instructions IR input value IR output value DR Size DR input value DR output value Details 00010 (0x02) p0001 Depending on boundary-scan chain, see BSDL-file. Depending on boundary-scan chain, see BSDL-file. Depending on boundary-scan chain, see BSDL-file. 27.5.2.3 EXTEST This instruction selects the boundary-scan chain as Data Register for testing circuitry external to the 32-bit AVR package. The contents of the latched outputs of the boundary-scan chain is driven out as soon as the JTAG IR-register is loaded with the EXTEST instruction. Starting in Run-Test/Idle, the EXTEST instruction is accessed the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. In Update-IR: The data from the boundary-scan chain is applied to the output pins. 5. Return to Run-Test/Idle. 6. Select the DR Scan path. 7. In Capture-DR: The data on the external pins is sampled into the boundary-scan chain. 8. In Shift-DR: The boundary-scan chain is shifted by the TCK input. 9. In Update-DR: The data from the scan chain is applied to the output pins. 10. Return to Run-Test/Idle. Table 27-12. EXTEST Details Instructions IR input value IR output value DR Size DR input value DR output value Details 00011 (0x03) p0001 Depending on boundary-scan chain, see BSDL-file. Depending on boundary-scan chain, see BSDL-file. Depending on boundary-scan chain, see BSDL-file. 601 32059I-06/2010 AT32UC3B 27.5.2.4 INTEST This instruction selects the boundary-scan chain as Data Register for testing internal logic in the device. The logic inputs are determined by the boundary-scan chain, and the logic outputs are captured by the boundary-scan chain. The device output pins are driven from the boundary-scan chain. Starting in Run-Test/Idle, the INTEST instruction is accessed the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. In Update-IR: The data from the boundary-scan chain is applied to the internal logic inputs. 5. Return to Run-Test/Idle. 6. Select the DR Scan path. 7. In Capture-DR: The data on the internal logic is sampled into the boundary-scan chain. 8. In Shift-DR: The boundary-scan chain is shifted by the TCK input. 9. In Update-DR: The data from the boundary-scan chain is applied to internal logic inputs. 10. Return to Run-Test/Idle. Table 27-13. INTEST Details Instructions IR input value IR output value DR Size DR input value DR output value Details 00100 (0x04) p0001 Depending on boundary-scan chain, see BSDL-file. Depending on boundary-scan chain, see BSDL-file. Depending on boundary-scan chain, see BSDL-file. 27.5.2.5 CLAMP This instruction selects the Bypass register as Data Register. The device output pins are driven from the boundary-scan chain. Starting in Run-Test/Idle, the CLAMP instruction is accessed the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. In Update-IR: The data from the boundary-scan chain is applied to the output pins. 5. Return to Run-Test/Idle. 6. Select the DR Scan path. 7. In Capture-DR: A logic `0' is loaded into the Bypass Register. 8. In Shift-DR: Data is scanned from TDI to TDO through the Bypass register. 602 32059I-06/2010 AT32UC3B 9. Return to Run-Test/Idle. Table 27-14. CLAMP Details Instructions IR input value IR output value DR Size DR input value DR output value Details 00110 (0x06) p0001 1 x x 27.5.2.6 BYPASS This instruction selects the 1-bit Bypass Register as Data Register. Starting in Run-Test/Idle, the CLAMP instruction is accessed the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 6. In Capture-DR: A logic `0' is loaded into the Bypass Register. 7. In Shift-DR: Data is scanned from TDI to TDO through the Bypass register. 8. Return to Run-Test/Idle. Table 27-15. BYPASS Details Instructions IR input value IR output value DR Size DR input value DR output value Details 11111 (0x1F) p0001 1 x x 27.5.3 Private JTAG Instructions The 32-bit AVR defines a number of private JTAG instructions, not defined by the JTAG standard. Each instruction is briefly described in text, with details following in table form. 27.5.3.1 NEXUS_ACCESS This instruction allows Nexus-compliant access to the On-Chip Debug registers through the SAB. The 7-bit register index, a read/write control bit, and the 32-bit data is accessed through the JTAG port. The data register is alternately interpreted by the SAB as an address register and a data register. The SAB starts in address mode after the NEXUS_ACCESS instruction is selected, and toggles between address and data mode each time a data scan completes with the busy bit cleared. 603 32059I-06/2010 AT32UC3B NOTE: The polarity of the direction bit is inverse of the Nexus standard. Starting in Run-Test/Idle, OCD registers are accessed in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 6. In Shift-DR: Scan in the direction bit (1=read, 0=write) and the 7-bit address for the OCD register. 7. Go to Update-DR and re-enter Select-DR Scan. 8. In Shift-DR: For a read operation, scan out the contents of the addressed register. For a write operation, scan in the new contents of the register. 9. Return to Run-Test/Idle. For any operation, the full 7 bits of the address must be provided. For write operations, 32 data bits must be provided, or the result will be undefined. For read operations, shifting may be terminated once the required number of bits have been acquired. Table 27-16. NEXUS_ACCESS Details Instructions IR input value IR output value DR Size DR input value (Address phase) DR input value (Data read phase) DR input value (Data write phase) DR output value (Address phase) DR output value (Data read phase) DR output value (Data write phase) Details 10000 (0x10) peb01 34 bits aaaaaaar xxxxxxxx xxxxxxxx xxxxxxxx xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xx dddddddd dddddddd dddddddd dddddddd xx xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb eb dddddddd dddddddd dddddddd dddddddd xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb 27.5.3.2 MEMORY_SERVICE This instruction allows access to registers in an optional Memory Service Unit. The 7-bit register index, a read/write control bit, and the 32-bit data is accessed through the JTAG port. The data register is alternately interpreted by the SAB as an address register and a data register. The SAB starts in address mode after the MEMORY_SERVICE instruction is selected, and toggles between address and data mode each time a data scan completes with the busy bit cleared. Starting in Run-Test/Idle, Memory Service registers are accessed in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 604 32059I-06/2010 AT32UC3B 6. In Shift-DR: Scan in the direction bit (1=read, 0=write) and the 7-bit address for the Memory Service register. 7. Go to Update-DR and re-enter Select-DR Scan. 8. In Shift-DR: For a read operation, scan out the contents of the addressed register. For a write operation, scan in the new contents of the register. 9. Return to Run-Test/Idle. For any operation, the full 7 bits of the address must be provided. For write operations, 32 data bits must be provided, or the result will be undefined. For read operations, shifting may be terminated once the required number of bits have been acquired. Table 27-17. MEMORY_SERVICE Details Instructions IR input value IR output value DR Size DR input value (Address phase) DR input value (Data read phase) DR input value (Data write phase) DR output value (Address phase) DR output value (Data read phase) DR output value (Data write phase) Details 10100 (0x14) peb01 34 bits aaaaaaar xxxxxxxx xxxxxxxx xxxxxxxx xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xx dddddddd dddddddd dddddddd dddddddd xx xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb eb dddddddd dddddddd dddddddd dddddddd xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb 27.5.3.3 MEMORY_SIZED_ACCESS This instruction allows access to the entire Service Access Bus data area. Data is accessed through a 36-bit byte index, a 2-bit size, a direction bit, and 8, 16, or 32 bits of data. Not all units mapped on the SAB bus may support all sizes of accesses, e.g., some may only support word accesses. The data register is alternately interpreted by the SAB as an address register and a data register. The SAB starts in address mode after the MEMORY_SIZED_ACCESS instruction is selected, and toggles between address and data mode each time a data scan completes with the busy bit cleared. 605 32059I-06/2010 AT32UC3B The size field is encoded as i Table 27-18. Table 27-18. Size Field Semantics Size field value Access size Data alignment Address modulo 4 : data alignment 0: dddddddd xxxxxxxx xxxxxxxx xxxxxxxx 1: xxxxxxxx dddddddd xxxxxxxx xxxxxxxx 2: xxxxxxxx xxxxxxxx dddddddd xxxxxxxx 3: xxxxxxxx xxxxxxxx xxxxxxxx dddddddd Address modulo 4 : data alignment 0: dddddddd dddddddd xxxxxxxx xxxxxxxx 1: Not allowed 2: xxxxxxxx xxxxxxxx dddddddd dddddddd 3: Not allowed Address modulo 4 : data alignment 0: dddddddd dddddddd dddddddd dddddddd 1: Not allowed 2: Not allowed 3: Not allowed N/A 00 Byte (8 bits) 01 Halfword (16 bits) 10 Word (32 bits) 11 Reserved Starting in Run-Test/Idle, SAB data is accessed in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 6. In Shift-DR: Scan in the direction bit (1=read, 0=write), 2-bit access size, and the 36-bit address of the data to access. 7. Go to Update-DR and re-enter Select-DR Scan. 8. In Shift-DR: For a read operation, scan out the contents of the addressed area. For a write operation, scan in the new contents of the area. 9. Return to Run-Test/Idle. For any operation, the full 36 bits of the address must be provided. For write operations, 32 data bits must be provided, or the result will be undefined. For read operations, shifting may be terminated once the required number of bits have been acquired. Table 27-19. MEMORY_SIZED_ACCESS Details Instructions IR input value IR output value DR Size DR input value (Address phase) DR input value (Data read phase) DR input value (Data write phase) Details 10101 (0x15) peb01 39 bits aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaassr xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxx dddddddd dddddddd dddddddd dddddddd xxxxxxx 606 32059I-06/2010 AT32UC3B Table 27-19. MEMORY_SIZED_ACCESS Details (Continued) Instructions DR output value (Address phase) DR output value (Data read phase) DR output value (Data write phase) Details xxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb xxxxxeb dddddddd dddddddd dddddddd dddddddd xxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb 27.5.3.4 MEMORY_WORD_ACCESS This instruction allows access to the entire Service Access Bus data area. Data is accessed through the 34 MSB of the SAB address, a direction bit, and 32 bits of data. This instruction is identical to MEMORY_SIZED_ACCESS except that it always does word sized accesses. The size field is implied, and the two lowest address bits are removed and not scanned in. Note: This instruction was previously known as MEMORY_ACCESS, and is provided for backwards compatibility. The data register is alternately interpreted by the SAB as an address register and a data register. The SAB starts in address mode after the MEMORY_WORD_ACCESS instruction is selected, and toggles between address and data mode each time a data scan completes with the busy bit cleared. Starting in Run-Test/Idle, SAB data is accessed in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 6. In Shift-DR: Scan in the direction bit (1=read, 0=write) and the 34-bit address of the data to access. 7. Go to Update-DR and re-enter Select-DR Scan. 8. In Shift-DR: For a read operation, scan out the contents of the addressed area. For a write operation, scan in the new contents of the area. 9. Return to Run-Test/Idle. For any operation, the full 34 bits of the address must be provided. For write operations, 32 data bits must be provided, or the result will be undefined. For read operations, shifting may be terminated once the required number of bits have been acquired. Table 27-20. MEMORY_WORD_ACCESS Details Instructions IR input value IR output value DR Size DR input value (Address phase) DR input value (Data read phase) DR input value (Data write phase) Details 10001 (0x11) peb01 35 bits aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aar xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxx dddddddd dddddddd dddddddd dddddddd xxx 607 32059I-06/2010 AT32UC3B Table 27-20. MEMORY_WORD_ACCESS Details (Continued) Instructions DR output value (Address phase) DR output value (Data read phase) DR output value (Data write phase) Details xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xeb xeb dddddddd dddddddd dddddddd dddddddd xxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb 27.5.3.5 MEMORY_BLOCK_ACCESS This instruction allows access to the entire SAB data area. Up to 32 bits of data is accessed at a time, while the address is sequentially incremented from the previously used address. In this mode, the SAB address, size, and access direction is not provided with each access. Instead, the previous address is auto-incremented depending on the specified size and the previous operation repeated. The address must be set up in advance with MEMORY_SIZE_ACCESS or MEMORY_WORD_ACCESS. It is allowed, but not required, to shift data after shifting the address. This instruction is primarily intended to speed up large quantities of sequential word accesses. It is possible to use it also for byte and halfword accesses, but the overhead in this is case much larger as 32 bits must still be shifted for each access. The following sequence should be used: 1. Use the MEMORY_SIZE_ACCESS or MEMORY_WORD_ACCESS to read or write the first location. 2. Return to Run-Test/Idle. 3. Select the IR Scan path. 4. In Capture-IR: The IR output value is latched into the shift register. 5. In Shift-IR: The instruction register is shifted by the TCK input. 6. Return to Run-Test/Idle. 7. Select the DR Scan path. The address will now have incremented by 1, 2, or 4 (corresponding to the next byte, halfword, or word location). 8. In Shift-DR: For a read operation, scan out the contents of the next addressed location. For a write operation, scan in the new contents of the next addressed location. 9. Go to Update-DR. 10. If the block access is not complete, return to Select-DR Scan and repeat the access. 11. If the block access is complete, return to Run-Test/Idle. For write operations, 32 data bits must be provided, or the result will be undefined. For read operations, shifting may be terminated once the required number of bits have been acquired. Table 27-21. MEMORY_BLOCK_ACCESS Details Instructions IR input value IR output value DR Size DR input value (Data read phase) Details 10010 (0x12) peb01 34 bits xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xx 608 32059I-06/2010 AT32UC3B Table 27-21. MEMORY_BLOCK_ACCESS Details (Continued) Instructions DR input value (Data write phase) DR output value (Data read phase) DR output value (Data write phase) Details dddddddd dddddddd dddddddd dddddddd xx eb dddddddd dddddddd dddddddd dddddddd xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxeb The overhead using block word access is 4 cycles per 32 bits of data, resulting in an 88% transfer efficiency, or 2.1 MBytes per second with a 20 MHz TCK frequency. 27.5.3.6 CANCEL_ACCESS If a very slow memory location is accessed during a SAB memory access, it could take a very long time until the busy bit is cleared, and the SAB becomes ready for the next operation. The CANCEL_ACCESS instruction provides a possibility to abort an ongoing transfer and report a timeout to the JTAG master. When the CANCEL_ACCESS instruction is selected, the current access will be terminated as soon as possible. There are no guarantees about how long this will take, as the hardware may not always be able to cancel the access immediately. The SAB is ready to respond to a new command when the busy bit clears. Starting in Run-Test/Idle, CANCEL_ACCESS is accessed in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. Table 27-22. CANCEL_ACCESS Details Instructions IR input value IR output value DR Size DR input value DR output value Details 10011 (0x13) peb01 1 x 0 27.5.3.7 SYNC This instruction allows external debuggers and testers to measure the ratio between the external JTAG clock and the internal system clock. The SYNC data register is a 16-bit counter that counts down to zero using the internal system clock. The busy bit stays high until the counter reaches zero. Starting in Run-Test/Idle, SYNC instruction is used in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 609 32059I-06/2010 AT32UC3B 6. Scan in an 16-bit counter value. 