Features • High-performance, Low-power 32-bit Atmel® AVR® Microcontroller • • • • • • • • • • • • • • • • • – Compact Single-cycle RISC Instruction Set Including DSP Instructions – Read-modify-write Instructions and Atomic Bit Manipulation – Performance • Up to 64DMIPS Running at 50MHz from Flash (1 Flash Wait State) • Up to 36DMIPS Running at 25MHz from Flash (0 Flash Wait State) – Memory Protection Unit (MPU) • Secure Access Unit (SAU) providing User-defined Peripheral Protection picoPower® Technology for Ultra-low Power Consumption Multi-hierarchy Bus System – High-performance Data Transfers on Separate Buses for Increased Performance – 12 Peripheral DMA Channels Improve Speed for Peripheral Communication Internal High-speed Flash – 64Kbytes, 32Kbytes, and 16Kbytes Versions – Single-cycle Access up to 25MHz – FlashVault Technology Allows Pre-programmed Secure Library Support for End User Applications – Prefetch Buffer Optimizing Instruction Execution at Maximum Speed – 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 – 16Kbytes (64Kbytes and 32Kbytes Flash), or 8Kbytes (16Kbytes Flash) Interrupt Controller (INTC) – Autovectored Low-latency Interrupt Service with Programmable Priority External Interrupt Controller (EIC) Peripheral Event System for Direct Peripheral to Peripheral Communication System Functions – Power and Clock Manager – SleepWalking Power Saving Control – Internal System RC Oscillator (RCSYS) – 32KHz Oscillator – Multipurpose Oscillator and Digital Frequency Locked Loop (DFLL) Windowed Watchdog Timer (WDT) Asynchronous Timer (AST) with Real-time Clock Capability – Counter or Calendar Mode Supported Frequency Meter (FREQM) for Accurate Measuring of Clock Frequency Six 16-bit Timer/Counter (TC) Channels – External Clock Inputs, PWM, Capture, and Various Counting Capabilities 36 PWM Channels (PWMA) – 8-bit PWM with a Source Clock up to 150MHz Four Universal Synchronous/Asynchronous Receiver/Transmitters (USART) – Independent Baudrate Generator, Support for SPI – Support for Hardware Handshaking One Master/Slave Serial Peripheral Interfaces (SPI) with Chip Select Signals – Up to 15 SPI Slaves can be Addressed Two Master and Two Slave Two-wire Interface (TWI), 400kbit/s I2C-compatible One 8-channel Analog-to-digital Converter (ADC) with up to 12 Bits Resolution – Internal Temperature Sensor 32-bit Atmel AVR Microcontroller AT32UC3L064 AT32UC3L032 AT32UC3L016 Summary 32099IS–01/2012 AT32UC3L016/32/64 • Eight Analog Comparators (AC) with Optional Window Detection • Capacitive Touch (CAT) Module • • • • • – Hardware-assisted Atmel® AVR® QTouch® and Atmel® AVR® QMatrix Touch Acquisition – Supports QTouch and QMatrix Capture from Capacitive Touch Sensors QTouch Library Support – Capacitive Touch Buttons, Sliders, and Wheels – QTouch and QMatrix Acquisition On-chip Non-intrusive Debug System – Nexus Class 2+, Runtime Control, Non-intrusive Data and Program Trace – aWire Single-pin Programming Trace and Debug Interface Muxed with Reset Pin – NanoTrace Provides Trace Capabilities through JTAG or aWire Interface 48-pin TQFP/QFN/TLLGA (36 GPIO Pins) Five High-drive I/O Pins Single 1.62-3.6 V Power Supply 2 32099IS–01/2012 AT32UC3L016/32/64 1. Description The Atmel® AVR® AT32UC3L016/32/64 is a complete system-on-chip microcontroller based on the AVR32 UC RISC processor running at frequencies up to 50MHz. 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 and real-time operating systems. The Secure Access Unit (SAU) is used together with the MPU to provide the required security and integrity. Higher computation capability is achieved using a rich set of DSP instructions. The AT32UC3L016/32/64 embeds state-of-the-art picoPower technology for ultra-low power consumption. Combined power control techniques are used to bring active current consumption down to 165 µA/MHz, and leakage down to 9nA while still retaining a bank of backup registers. The device allows a wide range of trade-offs between functionality and power consumption, giving the user the ability to reach the lowest possible power consumption with the feature set required for the application. The Peripheral Direct Memory Access (DMA) controller enables data transfers between peripherals and memories without processor involvement. The Peripheral DMA controller drastically reduces processing overhead when transferring continuous and large data streams. The AT32UC3L016/32/64 incorporates on-chip Flash and SRAM memories for secure and fast access. The FlashVault technology allows secure libraries to be programmed into the device. The secure libraries can be executed while the CPU is in Secure State, but not read by nonsecure software in the device. The device can thus be shipped to end customers, who will be able to program their own code into the device to access the secure libraries, but without risk of compromising the proprietary secure code. 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. The Peripheral Event System allows peripherals to receive, react to, and send peripheral events without CPU intervention. Asynchronous interrupts allow advanced peripheral operation in low power sleep modes. The Power Manager (PM) improves design flexibility and security. The Power Manager supports SleepWalking functionality, by which a module can be selectively activated based on peripheral events, even in sleep modes where the module clock is stopped. Power monitoring is supported by on-chip Power-on Reset (POR), Brown-out Detector (BOD), and Supply Monitor (SM). The device features several oscillators, such as Digital Frequency Locked Loop (DFLL), Oscillator 0 (OSC0), and system RC oscillator (RCSYS). Either of these oscillators can be used as source for the system clock. The DFLL is a programmable internal oscillator from 40 to 150MHz. It can be tuned to a high accuracy if an accurate refernce clock is running, e.g. the 32 KHz crystal oscillator. The Watchdog Timer (WDT) will reset the device unless it is periodically serviced by the software. This allows the device to recover from a condition that has caused the system to be unstable. The Asynchronous Timer (AST) combined with the 32KHz crystal oscillator supports powerful real-time clock capabilities, with a maximum timeout of up to 136 years. The AST can operate in counter mode or calendar mode. 3 32099IS–01/2012 AT32UC3L016/32/64 The Frequency Meter (FREQM) allows accurate measuring of a clock frequency by comparing it to a known reference clock. The device includes six identical 16-bit Timer/Counter (TC) channels. Each channel can be independently programmed to perform frequency measurement, event counting, interval measurement, pulse generation, delay timing, and pulse width modulation. The Pulse Width Modulation controller (PWMA) provides 8-bit PWM channels which can be synchronized and controlled from a common timer. One PWM channel is available for each I/O pin on the device, enabling applications that require multiple PWM outputs, such as LCD backlight control. The PWM channels can operate independently, with duty cycles set independently from each other, or in interlinked mode, with multiple channels changed at the same time. The AT32UC3L016/32/64 also features many communication interfaces, like USART, SPI, and TWI, for communication intensive applications. The USART supports different communication modes, like SPI Mode and LIN Mode. A general purpose 8-channel ADC is provided, as well as eight analog comparators (AC). The ADC can operate in 10-bit mode at full speed or in enhanced mode at reduced speed, offering up to 12-bit resolution. The ADC also provides an internal temperature sensor input channel. The analog comparators can be paired to detect when the sensing voltage is within or outside the defined reference window. The Capacitive Touch (CAT) module senses touch on external capacitive touch sensors, using the QTouch technology. Capacitive touch sensors use no external mechanical components, unlike normal push buttons, and therefore demand less maintenance in the user application. The CAT module allows up to 17 touch sensors, or up to 16 by 8 matrix sensors to be interfaced. One touch sensor can be configured to operate autonomously without software interaction, allowing wakeup from sleep modes when activated. 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 includes fully debounced reporting of touch keys as well as Adjacent Key Suppression® (AKS®) 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. The AT32UC3L016/32/64 integrates a class 2+ Nexus 2.0 On-chip Debug (OCD) System, with non-intrusive real-time trace and full-speed read/write memory access, in addition to basic runtime control. The NanoTrace interface enables trace feature for aWire- or JTAG-based debuggers. The single-pin aWire interface allows all features available through the JTAG interface to be accessed through the RESET pin, allowing the JTAG pins to be used for GPIO or peripherals. 4 32099IS–01/2012 AT32UC3L016/32/64 2. Overview Block Diagram DATAOUT aWire NEXUS CLASS 2+ OCD MEMORY PROTECTION UNIT INSTR INTERFACE DATA INTERFACE M M M SAU S HSB-PB BRIDGE B GENERALPURPOSE I/Os S POWER MANAGER CLOCK CONTROLLER SLEEP CONTROLLER RESET CONTROLLER 64/32/16 KB FLASH M S CONFIGURATION PA PB S HIGH SPEED BUS MATRIX S/M LOCAL BUS 16/8 KB SRAM REGISTERS BUS PERIPHERAL DMA CONTROLLER HSB-PB BRIDGE A DMA RESET_N JTAG INTERFACE LOCAL BUS INTERFACE CAPACITIVE TOUCH MODULE DMA EVTO_N TCK TDO TDI TMS AVR32UC CPU FLASH CONTROLLER MCKO MDO[5..0] MSEO[1..0] EVTI_N MEMORY INTERFACE Block Diagram USART0 USART1 USART2 USART3 DMA Figure 2-1. SPI DMA 2.1 TWI MASTER 0 TWI MASTER 1 DIS VDIVEN CSA[16:0] CSB[16:0] SMP SYNC RXD TXD CLK RTS, CTS RCSYS RC32K XIN32 XOUT32 OSC32K XIN0 XOUT0 OSC0 SYSTEM CONTROL INTERFACE DFLL INTERRUPT CONTROLLER EXTINT[5..1] NMI PWMA[35..0] NPCS[3..0] TWCK TWD TWALM TWCK DMA RC120M TWI SLAVE 0 TWI SLAVE 1 8-CHANNEL ADC INTERFACE FREQUENCY METER TRIGGER AD[8..0] A[2..0] TIMER/COUNTER 0 TIMER/COUNTER 1 B[2..0] CLK[2..0] ASYNCHRONOUS TIMER WATCHDOG TIMER TWALM ADVREFP EXTERNAL INTERRUPT CONTROLLER PWM CONTROLLER TWD PA PB ADP[1..0] DMA RC32OUT MISO, MOSI GENERAL PURPOSE I/Os SCK GCLK[4..0] AC INTERFACE GLUE LOGIC CONTROLLER ACBP[3..0] ACBN[3..0] ACAP[3..0] ACAN[3..0] ACREFN OUT[1:0] IN[7..0] 5 32099IS–01/2012 AT32UC3L016/32/64 2.2 Configuration Summary Table 2-1. Configuration Summary Feature AT32UC3L064 AT32UC3L032 AT32UC3L016 Flash 64KB 32KB 16KB SRAM 16KB 16KB 8KB GPIO 36 High-drive pins 5 External Interrupts 6 TWI 2 USART 4 Peripheral DMA Channels 12 Peripheral Event System 1 SPI 1 Asynchronous Timers 1 Timer/Counter Channels 6 PWM channels 36 Frequency Meter 1 Watchdog Timer 1 Power Manager 1 Secure Access Unit 1 Glue Logic Controller 1 Oscillators ADC Digital Frequency Locked Loop 40-150MHz (DFLL) Crystal Oscillator 3-16MHz (OSC0) Crystal Oscillator 32KHz (OSC32K) RC Oscillator 120MHz (RC120M) RC Oscillator 115kHz (RCSYS) RC Oscillator 32kHz (RC32K) 8-channel 12-bit Temperature Sensor 1 Analog Comparators 8 Capacitive Touch Module 1 JTAG 1 aWire 1 Max Frequency Packages 50MHz TQFP48/QFN48/TLLGA48 6 32099IS–01/2012 AT32UC3L016/32/64 3. Package and Pinout 3.1 Package The device pins are multiplexed with peripheral functions as described in Section 3.2. TQFP48/QFN48 Pinout 36 35 34 33 32 31 30 29 28 27 26 25 PA14 VDDANA ADVREFP GNDANA PB08 PB07 PB06 PB09 PA04 PA11 PA13 PA20 Figure 3-1. PA15 PA16 PA17 PA19 PA18 VDDIO GND PB11 GND PA10 PA12 VDDIO 37 38 39 40 41 42 43 44 45 46 47 48 24 23 22 21 20 19 18 17 16 15 14 13 PA21 PB10 RESET_N PB04 PB05 GND VDDCORE VDDIN PB01 PA07 PA01 PA02 12 11 10 9 8 7 6 5 4 3 2 1 PA05 PA00 PA06 PA22 PB03 PB02 PB00 PB12 PA03 PA08 PA09 GND 7 32099IS–01/2012 AT32UC3L016/32/64 TLLGA48 Pinout 37 36 35 34 33 32 31 30 29 28 27 26 25 PA15 PA14 VDDANA ADVREFP GNDANA PB08 PB07 PB06 PB09 PA04 PA11 PA13 PA20 Figure 3-2. PA16 PA17 PA19 PA18 VDDIO GND PB11 GND PA10 PA12 VDDIO 24 23 22 21 20 19 18 17 16 15 14 38 39 40 41 42 43 44 45 46 47 48 PA21 PB10 RESET_N PB04 PB05 GND VDDCORE VDDIN PB01 PA07 PA01 13 12 11 10 9 8 7 6 5 4 3 2 1 PA02 PA05 PA00 PA06 PA22 PB03 PB02 PB00 PB12 PA03 PA08 PA09 GND 8 32099IS–01/2012 AT32UC3L016/32/64 3.2 Peripheral Multiplexing on I/O lines 3.2.1 Multiplexed signals Each GPIO line can be assigned to one of the peripheral functions.The following table describes the peripheral signals multiplexed to the GPIO lines. Table 3-1. GPIO Controller Function Multiplexing 48pin PIN G P I O 11 PA00 0 14 PA01 13 GPIO Function Pin Type A B C VDDIO Normal I/O USART0 TXD USART1 RTS SPI NPCS[2] 1 VDDIO Normal I/O USART0 RXD USART1 CTS SPI NPCS[3] USART1 CLK PWMA PWMA[1] PA02 2 VDDIO Highdrive I/O USART0 RTS ADCIFB TRIGGER USART2 TXD TC0 A0 4 PA03 3 VDDIO Normal I/O USART0 CTS SPI NPCS[1] USART2 TXD 28 PA04 4 VDDIO Normal I/O SPI MISO TWIMS0 TWCK 12 PA05 5 VDDIO Normal I/O (TWI) SPI MOSI 10 PA06 6 VDDIO Highdrive I/O, 5V tolerant 15 PA07 7 VDDIO 3 PA08 8 2 PA09 46 Supply D E F G H SCIF GCLK[0] CAT CSA[2] ACIFB ACAP[0] TWIMS0 TWALM CAT CSA[1] PWMA PWMA[2] ACIFB ACBP[0] USART0 CLK CAT CSA[3] TC0 B0 PWMA PWMA[3] ACIFB ACBN[3] USART0 CLK CAT CSB[3] USART1 RXD TC0 B1 PWMA PWMA[4] ACIFB ACBP[1] TWIMS1 TWCK USART1 TXD TC0 A1 PWMA PWMA[5] ACIFB ACBN[0] SPI SCK USART2 TXD USART1 CLK TC0 B0 PWMA PWMA[6] Normal I/O (TWI) SPI NPCS[0] USART2 RXD TWIMS1 TWALM TWIMS0 TWCK PWMA PWMA[7] VDDIO Highdrive I/O USART1 TXD SPI NPCS[2] TC0 A2 ADCIFB ADP[0] PWMA PWMA[8] 9 VDDIO Highdrive I/O USART1 RXD SPI NPCS[3] TC0 B2 ADCIFB ADP[1] PWMA PWMA[9] SCIF GCLK[2] EIC EXTINT[1] CAT CSB[4] PA10 10 VDDIO Normal I/O TWIMS0 TWD PWMA PWMA[10] ACIFB ACAP[1] SCIF GCLK[2] CAT CSA[5] 27 PA11 11 VDDIN Normal I/O 47 PA12 12 VDDIO Normal I/O 26 PA13 13 VDDIN Normal I/O 36 PA14 14 VDDIO 37 PA15 15 38 PA16 39 41 PWMA PWMA[0] TC0 A0 ACIFB ACAN[0] CAT CSA[7] TWIMS0 TWD CAT CSB[7] SCIF GCLK[1] CAT CSB[1] EIC EXTINT[0] CAT CSB[2] CAT CSA[4] PWMA PWMA[11] USART2 CLK TC0 CLK1 CAT SMP PWMA PWMA[12] ACIFB ACAN[1] SCIF GCLK[3] CAT CSB[5] GLOC OUT[0] GLOC IN[7] TC0 A0 SCIF GCLK[2] PWMA PWMA[13] CAT SMP EIC EXTINT[2] CAT CSA[0] Normal I/O ADCIFB AD[0] TC0 CLK2 USART2 RTS CAT SMP PWMA PWMA[14] SCIF GCLK[4] CAT CSA[6] VDDIO Normal I/O ADCIFB AD[1] TC0 CLK1 GLOC IN[6] PWMA PWMA[15] CAT SYNC EIC EXTINT[3] CAT CSB[6] 16 VDDIO Normal I/O ADCIFB AD[2] TC0 CLK0 GLOC IN[5] PWMA PWMA[16] ACIFB ACREFN EIC EXTINT[4] CAT CSA[8] PA17 17 VDDIO Normal I/O (TWI) TWIMS1 TWD PWMA PWMA[17] CAT SMP CAT DIS CAT CSB[8] PA18 18 VDDIO Normal I/O GLOC IN[4] PWMA PWMA[18] CAT SYNC EIC EXTINT[5] CAT CSB[0] TC0 A1 ADCIFB AD[4] TC0 B1 USART2 CTS 9 32099IS–01/2012 AT32UC3L016/32/64 Table 3-1. GPIO Controller Function Multiplexing 40 PA19 19 VDDIO Normal I/O ADCIFB AD[5] TC0 A2 TWIMS1 TWALM PWMA PWMA[19] CAT SYNC 25 PA20 20 VDDIN Normal I/O USART2 TXD TC0 A1 GLOC IN[3] PWMA PWMA[20] SCIF RC32OUT USART2 RXD TWIMS0 TWD TC0 B1 ADCIFB TRIGGER PWMA PWMA[21] PWMA PWMAOD[21] 24 PA21 21 VDDIN Normal I/O (TWI, 5V tolerant SMBus) 9 PA22 22 VDDIO Normal I/O USART0 CTS USART2 CLK TC0 B2 CAT SMP PWMA PWMA[22] ACIFB ACBN[2] 6 PB00 32 VDDIO Normal I/O USART3 TXD ADCIFB ADP[0] SPI NPCS[0] TC0 A1 PWMA PWMA[23] ACIFB ACAP[2] 16 PB01 33 VDDIO Highdrive I/O USART3 RXD ADCIFB ADP[1] SPI SCK TC0 B1 PWMA PWMA[24] 7 PB02 34 VDDIO Normal I/O USART3 RTS USART3 CLK SPI MISO TC0 A2 PWMA PWMA[25] 8 PB03 35 VDDIO Normal I/O USART3 CTS USART3 CLK SPI MOSI TC0 B2 VDDIN Normal I/O (TWI, 5V tolerant SMBus) TC1 A0 USART1 RTS USART1 CLK TC1 B0 USART1 CTS 21 PB04 36 CAT CSA[10] CAT CSA[12] SCIF GCLK[0] CAT SMP CAT CSB[10] TC1 A0 CAT CSA[9] TC1 A1 CAT CSB[9] ACIFB ACAN[2] SCIF GCLK[1] CAT CSB[11] PWMA PWMA[26] ACIFB ACBP[2] TC1 A2 CAT CSA[11] TWIMS0 TWALM PWMA PWMA[27] PWMA PWMAOD[27] TWIMS1 TWCK CAT CSA[14] USART1 CLK TWIMS0 TWCK PWMA PWMA[28] PWMA PWMAOD[28] SCIF GCLK[3] CAT CSB[14] 20 PB05 37 VDDIN Normal I/O (TWI, 5V tolerant SMBus) 30 PB06 38 VDDIO Normal I/O TC1 A1 USART3 TXD ADCIFB AD[6] GLOC IN[2] PWMA PWMA[29] ACIFB ACAN[3] EIC EXTINT[0] CAT CSB[13] 31 PB07 39 VDDIO Normal I/O TC1 B1 USART3 RXD ADCIFB AD[7] GLOC IN[1] PWMA PWMA[30] ACIFB ACAP[3] EIC EXTINT[1] CAT CSA[13] 32 PB08 40 VDDIO Normal I/O TC1 A2 USART3 RTS ADCIFB AD[8] GLOC IN[0] PWMA PWMA[31] CAT SYNC EIC EXTINT[2] CAT CSB[12] 29 PB09 41 VDDIO Normal I/O TC1 B2 USART3 CTS USART3 CLK PWMA PWMA[32] ACIFB ACBN[1] EIC EXTINT[3] CAT CSB[15] 23 PB10 42 VDDIN Normal I/O TC1 CLK0 USART1 TXD USART3 CLK EIC EXTINT[4] CAT CSB[16] 44 PB11 43 VDDIO Normal I/O TC1 CLK1 USART1 RXD 5 PB12 44 VDDIO Normal I/O TC1 CLK2 TWIMS1 TWALM GLOC OUT[1] PWMA PWMA[33] ADCIFB TRIGGER PWMA PWMA[34] CAT VDIVEN EIC EXTINT[5] CAT CSA[16] CAT SYNC PWMA PWMA[35] ACIFB ACBP[3] SCIF GCLK[4] CAT CSA[15] See Section 3.3 for a description of the various peripheral signals. Refer to ”Electrical Characteristics” on page 41 for a description of the electrical properties of the pin types used. 3.2.1.1 TWI, 5V Tolerant, and SMBUS Pins Some Normal I/O pins offer TWI, 5V Tolerant, and SMBUS features. These features are only available when either of the TWI functions or the PWMAOD function in the PWMA are selected for these pins. 10 32099IS–01/2012 AT32UC3L016/32/64 Refer to the ”TWI Pin Characteristics(1)” on page 49 for a description of the electrical properties of the TWI, 5V Tolerant, and SMBUS pins. 3.2.2 Peripheral Functions Each GPIO line can be assigned to one of several peripheral functions. The following table describes how the various peripheral functions are selected. The last listed function has priority in case multiple functions are enabled on the same pin. Table 3-2. 3.2.3 Function Description GPIO Controller Function multiplexing GPIO and GPIO peripheral selection A to H Nexus OCD AUX port connections OCD trace system aWire DATAOUT aWire output in two-pin mode JTAG port connections JTAG debug port Oscillators OSC0, OSC32 JTAG Port Connections If the JTAG is enabled, the JTAG will take control over a number of pins, irrespectively of the I/O Controller configuration. Table 3-3. 3.2.4 Peripheral Functions JTAG Pinout 48-pin Pin Name JTAG Pin 11 PA00 TCK 14 PA01 TMS 13 PA02 TDO 4 PA03 TDI 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 I/O Controller configuration. Two different OCD trace pin mappings are possible, depending on the configuration of the OCD AXS register. For details, see the AVR32 UC Technical Reference Manual. Table 3-4. Nexus OCD AUX Port Connections Pin AXS=1 AXS=0 EVTI_N PA05 PB08 MDO[5] PA10 PB00 MDO[4] PA18 PB04 MDO[3] PA17 PB05 MDO[2] PA16 PB03 11 32099IS–01/2012 AT32UC3L016/32/64 Table 3-4. 3.2.5 Pin AXS=1 AXS=0 MDO[1] PA15 PB02 MDO[0] PA14 PB09 EVTO_N PA04 PA04 MCKO PA06 PB01 MSEO[1] PA07 PB11 MSEO[0] PA11 PB12 Oscillator Pinout The oscillators are not mapped to the normal GPIO functions and their muxings are controlled by registers in the System Control Interface (SCIF). Please refer to the SCIF chapter for more information about this. Table 3-5. 