ST92163 8/16-BIT FULL SPEED USB MCU FOR COMPOSITE DEVICES WITH 16 ENDPOINTS, 20K ROM, 2K RAM, I2C, SCI, & MFT PRELIMINARY DATA ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ Internal Memories: 20 Kbytes ROM/EPROM/ OTP, 2 Kbytes RAM Register oriented 8/16 bit core 224 general purpose registers available as RAM, accumulators or index pointers Minimum instruction cycle time: 167 ns (@24 MHz CPU frequency) Low power modes: WFI, SLOW, HALT and STOP DMA controller for reduced processor overhead Full speed USB interface with DMA, compliant with USB specifications version 1.1 (in normal voltage mode) USB Embedded Functions with 16 fully configurable endpoints (buffer size programmable), supporting all USB data transfer types (Isochronous included) On-chip USB transceiver and 3.3 voltage regulator Multimaster I2C-bus serial interface up to 400KHz. with DMA capability Serial Communications Interface (SCI) with DMA capability: – Asynchronous mode up to 315 Kb/s – Synchronous mode up to 3 MHz External memory interface (8-bit data/16-bit address) with DMA capability from the USB 16-bit Multi-Function Timer (12 operating modes) with DMA capability 16-bit Timer with 8-bit prescaler and Watchdog 6-channel, 8-bit A/D Converter (ADC) 15 interrupt pins on 8 interrupt channels 14 pins programmable as wake-up or additional external interrupts 42 (DIP56) or 44 (QFP64) fully programmable I/Os with 6 or 8 high sink pads (10 mA @ 1 V) Programmable PLL clock generator (RCCU) using a low frequency external quartz (8 MHz) On-chip RC oscillator for low power operation Note 1: Refer to “Device Summary” on page 6 PSDIP56 TQFP64 ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ Low Voltage Detector Reset on some devices1 Rich instruction set with 14 addressing modes Several operating voltage modes available on some devices1: – Normal Voltage Mode – 8-MHz Low Voltage Mode – 16-MHz Low Voltage Mode 0 - 24 MHz CPU clock operation @ 4.0-5.5 V (all devices) 0 - 8 MHz CPU clock operation @ 3.0-4.0 V (8MHz and 16-MHz Low Voltage devices) 0 - 16 MHz CPU clock operation @ 3.0-4.0 V (16-MHz Low Voltage devices only) Division-by-zero trap generation 0 oC to 70 oC temperature range Low EMI design supporting single sided PCB Complete development tools, including assembler, linker, C-compiler, archiver, source level debugger and hardware emulators, and Real Time Operating System Rev. 1.9 January 2000 This is preliminary information on a new product in development orundergoing evaluation. Details are subject to change without notice. 1/224 1 Table of Contents ST92163 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 GENERAL DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.1.1 Core Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.1.2 Instruction Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.1.3 External MEMORY INTERFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.1.4 OPERATING MODES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.1.5 On-chip Peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.2 PIN DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.3 I/O PORT PINS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.4 MEMORY MAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1.5 ST92163 REGISTER MAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2 DEVICE ARCHITECTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.1 CORE ARCHITECTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.2 MEMORY SPACES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.2.1 Register File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.2.2 Register Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3 SYSTEM REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 2.3.1 Central Interrupt Control Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Flag Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Register Pointing Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Paged Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.5 Mode Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.6 Stack Pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 MEMORY ORGANIZATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 31 32 35 35 36 38 2.5 MEMORY MANAGEMENT UNIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.6 ADDRESS SPACE EXTENSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 2.6.1 2.6.2 2.7 MMU 2.7.1 2.7.2 2.7.3 2.7.4 2.8 MMU Addressing 16-Kbyte Pages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Addressing 64-Kbyte Segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DPR[3:0]: Data Page Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CSR: Code Segment Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ISR: Interrupt Segment Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DMASR: DMA Segment Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . USAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 41 41 41 43 43 43 45 2.8.1 Normal Program Execution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8.2 Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8.3 DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 INTERRUPTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 45 45 46 46 3.2 INTERRUPT VECTORING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3.2.1 Divide by Zero trap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3.2.2 Segment Paging During Interrupt Routines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 3.3 INTERRUPT PRIORITY LEVELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 2/224 Table of Contents 3.4 PRIORITY LEVEL ARBITRATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 3.4.1 Priority level 7 (Lowest) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.2 Maximum depth of nesting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.3 Simultaneous Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.4 Dynamic Priority Level Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 ARBITRATION MODES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 48 48 49 49 3.5.1 Concurrent Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.2 Nested Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 EXTERNAL INTERRUPTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7 MANAGEMENT OF WAKE-UP LINES AND EXTERNAL INTERRUPT LINES . . . . . . . . . 49 52 54 56 3.8 TOP LEVEL INTERRUPT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 3.9 ON-CHIP PERIPHERAL INTERRUPTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 3.10 INTERRUPT RESPONSE TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.11 INTERRUPT REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 3.12 WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (WUIMU) . . . . . . . . . . . . . . . . . . 3.12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.12.2 Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.12.3 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.12.4 Programming Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.12.5 Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 ON-CHIP DIRECT MEMORY ACCESS (DMA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 63 63 64 66 67 70 70 4.2 DMA PRIORITY LEVELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 4.3 DMA TRANSACTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.4 DMA CYCLE TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 4.5 SWAP MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 4.6 DMA REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 5 RESET AND CLOCK CONTROL UNIT (RCCU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5.2 CLOCK CONTROL UNIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5.2.1 Clock Control Unit Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 CLOCK MANAGEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 PLL Clock Multiplier Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 CPU Clock Prescaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3 Peripheral Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.4 Low Power Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.5 Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 CLOCK CONTROL REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 77 78 78 78 79 79 81 5.5 OSCILLATOR CHARACTERISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 5.6 RESET/STOP MANAGER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 5.6.1 Reset Pin Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 5.7 STOP MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 5.8 LOW VOLTAGE DETECTOR (LVD) RESET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 . . . . 88 3/224 Table of Contents 6 EXTERNAL MEMORY INTERFACE (EXTMI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 6.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 6.2 EXTERNAL MEMORY SIGNALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 6.2.1 AS: Address Strobe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.2 DS: Data Strobe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3 DS2: Data Strobe 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.4 RW: Read/Write . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.5 BREQ, BACK: Bus Request, Bus Acknowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.6 PORT 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.7 PORT 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.8 WAIT: External Memory Wait . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 REGISTER DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 90 90 93 93 94 94 94 95 7 I/O PORTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 7.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 7.2 SPECIFIC PORT CONFIGURATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 7.3 PORT CONTROL REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 7.4 INPUT/OUTPUT BIT CONFIGURATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 7.5 ALTERNATE FUNCTION ARCHITECTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 7.5.1 Pin Declared as I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.2 Pin Declared as an Alternate Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.3 Pin Declared as an Alternate Function Output . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6 I/O STATUS AFTER WFI, HALT AND RESET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 ON-CHIP PERIPHERALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 TIMER/WATCHDOG (WDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.2 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.3 Watchdog Timer Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.4 WDT Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.5 Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 MULTIFUNCTION TIMER (MFT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 103 103 103 104 104 104 105 106 108 109 111 8.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.2 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.3 Input Pin Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.4 Output Pin Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.5 Interrupt and DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.6 Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3 USB PERIPHERAL (USB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 113 116 120 122 124 135 8.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.2 Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.3 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.4 Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.5 Register pages summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4 SERIAL COMMUNICATIONS INTERFACE (SCI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.2 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.3 SCI Operating Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 135 135 138 148 150 150 151 152 4/224 Table of Contents 8.4.4 Serial Frame Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.5 Clocks And Serial Transmission Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.6 SCI Initialization Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.7 Input Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.8 Output Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.9 Interrupts and DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.10 Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5 I2C BUS INTERFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 158 158 160 160 161 164 175 8.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.2 Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.3 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.4 I2C State Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.5 Interrupt Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.6 DMA Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.7 Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6 A/D CONVERTER (A/D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 175 176 178 183 184 186 197 8.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6.2 Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6.3 General Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6.4 Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 ELECTRICAL CHARACTERISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 GENERAL INFORMATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.1 EPROM/OTP PROGRAMMING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 197 197 199 201 218 218 10.2 PACKAGE DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 10.3 ORDERING INFORMATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 10.4 TRANSFER OF CUSTOMER CODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 224 5/224 ST92163 - GENERAL DESCRIPTION 1 GENERAL DESCRIPTION 1.1 INTRODUCTION The ST9216x family brings the enhanced ST9 register-based architecture to a new range of highperformance microcontrollers specifically designed for USB (Universal Serial Bus) applications. Their performance derives from the use of a flexible 256-register programming model for ultrafast context switching and real-time event response. The intelligent on-chip peripherals offload the ST9 core from I/O and data management processing tasks allowing critical application tasks to get the maximum use of core resources. The ST9 MCU devices support low power consumption and low voltage operation for power-efficient and low-cost embedded systems. In the ST92163 family, four different types of device are available: Normal Voltage Devices with LVD function They operate in Normal Voltage Mode only (4.05.5V @ 24MHz) and include the Low Voltage Detector (LVD) function. Normal Voltage Devices without LVD function They operate in Normal Voltage Mode only (4.05.5V @ 24MHz) and do not include the Low Voltage Detector (LVD) function. 8-MHz Low Voltage Devices They do not include the Low Voltage Detector (LVD) function and they support two operating voltage modes: – Normal Voltage mode (4.0-5.5V @ 24MHz) with full functionality including USB. – 8-MHz Low Voltage mode (3.0-4.0V @ 8MHz) without the USB interface. 16-MHz Low Voltage Devices They do not include the Low Voltage Detector (LVD) function and they support three operating voltage modes: – Normal Voltage mode (4.0-5.5V @ 24MHz) with full functionality including USB. – 8-MHz Low Voltage mode (3.0-4.0V @ 8MHz) without the USB interface. – 16-MHz Low Voltage mode (3.0-4.0V @ 16MHz) without the USB interface. Figure 1, on page 7 shows the operating range of the ST92163 devices. Device Summary Device Package Program Memory ST92163 1 ST92T163 PSDIP56/ TQFP64 CSDIP56/ CQFP64 PSDIP56/ TQFP64 CSDIP56/ CQFP64 20K ROM 20K OTP ST92E163 1 ST92163E ST92T163E ST92E163E ST92163L 1 ST92T163L ST92E163L ST92163V 1 ST92T163V1 ST92E163V 1 1 TQFP64 CQFP64 TQFP64 CQFP64 16-MHz 8-MHz Low Voltage Low Voltage Mode Mode LVD 20K ROM 20K OTP 20K EPROM 20K ROM 20K OTP 20K EPROM Yes No 20K ROM 20K OTP 20K EPROM USB Yes 20K EPROM Contact sales office for availability 6/224 RAM No 2K No Yes Yes In Normal Mode only ST92163 - GENERAL DESCRIPTION INTRODUCTION (Cont’d) Figure 1. Maximum Operating Frequency (fMAX) versus Supply Voltage (VDD) MAX FREQUE NCY (MHz) FUNCTIONALITY IS NOT GUARANTEED IN THIS AREA 24 20 16 12 8 4 NORMAL VOLTAGE MODE 3) 16-MHz LOW VOLTAGE MODE 1) 8-MHz LOW VOLTAGE MODE 2) 0 2.5 3.0 4 4.5 5 5.5 SUPPLY VOLTAGE (V) Notes: 1) This mode is supported by 16-MHz Low Voltage devices only 2) This mode is supported by 8-MHz Low Voltage devices and 16-MHz Low Voltage devices 3) This mode is supported by all devices 7/224 ST92163 - GENERAL DESCRIPTION INTRODUCTION (Cont’d) Figure 2. ST92163 Architectural Block Diagram 20K ROM/ EPROM/OTP External Memory Interface 2K RAM MF TIMER TINA TINB TOUTA TOUTB A/D Converter AIN[5:0] EXTRG DMA USB with 16 endpoints 9V/3.3V Voltage Regulator 256 bytes Register File 8/16-bit CPU INT[7:0] NMI MEMORY BUS USBGND USBVCC USBDM0 USBDP0 USBOE USBSOF Fully Prog. I/Os Low Voltage Detector LVD** Interrupt Management AS BACK BREQ DS WAIT RW A[15:0] D[7:0] P0[7:0] P1[7:0] P3[7:0] P4[3:0] P5[7:0] P6[5:0] P6[7:6]* DMA SCI WKUP[14:0] OSCIN OSCOUT RESET INTCLK WDIN WDOUT Wakeup and Interrupt Management REGISTER BUS ST9+ CORE RCCU I2C BUS WATCHDOG TIMER MIRROR REGISTER All alternate functions (Italic characters ) are mapped on Ports 0,1, 3, 4, 5 and 6 *64-pin devices only **on some devices only (refer to “Device Summary” on page 6) 8/224 TXCLK RXCLK SIN DCD SOUT CLKOUT RTS SDS SDA SCL ST92163 - GENERAL DESCRIPTION INTRODUCTION (Cont’d) 1.1.1 Core Architecture The nucleus of the ST92163 is the enhanced ST9 Core that includes the Central Processing Unit (CPU), the register file, the interrupt and DMA controller, and the Memory Management Unit (MMU). Three independent buses are controlled by the Core: a 22-bit memory bus, an 8-bit register addressing bus and a 6-bit interrupt/DMA bus which connects the interrupt and DMA controllers in the on-chip peripherals with the core. This multiple bus architecture makes the ST9 family devices highly efficient for accessing on and off-chip memory and fast exchange of data with the on-chip peripherals. The general-purpose registers can be used as accumulators, index registers, or address pointers. Adjacent register pairs make up 16-bit registers for addressing or 16-bit processing. Although the ST9 has an 8-bit ALU, the chip handles 16-bit operations, including arithmetic, loads/stores, and memory/register and memory/memory exchanges. Many opcodes specify byte or word operations, the hardware automatically handles 16-bit operations and accesses. For interrupts or subroutine calls, the CPU uses a system stack in conjunction with the stack pointer (SP). A separate user stack has its own SP. The separate stacks, without size limitations, can be in on-chip RAM (or in Register File) or off-chip memory. 1.1.2 Instruction Set The ST9 instruction set consists of 94 instruction types, including instructions for bit handling, byte (8-bit) and word (16-bit) data, as well as BCD and Boolean formats. Instructions have been added to facilitate large program and data handling through the MMU, as well as to improve the performance and code density of C Function calls. 14 addressing modes are available, including powerful indirect addressing capabilities. The bit-manipulation instructions of the ST9 are set, clear, complement, test and set, load, and various logic instructions (AND, OR, and XOR). Math functions include add, subtract, increment, decrement, decimal adjust, multiply and divide. 1.1.3 External MEMORY INTERFACE The ST92163 device has a 16-bit external address bus allowing it to address up to 64K bytes of external memory. 1.1.4 OPERATING MODES To optimize performance versus the power consumption of the device, ST9 devices now support a range of operating modes that can be dynamically selected depending on the performance and functionality requirements of the application at a given moment. Run Mode. This is the full speed execution mode with CPU and peripherals running at the maximum clock speed delivered by the Phase Locked Loop (PLL) of the Clock Control Unit (CCU). Slow Mode. Power consumption can be significantly reduced by running the CPU and the peripherals at reduced clock speed using the CPU Prescaler and CCU Clock Divider. Wait For Interrupt Mode. The Wait For Interrupt (WFI) instruction suspends program execution until an interrupt request is acknowledged. During WFI, the CPU clock is halted while the peripheral and interrupt controller keep running at a frequency programmable via the CCU. In this mode, the power consumption of the device can be reduced by more than 95% (LP WFI). Halt Mode. When executing the HALT instruction, and if the Watchdog is not enabled, the CPU and its peripherals stop operating and the status of the machine remains frozen (the clock is also stopped). A reset is necessary to exit from Halt mode. Stop Mode. Under user program control, (see Wake-up and Interrupt Management Unit), the CPU and its peripherals stop operating and the status of the machine remains frozen (the clock is also stopped) until program execution is woken up by an event on an external Wake-up pin. 9/224 ST92163 - GENERAL DESCRIPTION INTRODUCTION (Cont’d) 1.1.5 On-chip Peripherals USB Interface The USB interface provides a full speed USB 1.1 compliant port with embedded transceiver and voltage regulator. Up to 16 endpoints are available supporting up to 8 USB devices. Separate transmit and receive DMA channels are available for each device for fast data transfers with internal RAM. Parallel I/O Ports The ST9 is provided with dedicated lines for input/ output. These lines, grouped into 8-bit ports, can be independently programmed to provide parallel input/output or to carry input/output signals to or from the on-chip peripherals and core. All ports have active pull-ups and pull-down resistors compatible with TTL loads. In addition pull-ups can be turned off for open drain operation and weak pullups can be turned on to save chip resistive pullups. Input buffers can be either TTL or CMOS compatible. High Current (10 mA) outputs are available for driving external devices such as LEDs. Multifunction Timer The Multifunction Timer has a 16-bit Up/Down counter supported by two 16-bit compare registers, two 16-bit input capture registers and two DMA channels. Timing resolution can be programmed using an 8-bit prescaler. 12 operating modes allow a range of different timing functions to be easily performed such as complex waveform generatation, measurement or PWM output. 10/224 16-bit Timer/Watchdog The Timer/Watchdog peripheral can be used as a watchdog or for a wide range of other timing functions such as generating periodic interrupts, measuring input signal pulse widths, requesting an interrupt after a set number of events. It can also generate a square wave or PWM output signal. Serial Communications Controller The SCI provides a synchronous or asynchronous serial I/O port using two DMA channels. Baud rates and data formats are programmable. Controller applications can further benefit from the self test and address wake-up facility offered by the character search mode. I2C Bus Interface The I2C bus is a synchronous serial bus for connecting multiple devices using a data line and a clock line. Multimaster and slave modes are supported. Data transfer between the bus and memory is performed by DMA. The I2C interface supports 7 and 10-bit addressing. It operates in multimaster or slave mode and supports speeds of up to 400 KHz. Bus events (Bus busy, slave address recognized) and error conditions are automatically flagged in peripheral registers and interrupts are optionally generated. Analog/Digital Converter The ADC provides up to 6 analog inputs with onchip sample and hold, fast conversion time and 8bit resolution. Conversion can be triggered by a signal from the Multifunction Timer (MFT). ST92163 - GENERAL DESCRIPTION 1.2 PIN DESCRIPTION P1.3/A11/WKUP14 P1.4/A12/WKUP14 P1.5/A13/WKUP14 P1.6/A14/WKUP14 P1.7/A15/WKUP14 N.C. N.C. V SS V DD P4.0/BREQ P4.1/WAIT P4.2 P4.3//BACK USBDM0 USBDP0 N.C. Figure 3. 64-Pin Package Pin-Out 1 64 16 48 32 N.C. USBVCC USBGND DS P3.0/INT7/SOUT P3.1/INT7/RTS P3.2/INT7/TXCLK/CLKOUT P3.3/INT7/RXCLK P3.4/INT7/DCD P3.5/INT7/SIN P3.6/INT7/AS P3.7/INT7/SDS V PP RESET P5.0/INT1/TINA N.C. AVDD WKUP12/AIN0/INTCLK/P6.2 WKUP11/SCL/EXTRG/INT6/P6.1 WKUP10/SDA/INT5/P6.0 VDD OSCIN VSS OSCOUT WDOUT/NMI/P5.7 WKUP9/TOUTB/P5.6 RW/WDIN/INT0/P5.5 USBOE/WKUP8/P5.4 TOUTA/INT2/P5.3 INT3/P5.2 TINB/INT4/P5.1 N.C. WKUP14/A10/P1.2 WKUP14/A9/P1.1 WKUP14/A8/P1.0 D7/A7/P0.7 D6/A6/P0.6 D5/A5/P0.5 D4/A4/P0.4 D3/A3/P0.3 D2/A2/P0.2 D1/A1/P0.1 D0/A0/P0.0 AIN5/P6.7 AIN4/P6.6 USBSOF/AIN3/P6.5 USBSOF/AIN2/P6.4 WKUP13/AIN1/P6.3 N.C. = Not connected 11/224 ST92163 - GENERAL DESCRIPTION Figure 4. 