7. Go to Update-DR and re-enter Select-DR Scan. 8. In Shift-DR: Scan out the busy bit, and until the busy bit clears goto 7. 9. Calculate an approximation to the internal clock speed using the elapsed time and the counter value. 10. Return to Run-Test/Idle. The full 16-bit counter value must be provided when starting the synch operation, or the result will be undefined. When reading status, shifting may be terminated once the required number of bits have been acquired. Table 27-23. SYNC_ACCESS Details Instructions IR input value IR output value DR Size DR input value DR output value Details 10111 (0x17) peb01 16 bits dddddddd dddddddd xxxxxxxx xxxxxxeb 27.5.3.8 AVR_RESET This instruction allows a debugger or tester to directly control separate reset domains inside the chip. The shift register contains one bit for each controllable reset domain. Setting a bit to one resets that domain and holds it in reset. Setting a bit to zero releases the reset for that domain. The AVR_RESET instruction can be used in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 6. In Shift-DR: Scan in the value corresponding to the reset domains the JTAG master wants to reset into the data register. 7. Return to Run-Test/Idle. 8. Stay in run test idle for at least 10 TCK clock cycles to let the reset propagate to the system. See the device specific documentation for the number of reset domains, and what these domains are. For any operation, all bits must be provided or the result will be undefined. Table 27-24. AVR_RESET Details Instructions IR input value IR output value Details 01100 (0x0C) p0001 610 32059I-06/2010 AT32UC3B Table 27-24. AVR_RESET Details (Continued) Instructions DR Size DR input value DR output value Details Device specific. Device specific. Device specific. 27.5.3.9 CHIP_ERASE This instruction allows a programmer to completely erase all nonvolatile memories in a chip. This will also clear any security bits that are set, so the device can be accessed normally. In devices without non-volatile memories this instruction does nothing, and appears to complete immediately. The erasing of non-volatile memories starts as soon as the CHIP_ERASE instruction is selected. The CHIP_ERASE instruction selects a 1 bit bypass data register. A chip erase operation should be performed as: 1. Reset the system and stop the CPU from executing. 2. Select the IR Scan path. 3. In Capture-IR: The IR output value is latched into the shift register. 4. In Shift-IR: The instruction register is shifted by the TCK input. 5. Check the busy bit that was scanned out during Shift-IR. If the busy bit was set goto 2. 6. Return to Run-Test/Idle. Table 27-25. CHIP_ERASE Details Instructions IR input value IR output value DR Size DR input value DR output value Details 01111 (0x0F) p0b01 Where b is the busy bit. 1 bit x 0 27.5.3.10 HALT This instruction allows a programmer to easily stop the CPU to ensure that it does not execute invalid code during programming. This instruction selects a 1-bit halt register. Setting this bit to one resets the device and halts the CPU. Setting this bit to zero resets the device and releases the CPU to run normally. The value shifted out from the data register is one if the CPU is halted. The HALT instruction can be used in the following way: 1. Select the IR Scan path. 2. In Capture-IR: The IR output value is latched into the shift register. 3. In Shift-IR: The instruction register is shifted by the TCK input. 4. Return to Run-Test/Idle. 5. Select the DR Scan path. 611 32059I-06/2010 AT32UC3B 6. In Shift-DR: Scan in the value 1 to halt the CPU, 0 to start CPU execution. 7. Return to Run-Test/Idle. Table 27-26. HALT Details Instructions IR input value IR output value DR Size DR input value DR output value Details 11100 (0x1C) p0001 1 bit d d 612 32059I-06/2010 AT32UC3B 27.6 JTAG Data Registers The following device specific registers can be selected as JTAG scan chain depending on the instruction loaded in the JTAG Instruction Register. Additional registers exist, but are implicitly described in the functional description of the relevant instructions. 27.6.1 Device Identification Register The Device Identification Register contains a unique identifier for each product. The register is selected by the IDCODE instruction, which is the default instruction after a JTAG reset. MS B Bit Device ID 31 28 27 Part Number LSB 12 11 Manufacturer ID 1 0 1 Revision 4 bits 16 bits 11 bits 1 bit Revision Part Number Manufacturer ID This is a 4 bit number identifying the revision of the component. Rev A = 0x0, B = 0x1, etc. The part number is a 16 bit code identifying the component. The Manufacturer ID is a 11 bit code identifying the manufacturer. The JTAG manufacturer ID for ATMEL is 0x01F. 27.6.1.1 Device specific ID codes The different device configurations have different JTAG ID codes, as shown in Table 27-27. Note that if the flash controller is statically reset, the ID code will be undefined. Table 27-27. Device and JTAG ID Device name AT32UC3B0512 AT32UC3B1512 AT32UC3B0256 AT32UC3B1256 AT32UC3B0128 AT32UC3B1128 AT32UC3B064 AT32UC3B164 JTAG ID code (r is the revision number) 0xr205003F 0xr205203F 0xr1EE403F 0xr1EE503F 0xr1EE603F 0xr1EE903F 0xr1EEA03F 0xr1EEB03F 613 32059I-06/2010 AT32UC3B 27.6.2 Reset register The reset register is selected by the AVR_RESET instruction and contains one bit for each reset domain in the device. Setting each bit to one will keep that domain reset until the bit is cleared. LSB Bit Device ID 4 OCD 3 APP 2 RESERVED 1 RESERVED 0 CPU CPU APP OCD RSERVED CPU HSB and PB buses On-Chip Debug logic and registers No effect Note: This register is primarily intended for compatibility with other 32-bit AVR devices. Certain operations may not function correctly when parts of the system are reset. It is generally recommended to only write 0x11111 or 0x00000 to these bits to ensure no unintended side effects occur. 27.6.3 Boundary-Scan Chain The Boundary-Scan Chain has the capability of driving and observing the logic levels on the digital I/O pins, as well as driving and observing the logic levels between the digital I/O pins and the internal logic. Typically, output value, output enable, and input data are all available in the boundary scan chain. The boundary scan chain is described in the BDSL (Boundary Scan Description Language) file available at the Atmel web site. 27.7 SAB address map The Service Access Bus (SAB) gives the user access to the internal address space and other features through a 36 bits address space. The 4 MSBs identify the slave number, while the 32 LSBs are decoded within the slave's address space. The SAB slaves are shown in Table 27-28. Table 27-28. SAB Slaves, addresses and descriptions. Slave Unallocated OCD HSB HSB Memory Service Unit Reserved Address [35:32] 0x0 0x1 0x4 0x5 0x6 Other Description Intentionally unallocated OCD registers HSB memory space, as seen by the CPU Alternative mapping for HSB space, for compatibility with other 32-bit AVR devices. Memory Service Unit registers Unused 614 32059I-06/2010 AT32UC3B 28. Electrical Characteristics 28.1 Absolute Maximum Ratings* *NOTICE: Stresses beyond those listed under "Absolute Maximum Ratings" may cause permanent damage to the device. This is a stress rating only and functional operation of the device at these or other conditions beyond those indicated in the operational sections of this specification is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. Operating Temperature.................................... -40C to +85C Storage Temperature ..................................... -60C to +150C Voltage on GPIO Pins with respect to Ground for TCK, RESET_N, PA03, PA04, PA05, PA06, PA07, PA08, PA11, PA12, PA18, PA19, PA28, PA29, PA30, PA31 ............................................................ -0.3 to 3.6V Voltage on GPIO Pins with respect to Ground except for TCK, RESET_N, PA03, PA04, PA05, PA06, PA07, PA08, PA11, PA12, PA18, PA19, PA28, PA29, PA30, PA31 ....................................... -0.3 to 5.5V Maximum Operating Voltage (VDDCORE, VDDPLL) ..... 1.95V Maximum Operating Voltage (VDDIO,VDDIN,VDDANA).. 3.6V Total DC Output Current on all I/O Pin for 48-pin package ....................................................... 200 mA for 64-pin package ....................................................... 265 mA 615 32059I-06/2010 AT32UC3B 28.2 DC Characteristics The following characteristics are applicable to the operating temperature range: TA = -40C to 85C, unless otherwise specified and are certified for a junction temperature up to TJ = 100C. Table 28-1. Symbol VVDDCORE VVDDPLL VVDDIO DC Characteristics Parameter DC Supply Core DC Supply PLL DC Supply Peripheral I/Os AT32UC3B064 AT32UC3B0128 AT32UC3B0256 AT32UC3B164 AT32UC3B1128 AT32UC3B1256 AT32UC3B0512 AT32UC3B1512 Conditions Min. 1.65 1.65 3.0 Typ. Max. 1.95 1.95 3.6 Unit V V V All pins -0.3 +0.8 V VIL Input Low-level Voltage All pins exept Reset_N Reset_N All I/O except TCK, RESET_N, PA03, PA04, PA05, PA06, PA07, PA08, PA11, PA12, PA18, PA19, PA28, PA29, PA30, PA31 TCK, RESET_N, PA03, PA04, PA05, PA06, PA07, PA08, PA11, PA12, PA18, PA19, PA28, PA29, PA30, PA31 All I/O except TCK, RESET_N, PA03, PA04, PA05, PA06, PA07, PA08, PA11, PA12, PA18, PA19, PA28, PA29, PA30, PA31 TCK, RESET_N PA03, PA04, PA05, PA06, PA07, PA08, PA11, PA12, PA18, PA19, PA28, PA29, PA30, PA31 -0.3 -0.3 +0.8 +0.3 V V AT32UC3B064 AT32UC3B0128 AT32UC3B0256 AT32UC3B164 AT32UC3B1128 AT32UC3B1256 2.0 5.5 V 2.0 3.6 V VIH Input High-level Voltage 2.0 5.5 V AT32UC3B0512 AT32UC3B1512 2.5 3.6 V 2.0 3.6 V VOL Output Low-level Voltage IOL= -4mA for all I/O except for PA20, PA21, PA22, PA23 IOL= -8mA for I/O PA20, PA21, PA22, PA23 IOL= -4mA for all I/O except for PA20, PA21, PA22, PA23 VVDDIO -0.4 VVDDIO -0.4 0.4 0.4 V V V V VOH Output High-level Voltage IOL= -8mA for I/O PA20, PA21, PA22, PA23 616 32059I-06/2010 AT32UC3B Table 28-1. Symbol IOL DC Characteristics Parameter Output Low-level Current I/O PA20, PA21, PA22, PA23 Alll I/O except for PA20, PA21, PA22, PA23 Output High-level Current I/O PA20, PA21, PA22, PA23 Input Leakage Current Pullup resistors disabled QFP64 QFP48 8 1 7 7 7 7 All I/O pins except RESET_N, TCK, TDI, TMS pins RESET_N pin, TCK, TDI, TMS pins All I/O pins except PA20, PA21, PA22, PA23, RESET_N, TCK, TDI, TMS pins PA20, PA21, PA22, PA23 RESET_N pin, TCK, TDI, TMS pins AT32UC3B064 AT32UC3B0128 AT32UC3B0256 AT32UC3B164 AT32UC3B1128 AT32UC3B1256 On VVDDCORE = 1.8V, device in static mode All inputs driven including JTAG; RESET_N=1 On VVDDCORE = 1.8V, device in static mode All inputs driven including JTAG; RESET_N=1 TA = 25C 5 13 19 25 mA A pF pF pF pF K -8 4 mA mA Conditions Alll GPIOs except for PA20, PA21, PA22, PA23 Min. Typ. Max. -4 Unit mA IOH ILEAK CIN Input Capacitance QFN64 QFN48 AT32UC3B064 AT32UC3B0128 AT32UC3B0256 AT32UC3B164 AT32UC3B1128 AT32UC3B1256 5 12 25 K RPULLUP Pull-up Resistance 13 15 25 K AT32UC3B0512 AT32UC3B1512 8 10 25 K K 6 A TA = 85C 42.5 A ISC Static Current TA = 25C 7.5 A AT32UC3B0512 AT32UC3B1512 TA = 85C 62.5 A 617 32059I-06/2010 AT32UC3B 28.3 Regulator Characteristics Electrical Characteristics Parameter Supply voltage (input) Supply voltage (output) Maximum DC output current Static Current of internal regulator VVDDIN = 3.3V Low Power mode (stop, deep stop or static) at TA = 25C 10 Conditions Min. 3 1.70 Typ. 3.3 1.8 Max. 3.6 1.85 100 Unit V V mA A Table 28-2. Symbol VVDDIN VVDDOUT IOUT ISCR Table 28-3. Symbol CIN1 CIN2 COUT1 COUT2 Decoupling Requirements Parameter Input Regulator Capacitor 1 Input Regulator Capacitor 2 Output Regulator Capacitor 1 Output Regulator Capacitor 2 Conditions Typ. 1 4.7 470 2.2 Technology NPO X7R NPO X7R Unit nF F pF F 28.4 28.4.1 Analog Characteristics ADC Reference Electrical Characteristics Parameter Analog voltage reference (input) Conditions Min. 2.6 Typ. Max. 3.6 Unit V Table 28-4. Symbol VADVREF Table 28-5. Symbol CVREF1 CVREF2 Decoupling Requirements Parameter Voltage reference Capacitor 1 Voltage reference Capacitor 2 Conditions Typ. 10 1 Technology NPO NPO Unit nF uF 28.4.2 BOD BOD Level Values Parameter Value 00 0000b 01 0111b Conditions Min. Typ. 1.44 1.52 1.61 1.71 Max. Unit V V V V Table 28-6. Symbol BODLEVEL 01 1111b 10 0111b Table 28-6 describes the values of the BODLEVEL field in the flash FGPFR register. 618 32059I-06/2010 AT32UC3B Table 28-7. Symbol TBOD BOD Timing Parameter Minimum time with VDDCORE < VBOD to detect power failure Conditions Falling VDDCORE from 1.8V to 1.1V Min. Typ. 300 Max. 800 Unit ns 28.4.3 Reset Sequence Electrical Characteristics Parameter VDDCORE rise rate to ensure poweron-reset VDDCORE fall rate to ensure poweron-reset Rising threshold voltage: voltage up to which device is kept under reset by POR on rising VDDCORE Falling threshold voltage: voltage when POR resets device on falling VDDCORE On falling VDDCORE, voltage must go down to this value before supply can rise again to ensure reset signal is released at VPOR+ Minimum time with VDDCORE < VPORTime for reset signal to be propagated to system Time for Cold System Startup: Time for CPU to fetch its first instruction (RCosc not calibrated) Time for Hot System Startup: Time for CPU to fetch its first instruction (RCosc calibrated) 480 Rising VDDCORE: VRESTART -> VPOR+ Falling VDDCORE: 1.8V -> VPOR+ Conditions Min. 0.01 0.01 400 Typ. Max. Unit V/ms V/ms Table 28-8. Symbol VDDRR VDDFR VPOR+ 1.4 1.55 1.65 V VPOR- 1.2 1.3 1.4 V VRESTART Falling VDDCORE: 1.8V -> VRESTART Falling VDDCORE: 1.8V -> 1.1V -0.1 0.5 V TPOR TRST 15 200 400 s s TSSU1 960 s TSSU2 420 s 619 32059I-06/2010 AT32UC3B Figure 28-1. MCU Cold Start-Up RESET_N tied to VDDIN VDDCORE VPORVRESTART VPOR+ RESET_N Internal POR Reset TPOR Internal MCU Reset TRST TSSU1 Figure 28-2. MCU Cold Start-Up RESET_N Externally Driven VDDCORE VPORVRESTART VPOR+ RESET_N Internal POR Reset TPOR Internal MCU Reset TRST TSSU1 Figure 28-3. MCU Hot Start-Up VDDCORE RESET_N BOD Reset WDT Reset TSSU2 Internal MCU Reset 28.4.4 RESET_N Characteristics RESET_N Waveform Parameters Parameter RESET_N minimum pulse width Conditions Min. 10 Typ. Max. Unit ns Table 28-9. Symbol tRESET 620 32059I-06/2010 AT32UC3B 28.5 Power Consumption The values in Table 28-10, Table 28-11 on page 622 and Table 28-12 on page 623 are measured values of power consumption with operating conditions as follows: *VDDIO = VDDANA = 3.3V *VDDCORE = VDDPLL = 1.8V *TA = 25C, TA = 85C *I/Os are configured in input, pull-up enabled. Figure 28-4. Measurement Setup VDDANA VDDIO Amp0 VDDIN Internal Voltage Regulator VDDOUT Amp1 VDDCORE VDDPLL The following tables represent the power consumption measured on the power supplies. 