3.2.6 Nexus OCD AUX Port Connections Oscillator Pinout 48-pin Pin Oscillator Function 3 PA08 XIN0 46 PA10 XIN32 26 PA13 XIN32_2 2 PA09 XOUT0 47 PA12 XOUT32 25 PA20 XOUT32_2 Other Functions The functions listed in Table 3-6 are not mapped to the normal GPIO functions. The aWire DATA pin will only be active after the aWire is enabled. The aWire DATAOUT pin will only be active after the aWire is enabled and the 2_PIN_MODE command has been sent. The WAKE_N pin is always enabled. Please refer to Section 6.1.4 on page 40 for constraints on the WAKE_N pin. Table 3-6. Other Functions 48-pin Pin Function 27 PA11 WAKE_N 22 RESET_N aWire DATA 11 PA00 aWire DATAOUT 12 32099IS–01/2012 AT32UC3L016/32/64 3.3 Signal Descriptions The following table gives details on signal name classified by peripheral. Table 3-7. Signal Descriptions List Signal Name Function Type Active Level Comments Analog Comparator Interface - ACIFB ACAN3 - ACAN0 Negative inputs for comparators "A" Analog ACAP3 - ACAP0 Positive inputs for comparators "A" Analog ACBN3 - ACBN0 Negative inputs for comparators "B" Analog ACBP3 - ACBP0 Positive inputs for comparators "B" Analog ACREFN Common negative reference Analog ADC Interface - ADCIFB AD8 - AD0 Analog Signal Analog ADP1 - ADP0 Drive Pin for resistive touch screen Output TRIGGER External trigger Input aWire - AW DATA aWire data I/O DATAOUT aWire data output for 2-pin mode I/O Capacitive Touch Module - CAT CSA16 - CSA0 Capacitive Sense A I/O CSB16 - CSB0 Capacitive Sense B I/O DIS Discharge current control Analog SMP SMP signal Output SYNC Synchronize signal VDIVEN Voltage divider enable Input Output External Interrupt Controller - EIC NMI Non-Maskable Interrupt Input EXTINT5 - EXTINT1 External interrupt Input Glue Logic Controller - GLOC IN7 - IN0 Inputs to lookup tables OUT1 - OUT0 Outputs from lookup tables Input Output JTAG module - JTAG TCK Test Clock Input TDI Test Data In Input TDO Test Data Out TMS Test Mode Select Output Input 13 32099IS–01/2012 AT32UC3L016/32/64 Table 3-7. Signal Descriptions List Power Manager - PM RESET_N Reset Input Low Pulse Width Modulation Controller - PWMA PWMA35 - PWMA0 PWMA channel waveforms Output PWMAOD35 PWMAOD0 PWMA channel waveforms, open drain mode Output Not all channels support open drain mode System Control Interface - SCIF GCLK4 - GCLK0 Generic Clock Output Output RC32OUT RC32K output at startup Output XIN0 Crystal 0 Input Analog/ Digital XIN32 Crystal 32 Input (primary location) Analog/ Digital XIN32_2 Crystal 32 Input (secondary location) Analog/ Digital XOUT0 Crystal 0 Output Analog XOUT32 Crystal 32 Output (primary location) Analog XOUT32_2 Crystal 32 Output (secondary location) Analog Serial Peripheral Interface - SPI MISO Master In Slave Out I/O MOSI Master Out Slave In I/O NPCS3 - NPCS0 SPI Peripheral Chip Select I/O SCK Clock I/O Low Timer/Counter - TC0, TC1 A0 Channel 0 Line A I/O A1 Channel 1 Line A I/O A2 Channel 2 Line A I/O B0 Channel 0 Line B I/O B1 Channel 1 Line B I/O B2 Channel 2 Line B I/O CLK0 Channel 0 External Clock Input Input CLK1 Channel 1 External Clock Input Input CLK2 Channel 2 External Clock Input Input Two-wire Interface - TWIMS0, TWIMS1 TWALM SMBus SMBALERT I/O TWCK Two-wire Serial Clock I/O TWD Two-wire Serial Data I/O Low Universal Synchronous/Asynchronous Receiver/Transmitter - USART0, USART1, USART2, USART3 14 32099IS–01/2012 AT32UC3L016/32/64 Table 3-7. Signal Descriptions List CLK Clock CTS Clear To Send RTS Request To Send RXD Receive Data Input TXD Transmit Data Output Note: I/O Input Low Output Low 1. ADCIFB: AD3 does not exist. Table 3-8. Signal Description List, Continued Signal Name Function Type Active Level Comments Power VDDCORE Core Power Supply / Voltage Regulator Output Power Input/Output 1.62V to 1.98V VDDIO I/O Power Supply Power Input 1.62V to 3.6V. VDDIO should always be equal to or lower than VDDIN. VDDANA Analog Power Supply Power Input 1.62V to 1.98V ADVREFP Analog Reference Voltage Power Input 1.62V to 1.98V VDDIN Voltage Regulator Input Power Input 1.62V to 3.6V (1) GNDANA Analog Ground Ground GND Ground Ground Auxiliary Port - AUX MCKO Trace Data Output Clock Output MDO5 - MDO0 Trace Data Output Output MSEO1 - MSEO0 Trace Frame Control Output EVTI_N Event In EVTO_N Event Out Input Low Output Low General Purpose I/O pin PA22 - PA00 Parallel I/O Controller I/O Port 0 I/O PB12 - PB00 Parallel I/O Controller I/O Port 1 I/O 1. See Section 6.1 on page 36 15 32099IS–01/2012 AT32UC3L016/32/64 3.4 3.4.1 I/O Line Considerations JTAG Pins The JTAG is enabled if TCK is low while the RESET_N pin is released. The TCK, TMS, and TDI pins have pull-up resistors when JTAG is enabled. The TCK pin always has pull-up enabled during reset. The TDO pin is an output, driven at VDDIO, and has no pull-up resistor. The JTAG pins can be used as GPIO pins and multiplexed with peripherals when the JTAG is disabled. Please refer to Section 3.2.3 on page 11 for the JTAG port connections. 3.4.2 PA00 Note that PA00 is multiplexed with TCK. PA00 GPIO function must only be used as output in the application. 3.4.3 RESET_N Pin The RESET_N pin is a schmitt input and integrates a permanent pull-up resistor to VDDIN. As the product integrates a power-on reset detector, the RESET_N pin can be left unconnected in case no reset from the system needs to be applied to the product. The RESET_N pin is also used for the aWire debug protocol. When the pin is used for debugging, it must not be driven by external circuitry. 3.4.4 TWI Pins PA21/PB04/PB05 When these pins are used for TWI, the pins are open-drain outputs with slew-rate limitation and inputs with spike filtering. When used as GPIO pins or used for other peripherals, the pins have the same characteristics as other GPIO pins. Selected pins are also SMBus compliant (refer to Section 3.2 on page 9). As required by the SMBus specification, these pins provide no leakage path to ground when the AT32UC3L016/32/64 is powered down. This allows other devices on the SMBus to continue communicating even though the AT32UC3L016/32/64 is not powered. After reset a TWI function is selected on these pins instead of the GPIO. Please refer to the GPIO Module Configuration chapter for details. 3.4.5 TWI Pins PA05/PA07/PA17 When these pins are used for TWI, the pins are open-drain outputs with slew-rate limitation and inputs with spike filtering. When used as GPIO pins or used for other peripherals, the pins have the same characteristics as other GPIO pins. After reset a TWI function is selected on these pins instead of the GPIO. Please refer to the GPIO Module Configuration chapter for details. 3.4.6 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 Controller. After reset, I/O lines default as inputs with pull-up resistors disabled, except PA00. PA20 selects SCIF-RC32OUT (GPIO Function F) as default enabled after reset. 3.4.7 High-Drive Pins The five pins PA02, PA06, PA08, PA09, and PB01 have high-drive output capabilities. Refer to Section 7. on page 41 for electrical characteristics. 16 32099IS–01/2012 AT32UC3L016/32/64 3.4.8 RC32OUT Pin 3.4.8.1 Clock output at startup After power-up, the clock generated by the 32kHz RC oscillator (RC32K) will be output on PA20, even when the device is still reset by the Power-On Reset Circuitry. This clock can be used by the system to start other devices or to clock a switching regulator to rise the power supply voltage up to an acceptable value. The clock will be available on PA20, but will be disabled if one of the following conditions are true: • PA20 is configured to use a GPIO function other than F (SCIF-RC32OUT) • PA20 is configured as a General Purpose Input/Output (GPIO) • The bit FRC32 in the Power Manager PPCR register is written to zero (refer to the Power Manager chapter) The maximum amplitude of the clock signal will be defined by VDDIN. Once the RC32K output on PA20 is disabled it can never be enabled again. 3.4.8.2 3.4.9 XOUT32_2 function PA20 selects RC32OUT as default enabled after reset. This function is not automatically disabled when the user enables the XOUT32_2 function on PA20. This disturbs the oscillator and may result in the wrong frequency. To avoid this, RC32OUT must be disabled when XOUT32_2 is enabled. ADC Input Pins These pins are regular I/O pins powered from the VDDIO. However, when these pins are used for ADC inputs, the voltage applied to the pin must not exceed 1.98V. Internal circuitry ensures that the pin cannot be used as an analog input pin when the I/O drives to VDD. When the pins are not used for ADC inputs, the pins may be driven to the full I/O voltage range. 17 32099IS–01/2012 AT32UC3L016/32/64 4. Processor and Architecture Rev: 2.1.0.0 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. 4.1 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 extension with saturating arithmetic, and a wide variety of multiply instructions • 3-stage pipeline allowing 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 • Secure State for supporting FlashVault technology 4.2 AVR32 Architecture AVR32 is a new, high-performance 32-bit RISC microprocessor architecture, designed for costsensitive 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 18 32099IS–01/2012 AT32UC3L016/32/64 single cycle. Load and store instructions have several different formats in order to reduce code size and speed up execution. 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. 4.3 The AVR32UC CPU The AVR32UC CPU targets low- and medium-performance applications, and provides an advanced On-Chip Debug (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 I/O controller ports. This local bus has to be enabled by writing a one to 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 CPU Local Bus section in the Memories chapter. Figure 4-1 on page 20 displays the contents of AVR32UC. 19 32099IS–01/2012 AT32UC3L016/32/64 OCD interface Reset interface Overview of the AVR32UC CPU Interrupt controller interface Figure 4-1. OCD system Power/ Reset control AVR32UC CPU pipeline MPU 4.3.1 High Speed Bus slave CPU Local Bus master CPU Local Bus High Speed Bus master High Speed Bus High Speed Bus High Speed Bus master High Speed Bus Data memory controller Instruction memory controller CPU RAM 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 4-2 on page 21 shows an overview of the AVR32UC pipeline stages. 20 32099IS–01/2012 AT32UC3L016/32/64 Figure 4-2. The AVR32UC Pipeline MUL IF ID Prefetch unit Decode unit Regfile Read ALU LS 4.3.2 Multiply unit Regfile write ALU unit Load-store unit 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. 4.3.2.1 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. 4.3.2.2 Java Support AVR32UC does not provide Java hardware acceleration. 4.3.2.3 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. 4.3.2.4 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 21 32099IS–01/2012 AT32UC3L016/32/64 address exception. Doubleword-sized accesses with word-aligned pointers will automatically be performed as two word-sized accesses. The following table shows the instructions with support for unaligned addresses. All other instructions require aligned addresses. Table 4-1. 4.3.2.5 Instructions with Unaligned Reference Support Instruction Supported Alignment ld.d Word st.d Word 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 4.3.2.6 CPU and Architecture Revision Three major revisions of the AVR32UC CPU currently exist. The device described in this datasheet uses CPU revision 3. 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. 22 32099IS–01/2012 AT32UC3L016/32/64 4.4 4.4.1 Programming Model Register File Configuration The AVR32UC register file is shown below. Figure 4-3. The AVR32UC Register File Application Supervisor INT0 Bit 31 Bit 31 Bit 31 Bit 0 Bit 0 INT1 Bit 0 INT2 Bit 31 Bit 0 INT3 Bit 31 Bit 0 Bit 31 Bit 0 Exception NMI Bit 31 Bit 31 Bit 0 Secure Bit 0 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 PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 PC LR SP_SYS R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 PC LR SP_SEC R12 R11 R10 R9 R8 INT0PC R7 INT1PC R6 FINTPC R5 SMPC R4 R3 R2 R1 R0 SR SR SR SR SR SR SR SR SR SS_STATUS SS_ADRF SS_ADRR SS_ADR0 SS_ADR1 SS_SP_SYS SS_SP_APP SS_RAR SS_RSR 4.4.2 Status Register Configuration The Status Register (SR) is split into two halfwords, one upper and one lower, see Figure 4-4 and Figure 4-5. 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 4-4. The Status Register High Halfword Bit 31 Bit 16 SS LC 1 - - DM D - M2 M1 M0 EM I3M I2M FE I1M I0M GM 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 Bit nam e Initial value G lobal Interrupt M ask Interrupt Level 0 M ask Interrupt Level 1 M ask Interrupt Level 2 M ask Interrupt Level 3 M ask Exception M ask M ode Bit 0 M ode Bit 1 M ode Bit 2 Reserved Debug State Debug State M ask Reserved Secure State 23 32099IS–01/2012 AT32UC3L016/32/64 Figure 4-5. The Status Register Low Halfword Bit 15 Bit 0 - T - - - - - - - - L Q V N Z C Bit name 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Initial value Carry Zero Sign Overflow Saturation Lock Reserved Scratch Reserved 4.4.3 Processor States 4.4.3.1 Normal RISC State The AVR32 processor supports several different execution contexts as shown in Table 4-2. Table 4-2. Overview of Execution Modes, their Priorities and Privilege Levels. Priority Mode Security Description 1 Non Maskable Interrupt Privileged Non Maskable high priority interrupt mode 2 Exception Privileged Execute exceptions 3 Interrupt 3 Privileged General purpose interrupt mode 4 Interrupt 2 Privileged General purpose interrupt mode 5 Interrupt 1 Privileged General purpose interrupt mode 6 Interrupt 0 Privileged General purpose interrupt mode N/A Supervisor Privileged Runs supervisor calls N/A Application Unprivileged 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. 4.4.3.2 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. 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. 24 32099IS–01/2012 AT32UC3L016/32/64 Debug state can be entered as described in the AVR32UC Technical Reference Manual. Debug state is exited by the retd instruction. 4.4.3.3 4.4.4 Secure State The AVR32 can be set in a secure state, that allows a part of the code to execute in a state with higher security levels. The rest of the code can not access resources reserved for this secure code. Secure State is used to implement FlashVault technology. Refer to the AVR32UC Technical Reference Manual for details. 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 4-3. System Registers Reg # Address Name Function 0 0 SR Status Register 1 4 EVBA Exception Vector Base Address 2 8 ACBA Application Call Base Address 3 12 CPUCR CPU Control Register 4 16 ECR Exception Cause Register 5 20 RSR_SUP Unused in AVR32UC 6 24 RSR_INT0 Unused in AVR32UC 7 28 RSR_INT1 Unused in AVR32UC 8 32 RSR_INT2 Unused in AVR32UC 9 36 RSR_INT3 Unused in AVR32UC 10 40 RSR_EX Unused in AVR32UC 11 44 RSR_NMI Unused in AVR32UC 12 48 RSR_DBG Return Status Register for Debug mode 13 52 RAR_SUP Unused in AVR32UC 14 56 RAR_INT0 Unused in AVR32UC 15 60 RAR_INT1 Unused in AVR32UC 16 64 RAR_INT2 Unused in AVR32UC 17 68 RAR_INT3 Unused in AVR32UC 18 72 RAR_EX Unused in AVR32UC 19 76 RAR_NMI Unused in AVR32UC 20 80 RAR_DBG Return Address Register for Debug mode 21 84 JECR Unused in AVR32UC 22 88 JOSP Unused in AVR32UC 23 92 JAVA_LV0 Unused in AVR32UC 25 32099IS–01/2012 AT32UC3L016/32/64 Table 4-3. System Registers (Continued) Reg # Address Name Function 24 96 JAVA_LV1 Unused in AVR32UC 25 100 JAVA_LV2 Unused in AVR32UC 26 104 JAVA_LV3 Unused in AVR32UC 27 108 JAVA_LV4 Unused in AVR32UC 28 112 JAVA_LV5 Unused in AVR32UC 29 116 JAVA_LV6 Unused in AVR32UC 30 120 JAVA_LV7 Unused in AVR32UC 31 124 JTBA Unused in AVR32UC 32 128 JBCR Unused in AVR32UC 33-63 132-252 Reserved Reserved for future use 64 256 CONFIG0 Configuration register 0 65 260 CONFIG1 Configuration register 1 66 264 COUNT Cycle Counter register 67 268 COMPARE Compare register 68 272 TLBEHI Unused in AVR32UC 69 276 TLBELO Unused in AVR32UC 70 280 PTBR Unused in AVR32UC 71 284 TLBEAR Unused in AVR32UC 72 288 MMUCR Unused in AVR32UC 73 292 TLBARLO Unused in AVR32UC 74 296 TLBARHI Unused in AVR32UC 75 300 PCCNT Unused in AVR32UC 76 304 PCNT0 Unused in AVR32UC 77 308 PCNT1 Unused in AVR32UC 78 312 PCCR Unused in AVR32UC 79 316 BEAR Bus Error Address Register 80 320 MPUAR0 MPU Address Register region 0 81 324 MPUAR1 MPU Address Register region 1 82 328 MPUAR2 MPU Address Register region 2 83 332 MPUAR3 MPU Address Register region 3 84 336 MPUAR4 MPU Address Register region 4 85 340 MPUAR5 MPU Address Register region 5 86 344 MPUAR6 MPU Address Register region 6 87 348 MPUAR7 MPU Address Register region 7 88 352 MPUPSR0 MPU Privilege Select Register region 0 89 356 MPUPSR1 MPU Privilege Select Register region 1 26 32099IS–01/2012 AT32UC3L016/32/64 Table 4-3. 4.5 System Registers (Continued) Reg # Address Name Function 90 360 MPUPSR2 MPU Privilege Select Register region 2 91 364 MPUPSR3 MPU Privilege Select Register region 3 92 368 MPUPSR4 MPU Privilege Select Register region 4 93 372 MPUPSR5 MPU Privilege Select Register region 5 94 376 MPUPSR6 MPU Privilege Select Register region 6 95 380 MPUPSR7 MPU Privilege Select Register region 7 96 384 MPUCRA Unused in this version of AVR32UC 97 388 MPUCRB Unused in this version of AVR32UC 98 392 MPUBRA Unused in this version of AVR32UC 99 396 MPUBRB Unused in this version of AVR32UC 100 400 MPUAPRA MPU Access Permission Register A 101 404 MPUAPRB MPU Access Permission Register B 102 408 MPUCR MPU Control Register 103 412 SS_STATUS Secure State Status Register 104 416 SS_ADRF Secure State Address Flash Register 105 420 SS_ADRR Secure State Address RAM Register 106 424 SS_ADR0 Secure State Address 0 Register 107 428 SS_ADR1 Secure State Address 1 Register 108 432 SS_SP_SYS Secure State Stack Pointer System Register 109 436 SS_SP_APP Secure State Stack Pointer Application Register 110 440 SS_RAR Secure State Return Address Register 111 444 SS_RSR Secure State Return Status Register 112-191 448-764 Reserved Reserved for future use 192-255 768-1020 IMPL IMPLEMENTATION DEFINED Exceptions and Interrupts In the AVR32 architecture, events are used as a common term for exceptions and interrupts. AVR32UC incorporates a powerful event handling scheme. The different event sources, like Illegal Op-code and interrupt requests, have different priority levels, ensuring a well-defined behavior when multiple events are received simultaneously. Additionally, pending events of a higher priority class may preempt handling of ongoing events of a lower priority class. When an event occurs, the execution of the instruction stream is halted, and execution is passed to an event handler at an address specified in Table 4-4 on page 31. 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 interrupt sources have autovectored interrupt service routine (ISR) addresses. This allows the interrupt controller to directly specify the ISR address as an address 27 32099IS–01/2012 AT32UC3L016/32/64 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 interrupt requests, yielding a uniform event handling scheme. An interrupt controller does the priority handling of the interrupts and provides the autovector offset to the CPU. 4.5.1 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. 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. 4.5.2 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 4-4 on page 31, 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. 28 32099IS–01/2012 AT32UC3L016/32/64 4.5.3 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. 4.5.4 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 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. 4.5.5 Entry Points for Events Several different event handler entry points exist. In AVR32UC, the reset address is 0x80000000. 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 interrupt requests have entry points located at an offset relative to EVBA. This autovector offset is specified by an 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 4-4 on page 31. 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 29 32099IS–01/2012 AT32UC3L016/32/64 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 4-4 on page 31. Some of the exceptions are unused in AVR32UC since it has no MMU, coprocessor interface, or floatingpoint unit. 30 32099IS–01/2012 AT32UC3L016/32/64 Table 4-4. Priority and Handler Addresses for Events Priority Handler Address Name Event source Stored Return Address 1 0x80000000 Reset External input Undefined 2 Provided by OCD system OCD Stop CPU OCD system First non-completed instruction 3 EVBA+0x00 Unrecoverable exception Internal PC of offending instruction 4 EVBA+0x04 TLB multiple hit MPU PC of offending instruction 5 EVBA+0x08 Bus error data fetch Data bus First non-completed instruction 6 EVBA+0x0C Bus error instruction fetch Data bus First non-completed instruction 7 EVBA+0x10 NMI External input First non-completed instruction 8 Autovectored Interrupt 3 request External input First non-completed instruction 9 Autovectored Interrupt 2 request External input First non-completed instruction 10 Autovectored Interrupt 1 request External input First non-completed instruction 11 Autovectored Interrupt 0 request External input First non-completed instruction 12 EVBA+0x14 Instruction Address CPU PC of offending instruction 13 EVBA+0x50 ITLB Miss MPU PC of offending instruction 14 EVBA+0x18 ITLB Protection MPU PC of offending instruction 15 EVBA+0x1C Breakpoint OCD system First non-completed instruction 16 EVBA+0x20 Illegal Opcode Instruction PC of offending instruction 17 EVBA+0x24 Unimplemented instruction Instruction PC of offending instruction 18 EVBA+0x28 Privilege violation Instruction PC of offending instruction 19 EVBA+0x2C Floating-point UNUSED 20 EVBA+0x30 Coprocessor absent Instruction PC of offending instruction 21 EVBA+0x100 Supervisor call Instruction PC(Supervisor Call) +2 22 EVBA+0x34 Data Address (Read) CPU PC of offending instruction 23 EVBA+0x38 Data Address (Write) CPU PC of offending instruction 24 EVBA+0x60 DTLB Miss (Read) MPU PC of offending instruction 25 EVBA+0x70 DTLB Miss (Write) MPU PC of offending instruction 26 EVBA+0x3C DTLB Protection (Read) MPU PC of offending instruction 27 EVBA+0x40 DTLB Protection (Write) MPU PC of offending instruction 28 EVBA+0x44 DTLB Modified UNUSED 31 32099IS–01/2012 AT32UC3L016/32/64 5. Memories 5.1 Embedded Memories • Internal high-speed flash – 64Kbytes (AT32UC3L064) – 32Kbytes (AT32UC3L032) – 16Kbytes (AT32UC3L016) • 0 wait state access at up to 25mhz in worst case conditions • 1 wait state access at up to 50mhz in worst case conditions • Pipelined flash architecture, allowing burst reads from sequential Flash locations, hiding penalty of 1 wait state access • Pipelined flash architecture typically reduces the cycle penalty of 1 wait state operation to only 8% compared to 0 wait state operation • 100 000 write cycles, 15-year data retention capability • 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 – 16Kbytes (AT32UC3L064, AT32UC3L032) – 8Kbytes (AT32UC3L016) 5.2 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 AVR32 Architecture Manual. The 32-bit physical address space is mapped as follows: Table 5-1. AT32UC3L016/32/64 Physical Memory Map Device Table 5-2. Start Address Size AT32UC3L064 AT32UC3L032 AT32UC3L016 Embedded SRAM 0x00000000 16Kbytes 16Kbytes 8Kbytes Embedded Flash 0x80000000 64Kbytes 32Kbytes 16Kbytes SAU Channels 0x90000000 256 bytes 256 bytes 256 bytes HSB-PB Bridge B 0xFFFE0000 64Kbytes 64Kbytes 64Kbytes HSB-PB Bridge A 0xFFFF0000 64Kbytes 64Kbytes 64Kbytes Flash Memory Parameters Part Number Flash Size (FLASH_PW) Number of pages (FLASH_P) Page size (FLASH_W) AT32UC3L064 64Kbytes 256 256 bytes AT32UC3L032 32Kbytes 128 256 bytes AT32UC3L016 16Kbytes 64 256 bytes 32 32099IS–01/2012 AT32UC3L016/32/64 5.3 Peripheral Address Map Table 5-3. Peripheral Address Mapping Address Peripheral Name 0xFFFE0000 FLASHCDW Flash Controller - FLASHCDW 0xFFFE0400 HMATRIX HSB Matrix - HMATRIX 0xFFFE0800 SAU Secure Access Unit - SAU 0xFFFF0000 PDCA Peripheral DMA Controller - PDCA INTC Interrupt controller - INTC 0xFFFF1000 0xFFFF1400 PM Power Manager - PM 0xFFFF1800 SCIF System Control Interface - SCIF AST Asynchronous Timer - AST WDT Watchdog Timer - WDT EIC External Interrupt Controller - EIC 0xFFFF1C00 0xFFFF2000 0xFFFF2400 0xFFFF2800 FREQM Frequency Meter - FREQM 0xFFFF2C00 GPIO 0xFFFF3000 General Purpose Input/Output Controller - GPIO USART0 Universal Synchronous/Asynchronous Receiver/Transmitter - USART0 USART1 Universal Synchronous/Asynchronous Receiver/Transmitter - USART1 USART2 Universal Synchronous/Asynchronous Receiver/Transmitter - USART2 USART3 Universal Synchronous/Asynchronous Receiver/Transmitter - USART3 0xFFFF3400 0xFFFF3800 0xFFFF3C00 0xFFFF4000 SPI Serial Peripheral Interface - SPI 0xFFFF4400 TWIM0 Two-wire Master Interface - TWIM0 33 32099IS–01/2012 AT32UC3L016/32/64 Table 5-3. Peripheral Address Mapping 0xFFFF4800 TWIM1 Two-wire Master Interface - TWIM1 TWIS0 Two-wire Slave Interface - TWIS0 TWIS1 Two-wire Slave Interface - TWIS1 PWMA Pulse Width Modulation Controller - PWMA 0xFFFF4C00 0xFFFF5000 0xFFFF5400 0xFFFF5800 TC0 Timer/Counter - TC0 TC1 Timer/Counter - TC1 0xFFFF5C00 0xFFFF6000 ADCIFB ADC Interface - ADCIFB 0xFFFF6400 ACIFB Analog Comparator Interface - ACIFB 0xFFFF6800 CAT Capacitive Touch Module - CAT 0xFFFF6C00 GLOC Glue Logic Controller - GLOC 0xFFFF7000 AW 5.4 aWire - AW 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. 