56-Pin Package Pin-Out WKUP3/RXCLK/INT7/P3.3 WKUP2/CLKOUT/TXCLK/INT7/P3.2 WKUP1/RTS/INT7/P3.1 WKUP0/SOUT/INT7/P3.0 DS USBGND USBVCC USBDP0 USBDM0 BACK/P4.3 P4.2 WAIT/P4.1 BREQ/P4.0 VDD VSS WKUP14/A15/P1.7 WKUP14/A14/P1.6 WKUP14/A13/P1.5 WKUP14/A12/P1.4 WKUP14/A11/P1.3 WKUP14/A10/P1.2 WKUP14/A9/P1.1 WKUP14/A8/P1.0 D7/A7/P0.7 D6/A6/P0.6 D5/A5/P0.5 D4/A4/P0.4 D3/A3/P0.3 1 56 28 29 AVDD VPP 12/224 QFP64 VSS DIP56 VDD QFP64 Name Table 2. Primary Function pins DIP56 Table 1. Power Supply Pins P3.4/INT7/DCD/WKUP4 P3.5/INT7/SIN/WKUP5 P3.6/INT7/ASN/WKUP6 P3.7/INT7/SDS/WKUP7 VPP RESET P5.0/INT1/TI NA P5.1/INT4/TI NB P5.2/INT3 P5.3/INT2/TO UTA P5.4/WKUP8/USBOE P5.5/INT0/W DIN/RW P5.6/TOUTB /WKUP9 P5.7/NMI/W DOUT OSCOUT VSS OSCIN VDD P6.0/INT5/SDA/WKUP10 P6.1/INT6/EXTRG/SCL/WKUP11 P6.2/INTCLK/AIN0/WKUP12 AVDD P6.3/AIN1/WKUP13 P6.4/AIN2/USBSOF P6.5/AIN3/USBSOF P0.0/A0/D0 P0.1/A1/D1 P0.2/A2/D2 Main Power Supply Voltage 14 21 DS Data Strobe 5 45 (2 pins internally connected) 39 56 OSCIN Oscillator Input 40 22 Digital Circuit Ground 15 23 OSCOUT Oscillator Output 42 24 (2 pins internally connected) 41 57 RESET Reset to initialize the ST9 51 35 Analog Circuit Supply Voltage 35 17 USBGND USB bus ground level 6 46 USBVCC USB voltage regulator output 7 47 USBDM0 USB Upstream port Data- line 9 51 USBDP0 USB Upstream port Data+ line 8 50 Function EPROM Programming Voltage. Must be connected to ground in normal operating mode. 52 36 Name Function ST92163 - GENERAL DESCRIPTION 1.3 I/O Port Pins All the ports of the device can be programmed as Input/Output or in Input mode, compatible with TTL or CMOS levels (except where Schmitt Trigger is present). Each bit can be programmed individually (Refer to the I/O ports chapter). TTL/CMOS Input For all those port bits where no input schmitt trigger is implemented, it is always possible to program the input level as TTL or CMOS compatible by programming the relevant PxC2.n control bit. Refer I/O Ports Chapter to the section titled “Input/ Output Bit Configuration”. Push-Pull/OD Output The output buffer can be programmed as pushpull or open-drain: attention must be paid to the fact that the open-drain option corresponds only to a disabling of P-channel MOS transistor of the buffer itself: it is still present and physically connected to the pin. Consequently it is not possible to increase the output voltage on the pin over VDD+0.3 Volt, to avoid direct junction biasing. Pure Open-drain Output The user can increase the voltage on an I/O pin over VDD+0.3 Volt where the P-channel MOS transistor is physically absent: this is allowed on all “Pure Open Drain” pins. Of course, in this case the push-pull option is not available and any weak pull-up must implemented externally. Table 3. I/O Port Characteristics Port Port Port Port Port Port Port Port Port 0[7:0] 1[7:0] 3[7:0] 4[3:0] 5[7:0] 6[1:0] 6[5:2] 6.6 6.7 Input TTL/CMOS TTL/CMOS Schmitt trigger Schmitt trigger Schmitt trigger Schmitt trigger TTL/CMOS Schmitt trigger TTL/CMOS Output Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Pure Open Drain with high sink capability Push-Pull/OD with high sink capability Push-Pull/OD with high sink capability Push-Pull/OD with high sink capability Weak Pull-Up Yes Yes Yes Yes Yes No Yes No No Reset State Bidirectional WPU Bidirectional WPU Bidirectional WPU Bidirectional WPU Bidirectional WPU Bidirectional Bidirectional WPU Bidirectional Bidirectional Legend: WPU = Weak Pull-Up, OD = Open Drain 13/224 ST92163 - GENERAL DESCRIPTION Table 4. ST92163 Alternate Functions Alternate Functions QFP64 General Purpose I/O DIP56 Port Name Pin No. P0.0 31 11 A0/D0 I/O Ext. Mem. Address/Data bit 0 P0.1 30 10 A1/D1 I/O Ext. Mem. Address/Data bit 1 P0.2 29 9 A2/D2 I/O Ext. Mem. Address/Data bit 2 P0.3 28 8 A3/D3 I/O Ext. Mem. Address/Data bit 3 P0.4 27 7 A4/D4 I/O Ext. Mem. Address/Data bit 4 P0.5 26 6 A5/D5 I/O Ext. Mem. Address/Data bit 5 P0.6 25 5 A6/D6 I/O Ext. Mem. Address/Data bit 6 P0.7 24 4 P1.0 23 3 P1.1 22 2 P1.2 P1.3 21 1 All ports useable for general purpose I/O (input, output or bidirectional) 20 64 P1.4 19 63 P1.5 18 62 P1.6 17 61 P1.7 16 60 P3.0 P3.1 14/224 4 3 A7/D7 I/O Ext. Mem. Address/Data bit 7 A8 I/O Ext. Mem. Address bit 8 WKUP14 A9 WKUP14 A10 WKUP14 A11 WKUP14 A12 WKUP14 A13 WKUP14 A14 WKUP14 A15 I Wakeup Line 14 (***) I/O Ext. Mem. Address bit 9 I Wakeup Line 14 (***) I/O Ext. Mem. Address bit 10 I Wakeup Line 14 (***) I/O Ext. Mem. Address bit 11 I Wakeup Line 14 (***) I/O Ext. Mem. Address bit 12 I Wakeup Line 14 (***) I/O Ext. Mem. Address bit 13 I Wakeup Line 14 (***) I/O Ext. Mem. Address bit 14 I Wakeup Line 14 (***) I/O Ext. Mem. Address bit 15 WKUP14 I Wakeup Line 14 (***) WKUP0 I Wakeup Line 0 I External Interrupt 7 (*) SOUT O SCI Data Output WKUP1 44 INT7 O Wakeup Line 1 43 INT7 I External Interrupt 7 (*) RTS O SCI Request to Send ST92163 - GENERAL DESCRIPTION P3.2 P3.3 2 1 Alternate Functions QFP64 General Purpose I/O DIP56 Port Name Pin No. 42 41 WKUP2 I Wakeup Line 2 INT7 I External Interrupt 7 (*) TXCLK I SCI Transmit CK Input CLKOUT O SCI Clock Output WKUP3 I Wakeup Line 3 INT7 I External Interrupt 7 (*) RXCLK I SCI Receive CK Input O P3.4 56 40 WKUP4 I Wakeup Line 4 INT7 I External Interrupt 7 (*) DCD I SCI Data Carrier Detect O P3.5 P3.6 P3.7 WKUP5 All ports useable INT7 for general pur55 39 SIN pose I/O (input, output or bidirectional) WKUP6 I Wakeup Line 5 I External Interrupt 7 (*) I SCI Data Input O I Wakeup Line 6 I External Interrupt 7 (*) AS (**) O Ext. Mem. Address Strobe WKUP7 I Wakeup Line 7 54 38 INT7 53 37 INT7 I External Interrupt 7 (*) SDS O SCI Synchronous Data Send I Ext. Mem. Bus Request WAIT I Ext. Mem. Wait Input RW O Ext. Mem. Read/Write Mode Select P4.0 13 55 BREQ P4.1 12 54 P4.2 11 53 P4.3 10 52 I AS (**) O Ext. Mem. Address Strobe I BACK O Ext. Mem. bus acknow 15/224 ST92163 - GENERAL DESCRIPTION P5.0 Alternate Functions QFP64 General Purpose I/O DIP56 Port Name Pin No. INT1 I External Interrupt 1 50 34 TINA I MF Timer Input A O INT4 P5.1 49 31 TINB I External Interrupt 4 I MF Timer Input B O P5.2 48 30 INT3 I External Interrupt 3 P5.3 47 29 INT2 I External Interrupt 2 TOUTA O MF Timer Output A WKUP8 I Wakeup Line 8 USBOE O USB Output enable WDIN I Watchdog Timer Input P5.4 P5.5 46 28 45 27 INT0 RW P5.6 P5.7 P6.0 P6.1 P6.2 P6.3 WKUP9 All ports useable 44 26 TOUTB for general purNMI pose I/O (input, output or bidirec- 43 25 WDOUT tional) WKUP10 I External Interrupt 0 O Ext. Mem. Read/Write Mode Select I Wakeup Line 9 O MF Timer Output B I Non Maskable Interrupt O Watchdog Timer Output I Wakeup Line 10 INT5 I External Interrupt 5 SDAI I I2C Bus Data In SDAO O I2C Bus Data Out WKUP11 I Wakeup Line 11 INT6 I External Interrupt 6 37 19 SCLI I I2C Bus Clock In EXTRG I A/D External Trigger SCLO O I2C Bus Clock Out AIN0 I A/D Analog Input 0 38 20 36 18 WKUP12 I Wakeup Line 12 INTCLK O Internal Clock WKUP13 I Wakeup Line 13 I A/D Analog Input 1 34 16 AIN1 O 16/224 ST92163 - GENERAL DESCRIPTION P6.4 P6.5 P6.6 P6.7 Alternate Functions QFP64 General Purpose I/O DIP56 Port Name Pin No. AIN2 I A/D Analog Input 2 USBSOF O USB SOF Synchro All ports useable 32 14 AIN3 for general purUSBSOF pose I/O (input, AIN4 output or bidirec13 tional) I A/D Analog Input 3 O USB SOF Synchro I A/D Analog Input 4 33 15 - 12 AIN5 O I A/D Analog Input 5 O *Eight interrupt lines internally connected to INT7 through a boolean AND function. ** AS cannot be disabled by software if the ASAF bit is set (Page Register 245) once the corresponding P3.6 bit is configured as an Alternate Function output. ***Eight wakeup lines internally connected to WKUP14 through a boolean AND function. Note: The reset state of Port 0 and Port 1 is Input, Weak Pull-Up. To interface external memory, the ports must be configured by software as alternate function output. 17/224 ST92163 - GENERAL DESCRIPTION How to configure the I/O ports To configure the I/O ports, use the information in Table 3 and Table 4 and the Port Bit Configuration Table in the I/O Ports Chapter on page 100. I/O Note = the hardware characteristics fixed for each port line. Inputs: – If I/O note = TTL/CMOS, either TTL or CMOS input level can be selected by software. – If I/O note = Schmitt trigger, selecting CMOS or TTL input by software has no effect, the input will always be Schmitt Trigger. Outputs: – If I/O note = Push-Pull, either Push Pull or Open Drain can be selected by software. – If I/O note = Open Drain, selecting Push-Pull by software has no effect, the input will always be Open Drain. Alternate Functions (AF) = More than one AF cannot be assigned to an external pin at the same time: it can be selected as follows, but simultaneous availability of several functions of one pin is obviously impossible. AF Inputs: – AF is selected implicitly by enabling the corresponding peripheral. Exceptions to this are ADC 18/224 inputs which are selected explicitly as AF by software. AF Outputs or Bidirectional Lines: – In the case of Outputs or I/Os, AF is selected explicitly by software. Example 1: Timer/Watchdog input AF: WDIN, Port: P5.5, I/O note: Input Schmitt Trigger. Write the port configuration bits: P5C2.5=1 P5C1.5=0 P5C0.5=1 Enable the WDT peripheral by software as described in the WDT chapter. Example 2: Timer/Watchdog output AF: WDOUT, Port: P5.7, I/O note: None Write the port configuration bits: P5C2.7=0 P5C1.7=1 P5C0.7=1 Example 3: ADC input AF: AIN0, Port: P6.2, I/O note: Does not apply to ADC Write the port configuration bits: P6C2.2=1 P6C1.2=1 P6C0.2=1 ST92163 - GENERAL DESCRIPTION 1.4 MEMORY MAP Figure 5. ST92163 Memory Map 3FFFFFh Upper Memory (usually external RAM mapped in Segment 23h) External Memory 220000h 21FFFFh SEGMENT 21h 64 Kbytes Internal RAM 2 Kbytes Reserved 210000h 20FFFFh 20FFFFh PAGE 83 - 16 Kbytes 20C000h 20BFFFh 20F800h PAGE 82 - 16 Kbytes SEGMENT 20h 64 Kbytes 208000h 207FFFh PAGE 81 - 16 Kbytes 204000h 203FFFh Note: Internal RAM addresses are repeated each 2 Kbytes inside segment 20h. PAGE 80 - 16 Kbytes 200000h 1FFFFFh Lower Memory (usually external ROM/EPROM mapped in Segment 1) External Memory 010000h 00FFFFh Reserved 00C000h 00BFFFh 008000h 007FFFh Internal ROM/EPROM 20 Kbytes 004FFFh Internal ROM/EPROM 000000h 004000h 003FFFh PAGE 3 - 16 Kbytes PAGE 2 - 16 Kbytes SEGMENT 0 PAGE 1 - 16 Kbytes 64 Kbytes PAGE 0 - 16 Kbytes 000000h Note: The total amount of external memory is 64 Kbytes. 19/224 ST92163 - GENERAL DESCRIPTION 1.5 ST92163 REGISTER MAP Table 6 contains the map of the group F peripheral pages. The common registers used by each peripheral are listed in Table 5. Be very careful to correctly program both: – The set of registers dedicated to a particular function or peripheral. – Registers common to other functions. – In particular, double-check that any registers with “undefined” reset values have been correctly initialized. Warning: Note that in the EIVR and each IVR register, all bits are significant. Take care when defining base vector addresses that entries in the Interrupt Vector table do not overlap. Table 5. Common Registers Function or Peripheral SCI, MFT Common Registers CICR + NICR + DMA REGISTERS + I/O PORT REGISTERS ADC CICR + NICR + I/O PORT REGISTERS WDT CICR + NICR + EXTERNAL INTERRUPT REGISTERS + I/O PORT REGISTERS I/O PORTS EXTERNAL INTERRUPT RCCU I/O PORT REGISTERS + MODER INTERRUPT REGISTERS + I/O PORT REGISTERS INTERRUPT REGISTERS + MODER Figure 6. ST92163 Register Groups REGISTER FILE 255 240 F PAGED REGISTERS 239 E SYSTEM REGISTERS 224 223 D C B A 9 8 7 6 These register groups (16 registers per group) are potentially reserved for USB DMA. 5 4 The amount of reserved registers depends on the number of endpoints used in the program. (8 registers are used per endpoint). 3 2 1 0 0 20/224 15 0 ST92163 - GENERAL DESCRIPTION Table 6. Group F Pages Register Map Resources available on the ST92163 device: Page Register 0 R255 R254 3 4 5 9 10 15 20 21 24 43 55 57 62 Res. Res. Port 3 R253 R252 2 Port 9 Res. WUI MU WCR R251 Res. R250 Port 6 WDT R249 R248 USB Res. R247 Port 8 Res. MFT Endpoints USB Common I2C MMU ADC SCI R246 Res. R245 R244 EXT INT Port 5 Res. Res. R243 Res. RCCU R242 Port 4 R241 MFT Res. R240 21/224 ST92163 - GENERAL DESCRIPTION Table 7. Detailed Register Map Page No. Block Reg. No. Register Name Description Reset Value Hex. I/O R227 P3DR Port 3 Data Register FF Port R228 P4DR Port 4 Data Register FF 3:5 R229 P5DR Port 5 Data Register FF R230 CICR Central Interrupt Control Register 87 30 R231 FLAGR Flag Register 00 31 R232 RP0 Pointer 0 Register 00 33 R233 RP1 Pointer 1 Register 00 33 R234 PPR Page Pointer Register 54 35 R235 MODER Mode Register E0 35 R236 USPHR User Stack Pointer High Register xx 37 R237 USPLR User Stack Pointer Low Register xx 37 R238 SSPHR System Stack Pointer High Reg. xx 37 R239 SSPLR System Stack Pointer Low Reg. xx 37 R242 EITR External Interrupt Trigger Register 00 59 System Core INT 0 WDT 2 3 External Interrupt Pending Reg. 00 60 EIMR External Interrupt Mask-bit Reg. 00 60 R245 EIPLR External Interrupt Priority Level Reg. FF 60 R246 EIVR External Interrupt Vector Register x6 61 R247 NICR Nested Interrupt Control 00 61 R248 WDTHR Watchdog Timer High Register FF 109 R249 WDTLR Watchdog Timer Low Register FF 109 R250 WDTPR Watchdog Timer Prescaler Reg. FF 109 R251 WDTCR Watchdog Timer Control Register 12 109 110 WCR Wait Control Register 7F R252 P3C0 Port 3 Configuration Register 0 00 Port R253 P3C1 Port 3 Configuration Register 1 00 3 R254 P3C2 Port 3 Configuration Register 2 00 I/O R240 P4C0 Port 4 Configuration Register 0 00 Port R241 P4C1 Port 4 Configuration Register 1 00 4 R242 P4C2 Port 4 Configuration Register 2 00 I/O R244 P5C0 Port 5 Configuration Register 0 00 Port R245 P5C1 Port 5 Configuration Register 1 00 5 R246 P5C2 Port 5 Configuration Register 2 00 R248 P6C0 Port 6 Configuration Register 0 00 R249 P6C1 Port 6 Configuration Register 1 00 R250 P6C2 Port 6 Configuration Register 2 00 R251 P6DR Port 6 Data Register FF 6 22/224 EIPR R244 R252 Port 98 R243 I/O I/O Doc. Page 98 98 ST92163 - GENERAL DESCRIPTION Page No. Block 4 USB End Points 5 9 MFT Reg. No. Register Name Description Reset Value Hex. R240 EP0RA Endpoint 0 Register A (Transmission) 00 R241 EP0RB Endpoint 0 Register B (Reception) 00 R242 EP1RA Endpoint 1 Register A (Transmission) 00 R243 EP1RB Endpoint 1 Register B (Reception) 00 R244 EP2RA Endpoint 2 Register A (Transmission) 00 R245 EP2RB Endpoint 2 Register B (Reception) 00 R246 EP3RA Endpoint 3 Register A (Transmission) 00 R247 EP3RB Endpoint 3 Register B (Reception) 00 R248 EP4RA Endpoint 4 Register A (Transmission) 00 R249 EP4RB Endpoint 4 Register B (Reception) 00 R250 EP5RA Endpoint 5 Register A (Transmission) 00 R251 EP5RB Endpoint 5 Register B (Reception) 00 R252 EP6RA Endpoint 6 Register A (Transmission) 00 R253 EP6RB Endpoint 6 Register B (Reception) 00 R254 EP7RA Endpoint 7 Register A (Transmission) 00 R255 EP7RB Endpoint 7 Register B (Reception) 00 R240 EP8RA Endpoint 8 Register A (Transmission) 00 R241 EP8RB Endpoint 8 Register B (Reception) 00 R242 EP9RA Endpoint 9 Register A (Transmission) 00 Doc. Page 143 R243 EP9RB Endpoint 9 Register B (Reception) 00 R244 EP10RA Endpoint 10 Register A (Transmission) 00 R245 EP10RB Endpoint 10 Register B (Reception) 00 R246 EP11RA Endpoint 11 Register A (Transmission) 00 R247 EP11RB Endpoint 11 Register B (Reception) 00 R248 EP12RA Endpoint 12 Register A (Transmission) 00 R249 EP12RB Endpoint 12 Register B (Reception) 00 R250 EP13RA Endpoint 13 Register A (Transmission) 00 R251 EP13RB Endpoint 13 Register B (Reception) 00 R252 EP14RA Endpoint 14 Register A (Transmission) 00 R253 EP14RB Endpoint 14 Register B (Reception) 00 R254 EP15RA Endpoint 15 Register A (Transmission) 00 R255 EP15RB Endpoint 15 Register B (Reception) 00 R240 DCPR DMA Counter Pointer Register xx 132 R241 DAPR DMA Address Pointer Register xx 133 R242 T_IVR Interrupt Vector Register xx 133 R243 IDCR Interrupt/DMA Control Register C7 134 R248 IOCR I/O Connection Register FC 134 23/224 ST92163 - GENERAL DESCRIPTION Page No. 10 15 24/224 Block Reg. No. Register Name Description Reset Value Hex. Doc. Page R240 REG0HR Capture Load Register 0 High xx 125 R241 REG0LR Capture Load Register 0 Low xx 125 R242 REG1HR Capture Load Register 1 High xx 125 R243 REG1LR Capture Load Register 1 Low xx 125 R244 CMP0HR Compare 0 Register High 00 125 125 R245 CMP0LR Compare 0 Register Low 00 R246 CMP1HR Compare 1 Register High 00 125 R247 CMP1LR Compare 1 Register Low 00 125 R248 TCR Timer Control Register 0x 126 R249 TMR Timer Mode Register 00 127 R250 T_ICR External Input Control Register 0x 128 R251 PRSR Prescaler Register 00 128 R252 OACR Output A Control Register xx 129 R253 OBCR Output B Control Register xx 130 R254 T_FLAGR Flags Register 00 31 R255 IDMR Interrupt/DMA Mask Register 00 132 R240 DADDR0 Device Address Register 0 00 R241 DADDR1 Device Address Register 1 00 R242 DADDR2 Device Address Register 2 00 R243 DADDR3 Device Address Register 3 00 R244 DADDR4 Device Address Register 4 00 R245 DADDR5 Device Address Register 5 00 R246 DADDR6 Device Address Register 6 00 USB R247 DADDR7 Device Address Register 7 00 Common R248 USBIVR USB Interrupt Vector Register xx 139 R249 USBISTR USB Interrupt Status Register 00 139 R250 USBIMR USB Interrupt Mask Register 00 140 R251 USBIPR USB Interrupt Priority Register xx 140 R252 USBCTLR USB Control Register 17 141 R253 CTRINF CTR Interrrupt Flags xx 142 R254 FNRH Frame Number Register High 0x 142 R255 FNRL Frame Number Register Low xx 142 MFT 143 ST92163 - GENERAL DESCRIPTION Page No. 20 Block I2C MMU 21 EXTMI 24 SCI Reg. No. Register Name Description Reset Value Hex. Doc. Page R240 I2CCR I2C Control Register 00 186 R241 I2CSR1 I2C Status Register 1 00 187 R242 I2CSR2 I2C Status Register 2 00 189 2 R243 I2CCCR I C Clock Control Register 00 190 R244 I2COAR1 I2C Own Address Register 1 00 190 191 2 R245 I2COAR2 I C Own Address Register 2 00 R246 I2CDR I2C Data Register 00 191 R247 I2CADR I2C General Call Address A0 191 R248 I2CISR I2C Interrupt Status Register xx 192 R249 I2CIVR I2C Interrupt Vector Register xx 193 R250 I2CRDAP Receiver DMA Source Addr. Pointer xx 193 R251 I2CRDC Receiver DMA Transaction Counter xx 193 R252 I2CTDAP Transmitter DMA Source Addr. Pointer xx 194 R253 I2CTDC Transmitter DMA Transaction Counter xx 194 2 R254 I2CECCR I C Extended Clock Control Register 00 194 R255 I2CIMR I2C Interrupt Mask Register x0 195 R240 DPR0 Data Page Register 0 00 42 R241 DPR1 Data Page Register 1 01 42 R242 DPR2 Data Page Register 2 02 42 R243 DPR3 Data Page Register 3 83 42 R244 CSR Code Segment Register 00 43 R248 ISR Interrupt Segment Register x0 43 R249 DMASR DMA Segment Register x0 43 R245 EMR1 External Memory Register 1 80 95 R246 EMR2 External Memory Register 2 0F 96 R240 RDCPR Receiver DMA Transaction Counter Pointer xx 165 R241 RDAPR Receiver DMA Source Address Pointer xx 165 R242 TDCPR Transmitter DMA Transaction Counter Pointer xx 165 R243 TDAPR Transmitter DMA Destination Address Pointer xx 165 R244 S_IVR Interrupt Vector Register xx 166 R245 ACR Address/Data Compare Register xx 166 R246 IMR Interrupt Mask Register x0 167 R247 S_ISR Interrupt Status Register xx 43 R248 RXBR Receive Buffer Register xx 169 R248 TXBR Transmitter Buffer Register xx 169 R249 IDPR Interrupt/DMA Priority Register xx 170 R250 CHCR Character Configuration Register xx 171 R251 CCR Clock Configuration Register 00 172 R252 BRGHR Baud Rate Generator High Reg. xx 173 R253 BRGLR Baud Rate Generator Low Register xx 173 R254 SICR Synchronous Input Control 03 173 R255 SOCR Synchronous Output Control 01 174 25/224 ST92163 - GENERAL DESCRIPTION Page No. Block I/O Port 8 43 I/O Port 9 55 59 60 62 RCCU WUIMU USB ADC Reg. No. Register Name Description Reset Value Hex. R248 P8C0 Port 8 Configuration Register 0 00 R249 P8C1 Port 8 Configuration Register 1 00 R250 P8C2 Port 8 Configuration Register 2 00 R251 P8DR Port 8 Data Register FF R252 P9C0 Port 9 Configuration Register 0 00 R253 P9C1 Port 9 Configuration Register 1 00 R254 P9C2 Port 9 Configuration Register 2 00 R255 P9DR Port 9 Data Register FF R240 CLKCTL Clock Control Register 00 81 82 Doc. Page 98 R242 CLK_FLAG Clock Flag Register 48, 28 or 08 R246 PLLCONF PLL Configuration Register xx 83 R249 WUCTRL Wake-Up Control Register 00 67 R250 WUMRH Wake-Up Mask Register High 00 68 R251 WUMRL Wake-Up Mask Register Low 00 68 R252 WUTRH Wake-Up Trigger Register High 00 69 69 R253 WUTRL Wake-Up Trigger Register Low 00 R254 WUPRH Wake-Up Pending Register High 00 69 R255 WUPRL Wake-Up Pending Register Low 00 69 R244 DEVCONF1 USB device configuration 1 0F 146 R245 DEVCONF2 USB device configuration 2 00 146 R246 MIRRA Mirror Register A xx 147 R247 MIRRB Mirror Register B xx 147 R240 ADDTR Channel i Data Register xx 199 R241 ADCLR Control Logic Register 00 199 R242 ADINT AD Interrupt Register 01 200 Note: xx denotes a byte with an undefined value, but some bits may have defined values. See register description for details. 26/224 ST92163 - DEVICE ARCHITECTURE 2 DEVICE ARCHITECTURE 2.1 CORE ARCHITECTURE The ST9+ Core or Central Processing Unit (CPU) features a highly optimised instruction set, capable of handling bit, byte (8-bit) and word (16-bit) data, as well as BCD and Boolean formats; 14 addressing modes are available. Four independent buses are controlled by the Core: a 16-bit Memory bus, an 8-bit Register data bus, an 8-bit Register address bus and a 6-bit Interrupt/DMA bus which connects the interrupt and DMA controllers in the on-chip peripherals with the Core. This multiple bus architecture affords a high degree of pipelining and parallel operation, thus making the ST9+ family devices highly efficient, both for numerical calculation, data handling and with regard to communication with on-chip peripheral resources. which hold data and control bits for the on-chip peripherals and I/Os. – A single linear memory space accommodating both program and data. All of the physically separate memory areas, including the internal ROM, internal RAM and external memory are mapped in this common address space. The total addressable memory space of 4 Mbytes (limited by the size of on-chip memory and the number of external address pins) is arranged as 64 segments of 64 Kbytes. Each segment is further subdivided into four pages of 16 Kbytes, as illustrated in Figure 1. A Memory Management Unit uses a set of pointer registers to address a 22-bit memory field using 16-bit address-based instructions. 2.2.1 Register File The Register File consists of (see Figure 2): 2.2 MEMORY SPACES – 224 general purpose registers (Group 0 to D, There are two separate memory spaces: registers R0 to R223) – The Register File, which comprises 240 8-bit – 6 system registers in the System Group (Group registers, arranged as 15 groups (Group 0 to E), E, registers R224 to R239) each containing sixteen 8-bit registers plus up to – Up to 64 pages, depending on device configura64 pages of 16 registers mapped in Group F, tion, each containing up to 16 registers, mapped to Group F (R240 to R255), see Figure 3. Figure 7. Single Program and Data Memory Address Space Data 16K Pages Address 255 254 253 252 251 250 249 248 247 3FFFFFh 3F0000h 3EFFFFh 3E0000h Code 64K Segments 63 62 up to 4 Mbytes 21FFFFh 210000h 20FFFFh 02FFFFh 020000h 01FFFFh 010000h 00FFFFh 000000h Reserved 135 134 133 132 11 10 9 8 7 6 5 4 3 2 1 0 33 2 1 0 27/224 ST92163 - DEVICE ARCHITECTURE MEMORY SPACES (Cont’d) Figure 8. Register Groups Figure 9. Page Pointer for Group F mapping PAGE 63 UP TO 64 PAGES 255 240 F PAGED REGISTERS 239 E SYSTEM REGISTERS 224 223 D PAGE 5 R255 PAGE 0 C B A R240 9 R234 8 224 GENERAL PURPOSE REGISTERS 7 6 PAGE POINTER R224 5 4 3 2 1 0 15 0 0 VA00432 R0 VA00433 Figure 10. Addressing the Register File REGISTER FILE 255 240 F PAGED REGISTERS 239 E SYSTEM REGISTERS 224 223 D GROUP D C R195 (R0C3h) B R207 A 9 (1100) (0011) 8 GROUP C 7 6 R195 5 4 R192 3 GROUP B 2 1 0 0 15 0 VR000118 28/224 ST92163 - DEVICE ARCHITECTURE MEMORY SPACES (Cont’d) 2.2.2 Register Addressing Register File registers, including Group F paged registers (but excluding Group D), may be addressed explicitly by means of a decimal, hexadecimal or binary address; thus R231, RE7h and R11100111b represent the same register (see Figure 4). Group D registers can only be addressed in Working Register mode. Note that an upper case “R” is used to denote this direct addressing mode. Working Registers Certain types of instruction require that registers be specified in the form “rx”, where x is in the range 0 to 15: these are known as Working Registers. Note that a lower case “r” is used to denote this indirect addressing mode. Two addressing schemes are available: a single group of 16 working registers, or two separately mapped groups, each consisting of 8 working registers. These groups may be mapped starting at any 8 or 16 byte boundary in the register file by means of dedicated pointer registers. This technique is described in more detail in Section 1.3.3, and illustrated in Figure 5 and in Figure 6. System Registers The 16 registers in Group E (R224 to R239) are System registers and may be addressed using any of the register addressing modes. These registers are described in greater detail in Section 1.3. Paged Registers Up to 64 pages, each containing 16 registers, may be mapped to Group F. These are addressed using any register addressing mode, in conjunction with the Page Pointer register, R234, which is one of the System registers. This register selects the page to be mapped to Group F and, once set, does not need to be changed if two or more registers on the same page are to be addressed in succession. Therefore if the Page Pointer, R234, is set to 5, the instructions: spp #5 ld R242, r4 will load the contents of working register r4 into the third register of page 5 (R242). These paged registers hold data and control information relating to the on-chip peripherals, each peripheral always being associated with the same pages and registers to ensure code compatibility between ST9+ devices. The number of these registers therefore depends on the peripherals which are present in the specific ST9+ family device. In other words, pages only exist if the relevant peripheral is present. Table 8. Register File Organization Hex. Address Decimal Address Function Register File Group F0-FF 240-255 Paged Registers Group F E0-EF 224-239 System Registers Group E D0-DF 208-223 Group D C0-CF 192-207 Group C B0-BF 176-191 Group B A0-AF 160-175 Group A 90-9F 144-159 Group 9 80-8F 128-143 Group 8 General Purpose Registers 70-7F 112-127 60-6F 96-111 Group 7 50-5F 80-95 Group 5 40-4F 64-79 Group 4 30-3F 48-63 Group 3 20-2F 32-47 Group 2 10-1F 16-31 Group 1 00-0F 00-15 Group 0 Group 6 29/224 ST92163 - DEVICE ARCHITECTURE 2.3 SYSTEM REGISTERS The System registers are listed in Table 2 System Registers (Group E). They are used to perform all the important system settings. Their purpose is described in the following pages. Refer to the chapter dealing with I/O for a description of the PORT[5:0] Data registers. Table 9. System Registers (Group E) R239 (EFh) SSPLR R238 (EEh) SSPHR R237 (EDh) USPLR R236 (ECh) USPHR R235 (EBh) MODE REGISTER R234 (EAh) PAGE POINTER REGISTER R233 (E9h) REGISTER POINTER 1 R232 (E8h) REGISTER POINTER 0 R231 (E7h) FLAG REGISTER R230 (E6h) CENTRAL INT. CNTL REG R229 (E5h) PORT5 DATA REG. R228 (E4h) PORT4 DATA REG. R227 (E3h) PORT3 DATA REG. R226 (E2h) PORT2 DATA REG. R225 (E1h) PORT1 DATA REG. R224 (E0h) PORT0 DATA REG. GCEN TLIP 0 TLI IEN IAM CPL2 CPL1 CPL0 Bit 7 = GCEN: Global Counter Enable. This bit is the Global Counter Enable of the Multifunction Timers. The GCEN bit is ANDed with the CE bit in the TCR Register (only in devices featuring the MFT Multifunction Timer) in order to enable the Timers when both bits are set. This bit is set after the Reset cycle. 