621 32059I-06/2010 AT32UC3B 28.5.1 Power Consumtion for Different Sleep Modes Table 28-10. Power Consumption for Different Sleep Modes for AT32UC3B064, AT32UC3B0128, AT32UC3B0256, AT32UC3B164, AT32UC3B1128, AT32UC3B1256 Mode Conditions -CPU running a recursive Fibonacci Algorithm from flash and clocked from PLL0 at f MHz. -Voltage regulator is on. -XIN0 : external clock. Xin1 Stopped. XIN32 stopped. -All peripheral clocks activated with a division by 8. - GPIOs are inactive with internal pull-up, JTAG unconnected with external pullup and Input pins are connected to GND Same conditions at 60 MHz See Active mode conditions Idle Same conditions at 60 MHz See Active mode conditions Frozen Same conditions at 60 MHz See Active mode conditions Standby Same conditions at 60 MHz -CPU running in sleep mode -XIN0, Xin1 and XIN32 are stopped. -All peripheral clocks are desactived. - GPIOs are inactive with internal pull-up, JTAG unconnected with external pullup and Input pins are connected to GND. See Stop mode conditions Voltage Regulator On Static See Stop mode conditions Voltage Regulator Off Notes: 8.9 A 2.7 mA 3.6 0.042xf(MHz)+0.115 mA mA/MHz 7.3 0.058xf(MHz)+0.115 mA mA/MHz Typ. Unit 0.3xf(MHz)+0.443 mA/MHz Active 18.5 0.117xf(MHz)+0.28 mA mA/MHz Stop 37.8 A Deepstop 24.9 13.9 A A 1. Core frequency is generated from XIN0 using the PLL so that 140 MHz < fPLL0 < 160 MHz and 10 MHz < fXIN0 < 12 MHz. Table 28-11. Power Consumption for Different Sleep Modes for AT32UC3B0512, AT32UC3B1512 Mode Conditions -CPU running a recursive Fibonacci Algorithm from flash and clocked from PLL0 at f MHz. -Voltage regulator is on. -XIN0 : external clock. Xin1 Stopped. XIN32 stopped. -All peripheral clocks activated with a division by 8. - GPIOs are inactive with internal pull-up, JTAG unconnected with external pullup and Input pins are connected to GND Same conditions at 60 MHz See Active mode conditions Idle Same conditions at 60 MHz See Active mode conditions Frozen Same conditions at 60 MHz 4.3 mA 9 0.07xf(MHz)+0.094 mA mA/MHz Typ. Unit 0.388xf(MHz)+0.781 mA/MHz Active 24 0.143xf(MHz)+0.248 mA mA/MHz 622 32059I-06/2010 AT32UC3B Table 28-11. Power Consumption for Different Sleep Modes for AT32UC3B0512, AT32UC3B1512 Mode Standby Same conditions at 60 MHz -CPU running in sleep mode -XIN0, Xin1 and XIN32 are stopped. -All peripheral clocks are desactived. - GPIOs are inactive with internal pull-up, JTAG unconnected with external pullup and Input pins are connected to GND. See Stop mode conditions Voltage Regulator On Static Notes: See Stop mode conditions Voltage Regulator Off 7.5 3.2 mA Conditions See Active mode conditions Typ. 0.052xf(MHz)+0.091 Unit mA/MHz Stop 63.2 A Deepstop 30.3 16.5 A A 1. Core frequency is generated from XIN0 using the PLL so that 140 MHz < fPLL0 < 160 MHz and 10 MHz < fXIN0 < 12 MHz. Table 28-12. Peripheral Interface Power Consumption in Active Mode Peripheral INTC GPIO PDCA USART USB ADC TWI PWM SPI SSC TC ABDAC AT32UC3B0512 AT32UC3B1512 AT32UC3B064 AT32UC3B0128 AT32UC3B0256 AT32UC3B164 AT32UC3B1128 AT32UC3B1256 AT32UC3B0512 AT32UC3B1512 Conditions Consumption 20 16 12 14 23 8 7 18 8 11 11 6.5 A/MHz Unit 623 32059I-06/2010 AT32UC3B 28.6 System Clock Characteristics These parameters are given in the following conditions: * VDDCORE = 1.8V * Ambient Temperature = 25C 28.6.1 CPU/HSB Clock Characteristics Table 28-13. Core Clock Waveform Parameters Symbol 1/(tCPCPU) tCPCPU Parameter CPU Clock Frequency CPU Clock Period 16.6 Conditions Min. Typ. Max. 60 Unit MHz ns 28.6.2 PBA Clock Characteristics Table 28-14. PBA Clock Waveform Parameters Symbol 1/(tCPPBA) tCPPBA Parameter PBA Clock Frequency PBA Clock Period 16.6 Conditions Min. Typ. Max. 60 Unit MHz ns 28.6.3 PBB Clock Characteristics Table 28-15. PBB Clock Waveform Parameters Symbol 1/(tCPPBB) tCPPBB Parameter PBB Clock Frequency PBB Clock Period 16.6 Conditions Min. Typ. Max. 60 Unit MHz ns 624 32059I-06/2010 AT32UC3B 28.7 Oscillator Characteristics The following characteristics are applicable to the operating temperature range: TA = -40C to 85C and worst case of power supply, unless otherwise specified. 28.7.1 Slow Clock RC Oscillator Table 28-16. RC Oscillator Frequency Symbol Parameter Conditions Calibration point: TA = 85C FRC RC Oscillator Frequency TA = 25C TA = -40C 105 Min. Typ. 115.2 112 108 Max. 116 Unit KHz KHz KHz 28.7.2 32 KHz Oscillator Table 28-17. 32 KHz Oscillator Characteristics Symbol 1/(tCP32KHz) CL ESR tST tCH tCL CIN IOSC Note: Parameter Oscillator Frequency Crystal Equivalent Load Capacitance Crystal Equivalent Series Resistance Startup Time XIN32 Clock High Half-period XIN32 Clock Low Half-period XIN32 Input Capacitance Active mode Current Consumption Standby mode 1. CL is the equivalent load capacitance. 0.1 A CL = 6pF(1) CL = 12.5pF(1) 0.4 tCP 0.4 tCP 6 32 768 12.5 100 600 1200 0.6 tCP 0.6 tCP 5 1.8 pF A Hz pF K ms Conditions External clock on XIN32 Min. Typ. Max. 30 Unit MHz 625 32059I-06/2010 AT32UC3B 28.7.3 Main Oscillators Table 28-18. Main Oscillators Characteristics Symbol 1/(tCPMAIN) CL1, CL2 ESR Parameter Oscillator Frequency Crystal Internal Load Capacitance (CL1 = CL2) Crystal Equivalent Series Resistance Duty Cycle tST tCH tCL CIN IOSC Startup Time XIN Clock High Half-period XIN Clock Low Half-period XIN Input Capacitance Active mode at 450 KHz. Gain = G0 Current Consumption Active mode at 16 MHz. Gain = G3 325 A f = 3 MHz f = 8 MHz f = 16 MHz 0.4 tCP 0.4 tCP 12 25 40 50 0.4 12 75 60 14.5 4 1.4 0.6 tCP 0.6 tCP pF A 20 MHz pF % ms Conditions External clock on XIN Min. Typ. Max. 50 Unit MHz 28.7.4 Phase Lock Loop Table 28-19. Phase Lock Loop Characteristics Symbol FOUT FIN Parameter VCO Output Frequency Input Frequency Active mode FVCO@96 MHz Active mode FVCO@128 MHz Active mode FVCO@160 MHz Standby mode Conditions Min. 80 4 320 410 450 5 Typ. Max. 240 16 Unit MHz MHz A A IPLL Current Consumption 626 32059I-06/2010 AT32UC3B 28.8 ADC Characteristics Table 28-20. Channel Conversion Time and ADC Clock Parameter ADC Clock Frequency ADC Clock Frequency Startup Time Track and Hold Acquisition Time Track and Hold Input Resistor Track and Hold Capacitor ADC Clock = 5 MHz Conversion Time ADC Clock = 8 MHz ADC Clock = 5 MHz Throughput Rate ADC Clock = 8 MHz Notes: 1.25 384 533 (1) (2) Conditions 10-bit resolution mode 8-bit resolution mode Return from Idle Mode Min. Typ. Max. 5 8 20 Unit MHz MHz s ns 600 350 12 2 pF s s kSPS kSPS 1. Corresponds to 13 clock cycles: 3 clock cycles for track and hold acquisition time and 10 clock cycles for conversion. 2. Corresponds to 15 clock cycles: 5 clock cycles for track and hold acquisition time and 10 clock cycles for conversion. Table 28-21. External Voltage Reference Input Parameter ADVREF Input Voltage Range ADVREF Average Current Current Consumption on VDDANA Note: Conditions (1) Min. 2.6 Typ. Max. VDDANA Unit V A mA On 13 samples with ADC Clock = 5 MHz On 13 samples with ADC Clock = 5 MHz 200 250 1 1. ADVREF should be connected to GND to avoid extra consumption in case ADC is not used. Table 28-22. Analog Inputs Parameter Input Voltage Range Input Leakage Current Input Capacitance 7 Conditions Min. 0 Typ. Max. VADVREF 1 Unit V A pF Table 28-23. Transfer Characteristics in 8-bit Mode Parameter Resolution ADC Clock = 5 MHz Absolute Accuracy ADC Clock = 8 MHz ADC Clock = 5 MHz Integral Non-linearity ADC Clock = 8 MHz 0.5 1.0 LSB 0.35 1.5 0.5 LSB LSB Conditions Min. Typ. 8 0.8 Max. Unit Bit LSB 627 32059I-06/2010 AT32UC3B Table 28-23. Transfer Characteristics in 8-bit Mode Parameter Differential Non-linearity ADC Clock = 8 MHz Offset Error Gain Error ADC Clock = 5 MHz ADC Clock = 5 MHz -0.5 -0.5 0.5 1.0 0.5 0.5 LSB LSB LSB Conditions ADC Clock = 5 MHz Min. Typ. 0.3 Max. 0.5 Unit LSB Table 28-24. Transfer Characteristics in 10-bit Mode Parameter Resolution Absolute Accuracy Integral Non-linearity Differential Non-linearity Offset Error Gain Error ADC Clock = 5 MHz ADC Clock = 5 MHz ADC Clock = 5 MHz ADC Clock = 2.5 MHz ADC Clock = 5 MHz ADC Clock = 5MHz -2 -2 1.5 1 0.6 Conditions Min. Typ. 10 3 2 2 1 2 2 Max. Unit Bit LSB LSB LSB LSB LSB LSB 28.9 28.9.1 JTAG Characteristics JTAG Interface Signals Table 28-25. JTAG Interface Timing Specification Symbol JTAG0 JTAG1 JTAG2 JTAG3 JTAG4 JTAG5 JTAG6 JTAG7 JTAG8 JTAG9 JTAG10 Notes: Parameter TCK Low Half-period TCK High Half-period TCK Period TDI, TMS Setup before TCK High TDI, TMS Hold after TCK High TDO Hold Time TCK Low to TDO Valid Device Inputs Setup Time Device Inputs Hold Time Device Outputs Hold Time TCK to Device Outputs Valid Conditions (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) (1) Min. 6 3 9 1 0 4 Max. Unit ns ns ns ns ns ns 6 ns ns ns ns ns 1. VVDDIO from 3.0V to 3.6V, maximum external capacitor = 40pF 628 32059I-06/2010 AT32UC3B Figure 28-5. JTAG Interface Signals JTAG2 TCK JTAG JTAG1 0 TMS/TDI JTAG3 JTAG4 TDO JTAG5 JTAG6 Device Inputs JTAG7 JTAG8 Device Outputs JTAG9 JTAG10 28.10 SPI Characteristics Figure 28-6. SPI Master mode with (CPOL = NCPHA = 0) or (CPOL= NCPHA= 1) SPCK SPI0 MISO SPI1 SPI2 MOSI 629 32059I-06/2010 AT32UC3B Figure 28-7. SPI Master mode with (CPOL=0 and NCPHA=1) or (CPOL=1 and NCPHA=0) SPCK SPI3 MISO SPI4 SPI5 MOSI Figure 28-8. SPI Slave mode with (CPOL=0 and NCPHA=1) or (CPOL=1 and NCPHA=0) SPCK SPI6 MISO SPI7 MOSI SPI8 Figure 28-9. SPI Slave mode with (CPOL = NCPHA = 0) or (CPOL= NCPHA= 1) SPCK SPI9 MISO SPI10 MOSI SPI11 630 32059I-06/2010 AT32UC3B Table 28-26. SPI Timings Symbol SPI0 SPI1 SPI2 SPI3 SPI4 SPI5 SPI6 SPI7 SPI8 SPI9 SPI10 SPI11 Notes: Parameter MISO Setup time before SPCK rises (master) MISO Hold time after SPCK rises (master) SPCK rising to MOSI Delay (master) MISO Setup time before SPCK falls (master) MISO Hold time after SPCK falls (master) SPCK falling to MOSI Delay master) SPCK falling to MISO Delay (slave) MOSI Setup time before SPCK rises (slave) MOSI Hold time after SPCK rises (slave) SPCK rising to MISO Delay (slave) MOSI Setup time before SPCK falls (slave) MOSI Hold time after SPCK falls (slave) Conditions 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 3.3V domain(1) 0 1 0 1.5 27 22 + (tCPMCK)/2(2) 0 7 26.5 Min. 22 + (tCPMCK)/2(2) 0 7 Max. Unit ns ns ns ns ns ns ns ns ns ns ns ns 1. 3.3V domain: VVDDIO from 3.0V to 3.6V, maximum external capacitor = 40 pF. 2. tCPMCK: Master Clock period in ns. 631 32059I-06/2010 AT32UC3B 28.11 Flash Memory Characteristics The following table gives the device maximum operating frequency depending on the field FWS of the Flash FSR register. This field defines the number of wait states required to access the Flash Memory. Flash operating frequency equals the CPU/HSB frequency. Table 28-27. Flash Operating Frequency Symbol FFOP Parameter Flash Operating Frequency FWS = 1 60 MHz Conditions FWS = 0 Min. Typ. Max. 33 Unit MHz Table 28-28. Programming TIme Symbol TFPP TFFP TFCE Parameter Page Programming Time Fuse Programming Time Chip Erase Time Conditions Min. Typ. 4 0.5 4 Max. Unit ms ms ms Table 28-29. Flash Parameters Symbol NFARRAY NFFUSE TFDR Parameter Flash Array Write/Erase cycle General Purpose Fuses write cycle Flash Data Retention Time 15 Conditions Min. Typ. Max. 100K 10K Unit Cycle Cycle Year 632 32059I-06/2010 AT32UC3B 29. Mechanical Characteristics 29.1 29.1.1 Thermal Considerations Thermal Data Table 29-1 summarizes the thermal resistance data depending on the package. Table 29-1. Symbol JA JC JA JC Thermal Resistance Data Parameter Junction-to-ambient thermal resistance Junction-to-case thermal resistance Junction-to-ambient thermal resistance Junction-to-case thermal resistance Still Air Condition Still Air Package TQFP64 TQFP64 TQFP48 TQFP48 Typ 49.6 13.5 51.1 13.7 C/W Unit C/W 29.1.2 Junction Temperature The average chip-junction temperature, TJ, in C can be obtained from the following: 1. 2. T J = T A + ( P D x JA ) T J = T A + ( P D x ( HEATSINK + JC ) ) where: * JA = package thermal resistance, Junction-to-ambient (C/W), provided in Table 29-1 on page 633. * JC = package thermal resistance, Junction-to-case thermal resistance (C/W), provided in Table 29-1 on page 633. * HEAT SINK = cooling device thermal resistance (C/W), provided in the device datasheet. * PD = device power consumption (W) estimated from data provided in the section "Power Consumption" on page 621. * TA = ambient temperature (C). From the first equation, the user can derive the estimated lifetime of the chip and decide if a cooling device is necessary or not. If a cooling device is to be fitted on the chip, the second equation should be used to compute the resulting average chip-junction temperature TJ in C. 633 32059I-06/2010 AT32UC3B 29.2 Package Drawings Figure 29-1. TQFP-64 package drawing Table 29-2. Weight Device and Package Maximum Weight 300 mg Table 29-3. Package Characteristics Jedec J-STD-20D-MSL3 Moisture Sensitivity Level Table 29-4. Package Reference MS-026 e3 JEDEC Drawing Reference JESD97 Classification 634 32059I-06/2010 AT32UC3B Figure 29-2. TQFP-48 package drawing Table 29-5. Weight Device and Package Maximum Weight 100 mg Table 29-6. Package Characteristics Jedec J-STD-20D-MSL3 Moisture Sensitivity Level Table 29-7. Package Reference MS-026 e3 JEDEC Drawing Reference JESD97 Classification 635 32059I-06/2010 AT32UC3B Figure 29-3. QFN-64 package drawing Table 29-8. Weight Device and Package Maximum Weight 200 mg Table 29-9. Package Characteristics Jedec J-STD-20D-MSL3 Moisture Sensitivity Level Table 29-10. Package Reference JEDEC Drawing Reference JESD97 Classification M0-220 e3 636 32059I-06/2010 AT32UC3B Figure 29-4. QFN-48 package drawing Table 29-11. Device and Package Maximum Weight Weight 100 mg Table 29-12. Package Characteristics Moisture Sensitivity Level Jedec J-STD-20D-MSL3 Table 29-13. Package Reference JEDEC Drawing Reference JESD97 Classification M0-220 e3 637 32059I-06/2010 AT32UC3B 29.3 Soldering Profile Table 29-14 gives the recommended soldering profile from J-STD-20. Table 29-14. Soldering Profile Profile Feature Average Ramp-up Rate (217C to Peak) Preheat Temperature 175C 25C Temperature Maintained Above 217C Time within 5C of Actual Peak Temperature Peak Temperature Range Ramp-down Rate Time 25C to Peak Temperature Note: Green Package 3C/s Min. 150C, Max. 200C 60-150s 30s 260C 6C/s Max. 8mn It is recommended to apply a soldering temperature higher than 250C. A maximum of three reflow passes is allowed per component. 638 32059I-06/2010 AT32UC3B 30. Ordering Information Temperature Operating Range Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Industrial (-40C to 85C) Device AT32UC3B0512 Ordering Code AT32UC3B0512-A2UES AT32UC3B0512-A2UR AT32UC3B0512-A2UT AT32UC3B0512-Z2UES AT32UC3B0512-Z2UR AT32UC3B0512-Z2UT Package TQFP 64 TQFP 64 TQFP 64 QFN 64 QFN 64 QFN 64 TQFP 64 TQFP 64 QFN 64 QFN 64 TQFP 64 TQFP 64 QFN 64 QFN 64 TQFP 64 TQFP 64 QFN 64 QFN 64 QFN 48 QFN 48 TQFP 48 TQFP 48 QFN 48 QFN 48 TQFP 48 TQFP 48 QFN 48 QFN 48 TQFP 48 TQFP 48 QFN 48 QFN 48 Conditioning Reel Tray Reel Tray Tray Reel Tray Reel Tray Reel Tray Reel Tray Reel Tray Reel Tray Reel Tray Reel Tray Reel Tray Reel Tray Reel Tray Reel AT32UC3B0256 AT32UC3B0256-A2UT AT32UC3B0256-A2UR AT32UC3B0256-Z2UT AT32UC3B0256-Z2UR AT32UC3B0128 AT32UC3B0128-A2UT AT32UC3B0128-A2UR AT32UC3B0128-Z2UT AT32UC3B0128-Z2UR AT32UC3B064 AT32UC3B064-A2UT AT32UC3B064-A2UR AT32UC3B064-Z2UT AT32UC3B064-Z2UR AT32UC3B1512 AT32UC3B1256 AT32UC3B1512-Z1UT AT32UC3B1512-Z1UR AT32UC3B1256-AUT AT32UC3B1256-AUR AT32UC3B1256-Z1UT AT32UC3B1256-Z1UR AT32UC3B1128 AT32UC3B1128-AUT AT32UC3B1128-AUR AT32UC3B1128-Z1UT AT32UC3B1128-Z1UR AT32UC3B164 AT32UC3B164-AUT AT32UC3B164-AUR AT32UC3B164-Z1UT AT32UC3B164-Z1UR 639 32059I-06/2010 AT32UC3B 31. Errata 31.1 31.1.1 31.1.1.1 AT32UC3B0512, AT32UC3B1512 Rev D PWM 1. PWM channel interrupt enabling triggers an interrupt When enabling a PWM channel that is configured with center aligned period (CALG=1), an interrupt is signalled. Fix/Workaround When using center aligned mode, enable the channel and read the status before channel interrupt is enabled. 2. PWM counter restarts at 0x0001 The PWM counter restarts at 0x0001 and not 0x0000 as specified. Because of this the first PWM period has one more clock cycle. Fix/Workaround - The first period is 0x0000, 0x0001, ..., period - Consecutive periods are 0x0001, 0x0002, ..., period 3. PWM update period to a 0 value does not work It is impossible to update a period equal to 0 by the using the PWM update register (PWM_CUPD). Fix/Workaround Do not update the PWM_CUPD register with a value equal to 0. 31.1.1.2 SPI 1. SPI Slave / PDCA transfer: no TX UNDERRUN flag There is no TX UNDERRUN flag available, therefore in SPI slave mode, there is no way to be informed of a character lost in transmission. Fix/Workaround For PDCA transfer: none. 2. SPI Bad Serial Clock Generation on 2nd chip_select when SCBR = 1, CPOL=1 and NCPHA=0 When multiple CS are in use, if one of the baudrate equals to 1 and one of the others doesn't equal to 1, and CPOL=1 and CPHA=0, then an aditional pulse will be generated on SCK. Fix/workaround When multiple CS are in use, if one of the baudrate equals 1, the other must also equal 1 if CPOL=1 and CPHA=0. 3. SPI Glitch on RXREADY flag in slave mode when enabling the SPI or during the first transfer In slave mode, the SPI can generate a false RXREADY signal during enabling of the SPI or during the first transfer. Fix/Workaround 1. Set slave mode, set required CPOL/CPHA. 2. Enable SPI. 3. Set the polarity CPOL of the line in the opposite value of the required one. 4. Set the polarity CPOL to the required one. 640 32059I-06/2010 AT32UC3B 5. Read the RXHOLDING register. Transfers can now befin and RXREADY will now behave as expected. 4. SPI Disable does not work in Slave mode SPI Disable does not work in Slave mode. Fix/workaround Read the last received data then perform a Software reset. 