34 32099IS–01/2012 AT32UC3L016/32/64 The following GPIO registers are mapped on the local bus: Table 5-4. Local Bus Mapped GPIO Registers Port Register Mode Local Bus Address Access 0 Output Driver Enable Register (ODER) WRITE 0x40000040 Write-only SET 0x40000044 Write-only CLEAR 0x40000048 Write-only TOGGLE 0x4000004C Write-only WRITE 0x40000050 Write-only SET 0x40000054 Write-only CLEAR 0x40000058 Write-only TOGGLE 0x4000005C Write-only Pin Value Register (PVR) - 0x40000060 Read-only Output Driver Enable Register (ODER) WRITE 0x40000140 Write-only SET 0x40000144 Write-only CLEAR 0x40000148 Write-only TOGGLE 0x4000014C Write-only WRITE 0x40000150 Write-only SET 0x40000154 Write-only CLEAR 0x40000158 Write-only TOGGLE 0x4000015C Write-only - 0x40000160 Read-only Output Value Register (OVR) 1 Output Value Register (OVR) Pin Value Register (PVR) 35 32099IS–01/2012 AT32UC3L016/32/64 6. Supply and Startup Considerations 6.1 6.1.1 Supply Considerations Power Supplies The AT32UC3L016/32/64 has several types of power supply pins: •VDDIO: Powers I/O lines. Voltage is 1.8 to 3.3V nominal. •VDDIN: Powers I/O lines and the internal regulator. Voltage is 1.8 to 3.3V nominal. •VDDANA: Powers the ADC. Voltage is 1.8V nominal. •VDDCORE: Powers the core, memories, and peripherals. Voltage is 1.8V nominal. The ground pins GND are common to VDDCORE, VDDIO, and VDDIN. The ground pin for VDDANA is GNDANA. When VDDCORE is not connected to VDDIN, the VDDIN voltage must be higher than 1.98V. Refer to Section 7. on page 41 for power consumption on the various supply pins. For decoupling recommendations for the different power supplies, please refer to the schematic checklist. 6.1.2 Voltage Regulator The AT32UC3L016/32/64 embeds a voltage regulator that converts from 3.3V nominal to 1.8V with a load of up to 60mA. The regulator supplies the output voltage on VDDCORE. The regulator may only be used to drive internal circuitry in the device. VDDCORE should be externally connected to the 1.8V domains. See Section 6.1.3 for regulator connection figures. Adequate output supply decoupling is mandatory for VDDCORE to reduce ripple and avoid oscillations. The best way to achieve this is to use two capacitors in parallell between VDDCORE and GND as close to the device as possible. Please refer to Section 7.8.1 on page 55 for decoupling capacitors values and regulator characteristics. Figure 6-1. Supply Decoupling 3.3V VDDIN C IN3 CIN2 CIN1 1.8V VDDCORE COUT2 6.1.3 1.8V Regulator COUT1 Regulator Connection The AT32UC3L016/32/64 supports three power supply configurations: • 3.3V single supply mode • 1.8V single supply mode • 3.3V supply mode, with 1.8V regulated I/O lines 36 32099IS–01/2012 AT32UC3L016/32/64 6.1.3.1 3.3V Single Supply Mode In 3.3V single supply mode the internal regulator is connected to the 3.3V source (VDDIN pin) and its output feeds VDDCORE. Figure 6-2 shows the power schematics to be used for 3.3V single supply mode. All I/O lines will be powered by the same power (VDDIN=VDDIO). Figure 6-2. 3.3V Single Supply Mode + 1.98-3.6V - VDDIN VDDIO I/O Pins I/O Pins VDDCORE OSC32K RC32K AST Wake POR33 SM33 Linear regulator VDDANA GND ADC CPU, Peripherals, Memories, SCIF, BOD, RCSYS, DFLL GNDANA 37 32099IS–01/2012 AT32UC3L016/32/64 6.1.3.2 1.8V Single Supply Mode In 1.8V single supply mode the internal regulator is not used, and VDDIO and VDDCORE are powered by a single 1.8V supply as shown in Figure 6-3. All I/O lines will be powered by the same power (VDDIN = VDDIO = VDDCORE). Figure 6-3. 1.8V Single Supply Mode. + 1.62-1.98V - VDDIN VDDIO I/O Pins I/O Pins VDDCORE Linear Regulator VDDANA OSC32K RC32K AST Wake POR33 SM33 ADC GNDANA GND CPU, Peripherals, Memories, SCIF, BOD, RCSYS, DFLL 38 32099IS–01/2012 AT32UC3L016/32/64 6.1.3.3 3.3V Supply Mode with 1.8V Regulated I/O Lines In this mode, the internal regulator is connected to the 3.3V source and its output is connected to both VDDCORE and VDDIO as shown in Figure 6-4. This configuration is required in order to use Shutdown mode. Figure 6-4. 3.3V Supply Mode with 1.8V Regulated I/O Lines 1.98-3.6V + - VDDIN VDDIO I/O Pins VDDCORE I/O Pins Linear Regulator VDDANA OSC32K RC32K AST Wake POR33 SM33 ADC GNDANA GND CPU, Peripherals, Memories, SCIF, BOD, RCSYS, DFLL In this mode, some I/O lines are powered by VDDIN while other I/O lines are powered by VDDIO. Refer to Section 3.2 on page 9 for description of power supply for each I/O line. Refer to the Power Manager chapter for a description of what parts of the system are powered in Shutdown mode. Important note: As the regulator has a maximum output current of 60mA, this mode can only be used in applications where the maximum I/O current is known and compatible with the core and peripheral power consumption. Typically, great care must be used to ensure that only a few I/O lines are toggling at the same time and drive very small loads. 39 32099IS–01/2012 AT32UC3L016/32/64 6.1.4 Power-up Sequence 6.1.4.1 Maximum Rise Rate To avoid risk of latch-up, the rise rate of the power supplies must not exceed the values described in Table 7-3 on page 42. Recommended order for power supplies is also described in this chapter. 6.1.4.2 Minimum Rise Rate The integrated Power-on Reset (POR33) circuitry monitoring the VDDIN powering supply requires a minimum rise rate for the VDDIN power supply. See Table 7-3 on page 42 for the minimum rise rate value. If the application can not ensure that the minimum rise rate condition for the VDDIN power supply is met, one of the following configurations can be used: • A logic “0” value is applied during power-up on pin PA11 until VDDIN rises above 1.2V. • A logic “0” value is applied during power-up on pin RESET_N until VDDIN rises above 1.2V. 6.2 Startup Considerations This chapter summarizes the boot sequence of the AT32UC3L016/32/64. The behavior after power-up is controlled by the Power Manager. For specific details, refer to the Power Manager chapter. 6.2.1 Starting of Clocks After power-up, the device will be held in a reset state by the Power-on Reset (POR18 and POR33) circuitry for a short time to allow the power to stabilize throughout the device. After reset, the device will use the System RC Oscillator (RCSYS) as clock source. Please refer to Table 7-17 on page 54 for the frequency for this oscillator. On system start-up, the DFLL is disabled. All clocks to all modules are running. No clocks have a divided frequency; all parts of the system receive a clock with the same frequency as the System RC Oscillator. When powering up the device, there may be a delay before the voltage has stabilized, depending on the rise time of the supply used. The CPU can start executing code as soon as the supply is above the POR18 and POR33 thresholds, and before the supply is stable. Before switching to a high-speed clock source, the user should use the BOD to make sure the VDDCORE is above the minimum level (1.62V). 6.2.2 Fetching of Initial Instructions After reset has been released, the AVR32 UC CPU starts fetching instructions from the reset address, which is 0x80000000. This address points to the first address in the internal flash. The code read from the internal flash is free to configure the clock system and clock sources . Please refer to the Power Manager and SCIF chapters for details. 40 32099IS–01/2012 AT32UC3L016/32/64 7. Electrical Characteristics 7.1 Absolute Maximum Ratings* Table 7-1. Absolute Maximum Ratings Operating temperature..................................... -40°C to +85°C *NOTICE: Storage temperature...................................... -60°C to +150°C Voltage on input pins (except for 5V pins) with respect to ground .................................................................-0.3V to VVDD(2)+0.3V Voltage on 5V tolerant(1) pins with respect to ground ............... .............................................................................-0.3V to 5.5V Total DC output current on all I/O pins - VDDIO ........... 120mA Total DC output current on all I/O pins - VDDIN ............. 36mA 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. Maximum operating voltage VDDCORE......................... 1.98V Maximum operating voltage VDDIO, VDDIN .................... 3.6V Notes: 1. 5V tolerant pins, see Section 3.2 ”Peripheral Multiplexing on I/O lines” on page 9 2. VVDD corresponds to either VVDDIN or VVDDIO, depending on the supply for the pin. Refer to Section 3.2 on page 9 for details. 7.2 Supply Characteristics The following characteristics are applicable to the operating temperature range: TA = -40°C to 85°C, unless otherwise specified and are valid for a junction temperature up to T J = 100°C. Please refer to Section 6. ”Supply and Startup Considerations” on page 36. Table 7-2. Supply Characteristics Voltage Symbol Parameter Min Max Unit VVDDIO DC supply peripheral I/Os 1.62 3.6 V DC supply peripheral I/Os, 1.8V single supply mode 1.62 1.98 V DC supply peripheral I/Os and internal regulator, 3.3V supply mode 1.98 3.6 V VVDDCORE DC supply core 1.62 1.98 V VVDDANA Analog supply voltage 1.62 1.98 V VVDDIN 41 32099IS–01/2012 AT32UC3L016/32/64 Table 7-3. Supply Rise Rates and Order(1) Rise Rate Symbol Parameter Min Max Unit VVDDIO DC supply peripheral I/Os 0 2.5 V/µs VVDDIN DC supply peripheral I/Os and internal regulator 0.002 2.5 V/µs Slower rise time requires external power-on reset circuit. VVDDCORE DC supply core 0 2.5 V/µs Rise before or at the same time as VDDIO VVDDANA Analog supply voltage 0 2.5 V/µs Rise together with VDDCORE Note: 7.3 Comment 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. Maximum Clock Frequencies These parameters are given in the following conditions: • VVDDCORE = 1.62V to 1.98V • Temperature = -40°C to 85°C Table 7-4. 7.4 Clock Frequencies Symbol Parameter fCPU Conditions Min Max Units CPU clock frequency 50 MHz fPBA PBA clock frequency 50 MHz fPBB PBB clock frequency 50 MHz fGCLK0 GCLK0 clock frequency DFLLIF main reference, GCLK0 pin 150 MHz fGCLK1 GCLK1 clock frequency DFLLIF dithering and ssg reference, GCLK1 pin 150 MHz fGCLK2 GCLK2 clock frequency AST, GCLK2 pin 80 MHz fGCLK3 GCLK3 clock frequency PWMA, GCLK3 pin 110 MHz fGCLK4 GCLK4 clock frequency CAT, ACIFB, GCLK4 pin 110 MHz fGCLK5 GCLK5 clock frequency GLOC 80 MHz Power Consumption The values in Table 7-5 are measured values of power consumption under the following conditions, except where noted: • Operating conditions internal core supply (Figure 7-1) - this is the default configuration – VVDDIN = 3.0V – VVDDCORE = 1.62V, supplied by the internal regulator 42 32099IS–01/2012 AT32UC3L016/32/64 – Corresponds to the 3.3V supply mode with 1.8V regulated I/O lines, please refer to the Supply and Startup Considerations section for more details • Equivalent to the 3.3V single supply mode • Consumption in 1.8V single supply mode can be estimated by subtracting the regulator static current • Operating conditions external core supply (Figure 7-2) - used only when noted – VVDDIN = VVDDCORE = 1.8V – Corresponds to the 1.8V single supply mode, please refer to the Supply and Startup Considerations section for more details • TA = 25°C • Oscillators – OSC0 (crystal oscillator) stopped – OSC32K (32KHz crystal oscillator) running with external 32KHz crystal – DFLL running at 50MHz with OSC32K as reference • Clocks – DFLL used as main clock source – CPU, HSB, and PBB clocks undivided – PBA clock divided by 4 – The following peripheral clocks running • PM, SCIF, AST, FLASHCDW, PBA bridge – All other peripheral clocks stopped • I/Os are inactive with internal pull-up • Flash enabled in high speed mode • POR33 disabled 43 32099IS–01/2012 AT32UC3L016/32/64 Table 7-5. Mode Power Consumption for Different Operating Modes Conditions Active(1) Measured on Consumption Typ -CPU running a recursive Fibonacci algorithm 260 -CPU running a division algorithm 165 Idle(1) 92 (1) (1) 47 Stop 37 DeepStop 23 -OSC32K and AST stopped -Internal core supply Static Shutdown Amp0 10 µA -OSC32K running -AST running at 1KHz -External core supply (Figure 7-2) 5.3 -OSC32K and AST stopped -External core supply (Figure 7-2) 4.7 -OSC32K running -AST running at 1KHz 600 nA -AST and OSC32K stopped Note: µA/MHz 58 Frozen Standby Unit 9 1. These numbers are valid for the measured condition only and must not be extrapolated to other frequencies. Figure 7-1. Measurement Schematic, Internal Core Supply Amp0 VDDIN VDDIO VDDCORE VDDANA 44 32099IS–01/2012 AT32UC3L016/32/64 Figure 7-2. Measurement Schematic, External Core Supply Amp0 VDDIN VDDIO VDDCORE VDDANA 7.4.1 Peripheral Power Consumption The values in Table 7-6 are measured values of power consumption under the following conditions. • Operating conditions internal core supply (Figure 7-1) – VVDDIN = 3.0V – VVDDCORE = 1.62V, supplied by the internal regulator – Corresponds to the 3.3V supply mode with 1.8V regulated I/O lines, please refer to the Supply and Startup Considerations section for more details • TA = 25°C • Oscillators – OSC0 (crystal oscillator) stopped – OSC32K (32KHz crystal oscillator) running with external 32KHz crystal – DFLL running at 50MHz with OSC32K as reference • Clocks – DFLL used as main clock source – CPU, HSB, and PB clocks undivided • I/Os are inactive with internal pull-up • Flash enabled in high speed mode • POR33 disabled Consumption active is the added current consumption when the module clock is turned on and the module is doing a typical set of operations. 45 32099IS–01/2012 AT32UC3L016/32/64 Table 7-6. Peripheral Typical Current Consumption by Peripheral(2) Typ Consumption Active ACIFB 14.0 ADCIFB(1) 14.9 AST 5.6 AW USART 6.8 CAT 12.4 EIC 1.3 FREQM 3.2 GLOC 0.4 GPIO 15.9 PWMA 2.5 SPI 7.6 TC 7.2 TWIM 5.1 TWIS 3.2 USART 12.3 WDT 2.3 Unit µA/MHz Notes: 1. Includes the current consumption on VDDANA and ADVREFP. 2. These numbers are valid for the measured condition only and must not be extrapolated to other frequencies. 46 32099IS–01/2012 AT32UC3L016/32/64 7.5 I/O Pin Characteristics Table 7-7. Normal I/O Pin Characteristics(1) Symbol Parameter RPULLUP Pull-up resistance VIL Input low-level voltage VIH Input high-level voltage VOL Output low-level voltage VOH Output high-level voltage fMAX Output frequency(2) tRISE Rise time(2) Condition Min Typ Max Units 75 100 145 kOhm VVDD = 3.0V -0.3 0.3*VVDD VVDD = 1.62V -0.3 0.3*VVDD VVDD = 3.6V 0.7*VVDD VVDD + 0.3 VVDD = 1.98V 0.7*VVDD VVDD + 0.3 VVDD = 3.0V, IOL = 3mA 0.4 VVDD = 1.62V, IOL = 2mA 0.4 VVDD = 3.0V, IOH = 3mA VVDD - 0.4 VVDD = 1.62V, IOH = 2mA VVDD - 0.4 V VVDD = 3.0V, load = 10pF 45 VVDD = 3.0V, load = 30pF 23 VVDD = 3.0V, load = 10pF 4.7 VVDD = 3.0V, load = 30pF 11.5 VVDD = 3.0V, load = 10pF 4.8 VVDD = 3.0V, load = 30pF 12 1 MHz ns Fall time(2) ILEAK Input leakage current Pull-up resistors disabled TQFP48 package 1.4 CIN Input capacitance, all normal I/O pins except PA05, PA07, PA17, PA20, PA21, PB04, PB05 QFN48 package 1.1 TLLGA 48 package 1.1 TQFP48 package 2.7 QFN48 package 2.4 TLLGA 48 package 2.4 TQFP48 package 3.8 QFN48 package 3.5 TLLGA 48 package 3.5 Input capacitance, PA20 Input capacitance, PA05, PA07, PA17, PA21, PB04, PB05 CIN Notes: V V tFALL CIN V µA pF 1. VVDD corresponds to either VVDDIN or VVDDIO, depending on the supply for the pin. Refer to Section 3.2 on page 9 for details. 2. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. Table 7-8. Symbol RPULLUP High-drive I/O Pin Characteristics(1) Parameter Pull-up resistance Condition Min Typ Max PA06 30 50 110 PA02, PB01, RESET 75 100 145 PA08, PA09 10 20 45 Units kOhm 47 32099IS–01/2012 AT32UC3L016/32/64 Table 7-8. High-drive I/O Pin Characteristics(1) Symbol Parameter Condition Min Typ Max VIL Input low-level voltage VVDD = 3.0V -0.3 0.3*VVDD VVDD = 1.62V -0.3 0.3*VVDD VIH Input high-level voltage VVDD = 3.6V 0.7*VVDD VVDD + 0.3 VVDD = 1.98V 0.7*VVDD VVDD + 0.3 VOL Output low-level voltage VOH Output high-level voltage Output frequency, all Highdrive I/O pins, except PA08 and PA09(2) VVDD = 3.0V, load = 10pF 45 fMAX VVDD = 3.0V, load = 30pF 23 Rise time, all High-drive I/O pins, except PA08 and PA09(2) VVDD = 3.0V, load = 10pF 4.7 tRISE VVDD = 3.0V, load = 30pF 11.5 VVDD = 3.0V, load = 10pF 4.8 tFALL Fall time, all High-drive I/O pins, except PA08 and PA09(2) VVDD = 3.0V, load = 30pF 12 Output frequency, PA08 and PA09(2) VVDD = 3.0V, load = 10pF 52 fMAX VVDD = 3.0V, load = 30pF 39 Rise time, PA08 and PA09(2) VVDD = 3.0V, load = 10pF 2.9 tRISE VVDD = 3.0V, load = 30pF 4.9 VVDD = 3.0V, load = 10pF 2.5 tFALL Fall time, PA08 and PA09(2) VVDD = 3.0V, load = 30pF 4.6 ILEAK Input leakage current Pull-up resistors disabled 1 CIN Input capacitance, all High-drive I/O pins, except PA08 and PA09 VVDD = 3.0V, IOL = 6mA 0.4 VVDD = 1.62V, IOL = 4mA 0.4 Units V V V VVDD = 3.0V, IOH = 6mA VVDD-0.4 VVDD = 1.62V, IOH = 4mA VVDD-0.4 V MHz ns MHz ns TQFP48 package 2,2 QFN48 package 2.0 TLLGA 48 package 2.0 TQFP48 package 7.0 QFN48 package 6.7 TLLGA 48 package 6.7 µA pF Input capacitance, PA08 and PA09 CIN Notes: 1. VVDD corresponds to either VVDDIN or VVDDIO, depending on the supply for the pin. Refer to Section 3.2 on page 9 for details. 2. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. Table 7-9. High-drive I/O, 5V Tolerant, Pin Characteristics(1) Symbol Parameter RPULLUP Pull-up resistance VIL Input low-level voltage Condition Min Typ Max Units 30 50 110 kOhm VVDD = 3.0V -0.3 0.3*VVDD VVDD = 1.62V -0.3 0.3*VVDD V 48 32099IS–01/2012 AT32UC3L016/32/64 Table 7-9. High-drive I/O, 5V Tolerant, Pin Characteristics(1) Symbol Parameter VIH Input high-level voltage VOL Output low-level voltage VOH Output high-level voltage fMAX Output frequency(2) tRISE Rise time(2) (2) tFALL Fall time ILEAK Input leakage current CIN Notes: Input capacitance Condition Min Typ Max VVDD = 3.6V 0.7*VVDD 5.5 VVDD = 1.98V 0.7*VVDD 5.5 Units V VVDD = 3.0V, IOL = 6mA 0.4 VVDD = 1.62V, IOL = 4mA 0.4 V VVDD = 3.0V, IOH = 6mA VVDD-0.4 VVDD = 1.62V, IOH = 4mA VVDD-0.4 V VVDD = 3.0V, load = 10pF 87 VVDD = 3.0V, load = 30pF 58 VVDD = 3.0V, load = 10pF 2.3 VVDD = 3.0V, load = 30pF 4.3 VVDD = 3.0V, load = 10pF 1.9 VVDD = 3.0V, load = 30pF 3.7 5.5V, pull-up resistors disabled 10 MHz ns TQFP48 package 4.5 QFN48 package 4.2 TLLGA48 package 4.2 µA pF 1. VVDD corresponds to either VVDDIN or VVDDIO, depending on the supply for the pin. Refer to Section 3.2 on page 9 for details. 2. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. Table 7-10. TWI Pin Characteristics(1) Symbol Parameter RPULLUP Pull-up resistance VIL Input low-level voltage Input high-level voltage VIH Condition Min Typ Max Units 25 35 60 kOhm VVDD = 3.0V -0.3 0.3*VVDD VVDD = 1.62V -0.3 0.3*VVDD VVDD = 3.6V 0.7*VVDD VVDD + 0.3 VVDD = 1.98V 0.7*VVDD VVDD + 0.3 Input high-level voltage, 5V tolerant SMBUS compliant pins VVDD = 3.6V 0.7*VVDD 5.5 VVDD = 1.98V 0.7*VVDD 5.5 VOL Output low-level voltage IOL = 3mA ILEAK Input leakage current Pull-up resistors disabled IIL Input low leakage 1 IIH Input high leakage 1 CIN Input capacitance V V V 0.4 V 1 TQFP48 package 3.8 QFN48 package 3.5 TLLGA48 package 3.5 µA pF 49 32099IS–01/2012 AT32UC3L016/32/64 Table 7-10. TWI Pin Characteristics(1) Symbol Parameter tFALL Fall time fMAX Max frequency Condition Min Typ Cbus = 400pF, VVDD > 2.0V 250 Cbus = 400pF, VVDD > 1.62V 470 Cbus = 400pF, VVDD > 2.0V Max Units ns 400 kHz Note: 1. VVDD corresponds to either VVDDIN or VVDDIO, depending on the supply for the pin. Refer to Section 3.2 on page 9 for details. 7.6 Oscillator Characteristics 7.6.1 Oscillator 0 (OSC0) Characteristics 7.6.1.1 Digital Clock Characteristics The following table describes the characteristics for the oscillator when a digital clock is applied on XIN. Table 7-11. Digital Clock Characteristics Symbol Parameter fCPXIN XIN clock frequency tCPXIN XIN clock duty cycle tSTARTUP Startup time CIN 7.6.1.2 XIN input capacitance Conditions Min Typ Max 40 Units 50 MHz 60 % 0 TQFP48 package 7.0 QFN48 package 6.7 TLLGA48 package 6.7 cycles pF Crystal Oscillator Characteristics The following table describes the characteristics for the oscillator when a crystal is connected between XIN and XOUT as shown in Figure 7-3. The user must choose a crystal oscillator where the crystal load capacitance CL is within the range given in the table. The exact value of CL can be found in the crystal datasheet. The capacitance of the external capacitors (CLEXT) can then be computed as follows: C LEXT = 2 ( C L – C i ) – C PCB where CPCB is the capacitance of the PCB and Ci is the internal equivalent load capacitance. Table 7-12. Crystal Oscillator Characteristics Symbol Parameter fOUT Crystal oscillator frequency CL Crystal load capacitance Ci Internal equivalent load capacitance Conditions Min Typ Max Unit 0.45 10 16 MHz 6 18 pF 2 50 32099IS–01/2012 AT32UC3L016/32/64 Table 7-12. Symbol Crystal Oscillator Characteristics Parameter tSTARTUP Conditions Min (1) Startup time SCIF.OSCCTRL.GAIN = 2 Current consumption Notes: Max (2) 30 000 Active mode, f = 0.45MHz, SCIF.OSCCTRL.GAIN = 0 IOSC Typ Unit cycles 30 µA Active mode, f = 10MHz, SCIF.OSCCTRL.GAIN = 2 170 1. Please refer to the SCIF chapter for details. 2. Nominal crystal cycles. 3. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. Figure 7-3. Oscillator Connection CLEXT XOUT UC3L Ci CL XIN CLEXT 7.6.2 32KHz Crystal Oscillator (OSC32K) Characteristics Figure 7-3 and the equation above also applies to the 32 KHz oscillator connection. The user must choose a crystal oscillator where the crystal load capacitance CL is within the range given in the table. The exact value of CL can then be found in the crystal datasheet. Table 7-13. 32 KHz Crystal Oscillator Characteristics Symbol Parameter fOUT Crystal oscillator frequency tSTARTUP Startup time CL Crystal load capacitance(2) Ci Internal equivalent load capacitance IOSC32 Current consumption RS Equivalent series resistance Conditions Min Typ Max 32 768 Hz (1) RS = 60kOhm, CL = 9pF 30 000 6 Unit cycles 12.5 pF 2 0.9 (2) 32 768Hz 35 µA 85 kOhm 51 32099IS–01/2012 AT32UC3L016/32/64 Notes: 1. Nominal crystal cycles. 2. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 7.6.3 Digital Frequency Locked Loop (DFLL) Characteristics Table 7-14. Symbol Digital Frequency Locked Loop Characteristics Parameter Conditions Output frequency fOUT (2) Reference frequency fREF (2) FINE resolution FINE > 100, all COARSE values Frequency drift over voltage and temperature Accuracy(2) tSTARTUP Startup time tLOCK Lock time Notes: (2) Typ Max Unit 40 150 MHz 8 150 kHz 0.25 % See Figure 7-4 Fine lock, fREF = 32kHz, SSG disabled 0.1 0.5 Accurate lock, fREF = 32kHz, dither clk RCSYS/2, SSG disabled 0.06 0.5 Fine lock, fREF = 8-150kHz, SSG disabled 0.2 1 Accurate lock, fREF = 8-150kHz, dither clk RCSYS/2, SSG disabled 0.1 1 Power consumption IDFLL Min % 22 Within 90% of final values µA/MHz 100 fREF = 32kHz, fine lock, SSG disabled 600 fREF = 32kHz, accurate lock, dithering clock = RCSYS/2, SSG disabled 1100 µs 1. Spread Spectrum Generator (SSG) is disabled by writing a zero to the EN bit in the SCIF.DFLL0SSG register. 2. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 52 32099IS–01/2012 AT32UC3L016/32/64 Figure 7-4. DFLL Open Loop Frequency Variation(1) DFLL Open Loop Frequency variation 160 150 Frequencies (MHz) 140 130 1,98V 120 1,8V 1.62V 110 100 90 80 -40 -20 0 20 40 60 80 Tem pera ture Note: 1. The plot shows a typical behaviour for coarse = 99 and fine = 255 in open loop mode. 7.6.4 120MHz RC Oscillator (RC120M) Characteristics Table 7-15. Symbol Internal 120MHz RC Oscillator Characteristics Parameter Conditions (1) fOUT Output frequency IRC120M Current consumption tSTARTUP Startup time Note: VVDDCORE = 1.8V Min Typ Max Unit 88 120 152 MHz 1.85 mA 3 µs 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 53 32099IS–01/2012 AT32UC3L016/32/64 7.6.5 32kHz RC Oscillator (RC32K) Characteristics Table 7-16. Symbol 32kHz RC Oscillator Characteristics Parameter Conditions (1) Min Typ Max Unit 20 32 44 kHz fOUT Output frequency IRC32K Current consumption 0.6 µA tSTARTUP Startup time 100 µs Note: 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 7.6.6 System RC Oscillator (RCSYS) Characteristics Table 7-17. System RC Oscillator Characteristics Symbol Parameter Conditions fOUT Output frequency Calibrated at 85°C 7.7 Min Typ Max Unit 111.6 115 118.4 kHz Flash Characteristics Table 7-18 gives the device maximum operating frequency depending on the number of flash wait states and the flash read mode. The FSW bit in the FLASHCDW FSR register controls the number of wait states used when accessing the flash memory. Table 7-18. Maximum Operating Frequency Flash Wait States Read Mode Maximum Operating Frequency 1 50MHz High speed read mode 0 25MHz 1 30MHz Normal read mode 0 Table 7-19. 15MHz Flash Characteristics Symbol Parameter tFPP Page programming time tFPE Page erase time tFFP Fuse programming time tFEA Full chip erase time (EA) tFCE JTAG chip erase time (CHIP_ERASE) Conditions Min Typ Max Unit 5 5 fCLK_HSB = 50MHz 1 ms 5 fCLK_HSB = 115kHz 170 54 32099IS–01/2012 AT32UC3L016/32/64 Table 7-20. Flash Endurance and Data Retention Symbol Parameter NFARRAY Array endurance (write/page) 100k NFFUSE General Purpose fuses endurance (write/bit) 10k tRET Data retention 15 7.8 Min Typ Max Unit cycles years Analog Characteristics 7.8.1 Voltage Regulator Characteristics Table 7-21. VREG Electrical Characteristics Symbol Parameter VVDDIN Input voltage range VVDDCORE Output voltage, calibrated value Condition Min Typ Max 1.98 3.3 3.6 Units V Output voltage accuracy IOUT DC output current(1) IVREG Static current of internal regulator Note: Conditions VVDDIN >= 1.98V 1.8 IOUT = 0.1mA to 60mA, VVDDIN > 2.2V 2 IOUT = 0.1mA to 60mA, VVDDIN = 1.98V to 2.2V 4 % Normal mode 60 Low power mode 1 mA Normal mode 20 Low power mode 6 µA 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. Table 7-22. Decoupling Requirements Symbol Parameter CIN1 Input regulator capacitor 1 33 CIN2 Input regulator capacitor 2 100 CIN3 Input regulator capacitor 3 10 µF COUT1 Output regulator capacitor 1 100 nF COUT2 Output regulator capacitor 2 2.2 Note: Condition Typ Techno. Units nF Tantalum 0.5<ESR<10Ohm µF 1. Refer to Section 6.1.2 on page 36. 55 32099IS–01/2012 AT32UC3L016/32/64 7.8.2 Power-on Reset 18 Characteristics Table 7-23. POR18 Characteristics Symbol Parameter Condition VPOT+ Voltage threshold on VVDDCORE rising VPOT- Voltage threshold on VVDDCORE falling tDET Detection time Typ Max 1.45 1.58 Units V 1.2 Time with VDDCORE < VPOTnecessary to generate a reset signal 1.32 460 µs POR18 Operating Principles VVDDCORE Figure 7-5. Min VPOT+ Reset VPOT- 56 32099IS–01/2012 AT32UC3L016/32/64 7.8.3 Power-on Reset 33 Characteristics Table 7-24. POR33 Characteristics Symbol Parameter Condition VPOT+ Voltage threshold on VVDDIN rising VPOT- Voltage threshold on VVDDIN falling tDET Detection time Time with VDDIN < VPOTnecessary to generate a reset signal 460 µs IPOR33 Current consumption After tRESET 15 µA tSTARTUP Startup time 400 µs Typ Max 1.49 1.58 Units V 1.3 1.45 POR33 Operating Principles VVDDIN Figure 7-6. Min Reset VPOT+ VPOT- 7.8.4 Brown Out Detector Characteristics The values in Table 7-25 describe the values of the BODLEVEL in the flash General Purpose Fuse register. Table 7-25. BODLEVEL Values BODLEVEL Value Min Typ 011111 binary (31) 0x1F 1.56 100111 binary (39) 0x27 1.65 Max Units V 57 32099IS–01/2012 AT32UC3L016/32/64 Table 7-26. BOD Characteristics Symbol Parameter Condition VHYST BOD hysteresis T = 25°C 10 mV tDET Detection time Time with VDDCORE < BODLEVEL necessary to generate a reset signal 1 µs IBOD Current consumption 16 µA tSTARTUP Startup time 5 µs 7.8.5 Min Typ Max Units Supply Monitor 33 Characteristics Table 7-27. Symbol VTH SM33 Characteristics Parameter Voltage threshold Condition (1) Calibrated , T = 25°C Min Typ Max Units 1.675 1.75 1.825 V Step size, between adjacent values in SCIF.SM33.CALIB 11 VHYST Hysteresis 30 tDET Detection time Time with VDDIN < VTH necessary to generate a reset signal 280 µs ISM33 Current consumption Normal mode 15 µA tSTARTUP Startup time Normal mode 140 µs Note: mV 1. Calibration value can be read from the SCIF.SM33.CALIB field. This field is updated by the flash fuses after a reset. Refer to SCIF chapter for details. 58 32099IS–01/2012 AT32UC3L016/32/64 7.8.6 Analog to Digital Converter Characteristics Table 7-28. ADC Characteristics Symbol Parameter Conditions fADC ADC clock frequency tSTARTUP Startup time Return from Idle Mode tCONV Conversion time (latency) fADC = 6MHz Throughput rate Min Typ 10-bit resolution mode 6 8-bit resolution mode 6 15 11 VVDD > 3.0V, fADC = 6MHz, 10-bit resolution mode, low impedance source 460 VVDD > 3.0V, fADC = 6MHz, 8-bit resolution mode, low impedance source MHz 460 cycles kSPS Reference voltage range VADVREFP = VVDDANA IADC Current consumption on VVDDANA ADC Clock = 6MHz 300 IADVREFP Current consumption on ADVREFP pin fADC = 6MHz 250 1.62 1.98 V µA These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 7.8.6.1 Inputs and Sample and Hold Aquisition Time Table 7-29. Analog Inputs Symbol Parameter VADn Input Voltage Range CONCHIP Internal Capacitance(1) RONCHIP Note: Units µs 26 VADVREFP Note: Max Conditions Internal Resistance (1) 10-bit mode 8-bit mode Min 0 Typ Max Units VADVREFP V 21.5 pF VVDDIO = 3.0V to 3.6V, VVDDCORE = 1.8V 2.55 VVDDIO = VVDDCORE = 1.62V to 1.98V 55.3 kOhm 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. An analog voltage input must be able to charge the sample and hold (S/H) capacitor in the ADC in order to achieve maximum accuracy. Seen externally the ADC input consists of a resistor ( R ONCHIP ) and a capacitor ( CONCHIP ). In addition the resistance ( R SOURCE ) and capacitance ( C SOURCE ) of the PCB and source must be taken into account when calculating the sample and hold time. Figure 7-7 shows the ADC input channel equivalent circuit. 59 32099IS–01/2012 AT32UC3L016/32/64 Figure 7-7. ADC Input RSOURCE Positive Input RONCHIP CSOURCE VIN CONCHIP ADCVREFP/2 The minimum sample and hold time (in ns) can be found using this formula: t SAMPLEHOLD ≥ ( R ONCHIP + R OFFCHIP ) × ( C ONCHIP + C OFFCHIP ) × ln ( 2 n+1 ) Where n is the number of bits in the conversion. t SAMPLEHOLD is defined by the SHTIM field in the ADCIFB ACR register. Please refer to the ADCIFB chapter for more information. 7.8.6.2 Table 7-30. Applicable Conditions and Derating Data Transfer Characteristics 10-bit Resolution Mode Parameter Conditions Min Resolution Differential non-linearity Offset error Bit ADC clock frequency = 6MHz 1 +/-4 LSB +/-4 Transfer Characteristics 8-bit Resolution Mode Parameter Conditions Min Resolution Differential non-linearity Typ Max 8 Integral non-linearity Gain error Units +/-2 -0.9 Gain error Offset error Max 10 Integral non-linearity Table 7-31. Typ Units Bit +/-0.5 ADC clock frequency = 6MHz -0.23 0.25 +/-1 LSB +/-1 60 32099IS–01/2012 AT32UC3L016/32/64 7.8.7 Temperature Sensor Characteristics Temperature Sensor Characteristics(1) Table 7-32. Symbol Parameter Condition Min Typ Max Units Gradient 1 mV/°C ITS Current consumption 0.5 µA tSTARTUP Startup time 0 µs Note: 1. The Temperature Sensor is not calibrated. The accuracy of the Temperature Sensor is governed by the ADC accuracy. 7.8.8 Analog Comparator Characteristics Table 7-33. Symbol Analog Comparator Characteristics Parameter Condition Min Typ Max Positive input voltage range -0.2 VVDDIO + 0.3 Negative input voltage range -0.2 VVDDIO - 0.6 Units V Statistical offset VACREFN = 1.0V, fAC = 12MHz, filter length = 2, hysteresis = 0(1) Clock frequency for GCLK4 fAC Throughput rate(3) fAC = 12MHz Propagation delay Delay from input change to Interrupt Status Register Changes IAC Current consumption All channels, VDDIO = 3.3V, fA = 3MHz tSTARTUP Startup time Input current per pin Notes: 20 1 ⎛ + 3⎞⎠ × t CLKACIFB ⎝ t---------------------------------------CLKACIFB × f AC mV 12 MHz 12 000 000 Comparisons per second ns 420 µA 3 cycles 0.2 µA/MHz(2) 1. AC.CONFn.FLEN and AC.CONFn.HYS fields, refer to the Analog Comparator Interface chapter. 2. Referring to fAC. 3. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 61 32099IS–01/2012 AT32UC3L016/32/64 7.8.9 Capacitive Touch Characteristics 7.8.9.1 Table 7-34. Discharge Current Source DICS Characteristics Symbol Parameter RREF Internal resistor 120 kOhm k Trim step size 0.7 % 7.8.9.2 Table 7-35. Min Typ Max Unit Strong Pull-up Pull-down Strong Pull-up Pull-down Parameter Min Typ Pull-down resistor 1 Pull-up resistor 1 Max Unit kOhm 62 32099IS–01/2012 AT32UC3L016/32/64 7.9 Timing Characteristics 7.9.1 Startup, Reset, and Wake-up Timing The startup, reset, and wake-up timings are calculated using the following formula: t = t CONST + N CPU × t CPU Where t CONST and N CPU are found in Table 7-36. t CPU is the period of the CPU clock. If another clock source than RCSYS is selected as CPU clock the startup time of the oscillator, t OSCSTART , must added to the wake-up time in the stop, deepstop, and static sleep modes. Please refer to the source for the CPU clock in the ”Oscillator Characteristics” on page 50 for more details about oscillator startup times. Table 7-36. Maximum Reset and Wake-up Timing(1) Max t CONST (in µs) Max N CPU Parameter Measuring Startup time from power-up, using regulator Time from VDDIN crossing the VPOT+ threshold of POR33 to the first instruction entering the decode stage of CPU. VDDCORE is supplied by the internal regulator. 2210 0 Startup time from power-up, no regulator Time from VDDIN crossing the VPOT+ threshold of POR33 to the first instruction entering the decode stage of CPU. VDDCORE is connected to VDDIN. 1810 0 Startup time from reset release Time from releasing a reset source (except POR18, POR33, and SM33) to the first instruction entering the decode stage of CPU. 170 0 Idle 0 19 Frozen 0 110 0 110 27 + t OSCSTART 116 Deepstop 27 + t OSCSTART 116 Static 97 + t OSCSTART 116 1180 0 Standby Wake-up Stop Wake-up from shutdown Note: From wake-up event to the first instruction of an interrupt routine entering the decode stage of the CPU. From wake-up event to the first instruction entering the decode stage of the CPU. 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 7.9.2 RESET_N Timing Table 7-37. RESET_N Waveform Parameters(1) Symbol Parameter tRESET RESET_N minimum pulse length Note: Conditions Min 10 Max Units ns 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 63 32099IS–01/2012 AT32UC3L016/32/64 7.9.3 USART in SPI Mode Timing 7.9.3.1 Master mode Figure 7-8. USART in SPI Master Mode With (CPOL= CPHA= 0) or (CPOL= CPHA= 1) SPCK MISO USPI0 USPI1 MOSI USPI2 Figure 7-9. USART in SPI Master Mode With (CPOL= 0 and CPHA= 1) or (CPOL= 1 and CPHA= 0) SPCK MISO USPI3 USPI4 MOSI USPI5 Table 7-38. Symbol USART in SPI Mode Timing, Master Mode(1) Parameter Conditions USPI0 MISO setup time before SPCK rises USPI1 MISO hold time after SPCK rises USPI2 SPCK rising to MOSI delay USPI3 MISO setup time before SPCK falls USPI4 MISO hold time after SPCK falls USPI5 SPCK falling to MOSI delay Notes: Min 30.0+ VVDDIO from 3.0V to 3.6V, maximum external capacitor = 40pF Max Units tSAMPLE(2) 0 8.5 ns 25.5 + tSAMPLE(2) 0 13.6 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. t SPCK 1⎞ 2. Where: t SAMPLE = t SPCK – ⎛ -------------------------------------- × t CLKUSART ⎝ 2×t 2⎠ CLKUSART 64 32099IS–01/2012 AT32UC3L016/32/64 Maximum SPI Frequency, Master Output The maximum SPI master output frequency is given by the following formula: 1 f CLKSPI × 2 f SPCKMAX = MIN (f PINMAX,------------, -----------------------------) SPIn 9 Where SPIn is the MOSI delay, USPI2 or USPI5 depending on CPOL and NCPHA. f PINMAX is the maximum frequency of the SPI pins. Please refer to the I/O Pin Characteristics section for the maximum frequency of the pins. f CLKSPI is the maximum frequency of the CLK_SPI. Refer to the SPI chapter for a description of this clock. Maximum SPI Frequency, Master Input The maximum SPI master input frequency is given by the following formula: f CLKSPI × 2 1 f SPCKMAX = MIN (------------------------------------,-----------------------------) SPIn + t VALID 9 Where SPIn is the MISO setup and hold time, USPI0 + USPI1 or USPI3 + USPI4 depending on CPOL and NCPHA. T VALID is the SPI slave response time. Please refer to the SPI slave datasheet for T VALID . f CLKSPI is the maximum frequency of the CLK_SPI. Refer to the SPI chapter for a description of this clock. 7.9.3.2 Slave mode Figure 7-10. USART in SPI Slave Mode With (CPOL= 0 and CPHA= 1) or (CPOL= 1 and CPHA= 0) SPCK MISO USPI6 MOSI USPI7 USPI8 Figure 7-11. USART in SPI Slave Mode With (CPOL= CPHA= 0) or (CPOL= CPHA= 1) SPCK MISO USPI9 MOSI USPI10 USPI11 65 32099IS–01/2012 AT32UC3L016/32/64 Figure 7-12. USART in SPI Slave Mode NPCS Timing USPI12 USPI13 USPI14 USPI15 SPCK, CPOL=0 SPCK, CPOL=1 NSS Table 7-39. USART in SPI mode Timing, Slave Mode(1) Symbol Parameter Conditions USPI6 SPCK falling to MISO delay Max Units 27.6 USPI7 MOSI setup time before SPCK rises USPI8 MOSI hold time after SPCK rises USPI9 SPCK rising to MISO delay USPI10 MOSI setup time before SPCK falls USPI11 MOSI hold time after SPCK falls USPI12 NSS setup time before SPCK rises USPI13 NSS hold time after SPCK falls USPI14 NSS setup time before SPCK falls USPI15 NSS hold time after SPCK rises Notes: Min tSAMPLE(2) + tCLK_USART 0 VVDDIO from 3.0V to 3.6V, maximum external capacitor = 40pF 27.2 tSAMPLE(2) + tCLK_USART ns 0 25.0 0 25.0 0 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. t SPCK 1 2. Where: t SAMPLE = t SPCK – ⎛ -----------------------------------+ ---⎞ × t CLKUSART ⎝ 2×t 2⎠ CLKUSART Maximum SPI Frequency, Slave Input Mode The maximum SPI slave input frequency is given by the following formula: f CLKSPI × 2 1 f SPCKMAX = MIN (-----------------------------,------------) 9 SPIn Where SPIn is the MOSI setup and hold time, USPI7 + USPI8 or USPI10 + USPI11 depending on CPOL and NCPHA. f CLKSPI is the maximum frequency of the CLK_SPI. Refer to the SPI chapter for a description of this clock. Maximum SPI Frequency, Slave Output Mode 66 32099IS–01/2012 AT32UC3L016/32/64 The maximum SPI slave output frequency is given by the following formula: f CLKSPI × 2 1 f SPCKMAX = MIN (-----------------------------, f PINMAX,------------------------------------) 9 SPIn + t SETUP Where SPIn is the MISO delay, USPI6 or USPI9 depending on CPOL and NCPHA. T SETUP is the SPI master setup time. Please refer to the SPI master datasheet for T SETUP . f CLKSPI is the maximum frequency of the CLK_SPI. Refer to the SPI chapter for a description of this clock. f PINMAX is the maximum frequency of the SPI pins. Please refer to the I/O Pin Characteristics section for the maximum frequency of the pins. 7.9.4 SPI Timing 7.9.4.1 Master mode Figure 7-13. SPI Master Mode With (CPOL= NCPHA= 0) or (CPOL= NCPHA= 1) SPCK MISO SPI0 SPI1 MOSI SPI2 Figure 7-14. SPI Master Mode With (CPOL= 0 and NCPHA= 1) or (CPOL= 1 and NCPHA= 0) SPCK MISO SPI3 SPI4 MOSI SPI5 67 32099IS–01/2012 AT32UC3L016/32/64 Table 7-40. SPI Timing, Master Mode(1) Symbol Parameter SPI0 MISO setup time before SPCK rises SPI1 MISO hold time after SPCK rises SPI2 SPCK rising to MOSI delay SPI3 MISO setup time before SPCK falls SPI4 MISO hold time after SPCK falls SPI5 SPCK falling to MOSI delay Note: Conditions Min Max Units 28.4 + (tCLK_SPI)/2 VVDDIO from 3.0V to 3.6V, maximum external capacitor = 40pF 0 7.1 ns 22.8 + (tCLK_SPI)/2 0 11.0 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. Maximum SPI Frequency, Master Output The maximum SPI master output frequency is given by the following formula: 1 f SPCKMAX = MIN (f PINMAX,------------) SPIn Where SPIn is the MOSI delay, SPI2 or SPI5 depending on CPOL and NCPHA. f PINMAX is the maximum frequency of the SPI pins. Please refer to the I/O Pin Characteristics section for the maximum frequency of the pins. Maximum SPI Frequency, Master Input The maximum SPI master input frequency is given by the following formula: 1 f SPCKMAX = -----------------------------------SPIn + t VALID Where SPIn is the MISO setup and hold time, SPI0 + SPI1 or SPI3 + SPI4 depending on CPOL and NCPHA. t VALID is the SPI slave response time. Please refer to the SPI slave datasheet for t VALID . 