30/224 Bit 6 = TLIP: Top Level Interrupt Pending . This bit is set by hardware when a Top Level Interrupt Request is recognized. This bit can also be set by software to simulate a Top Level Interrupt Request. 0: No Top Level Interrupt pending 1: Top Level Interrupt pending Bit 5 = TLI: Top Level Interrupt bit. 0: Top Level Interrupt is acknowledged depending on the TLNM bit in the NICR Register. 1: Top Level Interrupt is acknowledged depending on the IEN and TLNM bits in the NICR Register (described in the Interrupt chapter). 2.3.1 Central Interrupt Control Register Please refer to the ”INTERRUPT” chapter for a detailed description of the ST9 interrupt philosophy. CENTRAL INTERRUPT CONTROL REGISTER (CICR) R230 - Read/Write Register Group: E (System) Reset Value: 1000 0111 (87h) 7 Note: If an MFT is not included in the ST9 device, then this bit has no effect. Bit 4 = IEN: Interrupt Enable . This bit is cleared by interrupt acknowledgement, and set by interrupt return (iret). IEN is modified implicitly by iret, ei and di instructions or by an interrupt acknowledge cycle. It can also be explicitly written by the user, but only when no interrupt is pending. Therefore, the user should execute a di instruction (or guarantee by other means that no interrupt request can arrive) before any write operation to the CICR register. 0: Disable all interrupts except Top Level Interrupt. 1: Enable Interrupts Bit 3 = IAM: Interrupt Arbitration Mode. This bit is set and cleared by software to select the arbitration mode. 0: Concurrent Mode 1: Nested Mode. Bit 2:0 = CPL[2:0]: Current Priority Level. These three bits record the priority level of the routine currently running (i.e. the Current Priority Level, CPL). The highest priority level is represented by 000, and the lowest by 111. The CPL bits can be set by hardware or software and provide the reference according to which subsequent interrupts are either left pending or are allowed to interrupt the current interrupt service routine. When the current interrupt is replaced by one of a higher priority, the current priority value is automatically stored until required in the NICR register. ST92163 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) 2.3.2 Flag Register The Flag Register contains 8 flags which indicate the CPU status. During an interrupt, the flag register is automatically stored in the system stack area and recalled at the end of the interrupt service routine, thus returning the CPU to its original status. This occurs for all interrupts and, when operating in nested mode, up to seven versions of the flag register may be stored. FLAG REGISTER (FLAGR) R231- Read/Write Register Group: E (System) Reset value: 0000 0000 (00h) 7 C 0 Z S V DA H - DP Bit 7 = C: Carry Flag . The carry flag is affected by: Addition (add, addw, adc, adcw), Subtraction (sub, subw, sbc, sbcw), Compare (cp, cpw), Shift Right Arithmetic (sra, sraw), Shift Left Arithmetic (sla, slaw), Swap Nibbles (swap), Rotate (rrc, rrcw, rlc, rlcw, ror, rol), Decimal Adjust (da), Multiply and Divide (mul, div, divws). When set, it generally indicates a carry out of the most significant bit position of the register being used as an accumulator (bit 7 for byte operations and bit 15 for word operations). The carry flag can be set by the Set Carry Flag (scf) instruction, cleared by the Reset Carry Flag (rcf) instruction, and complemented by the Complement Carry Flag (ccf) instruction. Bit 6 = Z: Zero Flag. The Zero flag is affected by: Addition (add, addw, adc, adcw), Subtraction (sub, subw, sbc, sbcw), Compare (cp, cpw), Shift Right Arithmetic (sra, sraw), Shift Left Arithmetic (sla, slaw), Swap Nibbles (swap), Rotate (rrc, rrcw, rlc, rlcw, ror, rol), Decimal Adjust (da), Multiply and Divide (mul, div, divws), Logical (and, andw, or, orw, xor, xorw, cpl), Increment and Decrement (inc, incw, dec, decw), Test (tm, tmw, tcm, tcmw, btset). In most cases, the Zero flag is set when the contents of the register being used as an accumulator become zero, following one of the above operations. Bit 5 = S: Sign Flag. The Sign flag is affected by the same instructions as the Zero flag. The Sign flag is set when bit 7 (for a byte operation) or bit 15 (for a word operation) of the register used as an accumulator is one. Bit 4 = V: Overflow Flag. The Overflow flag is affected by the same instructions as the Zero and Sign flags. When set, the Overflow flag indicates that a two’scomplement number, in a result register, is in error, since it has exceeded the largest (or is less than the smallest), number that can be represented in two’s-complement notation. Bit 3 = DA: Decimal Adjust Flag. The DA flag is used for BCD arithmetic. Since the algorithm for correcting BCD operations is different for addition and subtraction, this flag is used to specify which type of instruction was executed last, so that the subsequent Decimal Adjust (da) operation can perform its function correctly. The DA flag cannot normally be used as a test condition by the programmer. Bit 2 = H: Half Carry Flag. The H flag indicates a carry out of (or a borrow into) bit 3, as the result of adding or subtracting two 8-bit bytes, each representing two BCD digits. The H flag is used by the Decimal Adjust (da) instruction to convert the binary result of a previous addition or subtraction into the correct BCD result. Like the DA flag, this flag is not normally accessed by the user. Bit 1 = Reserved bit (must be 0). Bit 0 = DP: Data/Program Memory Flag. This bit indicates the memory area addressed. Its value is affected by the Set Data Memory (sdm) and Set Program Memory (spm) instructions. Refer to the Memory Management Unit for further details. 31/224 ST92163 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) If the bit is set, data is accessed using the Data Pointers (DPRs registers), otherwise it is pointed to by the Code Pointer (CSR register); therefore, the user initialization routine must include a Sdm instruction. Note that code is always pointed to by the Code Pointer (CSR). Note: In the ST9+, the DP flag is only for compatibility with software developed for the first generation of ST9 devices. With the single memory addressing space, its use is now redundant. It must be kept to 1 with a Sdm instruction at the beginning of the program to ensure a normal use of the different memory pointers. 2.3.3 Register Pointing Techniques Two registers within the System register group, are used as pointers to the working registers. Register Pointer 0 (R232) may be used on its own as a single pointer to a 16-register working space, or in conjunction with Register Pointer 1 (R233), to point to two separate 8-register spaces. For the purpose of register pointing, the 16 register groups of the register file are subdivided into 32 8register blocks. The values specified with the Set Register Pointer instructions refer to the blocks to be pointed to in twin 8-register mode, or to the lower 8-register block location in single 16-register mode. The Set Register Pointer instructions srp, srp0 and srp1 automatically inform the CPU whether the Register File is to operate in single 16-register mode or in twin 8-register mode. The srp instruction selects the single 16-register group mode and 32/224 specifies the location of the lower 8-register block, while the srp0 and srp1 instructions automatically select the twin 8-register group mode and specify the locations of each 8-register block. There is no limitation on the order or position of these register groups, other than that they must start on an 8-register boundary in twin 8-register mode, or on a 16-register boundary in single 16register mode. The block number should always be an even number in single 16-register mode. The 16-register group will always start at the block whose number is the nearest even number equal to or lower than the block number specified in the srp instruction. Avoid using odd block numbers, since this can be confusing if twin mode is subsequently selected. Thus: srp #3 will be interpreted as srp #2 and will allow using R16 ..R31 as r0 .. r15. In single 16-register mode, the working registers are referred to as r0 to r15. In twin 8-register mode, registers r0 to r7 are in the block pointed to by RP0 (by means of the srp0 instruction), while registers r8 to r15 are in the block pointed to by RP1 (by means of the srp1 instruction). Caution: Group D registers can only be accessed as working registers using the Register Pointers, or by means of the Stack Pointers. They cannot be addressed explicitly in the form “Rxxx”. ST92163 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) POINTER 0 REGISTER (RP0) R232 - Read/Write Register Group: E (System) Reset Value: xxxx xx00 (xxh) POINTER 1 REGISTER (RP1) R233 - Read/Write Register Group: E (System) Reset Value: xxxx xx00 (xxh) 7 RG4 RG3 RG2 RG1 RG0 RPS 0 0 7 0 RG4 Bit 7:3 = RG[4:0]: Register Group number. These bits contain the number (in the range 0 to 31) of the register block specified in the srp0 or srp instructions. In single 16-register mode the number indicates the lower of the two 8-register blocks to which the 16 working registers are to be mapped, while in twin 8-register mode it indicates the 8-register block to which r0 to r7 are to be mapped. Bit 2 = RPS: Register Pointer Selector. This bit is set by the instructions srp0 and srp1 to indicate that the twin register pointing mode is selected. The bit is reset by the srp instruction to indicate that the single register pointing mode is selected. 0: Single register pointing mode 1: Twin register pointing mode 0 RG3 RG2 RG1 RG0 RPS 0 0 This register is only used in the twin register pointing mode. When using the single register pointing mode, or when using only one of the twin register groups, the RP1 register must be considered as RESERVED and may NOT be used as a general purpose register. Bit 7:3 = RG[4:0]: Register Group number. These bits contain the number (in the range 0 to 31) of the 8-register block specified in the srp1 instruction, to which r8 to r15 are to be mapped. Bit 2 = RPS: Register Pointer Selector . This bit is set by the srp0 and srp1 instructions to indicate that the twin register pointing mode is selected. The bit is reset by the srp instruction to indicate that the single register pointing mode is selected. 0: Single register pointing mode 1: Twin register pointing mode Bit 1:0: Reserved. Forced by hardware to zero. Bit 1:0: Reserved. Forced by hardware to zero. 33/224 ST92163 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) Figure 11. Pointing to a single group of 16 registers REGIST ER GROUP BLOCK NUMBER REGISTER GROUP BLOCK NUMBER Figure 12. Pointing to two groups of 8 registers REGISTER FILE REGIST ER FILE 31 REGISTER POINTE R 0 & REGIST ER POINTE R 1 F 31 REGISTER POINTER 0 set by: F 30 srp #2 29 instruction E 30 29 E set by: 28 srp0 #2 28 & points to: 27 D 27 D srp1 #7 instructions 26 point to: 26 25 25 addressed by BLOCK 7 9 4 9 8 4 r15 8 7 GROUP 3 3 7 r8 6 3 6 5 2 5 4 2 4 3 r15 1 3 1 GROUP 1 r0 2 r0 1 0 0 34/224 r7 2 addressed by BLOCK 2 1 0 0 GROUP 1 addressed by BLOCK 2 ST92163 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) 2.3.4 Paged Registers Up to 64 pages, each containing 16 registers, may be mapped to Group F. These paged registers hold data and control information relating to the on-chip peripherals, each peripheral always being associated with the same pages and registers to ensure code compatibility between ST9+ devices. The number of these registers depends on the peripherals present in the specific ST9 device. In other words, pages only exist if the relevant peripheral is present. The paged registers are addressed using the normal register addressing modes, in conjunction with the Page Pointer register, R234, which is one of the System registers. This register selects the page to be mapped to Group F and, once set, does not need to be changed if two or more registers on the same page are to be addressed in succession. Thus the instructions: spp #5 ld R242, r4 will load the contents of working register r4 into the third register of page 5 (R242). Warning: During an interrupt, the PPR register is not saved automatically in the stack. If needed, it should be saved/restored by the user within the interrupt routine. PAGE POINTER REGISTER (PPR) R234 - Read/Write Register Group: E (System) Reset value: xxxx xx00 (xxh) 7 PP5 0 PP4 PP3 PP2 PP1 PP0 0 0 Bit 7:2 = PP[5:0]: Page Pointer. These bits contain the number (in the range 0 to 63) of the page specified in the spp instruction. Once the page pointer has been set, there is no need to refresh it unless a different page is required. Bit 1:0: Reserved. Forced by hardware to 0. 2.3.5 Mode Register The Mode Register allows control of the following operating parameters: – Selection of internal or external System and User Stack areas, – Management of the clock frequency, – Enabling of Bus request and Wait signals when interfacing to external memory. MODE REGISTER (MODER) R235 - Read/Write Register Group: E (System) Reset value: 1110 0000 (E0h) 7 SSP 0 USP DIV2 PRS2 PRS1 PRS0 BRQEN HIMP Bit 7 = SSP: System Stack Pointer. This bit selects an internal or external System Stack area. 0: External system stack area, in memory space. 1: Internal system stack area, in the Register File (reset state). Bit 6 = USP: User Stack Pointer. This bit selects an internal or external User Stack area. 0: External user stack area, in memory space. 1: Internal user stack area, in the Register File (reset state). Bit 5 = DIV2: OSCIN Clock Divided by 2. This bit controls the divide-by-2 circuit operating on OSCIN. 0: Clock divided by 1 1: Clock divided by 2 Bit 4:2 = PRS[2:0]: CPUCLK Prescaler. These bits load the prescaler division factor for the internal clock (INTCLK). The prescaler factor selects the internal clock frequency, which can be divided by a factor from 1 to 8. Refer to the Reset and Clock Control chapter for further information. Bit 1 = BRQEN: Bus Request Enable. 0: External Memory Bus Request disabled 1: External Memory Bus Request enabled on the BREQ pin (where available). Bit 0 = HIMP: High Impedance Enable. When any of Ports 0, 1, 2 or 6 depending on device configuration, are programmed as Address and Data lines to interface external Memory, these lines and the Memory interface control lines (AS, 35/224 ST92163 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) DS, R/W) can be forced into the High Impedance state by setting the HIMP bit. When this bit is reset, it has no effect. Setting the HIMP bit is recommended for noise reduction when only internal Memory is used. If Port 1 and/or 2 are declared as an address AND as an I/O port (for example: P10... P14 = Address, and P15... P17 = I/O), the HIMP bit has no effect on the I/O lines. 2.3.6 Stack Pointers Two separate, double-register stack pointers are available: the System Stack Pointer and the User Stack Pointer, both of which can address registers or memory. The stack pointers point to the “bottom” of the stacks which are filled using the push commands and emptied using the pop commands. The stack pointer is automatically pre-decremented when data is “pushed” in and post-incremented when data is “popped” out. The push and pop commands used to manage the System Stack may be addressed to the User Stack by adding the suffix “u”. To use a stack instruction for a word, the suffix “w” is added. These suffixes may be combined. When bytes (or words) are “popped” out from a stack, the contents of the stack locations are unchanged until fresh data is loaded. Thus, when data is “popped” from a stack area, the stack contents remain unchanged. Note: Instructions such as: pushuw RR236 or pushw RR238, as well as the corresponding pop instructions (where R236 & R237, and R238 & R239 are themselves the user and system stack pointers respectively), must not be used, since the pointer values are themselves automatically changed by the push or pop instruction, thus corrupting their value. System Stack The System Stack is used for the temporary storage of system and/or control data, such as the Flag register and the Program counter. The following automatically push data onto the System Stack: – Interrupts When entering an interrupt, the PC and the Flag Register are pushed onto the System Stack. If the ENCSR bit in the EMR2 register is set, then the 36/224 Code Segment Register is also pushed onto the System Stack. – Subroutine Calls When a call instruction is executed, only the PC is pushed onto stack, whereas when a calls instruction (call segment) is executed, both the PC and the Code Segment Register are pushed onto the System Stack. – Link Instruction The link or linku instructions create a C language stack frame of user-defined length in the System or User Stack. All of the above conditions are associated with their counterparts, such as return instructions, which pop the stored data items off the stack. User Stack The User Stack provides a totally user-controlled stacking area. The User Stack Pointer consists of two registers, R236 and R237, which are both used for addressing a stack in memory. When stacking in the Register File, the User Stack Pointer High Register, R236, becomes redundant but must be considered as reserved. Stack Pointers Both System and User stacks are pointed to by double-byte stack pointers. Stacks may be set up in RAM or in the Register File. Only the lower byte will be required if the stack is in the Register File. The upper byte must then be considered as reserved and must not be used as a general purpose register. The stack pointer registers are located in the System Group of the Register File, this is illustrated in Table 2 System Registers (Group E). Stack location Care is necessary when managing stacks as there is no limit to stack sizes apart from the bottom of any address space in which the stack is placed. Consequently programmers are advised to use a stack pointer value as high as possible, particularly when using the Register File as a stacking area. Group D is a good location for a stack in the Register File, since it is the highest available area. The stacks may be located anywhere in the first 14 groups of the Register File (internal stacks) or in RAM (external stacks). Note. Stacks must not be located in the Paged Register Group or in the System Register Group. ST92163 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) USER STACK POINTER HIGH REGISTER (USPHR) R236 - Read/Write Register Group: E (System) Reset value: undefined SYSTEM STACK POINTER HIGH REGISTER (SSPHR) R238 - Read/Write Register Group: E (System) Reset value: undefined 7 0 USP15 USP14 USP13 USP12 USP11 USP10 USP9 USP8 USER STACK POINTER LOW REGISTER (USPLR) R237 - Read/Write Register Group: E (System) Reset value: undefined USP6 USP5 USP4 USP3 USP2 USP1 SSP15 SSP14 SSP13 SSP12 SSP11 SSP10 SSP9 0 7 USP0 SSP7 Figure 13. Internal Stack Mode 0 SSP6 SSP5 REGISTER FILE STACK POINTER (LOW) F SSP8 SSP4 SSP3 SSP2 SSP1 SSP0 Figure 14. External Stack Mode REGIST ER FILE points to: 0 SYSTEM STACK POINTER LOW REGISTER (SSPLR) R239 - Read/Write Register Group: E (System) Reset value: undefined 7 USP7 7 F STACK POINTER (LOW) & STACK POINTER (HIGH) point to: MEMORY E E STACK D D 4 4 3 3 2 2 1 1 0 0 STACK 37/224 ST92163 - DEVICE ARCHITECTURE 2.4 MEMORY ORGANIZATION Code and data are accessed within the same linear address space. All of the physically separate memory areas, including the internal ROM, internal RAM and external memory are mapped in a common address space. The ST9+ provides a total addressable memory space of 4 Mbytes. This address space is arranged as 64 segments of 64 Kbytes; each segment is again subdivided into four 16 Kbyte pages. 38/224 The mapping of the various memory areas (internal RAM or ROM, external memory) differs from device to device. Each 64-Kbyte physical memory segment is mapped either internally or externally; if the memory is internal and smaller than 64 Kbytes, the remaining locations in the 64-Kbyte segment are not used (reserved). Refer to the Register and Memory Map Chapter for more details on the memory map. ST92163 - DEVICE ARCHITECTURE 2.5 MEMORY MANAGEMENT UNIT The CPU Core includes a Memory Management Unit (MMU) which must be programmed to perform memory accesses (even if external memory is not used). The MMU is controlled by 7 registers and 2 bits (ENCSR and DPRREM) present in EMR2, which may be written and read by the user program. These registers are mapped within group F, Page 21 of the Register File. The 7 registers may be Figure 15. Page 21 Registers sub-divided into 2 main groups: a first group of four 8-bit registers (DPR[3:0]), and a second group of three 6-bit registers (CSR, ISR, and DMASR). The first group is used to extend the address during Data Memory access (DPR[3:0]). The second is used to manage Program and Data Memory accesses during Code execution (CSR), Interrupts Service Routines (ISR or CSR), and DMA transfers (DMASR or ISR). Page 21 FFh R255 FEh R254 FDh R253 FCh R252 FBh R251 FAh R250 F9h DMASR R249 F8h ISR R248 F7h Relocation of P[3:0] and DPR[3:0] Registers MMU R247 F6h EMR2 R246 F5h EMR1 R245 F4h CSR R244 F3h DPR3 R243 F2h DPR2 R242 F1h DPR1 R241 F0h DPR0 R240 EM MMU MMU SSPLR SSPHR USPLR USPHR MODER PPR RP1 RP0 FLAGR CICR P5DR P4DR P3DR P2DR P1DR P0DR DMASR ISR EMR2 EMR1 CSR DPR3 DPR2 1 DPR0 Bit DPRREM=0 (default setting) SSPLR SSPHR USPLR USPHR MODER PPR RP1 RP0 FLAGR CICR P5DR P4DR DPR3 DPR2 DPR1 DPR0 DMASR ISR EMR2 EMR1 CSR P3DR P2DR P1DR P0DR Bit DPRREM=1 39/224 ST92163 - DEVICE ARCHITECTURE 2.6 ADDRESS SPACE EXTENSION To manage 4 Mbytes of addressing space it is necessary to have 22 address bits. The MMU adds 6 bits to the usual 16-bit address, thus translating a 16-bit virtual address into a 22-bit physical address. There are 2 different ways to do this depending on the memory involved and on the operation being performed. 2.6.1 Addressing 16-Kbyte Pages This extension mode is implicitly used to address Data memory space if no DMA is being performed. The Data memory space is divided into 4 pages of 16 Kbytes. Each one of the four 8-bit registers (DPR[3:0], Data Page Registers) selects a different 16-Kbyte page. The DPR registers allow access to the entire memory space which contains 256 pages of 16 Kbytes. Data paging is performed by extending the 14 LSB of the 16-bit address with the contents of a DPR register. The two MSBs of the 16-bit address are interpreted as the identification number of the DPR register to be used. Therefore, the DPR registers Figure 16. Addressing via DPR[3:0] are involved in the following virtual address ranges: DPR0: from 0000h to 3FFFh; DPR1: from 4000h to 7FFFh; DPR2: from 8000h to BFFFh; DPR3: from C000h to FFFFh. The contents of the selected DPR register specify one of the 256 possible data memory pages. This 8-bit data page number, in addition to the remaining 14-bit page offset address forms the physical 22-bit address (see Figure 10). A DPR register cannot be modified via an addressing mode that uses the same DPR register. For instance, the instruction “POPW DPR0” is legal only if the stack is kept either in the register file or in a memory location above 8000h, where DPR2 and DPR3 are used. Otherwise, since DPR0 and DPR1 are modified by the instruction, unpredictable behaviour could result. 16-bit virtual address MMU registers DPR0 DPR1 DPR2 DPR3 00 01 10 11 8 bits 14 LSB 22-bit physical address 40/224 2M SB ST92163 - DEVICE ARCHITECTURE ADDRESS SPACE EXTENSION (Cont’d) 2.6.2 Addressing 64-Kbyte Segments This extension mode is used to address Data memory space during a DMA and Program memory space during any code execution (normal code and interrupt routines). Three registers are used: CSR, ISR, and DMASR. The 6-bit contents of one of the registers CSR, ISR, or DMASR define one out of 64 Memory segments of 64 Kbytes within the 4 Mbytes address space. The register contents represent the 6 MSBs of the memory address, whereas the 16 LSBs of the address (intra-segment address) are given by the virtual 16-bit address (see Figure 11). 2.7 MMU REGISTERS The MMU uses 7 registers mapped into Group F, Page 21 of the Register File and 2 bits of the EMR2 register. Most of these registers do not have a default value after reset. 2.7.1 DPR[3:0]: Data Page Registers The DPR[3:0] registers allow access to the entire 4 Mbyte memory space composed of 256 pages of 16 Kbytes. 