5. Module hangs when CSAAT=1 on CS0 SPI data transfer hangs with CSAAT=1 in CSR0 and MODFDIS=0 in MR. When CSAAT=1 in CSR0 and mode fault detection is enabled (MODFDIS=0 in MR), the SPI module will not start a data transfer. Fix/Workaround Disable mode fault detection by writing a one to MODFDIS in MR. 6. Disabling SPI has no effect on flag TDRE flag Disabling SPI has no effect on TDRE whereas the write data command is filtered when SPI is disabled. This means that as soon as the SPI is disabled it becomes impossible to reset the TDRE flag by writing in the TDR. So if the SPI is disabled during a PDCA transfer, the PDCA will continue to write data in the TDR (as TDRE stays high) until its buffer is empty, and all data written after the disable command is lost. Fix/Workaround Disable the PDCA, 2 NOP (minimum), disable SPI. When you want to continue the transfer: Enable SPI, enable PDCA. 31.1.1.3 Power Manager 1. If the BOD level is higher than VDDCORE, the part is constantly under reset If the BOD level is set to a value higher than VDDCORE and enabled by fuses, the part will be in constant reset. Fix/Workaround Apply an external voltage on VDDCORE that is higher than the BOD level and is lower than VDDCORE max and disable the BOD. 2. When the main clock is RCSYS, TIMER_CLOCK5 is equal to PBA clock When the main clock is generated from RCSYS, TIMER_CLOCK5 is equal to PBA Clock and not PBA Clock / 128. Fix/workaround: None. 3. Clock sources will not be stopped in STATIC sleep mode if the difference between CPU and PBx division factor is too big If the division factor between the CPU/HSB and PBx frequencies is more than 4 when going to a sleep mode where the system RC oscillator is turned off, then high speed clock sources will not be turned off. This will result in a significantly higher power consumption during the sleep mode. Fix/Workaround Before going to sleep modes where the system RC oscillator is stopped, make sure that the factor between the CPU/HSB and PBx frequencies is less than or equal to 4. 4. Increased Power Consunption in VDDIO in sleep modes If the OSC0 is enabled in crystal mode when entering a sleep mode where the OSC0 is disabled, this will lead to an increased power consumption in VDDIO. 641 32059I-06/2010 AT32UC3B Fix/Workaround Disable the OSC0 through the Power Manager (PM) before going to any sleep mode where the OSC0 is disabled, or pull down or up XIN0 and XOUT0 with 1Mohm resistor. 31.1.1.4 SSC 1. Additional delay on TD output A delay from 2 to 3 system clock cycles is added to TD output when: TCMR.START = Receive Start, TCMR.STTDLY = more than ZERO, RCMR.START = Start on falling edge / Start on Rising edge / Start on any edge, RFMR.FSOS = None (input). Fix/Workaround None. 2. TF output is not correct TF output is not correct (at least emitted one serial clock cycle later than expected) when: TFMR.FSOS = Driven Low during data transfer/ Driven High during data transfer TCMR.START = Receive start RFMR.FSOS = None (Input) RCMR.START = any on RF (edge/level) Fix/Workaround None. 3. Frame Synchro and Frame Synchro Data are delayed by one clock cycle. The frame synchro and the frame synchro data are delayed from 1 SSC_CLOCK when: Clock is CKDIV The START is selected on either a frame synchro edge or a level, Frame synchro data is enabled, Transmit clock is gated on output (through CKO field). Fix/Workaround Transmit or receive CLOCK must not be gated (by the mean of CKO field) whenSTART condition is performed on a generated frame synchro. 31.1.1.5 USB 1. For isochronous pipe, the INTFRQ is irrelevant IN and OUT tokens are sent every 1 ms (Full Speed). Fix/Workaround For longer polling time, the software must freeze the pipe for the desired period in order to prevent any "extra" token. 31.1.1.6 ADC 1. Sleep Mode activation needs additional A to D conversion If the ADC sleep mode is activated when the ADC is idle the ADC will not enter sleep mode before after the next AD conversion. Fix/Workaround Activate the sleep mode in the mode register and then perform an AD conversion. 31.1.1.7 PDCA 1. Wrong PDCA behavior when using two PDCA channels with the same PID Wrong PDCA behavior when using two PDCA channels with the same PID. Fix/Workaround 642 32059I-06/2010 AT32UC3B The same PID should not be assigned to more than one channel. 2. Transfer error will stall a transmit peripheral handshake interface If a tranfer error is encountered on a channel transmitting to a peripheral, the peripheral handshake of the active channel will stall and the PDCA will not do any more transfers on the affected peripheral handshake interface. Fix/workaround: Disable and then enable the peripheral after the transfer error. 31.1.1.8 TWI 1. The TWI RXRDY flag in SR register is not reset when a software reset is performed The TWI RXRDY flag in SR register is not reset when a software reset is performed. Fix/Workaround After a Software Reset, the register TWI RHR must be read. 2. TWI in master mode will continue to read data TWI in master mode will continue to read data on the line even if the shift register and the RHR register are full. This will generate an overrun error. Fix/workaround To prevent this, read the RHR register as soon as a new RX data is ready. 3. TWI slave behaves improperly if master acknowledges the last transmitted data byte before a STOP condition In I2C slave transmitter mode, if the master acknowledges the last data byte before a STOP condition (what the master is not supposed to do), the following TWI slave receiver mode frame may contain an inappropriate clock stretch. This clock stretch can only be stopped by resetting the TWI. Fix/Workaround If the TWI is used as a slave transmitter with a master that acknowledges the last data byte before a STOP condition, it is necessary to reset the TWI beforeentering slave receiver mode. 31.1.1.9 Processor and Architecture 1. LDM instruction with PC in the register list and without ++ increments Rp For LDM with PC in the register list: the instruction behaves as if the ++ field is always set, ie the pointer is always updated. This happens even if the ++ field is cleared. Specifically, the increment of the pointer is done in parallel with the testing of R12. Fix/Workaround None. 2. RETE instruction does not clear SREG[L] from interrupts The RETE instruction clears SREG[L] as expected from exceptions. Fix/Workaround When using the STCOND instruction, clear SREG[L] in the stacked value of SR before returning from interrupts with RETE. 3. Privilege violation when using interrupts in application mode with protected system stack If the system stack is protected by the MPU and an interrupt occurs in application mode, an MPU DTLB exception will occur. Fix/Workaround 643 32059I-06/2010 AT32UC3B Make a DTLB Protection (Write) exception handler which permits the interrupt request to be handled in privileged mode. 31.1.1.10 USART 1. ISO7816 info register US_NER cannot be read The NER register always returns zero. Fix/Workaround None. 2. ISO7816 Mode T1: RX impossible after any TX RX impossible after any TX. Fix/Workaround SOFT_RESET on RX+ Config US_MR + Config_US_CR 3. The RTS output does not function correctly in hardware handshaking mode The RTS signal is not generated properly when the USART receives data in hardware handshaking mode. When the Peripheral DMA receive buffer becomes full, the RTS output should go high, but it will stay low. Fix/Workaround Do not use the hardware handshaking mode of the USART. If it is necessary to drive the RTS output high when the Peripheral DMA receive buffer becomes full, use the normal mode of the USART. Configure the Peripheral DMA Controller to signal an interrupt when the receive buffer is full. In the interrupt handler code, write a one to the RTSDIS bit in the USART Control Register (CR). This will drive the RTS output high. After the next DMA transfer is started and a receive buffer is available, write a one to the RTSEN bit in the USART CR so that RTS will be driven low. 4. Corruption after receiving too many bits in SPI slave mode If the USART is in SPI slave mode and receives too much data bits (ex: 9bitsinstead of 8 bits) by the the SPI master, an error occurs . After that, the next reception may be corrupted even if the frame is correct and the USART has been disabled, reseted by a soft reset and re-enabled. Fix/Workaround None. 5. USART slave synchronous mode external clock must be at least 9 times lower in frequency than CLK_USART When the USART is operating in slave synchronous mode with an external clock, the frequency of the signal provided on CLK must be at least 9 times lower than CLK_USART. Fix/Workaround When the USART is operating in slave synchronous mode with an external clock, provide a signal on CLK that has a frequency at least 9 times lower than CLK_USART. 31.1.1.11 HMATRIX 1. In the HMATRIX PRAS and PRBS registers MxPR fields are only two bits In the HMATRIX PRAS and PRBS registers MxPR fields are only two bits wide, instead of four bits. The unused bits are undefined when reading the registers. Fix/Workaround Mask undefined bits when reading PRAS and PRBS. 644 32059I-06/2010 AT32UC3B 31.1.1.12 DSP Operations 1. Instruction breakpoints affected on all MAC instruction Hardware breakpoints on MAC instructions may corrupt the destination register of the MAC instruction. Fix/Workaround Place breakpoints on earlier or later instructions. 645 32059I-06/2010 AT32UC3B 31.1.2 31.1.2.1 Rev C PWM 1. PWM channel interrupt enabling triggers an interrupt When enabling a PWM channel that is configured with center aligned period (CALG=1), an interrupt is signalled. Fix/Workaround When using center aligned mode, enable the channel and read the status before channel interrupt is enabled. 2. PWM counter restarts at 0x0001 The PWM counter restarts at 0x0001 and not 0x0000 as specified. Because of this the first PWM period has one more clock cycle. Fix/Workaround - The first period is 0x0000, 0x0001, ..., period - Consecutive periods are 0x0001, 0x0002, ..., period 3. PWM update period to a 0 value does not work It is impossible to update a period equal to 0 by the using the PWM update register (PWM_CUPD). Fix/Workaround Do not update the PWM_CUPD register with a value equal to 0. 31.1.2.2 SPI 1. SPI Slave / PDCA transfer: no TX UNDERRUN flag There is no TX UNDERRUN flag available, therefore in SPI slave mode, there is no way to be informed of a character lost in transmission. Fix/Workaround For PDCA transfer: none. 2. SPI Bad Serial Clock Generation on 2nd chip_select when SCBR = 1, CPOL=1 and NCPHA=0 When multiple CS are in use, if one of the baudrate equals to 1 and one of the others doesn't equal to 1, and CPOL=1 and CPHA=0, then an aditional pulse will be generated on SCK. Fix/workaround When multiple CS are in use, if one of the baudrate equals 1, the other must also equal 1 if CPOL=1 and CPHA=0. 3. SPI Glitch on RXREADY flag in slave mode when enabling the SPI or during the first transfer In slave mode, the SPI can generate a false RXREADY signal during enabling of the SPI or during the first transfer. Fix/Workaround 1. Set slave mode, set required CPOL/CPHA. 2. Enable SPI. 3. Set the polarity CPOL of the line in the opposite value of the required one. 4. Set the polarity CPOL to the required one. 5. Read the RXHOLDING register. Transfers can now befin and RXREADY will now behave as expected. 4. SPI Disable does not work in Slave mode 646 32059I-06/2010 AT32UC3B SPI Disable does not work in Slave mode. Fix/workaround Read the last received data then perform a Software reset. 5. Module hangs when CSAAT=1 on CS0 SPI data transfer hangs with CSAAT=1 in CSR0 and MODFDIS=0 in MR. When CSAAT=1 in CSR0 and mode fault detection is enabled (MODFDIS=0 in MR), the SPI module will not start a data transfer. Fix/Workaround Disable mode fault detection by writing a one to MODFDIS in MR. 6. Disabling SPI has no effect on flag TDRE flag Disabling SPI has no effect on TDRE whereas the write data command is filtered when SPI is disabled. This means that as soon as the SPI is disabled it becomes impossible to reset the TDRE flag by writing in the TDR. So if the SPI is disabled during a PDCA transfer, the PDCA will continue to write data in the TDR (as TDRE stays high) until its buffer is empty, and all data written after the disable command is lost. Fix/Workaround Disable the PDCA, 2 NOP (minimum), disable SPI. When you want to continue the transfer: Enable SPI, enable PDCA. 31.1.2.3 Power Manager 1. If the BOD level is higher than VDDCORE, the part is constantly under reset If the BOD level is set to a value higher than VDDCORE and enabled by fuses, the part will be in constant reset. Fix/Workaround Apply an external voltage on VDDCORE that is higher than the BOD level and is lower than VDDCORE max and disable the BOD. 2. When the main clock is RCSYS, TIMER_CLOCK5 is equal to PBA clock When the main clock is generated from RCSYS, TIMER_CLOCK5 is equal to PBA Clock and not PBA Clock / 128. Fix/workaround: None. 3. VDDCORE power supply input needs to be 1.95V When used in dual power supply, VDDCORE needs to be 1.95V. Fix/workaround: When used in single power supply, VDDCORE needs to be connected to VDDOUT, which is configured on revision C at 1.95V (typ.). 4. Clock sources will not be stopped in STATIC sleep mode if the difference between CPU and PBx division factor is too big If the division factor between the CPU/HSB and PBx frequencies is more than 4 when going to a sleep mode where the system RC oscillator is turned off, then high speed clock sources will not be turned off. This will result in a significantly higher power consumption during the sleep mode. Fix/Workaround Before going to sleep modes where the system RC oscillator is stopped, make sure that the factor between the CPU/HSB and PBx frequencies is less than or equal to 4. 5. Increased Power Consunption in VDDIO in sleep modes 647 32059I-06/2010 AT32UC3B If the OSC0 is enabled in crystal mode when entering a sleep mode where the OSC0 is disabled, this will lead to an increased power consumption in VDDIO. Fix/Workaround Disable the OSC0 through the Power Manager (PM) before going to any sleep mode where the OSC0 is disabled, or pull down or up XIN0 and XOUT0 with 1Mohm resistor. 31.1.2.4 SSC 1. Additional delay on TD output A delay from 2 to 3 system clock cycles is added to TD output when: TCMR.START = Receive Start, TCMR.STTDLY = more than ZERO, RCMR.START = Start on falling edge / Start on Rising edge / Start on any edge, RFMR.FSOS = None (input). Fix/Workaround None. 2. TF output is not correct TF output is not correct (at least emitted one serial clock cycle later than expected) when: TFMR.FSOS = Driven Low during data transfer/ Driven High during data transfer TCMR.START = Receive start RFMR.FSOS = None (Input) RCMR.START = any on RF (edge/level) Fix/Workaround None. 3. Frame Synchro and Frame Synchro Data are delayed by one clock cycle. The frame synchro and the frame synchro data are delayed from 1 SSC_CLOCK when: Clock is CKDIV The START is selected on either a frame synchro edge or a level, Frame synchro data is enabled, Transmit clock is gated on output (through CKO field). Fix/Workaround Transmit or receive CLOCK must not be gated (by the mean of CKO field) whenSTART condition is performed on a generated frame synchro. 31.1.2.5 USB 1. For isochronous pipe, the INTFRQ is irrelevant IN and OUT tokens are sent every 1 ms (Full Speed). Fix/Workaround For longer polling time, the software must freeze the pipe for the desired period in order to prevent any "extra" token. 31.1.2.6 ADC 1. Sleep Mode activation needs additional A to D conversion If the ADC sleep mode is activated when the ADC is idle the ADC will not enter sleep mode before after the next AD conversion. Fix/Workaround Activate the sleep mode in the mode register and then perform an AD conversion. 31.1.2.7 PDCA 1. Wrong PDCA behavior when using two PDCA channels with the same PID 648 32059I-06/2010 AT32UC3B Wrong PDCA behavior when using two PDCA channels with the same PID. Fix/Workaround The same PID should not be assigned to more than one channel. 2. Transfer error will stall a transmit peripheral handshake interface If a tranfer error is encountered on a channel transmitting to a peripheral, the peripheral handshake of the active channel will stall and the PDCA will not do any more transfers on the affected peripheral handshake interface. Fix/workaround: Disable and then enable the peripheral after the transfer error. 