7.9.4.2 Slave mode Figure 7-15. SPI Slave Mode With (CPOL= 0 and NCPHA= 1) or (CPOL= 1 and NCPHA= 0) SPCK MISO SPI6 MOSI SPI7 SPI8 68 32099IS–01/2012 AT32UC3L016/32/64 Figure 7-16. SPI Slave Mode With (CPOL= NCPHA= 0) or (CPOL= NCPHA= 1) SPCK MISO SPI9 MOSI SPI10 Figure 7-17. SPI11 SPI Slave Mode NPCS Timing SPI12 SPI13 SPI14 SPI15 SPCK, CPOL=0 SPCK, CPOL=1 NPCS Table 7-41. SPI Timing, Slave Mode(1) Symbol Parameter SPI6 SPCK falling to MISO delay SPI7 MOSI setup time before SPCK rises SPI8 MOSI hold time after SPCK rises SPI9 SPCK rising to MISO delay SPI10 MOSI setup time before SPCK falls SPI11 MOSI hold time after SPCK falls SPI12 NPCS setup time before SPCK rises SPI13 NPCS hold time after SPCK falls 0.2 SPI14 NPCS setup time before SPCK falls 2.2 SPI15 NPCS hold time after SPCK rises Note: Conditions Min Max Units 30.8 0 4.1 VVDDIO from 3.0V to 3.6V, maximum external capacitor = 40pF 29.9 0 ns 3.5 1.9 0 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. Maximum SPI Frequency, Slave Input Mode 69 32099IS–01/2012 AT32UC3L016/32/64 The maximum SPI slave input frequency is given by the following formula: 1 f SPCKMAX = MIN (f CLKSPI,------------) SPIn Where SPIn is the MOSI setup and hold time, SPI7 + SPI8 or SPI10 + SPI11 depending on CPOL and NCPHA. f CLKSPI is the maximum frequency of the CLK_SPI. Refer to the SPI chapter for a description of this clock. Maximum SPI Frequency, Slave Output Mode The maximum SPI slave output frequency is given by the following formula: 1 f SPCKMAX = MIN (f PINMAX,------------------------------------) SPIn + t SETUP Where SPIn is the MISO delay, SPI6 or SPI9 depending on CPOL and NCPHA. t SETUP is the SPI master setup time. Please refer to the SPI master datasheet for t SETUP . f PINMAX is the maximum frequency of the SPI pins. Please refer to the I/O Pin Characteristics section for the maximum frequency of the pins. 7.9.5 TWIM/TWIS Timing Figure 7-42 shows the TWI-bus timing requirements and the compliance of the device with them. Some of these requirements (tr and tf) are met by the device without requiring user intervention. Compliance with the other requirements (tHD-STA, tSU-STA, tSU-STO, tHD-DAT, tSU-DAT-TWI, tLOWTWI, tHIGH, and fTWCK) requires user intervention through appropriate programming of the relevant TWIM and TWIS user interface registers. Please refer to the TWIM and TWIS sections for more information. Table 7-42. TWI-Bus Timing Requirements Minimum Symbol Parameter Mode Requirement Standard(1) tr TWCK and TWD rise time tf TWCK and TWD fall time tHD-STA (Repeated) START hold time tSU-STA (Repeated) START set-up time tSU-STO STOP set-up time tHD-DAT Data hold time Maximum Device Requirement Device - 1000 20 + 0.1Cb 300 - 300 20 + 0.1Cb 300 Unit ns Fast(1) Standard ns Fast Standard 4 Fast 0.6 Standard 4.7 Fast 0.6 Standard 4.0 Fast 0.6 Standard Fast 0.3(2) tclkpb - μs tclkpb - μs 4tclkpb - μs 3.45() 2tclkpb 0.9() 15tprescaled + tclkpb μs 70 32099IS–01/2012 AT32UC3L016/32/64 Table 7-42. TWI-Bus Timing Requirements Minimum Symbol Parameter tSU-DAT-TWI Data set-up time tSU-DAT tLOW-TWI Mode Requirement Standard 250 Fast 100 - Device - Standard 4.7 Fast 1.3 TWCK LOW period tLOW - tHIGH TWCK HIGH period fTWCK TWCK frequency - Standard 4.0 Fast 0.6 Standard Notes: Maximum Requirement Unit 2tclkpb - ns tclkpb - - 4tclkpb - μs tclkpb - - 8tclkpb - μs 100 - Fast Device 400 1 -----------------------12t clkpb kHz 1. Standard mode: f TWCK ≤ 100 kHz ; fast mode: f TWCK > 100 kHz . 2. A device must internally provide a hold time of at least 300 ns for TWD with reference to the falling edge of TWCK. Notations: Cb = total capacitance of one bus line in pF tclkpb = period of TWI peripheral bus clock tprescaled = period of TWI internal prescaled clock (see chapters on TWIM and TWIS) The maximum tHD;DAT has only to be met if the device does not stretch the LOW period (tLOW-TWI) of TWCK. 71 32099IS–01/2012 AT32UC3L016/32/64 7.9.6 JTAG Timing Figure 7-18. JTAG Interface Signals JTAG2 TCK JTAG0 JTAG1 TMS/TDI JTAG3 JTAG4 JTAG7 JTAG8 TDO JTAG5 JTAG6 Boundary Scan Inputs Boundary Scan Outputs JTAG9 JTAG10 Table 7-43. JTAG Timings(1) Symbol Parameter JTAG0 TCK Low Half-period 23.2 JTAG1 TCK High Half-period 8.8 JTAG2 TCK Period 32.0 JTAG3 TDI, TMS Setup before TCK High JTAG4 TDI, TMS Hold after TCK High JTAG5 TDO Hold Time JTAG6 TCK Low to TDO Valid JTAG7 Boundary Scan Inputs Setup Time JTAG8 Boundary Scan Inputs Hold Time 5.0 JTAG9 Boundary Scan Outputs Hold Time 8.7 JTAG10 TCK to Boundary Scan Outputs Valid Note: Conditions VVDDIO from 3.0V to 3.6V, maximum external capacitor = 40pF Min Max Units 3.9 0.6 ns 4.5 23.2 0 17.7 1. These values are based on simulation and characterization of other AVR microcontrollers manufactured in the same process technology. These values are not covered by test limits in production. 72 32099IS–01/2012 AT32UC3L016/32/64 8. Mechanical Characteristics 8.1 8.1.1 Thermal Considerations Thermal Data Table 8-1 summarizes the thermal resistance data depending on the package. Table 8-1. 8.1.2 Thermal Resistance Data Symbol Parameter Condition Package Typ θJA Junction-to-ambient thermal resistance Still Air TQFP48 63.2 θJC Junction-to-case thermal resistance TQFP48 21.8 θJA Junction-to-ambient thermal resistance QFN48 28.3 θJC Junction-to-case thermal resistance QFN48 2.5 θJA Junction-to-ambient thermal resistance TLLGA48 25.4 θJC Junction-to-case thermal resistance TLLGA48 12.7 Still Air Still Air Unit °C/W °C/W °C/W Junction Temperature The average chip-junction temperature, TJ, in °C can be obtained from the following: 1. T J = T A + ( P D × θ JA ) 2. T J = T A + ( P D × ( θ HEATSINK + θ JC ) ) where: • θJA = package thermal resistance, Junction-to-ambient (°C/W), provided in Table 8-1. • θJC = package thermal resistance, Junction-to-case thermal resistance (°C/W), provided in Table 8-1. • θHEAT SINK = cooling device thermal resistance (°C/W), provided in the device datasheet. • PD = device power consumption (W) estimated from data provided in the Section 7.4 on page 42. • 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. 73 32099IS–01/2012 AT32UC3L016/32/64 8.2 Package Drawings Figure 8-1. TQFP-48 Package Drawing Table 8-2. Device and Package Maximum Weight 140 Table 8-3. mg Package Characteristics Moisture Sensitivity Level Table 8-4. MSL3 Package Reference JEDEC Drawing Reference MS-026 JESD97 Classification E3 74 32099IS–01/2012 AT32UC3L016/32/64 Figure 8-2. Note: QFN-48 Package Drawing The exposed pad is not connected to anything internally, but should be soldered to ground to increase board level reliability. Table 8-5. Device and Package Maximum Weight 140 Table 8-6. mg Package Characteristics Moisture Sensitivity Level Table 8-7. MSL3 Package Reference JEDEC Drawing Reference M0-220 JESD97 Classification E3 75 32099IS–01/2012 AT32UC3L016/32/64 Figure 8-3. TLLGA-48 Package Drawing Table 8-8. Device and Package Maximum Weight 39.3 Table 8-9. mg Package Characteristics Moisture Sensitivity Level Table 8-10. MSL3 Package Reference JEDEC Drawing Reference N/A JESD97 Classification E4 76 32099IS–01/2012 AT32UC3L016/32/64 8.3 Soldering Profile Table 8-11 gives the recommended soldering profile from J-STD-20. Table 8-11. Soldering Profile Profile Feature Green Package Average Ramp-up Rate (217°C to Peak) 3°C/s max Preheat Temperature 175°C ±25°C 150-200°C Time Maintained Above 217°C 60-120s Time within 5°C of Actual Peak Temperature 30s Peak Temperature Range 260°C Ramp-down Rate 6°C/s max Time 25°C to Peak Temperature 8 minutes max A maximum of three reflow passes is allowed per component. 77 32099IS–01/2012 AT32UC3L016/32/64 9. Ordering Information Table 9-1. Ordering Information Device Ordering Code Carrier Type AT32UC3L064-AUTES ES AT32UC3L064-AUT Tray AT32UC3L064-AUR Tape & Reel Package Package Type Temperature Operating Range TQFP 48 JESD97 Classification E3 AT32UC3L064 AT32UC3L064-ZAUES ES AT32UC3L064-ZAUT Tray AT32UC3L064-ZAUR Tape & Reel AT32UC3L064-D3HES ES AT32UC3L064-D3HT Tray AT32UC3L064-D3HR Tape & Reel AT32UC3L032-AUT Tray AT32UC3L032-AUR Tape & Reel QFN 48 TLLGA 48 JESD97 Classification E4 TQFP 48 Industrial (-40°C to 85°C) JESD97 Classification E3 AT32UC3L032-ZAUT Tray AT32UC3L032-ZAUR Tape & Reel AT32UC3L032-D3HT Tray AT32UC3L032-D3HR Tape & Reel AT32UC3L032 QFN 48 TLLGA 48 AT32UC3L016-AUT Tray AT32UC3L016-AUR Tape & Reel AT32UC3L016-ZAUT Tray AT32UC3L016-ZAUR Tape & Reel AT32UC3L016-D3HT Tray AT32UC3L016-D3HR Tape & Reel JESD97 Classification E4 TQFP 48 JESD97 Classification E3 AT32UC3L016 QFN 48 TLLGA 48 JESD97 Classification E4 78 32099IS–01/2012 AT32UC3L016/32/64 10. Errata 10.1 10.1.1 Rev. E Processor and Architecture 1. Hardware breakpoints may corrupt MAC results Hardware breakpoints on MAC instructions may corrupt the destination register of the MAC instruction. Fix/Workaround Place breakpoints on earlier or later instructions. 2. 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. 10.1.2 FLASHCDW 1. Flash self programming may fail in one wait state mode Writes in flash and user pages may fail if executing code is located in address space mapped to flash, and the flash controller is configured in one wait state mode (the Flash Wait State bit in the Flash Control Register (FCR.FWS) is one). Fix/Workaround Solution 1: Configure the flash controller in zero wait state mode (FCR.FWS=0). Solution 2: Configure the HMATRIX master 1 (CPU Instruction) to use the unlimited burst length transfer mode (MCFG1.ULBT=0), and the HMATRIX slave 0 (FLASHCDW) to use the maximum slot cycle limit (SCFG0.SLOT_CYCLE=255). 10.1.3 Power Manager 1. Clock sources will not be stopped in Static mode if the difference between CPU and PBx division factor is larger than 4 If the division factor between the CPU/HSB and PBx frequencies is more than 4 when entering a sleep mode where the system RC oscillator (RCSYS) is turned off, the 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 RCSYS is stopped, make sure the division factor between CPU/HSB and PBx frequencies is less than or equal to 4. 2. Clock Failure Detector (CFD) can be issued while turning off the CFD While turning off the CFD, the CFD bit in the Status Register (SR) can be set. This will change the main clock source to RCSYS. Fix/Workaround Solution 1: Enable CFD interrupt. If CFD interrupt is issues after turning off the CFD, switch back to original main clock source. Solution 2: Only turn off the CFD while running the main clock on RCSYS. 3. Sleepwalking in idle and frozen sleep mode will mask all other PB clocks 79 32099IS–01/2012 AT32UC3L016/32/64 If the CPU is in idle or frozen sleep mode and a module is in a state that triggers sleep walking, all PB clocks will be masked except the PB clock to the sleepwalking module. Fix/Workaround Mask all clock requests in the PM.PPCR register before going into idle or frozen mode. 10.1.4 SCIF 1. PCLKSR.OSC32RDY bit might not be cleared after disabling OSC32K In some cases the OSC32RDY bit in the PCLKSR register will not be cleared when OSC32K is disabled. Fix/Workaround When re-enabling the OSC32K, read the PCLKSR.OSC32RDY bit. If this bit is: 0: Follow normal procedures. 1: Ignore the PCLKSR.OSC32RDY and ISR.OSC32RDY bit. Use the Frequency Meter (FREQM) to determine if the OSC32K clock is ready. The OSC32K clock is ready when the FREQM measures a non-zero frequency. 2. The RC32K output on PA20 is not always permanently disabled The RC32K output on PA20 may sometimes re-appear. Fix/Workaround Before using RC32K for other purposes, the following procedure has to be followed in order to properly disable it: - Run the CPU on RCSYS - Disable the output to PA20 by writing a zero to PM.PPCR.RC32OUT - Enable RC32K by writing a one to SCIF.RC32KCR.EN, and wait for this bit to be read as one - Disable RC32K by writing a zero to SCIF.RC32KCR.EN, and wait for this bit to be read as zero. 10.1.5 AST 1. Reset may set status bits in the AST If a reset occurs and the AST is enabled, the SR.ALARM0, SR.PER0, and SR.OVF bits may be set. Fix/Workaround If the part is reset and the AST is used, clear all bits in the Status Register before entering sleep mode. 2. AST wake signal is released one AST clock cycle after the BUSY bit is cleared After writing to the Status Clear Register (SCR) the wake signal is released one AST clock cycle after the BUSY bit in the Status Register (SR.BUSY) is cleared. If entering sleep mode directly after the BUSY bit is cleared the part will wake up immediately. Fix/Workaround Read the Wake Enable Register (WER) and write this value back to the same register. Wait for BUSY to clear before entering sleep mode. 10.1.6 WDT 1. Clearing the Watchdog Timer (WDT) counter in second half of timeout period will issue a Watchdog reset If the WDT counter is cleared in the second half of the timeout period, the WDT will immediately issue a Watchdog reset. Fix/Workaround 80 32099IS–01/2012 AT32UC3L016/32/64 Use twice as long timeout period as needed and clear the WDT counter within the first half of the timeout period. If the WDT counter is cleared after the first half of the timeout period, you will get a Watchdog reset immediately. If the WDT counter is not cleared at all, the time before the reset will be twice as long as needed. 2. WDT Control Register does not have synchronization feedback When writing to the Timeout Prescale Select (PSEL), Time Ban Prescale Select (TBAN), Enable (EN), or WDT Mode (MODE) fieldss of the WDT Control Register (CTRL), a synchronizer is started to propagate the values to the WDT clcok domain. This synchronization takes a finite amount of time, but only the status of the synchronization of the EN bit is reflected back to the user. Writing to the synchronized fields during synchronization can lead to undefined behavior. Fix/Workaround -When writing to the affected fields, the user must ensure a wait corresponding to 2 clock cycles of both the WDT peripheral bus clock and the selected WDT clock source. -When doing writes that changes the EN bit, the EN bit can be read back until it reflects the written value. 10.1.7 GPIO 1. Clearing GPIO interrupt may fail Writing a one to the GPIO.IFRC register to clear an interrupt will be ignored if interrupt is enabled for the corresponding port. Fix/Workaround Disable the interrupt, clear it by writing a one to GPIO.IFRC, then enable the interrupt. 10.1.8 SPI 1. SPI data transfer hangs with CSR0.CSAAT==1 and MR.MODFDIS==0 When CSR0.CSAAT==1 and mode fault detection is enabled (MR.MODFDIS==0), the SPI module will not start a data transfer. Fix/Workaround Disable mode fault detection by writing a one to MR.MODFDIS. 2. Disabling SPI has no effect on the SR.TDRE bit Disabling SPI has no effect on the SR.TDRE bit whereas the write data command is filtered when SPI is disabled. Writing to TDR when SPI is disabled will not clear SR.TDRE. If SPI is disabled during a PDCA transfer, the PDCA will continue to write data to TDR until its buffer is empty, and this data will be lost. Fix/Workaround Disable the PDCA, add two NOPs, and disable the SPI. To continue the transfer, enable the SPI and PDCA. 3. 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 by writing a one to the Software Reset bit in the Control Register (CR.SWRST). 4. SPI bad serial clock generation on 2nd chip_select when SCBR=1, CPOL=1, and NCPHA=0 When multiple chip selects (CS) are in use, if one of the baudrates equal 1 while one (CSRn.SCBR=1) of the others do not equal 1, and CSRn.CPOL=1 and CSRn.NCPHA=0, then an additional pulse will be generated on SCK. 81 32099IS–01/2012 AT32UC3L016/32/64 Fix/Workaround When multiple CS are in use, if one of the baudrates equals 1, the others must also equal 1 if CSRn.CPOL=1 and CSRn.NCPHA=0. 5. SPI mode fault detection enable causes incorrect behavior When mode fault detection is enabled (MR.MODFDIS==0), the SPI module may not operate properly. Fix/Workaround Always disable mode fault detection before using the SPI by writing a one to MR.MODFDIS. 10.1.9 TWI 1. TWIM SR.IDLE goes high immediately when NAK is received When a NAK is received and there is a non-zero number of bytes to be transmitted, SR.IDLE goes high immediately and does not wait for the STOP condition to be sent. This does not cause any problem just by itself, but can cause a problem if software waits for SR.IDLE to go high and then immediately disables the TWIM by writing a one to CR.MDIS. Disabling the TWIM causes the TWCK and TWD pins to go high immediately, so the STOP condition will not be transmitted correctly. Fix/Workaround If possible, do not disable the TWIM. If it is absolutely necessary to disable the TWIM, there must be a software delay of at least two TWCK periods between the detection of SR.IDLE==1 and the disabling of the TWIM. 2. TWIM TWALM polarity is wrong The TWALM signal in the TWIM is active high instead of active low. Fix/Workaround Use an external inverter to invert the signal going into the TWIM. When using both TWIM and TWIS on the same pins, the TWALM cannot be used. 3. TWIS may not wake the device from sleep mode If the CPU is put to a sleep mode (except Idle and Frozen) directly after a TWI Start condition, the CPU may not wake upon a TWIS address match. The request is NACKed. Fix/Workaround When using the TWI address match to wake the device from sleep, do not switch to sleep modes deeper than Frozen. Another solution is to enable asynchronous EIC wake on the TWIS clock (TWCK) or TWIS data (TWD) pins, in order to wake the system up on bus events. 4. SMBALERT bit may be set after reset The SMBus Alert (SMBALERT) bit in the Status Register (SR) might be erroneously set after system reset. Fix/Workaround After system reset, clear the SR.SMBALERT bit before commencing any TWI transfer. 5. Clearing the NAK bit before the BTF bit is set locks up the TWI bus When the TWIS is in transmit mode, clearing the NAK Received (NAK) bit of the Status Register (SR) before the end of the Acknowledge/Not Acknowledge cycle will cause the TWIS to attempt to continue transmitting data, thus locking up the bus. Fix/Workaround Clear SR.NAK only after the Byte Transfer Finished (BTF) bit of the same register has been set. 82 32099IS–01/2012 AT32UC3L016/32/64 6. TWIS stretch on Address match error When the TWIS stretches TWCK due to a slave address match, it also holds TWD low for the same duration if it is to be receiving data. When TWIS releases TWCK, it releases TWD at the same time. This can cause a TWI timing violation. Fix/Workaround None. 10.1.10 PWMA 1. BUSY bit is never cleared after writes to the Control Register (CR) When writing a non-zero value to CR.TOP, CR.SPREAD, or CR.TCLR when the PWMA is disabled (CR.EN == 0), the BUSY bit in the Status Register (SR.BUSY) will be set, but never cleared. Fix/Workaround When writing a non-zero value to CR.TOP, CR.SPREAD, or CR.TCLR, make sure the PWMA is enabled, or simultaneously enable the PWMA by writing a one to CR.EN. 2. Incoming peripheral events are discarded during duty cycle register update Incoming peripheral events to all applied channels will be discarded if a duty cycle update is received from the user interface in the same PWMA clock period. Fix/Workaround Ensure that duty cycle writes from the user interface are not performed in a PWMA period when an incoming peripheral event is expected. 10.1.11 ADCIFB 1. Using STARTUPTIME larger than 0x1F will freeze the ADC Writing a value larger than 0x1F to the Startup Time field in the ADC Configuration Register (ACR.STARTUP) will freeze the ADC, and the Busy Status bit in the Status Register (SR.BUSY) will never be cleared. Fix/Workaround Do not write values larger than 0x1F to ACR.STARTUP. 