2.7.1.1 Data Page Register Relocation If these registers are to be used frequently, they may be relocated in register group E, by programming bit 5 of the EMR2-R246 register in page 21. If this bit is set, the DPR[3:0] registers are located at R224-227 in place of the Port 0-3 Data Registers, which are re-mapped to the default DPR’s locations: R240-243 page 21. Data Page Register relocation is illustrated in Figure 9. Figure 17. Addressing via CSR, ISR, and DMASR 16-bit virtual address MMU registers CSR 1 1 2 3 Fetching program instruction Data Memory accessed in DMA Fetching interrupt instruction or DMA access to Program Memory DMASR 2 ISR 3 6 bits 22-bit physical address 41/224 ST92163 - DEVICE ARCHITECTURE MMU REGISTERS (Cont’d) DATA PAGE REGISTER 0 (DPR0) R240 - Read/Write Register Page: 21 Reset value: undefined This register is relocated to R224 if EMR2.5 is set. 7 0 DATA PAGE REGISTER 2 (DPR2) R242 - Read/Write Register Page: 21 Reset value: undefined This register is relocated to R226 if EMR2.5 is set. 7 0 DPR0_7 DPR0_6 DPR0_5 DPR0_4 DPR0_3 DPR0_2 DPR0_1 DPR0_0 DPR2_7 DPR2_6 DPR2_5 DPR2_4 DPR2_3 DPR2_2 DPR2_1 DPR2_0 Bit 7:0 = DPR0_[7:0]: These bits define the 16Kbyte Data Memory page number. They are used as the most significant address bits (A21-14) to extend the address during a Data Memory access. The DPR0 register is used when addressing the virtual address range 0000h-3FFFh. Bit 7:0 = DPR2_[7:0]: These bits define the 16Kbyte Data memory page. They are used as the most significant address bits (A21-14) to extend the address during a Data memory access. The DPR2 register is involved when the virtual address is in the range 8000h-BFFFh. DATA PAGE REGISTER 1 (DPR1) R241 - Read/Write Register Page: 21 Reset value: undefined This register is relocated to R225 if EMR2.5 is set. DATA PAGE REGISTER 3 (DPR3) R243 - Read/Write Register Page: 21 Reset value: undefined This register is relocated to R227 if EMR2.5 is set. 7 0 7 0 DPR1_7 DPR1_6 DPR1_5 DPR1_4 DPR1_3 DPR1_2 DPR1_1 DPR1_0 DPR3_7 DPR3_6 DPR3_5 DPR3_4 DPR3_3 DPR3_2 DPR3_1 DPR3_0 Bit 7:0 = DPR1_[7:0]: These bits define the 16Kbyte Data Memory page number. They are used as the most significant address bits (A21-14) to extend the address during a Data Memory access. The DPR1 register is used when addressing the virtual address range 4000h-7FFFh. Bit 7:0 = DPR3_[7:0]: These bits define the 16Kbyte Data memory page. They are used as the most significant address bits (A21-14) to extend the address during a Data memory access. The DPR3 register is involved when the virtual address is in the range C000h-FFFFh. 42/224 ST92163 - DEVICE ARCHITECTURE MMU REGISTERS (Cont’d) 2.7.2 CSR: Code Segment Register This register selects the 64-Kbyte code segment being used at run-time to access instructions. It can also be used to access data if the spm instruction has been executed (or ldpp, ldpd, lddp). Only the 6 LSBs of the CSR register are implemented, and bits 6 and 7 are reserved. The CSR register allows access to the entire memory space, divided into 64 segments of 64 Kbytes. To generate the 22-bit Program memory address, the contents of the CSR register is directly used as the 6 MSBs, and the 16-bit virtual address as the 16 LSBs. Note: The CSR register should only be read and not written for data operations (there are some exceptions which are documented in the following paragraph). It is, however, modified either directly by means of the jps and calls instructions, or indirectly via the stack, by means of the rets instruction. CODE SEGMENT REGISTER (CSR) R244 - Read/Write Register Page: 21 Reset value: 0000 0000 (00h) 7 0 0 0 CSR_5 CSR_4 CSR_3 CSR_2 CSR_1 CSR_0 Bit 7:6 = Reserved, keep in reset state. Bit 5:0 = CSR_[5:0]: These bits define the 64Kbyte memory segment (among 64) which contains the code being executed. These bits are used as the most significant address bits (A21-16). 0 0 0 Bit 7:6 = Reserved, keep in reset state. Bit 5:0 = ISR_[5:0]: These bits define the 64-Kbyte memory segment (among 64) which contains the interrupt vector table and the code for interrupt service routines and DMA transfers (when the PS bit of the DAPR register is reset). These bits are used as the most significant address bits (A21-16). The ISR is used to extend the address space in two cases: – Whenever an interrupt occurs: ISR points to the 64-Kbyte memory segment containing the interrupt vector table and the interrupt service routine code. See also the Interrupts chapter. – During DMA transactions between the peripheral and memory when the PS bit of the DAPR register is reset : ISR points to the 64 K-byte Memory segment that will be involved in the DMA transaction. 2.7.4 DMASR: DMA Segment Register DMA SEGMENT REGISTER (DMASR) R249 - Read/Write Register Page: 21 Reset value: undefined 7 0 0 0 DMA SR_5 DMA SR_4 DMA SR_3 DMA SR_2 DMA SR_1 DMA SR_0 Bit 7:6 = Reserved, keep in reset state. 2.7.3 ISR: Interrupt Segment Register INTERRUPT SEGMENT REGISTER (ISR) R248 - Read/Write Register Page: 21 Reset value: undefined 7 ISR and ENCSR bit (EMR2 register) are also described in the chapter relating to Interrupts, please refer to this description for further details. ISR_5 ISR_4 ISR_3 ISR_2 ISR_1 ISR_0 Bit 5:0 = DMASR_[5:0]: These bits define the 64Kbyte Memory segment (among 64) used when a DMA transaction is performed between the peripheral’s data register and Memory, with the PS bit of the DAPR register set. These bits are used as the most significant address bits (A21-16). If the PS bit is reset, the ISR register is used to extend the address. 43/224 ST92163 - DEVICE ARCHITECTURE MMU REGISTERS (Cont’d) Figure 18. Memory Addressing Scheme (example) 4M bytes 3FFFFFh 16K 294000h DPR3 240000h 23FFFFh DPR2 DPR1 DPR0 16K 20C000h 16K 200000h 1FFFFFh 64K 040000h 03FFFFh 030000h DMASR 020000h 44/224 ISR 64K CSR 16K 64K 010000h 00C000h 000000h ST92163 - DEVICE ARCHITECTURE 2.8 MMU USAGE 2.8.1 Normal Program Execution Program memory is organized as a set of 64Kbyte segments. The program can span as many segments as needed, but a procedure cannot stretch across segment boundaries. jps, calls and rets instructions, which automatically modify the CSR, must be used to jump across segment boundaries. Writing to the CSR is forbidden during normal program execution because it is not synchronized with the opcode fetch. This could result in fetching the first byte of an instruction from one memory segment and the second byte from another. Writing to the CSR is allowed when it is not being used, i.e during an interrupt service routine if ENCSR is reset. Note that a routine must always be called in the same way, i.e. either always with call or always with calls, depending on whether the routine ends with ret or rets. This means that if the routine is written without prior knowledge of the location of other routines which call it, and all the program code does not fit into a single 64-Kbyte segment, then calls/rets should be used. In typical microcontroller applications, less than 64 Kbytes of RAM are used, so the four Data space pages are normally sufficient, and no change of DPR[3:0] is needed during Program execution. It may be useful however to map part of the ROM into the data space if it contains strings, tables, bit maps, etc. If there is to be frequent use of paging, the user can set bit 5 (DPRREM) in register R246 (EMR2) of Page 21. This swaps the location of registers DPR[3:0] with that of the data registers of Ports 03. In this way, DPR registers can be accessed without the need to save/set/restore the Page Pointer Register. Port registers are therefore moved to page 21. Applications that require a lot of paging typically use more than 64 Kbytes of external memory, and as ports 0, 1 and 2 are required to address it, their data registers are unused. 2.8.2 Interrupts The ISR register has been created so that the interrupt routines may be found by means of the same vector table even after a segment jump/call. When an interrupt occurs, the CPU behaves in one of 2 ways, depending on the value of the ENCSR bit in the EMR2 register (R246 on Page 21). If this bit is reset (default condition), the CPU works in original ST9 compatibility mode. For the duration of the interrupt service routine, the ISR is used instead of the CSR, and the interrupt stack frame is kept exactly as in the original ST9 (only the PC and flags are pushed). This avoids the need to save the CSR on the stack in the case of an interrupt, ensuring a fast interrupt response time. The drawback is that it is not possible for an interrupt service routine to perform segment calls/jps: these instructions would update the CSR, which, in this case, is not used (ISR is used instead). The code size of all interrupt service routines is thus limited to 64 Kbytes. If, instead, bit 6 of the EMR2 register is set, the ISR is used only to point to the interrupt vector table and to initialize the CSR at the beginning of the interrupt service routine: the old CSR is pushed onto the stack together with the PC and the flags, and then the CSR is loaded with the ISR. In this case, an iret will also restore the CSR from the stack. This approach lets interrupt service routines access the whole 4-Mbyte address space. The drawback is that the interrupt response time is slightly increased, because of the need to also save the CSR on the stack. Compatibility with the original ST9 is also lost in this case, because the interrupt stack frame is different; this difference, however, would not be noticeable for a vast majority of programs. Data memory mapping is independent of the value of bit 6 of the EMR2 register, and remains the same as for normal code execution: the stack is the same as that used by the main program, as in the ST9. If the interrupt service routine needs to access additional Data memory, it must save one (or more) of the DPRs, load it with the needed memory page and restore it before completion. 2.8.3 DMA Depending on the PS bit in the DAPR register (see DMA chapter) DMA uses either the ISR or the DMASR for memory accesses: this guarantees that a DMA will always find its memory segment(s), no matter what segment changes the application has performed. Unlike interrupts, DMA transactions cannot save/restore paging registers, so a dedicated segment register (DMASR) has been created. Having only one register of this kind means that all DMA accesses should be programmed in one of the two following segments: the one pointed to by the ISR (when the PS bit of the DAPR register is reset), and the one referenced by the DMASR (when the PS bit is set). 45/224 ST92163 - INTERRUPTS 3 INTERRUPTS 3.1 INTRODUCTION The ST9 responds to peripheral and external events through its interrupt channels. Current program execution can be suspended to allow the ST9 to execute a specific response routine when such an event occurs, providing that interrupts have been enabled, and according to a priority mechanism. If an event generates a valid interrupt request, the current program status is saved and control passes to the appropriate Interrupt Service Routine. The ST9 CPU can receive requests from the following sources: – On-chip peripherals – External pins – Top-Level Pseudo-non-maskable interrupt According to the on-chip peripheral features, an event occurrence can generate an Interrupt request which depends on the selected mode. Up to eight external interrupt channels, with programmable input trigger edge, are available. In addition, a dedicated interrupt channel, set to the Top-level priority, can be devoted either to the external NMI pin (where available) to provide a NonMaskable Interrupt, or to the Timer/Watchdog. In- 46/224 terrupt service routines are addressed through a vector table mapped in Memory. Figure 19. Interrupt Response NORMAL PROGRAM FLOW INTERRUPT INTERRUPT SERVICE ROUTINE CLEAR PENDING BIT IRET INSTRUCTION VR001833 ST92163 - INTERRUPTS INTERRUPTS (Cont’d) 3.2 INTERRUPT VECTORING The ST9 implements an interrupt vectoring structure which allows the on-chip peripheral to identify the location of the first instruction of the Interrupt Service Routine automatically. When an interrupt request is acknowledged, the peripheral interrupt module provides, through its Interrupt Vector Register (IVR), a vector to point into the vector table of locations containing the start addresses of the Interrupt Service Routines (defined by the programmer). Each peripheral has a specific IVR mapped within its Register File pages. The Interrupt Vector table, containing the addresses of the Interrupt Service Routines, is located in the first 256 locations of Memory pointed to by the ISR register, thus allowing 8-bit vector addressing. For a description of the ISR register refer to the chapter describing the MMU. The user Power on Reset vector is stored in the first two physical bytes in memory, 000000h and 000001h. The Top Level Interrupt vector is located at addresses 0004h and 0005h in the segment pointed to by the Interrupt Segment Register (ISR). With one Interrupt Vector register, it is possible to address several interrupt service routines; in fact, peripherals can share the same interrupt vector register among several interrupt channels. The most significant bits of the vector are user programmable to define the base vector address within the vector table, the least significant bits are controlled by the interrupt module, in hardware, to select the appropriate vector. Note: The first 256 locations of the memory segment pointed to by ISR can contain program code. 3.2.1 Divide by Zero trap The Divide by Zero trap vector is located at addresses 0002h and 0003h of each code segment; it should be noted that for each code segment a Divide by Zero service routine is required. Warning. Although the Divide by Zero Trap operates as an interrupt, the FLAG Register is not pushed onto the system Stack automatically. As a result it must be regarded as a subroutine, and the service routine must end with the RET instruction (not IRET ). Figure 20. Interrupt Vector Table REGISTE R FILE PROGRAM MEMORY F PAGE REGISTE RS USER ISR USER DIVIDE-BY -ZERO ISR USER MAIN PROGRAM INT. VECTOR REGISTER USER TOP LEVEL ISR R240 R239 0000FFh ODD LO EVEN HI ISR ADDRES S VECTOR LO 000004h HI LO 000002h HI LO 000000h HI TOP LEVEL INT. TABLE DIVIDE-B Y-ZERO POWER-ON RESET 47/224 ST92163 - INTERRUPTS INTERRUPTS (Cont’d) 3.2.2 Segment Paging During Interrupt Routines The ENCSR bit in the EMR2 register can be used to select between original ST9 backward compatibility mode and ST9+ interrupt management mode. ST9 backward compatibility mode (ENCSR = 0) If ENCSR is reset, the CPU works in original ST9 compatibility mode. For the duration of the interrupt service routine, ISR is used instead of CSR, and the interrupt stack frame is identical to that of the original ST9: only the PC and Flags are pushed. This avoids saving the CSR on the stack in the event of an interrupt, thus ensuring a faster interrupt response time. It is not possible for an interrupt service routine to perform inter-segment calls or jumps: these instructions would update the CSR, which, in this case, is not used (ISR is used instead). The code segment size for all interrupt service routines is thus limited to 64K bytes. ST9+ mode (ENCSR = 1) If ENCSR is set, ISR is only used to point to the interrupt vector table and to initialize the CSR at the beginning of the interrupt service routine: the old CSR is pushed onto the stack together with the PC and flags, and CSR is then loaded with the contents of ISR. In this case, iret will also restore CSR from the stack. This approach allows interrupt service routines to access the entire 4 Mbytes of address space. The drawback is that the interrupt response time is slightly increased, because of the need to also save CSR on the stack. Full compatibility with the original ST9 is lost in this case, because the interrupt stack frame is different. ENCSR Bit Mode 0 ST9 Compatible 1 ST9+ Pushed/Popp ed PC, FLAGR, PC, FLAGR Registers CSR Max. Code Size 64KB <4 MB for interrupt Within 1 segment Across segments service routine 48/224 3.3 INTERRUPT PRIORITY LEVELS The ST9 supports a fully programmable interrupt priority structure. Nine priority levels are available to define the channel priority relationships: – The on-chip peripheral channels and the eight external interrupt sources can be programmed within eight priority levels. Each channel has a 3bit field, PRL (Priority Level), that defines its priority level in the range from 0 (highest priority) to 7 (lowest priority). – The 9th level (Top Level Priority) is reserved for the Timer/Watchdog or the External Pseudo Non-Maskable Interrupt. An Interrupt service routine at this level cannot be interrupted in any arbitration mode. Its mask can be both maskable (TLI) or non-maskable (TLNM). 3.4 PRIORITY LEVEL ARBITRATION The 3 bits of CPL (Current Priority Level) in the Central Interrupt Control Register contain the priority of the currently running program (CPU priority). CPL is set to 7 (lowest priority) upon reset and can be modified during program execution either by software or automatically by hardware according to the selected Arbitration Mode. During every instruction, an arbitration phase takes place, during which, for every channel capable of generating an Interrupt, each priority level is compared to all the other requests (interrupts or DMA). If the highest priority request is an interrupt, its PRL value must be strictly lower (that is, higher priority) than the CPL value stored in the CICR register (R230) in order to be acknowledged. The Top Level Interrupt overrides every other priority. 3.4.1 Priority level 7 (Lowest) Interrupt requests at PRL level 7 cannot be acknowledged, as this PRL value (the lowest possible priority) cannot be strictly lower than the CPL value. This can be of use in a fully polled interrupt environment. 3.4.2 Maximum depth of nesting No more than 8 routines can be nested. If an interrupt routine at level N is being serviced, no other Interrupts located at level N can interrupt it. This guarantees a maximum number of 8 nested levels including the Top Level Interrupt request. 3.4.3 Simultaneous Interrupts If two or more requests occur at the same time and at the same priority level, an on-chip daisy chain, specific to every ST9 version, selects the channel ST92163 - INTERRUPTS with the highest position in the chain, as shown in Table 10. Table 10. Daisy Chain Priority Highest Position Lowest Position INTA0 INTA1 INTB0 INTB1 INTC0 INTC1 INTD0 INTD1 USB MFT SCI I2C INT0/WDT INT1/ADC INT2 INT3 INT4/ INT5 INT6/RCCU INT7/WKUP 3.4.4 Dynamic Priority Level Modification The main program and routines can be specifically prioritized. Since the CPL is represented by 3 bits in a read/write register, it is possible to modify dynamically the current priority value during program execution. This means that a critical section can have a higher priority with respect to other interrupt requests. Furthermore it is possible to prioritize even the Main Program execution by modifying the CPL during its execution. See Figure 21 Figure 21. Example of Dynamic priority level modification in Nested Mode INTERRUPT 6 HAS PRIORITY LEVEL 6 Priority Level CPL is set to 7 4 by MAIN program ei INT6 5 MAIN CPL is set to 5 CPL6 > CPL5: 6 INT6 pending 7 INT 6 CPL=6 MAIN CPL=7 3.5 ARBITRATION MODES The ST9 provides two interrupt arbitration modes: Concurrent mode and Nested mode. Concurrent mode is the standard interrupt arbitration mode. Nested mode improves the effective interrupt response time when service routine nesting is required, depending on the request priority levels. The IAM control bit in the CICR Register selects Concurrent Arbitration mode or Nested Arbitration Mode. 3.5.1 Concurrent Mode This mode is selected when the IAM bit is cleared (reset condition). The arbitration phase, performed during every instruction, selects the request with the highest priority level. The CPL value is not modified in this mode. Start of Interrupt Routine The interrupt cycle performs the following steps: – All maskable interrupt requests are disabled by clearing CICR.IEN. – The PC low byte is pushed onto system stack. – The PC high byte is pushed onto system stack. – If ENCSR is set, CSR is pushed onto system stack. – The Flag register is pushed onto system stack. – The PC is loaded with the 16-bit vector stored in the Vector Table, pointed to by the IVR. – If ENCSR is set, CSR is loaded with ISR contents; otherwise ISR is used in place of CSR until iret instruction. End of Interrupt Routine The Interrupt Service Routine must be ended with the iret instruction. The iret instruction executes the following operations: – The Flag register is popped from system stack. – If ENCSR is set, CSR is popped from system stack. – The PC high byte is popped from system stack. – The PC low byte is popped from system stack. – All unmasked Interrupts are enabled by setting the CICR.IEN bit. – If ENCSR is reset, CSR is used instead of ISR. Normal program execution thus resumes at the interrupted instruction. All pending interrupts remain pending until the next ei instruction (even if it is executed during the interrupt service routine). Note: In Concurrent mode, the source priority level is only useful during the arbitration phase, where it is compared with all other priority levels and with the CPL. No trace is kept of its value during the ISR. If other requests are issued during the interrupt service routine, once the global CICR.IEN is re-enabled, they will be acknowledged regardless of the interrupt service routine’s priority. This may cause undesirable interrupt response sequences. 49/224 ST92163 - INTERRUPTS ARBITRATION MODES (Cont’d) Examples In the following two examples, three interrupt requests with different priority levels (2, 3 & 4) occur simultaneously during the interrupt 5 service routine. Example 1 In the first example, (simplest case, Figure 22) the ei instruction is not used within the interrupt service routines. This means that no new interrupt can be serviced in the middle of the current one. The interrupt routines will thus be serviced one after another, in the order of their priority, until the main program eventually resumes. Figure 22. Simple Example of a Sequence of Interrupt Requests with: - Concurrent mode selected and - IEN unchanged by the interrupt routines 0 INTERRUPT 2 HAS PRIORITY LEVEL 2 Priority Level of Interrupt Request INTERRUPT 3 HAS PRIORITY LEVEL 3 INTERRUPT 4 HAS PRIORITY LEVEL 4 INTERRUPT 5 HAS PRIORITY LEVEL 5 1 2 INT 2 CPL = 7 3 INT 3 CPL = 7 INT 2 INT 3 INT 4 4 5 INT 4 CPL = 7 INT 5 ei CPL = 7 6 INT 5 7 MAIN CPL is set to 7 50/224 MAIN CPL = 7 ST92163 - INTERRUPTS ARBITRATION MODES (Cont’d) Example 2 In the second example, (more complex, Figure 23), each interrupt service routine sets Interrupt Enable with the ei instruction at the beginning of the routine. Placed here, it minimizes response time for requests with a higher priority than the one being serviced. The level 2 interrupt routine (with the highest priority) will be acknowledged first, then, when the ei instruction is executed, it will be interrupted by the level 3 interrupt routine, which itself will be interrupted by the level 4 interrupt routine. When the level 4 interrupt routine is completed, the level 3 interrupt routine resumes and finally the level 2 interrupt routine. This results in the three interrupt serv- ice routines being executed in the opposite order of their priority. It is therefore recommended to avoid inserting the ei instruction in the interrupt service routine in Concurrent mode. Use the ei instruction only in nested mode. WARNING: If, in Concurrent Mode, interrupts are nested (by executing ei in an interrupt service routine), make sure that either ENCSR is set or CSR=ISR, otherwise the iret of the innermost interrupt will make the CPU use CSR instead of ISR before the outermost interrupt service routine is terminated, thus making the outermost routine fail. Figure 23. Complex Example of a Sequence of Interrupt Requests with: - Concurrent mode selected - IEN set to 1 during interrupt service routine execution 0 Priority Level of Interrupt Request INTERRUP T 2 HAS PRIORITY LEVEL 2 INTERRUP T 3 HAS PRIORITY LEVEL 3 INTERRUP T 4 HAS PRIORITY LEVEL 4 1 INTERRUP T 5 HAS PRIORITY LEVEL 5 2 3 INT 2 INT 2 CPL = 7 CPL = 7 ei INT 2 INT 3 INT 4 4 5 INT 5 ei 6 CPL = 7 INT 3 CPL = 7 ei ei INT 3 CPL = 7 INT 4 CPL = 7 INT 5 CPL = 7 ei INT 5 7 MAIN CPL is set to 7 MAIN CPL = 7 51/224 ST92163 - INTERRUPTS ARBITRATION MODES (Cont’d) 3.5.2 Nested Mode The difference between Nested mode and Concurrent mode, lies in the modification of the Current Priority Level (CPL) during interrupt processing. The arbitration phase is basically identical to Concurrent mode, however, once the request is acknowledged, the CPL is saved in the Nested Interrupt Control Register (NICR) by setting the NICR bit corresponding to the CPL value (i.e. if the CPL is 3, the bit 3 will be set). The CPL is then loaded with the priority of the request just acknowledged; the next arbitration cycle is thus performed with reference to the priority of the interrupt service routine currently being executed. Start of Interrupt Routine The interrupt cycle performs the following steps: – All maskable interrupt requests are disabled by clearing CICR.IEN. – CPL is saved in the special NICR stack to hold the priority level of the suspended routine. – Priority level of the acknowledged routine is stored in CPL, so that the next request priority will be compared with the one of the routine currently being serviced. – The PC low byte is pushed onto system stack. – The PC high byte is pushed onto system stack. – If ENCSR is set, CSR is pushed onto system stack. – The Flag register is pushed onto system stack. – The PC is loaded with the 16-bit vector stored in the Vector Table, pointed to by the IVR. – If ENCSR is set, CSR is loaded with ISR contents; otherwise ISR is used in place of CSR until iret instruction. Figure 24. Simple Example of a Sequence of Interrupt Requests with: - Nested mode - IEN unchanged by the interrupt routines Priority Level of Interrupt Request INTE RRUPT 0 HAS PRIORITY LEVEL 0 INTE RRUPT 2 HAS PRIORITY LEVEL 2 1 INT0 2 INT 2 CPL=2 3 INTE RRUPT 4 HAS PRIORITY LEVEL 4 CPL6 > CPL3: INT6 pending INT2 INT3 INT4 5 ei INT 5 CPL=5 6 INT5 MAIN CPL is set to 7 52/224 CPL2 < CPL4: Serviced next INTE RRUPT 5 HAS PRIORITY LEVEL 5 INTE RRUPT 6 HAS PRIORITY LEVEL 6 INT 2 CPL=2 INT6 INT 3 CPL=3 4 7 INTE RRUPT 3 HAS PRIORITY LEVEL 3 INT 0 CPL=0 0 INT2 INT 4 CPL=4 INT 6 CPL=6 MAIN CPL=7 ST92163 - INTERRUPTS ARBITRATION MODES (Cont’d) End of Interrupt Routine The iret Interrupt Return instruction executes the following steps: – The Flag register is popped from system stack. – If ENCSR is set, CSR is popped from system stack. – The PC high byte is popped from system stack. – The PC low byte is popped from system stack. – All unmasked Interrupts are enabled by setting the CICR.IEN bit. – The priority level of the interrupted routine is popped from the special register (NICR) and copied into CPL. – If ENCSR is reset, CSR is used instead of ISR, unless the program returns to another nested routine. The suspended routine thus resumes at the interrupted instruction. Figure 24 contains a simple example, showing that if the ei instruction is not used in the interrupt service routines, nested and concurrent modes are equivalent. Figure 25 contains a more complex example showing how nested mode allows nested interrupt processing (enabled inside the interrupt service routines using the ei instruction) according to their priority level. Figure 25. Complex Example of a Sequence of Interrupt Requests with: - Nested mode - IEN set to 1 during the interrupt routine execution Priority Level of Interrupt Request 0 INTERRUPT 0 HAS PRIORI TY LEVEL 0 INTERRUPT 2 HAS PRIORI TY LEVEL 2 INT 0 CPL=0 1 INT0 2 INT 2 CPL=2 3 INT 5 CPL=5 ei ei INT5 MAIN CPL is set to 7 INTERRUPT 5 HAS PRIORI TY LEVEL 5 INTERRUPT 6 HAS PRIORI TY LEVEL 6 CPL6 > CPL3: INT6 pending INT 2 CPL=2 INT 2 CPL=2 INT6 INT 3 CPL=3 INT2 ei INT2 INT3 INT4 6 7 INTERRUPT 4 HAS PRIORI TY LEVEL 4 ei 4 5 INTERRUPT 3 HAS PRIORI TY LEVEL 3 CPL2 < CPL4: Serviced just after ei INT 4 CPL=4 ei INT 4 CPL=4 INT 5 CPL=5 INT 6 CPL=6 MAIN CPL=7 53/224 ST92163 - INTERRUPTS 3.6 EXTERNAL INTERRUPTS The standard ST9 core contains 8 external interrupts sources grouped into four pairs. INT7 is connected to 8 different I/O pins of Port 3. Once these pins are programmed as alternate function they are able to generate an interrupt. Figure 26. Priority Level Examples PL2 D PL1D PL2 C PL1 C PL2B PL1B PL2A PL1A 1 Channel INT7 INT6 INTD1 INTD0 INT5 INT4 INTC1 INTC0 INT3 INT2 INTB1 INTB0 INT1 INT0 INTA1 INTA0 INT0 .. 