31.1.2.8 TWI 1. The TWI RXRDY flag in SR register is not reset when a software reset is performed The TWI RXRDY flag in SR register is not reset when a software reset is performed. Fix/Workaround After a Software Reset, the register TWI RHR must be read. 2. TWI in master mode will continue to read data TWI in master mode will continue to read data on the line even if the shift register and the RHR register are full. This will generate an overrun error. Fix/workaround To prevent this, read the RHR register as soon as a new RX data is ready. 3. TWI slave behaves improperly if master acknowledges the last transmitted data byte before a STOP condition In I2C slave transmitter mode, if the master acknowledges the last data byte before a STOP condition (what the master is not supposed to do), the following TWI slave receiver mode frame may contain an inappropriate clock stretch. This clock stretch can only be stopped by resetting the TWI. Fix/Workaround If the TWI is used as a slave transmitter with a master that acknowledges the last data byte before a STOP condition, it is necessary to reset the TWI beforeentering slave receiver mode. 31.1.2.9 Processor and Architecture 1. LDM instruction with PC in the register list and without ++ increments Rp For LDM with PC in the register list: the instruction behaves as if the ++ field is always set, ie the pointer is always updated. This happens even if the ++ field is cleared. Specifically, the increment of the pointer is done in parallel with the testing of R12. Fix/Workaround None. 2. RETE instruction does not clear SREG[L] from interrupts The RETE instruction clears SREG[L] as expected from exceptions. Fix/Workaround When using the STCOND instruction, clear SREG[L] in the stacked value of SR before returning from interrupts with RETE. 3. Privilege violation when using interrupts in application mode with protected system stack If the system stack is protected by the MPU and an interrupt occurs in application mode, an MPU DTLB exception will occur. 649 32059I-06/2010 AT32UC3B Fix/Workaround Make a DTLB Protection (Write) exception handler which permits the interrupt request to be handled in privileged mode. 31.1.2.10 Flash 1. Reset vector is 80000020h rather than 80000000h Reset vector is 80000020h rather than 80000000h. Fix/Workaround The flash program code must start at the address 80000020h. The flash memory range 80000000h-80000020h must be programmed with 00000000h. 31.1.2.11 USART 1. ISO7816 info register US_NER cannot be read The NER register always returns zero. Fix/Workaround None. 2. ISO7816 Mode T1: RX impossible after any TX RX impossible after any TX. Fix/Workaround SOFT_RESET on RX+ Config US_MR + Config_US_CR 3. The RTS output does not function correctly in hardware handshaking mode The RTS signal is not generated properly when the USART receives data in hardware handshaking mode. When the Peripheral DMA receive buffer becomes full, the RTS output should go high, but it will stay low. Fix/Workaround Do not use the hardware handshaking mode of the USART. If it is necessary to drive the RTS output high when the Peripheral DMA receive buffer becomes full, use the normal mode of the USART. Configure the Peripheral DMA Controller to signal an interrupt when the receive buffer is full. In the interrupt handler code, write a one to the RTSDIS bit in the USART Control Register (CR). This will drive the RTS output high. After the next DMA transfer is started and a receive buffer is available, write a one to the RTSEN bit in the USART CR so that RTS will be driven low. 4. Corruption after receiving too many bits in SPI slave mode If the USART is in SPI slave mode and receives too much data bits (ex: 9bitsinstead of 8 bits) by the the SPI master, an error occurs . After that, the next reception may be corrupted even if the frame is correct and the USART has been disabled, reseted by a soft reset and re-enabled. Fix/Workaround None. 5. USART slave synchronous mode external clock must be at least 9 times lower in frequency than CLK_USART When the USART is operating in slave synchronous mode with an external clock, the frequency of the signal provided on CLK must be at least 9 times lower than CLK_USART. Fix/Workaround When the USART is operating in slave synchronous mode with an external clock, provide a signal on CLK that has a frequency at least 9 times lower than CLK_USART. 650 32059I-06/2010 AT32UC3B 31.1.2.12 HMATRIX 1. In the HMATRIX PRAS and PRBS registers MxPR fields are only two bits In the HMATRIX PRAS and PRBS registers MxPR fields are only two bits wide, instead of four bits. The unused bits are undefined when reading the registers. Fix/Workaround Mask undefined bits when reading PRAS and PRBS. 31.1.2.13 DSP Operations 1. Instruction breakpoints affected on all MAC instruction Hardware breakpoints on MAC instructions may corrupt the destination register of the MAC instruction. Fix/Workaround Place breakpoints on earlier or later instructions. 651 32059I-06/2010 AT32UC3B 31.2 AT32UC3B0256, AT32UC3B0128, AT32UC3B064, AT32UC3B1256, AT32UC3B1128, AT32UC3B164 All industrial parts labelled with -UES (for engineering samples) are revision B parts. 31.2.1 31.2.1.1 Rev. G PWM 1. PWM channel interrupt enabling triggers an interrupt When enabling a PWM channel that is configured with center aligned period (CALG=1), an interrupt is signalled. Fix/Workaround When using center aligned mode, enable the channel and read the status before channel interrupt is enabled. 2. PWM counter restarts at 0x0001 The PWM counter restarts at 0x0001 and not 0x0000 as specified. Because of this the first PWM period has one more clock cycle. Fix/Workaround - The first period is 0x0000, 0x0001, ..., period - Consecutive periods are 0x0001, 0x0002, ..., period 3. PWM update period to a 0 value does not work It is impossible to update a period equal to 0 by the using the PWM update register (PWM_CUPD). Fix/Workaround Do not update the PWM_CUPD register with a value equal to 0. 31.2.1.2 SPI 1. SPI Slave / PDCA transfer: no TX UNDERRUN flag There is no TX UNDERRUN flag available, therefore in SPI slave mode, there is no way to be informed of a character lost in transmission. Fix/Workaround For PDCA transfer: none. 2. SPI Bad Serial Clock Generation on 2nd chip_select when SCBR = 1, CPOL=1 and NCPHA=0 When multiple CS are in use, if one of the baudrate equals to 1 and one of the others doesn't equal to 1, and CPOL=1 and CPHA=0, then an aditional pulse will be generated on SCK. Fix/workaround When multiple CS are in use, if one of the baudrate equals 1, the other must also equal 1 if CPOL=1 and CPHA=0. 3. SPI Glitch on RXREADY flag in slave mode when enabling the SPI or during the first transfer In slave mode, the SPI can generate a false RXREADY signal during enabling of the SPI or during the first transfer. Fix/Workaround 1. Set slave mode, set required CPOL/CPHA. 2. Enable SPI. 652 32059I-06/2010 AT32UC3B 3. Set the polarity CPOL of the line in the opposite value of the required one. 4. Set the polarity CPOL to the required one. 5. Read the RXHOLDING register. Transfers can now befin and RXREADY will now behave as expected. 4. SPI Disable does not work in Slave mode SPI Disable does not work in Slave mode. Fix/workaround Read the last received data then perform a Software reset. 5. Disabling SPI has no effect on flag TDRE flag Disabling SPI has no effect on TDRE whereas the write data command is filtered when SPI is disabled. This means that as soon as the SPI is disabled it becomes impossible to reset the TDRE flag by writing in the TDR. So if the SPI is disabled during a PDCA transfer, the PDCA will continue to write data in the TDR (as TDRE stays high) until its buffer is empty, and all data written after the disable command is lost. Fix/Workaround Disable the PDCA, 2 NOP (minimum), disable SPI. When you want to continue the transfer: Enable SPI, enable PDCA. Power Manager 1. If the BOD level is higher than VDDCORE, the part is constantly under reset If the BOD level is set to a value higher than VDDCORE and enabled by fuses, the part will be in constant reset. Fix/Workaround Apply an external voltage on VDDCORE that is higher than the BOD level and is lower than VDDCORE max and disable the BOD. 2. When the main clock is RCSYS, TIMER_CLOCK5 is equal to PBA clock When the main clock is generated from RCSYS, TIMER_CLOCK5 is equal to PBA Clock and not PBA Clock / 128. Fix/workaround: None. 3. Clock sources will not be stopped in STATIC sleep mode if the difference between CPU and PBx division factor is too big If the division factor between the CPU/HSB and PBx frequencies is more than 4 when going to a sleep mode where the system RC oscillator is turned off, then high speed clock sources will not be turned off. This will result in a significantly higher power consumption during the sleep mode. Fix/Workaround Before going to sleep modes where the system RC oscillator is stopped, make sure that the factor between the CPU/HSB and PBx frequencies is less than or equal to 4. 4. Increased Power Consunption in VDDIO in sleep modes If the OSC0 is enabled in crystal mode when entering a sleep mode where the OSC0 is disabled, this will lead to an increased power consumption in VDDIO. Fix/Workaround Disable the OSC0 through the Power Manager (PM) before going to any sleep mode where the OSC0 is disabled, or pull down or up XIN0 and XOUT0 with 1Mohm resistor. 31.2.1.4 SSC 31.2.1.3 1. Frame Synchro and Frame Synchro Data are delayed by one clock cycle. 653 32059I-06/2010 AT32UC3B The frame synchro and the frame synchro data are delayed from 1 SSC_CLOCK when: Clock is CKDIV The START is selected on either a frame synchro edge or a level, Frame synchro data is enabled, Transmit clock is gated on output (through CKO field). Fix/Workaround Transmit or receive CLOCK must not be gated (by the mean of CKO field) whenSTART condition is performed on a generated frame synchro. 2. Additional delay on TD output A delay from 2 to 3 system clock cycles is added to TD output when: TCMR.START = Receive Start, TCMR.STTDLY = more than ZERO, RCMR.START = Start on falling edge / Start on Rising edge / Start on any edge, RFMR.FSOS = None (input). Fix/Workaround None. 3. TF output is not correct TF output is not correct (at least emitted one serial clock cycle later than expected) when: TFMR.FSOS = Driven Low during data transfer/ Driven High during data transfer TCMR.START = Receive start RFMR.FSOS = None (Input) RCMR.START = any on RF (edge/level) Fix/Workaround None. 31.2.1.5 USB 1. For isochronous pipe, the INTFRQ is irrelevant IN and OUT tokens are sent every 1 ms (Full Speed). Fix/Workaround For longer polling time, the software must freeze the pipe for the desired period in order to prevent any "extra" token. 31.2.1.6 ADC 1. Sleep Mode activation needs additional A to D conversion If the ADC sleep mode is activated when the ADC is idle the ADC will not enter sleep mode before after the next AD conversion. Fix/Workaround Activate the sleep mode in the mode register and then perform an AD conversion. 31.2.1.7 PDCA 1. Wrong PDCA behavior when using two PDCA channels with the same PID Wrong PDCA behavior when using two PDCA channels with the same PID. Fix/Workaround The same PID should not be assigned to more than one channel. 2. Transfer error will stall a transmit peripheral handshake interface If a tranfer error is encountered on a channel transmitting to a peripheral, the peripheral handshake of the active channel will stall and the PDCA will not do any more transfers on the affected peripheral handshake interface. Fix/workaround: 654 32059I-06/2010 AT32UC3B Disable and then enable the peripheral after the transfer error. 31.2.1.8 TWI 1. The TWI RXRDY flag in SR register is not reset when a software reset is performed The TWI RXRDY flag in SR register is not reset when a software reset is performed. Fix/Workaround After a Software Reset, the register TWI RHR must be read. 2. TWI in master mode will continue to read data TWI in master mode will continue to read data on the line even if the shift register and the RHR register are full. This will generate an overrun error. Fix/workaround To prevent this, read the RHR register as soon as a new RX data is ready. 3. TWI slave behaves improperly if master acknowledges the last transmitted data byte before a STOP condition In I2C slave transmitter mode, if the master acknowledges the last data byte before a STOP condition (what the master is not supposed to do), the following TWI slave receiver mode frame may contain an inappropriate clock stretch. This clock stretch can only be stopped by resetting the TWI. Fix/Workaround If the TWI is used as a slave transmitter with a master that acknowledges the last data byte before a STOP condition, it is necessary to reset the TWI beforeentering slave receiver mode. 31.2.1.9 GPIO 1. PA29 (TWI SDA) and PA30 (TWI SCL) GPIO VIH (input high voltage) is 3.6V max instead of 5V tolerant The following GPIOs are not 5V tolerant : PA29 and PA30. Fix/Workaround None. 31.2.1.10 OCD 1. The auxiliary trace does not work for CPU/HSB speed higher than 50MHz. The auxiliary trace does not work for CPU/HSB speed higher than 50MHz. Workaround: Do not use the auxiliary trace for CPU/HSB speed higher than 50MHz. 31.2.1.11 Processor and Architecture 1. LDM instruction with PC in the register list and without ++ increments Rp For LDM with PC in the register list: the instruction behaves as if the ++ field is always set, ie the pointer is always updated. This happens even if the ++ field is cleared. Specifically, the increment of the pointer is done in parallel with the testing of R12. Fix/Workaround None. 2. RETE instruction does not clear SREG[L] from interrupts The RETE instruction clears SREG[L] as expected from exceptions. Fix/Workaround When using the STCOND instruction, clear SREG[L] in the stacked value of SR before returning from interrupts with RETE. 655 32059I-06/2010 AT32UC3B 3. Exceptions when system stack is protected by MPU RETS behaves incorrectly when MPU is enabled and MPU is configured so thatsystem stack is not readable in unprivileged mode. Fix/Workaround Workaround 1: Make system stack readable in unprivileged mode, or Workaround 2: Return from supervisor mode using rete instead of rets. This requires: 1. Changing the mode bits from 001b to 110b before issuing the instruction. Updating the mode bits to the desired value must be done using a single mtsr instruction so it is done atomically. Even if this step is described in general as not safe in the UC technical reference guide, it is safe in this very specific case. 2. Execute the RETE instruction. 4. Privilege violation when using interrupts in application mode with protected system stack If the system stack is protected by the MPU and an interrupt occurs in application mode, an MPU DTLB exception will occur. Fix/Workaround Make a DTLB Protection (Write) exception handler which permits the interrupt request to be handled in privileged mode. 31.2.1.12 USART 1. ISO7816 info register US_NER cannot be read The NER register always returns zero. Fix/Workaround None. 2. ISO7816 Mode T1: RX impossible after any TX RX impossible after any TX. Fix/Workaround SOFT_RESET on RX+ Config US_MR + Config_US_CR 3. The RTS output does not function correctly in hardware handshaking mode The RTS signal is not generated properly when the USART receives data in hardware handshaking mode. When the Peripheral DMA receive buffer becomes full, the RTS output should go high, but it will stay low. Fix/Workaround Do not use the hardware handshaking mode of the USART. If it is necessary to drive the RTS output high when the Peripheral DMA receive buffer becomes full, use the normal mode of the USART. Configure the Peripheral DMA Controller to signal an interrupt when the receive buffer is full. In the interrupt handler code, write a one to the RTSDIS bit in the USART Control Register (CR). This will drive the RTS output high. After the next DMA transfer is started and a receive buffer is available, write a one to the RTSEN bit in the USART CR so that RTS will be driven low. 4. Corruption after receiving too many bits in SPI slave mode If the USART is in SPI slave mode and receives too much data bits (ex: 9bitsinstead of 8 bits) by the the SPI master, an error occurs . After that, the next reception may be corrupted even if the frame is correct and the USART has been disabled, reseted by a soft reset and re-enabled. Fix/Workaround None. 656 32059I-06/2010 AT32UC3B 5. USART slave synchronous mode external clock must be at least 9 times lower in frequency than CLK_USART When the USART is operating in slave synchronous mode with an external clock, the frequency of the signal provided on CLK must be at least 9 times lower than CLK_USART. Fix/Workaround When the USART is operating in slave synchronous mode with an external clock, provide a signal on CLK that has a frequency at least 9 times lower than CLK_USART. 