10.1.12 CAT 1. CAT asynchronous wake will be delayed by one AST event period If the CAT detects a condition the should asynchronously wake the device in static mode, the asynchronous wake will not occur until the next AST event. For example, if the AST is generating events to the CAT every 50ms, and the CAT detects a touch at t=9200ms, the asynchronous wake will occur at t=9250ms. Fix/Workaround None. 2. CAT QMatrix sense capacitors discharged prematurely At the end of a QMatrix burst charging sequence that uses different burst count values for different Y lines, the Y lines may be incorrectly grounded for up to n-1 periods of the peripheral bus clock, where n is the ratio of the PB clock frequency to the GCLK_CAT frequency. This results in premature loss of charge from the sense capacitors and thus increased variability of the acquired count values. Fix/Workaround Enable the 1kOhm drive resistors on all implemented QMatrix Y lines (CSA 1, 3, 5, 7, 9, 11, 13, and/or 15) by writing ones to the corresponding odd bits of the CSARES register. 3. CAT module does not terminate QTouch burst on detect 83 32099IS–01/2012 AT32UC3L016/32/64 The CAT module does not terminate a QTouch burst when the detection voltage is reached on the sense capacitor. This can cause the sense capacitor to be charged more than necessary. Depending on the dielectric absorption characteristics of the capacitor, this can lead to unstable measurements. Fix/Workaround Use the minimum possible value for the MAX field in the ATCFG1, TG0CFG1, and TG1CFG1 registers. 4. Autonomous CAT acquisition must be longer than AST source clock period When using the AST to trigger CAT autonomous touch acquisition in sleep modes where the CAT bus clock is turned off, the CAT will start several acquisitions if the period of the AST source clock is larger than one CAT acquisition. One AST clock period after the AST trigger, the CAT clock will automatically stop and the CAT acquisition can be stopped prematurely, ruining the result. Fix/Workaround Always ensure that the ATCFG1.max field is set so that the duration of the autonomous touch acquisition is greater than one clock period of the AST source clock. 10.1.13 aWire 1. aWire CPU clock speed robustness The aWire memory speed request command counter warps at clock speeds below approximately 5kHz. Fix/Workaround None. 2. The aWire debug interface is reset after leaving Shutdown mode If the aWire debug mode is used as debug interface and the program enters Shutdown mode, the aWire interface will be reset when the part receives a wakeup either from the WAKE_N pin or the AST. Fix/Workaround None. 10.1.14 CHIP 1. Increased Power Consumption in VDDIO in sleep modes If 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 Solution 1: Disable OSC0 by writing a zero to the Oscillator Enable bit in the System Control Interface (SCIF) Oscillator Control Register (SCIF.OSC0CTRL.OSCEN) before going to a sleep mode where OSC0 is disabled. Solution 2: Pull down or up XIN0 or XOUT0 with 1MOhm resistor. 10.1.15 I/O Pins 1. PA17 has low ESD tolerance PA17 only tolerates 500V ESD pulses (Human Body Model). Fix/Workaround Care must be taken during manufacturing and PCB design. 84 32099IS–01/2012 AT32UC3L016/32/64 10.2 10.2.1 Rev. D Processor and Architecture 1. Hardware breakpoints may corrupt MAC results Hardware breakpoints on MAC instructions may corrupt the destination register of the MAC instruction. Fix/Workaround Place breakpoints on earlier or later instructions. 2. 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. 10.2.2 FLASHCDW 1. Flash self programming may fail in one wait state mode Writes in flash and user pages may fail if executing code is located in address space mapped to flash, and the flash controller is configured in one wait state mode (the Flash Wait State bit in the Flash Control Register (FCR.FWS) is one). Fix/Workaround Solution 1: Configure the flash controller in zero wait state mode (FCR.FWS=0). Solution 2: Configure the HMATRIX master 1 (CPU Instruction) to use the unlimited burst length transfer mode (MCFG1.ULBT=0), and the HMATRIX slave 0 (FLASHCDW) to use the maximum slot cycle limit (SCFG0.SLOT_CYCLE=255). 10.2.3 Power Manager 1. Clock sources will not be stopped in Static mode if the difference between CPU and PBx division factor is larger than 4 If the division factor between the CPU/HSB and PBx frequencies is more than 4 when entering a sleep mode where the system RC oscillator (RCSYS) is turned off, the 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 RCSYS is stopped, make sure the division factor between CPU/HSB and PBx frequencies is less than or equal to 4. 2. External reset in Shutdown mode If an external reset is asserted while the device is in Shutdown mode, the Power Manager will register this as a Power-on reset (POR), and not as a SLEEP reset, in the Reset Cause register (RCAUSE) Fix/Workaround None. 3. Disabling POR33 may generate spurious resets Depending on operating conditions, POR33 may generate a spurious reset in one of the following cases: - When POR33 is disabled from the user interface - When SM33 supply monitor is enabled 85 32099IS–01/2012 AT32UC3L016/32/64 - When entering Shutdown mode while debugging the chip using JTAG or aWire interface In the listed cases, writing a one to the bit VREGCR.POR33MASK in the System Control Interface (SCIF) to mask the POR33 reset will be ineffective Fix/Workaround - Do not disable POR33 using the user interface - Do not use the SM33 supply monitor - Do not enter Shutdown mode if a debugger is connected to the chip 4. Instability when exiting sleep walking If all the following operating conditions are true, exiting sleep walking might lead to instability: - The OSC0 is enabled in external clock mode (OSCCTRL0.OSCEN == 1 and OSCCTRL0.MODE == 0) - A sleep mode where the OSC0 is automatically disabled is entered - The device enters sleep walking Fix/Workaround Do not run OSC0 in external clock mode if sleepwalking is expected to be used. 5. Clock Failure Detector (CFD) can be issued while turning off the CFD While turning off the CFD, the CFD bit in the Status Register (SR) can be set. This will change the main clock source to RCSYS. Fix/Workaround Solution 1: Enable CFD interrupt. If CFD interrupt is issues after turning off the CFD, switch back to original main clock source. Solution 2: Only turn off the CFD while running the main clock on RCSYS. 6. Sleepwalking in idle and frozen sleep mode will mask all other PB clocks If the CPU is in idle or frozen sleep mode and a module is in a state that triggers sleep walking, all PB clocks will be masked except the PB clock to the sleepwalking module. Fix/Workaround Mask all clock requests in the PM.PPCR register before going into idle or frozen mode. 10.2.4 SCIF 1. PCLKSR.OSC32RDY bit might not be cleared after disabling OSC32K In some cases the OSC32RDY bit in the PCLKSR register will not be cleared when OSC32K is disabled. Fix/Workaround When re-enabling the OSC32K, read the PCLKSR.OSC32RDY bit. If this bit is: 0: Follow normal procedures. 1: Ignore the PCLKSR.OSC32RDY and ISR.OSC32RDY bit. Use the Frequency Meter (FREQM) to determine if the OSC32K clock is ready. The OSC32K clock is ready when the FREQM measures a non-zero frequency. 2. The RC32K output on PA20 is not always permanently disabled The RC32K output on PA20 may sometimes re-appear. Fix/Workaround Before using RC32K for other purposes, the following procedure has to be followed in order to properly disable it: - Run the CPU on RCSYS - Disable the output to PA20 by writing a zero to PM.PPCR.RC32OUT - Enable RC32K by writing a one to SCIF.RC32KCR.EN, and wait for this bit to be read as one 86 32099IS–01/2012 AT32UC3L016/32/64 - Disable RC32K by writing a zero to SCIF.RC32KCR.EN, and wait for this bit to be read as zero. 10.2.5 AST 1. Reset may set status bits in the AST If a reset occurs and the AST is enabled, the SR.ALARM0, SR.PER0, and SR.OVF bits may be set. Fix/Workaround If the part is reset and the AST is used, clear all bits in the Status Register before entering sleep mode. 2. AST wake signal is released one AST clock cycle after the BUSY bit is cleared After writing to the Status Clear Register (SCR) the wake signal is released one AST clock cycle after the BUSY bit in the Status Register (SR.BUSY) is cleared. If entering sleep mode directly after the BUSY bit is cleared the part will wake up immediately. Fix/Workaround Read the Wake Enable Register (WER) and write this value back to the same register. Wait for BUSY to clear before entering sleep mode. 10.2.6 WDT 1. Clearing the Watchdog Timer (WDT) counter in second half of timeout period will issue a Watchdog reset If the WDT counter is cleared in the second half of the timeout period, the WDT will immediately issue a Watchdog reset. Fix/Workaround Use twice as long timeout period as needed and clear the WDT counter within the first half of the timeout period. If the WDT counter is cleared after the first half of the timeout period, you will get a Watchdog reset immediately. If the WDT counter is not cleared at all, the time before the reset will be twice as long as needed. 2. WDT Control Register does not have synchronization feedback When writing to the Timeout Prescale Select (PSEL), Time Ban Prescale Select (TBAN), Enable (EN), or WDT Mode (MODE) fieldss of the WDT Control Register (CTRL), a synchronizer is started to propagate the values to the WDT clcok domain. This synchronization takes a finite amount of time, but only the status of the synchronization of the EN bit is reflected back to the user. Writing to the synchronized fields during synchronization can lead to undefined behavior. Fix/Workaround -When writing to the affected fields, the user must ensure a wait corresponding to 2 clock cycles of both the WDT peripheral bus clock and the selected WDT clock source. -When doing writes that changes the EN bit, the EN bit can be read back until it reflects the written value. 10.2.7 GPIO 1. Clearing GPIO interrupt may fail Writing a one to the GPIO.IFRC register to clear an interrupt will be ignored if interrupt is enabled for the corresponding port. Fix/Workaround Disable the interrupt, clear it by writing a one to GPIO.IFRC, then enable the interrupt. 87 32099IS–01/2012 AT32UC3L016/32/64 10.2.8 SPI 1. SPI data transfer hangs with CSR0.CSAAT==1 and MR.MODFDIS==0 When CSR0.CSAAT==1 and mode fault detection is enabled (MR.MODFDIS==0), the SPI module will not start a data transfer. Fix/Workaround Disable mode fault detection by writing a one to MR.MODFDIS. 2. Disabling SPI has no effect on the SR.TDRE bit Disabling SPI has no effect on the SR.TDRE bit whereas the write data command is filtered when SPI is disabled. Writing to TDR when SPI is disabled will not clear SR.TDRE. If SPI is disabled during a PDCA transfer, the PDCA will continue to write data to TDR until its buffer is empty, and this data will be lost. Fix/Workaround Disable the PDCA, add two NOPs, and disable the SPI. To continue the transfer, enable the SPI and PDCA. 3. 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 by writing a one to the Software Reset bit in the Control Register (CR.SWRST). 4. SPI bad serial clock generation on 2nd chip_select when SCBR=1, CPOL=1, and NCPHA=0 When multiple chip selects (CS) are in use, if one of the baudrates equal 1 while one (CSRn.SCBR=1) of the others do not equal 1, and CSRn.CPOL=1 and CSRn.NCPHA=0, then an additional pulse will be generated on SCK. Fix/Workaround When multiple CS are in use, if one of the baudrates equals 1, the others must also equal 1 if CSRn.CPOL=1 and CSRn.NCPHA=0. 5. SPI mode fault detection enable causes incorrect behavior When mode fault detection is enabled (MR.MODFDIS==0), the SPI module may not operate properly. Fix/Workaround Always disable mode fault detection before using the SPI by writing a one to MR.MODFDIS. 10.2.9 TWI 1. TWIM SR.IDLE goes high immediately when NAK is received When a NAK is received and there is a non-zero number of bytes to be transmitted, SR.IDLE goes high immediately and does not wait for the STOP condition to be sent. This does not cause any problem just by itself, but can cause a problem if software waits for SR.IDLE to go high and then immediately disables the TWIM by writing a one to CR.MDIS. Disabling the TWIM causes the TWCK and TWD pins to go high immediately, so the STOP condition will not be transmitted correctly. Fix/Workaround If possible, do not disable the TWIM. If it is absolutely necessary to disable the TWIM, there must be a software delay of at least two TWCK periods between the detection of SR.IDLE==1 and the disabling of the TWIM. 2. TWIM TWALM polarity is wrong 88 32099IS–01/2012 AT32UC3L016/32/64 The TWALM signal in the TWIM is active high instead of active low. Fix/Workaround Use an external inverter to invert the signal going into the TWIM. When using both TWIM and TWIS on the same pins, the TWALM cannot be used. 3. TWIS may not wake the device from sleep mode If the CPU is put to a sleep mode (except Idle and Frozen) directly after a TWI Start condition, the CPU may not wake upon a TWIS address match. The request is NACKed. Fix/Workaround When using the TWI address match to wake the device from sleep, do not switch to sleep modes deeper than Frozen. Another solution is to enable asynchronous EIC wake on the TWIS clock (TWCK) or TWIS data (TWD) pins, in order to wake the system up on bus events. 4. SMBALERT bit may be set after reset The SMBus Alert (SMBALERT) bit in the Status Register (SR) might be erroneously set after system reset. Fix/Workaround After system reset, clear the SR.SMBALERT bit before commencing any TWI transfer. 5. Clearing the NAK bit before the BTF bit is set locks up the TWI bus When the TWIS is in transmit mode, clearing the NAK Received (NAK) bit of the Status Register (SR) before the end of the Acknowledge/Not Acknowledge cycle will cause the TWIS to attempt to continue transmitting data, thus locking up the bus. Fix/Workaround Clear SR.NAK only after the Byte Transfer Finished (BTF) bit of the same register has been set. 6. TWIS stretch on Address match error When the TWIS stretches TWCK due to a slave address match, it also holds TWD low for the same duration if it is to be receiving data. When TWIS releases TWCK, it releases TWD at the same time. This can cause a TWI timing violation. Fix/Workaround None. 10.2.10 PWMA 1. BUSY bit is never cleared after writes to the Control Register (CR) When writing a non-zero value to CR.TOP, CR.SPREAD, or CR.TCLR when the PWMA is disabled (CR.EN == 0), the BUSY bit in the Status Register (SR.BUSY) will be set, but never cleared. Fix/Workaround When writing a non-zero value to CR.TOP, CR.SPREAD, or CR.TCLR, make sure the PWMA is enabled, or simultaneously enable the PWMA by writing a one to CR.EN. 2. Incoming peripheral events are discarded during duty cycle register update Incoming peripheral events to all applied channels will be discarded if a duty cycle update is received from the user interface in the same PWMA clock period. Fix/Workaround Ensure that duty cycle writes from the user interface are not performed in a PWMA period when an incoming peripheral event is expected. 89 32099IS–01/2012 AT32UC3L016/32/64 10.2.11 ADCIFB 1. Using STARTUPTIME larger than 0x1F will freeze the ADC Writing a value larger than 0x1F to the Startup Time field in the ADC Configuration Register (ACR.STARTUP) will freeze the ADC, and the Busy Status bit in the Status Register (SR.BUSY) will never be cleared. Fix/Workaround Do not write values larger than 0x1F to ACR.STARTUP. 10.2.12 CAT 1. CAT asynchronous wake will be delayed by one AST event period If the CAT detects a condition the should asynchronously wake the device in static mode, the asynchronous wake will not occur until the next AST event. For example, if the AST is generating events to the CAT every 50ms, and the CAT detects a touch at t=9200ms, the asynchronous wake will occur at t=9250ms. Fix/Workaround None. 2. CAT QMatrix sense capacitors discharged prematurely At the end of a QMatrix burst charging sequence that uses different burst count values for different Y lines, the Y lines may be incorrectly grounded for up to n-1 periods of the peripheral bus clock, where n is the ratio of the PB clock frequency to the GCLK_CAT frequency. This results in premature loss of charge from the sense capacitors and thus increased variability of the acquired count values. Fix/Workaround Enable the 1kOhm drive resistors on all implemented QMatrix Y lines (CSA 1, 3, 5, 7, 9, 11, 13, and/or 15) by writing ones to the corresponding odd bits of the CSARES register. 3. CAT module does not terminate QTouch burst on detect The CAT module does not terminate a QTouch burst when the detection voltage is reached on the sense capacitor. This can cause the sense capacitor to be charged more than necessary. Depending on the dielectric absorption characteristics of the capacitor, this can lead to unstable measurements. Fix/Workaround Use the minimum possible value for the MAX field in the ATCFG1, TG0CFG1, and TG1CFG1 registers. 4. Autonomous CAT acquisition must be longer than AST source clock period When using the AST to trigger CAT autonomous touch acquisition in sleep modes where the CAT bus clock is turned off, the CAT will start several acquisitions if the period of the AST source clock is larger than one CAT acquisition. One AST clock period after the AST trigger, the CAT clock will automatically stop and the CAT acquisition can be stopped prematurely, ruining the result. Fix/Workaround Always ensure that the ATCFG1.max field is set so that the duration of the autonomous touch acquisition is greater than one clock period of the AST source clock. 10.2.13 aWire 1. aWire CPU clock speed robustness The aWire memory speed request command counter warps at clock speeds below approximately 5kHz. Fix/Workaround 90 32099IS–01/2012 AT32UC3L016/32/64 None. 2. The aWire debug interface is reset after leaving Shutdown mode If the aWire debug mode is used as debug interface and the program enters Shutdown mode, the aWire interface will be reset when the part receives a wakeup either from the WAKE_N pin or the AST. Fix/Workaround None. 10.2.14 CHIP 1. In 3.3V Single Supply Mode, the Analog Comparator inputs affects the device’s ability to start When using the 3.3V Single Supply Mode the state of the Analog Comparator input pins can affect the device’s ability to release POR reset. This is due to an interaction between the Analog Comparator input pins and the POR circuitry. The issue is not present in the 1.8V Supply Mode or the 3.3V Supply Mode with 1.8V Regulated I/O Lines. Fix/Workaround ACREFN (pin PA16) must be connected to GND until the POR reset is released and the Analog Comparator inputs should not be driven higher than 1.0 V until the POR reset is released. 2. Increased Power Consumption in VDDIO in sleep modes If 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 Solution 1: Disable OSC0 by writing a zero to the Oscillator Enable bit in the System Control Interface (SCIF) Oscillator Control Register (SCIF.OSC0CTRL.OSCEN) before going to a sleep mode where OSC0 is disabled. Solution 2: Pull down or up XIN0 or XOUT0 with 1MOhm resistor. 10.2.15 I/O Pins 1. PA17 has low ESD tolerance PA17 only tolerates 500V ESD pulses (Human Body Model). Fix/Workaround Care must be taken during manufacturing and PCB design. 10.3 Rev. C Not sampled. 10.4 10.4.1 Rev. B Processor and Architecture 1. RETS behaves incorrectly when MPU is enabled RETS behaves incorrectly when MPU is enabled and MPU is configured so that system stack is not readable in unprivileged mode. Fix/Workaround Make system stack readable in unprivileged mode, or return from supervisor mode using rete instead of rets. This requires: 1. Changing the mode bits from 001 to 110 before issuing the instruction. Updating the mode bits to the desired value must be done using a single mtsr instruction so it is done 91 32099IS–01/2012 AT32UC3L016/32/64 atomically. Even if this step is generally described as not safe in the UC technical reference manual, it is safe in this very specific case. 2. Execute the RETE instruction. 2. Hardware breakpoints may corrupt MAC results Hardware breakpoints on MAC instructions may corrupt the destination register of the MAC instruction. Fix/Workaround Place breakpoints on earlier or later instructions. 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 Make a DTLB Protection (Write) exception handler which permits the interrupt request to be handled in privileged mode. 10.4.2 PDCA 1. PCONTROL.CHxRES is non-functional PCONTROL.CHxRES is non-functional. Counters are reset at power-on, and cannot be reset by software. Fix/Workaround Software needs to keep history of performance counters. 2. Transfer error will stall a transmit peripheral handshake interface If a transfer 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. 3. VERSION register reads 0x120 The VERSION register reads 0x120 instead of 0x122. Fix/Workaround None. 10.4.3 FLASHCDW 1. Fuse Programming Programming fuses does not work. Fix/Workaround Do not program fuses. All fuses will be erased during chip erase command. 2. Chip Erase When performing a chip erase, the device may report that it is protected (IR=0x11) and that the erase failed, even if it was successful. Fix/Workaround Perform a reset before any further read and programming. 