6 have a trigger control bit TEA0,..TED1 (R242,EITR.0,..,7 Page 0) to select triggering on the rising or falling edge of the external pin. If the Trigger control bit is set to “1”, the corresponding pending bit IPA0,..,IPD1 (R243, EIPR.0,..,6 Page 0) is set on the input pin rising edge, if it is cleared, the pending bit is set on the falling edge of the input pin. Each source can be individually masked through the corresponding control bit IMA0,..,IMD1 (EIMR.6,..,0). See Figure 27. INT7 is falling edge sensitive only, bit EIMR.7 must always be cleared. The priority level of the external interrupt sources can be programmed among the eight priority levels with the control register EIPLR (R245). The priority level of each pair is software defined using the bits PRL2, PRL1. For each pair, the even channel (A0,B0,C0,D0) of the group has the even priority level and the odd channel (A1,B1,C1,D1) has the odd (lower) priority level. 54/224 0 0 1 SOURCE PRIORITY Table 11. External Interrupt Channel Grouping External Interrupt 0 0 0 1 EIPLR SOURCE PRIORITY INT.D0: 100=4 INT.A0: 010=2 INT.D1: 101=5 INT.A1: 011=3 INT.C0: 000=0 INT.B0: 100=4 INT.C1: 001=1 INT.B1: 101=5 VR000151 n Figure 26 shows an example of priority levels. Figure 27 gives an overview of the External interrupt control bits and vectors. – The source of the interrupt channel A0 can be selected between the external pin INT0 (when IA0S = “1”, the reset value) or the On-chip Timer/ Watchdog peripheral (when IA0S = “0”). – The source of the interrupt channel A1 can be selected between the external pin INT1 (when AD-INT=“0”) or the on-chip ADC peripheral (when AD-INT=“1”, the reset value). – The source of the interrupt channel D0 can be selected between the external pin INT6 (when INT_SEL = “0”) or the on-chip RCCU. Warning: When using channels shared by both external interrupts and peripherals, special care must be taken to configure their control registers for both peripherals and interrupts. Table 12. Multiplexed Interrupt Sources Channel Internal Interrupt Source External Interrupt Source INTA0 Timer/Watchdog INT0 INTA1 ADC INT1 INTD0 RCCU INT6 ST92163 - INTERRUPTS EXTERNAL INTERRUPTS (Cont’d) Figure 27. External Interrupts Control Bits and Vectors Watchdog/Timer IA0S End of count TEA0 INT 0 pin “0” V7 V6 V5 V4 0 0 VECTOR Priority level PL2A PL1A 0 “1” Mask bit IMA0 0 X INTA0 request Pending bit IPA0 * AD-INT TEA1 ADC V7 V6 V5 V4 0 0 VECTOR Priority level PL2A PL1A 1 “1” INT 1 pin Mask bit IMA1 “0” * 1 X INT A1 request Pending bit IPA1 TEB0 V7 V6 V5 V4 0 1 VECTO R PL2B PL1B 0 Priority level INT 2 pin Mask bit IMB0 0 X INT B0 request Pending bit IPB0 TEB1 V7 V6 V5 V4 0 1 VECTOR PL2B PL1B 1 Priority level INT 3 pin Mask bit IMB1 1 X INT B1 request Pending bit IPB1 TEC0 V7 V6 V5 V4 1 0 VECTO R PL2C PL1C 0 Priority level INT 4 pin Mask bit IMC0 0 X INT C0 request Pending bit IPC0 TEC1 V7 V6 V5 V4 1 0 VECTOR PL2C PL1C 1 Priority level INT 5 pin Mask bit IMC1 1 X INT C1 request Pending bit IPC1 INT_SEL TED0 RCCU INT 6 pin TED1 INT 7 pins P3.[7:0] WKUP [13:0] pins “1” V7 V6 V5 V4 1 1 VECTOR Priority level PL2D PL1D 0 “0” Mask bit IMD0 INT D0 request Pending bit IPD0 * ID1S “0” “1” Wake-up Controller 0 X V7 V6 V5 V4 1 1 VECTOR PL2D PL1D 1 Priority level Mask bit IMD1 1 X INT D1 request Pending bit IPD1 * WKUP14 pins P1.[7:0] * Shared channels, see warning 55/224 ST92163 - INTERRUPTS 3.7 MANAGEMENT OF WAKE-UP LINES AND EXTERNAL INTERRUPT LINES In the ST92163, fifteen Wake-up lines (WKUP[14:0]) are available on external pins. The WKUP[15] line is internally connected to the USB interfaceline. Figure 28 shows the connections of the External Interrupt Lines INT7[7:0] and the Wake-up/Interrupt Lines managed through the WUIMU on the INTD1 interrupt channel. Figure 28. Wake-Up Lines and External Interrupt Lines Management USBISTR Register ESUSP WUIMU WKUP[7:0] WKUP[13:8] WKUP15 WKUP14 USB INTERFACE INT7[7:0] Trigger Registers Pending Registers WKUP14[7:0] Mask Registers SW Setting 1 WUCTRL Register 0 ID1S bit STOP (to RCCU) 56/224 External Interrupt INTD1 (to CPU) ST92163 - INTERRUPTS 3.8 TOP LEVEL INTERRUPT The Top Level Interrupt channel can be assigned either to the external pin NMI or to the Timer/ Watchdog according to the status of the control bit EIVR.TLIS (R246.2, Page 0). If this bit is high (the reset condition) the source is the external pin NMI. If it is low, the source is the Timer/ Watchdog End Of Count. When the source is the NMI external pin, the control bit EIVR.TLTEV (R246.3; Page 0) selects between the rising (if set) or falling (if reset) edge generating the interrupt request. When the selected event occurs, the CICR.TLIP bit (R230.6) is set. Depending on the mask situation, a Top Level Interrupt request may be generated. Two kinds of masks are available, a Maskable mask and a Non-Maskable mask. The first mask is the CICR.TLI bit (R230.5): it can be set or cleared to enable or disable respectively the Top Level Interrupt request. If it is enabled, the global Enable Interrupt bit, CICR.IEN (R230.4; Page 0) must also be enabled in order to allow a Top Level Request. The second mask NICR.TLNM (R247.7; Page 0) is a set-only mask. Once set, it enables the Top Level Interrupt request independently of the value of CICR.IEN and it cannot be cleared by the program. Only the processor RESET cycle can clear this bit. This does not prevent the user from ignoring some sources due to a change in TLIS. The Top Level Interrupt Service Routine cannot be interrupted by any other interrupt or DMA request, in any arbitration mode, not even by a subsequent Top Level Interrupt request. Warning. The interrupt machine cycle of the Top Level Interrupt does not clear the CICR.IEN bit, and the corresponding iret does not set it. 3.9 ON-CHIP PERIPHERAL INTERRUPTS The general structure of the peripheral interrupt unit is described here, however each on-chip peripheral has its own specific interrupt unit containing one or more interrupt channels, or DMA channels. Please refer to the specific peripheral chapter for the description of its interrupt features and control registers. The on-chip peripheral interrupt channels provide the following control bits: – Interrupt Pending bit (IP). Set by hardware when the Trigger Event occurs. Can be set/ cleared by software to generate/cancel pending interrupts and give the status for Interrupt polling. – Interrupt Mask bit (IM). If IM = “0”, no interrupt request is generated. If IM =“1” an interrupt request is generated whenever IP = “1” and CICR.IEN = “1”. – Priority Level (PRL, 3 bits). These bits define the current priority level, PRL=0: the highest priority, PRL=7: the lowest priority (the interrupt cannot be acknowledged) – Interrupt Vector Register (IVR, up to 7 bits). The IVR points to the vector table which itself contains the interrupt routine start address. Figure 29. Top Level Interrupt Structure WATCHDOG ENABLE WDEN CORE RESET TLIP WATCHDOG TIMER END OF COUNT PENDING MUX MASK TOP LEVEL INTERRUPT REQUEST OR NMI TLIS TLTEV TLNM TLI IEN VA00294 57/224 ST92163 - INTERRUPTS 3.10 INTERRUPT RESPONSE TIME The interrupt arbitration protocol functions completely asynchronously from instruction flow, and requires 6 CPUCLK cycles to resolve the request’s priority. Requests are sampled every 5 CPUCLK cycles. If the interrupt request comes from an external pin, the trigger event must occur a minimum of one INTCLK cycle before the sampling time. When an arbitration results in an interrupt request being generated, the interrupt logic checks if the current instruction (which could be at any stage of execution) can be safely aborted; if this is the case, instruction execution is terminated immediately and the interrupt request is serviced; if not, the CPU waits until the current instruction is terminated and then services the request. Instruction execution can normally be aborted provided no write operation has been performed. For an interrupt deriving from an external interrupt channel, the response time between a user event and the start of the interrupt service routine can range from a minimum of 26 clock cycles to a maximum of 48 clock cycles. 58/224 For a non-maskable Top Level interrupt, the response time between a user event and the start of the interrupt service routine can range from a minimum of 22 clock cycles to a maximum of 48 clock cycles. In order to guarantee edge detection, input signals must be kept low/high for a minimum of one INTCLK cycle. An interrupt machine cycle requires a basic 18 internal clock cycles (CPUCLK), to which must be added a further 2 clock cycles if the stack is in the Register File. 2 more clock cycles must further be added if the CSR is pushed (ENCSR =1). The interrupt machine cycle duration forms part of the two examples of interrupt response time previously quoted; it includes the time required to push values on the stack, as well as interrupt vector handling. In Wait for Interrupt mode, a further cycle is required as wake-up delay. ST92163 - INTERRUPTS 3.11 INTERRUPT REGISTERS CENTRAL INTERRUPT CONTROL REGISTER (CICR) R230 - Read/Write Register Page: System Reset value: 1000 0111 (87h) 7 GCEN TLIP ple, if the state of IEN is not known in advance, and its value must be restored from a previous push of CICR on the stack, use the sequence DI; POP CICR to make sure that no interrupts are being arbitrated when CICR is modified. 0 TLI IEN IAM CPL2 CPL1 CPL0 Bit 7 = GCEN: Global Counter Enable. This bit enables the 16-bit Multifunction Timer peripheral. 0: MFT disabled 1: MFT enabled Bit 6 = TLIP: Top Level Interrupt Pending. This bit is set by hardware when Top Level Interrupt (TLI) trigger event occurs. It is cleared by hardware when a TLI is acknowledged. It can also be set by software to implement a software TLI. 0: No TLI pending 1: TLI pending Bit 5 = TLI: Top Level Interrupt. This bit is set and cleared by software. 0: Generate a Top Level Interrupt only if TLNM=1 1: Generate a Top Level Interrupt request when the IEN and TLIP bits=1. Bit 4 = IEN: Interrupt Enable. This bit is cleared by the interrupt machine cycle (except for a TLI). It is set by the iret instruction (except for a return from TLI). It is set by the EI instruction. It is cleared by the DI instruction. 0: Maskable interrupts disabled 1: Maskable Interrupts enabled Note: The IEN bit can also be changed by software using any instruction that operates on register CICR, however in this case, take care to avoid spurious interrupts, since IEN cannot be cleared in the middle of an interrupt arbitration. Only modify the IEN bit when interrupts are disabled or when no peripheral can generate interrupts. For exam- Bit 3 = IAM: Interrupt Arbitration Mode. This bit is set and cleared by software. 0: Concurrent Mode 1: Nested Mode Bit 2:0 = CPL[2:0]: Current Priority Level. These bits define the Current Priority Level. CPL=0 is the highest priority. CPL=7 is the lowest priority. These bits may be modified directly by the interrupt hardware when Nested Interrupt Mode is used. EXTERNAL INTERRUPT TRIGGER REGISTER (EITR) R242 - Read/Write Register Page: 0 Reset value: 0000 0000 (00h) 7 0 TED1 TED0 TEC1 TEC0 TEB1 TEB0 TEA1 TEA0 Bit 7 = TED1: INTD1 Trigger Event Must always stay cleared Bit 6 = TED0: INTD0 Trigger Event Bit 5 = TEC1: INTC1 Trigger Event Bit 4 = TEC0: INTC0 Trigger Event Bit 3 = TEB1: INTB1 Trigger Event Bit 2 = TEB0: INTB0 Trigger Event Bit 1 = TEA1: INTA1 Trigger Event Bit 0 = TEA0: INTA0 Trigger Event These bits are set and cleared by software. 0: Select falling edge as interrupt trigger event 1: Select rising edge as interrupt trigger event 59/224 ST92163 - INTERRUPTS INTERRUPT REGISTERS (Cont’d) EXTERNAL INTERRUPT PENDING REGISTER (EIPR) R243 - Read/Write Register Page: 0 Reset value: 0000 0000 (00h) 7 IPD1 IPD0 0 IPC1 IPC0 IPB1 IPB0 IPA1 IPA0 Bit 7 = IPD1: INTD1 Interrupt Pending bit Bit 6 = IPD0: INTD0 Interrupt Pending bit Bit 5 = IPC1: INTC1 Interrupt Pending bit Bit 4 = IPC0: INTC0 Interrupt Pending bit Bit 3 = IPB1: INTB1 Interrupt Pending bit Bit 2 = IPB0: INTB0 Interrupt Pending bit Bit 1 = IPA1: INTA1 Interrupt Pending bit Bit 0 = IPA0: INTA0 Interrupt Pending bit These bits are set by hardware on occurrence of a trigger event (as specified in the EITR register) and are cleared by hardware on interrupt acknowledge. They can also be set by software to implement a software interrupt. 0: No interrupt pending 1: Interrupt pending EXTERNAL INTERRUPT MASK-BIT REGISTER (EIMR) R244 - Read/Write Register Page: 0 Reset value: 0000 0000 (00h) 7 Bit 3 = IMB1: INTB1 Interrupt Mask Bit 2 = IMB0: INTB0 Interrupt Mask Bit 1 = IMA1: INTA1 Interrupt Mask Bit 0 = IMA0: INTA0 Interrupt Mask These bits are set and cleared by software. 0: Interrupt masked 1: Interrupt not masked (an interrupt is generated if the IPxx and IEN bits = 1) EXTERNAL INTERRUPT PRIORITY REGISTER (EIPLR) R245 - Read/Write Register Page: 0 Reset value: 1111 1111 (FFh) 7 0 PL2D PL1D PL2C PL1C PL2B PL1B PL2A PL1A Bit 7:6 = PL2D, PL1D: INTD0, D1 Priority Level. Bit 5:4 = PL2C, PL1C: INTC0, C1 Priority Level. Bit 3:2 = PL2B, PL1B: INTB0, B1 Priority Level. Bit 1:0 = PL2A, PL1A: INTA0, A1 Priority Level. These bits are set and cleared by software. The priority is a three-bit value. The LSB is fixed by hardware at 0 for Channels A0, B0, C0 and D0 and at 1 for Channels A1, B1, C1 and D1. PL2x PL1x 0 0 0 1 1 0 1 1 0 IMD1 IMD0 IMC1 IMC0 IMB1 IMB0 IMA1 IMA0 Bit 7 = IMD1: INTD1 Interrupt Mask Bit 6 = IMD0: INTD0 Interrupt Mask Bit 5 = IMC1: INTC1 Interrupt Mask Bit 4 = IMC0: INTC0 Interrupt Mask 60/224 LEVEL Hardware bit 0 1 0 1 0 1 0 1 Priority 0 (Highest) 1 2 3 4 5 6 7 (Lowest) ST92163 - INTERRUPTS INTERRUPT REGISTERS (Cont’d) EXTERNAL INTERRUPT VECTOR REGISTER (EIVR) R246 - Read/Write Register Page: 0 Reset value: xxxx 0110 (x6h) 7 V7 0: WAITN pin disabled 1: WAITN pin enabled (to stretch the external memory access cycle). Note: For more details on Wait mode refer to the section describing the WAITN pin in the External Memory Chapter. 0 V6 V5 V4 TLTEV TLIS IAOS EWEN Bit 7:4 = V[7:4]: Most significant nibble of External Interrupt Vector. These bits are not initialized by reset. For a representation of how the full vector is generated from V[7:4] and the selected external interrupt channel, refer to Figure 27. Bit 3 = TLTEV: Top Level Trigger Event bit. This bit is set and cleared by software. 0: Select falling edge as NMI trigger event 1: Select rising edge as NMI trigger event Bit 2 = TLIS: Top Level Input Selection. This bit is set and cleared by software. 0: Watchdog End of Count is TL interrupt source 1: NMI is TL interrupt source Bit 1 = IA0S: Interrupt Channel A0 Selection. This bit is set and cleared by software. 0: Watchdog End of Count is INTA0 source 1: External Interrupt pin is INTA0 source NESTED INTERRUPT CONTROL (NICR) R247 - Read/Write Register Page: 0 Reset value: 0000 0000 (00h) 7 TLNM HL6 0 HL5 HL4 HL3 HL2 HL1 HL0 Bit 7 = TLNM: Top Level Not Maskable. This bit is set by software and cleared only by a hardware reset. 0: Top Level Interrupt Maskable. A top level request is generated if the IEN, TLI and TLIP bits =1 1: Top Level Interrupt Not Maskable. A top level request is generated if the TLIP bit =1 Bit 6:0 = HL[6:0]: Hold Level x These bits are set by hardware when, in Nested Mode, an interrupt service routine at level x is interrupted from a request with higher priority (other than the Top Level interrupt request). They are cleared by hardware at the iret execution when the routine at level x is recovered. Bit 0 = EWEN: External Wait Enable. This bit is set and cleared by software. 61/224 ST92163 - INTERRUPTS INTERRUPT REGISTERS (Cont’d) EXTERNAL MEMORY REGISTER 2 (EMR2) R246 - Read/Write Register Page: 21 Reset value: 0000 1111 (0Fh) 7 0 0 ENCSR 0 0 1 1 1 1 Bit 7, 5:0 = Reserved, keep in reset state. Refer to the external Memory Interface Chapter. Bit 6 = ENCSR: Enable Code Segment Register. This bit is set and cleared by software. It affects the ST9 CPU behaviour whenever an interrupt request is issued. 0: The CPU works in original ST9 compatibility mode. For the duration of the interrupt service routine, ISR is used instead of CSR, and the interrupt stack frame is identical to that of the original ST9: only the PC and Flags are pushed. This avoids saving the CSR on the stack in the event of an interrupt, thus ensuring a faster in- 62/224 terrupt response time. The drawback is that it is not possible for an interrupt service routine to perform inter-segment calls or jumps: these instructions would update the CSR, which, in this case, is not used (ISR is used instead). The code segment size for all interrupt service routines is thus limited to 64K bytes. 1: ISR is only used to point to the interrupt vector table and to initialize the CSR at the beginning of the interrupt service routine: the old CSR is pushed onto the stack together with the PC and flags, and CSR is then loaded with the contents of ISR. In this case, iret will also restore the CSR from the stack. This approach allows interrupt service routines to access the entire 4 Mbytes of address space; the drawback is that the interrupt response time is slightly increased, because of the need to also save the CSR on the stack. Full compatibility with the original ST9 is lost in this case, because the interrupt stack frame is different; this difference, however, should not affect the vast majority of programs. ST92163 - INTERRUPTS 3.12 WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (WUIMU) 3.12.1 Introduction The Wake-up/Interrupt Management Unit extends the number of external interrupt lines from 8 to 23 (depending on the number of external interrupt lines mapped on external pins of the device). It allows the source of the INTD1 external interrupt channel to be selected between the INT7 pin and up to 16 additional external Wake-up/interrupt pins. These 16 WKUP pins can be programmed as external interrupt lines or as wake-up lines, able to exit the microcontroller from low power mode (STOP mode) (see Figure 30). Figure 30. Wake-Up Lines / Interrupt Management WKUP[ 7:0] STOP 3.12.2 Main Features ■ Supports up to 16 additional external wake-up or interrupt lines ■ Wake-Up lines can be used to wake-up the ST9 from STOP mode. ■ Programmable selection of wake-up or interrupt ■ Programmable wake-up trigger edge polarity ■ All Wake-Up Lines maskable Note: The number of available pins is device dependent. Refer to the device pinout description. Unit Block Diagram INT7 WKUP[15:8] WUTRH WUTRL TRIGGERING LEVEL REGISTERS WUPRH WUPRL PENDING REQUEST REGISTERS WUMRH WUMRL MASK REGISTERS Set Reset SW SETT ING WUCTR L STOP ID1S WKUP-INT Note: Reset Signal on stop bit is stronger than the set signal 1 0 TO CPU INTD1 - External Interrupt Channel TO RCCU - Stop Mode Control 63/224 ST92163 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) only to restart code execution. 3.12.3 Functional Description 5. Set the WKUP-INT bit in the WUCTRL register 3.12.3.1 Interrupt Mode to select Wake-up Mode. To configure the 16 wake-up lines as interrupt 6. Set the ID1S bit in the WUCTRL register to dissources, use the following procedure: able the INT7 external interrupt source and 1. Configure the mask bits of the 16 wake-up lines enable the 16 wake-up lines as external inter(WUMRL, WUMRH). rupt source lines. This is not mandatory if the 2. Configure the triggering edge registers of the wake-up event does not require an interrupt wake-up lines (WUTRL, WUTRH). response. 3. Set bit 7 of EIMR (R244 Page 0) and EITR 7. Write the sequence 1,0,1 to the STOP bit of the (R242 Page 0) registers of the CPU: so an WUCTRL register with three consecutive write interrupt coming from one of the 16 lines can be operations. This is the STOP bit setting correctly acknowledged. sequence. 4. Reset the WKUP-INT bit in the WUCTRL regisTo detect if STOP Mode was entered or not, imter to disable Wake-up Mode. mediately after the STOP bit setting sequence, poll the RCCU EX_STP bit (R242.7, Page 55) and 5. Set the ID1S bit in the WUCTRL register to disthe STOP bit itself. able the INT7 external interrupt source and enable the 16 wake-up lines as external interrupt source lines. 3.12.3.3 STOP Mode Entry Conditions To return to standard mode (INT7 external interAssuming the ST9 is in Run mode: during the rupt source enabled and 16 wake-up lines disaSTOP bit setting sequence the following cases bled) it is sufficient to reset the ID1S bit. may occur: Case 1: Wrong STOP bit setting sequence 3.12.3.2 Wake-up Mode Selection This can happen if an Interrupt/DMA request is acTo configure the 16 lines as wake-up sources, use knowledged during the STOP bit setting sethe following procedure: quence. In this case polling the STOP and EX_STP bits will give: 1. Configure the mask bits of the 16 wake-up lines (WUMRL, WUMRH). STOP = 0, EX_STP = 0 2. Configure the triggering edge registers of the This means that the ST9 did not enter STOP mode wake-up lines (WUTRL, WUTRH). due to a bad STOP bit setting sequence: the user must retry the sequence. 3. Set, as for Interrupt Mode selection, bit 7 of EIMR and EITR registers only if an interrupt Case 2: Correct STOP bit setting sequence routine is to be executed after a wake-up event. In this case the ST9 enters STOP mode. Otherwise, if the wake-up event only restarts To exit STOP mode, a wake-up interrupt must be the execution of the code from where it was acknowledged. That implies: stopped, the INTD1 interrupt channel must be masked or the external source must be STOP = 0, EX_STP = 1 selected by resetting the ID1S bit. This means that the ST9 entered and exited STOP 4. Since the RCCU can generate an interrupt mode due to an external wake-up line event. request when exiting from STOP mode, take care to mask it even if the wake-up event is 64/224 ST92163 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) Case 3: A wake-up event on the external wakeIf the MCU really exits from STOP Mode, the up lines occurs during the STOP bit setting seRCCU EX_STP bit is still set and must be reset by quence software. Otherwise, if an Interrupt/DMA request was acknowledged during the STOP bit setting seThere are two possible cases: quence, the RCCU EX_STP bit is reset. This 1. Interrupt requests to the CPU are disabled: in means that the MCU has filtered the STOP Mode this case the ST9 will not enter STOP mode, no entry request. interrupt service routine will be executed and The WKUP-INT bit can be used by an interrupt the program execution continues from the routine to detect and to distinguish events coming instruction following the STOP bit setting from Interrupt Mode or from Wake-up Mode, allowsequence. The status of STOP and EX_STP ing the code to execute different procedures. bits will be again: To exit STOP mode, it is sufficient that one of the STOP = 0, EX_STP = 0 16 wake-up lines (not masked) generates an The application can determine why the ST9 did event: the clock restarts after the delay needed for not enter STOP mode by polling the pending the oscillator to restart. bits of the external lines (at least one must be at 1). Note: After exiting from STOP Mode, the software 2. Interrupt requests to CPU are enabled: in this can successfully reset the pending bits (edge sencase the ST9 will not enter STOP mode and the sitive), even though the corresponding wake-up interrupt service routine will be executed. The line is still active (high or low, depending on the status of STOP and EX_STP bits will be again: Trigger Event register programming); the user STOP = 0, EX_STP = 0 must poll the external pin status to detect and distinguish a short event from a long one (for example The interrupt service routine can determine why keyboard input with keystrokes of varying length). the ST9 did not enter STOP mode by polling the pending bits of the external lines (at least one must be at 1). 65/224 ST92163 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) 9. Poll the wake-up pending bits to determine 3.12.4 Programming Considerations which wake-up line caused the exit from STOP The following paragraphs give some guidelines for mode. designing an application program. 