31.2.1.13 HMATRIX 31.2.1.14 1. In the HMATRIX PRAS and PRBS registers MxPR fields are only two bits In the HMATRIX PRAS and PRBS registers MxPR fields are only two bits wide, instead of four bits. The unused bits are undefined when reading the registers. Fix/Workaround Mask undefined bits when reading PRAS and PRBS. DSP Operations 1. Instruction breakpoints affected on all MAC instruction Hardware breakpoints on MAC instructions may corrupt the destination register of the MAC instruction. Fix/Workaround Place breakpoints on earlier or later instructions. 657 32059I-06/2010 AT32UC3B 31.2.2 31.2.2.1 Rev. F PWM 1. PWM channel interrupt enabling triggers an interrupt When enabling a PWM channel that is configured with center aligned period (CALG=1), an interrupt is signalled. Fix/Workaround When using center aligned mode, enable the channel and read the status before channel interrupt is enabled. 2. PWN counter restarts at 0x0001 The PWM counter restarts at 0x0001 and not 0x0000 as specified. Because of this the first PWM period has one more clock cycle. Fix/Workaround - The first period is 0x0000, 0x0001, ..., period - Consecutive periods are 0x0001, 0x0002, ..., period 3. PWM update period to a 0 value does not work It is impossible to update a period equal to 0 by the using the PWM update register (PWM_CUPD). Fix/Workaround Do not update the PWM_CUPD register with a value equal to 0. 31.2.2.2 SPI 1. SPI Slave / PDCA transfer: no TX UNDERRUN flag There is no TX UNDERRUN flag available, therefore in SPI slave mode, there is no way to be informed of a character lost in transmission. Fix/Workaround For PDCA transfer: none. 2. SPI disable does not work in SLAVE mode SPI disable does not work in SLAVE mode. Fix/Workaround Read the last received data, then perform a Software Reset. 3. SPI Bad Serial Clock Generation on 2nd chip_select when SCBR = 1, CPOL=1 and NCPHA=0 When multiple CS are in use, if one of the baudrate equals to 1 and one of the others doesn't equal to 1, and CPOL=1 and CPHA=0, then an aditional pulse will be generated on SCK. Fix/workaround When multiple CS are in use, if one of the baudrate equals 1, the other must also equal 1 if CPOL=1 and CPHA=0. 4. SPI Glitch on RXREADY flag in slave mode when enabling the SPI or during the first transfer In slave mode, the SPI can generate a false RXREADY signal during enabling of the SPI or during the first transfer. Fix/Workaround 1. Set slave mode, set required CPOL/CPHA. 2. Enable SPI. 3. Set the polarity CPOL of the line in the opposite value of the required one. 658 32059I-06/2010 AT32UC3B 4. Set the polarity CPOL to the required one. 5. Read the RXHOLDING register. Transfers can now befin and RXREADY will now behave as expected. 31.2.2.3 Power Manager 1. If the BOD level is higher than VDDCORE, the part is constantly resetted If the BOD level is set to a value higher than VDDCORE and enabled by fuses, the part will be in constant reset. Fix/Workaround Apply an external voltage on VDDCORE that is higher than the BOD level and is lower than VDDCORE max and disable the BOD. 31.2.2.4 ADC 1. Sleep Mode activation needs addtionnal A to D conversion If the ADC sleep mode is activated when the ADC is idle the ADC will not enter sleep mode before after the next AD conversion. Fix/Workaround Activate the sleep mode in the mode register and then perform an AD conversion. 31.2.2.5 Processor and Architecture 1. LDM instruction with PC in the register list and without ++ increments Rp For LDM with PC in the register list: the instruction behaves as if the ++ field is always set, ie the pointer is always updated. This happens even if the ++ field is cleared. Specifically, the increment of the pointer is done in parallel with the testing of R12. Fix/Workaround None. 2. RETE instruction does not clear SREG[L] from interrupts The RETE instruction clears SREG[L] as expected from exceptions. Fix/Workaround When using the STCOND instruction, clear SREG[L] in the stacked value of SR before returning from interrupts with RETE. 3. Exceptions when system stack is protected by MPU RETS behaves incorrectly when MPU is enabled and MPU is configured so thatsystem stack is not readable in unprivileged mode. Fix/Workaround Workaround 1: Make system stack readable in unprivileged mode, or Workaround 2: Return from supervisor mode using rete instead of rets. This requires: 1. Changing the mode bits from 001b to 110b before issuing the instruction. Updating the mode bits to the desired value must be done using a single mtsr instruction so it is done atomically. Even if this step is described in general as not safe in the UC technical reference guide, it is safe in this very specific case. 2. Execute the RETE instruction. 659 32059I-06/2010 AT32UC3B 31.2.3 31.2.3.1 Rev. B PWM 1. PWM counter restarts at 0x0001 The PWM counter restarts at 0x0001 and not 0x0000 as specified. Because of this the first PWM period has one more clock cycle. Fix/Workaround - The first period is 0x0000, 0x0001, ..., period - Consecutive periods are 0x0001, 0x0002, ..., period 2. PWM channel interrupt enabling triggers an interrupt When enabling a PWM channel that is configured with center aligned period (CALG=1), an interrupt is signalled. Fix/Workaround When using center aligned mode, enable the channel and read the status before channel interrupt is enabled. 3. PWM update period to a 0 value does not work It is impossible to update a period equal to 0 by the using the PWM update register (PWM_CUPD). Fix/Workaround Do not update the PWM_CUPD register with a value equal to 0. 4. PWM channel status may be wrong if disabled before a period has elapsed Before a PWM period has elapsed, the read channel status may be wrong. The CHIDx-bit for a PWM channel in the PWM Enable Register will read '1' for one full PWM period even if the channel was disabled before the period elapsed. It will then read '0' as expected. Fix/Workaround Reading the PWM channel status of a disabled channel is only correct after a PWM period has elapsed. 5. The following alternate C functions PWM[4] on PA16 and PWM[6] on PA31 are not available on Rev B The following alternate C functions PWM[4] on PA16 and PWM[6] on PA31 are not available on Rev B. Fix/Workaround Do not use these PWM alternate functions on these pins. 31.2.3.2 SPI 1. SPI Slave / PDCA transfer: no TX UNDERRUN flag There is no TX UNDERRUN flag available, therefore in SPI slave mode, there is no way to be informed of a character lost in transmission. Fix/Workaround For PDCA transfer: none. 2. SPI Bad serial clock generation on 2nd chip select when SCBR=1, CPOL=1 and CNCPHA=0 When multiple CS are in use, if one of the baudrate equals to 1 and one of the others doesn't equal to 1, and CPOL=1 and CPHA=0, then an additional pulse will be generated on SCK. 660 32059I-06/2010 AT32UC3B Fix/Workaround When multiple CS are in use, if one of the baudrate equals to 1, the other must also equal 1 if CPOL=1 and CPHA=0. 3. SPI Glitch on RXREADY flag in slave mode when enabling the SPI or during the first transfer In slave mode, the SPI can generate a false RXREADY signal during enabling of the SPI or during the first transfer. Fix/Workaround 1. Set slave mode, set required CPOL/CPHA. 2. Enable SPI. 3. Set the polarity CPOL of the line in the opposite value of the required one. 4. Set the polarity CPOL to the required one. 5. Read the RXHOLDING register. Transfers can now befin and RXREADY will now behave as expected. 4. SPI CSNAAT bit 2 in register CSR0...CSR3 is not available SPI CSNAAT bit 2 in register CSR0...CSR3 is not available. Fix/Workaround Do not use this bit. 5. SPI disable does not work in SLAVE mode SPI disable does not work in SLAVE mode. Fix/Workaround Read the last received data, then perform a Software Reset. 6. SPI Bad Serial Clock Generation on 2nd chip_select when SCBR = 1, CPOL=1 and NCPHA=0 When multiple CS are in use, if one of the baudrate equals to 1 and one of the others doesn't equal to 1, and CPOL=1 and CPHA=0, then an aditional pulse will be generated on SCK. Fix/workaround When multiple CS are in use, if one of the baudrate equals 1, the other must also equal 1 if CPOL=1 and CPHA=0. 31.2.3.3 Power Manager 1. PLL Lock control does not work PLL lock Control does not work. Fix/Workaround In PLL Control register, the bit 7 should be set in order to prevent unexpected behaviour. 2. Wrong reset causes when BOD is activated Setting the BOD enable fuse will cause the Reset Cause Register to list BOD reset as the reset source even though the part was reset by another source. Fix/Workaround Do not set the BOD enable fuse, but activate the BOD as soon as your program starts. 3. System Timer mask (Bit 16) of the PM CPUMASK register is not available System Timer mask (Bit 16) of the PM CPUMASK register is not available. Fix/Workaround Do not use this bit. 661 32059I-06/2010 AT32UC3B 31.2.3.4 SSC 1. SSC does not trigger RF when data is low The SSC cannot transmit or receive data when CKS = CKDIV and CKO = none, in TCMR or RCMR respectively. Fix/Workaround Set CKO to a value that is not "none" and bypass the output of the TK/RK pin with the GPIO. 31.2.3.5 USB 1. USB No end of host reset signaled upon disconnection In host mode, in case of an unexpected device disconnection whereas a usb reset is being sent by the usb controller, the UHCON.RESET bit may not been cleared by the hardware at the end of the reset. Fix/Workaround A software workaround consists in testing (by polling or interrupt) the disconnection (UHINT.DDISCI == 1) while waiting for the end of reset (UHCON.RESET == 0) to avoid being stuck. 2. USBFSM and UHADDR1/2/3 registers are not available Do not use USBFSM register. Fix/Workaround Do not use USBFSM register and use HCON[6:0] field instead for all the pipes. 31.2.3.6 Cycle counter 1. CPU Cycle Counter does not reset the COUNT system register on COMPARE match. The device revision B does not reset the COUNT system register on COMPARE match. In this revision, the COUNT register is clocked by the CPU clock, so when the CPU clock stops, so does incrementing of COUNT. Fix/Workaround None. 31.2.3.7 ADC 1. ADC possible miss on DRDY when disabling a channel The ADC does not work properly when more than one channel is enabled. Fix/Workaround Do not use the ADC with more than one channel enabled at a time. 2. ADC OVRE flag sometimes not reset on Status Register read The OVRE flag does not clear properly if read simultaneously to an end of conversion. Fix/Workaround None. 3. Sleep Mode activation needs addtionnal A to D conversion If the ADC sleep mode is activated when the ADC is idle the ADC will not enter sleep mode before after the next AD conversion. Fix/Workaround Activate the sleep mode in the mode register and then perform an AD conversion. 31.2.3.8 USART 1. USART Manchester Encoder Not Working 662 32059I-06/2010 AT32UC3B Manchester encoding/decoding is not working. Fix/Workaround Do not use manchester encoding. 2. USART RXBREAK problem when no timeguard In asynchronous mode the RXBREAK flag is not correctly handled when the timeguard is 0 and the break character is located just after the stop bit. Fix/Workaround If the NBSTOP is 1, timeguard should be different from 0. 3. USART Handshaking: 2 characters sent / CTS rises when TX If CTS switches from 0 to 1 during the TX of a character, if the Holding register is not empty, the TXHOLDING is also transmitted. Fix/Workaround None. 4. USART PDC and TIMEGUARD not supported in MANCHESTER Manchester encoding/decoding is not working. Fix/Workaround Do not use manchester encoding. 5. USART SPI mode is non functional on this revision USART SPI mode is non functional on this revision. Fix/Workaround Do not use the USART SPI mode. 31.2.3.9 HMATRIX 1. HMatrix fixed priority arbitration does not work Fixed priority arbitration does not work. Fix/Workaround Use Round-Robin arbitration instead. 31.2.3.10 Clock caracteristic 1. PBA max frequency The Peripheral bus A (PBA) max frequency is 30MHz instead of 60MHz. Fix/Workaround Do not set the PBA maximum frequency higher than 30MHz. 31.2.3.11 FLASHC 1. The address of Flash General Purpose Fuse Register Low (FGPFRLO) is 0xFFFE140C on revB instead of 0xFFFE1410 The address of Flash General Purpose Fuse Register Low (FGPFRLO) is 0xFFFE140C on revB instead of 0xFFFE1410. Fix/Workaround None. 2. The command Quick Page Read User Page(QPRUP) is not functional The command Quick Page Read User Page(QPRUP) is not functional. Fix/Workaround None. 663 32059I-06/2010 AT32UC3B 3. PAGEN Semantic Field for Program GP Fuse Byte is WriteData[7:0], ByteAddress[1:0] on revision B instead of WriteData[7:0], ByteAddress[2:0] PAGEN Semantic Field for Program GP Fuse Byte is WriteData[7:0], ByteAddress[1:0] on revision B instead of WriteData[7:0], ByteAddress[2:0]. Fix/Workaround None. 31.2.3.12 RTC 1. Writes to control (CTRL), top (TOP) and value (VAL) in the RTC are discarded if the RTC peripheral bus clock (PBA) is divided by a factor of four or more relative to the HSB clock Writes to control (CTRL), top (TOP) and value (VAL) in the RTC are discarded if the RTC peripheral bus clock (PBA) is divided by a factor of four or more relative to the HSB clock. Fix/Workaround Do not write to the RTC registers using the peripheral bus clock (PBA) divided by a factor of four or more relative to the HSB clock. 2. The RTC CLKEN bit (bit number 16) of CTRL register is not available The RTC CLKEN bit (bit number 16) of CTRL register is not available. Fix/Workaround Do not use the CLKEN bit of the RTC on Rev B. 31.2.3.13 OCD 1. Stalled memory access instruction writeback fails if followed by a HW breakpoint Consider the following assembly code sequence: A B If a hardware breakpoint is placed on instruction B, and instruction A is a memory access instruction, register file updates from instruction A can be discarded. Fix/Workaround Do not place hardware breakpoints, use software breakpoints instead. Alternatively, place a hardware breakpoint on the instruction before the memory access instruction and then single step over the memory access instruction. 31.2.3.14 Processor and Architecture 1. Local Busto fast GPIO not available on silicon Rev B Local bus is only available for silicon RevE and later. Fix/Workaround Do not use if silicon revison older than F. 2. Memory Protection Unit (MPU) is non functional Memory Protection Unit (MPU) is non functional. Fix/Workaround Do not use the MPU. 3. Bus error should be masked in Debug mode If a bus error occurs during debug mode, the processor will not respond to debug commands through the DINST register. Fix/Workaround A reset of the device will make the CPU respond to debug commands again. 664 32059I-06/2010 AT32UC3B 4. Read Modify Write (RMW) instructions on data outside the internal RAM does not work Read Modify Write (RMW) instructions on data outside the internal RAM does not work. Fix/Workaround Do not perform RMW instructions on data outside the internal RAM. 5. Need two NOPs instruction after instructions masking interrupts The instructions following in the pipeline the instruction masking the interrupt through SR may behave abnormally. Fix/Workaround Place two NOPs instructions after each SSRF or MTSR instruction setting IxM or GM in SR 6. Clock connection table on Rev B Here is the table of Rev B Figure 31-1. Timer/Counter clock connections on RevB Source Internal Name TIMER_CLOCK1 TIMER_CLOCK2 TIMER_CLOCK3 TIMER_CLOCK4 TIMER_CLOCK5 External XC0 XC1 XC2 Connection 32KHz Oscillator PBA Clock / 4 PBA Clock / 8 PBA Clock / 16 PBA Clock / 32 7. Spurious interrupt may corrupt core SR mode to exception If the rules listed in the chapter Masking interrupt requests in peripheral modules' of the AVR32UC Technical Reference Manual are not followed, a spurious interrupt may occur. An interrupt context will be pushed onto the stack while the core SR mode will indicate an exception. A RETE instruction would then corrupt the stack. Fix/Workaround Follow the rules of the AVR32UC Technical Reference Manual. To increase software robustness, if an exception mode is detected at the beginning of an interrupt handler, change the stack interrupt context to an exception context and issue a RETE instruction. 