3. Wait 500 ns before reading from the flash after switching read mode After switching between normal read mode and high-speed read mode, the application must wait at least 500ns before attempting any access to the flash. 92 32099IS–01/2012 AT32UC3L016/32/64 Fix/Workaround Solution 1: Make sure that the appropriate instructions are executed from RAM, and that a waiting-loop is executed from RAM waiting 500ns or more before executing from flash. Solution 2. Execute from flash with a clock with period longer than 500ns. This guarantees that no new read access is attempted before the flash has had time to settle in the new read mode. 4. Flash self programming may fail in one wait state mode Writes in flash and user pages may fail if executing code is located in address space mapped to flash, and the flash controller is configured in one wait state mode (the Flash Wait State bit in the Flash Control Register (FCR.FWS) is one). Fix/Workaround Solution 1: Configure the flash controller in zero wait state mode (FCR.FWS=0). Solution 2: Configure the HMATRIX master 1 (CPU Instruction) to use the unlimited burst length transfer mode (MCFG1.ULBT=0), and the HMATRIX slave 0 (FLASHCDW) to use the maximum slot cycle limit (SCFG0.SLOT_CYCLE=255). 5. VERSION register reads 0x100 The VERSION register reads 0x100 instead of 0x102. Fix/Workaround None. 10.4.4 SAU 1. The SR.IDLE bit reads as zero The IDLE bit in the Status Register (SR.IDLE) reads as zero. Fix/Workaround None. 2. Open Mode is not functional The Open Mode is not functional. Fix/Workaround None. 3. VERSION register reads 0x100 The VERSION register reads 0x100 instead of 0x110. Fix/Workaround None. 10.4.5 HMATRIX 1. In the PRAS and PRBS registers, the MxPR fields are only two bits In the PRAS and PRBS registers, the 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. 10.4.6 Power Manager 1. CONFIG register reads 0x4F The CONFIG register reads 0x4F instead of 0x43. Fix/Workaround None. 93 32099IS–01/2012 AT32UC3L016/32/64 2. It is not possible to mask the request clock requests It is not possible to mask the request clock requests using PPCR. Fix/Workaround None. 3. Static mode cannot be entered if the WDT is using OSC32 If the WDT is using OSC32 as clock source and the user tries to enter Static mode, the Deepstop mode will be entered instead. Fix/Workaround None. 4. Clock Failure Detector (CFD) does not work Clock Failure Detector (CFD) does not work. Fix/Workaround None. 5. WCAUSE register should not be used The WCAUSE register should not be used. Fix/Workaround None. 6. PB writes via debugger in sleep modes are blocked during sleepwalking During sleepwalking, PB writes performed by a debugger will be discarded by all PB modules except the module that is requesting the clock. Fix/Workaround None. 7. Clock sources will not be stopped in Static mode if the difference between CPU and PBx division factor is larger than 4 If the division factor between the CPU/HSB and PBx frequencies is more than 4 when entering a sleep mode where the system RC oscillator (RCSYS) is turned off, the 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 RCSYS is stopped, make sure the division factor between CPU/HSB and PBx frequencies is less than or equal to 4. 8. Disabling POR33 may generate spurious resets Depending on operating conditions, POR33 may generate a spurious reset in one of the following cases: - When POR33 is disabled from the user interface - When SM33 supply monitor is enabled - When entering Shutdown mode while debugging the chip using JTAG or aWire interface In the listed cases, writing a one to the bit VREGCR.POR33MASK in the System Control Interface (SCIF) to mask the POR33 reset will be ineffective Fix/Workaround - Do not disable POR33 using the user interface - Do not use the SM33 supply monitor - Do not enter Shutdown mode if a debugger is connected to the chip 9. Instability when exiting sleep walking If all the following operating conditions are true, exiting sleep walking might lead to instability: 94 32099IS–01/2012 AT32UC3L016/32/64 - The OSC0 is enabled in external clock mode (OSCCTRL0.OSCEN == 1 and OSCCTRL0.MODE == 0) - A sleep mode where the OSC0 is automatically disabled is entered - The device enters sleep walking Fix/Workaround Do not run OSC0 in external clock mode if sleepwalking is expected to be used. 10. Sleepwalking in idle and frozen sleep mode will mask all other PB clocks If the CPU is in idle or frozen sleep mode and a module is in a state that triggers sleep walking, all PB clocks will be masked except the PB clock to the sleepwalking module. Fix/Workaround Mask all clock requests in the PM.PPCR register before going into idle or frozen mode. 11. VERSION register reads 0x400 The VERSION register reads 0x400 instead of 0x411. Fix/Workaround None. 10.4.7 SCIF 1. The DFLL should be slowed down before disabling it The frequency of the DFLL should be set to minimum before disabling it. Fix/Workaround Before disabling the DFLL the value of the COARSE register should be zero. 2. Writing to ICR masks new interrupts received in the same clock cycle Writing to ICR masks any new SCIF interrupt received in the same clock cycle, regardless of write value. Fix/Workaround For every interrupt except BODDET, SM33DET, and VREGOK the PCLKSR register can be read to detect new interrupts. BODDET, SM33DET and VREGOK interrupts will not be generated if they occur whilst writing to the ICR register. 3. FINE value for DFLL is not correct when dithering is disabled In open loop mode, the FINE value used by the DFLL DAC is offset by two compared to the value written to the DFLL0CONF.FINE field. The value used by the DFLL DAC is DFLL0CONF.FINE-0x002. If DFLL0CONF.FINE is written to 0x000, 0x001 or 0x002 the value used by the DFLL DAC will be 0x1FE, 0x1FF, or 0x000 respectively. Fix/Workaround Write the desired value added by two to the DFLL0CONF.FINE field. 4. BODVERSION register reads 0x100 The BODVERSION register reads 0x100 instead of 0x101 Fix/Workaround None. 5. VREGCR.DEEPMODEDISABLE bit is not readable VREGCR.DEEPMODEDISABLE bit is not readable. Fix/Workaround None. 6. DFLL step size should be seven or lower when below 30MHz If max step size is above seven, the DFLL might not lock at the correct frequency if the target frequency is below 30MHz. 95 32099IS–01/2012 AT32UC3L016/32/64 Fix/Workaround If the target frequency is below 30MHz, use a max step size (DFLL0MAXSTEP.MAXSTEP) of seven or lower. 7. Generic clock sources are kept running in sleep modes If a clock is used as a source for a generic clock when going to a sleep mode where clock sources are stopped, the source of the generic clock will be kept running. Please refer to Power Manager chapter for details about sleep modes. Fix/Workaround Disable generic clocks before going to sleep modes where clock sources are stopped to save power. 8. DFLL clock is unstable with a fast reference clock The DFLL clock can be unstable when a fast clock is used as a reference clock in closed loop mode. Fix/Workaround Use the 32KHz crystal oscillator clock, or a clock with a similar frequency, as DFLLIF reference clock. 9. DFLLIF indicates coarse lock too early The DFLLIF might indicate coarse lock too early, the DFLL will lose coarse lock and regain it later. Fix/Workaround Use max step size (DFLL0MAXSTEP.MAXSTEP) of 4 or higher. 10. DFLLIF dithering does not work The DFLLIF dithering does not work. Fix/Workaround None. 11. DFLLIF might lose fine lock when dithering is disabled When dithering is disabled and fine lock has been acquired, the DFLL might lose the fine lock resulting in up to 20% over-/undershoot. Fix/Workaround Solution 1: When the DFLL is used as main clock source, the target frequency of the DFLL should be 20% below the maximum operating frequency of the CPU. Don’t use the DFLL as clock source for frequency sensitive applications. Solution 2: Do not use the DFLL in closed loop mode. 12. GCLK5 is non-functional GCLK5 is non-functional. Fix/Workaround None. 13. BRIFA is non-functional BRIFA is non-functional. Fix/Workaround None. 14. SCIF VERSION register reads 0x100 SCIFVERSION register reads 0x100 instead of 0x102. Fix/Workaround None. 96 32099IS–01/2012 AT32UC3L016/32/64 15. BODVERSION register reads 0x100 BODVERSION register reads 0x100 instead of 0x101. Fix/Workaround None. 16. DFLLVERSION register reads 0x200 DFLLVERSION register reads 0x200 instead of 0x201. Fix/Workaround None. 17. RCCRVERSION register reads 0x100 RCCRVERSION register reads 0x100 instead of 0x101. Fix/Workaround None. 18. OSC32VERSION register reads 0x100 OSC32VERSION register reads 0x100 instead of 0x101. Fix/Workaround None. 19. VREGVERSION register reads 0x100 VREGVERSION register reads 0x100 instead of 0x101. Fix/Workaround None. 20. RC120MVERSION register reads 0x100 RC120MVERSION register reads 0x100 instead of 0x101. Fix/Workaround None. 10.4.8 AST 1. AST wake signal is released one AST clock cycle after the BUSY bit is cleared After writing to the Status Clear Register (SCR) the wake signal is released one AST clock cycle after the BUSY bit in the Status Register (SR.BUSY) is cleared. If entering sleep mode directly after the BUSY bit is cleared the part will wake up immediately. Fix/Workaround Read the Wake Enable Register (WER) and write this value back to the same register. Wait for BUSY to clear before entering sleep mode. 10.4.9 WDT 1. Clearing the WDT in window mode In window mode, if the WDT is cleared 2TBAN CLK_WDT cycles after entering the window, the counter will be cleared, but will not exit the window. If this occurs, the SR.WINDOW bit will not be cleared after clearing the WDT. Fix/Workaround Check SR.WINDOW immediately after clearing the WDT. If set then clear the WDT once more. 2. Clearing the Watchdog Timer (WDT) counter in second half of timeout period will issue a Watchdog reset If the WDT counter is cleared in the second half of the timeout period, the WDT will immediately issue a Watchdog reset. 97 32099IS–01/2012 AT32UC3L016/32/64 Fix/Workaround Use twice as long timeout period as needed and clear the WDT counter within the first half of the timeout period. If the WDT counter is cleared after the first half of the timeout period, you will get a Watchdog reset immediately. If the WDT counter is not cleared at all, the time before the reset will be twice as long as needed. 3. VERSION register reads 0x400 The VERSION register reads 0x400 instead of 0x402. Fix/Workaround None. 4. WDT Control Register does not have synchronization feedback When writing to the Timeout Prescale Select (PSEL), Time Ban Prescale Select (TBAN), Enable (EN), or WDT Mode (MODE) fieldss of the WDT Control Register (CTRL), a synchronizer is started to propagate the values to the WDT clcok domain. This synchronization takes a finite amount of time, but only the status of the synchronization of the EN bit is reflected back to the user. Writing to the synchronized fields during synchronization can lead to undefined behavior. Fix/Workaround -When writing to the affected fields, the user must ensure a wait corresponding to 2 clock cycles of both the WDT peripheral bus clock and the selected WDT clock source. -When doing writes that changes the EN bit, the EN bit can be read back until it reflects the written value. 10.4.10 FREQM 1. Measured clock (CLK_MSR) sources 15-17 are shifted CLKSEL = 14 selects the RC120M AW clock, CLKSEL = 15 selects the RC120M clock, and CLKSEL = 16 selects the RC32K clock as source for the measured clock (CLK_MSR). Fix/Workaround None. 2. GCLK5 can not be used as source for the CLK_MSR The frequency for GCLK5 can not be measured by the FREQM. Fix/Workaround None. 10.4.11 GPIO 1. GPIO interrupt can not be cleared when interrupts are disabled The GPIO interrupt can not be cleared unless the interrupt is enabled for the pin. Fix/Workaround Enable interrupt for the corresponding pin, then clear the interrupt. 2. VERSION register reads 0x210 The VERSION register reads 0x210 instead of 0x211. Fix/Workaround None. 10.4.12 USART 1. The RTS output does not function correctly in hardware handshaking mode 98 32099IS–01/2012 AT32UC3L016/32/64 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. 10.4.13 SPI 1. SPI data transfer hangs with CSR0.CSAAT==1 and MR.MODFDIS==0 When CSR0.CSAAT==1 and mode fault detection is enabled (MR.MODFDIS==0), the SPI module will not start a data transfer. Fix/Workaround Disable mode fault detection by writing a one to MR.MODFDIS. 2. Disabling SPI has no effect on the SR.TDRE bit Disabling SPI has no effect on the SR.TDRE bit whereas the write data command is filtered when SPI is disabled. Writing to TDR when SPI is disabled will not clear SR.TDRE. If SPI is disabled during a PDCA transfer, the PDCA will continue to write data to TDR until its buffer is empty, and this data will be lost. Fix/Workaround Disable the PDCA, add two NOPs, and disable the SPI. To continue the transfer, enable the SPI and PDCA. 3. 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 by writing a one to the Software Reset bit in the Control Register (CR.SWRST). 4. SPI bad serial clock generation on 2nd chip_select when SCBR=1, CPOL=1, and NCPHA=0 When multiple chip selects (CS) are in use, if one of the baudrates equal 1 while one (CSRn.SCBR=1) of the others do not equal 1, and CSRn.CPOL=1 and CSRn.NCPHA=0, then an additional pulse will be generated on SCK. Fix/Workaround When multiple CS are in use, if one of the baudrates equals 1, the others must also equal 1 if CSRn.CPOL=1 and CSRn.NCPHA=0. 5. SPI mode fault detection enable causes incorrect behavior When mode fault detection is enabled (MR.MODFDIS==0), the SPI module may not operate properly. Fix/Workaround Always disable mode fault detection before using the SPI by writing a one to MR.MODFDIS. 10.4.14 TWI 1. TWI pins are not SMBus compliant 99 32099IS–01/2012 AT32UC3L016/32/64 The TWI pins draw current when they are supplied with 3.3V and the part is left unpowered. Fix/Workaround None. 2. PA21, PB04, and PB05 are not 5V tolerant Pins PA21, PB04, and PB05 are only 3.3V tolerant. Fix/Workaround None. 3. PB04 SMBALERT function should not be used The SMBALERT function from TWIMS0 should not be selected on pin PB04. Fix/Workaround None. 4. TWIM STOP bit in IMR always reads as zero The STOP bit in IMR always reads as zero. Fix/Workaround None. 5. Disabled TWIM drives TWD and TWCK low When the TWIM is disabled, it drives the TWD and TWCK signals with logic level zero. This can lead to communication problems with other devices on the TWI bus. Fix/Workaround Enable the TWIM first and then enable the TWD and TWCK peripheral pins in the GPIO controller. If it is necessary to disable the TWIM, first disable the TWD and TWCK peripheral pins in the GPIO controller and then disable the TWIM. 6. TWIM SR.IDLE goes high immediately when NAK is received When a NAK is received and there is a non-zero number of bytes to be transmitted, SR.IDLE goes high immediately and does not wait for the STOP condition to be sent. This does not cause any problem just by itself, but can cause a problem if software waits for SR.IDLE to go high and then immediately disables the TWIM by writing a one to CR.MDIS. Disabling the TWIM causes the TWCK and TWD pins to go high immediately, so the STOP condition will not be transmitted correctly. Fix/Workaround If possible, do not disable the TWIM. If it is absolutely necessary to disable the TWIM, there must be a software delay of at least two TWCK periods between the detection of SR.IDLE==1 and the disabling of the TWIM. 7. TWIM TWALM polarity is wrong The TWALM signal in the TWIM is active high instead of active low. Fix/Workaround Use an external inverter to invert the signal going into the TWIM. When using both TWIM and TWIS on the same pins, the TWALM cannot be used. 8. TWIS CR.STREN does not work in deep sleep modes When the device is in Stop, DeepStop, or Static mode, address reception will not wake device if both CR.SOAM and CR.STREN are one. Fix/Workaround Do not write both CR.STREN and CR.SOAM to one if the device needs to wake from deep sleep modes. 9. TWI0.TWCK on PB05 is non-functional TWI0.TWCK on PB05 is non-functional. 100 32099IS–01/2012 AT32UC3L016/32/64 Fix/Workaround Use TWI0.TWCK on other pins. 10. TWIM Version Register reads zero TWIM Version Register (VR) reads zero instead of 0x101 Fix/Workaround None. 11. TWIS Version Register reads zero TWIS Version Register (VR) reads zero instead of 0x112 Fix/Workaround None. 12. SMBALERT bit may be set after reset The SMBus Alert (SMBALERT) bit in the Status Register (SR) might be erroneously set after system reset. Fix/Workaround After system reset, clear the SR.SMBALERT bit before commencing any TWI transfer. 13. Clearing the NAK bit before the BTF bit is set locks up the TWI bus When the TWIS is in transmit mode, clearing the NAK Received (NAK) bit of the Status Register (SR) before the end of the Acknowledge/Not Acknowledge cycle will cause the TWIS to attempt to continue transmitting data, thus locking up the bus. Fix/Workaround Clear SR.NAK only after the Byte Transfer Finished (BTF) bit of the same register has been set. 14. TWIS stretch on Address match error When the TWIS stretches TWCK due to a slave address match, it also holds TWD low for the same duration if it is to be receiving data. When TWIS releases TWCK, it releases TWD at the same time. This can cause a TWI timing violation. Fix/Workaround None. 10.4.15 PWMA 1. PARAMETER register reads 0x2424 The PARAMETER register reads 0x2424 instead of 0x24. Fix/Workaround None. 2. Writing to the duty cycle registers when the timebase counter overflows can give an undefined result The duty cycle registers will be corrupted if written when the timebase counter overflows. If the duty cycle registers are written exactly when the timebase counter overflows at TOP, the duty cycle registers may become corrupted. Fix/Workaround Write to the duty cycle registers only directly after the Timebase Overflow bit in the status register is set. 3. Open Drain mode does not work The open drain mode does not work. Fix/Workaround None. 101 32099IS–01/2012 AT32UC3L016/32/64 4. BUSY bit is never cleared after writes to the Control Register (CR) When writing a non-zero value to CR.TOP, CR.SPREAD, or CR.TCLR when the PWMA is disabled (CR.EN == 0), the BUSY bit in the Status Register (SR.BUSY) will be set, but never cleared. Fix/Workaround When writing a non-zero value to CR.TOP, CR.SPREAD, or CR.TCLR, make sure the PWMA is enabled, or simultaneously enable the PWMA by writing a one to CR.EN. 5. Incoming peripheral events are discarded during duty cycle register update Incoming peripheral events to all applied channels will be discarded if a duty cycle update is received from the user interface in the same PWMA clock period. Fix/Workaround Ensure that duty cycle writes from the user interface are not performed in a PWMA period when an incoming peripheral event is expected. 6. VERSION register reads 0x100 The VERSION register reads 0x100 instead of 0x101. Fix/Workaround None. 10.4.16 TC 1. When the main clock is RCSYS, TIMER_CLOCK5 is equal to CLK_PBA When the main clock is generated from RCSYS, TIMER_CLOCK5 is equal to CLK_PBA and not CLK_PBA/128. Fix/Workaround None. 10.4.17 ADCIFB 1. Pendetect in sleep modes without CLK_ADCIFB will not wake the system The pendetect will not wake the system from a sleep mode if the clock for the ADCIFB (CLK_ADCIFB) is turned off. Fix/Workaround Use a sleep mode where CLK_ADCIFB is not turned off to wake the part using pendetect. 2. 8-bit mode is not working Do not use the ADCIFB 8-bit mode. Fix/Workaround Use the 10-bit mode and shift right by two bits. 3. Using STARTUPTIME larger than 0x1F will freeze the ADC Writing a value larger than 0x1F to the Startup Time field in the ADC Configuration Register (ACR.STARTUP) will freeze the ADC, and the Busy Status bit in the Status Register (SR.BUSY) will never be cleared. Fix/Workaround Do not write values larger than 0x1F to ACR.STARTUP. 4. ADC channels six to eight are non-functional ADC channels six to eight are non-functional. Fix/Workaround None. 5. VERSION register reads 0x100 102 32099IS–01/2012 AT32UC3L016/32/64 The VERSION register reads 0x100 instead of 0x101. Fix/Workaround None. 10.4.18 ACIFB 1. Generic clock sources in sleep modes. The ACIFB should not use RC32K or CLK_1K as generic clock source if the chip uses sleep modes. Fix/Workaround None. 2. Negative offset The static offset of the analog comparator is approximately -50mV Fix/Workaround None. 3. CONFW.WEVSRC and CONFW.WEVEN are not correctly described in the user interface CONFW.WEVSRC is only two bits instead of three bits wide. Only values 0, 1, and 2 can be written to this register. CONFW.WEVEN is in bit position 10 instead of 11. Fix/Workaround Only write values 0, 1, and 2 to CONFW.WEVSRC. When reading CONFW.WEVSRC, disregard the third bit. Read/write bit 10 to access CONFW.WEVEN. 