10.Clear the wake-up pending bit that was set. 3.12.4.1 Procedure for Entering/Exiting STOP 3.12.4.2 Simultaneous Setting of Pending Bits mode It is possible that several simultaneous events set 1. Program the polarity of the trigger event of different pending bits. In order to accept subseexternal wake-up lines by writing registers quent events on external wake-up/interrupt lines, it WUTRH and WUTRL. is necessary to clear at least one pending bit: this 2. Check that at least one mask bit (registers operation allows a rising edge to be generated on WUMRH, WUMRL) is equal to 1 (so at least the INTD1 line (if there is at least one more pendone external wake-up line is not masked). ing bit set and not masked) and so to set EIPR.7 3. Reset at least the unmasked pending bits: this bit again. A further interrupt on channel INTD1 will allows a rising edge to be generated on the be serviced depending on the status of bit EIMR.7. INTD1 channel when the trigger event occurs Two possible situations may arise: (an interrupt on channel INTD1 is recognized 1. The user chooses to reset all pending bits: no when a rising edge occurs). further interrupt requests will be generated on 4. Select the interrupt source of the INTD1 chanchannel INTD1. In this case the user has to: nel (see description of ID1S bit in the WUCTRL – Reset EIMR.7 bit (to avoid generating a spuriregister) and set the WKUP-INT bit. ous interrupt request during the next reset op5. To generate an interrupt on channel INTD1, bits eration on the WUPRH register) EITR.1 (R242.7, Page 0) and EIMR.1 (R244.7, – Reset WUPRH register using a read-modifyPage 0) must be set and bit EIPR.7 must be write instruction (AND, BRES, BAND) reset. Bits 7 and 6 of register R245, Page 0 – Clear the EIPR.7 bit must be written with the desired priority level for interrupt channel INTD1. – Reset the WUPRL register using a read-modify-write instruction (AND, BRES, BAND) 6. Reset the STOP bit in register WUCTRL and the EX_STP bit in the CLK_FLAG register 2. The user chooses to keep at least one pending (R242.7, Page 55). Refer to the RCCU chapter. bit active: at least one additional interrupt request will be generated on the INTD1 chan7. To enter STOP mode, write the sequence 1, 0, nel. In this case the user has to reset the 1 to the STOP bit in the WUCTRL register with desired pending bits with a read-modify-write three consecutive write operations. instruction (AND, BRES, BAND). This operation 8. The code to be executed just after the STOP will generate a rising edge on the INTD1 chansequence must check the status of the STOP nel and the EIPR.7 bit will be set again. An and RCCU EX_STP bits to determine if the ST9 interrupt on the INTD1 channel will be serviced entered STOP mode or not (See “Wake-up depending on the status of EIMR.7 bit. Mode Selection” on page 64. for details). If the ST9 did not enter in STOP mode it is necessary to reloop the procedure from the beginning, otherwise the procedure continues from next point. 66/224 ST92163 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) 3.12.5 Register Description sequence. If Interrupt or DMA requests (which always perform register write operations) are acWAKE-UP CONTROL REGISTER (WUCTRL) knowledged during the sequence, the ST9 will not R249 - Read/Write enter STOP mode: the user must re-enter the seRegister Page: 57 quence to set the STOP bit. Reset Value: 0000 0000 (00h) WARNING: Whenever a STOP request is issued 7 0 to the MCU, a few clock cycles are needed to enter STOP mode (see RCCU chapter for further deSTOP ID1S WKUP- INT tails). Hence the execution of the instruction following the STOP bit setting sequence might start before entering STOP mode: if such instruction Bit 2 = STOP: Stop bit. performs a register write operation, the ST9 will To enter STOP Mode, write the sequence 1,0,1 to not enter in STOP mode. In order to avoid to exethis bit with three consecutive write operations. cute register write instructions after a correct When a correct sequence is recognized, the STOP bit setting sequence and before entering STOP bit is set and the RCCU puts the MCU in the STOP mode, it is mandatory to execute 3 NOP STOP Mode. The software sequence succeeds instructions after the STOP bit setting sequence. only if the following conditions are true: – The WKUP-INT bit is 1, – All unmasked pending bits are reset, – At least one mask bit is equal to 1 (at least one external wake-up line is not masked). Otherwise the MCU cannot enter STOP mode, the program code continues executing and the STOP bit remains cleared. The bit is reset by hardware if, while the MCU is in STOP mode, a wake-up interrupt comes from any of the unmasked wake-up lines. The STOP bit is at 1 in the two following cases (See “Wake-up Mode Selection” on page 64. for details): – After the first write instruction of the sequence (a 1 is written to the STOP bit) – At the end of a successful sequence (i.e. after the third write instruction of the sequence) Note: The STOP request generated by the WUIMU (that allows the ST9 to enter STOP mode) is ORed with the external STOP pin (active low). This means that if the external STOP pin is forced low, the ST9 will enter STOP mode independently of the status of the STOP bit. WARNING: Writing the sequence 1,0,1 to the STOP bit will enter STOP mode only if no other register write instructions are executed during the Bit 1 = ID1S: Interrupt Channel INTD1 Source. This bit is set and cleared by software. 0: INT7 external interrupt source selected, excluding wake-up line interrupt requests 1: The 16 external wake-up lines enabled as interrupt sources, replacing the INT7 external pin function WARNING: To avoid spurious interrupt requests on the INTD1 channel due to changing the interrupt source, do the following before modifying the ID1S bit: 1. Mask the INTD1 interrupt channel (bit 7 of register EIMR - R244, Page 0 - reset to 0). 2. Program the ID1S bit as needed. 3. Clear the IPD1 interrupt pending bit (bit 7 of register EIPR - R243, Page 0). 4. Remove the mask on INTD1 (bit EIMR.7=1). Bit 0 = WKUP-INT: Wakeup Interrupt. This bit is set and cleared by software. 0: The 16 external wakeup lines can be used to generate interrupt requests 1: The 16 external wake-up lines to work as wakeup sources for exiting from STOP mode 67/224 ST92163 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) WAKE-UP MASK REGISTER LOW (WUMRL) WAKE-UP MASK REGISTER HIGH (WUMRH) R251 - Read/Write R250 - Read/Write Register Page: 57 Register Page: 57 Reset Value: 0000 0000 (00h) Reset Value: 0000 0000 (00h) 7 WUM15 WUM14 WUM13 WUM12 WUM11 WUM10 WUM9 0 7 WUM8 WUM7 Bit 7:0 = WUM[15:8]: Wake-Up Mask bits. If WUMx is set, an interrupt on channel INTD1 and/or a wake-up event (depending on ID1S and WKUP-INT bits) are generated if the corresponding WUPx pending bit is set. More precisely, if WUMx=1 and WUPx=1 then: – If ID1S=1 and WKUP-INT=1 then an interrupt on channel INTD1 and a wake-up event are generated. – If ID1S=1 and WKUP-INT=0 only an interrupt on channel INTD1 is generated. – If ID1S=0 and WKUP-INT=1 only a wake-up event is generated. – If ID1S=0 and WKUP-INT=0 neither interrupts on channel INTD1 nor wake-up events are generated. Interrupt requests on channel INTD1 may be generated only from external interrupt source INT7. If WUMx is reset, no wake-up events can be generated. Interrupt requests on channel INTD1 may be generated only from external interrupt source INT7 (resetting ID1S bit to 0). 68/224 0 WUM6 WUM5 WUM4 WUM3 WUM2 WUM1 WUM0 Bit 7:0 = WUM[7:0]: Wake-Up Mask bits. If WUMx is set, an interrupt on channel INTD1 and/or a wake-up event (depending on ID1S and WKUP-INT bits) are generated if the corresponding WUPx pending bit is set. More precisely, if WUMx=1 and WUPx=1 then: – If ID1S=1 and WKUP-INT=1 then an interrupt on channel INTD1 and a wake-up event are generated. – If ID1S=1 and WKUP-INT=0 only an interrupt on channel INTD1 is generated. – If ID1S=0 and WKUP-INT=1 only a wake-up event is generated. – If ID1S=0 and WKUP-INT=0 neither interrupts on channel INTD1 nor wake-up events are generated. Interrupt requests on channel INTD1 may be generated only from external interrupt source INT7. If WUMx is reset, no wake-up events can be generated. Interrupt requests on channel INTD1 may be generated only from external interrupt source INT7 (resetting ID1S bit to 0). ST92163 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) WAKE-UP TRIGGER REGISTER HIGH (WUTRH) WAKE-UP PENDING REGISTER HIGH R252 - Read/Write (WUPRH) Register Page: 57 R254 - Read/Write Reset Value: 0000 0000 (00h) Register Page: 57 7 0 Reset Value: 0000 0000 (00h) WUT15 WUT14 WUT13 WUT12 WUT11 WUT10 WUT9 WUT8 7 0 WUP15 WUP14 WUP13 WUP12 WUP11 WUP10 Bit 7:0 = WUT[15:8]: Wake-Up Trigger Polarity Bits These bits are set and cleared by software. 0: The corresponding WUPx pending bit will be set on the falling edge of the input wake-up line. 1: The corresponding WUPx pending bit will be set on the rising edge of the input wake-up line. WAKE-UP TRIGGER REGISTER LOW (WUTRL) R253 - Read/Write Register Page: 57 Reset Value: 0000 0000 (00h) 7 WUT7 0 WUT6 WUT5 WUT4 WUT3 WUT2 WUT1 WUT0 WUP9 Bit 7:0 = WUP[15:8]: Wake-Up Pending Bits These bits are set by hardware on occurrence of the trigger event on the corresponding wake-up line. They must be cleared by software. They can be set by software to implement a software interrupt. 0: No Wake-up Trigger event occurred 1: Wake-up Trigger event occured WAKE-UP PENDING REGISTER LOW (WUPRL) R255 - Read/Write Register Page: 57 Reset Value: 0000 0000 (00h) 7 Bit 7:0 = WUT[7:0]: Wake-Up Trigger Polarity Bits These bits are set and cleared by software. 0: The corresponding WUPx pending bit will be set on the falling edge of the input wake-up line. 1: The corresponding WUPx pending bit will be set on the rising edge of the input wake-up line. WARNING 1. As the external wake-up lines are edge triggered, no glitches must be generated on these lines. 2. If either a rising or a falling edge on the external wake-up lines occurs while writing the WUTRH or WUTRL registers, the pending bit will not be set. WUP8 WUP7 0 WUP6 WUP5 WUP4 WUP3 WUP2 WUP1 WUP0 Bit 7:0 = WUP[7:0]: Wake-Up Pending Bits These bits are set by hardware on occurrence of the trigger event on the corresponding wake-up line. They must be cleared by software. They can be set by software to implement a software interrupt. 0: No Wake-up Trigger event occurred 1: Wake-up Trigger event occured Note: To avoid losing a trigger event while clearing the pending bits, it is recommended to use read-modify-write instructions (AND, BRES, BAND) to clear them. 69/224 ST92163 - ON-CHIP DIRECT MEMORY ACCESS (DMA) 4 ON-CHIP DIRECT MEMORY ACCESS (DMA) 4.1 INTRODUCTION 4.2 DMA PRIORITY LEVELS The ST9 includes on-chip Direct Memory Access (DMA) in order to provide high-speed data transfer between peripherals and memory or Register File. Multi-channel DMA is fully supported by peripherals having their own controller and DMA channel(s). Each DMA channel transfers data to or from contiguous locations in the Register File, or in Memory. The maximum number of bytes that can be transferred per transaction by each DMA channel is 222 with the Register File, or 65536 with Memory. The DMA controller in the Peripheral uses an indirect addressing mechanism to DMA Pointers and Counter Registers stored in the Register File. This is the reason why the maximum number of transactions for the Register File is 222, since two Registers are allocated for the Pointer and Counter. Register pairs are used for memory pointers and counters in order to offer the full 65536 byte and count capability. The 8 priority levels used for interrupts are also used to prioritize the DMA requests, which are arbitrated in the same arbitration phase as interrupt requests. If the event occurrence requires a DMA transaction, this will take place at the end of the current instruction execution. When an interrupt and a DMA request occur simultaneously, on the same priority level, the DMA request is serviced before the interrupt. An interrupt priority request must be strictly higher than the CPL value in order to be acknowledged, whereas, for a DMA transaction request, it must be equal to or higher than the CPL value in order to be executed. Thus only DMA transaction requests can be acknowledged when the CPL=0. DMA requests do not modify the CPL value, since the DMA transaction is not interruptable. Figure 31. DMA Data Transfer REGISTER FILE REGISTER FILE OR MEMORY DF REGISTER FILE GROUP F PERIPHERAL PAGED REGISTERS COUNTER PERIPHERAL ADDRESS DATA 0 COUNTER VALUE TRANSFERRED DATA START ADDRESS VR001834 70/224 ST92163 - ON-CHIP DIRECT MEMORY ACCESS (DMA) 4.3 DMA TRANSACTIONS The purpose of an on-chip DMA channel is to transfer a block of data between a peripheral and the Register File, or Memory. Each DMA transfer consists of three operations: – A load from/to the peripheral data register to/ from a location of Register File (or Memory) addressed through the DMA Address Register (or Register pair) – A post-increment of the DMA Address Register (or Register pair) – A post-decrement of the DMA transaction counter, which contains the number of transactions that have still to be performed. If the DMA transaction is carried out between the peripheral and the Register File (Figure 32), one register is required to hold the DMA Address, and one to hold the DMA transaction counter. These two registers must be located in the Register File: the DMA Address Register in the even address register, and the DMA Transaction Counter in the next register (odd address). They are pointed to by the DMA Transaction Counter Pointer Register (DCPR), located in the peripheral’s paged registers. In order to select a DMA transaction with the Register File, the control bit DCPR.RM (bit 0 of DCPR) must be set. If the transaction is made between the peripheral and Memory, a register pair (16 bits) is required for the DMA Address and the DMA Transaction Counter (Figure 33). Thus, two register pairs must be located in the Register File. The DMA Transaction Counter is pointed to by the DMA Transaction Counter Pointer Register (DCPR), the DMA Address is pointed to by the DMA Address Pointer Register (DAPR),both DCPR and DAPR are located in the paged registers of the peripheral. Figure 32. DMA Between Register File and Peripheral IDCR IVR DAPR DCPR PAGED REGISTERS DATA F0h EFh DMA TRANSACTION PERIPHERAL PAGED REGISTERS DMA TABLE END OF BLOCK INTERRUPT SERVICE ROUTINE FFh 0100h SYSTEM ISR ADDRESS REGISTERS E0h DFh VECTOR TABLE 0000h MEMORY DATA ALREADY TRANSFERRED DMA COUNTER DMA ADDRESS REGISTER FILE 71/224 ST92163 - ON-CHIP DIRECT MEMORY ACCESS (DMA) DMA TRANSACTIONS (Cont’d) When selecting the DMA transaction with memory, bit DCPR.RM (bit 0 of DCPR) must be cleared. To select between using the ISR or the DMASR register to extend the address, (see Memory Management Unit chapter), the control bit DAPR.PS (bit 0 of DAPR) must be cleared or set respectively. The DMA transaction Counter must be initialized with the number of transactions to perform and will be decremented after each transaction. The DMA Address must be initialized with the starting address of the DMA table and is increased after each transaction. These two registers must be located between addresses 00h and DFh of the Register File. Once a DMA channel is initialized, a transfer can start. The direction of the transfer is automatically defined by the type of peripheral and programming mode. Once the DMA table is completed (the transaction counter reaches 0 value), an Interrupt request to the CPU is generated. When the Interrupt Pending (IP) bit is set by a hardware event (or by software), and the DMA Mask bit (DM) is set, a DMA request is generated. If the Priority Level of the DMA source is higher than, or equal to, the Current Priority Level (CPL), the DMA transfer is executed at the end of the current instruction. DMA transfers read/write data from/to the location pointed to by the DMA Address Register, the DMA Address register is incremented and the Transaction Counter Register is decremented. When the contents of the Transaction Counter are decremented to zero, the DMA Mask bit (DM) is cleared and an interrupt request is generated, according to the Interrupt Mask bit (End of Block interrupt). This End-of-Block interrupt request is taken into account, depending on the PRL value. WARNING. DMA requests are not acknowledged if the top level interrupt service is in progress. Figure 33. DMA Between Memory and Peripheral IDCR IVR DAPR DCPR DMA TRANSACTION FFh PAGED REGISTERS DATA PERIPHERAL PAGED REGISTERS F0h EFh DMA TABLE SYSTEM REGISTERS DATA ALREADY TRANSFERRED DMA TRANSACTION COUNTER END OF BLOCK INTERRUPT SERVICE ROUTINE E0h DFh 0100h DMA ADDRESS ISR ADDRESS 0000h REGISTER FILE n 72/224 MEMORY VECTOR TABLE ST92163 - ON-CHIP DIRECT MEMORY ACCESS (DMA) DMA TRANSACTIONS (Cont’d) 4.4 DMA CYCLE TIME The interrupt and DMA arbitration protocol functions completely asynchronously from instruction flow. Requests are sampled every 5 CPUCLK cycles. DMA transactions are executed if their priority allows it. A DMA transfer with the Register file requires 8 CPUCLK cycles. A DMA transfer with memory requires 16 CPUCLK cycles, plus any required wait states. 4.5 SWAP MODE An extra feature which may be found on the DMA channels of some peripherals (e.g. the MultiFunction Timer) is the Swap mode. This feature allows transfer from two DMA tables alternatively. All the DMA descriptors in the Register File are thus doubled. Two DMA transaction counters and two DMA address pointers allow the definition of two fully independent tables (they only have to belong to the same space, Register File or Memory). The DMA transaction is programmed to start on one of the two tables (say table 0) and, at the end of the block, the DMA controller automatically swaps to the other table (table 1) by pointing to the other DMA descriptors. In this case, the DMA mask (DM bit) control bit is not cleared, but the End Of Block interrupt request is generated to allow the optional updating of the first data table (table 0). Until the swap mode is disabled, the DMA controller will continue to swap between DMA Table 0 and DMA Table 1. n 73/224 ST92163 - ON-CHIP DIRECT MEMORY ACCESS (DMA) 4.6 DMA REGISTERS As each peripheral DMA channel has its own specific control registers, the following register list should be considered as a general example. The names and register bit allocations shown here may be different from those found in the peripheral chapters. DMA COUNTER POINTER REGISTER (DCPR) Read/Write Address set by Peripheral Reset value: undefined 7 C7 0 C6 C5 C4 C3 C2 C1 RM Bit 7:1 = C[7:1]: DMA Transaction Counter Pointer. Software should write the pointer to the DMA Transaction Counter in these bits. Bit 0 = RM: Register File/Memory Selector. This bit is set and cleared by software. 0: DMA transactions are with memory (see also DAPR.DP) 1: DMA transactions are with the Register File GENERIC EXTERNAL PERIPHERAL INTERRUPT AND DMA CONTROL (IDCR) Read/Write Address set by Peripheral Reset value: undefined 7 0 IP DM IM PRL2 PRL1 PRL0 Bit 3 = IM: End of block Interrupt Mask. This bit is set and cleared by software. 0: No End of block interrupt request is generated when IP is set 1: End of Block interrupt is generated when IP is set. DMA requests depend on the DM bit value as shown in the table below. DM IM Meaning A DMA request generated without End of Block 1 0 interrupt when IP=1 A DMA request generated with End of Block in1 1 terrupt when IP=1 No End of block interrupt or DMA request is 0 0 generated when IP=1 An End of block Interrupt is generated without 0 1 associated DMA request (not used) Bit 2:0 = PRL[2:0]: Source Priority Level. These bits are set and cleared by software. Refer to Section 4.2 for a description of priority levels. PRL2 PRL1 PRL0 Source Priority Level 0 0 0 0 Highest 0 0 0 1 1 1 1 0 1 1 0 0 1 1 1 0 1 0 1 0 1 1 2 3 4 5 6 7 Lowest DMA ADDRESS POINTER REGISTER (DAPR) Read/Write Address set by Peripheral Reset value: undefined 7 Bit 5 = IP: Interrupt Pending. This bit is set by hardware when the Trigger Event occurs. It is cleared by hardware when the request is acknowledged. It can be set/cleared by software in order to generate/cancel a pending request. 0: No interrupt pending 1: Interrupt pending Bit 4 = DM: DMA Request Mask. This bit is set and cleared by software. It is also cleared when the transaction counter reaches zero (unless SWAP mode is active). 0: No DMA request is generated when IP is set. 1: DMA request is generated when IP is set 74/224 A7 0 A6 A5 A4 A3 A2 A1 PS Bit 7:1 = A[7:1]: DMA Address Register(s) Pointer Software should write the pointer to the DMA Address Register(s) in these bits. Bit 0 = PS: Memory Segment Pointer Selector: This bit is set and cleared by software. It is only meaningful if DAPR.RM=0. 0: The ISR register is used to extend the address of data transferred by DMA (see MMU chapter). 1: The DMASR register is used to extend the address of data transferred by DMA (see MMU chapter). ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) 5 RESET AND CLOCK CONTROL UNIT (RCCU) 5.1 INTRODUCTION The Reset and Clock Control Unit (RCCU) comprises two distinct sections: – the Clock Control Unit, which generates and manages the internal clock signals. – the Reset/Stop Manager, which detects and flags Hardware, Software and Watchdog generated resets. On ST9 devices where the external Stop pin is available, this circuit also detects and manages the externally triggered Stop mode, during which all oscillators are frozen in order to achieve the lowest possible power consumption. 5.2 CLOCK CONTROL UNIT The Clock Control Unit generates the internal clocks for the CPU core (CPUCLK) and for the onchip peripherals (INTCLK). The Clock Control Unit may be driven by an external crystal circuit, connected to the OSCIN and OSCOUT pins, or by an external pulse generator, connected to OSCIN (see Figure 42 and Figure 44). Another clock source named CK_AF can be provided from the internal RC oscillator. 5.2.1 Clock Control Unit Overview As shown in Figure 34, a programmable divider can divide the CLOCK1 input clock signal by two. The resulting signal, CLOCK2, is the reference input clock to the programmable Phase Locked Loop frequency multiplier, which is capable of multiplying the clock frequency by a factor of 6, 8, 10 or 14; the multiplied clock is then divided by a pro- grammable divider, by a factor of 1 to 7. By this means, the ST9 can operate with cheaper, medium frequency (3-5 MHz) crystals, while still providing a high frequency internal clock for maximum system performance; the range of available multiplication and division factors allow a great number of operating clock frequencies to be derived from a single crystal frequency. The undivided PLL clock is also available for special purposes (high-speed peripheral). For low power operation, especially in Wait for Interrupt mode, the Clock Multiplier unit may be turned off, whereupon the output clock signal may be programmed as CLOCK2 divided by 16. For further power reduction, an internal RC oscillator with a frequency of 85KHZ (+/- 40%) is available to provide the CK_AF clock internally if the external clock source is not used. During the execution of a WFI in Low Power mode this clock is further divided by 16 to reduce power consumption (for the selection of this signal refer to the description the CK_AF clock source in the following sections). The internal system clock, INTCLK, is routed to all on-chip peripherals, as well as to the programmable Clock Prescaler Unit which generates the clock for the CPU core (CPUCLK). The Clock Prescaler is programmable and can slow the CPU clock by a factor of up to 8, allowing the programmer to reduce CPU processing speed, and thus power consumption, while maintaining a high speed clock to the peripherals. This is particularly useful when little actual processing is being done by the CPU and the peripherals are doing most of the work. Figure 34. Clock Control Unit Simplified Block Diagram 1/16 PLL Quartz oscillator 1/2 CLOCK1 CLOCK2 1/16 Internal RC oscillator CPU Clock Prescaler Clock Multiplier /Divider Unit CK_AF CPUCLK to CPU Core INTCLK to Peripherals 75/224 ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) Figure 35. ST92163 Clock Distribution Diagram Baud Rate Generato r 1/N N=2...(216 -1) SCI A/D 1/4 8-bit Prescaler 16-bit Up/Down Counter 1...2 56 16-bit Down Counter 8-bit Prescaler 1...2 56 WDG 1/3 (TxIN A/TxINB) (Max INTCLK/3) N=4,6, 8...258 STD MFT 24 MHz INTCLK 1/N FAST P6.2 Fscl ≤100 kHz Fscl > 100 kHz Fscl ≤ 400 kHz N=6,9,12...387 1/N I 2C 48 MHz 1...8 CPUCLK 3-bit Prescaler USB INTERFACE CPU EMBEDDED MEMORY WFI and LPOWFI=1 and WFI_CKSEL=1 or CK_ST=1 XT_DIV16=1 0 1 RAM 0 1 EPROM/ROM/OTP 1/16 CSU_CKSEL=1 DIV2=0 8 MHz Quartz Oscillator MX1=0 MX0=1 0 /2 1 CLOCK2 PLL x 6 RCCU Internal RC oscillator 76/224 0 1/2 1 DX2=0 DX1=0 DX0=2 1/16 0 1 CK_AF WFI and LPOWFI=1 and WFI_CKSEL = 1 ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) 5.3 CLOCK MANAGEMENT The various programmable features and operating modes of the CCU are handled by four registers: – MODER (Mode Register) – CLK_FLAG (Clock Flag Register) This is a System Register (R235, Group E). This is a Paged Register (R242, Page 55). The input clock divide-by-two and the CPU clock prescaler factors are handled by this register. This register contains various status flags, as well as control bits for clock selection. – CLKCTL (Clock Control Register) This is a Paged Register (R240, Page 55). – PLLCONF (PLL Configuration Register) This is a Paged Register (R246, Page 55). The low power modes and the interpretation of the HALT instruction are handled by this register. The PLL multiplication and division factors are programmed in this register. Figure 36. Clock Control Unit Programming XTSTOP DIV2 CSU_CKSEL CKAF_SEL (CLK_FLAG) (MODER) (CLK_FLAG) (CLKCTL) 1/16 0 0 Quartz oscillator 1/2 1 CLOCK2 PLL x 6/8/10/14 0 1/N 1 1 CLOCK1 1/16 Internal RC oscillator INTCLK 1 to Peripherals and CPU Clock Prescaler OUTPLL_2 (USB CLOCK) 0 1 0 CK_AF WFI and LPOWFI=1 and WFI_CKSEL = 1 MX(1:0) DX(2:0) (PLLCONF) XT_DIV16 CKAF_ST (CLK_FLAG) Wait for Interrupt and Low Power Modes: LPOWFI (CLKCTL) selects Low Power operation automatically on entering WFI mode. WFI_CKSEL (CLKCTL) selects the CK_AF clock automatically, if present, on entering WFI mode. XTSTOP (CLK_FLAG) automatically stops the Xtal oscillator when the CK_AF clock is present and selected. 