8. CPU cannot operate on a divided slow clock (internal RC oscillator) CPU cannot operate on a divided slow clock (internal RC oscillator). Fix/Workaround Do not run the CPU on a divided slow clock. 9. LDM instruction with PC in the register list and without ++ increments Rp For LDM with PC in the register list: the instruction behaves as if the ++ field is always set, ie the pointer is always updated. This happens even if the ++ field is cleared. Specifically, the increment of the pointer is done in parallel with the testing of R12. Fix/Workaround None. 665 32059I-06/2010 AT32UC3B 10. RETE instruction does not clear SREG[L] from interrupts The RETE instruction clears SREG[L] as expected from exceptions. Fix/Workaround When using the STCOND instruction, clear SREG[L] in the stacked value of SR before returning from interrupts with RETE. 11. Exceptions when system stack is protected by MPU RETS behaves incorrectly when MPU is enabled and MPU is configured so thatsystem stack is not readable in unprivileged mode. Fix/Workaround Workaround 1: Make system stack readable in unprivileged mode, or Workaround 2: Return from supervisor mode using rete instead of rets. This requires: 1. Changing the mode bits from 001b to 110b before issuing the instruction. Updating the mode bits to the desired value must be done using a single mtsr instruction so it is done atomically. Even if this step is described in general as not safe in the UC technical reference guide, it is safe in this very specific case. 2. Execute the RETE instruction. 666 32059I-06/2010 AT32UC3B 32. Datasheet Revision History Please note that the referring page numbers in this section are referred to this document. The referring revision in this section are referring to the document revision. 32.1 Rev. I - 06/2010 1. 2 Updated SPI section. Updated Electrical Characteristics section. 32.2 Rev. H - 10/2009 1. 2 Update datasheet architecture. Add AT32UC3B0512 and AT32UC3B1512 devices description. 32.3 Rev. G - 06/2009 1. 2 Open Drain Mode removed from GPIO section. Updated Errata section. 32.4 Rev. F - 04/2008 1. Updated Errata section. 32.5 Rev. E - 12/2007 1. Updated Memory Protection section. 32.6 Rev. D - 11/2007 1. 2. Updated Processor Architecture section. Updated Electrical Characteristics section. 667 32059I-06/2010 AT32UC3B 32.7 Rev. C - 10/2007 1. 2. 3. 4. 5. 6. 7. 8. 9. Updated Features sections. Updated block diagram with local bus figure Add schematic for HMatrix master/slave connection. Updated Features sections with local bus. Added SPI feature to USART section. Updated USBB section. Updated ADC trigger selection in ADC section. Updated JTAG and Boundary Scan section with programming procedure. Add description for silicon revision D 32.8 Rev. B - 07/2007 1. 2. Updated registered trademarks Updated address page. 32.9 Rev. A - 05/2007 1. Initial revision. 668 32059I-06/2010 AT32UC3B Table of Contents 1 2 Description ............................................................................................... 3 Overview ................................................................................................... 4 2.1Blockdiagram .............................................................................................................4 3 4 Configuration Summary .......................................................................... 5 Package and Pinout ................................................................................. 6 4.1Package ....................................................................................................................6 4.2Peripheral Multiplexing on I/O lines ...........................................................................7 4.3High Drive Current GPIO .........................................................................................10 5 Signals Description ............................................................................... 10 5.1JTAG pins ................................................................................................................13 5.2RESET_N pin ..........................................................................................................14 5.3TWI pins ..................................................................................................................14 5.4GPIO pins ................................................................................................................14 5.5High drive pins .........................................................................................................14 5.6Power Considerations .............................................................................................14 6 Processor and Architecture .................................................................. 17 6.1Features ..................................................................................................................17 6.2AVR32 Architecture .................................................................................................17 6.3The AVR32UC CPU ................................................................................................18 6.4Programming Model ................................................................................................22 6.5Exceptions and Interrupts ........................................................................................26 6.6Module Configuration ..............................................................................................30 7 Memories ................................................................................................ 31 7.1Embedded Memories ..............................................................................................31 7.2Physical Memory Map .............................................................................................31 7.3Peripheral Address Map ..........................................................................................32 7.4CPU Local Bus Mapping .........................................................................................33 8 Boot Sequence ....................................................................................... 34 8.1Starting of clocks .....................................................................................................34 8.2Fetching of initial instructions ..................................................................................34 9 Power Manager (PM) .............................................................................. 35 9.1Features ..................................................................................................................35 669 32059I-06/2010 AT32UC3B 9.2Description ..............................................................................................................35 9.3Block Diagram .........................................................................................................36 9.4Product Dependencies ............................................................................................37 9.5Functional Description .............................................................................................37 9.6User Interface ..........................................................................................................49 10 Real Time Counter (RTC) ...................................................................... 72 10.1Features ................................................................................................................72 10.2Overview ...............................................................................................................72 10.3Block Diagram .......................................................................................................72 10.4Product Dependencies ..........................................................................................72 10.5Functional Description ...........................................................................................73 10.6User Interface ........................................................................................................75 11 Watchdog Timer (WDT) ......................................................................... 84 11.1Features ................................................................................................................84 11.2Overview ...............................................................................................................84 11.3Block Diagram .......................................................................................................84 11.4Product Dependencies ..........................................................................................84 11.5Functional Description ...........................................................................................85 11.6User Interface ........................................................................................................85 12 Interrupt Controller (INTC) .................................................................... 88 12.1Features ................................................................................................................88 12.2Overview ...............................................................................................................88 12.3Block Diagram .......................................................................................................88 12.4Product Dependencies ..........................................................................................89 12.5Functional Description ...........................................................................................89 12.6User Interface ........................................................................................................92 12.7Interrupt Request Signal Map ................................................................................96 13 External Interrupt Controller (EIC) ....................................................... 98 13.1Features ................................................................................................................98 13.2Overview ...............................................................................................................98 13.3Block Diagram .......................................................................................................99 13.4I/O Lines Description .............................................................................................99 13.5Product Dependencies ..........................................................................................99 13.6Functional Description .........................................................................................100 13.7User Interface ......................................................................................................105 670 32059I-06/2010 AT32UC3B 14 Flash Controller (FLASHC) ................................................................. 121 14.1Features ..............................................................................................................121 14.2Overview .............................................................................................................121 14.3Product dependencies .........................................................................................121 14.4Functional description .........................................................................................122 14.5Flash commands .................................................................................................124 14.6General-purpose fuse bits ...................................................................................126 14.7Security bit ...........................................................................................................128 14.8User Interface ......................................................................................................129 14.9Fuses Settings .....................................................................................................137 14.10Module configuration .........................................................................................138 15 HSB Bus Matrix (HMATRIX) ................................................................ 139 15.1Features .............................................................................................................139 15.2Overview .............................................................................................................139 15.3Product Dependencies ........................................................................................139 15.4Functional Description .........................................................................................139 15.5User Interface ......................................................................................................143 15.6Bus Matrix Connections ......................................................................................151 16 Peripheral DMA Controller (PDCA) .................................................... 153 16.1Features ..............................................................................................................153 16.2Overview .............................................................................................................153 16.3Block Diagram .....................................................................................................154 16.4Product Dependencies ........................................................................................154 16.5Functional Description .........................................................................................155 16.6User Interface ......................................................................................................158 16.7Module Configuration ..........................................................................................171 17 General-Purpose Input/Output Controller (GPIO) ............................. 172 17.1Features ..............................................................................................................172 17.2Overview .............................................................................................................172 17.3Block Diagram .....................................................................................................172 17.4Product Dependencies ........................................................................................172 17.5Functional Description .........................................................................................173 17.6User Interface ......................................................................................................177 17.7Programming Examples ......................................................................................192 17.8Module Configuration ..........................................................................................194 671 32059I-06/2010 AT32UC3B 18 Serial Peripheral Interface (SPI) ......................................................... 196 18.1Features ..............................................................................................................196 18.2Overview .............................................................................................................196 18.3Block Diagram .....................................................................................................197 18.4Application Block Diagram ..................................................................................197 18.5Signal Description ..............................................................................................198 18.6Product Dependencies ........................................................................................198 18.7Functional Description .........................................................................................198 18.8User Interface ......................................................................................................208 19 Two-Wire Interface (TWI) ..................................................................... 