4. VERSION register reads 0x200 The VERSION register reads 0x200 instead of 0x212. Fix/Workaround None. 10.4.19 CAT 1. Switch off discharge current when reaching 0V The discharge current will switch off when reaching MGCFG1.MAX, not when reaching 0V. Fix/Workaround None. 2. CAT external capacitors are not clamped to ground when CAT is idle The CAT module does not clamp the external capacitors to ground when it is idle. The capacitors are left floating, so they could accumulate small amounts of charge. Fix/Workaround None. 3. DISHIFT field is stuck at zero The DISHIFT field in the MGCFG1, TGACFG1, TGBCFG1, and ATCFG1 registers is stuck at zero and cannot be written to a different value. Capacitor discharge time will only be determined by the DILEN field. Fix/Workaround None. 4. MGCFG2.CONSEN field is stuck at zero The CONSEN field in the MGCFG2 register is stuck at zero and cannot be written to a different value. The CAT consensus filter does not function properly, so termination of QMatrix data acquisition is controlled only by the MAX field in MGCFG1. 103 32099IS–01/2012 AT32UC3L016/32/64 Fix/Workaround None. 5. MGCFG2.ACCTRL bit is stuck at zero The ACCTRL bit in the MGCFG2 register is stuck at zero and cannot be written to one. The analog comparators will be constantly enabled. Fix/Workaround None. 6. CAT asynchronous wake will be delayed by one AST event period If the CAT detects a condition the should asynchronously wake the device in static mode, the asynchronous wake will not occur until the next AST event. For example, if the AST is generating events to the CAT every 50ms, and the CAT detects a touch at t=9200ms, the asynchronous wake will occur at t=9250ms. Fix/Workaround None. 7. VERSION register reads 0x100 The VERSION register reads 0x100 instead of 0x200. Fix/Workaround None. 8. CAT module does not terminate QTouch burst on detect The CAT module does not terminate a QTouch burst when the detection voltage is reached on the sense capacitor. This can cause the sense capacitor to be charged more than necessary. Depending on the dielectric absorption characteristics of the capacitor, this can lead to unstable measurements. Fix/Workaround Use the minimum possible value for the MAX field in the ATCFG1, TG0CFG1, and TG1CFG1 registers. 9. Autonomous CAT acquisition must be longer than AST source clock period When using the AST to trigger CAT autonomous touch acquisition in sleep modes where the CAT bus clock is turned off, the CAT will start several acquisitions if the period of the AST source clock is larger than one CAT acquisition. One AST clock period after the AST trigger, the CAT clock will automatically stop and the CAT acquisition can be stopped prematurely, ruining the result. Fix/Workaround Always ensure that the ATCFG1.max field is set so that the duration of the autonomous touch acquisition is greater than one clock period of the AST source clock. 10.4.20 GLOC 1. GLOC is non-functional Glue Logic Controller (GLOC) is non-functional. Fix/Workaround None. 10.4.21 aWire 1. SAB multiaccess reads are not working Reading more than one word, halfword, or byte in one command is not working correctly. Fix/Workaround Split the access into several single word, halfword, or byte accesses. 104 32099IS–01/2012 AT32UC3L016/32/64 2. If a reset happens during the last SAB write, the aWire will stall If a reset happens during the last word, halfword or byte write the aWire will wait forever for an acknowledge from the SAB. Fix/Workaround Reset the aWire by keeping the RESET_N line low for 100ms. 3. aWire enable does not work in Static mode aWire enable does not work in Static mode. Fix/Workaround None. 4. aWire CPU clock speed robustness The aWire memory speed request command counter warps at clock speeds below approximately 5kHz. Fix/Workaround None. 5. The aWire debug interface is reset after leaving Shutdown mode If the aWire debug mode is used as debug interface and the program enters Shutdown mode, the aWire interface will be reset when the part receives a wakeup either from the WAKE_N pin or the AST. Fix/Workaround None. 6. aWire PB mapping and PB clock mask number The aWire PB has a different PB address and PB clock mask number. Fix/Workaround Use aWire PB address 0xFFFF6C00 and PB clock (PBAMASK) 24. 7. VERSION register reads 0x200 The VERSION register reads 0x200 instead of 0x210. Fix/Workaround None. 10.4.22 Chip 1. WAKE_N pin can only wake up the chip from Shutdown mode It is not possible to wake up the chip from any other sleep mode than Shutdown using the WAKE_N pin. If the WAKE_N pin is asserted during a sleep mode other than Shutdown, nothing will happen. Fix/Workaround Use an EIC pin to wake up from sleep modes higher than Shutdown. 2. Power consumption in static mode is too high Power consumption in static mode is too high when PA21 is high. Fix/Workaround Ensure PA21 is low. 3. Shutdown mode is not functional Do not enter Shutdown mode. Fix/Workaround None. 4. VDDIN current consumption increase above 1.8V 105 32099IS–01/2012 AT32UC3L016/32/64 When VDDIN increases above 1.8V, current on VDDIN increases with up to 40uA. Fix/Workaround None. 5. Increased Power Consumption in VDDIO in sleep modes If 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 Solution 1: Disable OSC0 by writing a zero to the Oscillator Enable bit in the System Control Interface (SCIF) Oscillator Control Register (SCIF.OSC0CTRL.OSCEN) before going to a sleep mode where OSC0 is disabled. Solution 2: Pull down or up XIN0 and XOUT0 with 1MOhm resistor. 10.4.23 I/O Pins 1. PB10 is not 3.3V tolerant. PB10 should be grounded on the PCB and left unused. Fix/Workaround None. 2. Analog multiplexing consumes extra power Current consumption on VDDIO increases when the voltage on analog inputs is close to VDDIO/2. Fix/Workaround None. 3. PA02, PB01, PB04, PB05, and RESET_N have half of the pull-up strength Pins PA02, PB01, PB04, PB05, and RESET_N have half of the specified pull-up strength. Fix/Workaround None. 4. JTAG is enabled at power up The JTAG function on pins PA00, PA01, PA02, and PA03, are enabled after startup. Normal I/O module functionality is not possible on these pins. Fix/Workaround Add a 10kOhm pullup on the reset line. 5. MCKO and MDO[3] are swapped in the AUX1 mapping When using the OCD AUX1 mapping of trace signals MDO[3] is located on pin PB05 and MCKO is located on PB01. Fix/Workaround Swap pins PB01 and PB05 if using OCD AUX1. 106 32099IS–01/2012 AT32UC3L016/32/64 11. 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. 11.1 11.2 Rev. I - 01/2012 1. Overview - Block diagram: CAT SMP corrected from I/O to output. SPI NPCS corrected from output to I/O. 2. Package and Pinout: PRND signal removed from Signal Descriptions List table and GPIO Controller Function Multiplexing table 3. ADCIFB: PRND signal removed from block diagram. 4. Electrical Characteristics: Added pin input capacitance CIN for TLLGA package for “all normal I/O pins except PA05 , PA07, PA17, PA20, PA21, PB04, PB05”. 5. Errata: Added more errata for TWI and CAT modules. Rev. H - 12/2011 1. Mechanical Characteristics: Updated the note related to the QFN48 Package Drawing.Updated thermal reistance data for TLLGA48 package. Updated package drawings for TQFP48 and QFN48 packages. Soldering profile updated (Time maintained above 217°C.) 2. Memories: Local bus address map corrected: The address offset for port 1 registers is 0x100, not 0x200. 3. SCIF DFLL: Removed “not” from “DFLLnSTEP.CSTEP and DFLLnSTEP.FSTEP should not be lower than 50% of the maximum value of DFLLnCONF.COARSE and DFLLnCONF.FINE”. 4. SCIF VREG POR descriptions updated. POR33 bits added in VREGCR. 5. SCIF VREG: Removed reference to flash fuses for CALIB field, user is recommended not writing to SELVDD field. Removed references to Electrical Characteristics. 6. SCIF: Fuses text removed from some submodules (SM33, VREG). 7. SCIF VREG: Flash recalibration is always done at POR. 8. SCIF SM33: Enabling SM33 will disable the POR33 detector. 9, Erratum regarding OSC32 disabling is not valid for RevB. 10. Flash Controller: Serial number address updated. 11. Block diagram and ADCIFB: Removed PRND signal from block diagram and ADCIFB block diagram. 12. USART: Added CTSIC bit description. 13. Power Manager: Updated Clock division and Clock ready flag sections. 14. ADCIFB: Added DMA section in Product Dependencies. 107 32099IS–01/2012 AT32UC3L016/32/64 11.3 11.4 11.5 15. Electrcal Characteristics: Updated SPI timing data. 16. Electrical Characteristics, I/O Pin Characteristics: Added Input capacitance for TLLGA48 package. 17. Errata: Removed erratum regarding SPI RDR.PCS field, as the PCS field has been removed (refer to Section 11.8 on page 109). Rev. G - 06/2011 1. FLASHCDW: FSR register is a Read-only register. Added info about QPRUP. 2. PM: Clarified POR33 masking requirements before shutdown. Added more info about wakeup sources. Added AWEN description. PPCR register reset value corrected. 3. SAU: SR.IDLE definition and reset value corrected. 4. DFLL: Open- and closed-loop operation clarified. 5. OSC32: Added note about OSC32RDY bit not always cleared when disabling OSC32. 6. USART: Major cleanup. 1. Features: Removed superfluous R mark. 2. Package and Pinout, GPIO function multiplexing:TWIMS0-TWCK on PA20 removed. ADCIFBAD[3] on PA17 removed, number of ADC channels are 8, not 9. These were removed from rev. C, but reappeared in rev. E. 1. Overview: Added missing signals in block diagram. 2. Package and pinout: Added note about TWI, SMBUS and 5V tolerant pads in peripheral multiplexing. Added CAT DIS signal to signal descriptions. Removed TBD on ADVREFP minimum voltage. 3. Memories: Added SAU slave address to physical memory map. 4. Supply and startup considerations: VDDIN is using GND as ground pin. Clarified references to PORs in startup considerations. 5. FLASHCDW: Added serial number location to module configuration section. 6. PM: Added more info about the WAKE_N pin. Added info about CLK_PM, Updated the selection main clock source section. 7. SCIF: Major chapter update. 8. AST: Updated digital tuner formula and conditions. 9. GPIO: Updated GPER reset value and added more registers with non-zero reset value. 10. CAT: Added info about VDIVEN and discharge current formula. Rev. F- 11/2010 Rev. E- 10/2010 108 32099IS–01/2012 AT32UC3L016/32/64 11.6 11. ADCIFB: Fixed Sample and Hold time formula. 12. GLOC: Added info about pullup control and renamed LUTCR register to CR. 13. TC: Added features and version register. 14. SAU: Added OPEN bit to config register. Added description of unlock fields. 15. TWIS: SCR is Write-only. Improved explanation of slave transmitter mode. Updated data transfer diagrams. 16. Electrical Characteristics: Added more values. Added notes on simulated and characterized values. Added pin capacitance, rise, and fall times. Added timing characteristics. Removed all TBDs. Added ADC analog input characteristics. Symbol cleanup. 17. Errata: Updated errata list. Rev. D - 06/2010 1. 11.7 11.8 Ordering Information: Ordering code for TQFP ES changed from AT32UC3L064-AUES to AT32UC3L064-AUTES. TLLGA48 Tray option added. Rev. C - 06/2010 1. Features and Description: Added QTouch library support. 2. USART: Description of unimplemented features removed. 3. Electrical Characteristics: Power Consumption numbers updated. Flash timing numbers added. Rev. B - 05/2010 1. Package and Pinout: Added pinout figure for TLLGA48 package. 2. Package and Pinout, GPIO function multiplexing:TWIMS0-TWCK on PA20 removed. ADCIFBAD[3] on PA17 removed, number of ADC channels are 8, not 9. 3. I/O Lines Considerations: Added: Following pins have high-drive capability: PA02, PA06, PA08, PA09, and PB01. Some TWI0 pins are SMBUS compliant (PA21, PB04, PB05). 4. HMATRIX Masters: PDCA is master 4, not master 3. SAU is master 3, not master 4. 5. SAU: IDLE bit added in the Status Register. 6. PDCA: Number of PDCA performance monitors is device dependent. 7. Peripheral Event System: Chapter updated. 8. PM: Bits in RCAUSE registers removed and renamed (JTAGHARD and AWIREHARD renamed to JTAG and AWIRE respectively, JTAG and AWIRE removed. BOD33 bit removed). 9. PM: RCAUSE.BOD33 bit removed. SM33 reset will be detected as a POR reset. 10. PM: WDT can be used as wake-up source if WDT is clocked from 32KHz oscillator. 109 32099IS–01/2012 AT32UC3L016/32/64 11.9 11. PM: Entering Shutdown mode description updated. 12. SCIF: DFLL output frequency is 40-150MHz, not 20-150MHz or 30-150MHz. 13. SCIF: Temperature sensor is connected to ADC channel 9, not 7. 14. SCIF: Updated the oscillator connection figure for OSC0 15. GPIO: Removed unimplemented features (pull-down, buskeeper, drive strength, slew rate, Schmidt trigger, open drain). 16. SPI: RDR.PCS field removed (RDR[19:16]). 17. TWIS: Figures updated. 18. ADCIFB: The sample and hold time and the startup time formulas have been corrected (ADC Configuration Register). 19. ADCIFB: Updated ADC signal names. 20. ACIFB: CONFW.WEVSRC is bit 8-10, CONFW.EWEVEN is bit 11. CONF.EVENP and CONF.EVENN bits are swapped. 21. CAT: Matrix size is 16 by 8, not 18 by 8. 22. Electrical Characteristics: General update. 23. Mechanical Characteristics: Added numbers for package drawings. 24. Mechanical Characteristics: In the TQFP-48 package drawing the Lead Coplanarity is 0.102mm, not 0.080mm. 25. Ordering Information: Ordering code for TLLGA-48 package updated. Rev. A – 06/2009 1. Initial revision. 110 32099IS–01/2012 AT32UC3L016/32/64 Table of Contents Features ..................................................................................................... 1 1 Description ............................................................................................... 3 2 Overview ................................................................................................... 5 3 4 5 6 7 2.1 Block Diagram ...................................................................................................5 2.2 Configuration Summary .....................................................................................6 Package and Pinout ................................................................................. 7 3.1 Package .............................................................................................................7 3.2 Peripheral Multiplexing on I/O lines ...................................................................9 3.3 Signal Descriptions ..........................................................................................13 3.4 I/O Line Considerations ...................................................................................16 Processor and Architecture .................................................................. 18 4.1 Features ..........................................................................................................18 4.2 AVR32 Architecture .........................................................................................18 4.3 The AVR32UC CPU ........................................................................................19 4.4 Programming Model ........................................................................................23 4.5 Exceptions and Interrupts ................................................................................27 Memories ................................................................................................ 32 5.1 Embedded Memories ......................................................................................32 5.2 Physical Memory Map .....................................................................................32 5.3 Peripheral Address Map ..................................................................................33 5.4 CPU Local Bus Mapping .................................................................................34 Supply and Startup Considerations ..................................................... 36 6.1 Supply Considerations .....................................................................................36 6.2 Startup Considerations ....................................................................................40 Electrical Characteristics ...................................................................... 41 7.1 Absolute Maximum Ratings* ...........................................................................41 7.2 Supply Characteristics .....................................................................................41 7.3 Maximum Clock Frequencies ..........................................................................42 7.4 Power Consumption ........................................................................................42 7.5 I/O Pin Characteristics .....................................................................................47 7.6 Oscillator Characteristics .................................................................................50 7.7 Flash Characteristics .......................................................................................54 i 32099IS–01/2012 AT32UC3L016/32/64 8 9 7.8 Analog Characteristics .....................................................................................55 7.9 Timing Characteristics .....................................................................................63 Mechanical Characteristics ................................................................... 73 8.1 Thermal Considerations ..................................................................................73 8.2 Package Drawings ...........................................................................................74 8.3 Soldering Profile ..............................................................................................77 Ordering Information ............................................................................. 78 10 Errata ....................................................................................................... 79 10.1 Rev. E ..............................................................................................................79 10.2 Rev. D ..............................................................................................................85 10.3 Rev. C ..............................................................................................................91 10.4 Rev. B ..............................................................................................................91 11 Datasheet Revision History ................................................................ 107 11.1 Rev. I - 01/2012 .............................................................................................107 11.2 Rev. H - 12/2011 ...........................................................................................107 11.3 Rev. G - 06/2011 ...........................................................................................108 11.4 Rev. F- 11/2010 .............................................................................................108 11.5 Rev. E- 10/2010 .............................................................................................108 11.6 Rev. D - 06/2010 ...........................................................................................109 11.7 Rev. C - 06/2010 ...........................................................................................109 11.8 Rev. B - 05/2010 ............................................................................................109 11.9 Rev. A – 06/2009 ...........................................................................................110 Table of Contents....................................................................................... i ii 32099IS–01/2012 Atmel Corporation 2325 Orchard Parkway San Jose, CA 95131 USA Tel: (+1)(408) 441-0311 Fax: (+1)(408) 487-2600 www.atmel.com Atmel Asia Limited 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 Munich GmbH Business Campus Parkring 4 D-85748 Garching b. Munich GERMANY Tel: (+49) 89-31970-0 Fax: (+49) 89-3194621 Atmel Japan 16F, Shin Osaki Kangyo Bldg. 1-6-4 Osaka Shinagawa-ku Tokyo 104-0032 JAPAN Tel: (+81) 3-6417-0300 Fax: (+81) 3-6417-0370 © 2012 Atmel Corporation. All rights reserved. Atmel ®, logo and combinations thereof, AVR ®, picoPower ®, QTouch ®, AKS ® and others are registered trademarks or trademarks of Atmel Corporation or its subsidiaries. Other terms and product names may be trademarks of others. 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 THE ATMEL TERMS AND CONDITIONS OF SALES LOCATED ON THE ATMEL WEBSITE, 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 AND 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 products are not intended, authorized, or warranted for use as components in applications intended to support or sustain life. 32099IS–01/2012