77/224 ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) CLOCK MANAGEMENT (Cont’d) 5.3.1 PLL Clock Multiplier Programming The CLOCK1 signal generated by the oscillator drives a programmable divide-by-two circuit. If the DIV2 control bit in MODER is set (Reset Condition), CLOCK2, is equal to CLOCK1 divided by two; if DIV2 is reset, CLOCK2 is identical to CLOCK1. A CLOCK1 signal with a semiperiod (high or low) shorter than 40ns is forbidden if the divider by two is disabled. When the PLL is active, it multiplies CLOCK2 by 6, 8, 10 or 14, depending on the status of the MX0 -1 bits in PLLCONF. The multiplied clock is then divided by a factor in the range 1 to 7, determined by the status of the DX0-2 bits; when these bits are programmed to 111, the PLL is switched off. Following a RESET phase, programming bits DX0-2 to a value different from 111 will turn the PLL on. After allowing a stabilisation period for the PLL, setting the CSU_CKSEL bit in the CLK_FLAG Register selects the multiplier clock This peripheral contains a lock-in logic that verifies if the PLL is locked to the CLOCK2 frequency. The bit LOCK in CLK_FLAG register becomes 1 when this event occurs. The maximum frequency allowed for INTCLK is 25MHz for 5V operation, and 12MHz for 3V operation. Care is required, when programming the PLL multiplier and divider factors, not to exceed the maximum permissible operating frequency for INTCLK, according to supply voltage. The ST9 being a static machine, there is no lower limit for INTCLK. However, below 1MHz, A/D converter precision (if present) decreases. 5.3.2 CPU Clock Prescaling The system clock, INTCLK, which may be the output of the PLL clock multiplier, CLOCK2, CLOCK2/ 16 or CK_AF, drives a programmable prescaler which generates the basic time base, CPUCLK, for the instruction executer of the ST9 CPU core. This allows the user to slow down program execution during non processor intensive routines, thus reducing power dissipation. The internal peripherals are not affected by the CPUCLK prescaler and continue to operate at the full INTCLK frequency. This is particularly useful 78/224 when little processing is being done and the peripherals are doing most of the work. The prescaler divides the input clock by the value programmed in the control bits PRS2,1,0 in the MODER register. If the prescaler value is zero, no prescaling takes place, thus CPUCLK has the same period and phase as INTCLK. If the value is different from 0, the prescaling is equal to the value plus one, ranging thus from two (PRS2,1,0 = 1) to eight (PRS2,1,0 = 7). The clock generated is shown in Figure 37, and it will be noted that the prescaling of the clock does not preserve the 50% duty cycle, since the high level is stretched to replace the missing cycles. This is analogous to the introduction of wait cycles for access to external memory. When External Memory Wait or Bus Request events occur, CPUCLK is stretched at the high level for the whole period required by the function. Figure 37. CPU Clock Prescaling INTCLK PRS VALUE 000 001 010 011 CPUCLK 100 101 110 111 VA00260 5.3.3 Peripheral Clock The system clock, INTCLK, which may be the output of the PLL clock multiplier, CLOCK2, CLOCK2/ 16 or CK_AF, is also routed to all ST9 on-chip peripherals and acts as the central timebase for all timing functions. ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) CLOCK MANAGEMENT (Cont’d) 5.3.4 Low Power Modes The user can select an automatic slowdown of clock frequency during Wait for Interrupt operation, thus idling in low power mode while waiting for an interrupt. In WFI operation the clock to the CPU core (CPUCLK) is stopped, thus suspending program execution, while the clock to the peripherals (INTCLK) may be programmed as described in the following paragraphs. An example of Low Power operation in WFI is illustrated in Figure 38. If low power operation during WFI is disabled (LPOWFI bit = 0 in the CLKCTL Register), the CPU CLK is stopped but INTCLK is unchanged. If low power operation during Wait for Interrupt is enabled (LPOWFI bit = 1 in the CLKCTL Register), as soon as the CPU executes the WFI instruction, the PLL is turned off and the system clock will be forced to CLOCK2 divided by 16, or to CK_AF, if this has been selected by setting WFI_CKSEL, and providing CKAF_ST is set, thus indicating that the internal RC oscillator is selected. If the external clock source is used, the crystal oscillator may be stopped by setting the XTSTOP bit, providing that the CK_AK clock is present and selected, indicated by CKAF_ST being set. The crystal oscillator will be stopped automatically on entering WFI if the WFI_CKSEL bit has been set. It should be noted that selecting a non-existent CK_AF clock source is impossible, since such a selection requires that the auxiliary clock source be actually present and selected. In no event can a non-existent clock source be selected inadvertently. It is up to the user program to switch back to a faster clock on the occurrence of an interrupt, taking care to respect the oscillator and PLL stabilisation delays, as appropriate.It should be noted that any of the low power modes may also be selected explicitly by the user program even when not in Wait for Interrupt mode, by setting the appropriate bits. 5.3.5 Interrupt Generation System clock selection modifies the CLKCTL and CLK_FLAG registers. The clock control unit generates an external interrupt request when CK_AF and CLOCK2/16 are selected or deselected as system clock source, as well as when the system clock restarts after a hardware stop (when the STOP MODE feature is available on the specific device). This interrupt can be masked by resetting the INT_SEL bit in the CLKCTL register. In the RCCU the interrupt is generated with a high to low transition (see interrupt and DMA chapters for further information). Table 13. Summary of Operating Modes using main Crystal Controlled Oscillator MODE PLL x BY 14 PLL x BY 10 PLL x BY 8 PLL x BY 6 SLOW 1 SLOW 2 WAIT FOR INTERRUPT LOW POWER WAIT FOR INTERRUPT RESET EXAMPLE XTAL=4.4 MHz INTCLK XTAL/2 x (14/D) XTAL/2 x (10/D) XTAL/2 x (8/D) XTAL/2 x (6/D) XTAL/2 XTAL/32 CPUCLK DIV2 PRS0-2 CSU_CKSEL MX0-1 DX2-0 LPOWFI XT_DIV16 INTCLK/N 1 N-1 1 10 D-1 X 1 INTCLK/N 1 N-1 1 00 D-1 X 1 INTCLK/N 1 N-1 1 11 D-1 X 1 INTCLK/N 1 N-1 1 01 D-1 X 1 INTCLK/N INTCLK/N 1 1 N-1 N-1 X X X X 111 X X X 1 0 If LPOWFI=0, no changes occur on INTCLK, but CPUCLK is stopped anyway. XTAL/32 STOP 1 X X X X 1 1 XTAL/2 INTCLK 1 0 0 00 111 0 1 2.2*10/2 = 11MHz 11MHz 1 0 1 00 001 X 1 79/224 ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) Figure 38. Example of Low Power Mode programming COMMENTS PROGRAM FLOW SYSTEM CLOCK FREQUENCY FQ =8 MHz, Vcc=5 V and T=25°C Begin SET UP AFTER RESET PHASE: DIV2 = 0 Quartz not divided by 2 XTSTOP = 0 Xtal is selected to restart the PLL quickly PLL multiply factor fixed to 6. CSU_CKSEL = 0 MX(1:0) = 01 8 MHz Multiplier divider’s factor set to 2, and PLL turned ON DX2-0 ← 001 WAIT CSU_CKSEL Wait for the PLL locking (LOCK->1) WFI_CKSEL ← 1 1 PLL is system clock source T1* CK_AF clock selected when WFI XTSTOP ← 1 To stop PLL and Xtal when a WFI occurs LPOWFI ← 1 Low Power Mode enabled 24 MHz User’s Program WFI Interrupt WFI status Interrupt Routine WAIT XTSTOP ← 0 CKAF_SEL <- 0 Main code execution continued CSU_CKSEL<- 1 User’s Program End 80/224 Wait For Interrupt No code is executed until an interrupt is requested RC oscillator Interrupt served while the CK_AF is the system clock To restart Xtal and PLL The system CK switches to Xtal The PLL is locked and becomes the system clock T1* T2** RC oscillator 8 MHz 24 MHz * T1 = PLL lock-in time ** T2 = Quartz oscillator start-up time ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) 5.4 CLOCK CONTROL REGISTERS MODE REGISTER (MODER) R235 - Read/Write System Register Reset Value: 1110 0000 (E0h) CLOCK CONTROL REGISTER (CLKCTL) R240 - Read Write Register Page: 55 Reset Value: 0000 0000 (00h) 7 - 0 - DIV2 PRS2 PRS1 PRS0 - - 7 INT_S EL 0 - - - SRESEN CKAF_S WFI_CKS LPOW EL EL FI *Note: This register contains bits which relate to other functions; these are described in the chapter dealing with Device Architecture. Only those bits relating to Clock functions are described here. Bit 5 = DIV2: OSCIN Divided by 2. This bit controls the divide by 2 circuit which operates on the OSCIN Clock. 0: No division of the OSCIN Clock 1: OSCIN clock is internally divided by 2 Bit 7 = INT_SEL: Interrupt Selection . 0: Select the external interrupt pin as interrupt source (Reset state) 1: Select the internal RCCU interrupt (see Section 5.3.5) Bit 4:2 = PRS[2:0]: Clock Prescaling. These bits define the prescaler value used to prescale CPUCLK from INTCLK. When these three bits are reset, the CPUCLK is not prescaled, and is equal to INTCLK; in all other cases, the internal clock is prescaled by the value of these three bits plus one. Bit 3 = SRESEN: Software Reset Enable. 0: The HALT instruction turns off the quartz, the PLL and the CCU 1: A Reset is generated when HALT is executed Bit 6:4 = Reserved. Must be kept reset for normal operation. Bit 2 = CKAF_SEL: Alternate Function Clock Select. 0: CK_AF clock not selected 1: Select CK_AF clock Note: To check if the selection has actually occurred, check that CKAF_ST is set. If no CK_AF clock is present, the selection will not occur. Bit 1 = WFI_CKSEL: WFI Clock Select. This bit selects the clock used in Low power WFI mode if LPOWFI = 1. 0: INTCLK during WFI is CLOCK2/16 1: INTCLK during WFI is CK_AF, providing it is present. In effect this bit sets CKAF_SEL in WFI mode WARNING: When the CK_AF is selected as Low Power WFI clock but the XTAL is not turned off (R242.4 = 0), after exiting from the WFI, CK_AF will be still selected as system clock. In this case, reset the R240.2 bit to switch back to the XT. Bit 0 = LPOWFI: Low Power mode during Wait For Interrupt. 0: Low Power mode during WFI disabled. When WFI is executed, the CPUCLK is stopped and INTCLK is unchanged 1: The ST9 enters Low Power mode when the WFI instruction is executed. The clock during this state depends on WFI_CKSEL 81/224 ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) CLOCK CONTROL REGISTERS CLOCK FLAG REGISTER (CLK_FLAG) R242 -Read/Write Register Page: 55 Reset Value: 0100 1000 after a Watchdog Reset Reset Value: 0010 1000 after a Software Reset Reset Value: 0000 1000 after a Power-On Reset 7 EX_ STP 0 WDG RES SOFT RES XTSTOP XT_ CKAF_ LOC DIV16 ST K CSU_ CKSEL WARNING: If this register is accessed with a logical instruction, such as AND or OR, some bits may not be set as expected. WARNING: If you select the CK_AF as system clock and turn off the oscillator (bits R240.2 and R242.4 at 1), and then switch back to the XT clock by resetting the R240.2 bit, you must wait for the oscillator to restart correctly (12ms). Bit 7 = EX_STP: External Stop flag This bit is set by hardware and cleared by software. 0: No External Stop condition occurred 1: External Stop condition occurred Bit 6 = WDGRES: Watchdog reset flag. This bit is read only. 0: No Watchdog reset occurred 1: Watchdog reset occurred Bit 5 = SOFTRES: Software Reset Flag. This bit is read only. 0: No software reset occurred 1: Software reset occurred (HALT instruction) Bit 4 = XTSTOP: External Stop Enable. 0: External stop disabled 1: The Xtal oscillator will be stopped as soon as the CK_AF clock is present and selected, whether this is done explicitly by the user program, or as a result of WFI, if WFI_CKSEL has previously been set to select the CK_AF clock during WFI. WARNING: When the program writes ‘1’ to the XTSTOP bit, it will still be read as 0 and is only set when the CK_AF clock is running (CKAF_ST=1). 82/224 Take care, as any operation such as a subsequent AND with ‘ 1’ or an OR with ‘ 0’ to the XTSTOP bit will reset it and the oscillator will not be stopped even if CKAF_ST is subsequently set. Bit 3 = XT_DIV16: CLOCK/16 Selection This bit is set and cleared by software. An interrupt is generated when the bit is toggled. 0: CLOCK2/16 is selected and the PLL is off 1: The input is CLOCK2 (or the PLL output depending on the value of CSU_CKSEL) WARNING: After this bit is modified from 0 to 1, take care that the PLL lock-in time has elapsed before setting the CSU_CKSEL bit. Bit 2 = CKAF_ST: (Read Only) If set, indicates that the alternate function clock has been selected. If no CK_AF clock signal is present on the pin, the selection will not occur. If reset, the PLL clock, CLOCK2 or CLOCK2/16 is selected (depending on bit 0). Bit 1= LOCK: PLL locked-in This bit is read only. 0: The PLL is turned off or not locked and cannot be selected as system clock source. 1: The PLL is locked Bit 0 = CSU_CKSEL: CSU Clock Select This bit is set and cleared by software. It also cleared by hardware when: – bits DX[2:0] (PLLCONF) are set to 111; – the quartz is stopped (by hardware or software); – WFI is executed while the LPOWFI bit is set; – the XT_DIV16 bit (CLK_FLAG) is forced to ’0’. This prevents the PLL, when not yet locked, from providing an irregular clock. Furthermore, a ‘0’ stored in this bit speeds up the PLL’s locking. 0: CLOCK2 provides the system clock 1: The PLL Multiplier provides the system clock. NOTE: Setting the CKAF_SEL bit overrides any other clock selection. Resetting the XT_DIV16 bit overrides the CSU_CKSEL selection ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) Table 14. PLL Multiplication Factors PLL CONFIGURATION REGISTER (PLLCONF) R246 - Read/Write Register Page: 55 Reset Value: xx00 x111 7 - MX1 1 0 1 0 0 - MX1 MX0 - DX2 DX1 DX0 MX0 0 0 1 1 CLOCK2 x 14 10 8 6 Bit 7:6 = Reserved. Table 15. Divider Configuration Bit 5:4 = MX[1:0]: PLL Multiplication Factor. Refer to Table 14 for multiplier settings. Bit 3 = Reserved. Bit 2:0 = DX[2:0]: PLL output clock divider factor. Refer to Table 15 for divider settings. DX2 0 0 0 0 1 1 1 DX1 0 0 1 1 0 0 1 DX0 0 1 0 1 0 1 0 1 1 1 CK PLL CLOCK/1 PLL CLOCK/2 PLL CLOCK/3 PLL CLOCK/4 PLL CLOCK/5 PLL CLOCK/6 PLL CLOCK/7 CLOCK2 (PLL OFF, Reset State) Figure 39. RCCU General Timing PLL selected by user PLL turned on by user STOP pin External RESET Xtal clock Multiplier clock Internal reset Xtal/2 Xtal/2 PLL PLL INTCLK ((N-1)*512+510)xT Xtal(**) PLL 4 xTsys Lock-in time (*) if DIV2 =1 (**) +/- 1 TXtal Exit from RESET 510 x TXtal(*) Quartz start-up STOP Acknowledged PLL Lock-in time Switch to PLL clock 83/224 ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) Figure 40. RCCU Timing during STOP (CK_AF System Clock) CKAF_SEL<-1 STOP pin Xtal clock RC osc clock INTCLK ((N-1)*512+510) x TXtal(**) Exit from RESET CK_AF selected (*) if DIV2 =1 4 xTsys STOP Acknowledged (**) +/- 1 TXtal 510 x TXtal(*) Quartz start-up Figure 41. Low Power Mode with a Stopped Quartz Oscillator WFI state XTSTOP<-0 CKAF_SEL<-0 CSU_CKSEL<-1 Xtal clock Multiplier clock RC osc clock PLL (Tsys = TCK_AF) Tsys =2 xTXtal INTCLK Xtal restart time With: DIV2=1 84/224 Interrupt PLL Lock-in time PLLl ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) 5.5 OSCILLATOR CHARACTERISTICS The oscillator circuit uses an inverting gate circuit with tri-state output. Notes: Owing to the Q factor required, Ceramic Resonators may not provide a reliable oscillator source. OSCOUT must not be directly used to drive external circuits. When the oscillator is stopped, OSCOUT goes high impedance. In Halt mode, set by means of the HALT instruction, the parallel resistor, R, is disconnected and the oscillator is disabled, forcing the internal clock, CLOCK1, to a high level, and OSCOUT to a high impedance state. To exit the HALT condition and restart the oscillator, an external RESET pulse is required. It should be noted that, if the Watchdog function is enabled, a HALT instruction will not disable the oscillator. This to avoid stopping the Watchdog if a HALT code is executed in error. When this occurs, the CPU will be reset when the Watchdog times out or when an external reset is applied. Symbol Condition s C1=C 2 C 1=C2 C 1=C 2 C 1=C2 = 56pF = 47pF = 33pF = 22pF 4.0-5.5V 50 Ω 65Ω 120Ω 180Ω Rsmax Freq.=8MHz (ohm) 3.0-4.0V 30 Ω 40Ω 70Ω 110Ω Freq.=8MHz Legend: C1, C2: Maximum Total Capacitances on pins OSCIN and OSCOUT (the value includes the external capacitance tied to the pin CL1 and CL2 plus the parasitic capacitance of the board and of the device). Rsmax: The equivalent serial resistor of the crystal. Note 1: The tables are relative to the fundamental quartz crystal only (not ceramic resonator). Note 2: To reduce the parasitic capacitance, it is recommended to place the crystal as close to the ST9 MCU as possible. WARNING: At low temperature, frost and humidity might prevent the correct start-up of the oscillator. Figure 43. Internal Oscillator Schematic HALT Table 16. Oscillator Transconductance Voltage Symbol range 4.0-5.5V gm 3.0-4.0V Table 17. Crystal Internal Resistance CLOCK1 Min Typ Max Unit 0.77 0.5 1.5 0.73 2.4 1.47 mA/V R Figure 42. Crystal Oscillator ROUT R IN CRYSTAL CLOCK OSCIN OSCOUT OSCIN ST9 VR02086A Figure 44. External Clock OSCOUT EXTERNAL CLOCK CL1 ST9 CL2 OSCIN 1M* *Recommended for oscillator stability VR02116A OSCOUT NC CLOCK INPUT VR02116B 85/224 ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) 5.6 RESET/STOP MANAGER The Reset/Stop Manager resets the MCU when TRES or the WDGRES bits respectively; a hardone of the three following events occurs: ware initiated reset will leave both these bits reset. – A Hardware reset, initiated by a low level on the The hardware reset overrides all other conditions Reset pin. and forces the ST9 to the reset state. During Reset, the internal registers are set to their reset val– A Software reset, initiated by a HALT instruction ues, where these are defined, and the I/O pins are (when enabled). set to the Bidirectional Weak Pull-up mode. – A Watchdog end of count condition. Reset is asynchronous: as soon as the reset pin is The event which caused the last Reset is flagged driven low, a Reset cycle is initiated. in the CLK_FLAG register, by setting the SOFFigure 45. Oscillator Start-up Sequence and Reset Timing VDD MAX VDD MIN OSCIN OSCOUT 10ms INTCLK RESET PIN 86/224 VR02085A ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) RESET/STOP MANAGER (Cont’d) The on-chip Timer/Watchdog generates a reset condition if the Watchdog mode is enabled (WCR.WDEN cleared, R252 page 0), and if the programmed period elapses without the specific code (AAh, 55h) written to the appropriate register. The input pin RESET is not driven low by the onchip reset generated by the Timer/Watchdog. When the Reset pin goes high again, a delay of 10 ms occurs before exiting the Reset state (+-1 CLOCK1 period, depending on the delay between the rising edge of the Reset pin and the first rising edge of CLOCK1). Subsequently a short Boot routine is executed from the device internal Boot ROM, and control then passes to the user program. The Boot routine sets the device characteristics and loads the correct values in the Memory Management Unit’s pointer registers, so that these point to the physical memory areas as mapped in the specific device. The precise duration of this short Boot routine varies from device to device, depending on the Boot ROM contents. At the end of the Boot routine the Program Counter will be set to the location specified in the Reset Vector located in the lowest two bytes of memory. 5.6.1 Reset Pin Timing To improve the noise immunity of the device, the Reset pin has a Schmitt trigger input circuit with hysteresis. In addition, a filter will prevent an unwanted reset in case of a single glitch of less than 50 ns on the Reset pin. The device is certain to reset if a negative pulse of more than 20µs is applied. When the reset pin goes high again, a delay of up to 4µs (at 8 MHz.) will elapse before the RCCU detects this rising front. From this event on, 79870 (about 10 ms at 8 MHz.) oscillator clock cycles (CLOCK1) are counted before exiting the Reset state (+-1 CLOCK1 period depending on the delay between the positive edge the RCCU detects and the first rising edge of CLOCK1) If the ST9 is a ROMLESS version, without on-chip program memory, the memory interface ports are set to external memory mode (i.e Alternate Function) and the memory accesses are made to external Program memory with wait cycles insertion. Figure 46. Recommended Signal to be Applied on Reset Pin VRESET VCC 0.7 VCC 0.3 VCC 20 µs Minimum 5.7 STOP MODE In Stop mode, the Reset/Stop Manager can also stop all oscillators without resetting the device. For information on entering and exiting Stop Mode, refer to the Wake-Up/Interrupt lines management unit (WUIMU) chapter. In Stop Mode, all context information is preserved and the internal clock is frozen in the high state. On exiting Stop mode, the MCU resumes execution of the user program after a delay of 255 CLOCK2 periods, an interrupt is generated and the EX_STP bit in CLK_FLAG is set. 87/224 ST92163 - RESET AND CLOCK CONTROL UNIT (RCCU) 5.8 LOW VOLTAGE DETECTOR (LVD) RESET The on-chip Low Voltage Detector (LVD) generates a static reset when the supply voltage is below a reference value. The LVD works both during power-on as well as when the power supply drops (brown-out). The reference value for the voltage drop is lower than the reference value for poweron in order to avoid a parasitic reset when the MCU starts running and sinks current on the supply (hysteresis). The LVD circuitry generates a reset when VDD is below: – VLVDUP when VDD is rising – VLVDOWN when VDD is falling The Low Voltage Detector circuitry resets only the MCU and it does not change the external RESET pin status: no reset signal for an external application is generated. Figure 47. Low Voltage Detector Reset Function VDD Low Voltage Detector Reset Internal Reset RESET (External pin) Figure 48. Low Voltage Detector Reset Signal VLVDUP VLVDDOWN VDD Internal Reset Note: Refer to Electrical Characteristics for the values of VDD, VLVDUP and VLVDDOWN. 88/224 ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) 6 EXTERNAL MEMORY INTERFACE (EXTMI) 6.1 INTRODUCTION The ST9 External Memory Interface uses two registers (EMR1 and EMR2) to configure external memory accesses. Some interface signals are also affected by WCR - R252 Page 0. If the two registers EMR1 and EMR2 are set to the proper values, the ST9+ memory access cycle is similar to that of the original ST9, with the only exception that it is composed of just two system clock phases, named T1 and T2. During phase T1, the memory address is output on the AS falling edge and is valid on the rising edge of AS. Port0 and Port 1 maintain the address stable until the following T1 phase. Figure 49. Page 21 Registers During phase T2, two forms of behavior are possible. If the memory access is a Read cycle, Port 0 pins are released in high-impedance until the next T1 phase and the data signals are sampled by the ST9 on the rising edge of DS. If the memory access is a Write cycle, on the falling edge of DS, Port 0 outputs data to be written in the external memory. Those data signals are valid on the rising edge of DS and are maintained stable until the next address is output. Note that DS is pulled low at the beginning of phase T2 only during an external memory access. n Page 21 FFh R255 FEh R254 FDh R253 FCh R252 FBh R251 FAh R250 F9h F8h DMASR ISR F7h Relocation of P[3:0] and DPR[3:0] Registers R249 R248 MMU R247 F6h EMR2 R246 F5h EMR1 R245 F4h CSR R244 F3h DPR3 R243 F2h DPR2 R242 F1h DPR1 R241 F0h DPR0 R240 EXT.MEM MMU SSPL SSPH USPL USPH MODER PPR RP1 RP0 FLAGR CICR P5 P4 P3 P2 P1 P0 DMASR ISR EMR2 EMR1 CSR DPR3 DPR2 DPR1 DPR0 Bit DPRREM=0 SSPL SSPH USPL USPH MODER PPR RP1 RP0 FLAGR CICR P5 P4 DPR3 DPR2 DPR1 DPR0 DMASR ISR EMR2 EMR1 CSR P3 P2 P1 P0 Bit DPRREM=1 89/224 ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) 6.2 EXTERNAL MEMORY SIGNALS The access to external memory is made using the AS, DS, DS2, RW, Port 0, Port1, and WAIT signals described below. Refer to Figure 51 6.2.1 AS: Address Strobe AS (Output, Active low, Tristate) is active during the System Clock high-level phase of each T1 memory cycle: an AS rising edge indicates that Memory Address and Read/Write Memory control signals are valid. AS is released in high-impedance during the bus acknowledge cycle or under the processor control by setting the HIMP bit (MODER.0, R235). Depending on the device AS is available as Alternate Function or as a dedicated pin. Under Reset, AS is held high with an internal weak pull-up. The behavior of this signal is affected by the MC, ASAF, ETO, BSZ, LAS[1:0] and UAS[1:0] bits in the EMR1 or EMR2 registers. Refer to the Register description. 6.2.2 DS: Data Strobe DS (Output,Active low, Tristate) is active during the internal clock high-level phase of each T2 memory cycle. During an external memory read cycle, the data on Port 0 must be valid before the DS rising edge. During an external memory write cycle, the data on Port 0 are output on the falling edge of DS and they are valid on the rising edge of DS. When the internal memory is accessed DS is kept high during the whole memory cycle. DS is released in high-impedance during bus acknowledge cycle or 90/224 under processor control by setting the HIMP bit (MODER.0, R235). Under Reset status, DS is held high with an internal weak pull-up. The behavior of this signal is affected by the MC, DS2EN, and BSZ bits in the EMR1 register. Refer to the Register description. 