220 19.1Features ..............................................................................................................220 19.2Overview .............................................................................................................220 19.3List of Abbreviations ............................................................................................221 19.4Block Diagram .....................................................................................................221 19.5Application Block Diagram ..................................................................................222 19.6I/O Lines Description ...........................................................................................222 19.7Product Dependencies ........................................................................................222 19.8Functional Description .........................................................................................223 19.9Modes of Operation .............................................................................................223 19.10Master Mode .....................................................................................................224 19.11Using the Peripheral DMA Controller ................................................................228 19.12Multi-master Mode .............................................................................................236 19.13Slave Mode .......................................................................................................239 19.14User Interface ....................................................................................................247 20 Synchronous Serial Controller (SSC) ................................................ 262 20.1Features .............................................................................................................262 20.2Overview .............................................................................................................262 20.3Block Diagram .....................................................................................................263 20.4Application Block Diagram ..................................................................................263 20.5I/O Lines Description ...........................................................................................264 20.6Product Dependencies ........................................................................................264 20.7Functional Description .........................................................................................264 20.8SSC Application Examples ..................................................................................276 20.9User Interface ......................................................................................................278 21 Universal Synchronous Asynchronous Receiver Transmitter (USART) 672 32059I-06/2010 AT32UC3B 300 21.1Features ..............................................................................................................300 21.2Overview .............................................................................................................300 21.3Block Diagram .....................................................................................................301 21.4I/O Lines Description ..........................................................................................302 21.5Product Dependencies ........................................................................................303 21.6Functional Description .........................................................................................304 21.7User Interface ......................................................................................................340 21.8Module Configuration ..........................................................................................364 22 USB On-The-Go Interface (USBB) ...................................................... 365 22.1Features ..............................................................................................................365 22.2Overview .............................................................................................................365 22.3Block Diagram .....................................................................................................366 22.4Application Block Diagram ..................................................................................368 22.5I/O Lines Description ...........................................................................................370 22.6Product Dependencies ........................................................................................371 22.7Functional Description .........................................................................................372 22.8User Interface ......................................................................................................405 23 Timer/Counter (TC) .............................................................................. 483 23.1Features ..............................................................................................................483 23.2Overview .............................................................................................................483 23.3Block Diagram .....................................................................................................484 23.4I/O Lines Description ...........................................................................................484 23.5Product Dependencies ........................................................................................484 23.6Functional Description .........................................................................................485 23.7User Interface ......................................................................................................500 23.8Module Configuration ..........................................................................................520 24 Pulse Width Modulation Controller (PWM) ........................................ 521 24.1Features ..............................................................................................................521 24.2Description ..........................................................................................................521 24.3Block Diagram .....................................................................................................522 24.4I/O Lines Description ...........................................................................................522 24.5Product Dependencies ........................................................................................523 24.6Functional Description .........................................................................................524 24.7User Interface ......................................................................................................532 673 32059I-06/2010 AT32UC3B 25 Analog-to-Digital Converter (ADC) ..................................................... 547 25.1Features ..............................................................................................................547 25.2Overview .............................................................................................................547 25.3Block Diagram .....................................................................................................548 25.4I/O Lines Description ...........................................................................................548 25.5Product Dependencies ........................................................................................548 25.6Functional Description .........................................................................................549 25.7User Interface ......................................................................................................554 25.8Module Configuration ..........................................................................................567 26 Audio Bitstream DAC (ABDAC) .......................................................... 568 26.1Features ..............................................................................................................568 26.2Overview .............................................................................................................568 26.3Block Diagram .....................................................................................................569 26.4I/O Lines Description ...........................................................................................569 26.5Product Dependencies ........................................................................................569 26.6Functional Description .........................................................................................570 26.7User Interface ......................................................................................................573 27 Programming and Debugging ............................................................ 581 27.1Overview .............................................................................................................581 27.2Service Access Bus .............................................................................................581 27.3On-Chip Debug (OCD) ........................................................................................583 27.4JTAG and Boundary-Scan (JTAG) ......................................................................590 27.5JTAG Instruction Summary .................................................................................598 27.6JTAG Data Registers ..........................................................................................613 27.7SAB address map ...............................................................................................614 28 Electrical Characteristics .................................................................... 615 28.1Absolute Maximum Ratings* ...............................................................................615 28.2DC Characteristics ..............................................................................................616 28.3Regulator Characteristics ....................................................................................618 28.4Analog Characteristics ........................................................................................618 28.5Power Consumption ............................................................................................621 28.6System Clock Characteristics ..............................................................................624 28.7Oscillator Characteristics .....................................................................................625 28.8ADC Characteristics ............................................................................................627 28.9JTAG Characteristics ..........................................................................................628 674 32059I-06/2010 AT32UC3B 28.10SPI Characteristics ............................................................................................629 28.11Flash Memory Characteristics ...........................................................................632 29 Mechanical Characteristics ................................................................. 633 29.1Thermal Considerations ......................................................................................633 29.2Package Drawings ..............................................................................................634 29.3Soldering Profile ..................................................................................................638 30 Ordering Information ........................................................................... 639 31 Errata ..................................................................................................... 640 31.1AT32UC3B0512, AT32UC3B1512 ......................................................................640 31.2AT32UC3B0256, AT32UC3B0128, AT32UC3B064, AT32UC3B1256, AT32UC3B1128, AT32UC3B164 652 32 Datasheet Revision History ................................................................ 667 32.1Rev. I - 06/2010 ..................................................................................................667 32.2Rev. H - 10/2009 ................................................................................................667 32.3Rev. G - 06/2009 ................................................................................................667 32.4Rev. F - 04/2008 .................................................................................................667 32.5Rev. E - 12/2007 .................................................................................................667 32.6Rev. D - 11/2007 ................................................................................................667 32.7Rev. C - 10/2007 ................................................................................................668 32.8Rev. B - 07/2007 .................................................................................................668 32.9Rev. A - 05/2007 .................................................................................................668 675 32059I-06/2010 Headquarters Atmel Corporation 2325 Orchard Parkway San Jose, CA 95131 USA Tel: 1(408) 441-0311 Fax: 1(408) 487-2600 International Atmel Asia Unit 1-5 & 16, 19/F BEA Tower, Millennium City 5 418 Kwun Tong Road Kwun Tong, Kowloon Hong Kong Tel: (852) 2245-6100 Fax: (852) 2722-1369 Atmel Europe Le Krebs 8, Rue Jean-Pierre Timbaud BP 309 78054 Saint-Quentin-enYvelines Cedex France Tel: (33) 1-30-60-70-00 Fax: (33) 1-30-60-71-11 Atmel Japan 9F, Tonetsu Shinkawa Bldg. 1-24-8 Shinkawa Chuo-ku, Tokyo 104-0033 Japan Tel: (81) 3-3523-3551 Fax: (81) 3-3523-7581 Product Contact Web Site www.atmel.com Technical Support avr32@atmel.com Sales Contact www.atmel.com/contacts Literature Requests www.atmel.com/literature Disclaimer: The information in this document is provided in connection with Atmel products. No license, express or implied, by estoppel or otherwise, to any intellectual property right is granted by this document or in connection with the sale of Atmel products. EXCEPT AS SET FORTH IN ATMEL'S TERMS AND CONDITIONS OF SALE LOCATED ON ATMEL'S WEB SITE, ATMEL ASSUMES NO LIABILITY WHATSOEVER AND DISCLAIMS ANY EXPRESS, IMPLIED OR STATUTORY WARRANTY RELATING TO ITS PRODUCTS INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT. IN NO EVENT SHALL ATMEL BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE, SPECIAL OR INCIDENTAL DAMAGES (INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS OF PROFITS, BUSINESS INTERRUPTION, OR LOSS OF INFORMATION) ARISING OUT OF THE USE OR INABILITY TO USE THIS DOCUMENT, EVEN IF ATMEL HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Atmel makes no representations or warranties with respect to the accuracy or completeness of the contents of this document and reserves the right to make changes to specifications and product descriptions at any time without notice. Atmel does not make any commitment to update the information contained herein. Unless specifically provided otherwise, Atmel products are not suitable for, and shall not be used in, automotive applications. Atmel's products are not intended, authorized, or warranted for use as components in applications intended to support or sustain life. (c) 2010 Atmel Corporation. All rights reserved. Atmel (R), Atmel logo and combinations thereof, AVR(R), AVR(R) logo and others are registered trademarks or trademarks of Atmel Corporation or its subsidiaries. Other terms and product names may be trademarks of others. 32059I-06/2010 |
Price & Availability of AT32UC3B0128-Z2UR
 |
|
|
|
|
All Rights Reserved © IC-ON-LINE 2003 - 2022 |
| [Add Bookmark] [Contact Us] [Link exchange] [Privacy policy] |
|
Mirror Sites : [www.datasheet.hk]
[www.maxim4u.com] [www.ic-on-line.cn]
[www.ic-on-line.com] [www.ic-on-line.net]
[www.alldatasheet.com.cn]
[www.gdcy.com]
[www.gdcy.net] |