6.2.3 DS2: Data Strobe 2 This additional Data Strobe pin (Alternate Function Output, Active low, Tristate) is available on some ST9 devices only. It allows two external memories to be connected to the ST9, the upper memory block (A21=1 typically RAM) and the lower memory block (A21=0 typically ROM) without any external logic. The selection between the upper and lower memory blocks depends on the A21 address pin value. The upper memory block is controlled by the DS pin while the lower memory block is controlled by the DS2 pin. When the internal memory is addressed, DS2 is kept high during the whole memory cycle. DS2 is released in high-impedance during bus acknowledge cycle or under processor control by setting the HIMP bit (MODER.0, R235). DS2 is enabled via software as the Alternate Function output of the associated I/O port bit (refer to specific ST9 version to identify the specific port and pin). The behavior of this signal is affected by the DS2EN, and BSZ bits in the EMR1 register. Refer to the Register description. ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) EXTERNAL MEMORY SIGNALS (Cont’d) Figure 50. Effects of DS2EN on the behavior of DS and DS2 n T1 NO WAIT CYCLE T2 T1 SYSTEM CLOCK 1 DS WAIT CYCLE T2 DS STRETCH AS (MC=0) DS2EN=0 OR (DS2EN=1 AND UPPER MEMORY ADDRESSED): DS (MC=0) DS (MC=1, READ) DS (MC=1, WRITE) DS2 DS2EN=1 AND LOWER MEMORY ADDRESSED: DS DS2 (MC=0) DS2 (MC=1, READ) DS2 (MC=1, WRITE) 91/224 ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) EXTERNAL MEMORY SIGNALS (Cont’d) Figure 51. External memory Read/Write with a programmable wait n NO WAIT CYCLE T1 T2 SYSTEM CLOCK 1 DS WAIT CYCLE 1 AS WAIT CYCLE T1 T2 TWA TWD AS STRETCH DS STRETCH AS (MC=0) ALWAYS TAVQV ALE (MC=1) P1 ADDRESS ADDRESS DS (MC=0) ADDRESS DATA IN DATA IN ADDRESS READ P0 MULTIPLEXED RW (MC=0) DS (MC=1) RW (MC=1) P0 MULTIPLEXED ADDRESS DATA OUT DATA ADDRESS DS (MC=1) RW (MC=1) 92/224 TAVWL WRITE TAVWH RW (MC=0) ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) EXTERNAL MEMORY SIGNALS (Cont’d) 6.2.4 RW: Read/Write RW (Alternate Function Output, Active low, Tristate) identifies the type of memory cycle: RW=”1” identifies a memory read cycle, RW=”0” identifies a memory write cycle. It is defined at the beginning of each memory cycle and it remains stable until the following memory cycle. RW is released in high-impedance during bus acknowledge cycle or under processor control by setting the HIMP bit (MODER). RW is enabled via software as the Alternate Function output of the associated I/O port bit (refer to specific ST9 device to identify the port and pin). Under Reset status, the associated bit of the port is set into bidirectional weak pull-up mode. The behavior of this signal is affected by the MC, ETO and BSZ bits in the EMR1 register. Refer to the Register description. 6.2.5 BREQ, BACK: Bus Request, Bus Acknowledge Note: These pins are available only on some ST9 devices (see Pin description). BREQ (Alternate Function Input, Active low) indicates to the ST9 that a bus request has tried or is trying to gain control of the memory bus. Once enabled by setting the BRQEN bit (MODER.1, R235), BREQ is sampled with the falling edge of the processor internal clock during phase T2. n n Figure 52. External memory Read/Write sequence with external wait (WAIT pin) n T1 T2 T1 T2 T1 T2 WAI T P1 ADDRESS ALWAYS SYST EM CLOCK ADDRESS ADDRESS AS (MC=0) ALE (MC=1) DS (MC=0) P0 ADD. D.IN ADD. D.OUT ADDRESS D.IN ADD. D.IN READ MULTIPLEXED RW (MC=0) DS (MC=1) RW (MC=1) P0 RW (MC=0) DS (MC=1) ADDRESS D.OUT ADD. DATA OUT WRITE MULTIPLEXED RW (MC=1) 93/224 ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) EXTERNAL MEMORY SIGNALS (Cont’d) Whenever it is sampled low, the System Clock is stretched and the external memory signals (AS, DS, DS2, RW, P0 and P1) are released in high-impedance. The external memory interface pins are driven again by the ST9 as soon as BREQ is sampled high. BACK (Alternate Function Output, Active low) indicates that the ST9 has relinquished control of the memory bus in response to a bus request. BREQ is driven low when the external memory interface signals are released in high-impedance. At MCU reset, the bus request function is disabled. To enable it, configure the I/O port pins assigned to BREQ and BACK as Alternate Function and set the BRQEN bit in the MODER register. 6.2.6 PORT 0 If Port 0 (Input/Output, Push-Pull/Open-Drain/ Weak Pull-up) is used as a bit programmable parallel I/O port, it has the same features as a regular port. When set as an Alternate Function, it is used as the External Memory interface: it outputs the multiplexed Address 8 LSB: A[7:0] /Data bus D[7:0]. 6.2.7 PORT 1 If Port 1 (Input/Output, Push-Pull/Open-Drain/ Weak Pull-up) is used as a bit programmable parallel I/O port, it has the same features as a regular port. When set as an Alternate Function, it is used Figure 53. Application Example as the external memory interface to provide the 8 MSB of the address A[15:8]. The behavior of the Port 0 and 1 pins is affected by the BSZ and ETO bits in the EMR1 register. Refer to the Register description. 6.2.8 WAIT: External Memory Wait WAIT (Alternate Function Input, Active low) indicates to the ST9 that the external memory requires more time to complete the memory access cycle. If bit EWEN (EIVR) is set, the WAIT signal is sampled with the rising edge of the processor internal clock during phase T1 or T2 of every memory cycle. If the signal was sampled active, one more internal clock cycle is added to the memory cycle. On the rising edge of the added internal clock cycle, WAIT is sampled again to continue or finish the memory cycle stretching. Note that if WAIT is sampled active during phase T1 then AS is stretched, while if WAIT is sampled active during phase T2 then DS is stretched. WAIT is enabled via software as the Alternate Function input of the associated I/O port bit (refer to specific ST9 version to identify the specific port and pin). Under Reset status, the associated bit of the port is set to the bidirectional weak pull-up mode. Refer to Figure 52 RAM 64 Kbytes RW W G DS P0 A0-A7/D0-D7 ST9+ Q0-Q7 D1-D8 Q1-Q8 LE AS A0-A15 OE LATCH P1 94/224 A15-A8 E ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) 6.3 REGISTER DESCRIPTION EXTERNAL MEMORY REGISTER 1 (EMR1) R245 - Read/Write Register Page: 21 Reset value: 1000 0000 (80h) 7 x 0 MC DS2EN ASAF x ETO BSZ Bit 4 = ASAF: Address Strobe as Alternate Function. Depending on the device, AS can be either a dedicated pin or a port Alternate Function. This bit is used only in the second case. 0: AS Alternate function disabled. 1: AS Alternate Function enabled. X Bit 7 = Reserved. Bit 6 = MC: Mode Control. 0: AS, DS and RW pins keep the ST9OLD meaning. 1: AS pin becomes ALE, Address Load Enable (AS inverted); Thus Memory Adress, Read/ Write signals are valid whenever a falling edge of ALE occurs. DS becomes OEN, Output ENable: it keeps the ST9OLD meaning during external read operations, but is forced to “1” during external write operations. RW pin becomes WEN, Write ENable: it follows the ST9OLD DS meaning during external write operations, but is forced to “1” during external read operations. Bit 5 = DS2EN: Data Strobe 2 enable. 0: The DS2 pin is forced to “1” during the whole memory cycle. 1: If the lower memory block is addressed, the DS2 pin follows the ST9OLD DS meaning (if MC=0) or it becomes OEN (if MC=1). The DS pin is forced to 1 during the whole memory cycle. If the upper memory block is used, DS2 is forced to “1” during the whole memory cycle. The DS pin behaviour is not modified. Refer to Figure 50 Bit 2 = ETO: External toggle. 0: The external memory interface pins (AS, DS, DS2, RW, Port0, Port1) toggle only if an access to external memory is performed. 1: When the internal memory protection is disabled (mask option available on some devices only), the above pins (except DS and DS2 which never toggle during internal memory accesses) toggle during both internal and external memory accesses. Bit 1 = BSZ: Bus size. 0: All the I/O ports including the external memory interface pins use smaller, less noisy output buffers. This may limit the operation frequency of the device, unless the clock is slow enough or sufficient wait states are inserted. 1: All the I/O ports including the external memory interface pins (AS, DS, DS2, R/W, Port 0, 1) use larger, more noisy output buffers . Bit 0 = Reserved. WARNING: External memory must be correctly addressed before and after a write operation on the EMR1 register. For example, if code is fetched from external memory using the ST9OLD external memory interface configuration (MC=0), setting the MC bit will cause the device to behave unpredictably. 95/224 ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) REGISTER DESCRIPTION (Cont’d) EXTERNAL MEMORY REGISTER 2 (EMR2) R246 - Read/Write Register Page: 21 Reset value: 0000 1111 (0Fh) 7 - 0 ENCSR DPRREM MEM SEL LAS1 LAS0 UAS1 UAS0 the contents of ISR. In this case, iret will also restore CSR from the stack. This approach allows interrupt service routines to access the entire 4Mbytes of address space; the drawback is that the interrupt response time is slightly increased, because of the need to also save CSR on the stack. Full compatibility with the original ST9 is lost in this case, because the interrupt stack frame is different; this difference, however, should not affect the vast majority of programs. Bit 7 = Reserved. Bit 6 = ENCSR: Enable Code Segment Register. This bit affects the ST9 CPU behavior whenever an interrupt request is issued. 0: The CPU works in original ST9 compatibility mode concerning stack frame during interrupts. For the duration of the interrupt service routine, ISR is used instead of CSR, and the interrupt stack frame is identical to that of the original ST9: only the PC and Flags are pushed. This avoids saving the CSR on the stack in the event of an interrupt, thus ensuring a faster interrupt response time. The drawback is that it is not possible for an interrupt service routine to perform inter-segment calls or jumps: these instructions would update the CSR, which, in this case, is not used (ISR is used instead). The code segment size for all interrupt service routines is thus limited to 64K bytes. 1: If ENCSR is set, ISR is only used to point to the interrupt vector table and to initialize the CSR at the beginning of the interrupt service routine: the old CSR is pushed onto the stack together with the PC and flags, and CSR is then loaded with 96/224 Bit 5 = DPRREM: Data Page Registers remapping 0: The locations of the four MMU (Memory Management Unit) Data Page Registers (DPR0, DPR1, DPR2 and DPR3) are in page 21. 1: The four MMU Data Page Registers are swapped with that of the Data Registers of ports 0-3. Refer to Figure 49 Bit 4 = MEMSEL: Memory Selection. Warning: Must be set by the user when using the external memory interface (Reset value is 0). Bit 3:2 = LAS[1:0]: Lower memory address strobe stretch. These two bits contain the number of wait cycles (from 0 to 3) to add to the System Clock to stretch AS during external lower memory block accesses (MSB of 22-bit internal address=0). The reset value is 3. ST92163 - EXTERNAL MEMORY INTERFACE (EXTMI) REGISTER DESCRIPTION (Cont’d) Bit 1:0 = UAS[1:0]: Upper memory address strobe stretch. These two bits contain the number of wait cycles (from 0 to 3) to add to the System Clock to stretch AS during external upper memory block accesses (MSB of 22-bit internal address=1). The reset value is 3. WARNING: The EMR2 register cannot be written during an interrupt service routine. WAIT CONTROL REGISTER (WCR) R252 - Read/Write Register Page: 0 Reset Value: 0111 1111 (7Fh) 7 0 0 WDGEN UDS2 UDS1 UDS0 LDS2 LDS1 LDS0 Bit 7 = Reserved, forced by hardware to 0. Bit 6 = WDGEN: Watchdog Enable. For a description of this bit, refer to the Timer/ Watchdog chapter. WARNING: Clearing this bit has the effect of setting the Timer/Watchdog to Watchdog mode. Unless this is desired, it must be set to “1”. Bit 5:3 = UDS[2:0]: Upper memory data strobe stretch. These bits contain the number of INTCLK cycles to be added automatically to DS for external upper memory block accesses. UDS = 0 adds no addi- tional wait cycles. UDS = 7 adds the maximum 7 INTCLK cycles (reset condition). Bit 2:0 = LDS[2:0]: Lower memory data strobe stretch. These bits contain the number of INTCLK cycles to be added automatically to DS or DS2 (depending on the DS2EN bit of the EMR1 register) for external lower memory block accesses. LDS = 0 adds no additional wait cycles, LDS = 7 adds the maximum 7 INTCLK cycles (reset condition). Note 1: The number of clock cycles added refers to INTCLK and NOT to CPUCLK. Note 2: The distinction between the Upper memory block and the Lower memory block allows different wait cycles between the first 2 Mbytes and the second 2 Mbytes, and allows 2 different data strobe signals to be used to access 2 different memories. Typically, the RAM will be located above address 0x200000 and the ROM below address 0x1FFFFF, with different access times. No extra hardware is required as DS is used to access the upper memory block and DS2 is used to access the lower memory block. WARNING: The reset value of the Wait Control Register gives the maximum number of Wait cycles for external memory. To get optimum performance from the ST9, the user should write the UDS[2:0] and LDS[2:0] bits to 0, if the external addressed memories are fast enough. 97/224 ST92163 - I/O PORTS 7 I/O PORTS 7.1 INTRODUCTION 7.2 SPECIFIC PORT CONFIGURATIONS ST9 devices feature flexible individually programmable multifunctional input/output lines. Refer to the Pin Description Chapter for specific pin allocations. These lines, which are logically grouped as 8-bit ports, can be individually programmed to provide digital input/output and analog input, or to connect input/output signals to the on-chip peripherals as alternate pin functions. All ports can be individually configured as an input, bi-directional, output or alternate function. In addition, pull-ups can be turned off for open-drain operation, and weak pull-ups can be turned on in their place, to avoid the need for off-chip resistive pull-ups. Ports configured as open drain must never have voltage on the port pin exceeding VDD (refer to the Electrical Characteristics section). Input buffers can be either TTL or CMOS compatible. Alternatively some input buffers can be permanently forced by hardware to operate as Schmitt triggers. Refer to the Pin Description chapter for a list of the specific port styles and reset values. 7.3 PORT CONTROL REGISTERS Each port is associated with a Data register (PxDR) and three Control registers (PxC0, PxC1, PxC2). These define the port configuration and allow dynamic configuration changes during program execution. Port Data and Control registers are mapped into the Register File as shown in Figure 54. Port Data and Control registers are treated just like any other general purpose register. There are no special instructions for port manipulation: any instruction that can address a register, can address the ports. Data can be directly accessed in the port register, without passing through other memory or “accumulator” locations. Figure 54. I/O Register Map GROUP E System Registers E5h E4h E3h E2h E1h E0h 98/224 P5DR P4DR P3DR P2DR P1DR P0DR R229 R228 R227 R226 R225 R224 FFh FEh FDh FCh FBh FAh F9h F8h F7h F6h F5h F4h F3h F2h F1h F0h GROUP F PAGE 2 Reserved P3C2 P3C1 P3C0 Reserved P2C2 P2C1 P2C0 Reserved P1C2 P1C1 P1C0 Reserved P0C2 P0C1 P0C0 GROUP F PAGE 3 P7DR P7C2 P7C1 P7C0 P6DR P6C2 P6C1 P6C0 Reserved P5C2 P5C1 P5C0 Reserved P4C2 P4C1 P4C0 GROUP F PAGE 43 P9DR P9C2 P9C1 P9C0 P8DR P8C2 P8C1 P8C0 Reserved R255 R254 R253 R252 R251 R250 R249 R248 R247 R246 R245 R244 R243 R242 R241 R240 ST92163 - I/O PORTS PORT CONTROL REGISTERS (Cont’d) During Reset, ports with weak pull-ups are set in bidirectional/weak pull-up mode and the output Data Register is set to FFh. This condition is also held after Reset, except for Ports 0 and 1 in ROMless devices, and can be redefined under software control. Bidirectional ports without weak pull-ups are set in high impedance during reset. To ensure proper levels during reset, these ports must be externally connected to either VDD or VSS through external pull-up or pull-down resistors. Other reset conditions may apply in specific ST9 devices. 7.4 INPUT/OUTPUT BIT CONFIGURATION By programming the control bits PxC0.n and PxC1.n (see Figure 55) it is possible to configure bit Px.n as Input, Output, Bidirectional or Alternate Function Output, where X is the number of the I/O port, and n the bit within the port (n = 0 to 7). When programmed as input, it is possible to select the input level as TTL or CMOS compatible by programming the relevant PxC2.n control bit, except where the Schmitt trigger option is assigned to the pin. The output buffer can be programmed as pushpull or open-drain. A weak pull-up configuration can be used to avoid external pull-ups when programmed as bidirectional (except where the weak pull-up option has been permanently disabled in the pin hardware assignment). Each pin of an I/O port may assume software programmable Alternate Functions (refer to the device Pin Description and to Section 7.5). To output signals from the ST9 peripherals, the port must be configured as AF OUT. On ST9 devices with A/D Converter(s), configure the ports used for analog inputs as AF IN. The basic structure of the bit Px.n of a general purpose port Px is shown in Figure 56. Independently of the chosen configuration, when the user addresses the port as the destination register of an instruction, the port is written to and the data is transferred from the internal Data Bus to the Output Master Latches. When the port is addressed as the source register of an instruction, the port is read and the data (stored in the Input Latch) is transferred to the internal Data Bus. When Px.n is programmed as an Input: (See Figure 57). – The Output Buffer is forced tristate. – The data present on the I/O pin is sampled into the Input Latch at the beginning of each instruction execution. – The data stored in the Output Master Latch is copied into the Output Slave Latch at the end of the execution of each instruction. Thus, if bit Px.n is reconfigured as an Output or Bidirectional, the data stored in the Output Slave Latch will be reflected on the I/O pin. 99/224 ST92163 - I/O PORTS INPUT/OUTPUT BIT CONFIGURATION (Cont’d) Figure 55. Control Bits Bit 7 Bit n Bit 0 PxC2 PxC27 PxC2n PxC20 PxC1 PxC17 PxC1n PxC10 PxC0 PxC07 PxC0n PxC00 n Table 18. Port Bit Configuration Table (n = 0, 1... 7; X = port number) General Purpose I/O Pins PXC2n PXC1n PXC0n 0 0 0 1 0 0 0 1 0 1 1 0 PXn Configuration BID BID OUT OUT PXn Output Type WP OD OD PP OD TTL TTL TTL TTL CMOS TTL TTL TTL PXn Input Type (or Schmitt (or Schmitt (or Schmitt (or Schmitt (or Schmitt (or Schmitt (or Schmitt (or Schmitt Trigger) Tr igger) Tri gger) Trigger) Tr igger) Trigger) Trigger) Trigger) (1) For A/D Converter inputs. Legend: X = n = AF = BID = CMOS= HI-Z = IN = OD = OUT = PP = TTL = WP = 100/224 Port Bit Alternate Function Bidirectional CMOS Standard Input Levels High Impedance Input Open Drain Output Push-Pull TTL Standard Input Levels Weak Pull-up 0 0 1 A/D Pins 1 0 1 0 1 1 1 1 1 1 1 1 IN IN AF OUT AF OUT AF IN HI-Z HI-Z PP OD HI-Z(1) Analog Input ST92163 - I/O PORTS INPUT/OUTPUT BIT CONFIGURATION (Cont’d) Figure 56. Basic Structure of an I/O Port Pin I/O PIN PUSH-PULL TRISTATE OPEN DRAIN WEAK PULL-UP TTL / CMOS (or Schmitt Trigger) TO PERIPHERAL INPUTS AND INTERRUPTS OUTPUT SLAVE LATCH FROM PERIPHERAL OUTPUT ALTERNATE FUNCTION INPUT BIDIRECTIONAL ALTERNATE FUNCTION OUTPUT INPUT OUTPUT BIDIRECTIONAL OUTPUT MASTER LATCH INPUT LATCH INTERNAL DATA BUS Figure 57. Input Configuration Figure 58. Output Configuration I/O PIN I/O PIN OPEN DRAIN PUSH-PULL TTL / CMOS TRISTATE (or Schmitt Trigger) TO PERIPHERAL INPUTS AND OUTPUT SLAVE LATCH TTL (or Schmitt Trigger) TO PERIPHERAL INPUTS AND OUTPUT SLAVE LATCH INTERRUPTS INTERRUPTS OUTPUT MASTER LATCH INPUT LATCH OUTPUT MASTER LATCH INTERNAL DATA BUS n n INPUT LATCH INTERNAL DATA BUS n 101/224 ST92163 - I/O PORTS INPUT/OUTPUT BIT CONFIGURATION (Cont’d) When Px.n is programmed as an Output: (Figure 58) – The Output Buffer is turned on in an Open-drain or Push-pull configuration. – The data stored in the Output Master Latch is copied both into the Input Latch and into the Output Slave Latch, driving the I/O pin, at the end of the execution of the instruction. When Px.n is programmed as Bidirectional: (Figure 59) – The Output Buffer is turned on in an Open-Drain or Weak Pull-up configuration (except when disabled in hardware). – The data present on the I/O pin is sampled into the Input Latch at the beginning of the execution of the instruction. – The data stored in the Output Master Latch is copied into the Output Slave Latch, driving the I/ O pin, at the end of the execution of the instruction. WARNING: Due to the fact that in bidirectional mode the external pin is read instead of the output latch, particular care must be taken with arithmetic/logic and Boolean instructions performed on a bidirectional port pin. These instructions use a read-modify-write sequence, and the result written in the port register depends on the logical level present on the external pin. This may bring unwanted modifications to the port output register content. For example: Port register content, 0Fh external port value, 03h (Bits 3 and 2 are externally forced to 0) A bset instruction on bit 7 will return: Port register content, 83h external port value, 83h (Bits 3 and 2 have been cleared). To avoid this situation, it is suggested that all operations on a port, using at least one bit in bidirectional mode, are performed on a copy of the port register, then transferring the result with a load instruction to the I/O port. When Px.n is programmed as a digital Alternate Function Output: (Figure 60) – The Output Buffer is turned on in an Open-Drain or Push-Pull configuration. 102/224 – The data present on the I/O pin is sampled into the Input Latch at the beginning of the execution of the instruction. – The signal from an on-chip function is allowed to load the Output Slave Latch driving the I/O pin. Signal timing is under control of the alternate function. If no alternate function is connected to Px.n, the I/O pin is driven to a high level when in Push-Pull configuration, and to a high impedance state when in open drain configuration. Figure 59. Bidirectional Configuration I/O PIN WEAK PULL-UP OPEN DRAIN TTL (or Schmitt Trigger) TO PERIPHERAL INPUTS AND OUTPUT SLAVE LATCH INTERRUPTS OUTPUT MASTER LATCH INPUT LATCH INTERNAL DATA BUS n n Figure 60. Alternate Function Configuration I/O PIN OPEN DRAIN PUSH-PULL TTL (or Schmitt Trigger) TO PERIPHERAL INPUTS AND OUTPUT SLAVE LATCH INTERRUPTS FROM PERIPHERAL OUTPUT INPUT LATCH INTERNAL DATA BUS n n n n n n ST92163 - I/O PORTS 7.5 ALTERNATE FUNCTION ARCHITECTURE Each I/O pin may be connected to three different types of internal signal: – Data bus Input/Output – Alternate Function Input – Alternate Function Output 7.5.1 Pin Declared as I/O A pin declared as I/O, is connected to the I/O buffer. This pin may be an Input, an Output, or a bidirectional I/O, depending on the value stored in (PxC2, PxC1 and PxC0). 7.5.2 Pin Declared as an Alternate Input A single pin may be directly connected to several Alternate inputs. In this case, the user must select the required input mode (with the PxC2, PxC1, PxC0 bits) and enable the selected Alternate Function in the Control Register of the peripheral. No specific port configuration is required to enable an Alternate Function input, since the input buffer is directly connected to each alternate function module on the shared pin. As more than one module can use the same input, it is up to the user software to enable the required module as necessary. Parallel I/Os remain operational even when using an Alternate Function input. The exception to this is when an I/O port bit is permanently assigned by hardware as an A/D bit. In this case , after software programming of the bit in AF-OD-TTL, the Alternate function output is forced to logic level 1. The analog voltage level on the corresponding pin is directly input to the A/D. 7.5.3 Pin Declared as an Alternate Function Output The user must select the AF OUT configuration using the PxC2, PxC1, PxC0 bits. Several Alternate Function outputs may drive a common pin. In such case, the Alternate Function output signals are logically ANDed before driving the common pin. The user must therefore enable the required Alternate Function Output by software. WARNING: When a pin is connected both to an alternate function output and to an alternate function input, it should be noted that the output signal will always be present on the alternate function input. 7.6 I/O STATUS AFTER WFI, HALT AND RESET The status of the I/O ports during the Wait For Interrupt, Halt and Reset operational modes is shown in the following table. The External Memory Interface ports are shown separately. If only the internal memory is being used and the ports are acting as I/O, the status is the same as shown for the other I/O ports. Mode WFI HALT RESET Ext. Mem - I/O Ports P0 P1, P2, P6 High Impedance or next address (depending on Next the last Address memory operation performed on Port) High ImpedNext ance Address I/O Ports Not Affected (clock outputs running) Not Affected (clock outputs stopped) Bidirectional Weak Alternate function push- Pull-up (High impull (ROMless device) pedance when disabled in hardware). 103/224 ST92163 - TIMER/WATCHDOG (WDT) 8 ON-CHIP PERIPHERALS 8.1 TIMER/WATCHDOG (WDT) Important Note: This chapter is a generic description of the WDT peripheral. However depending on the ST9 device, some or all of WDT interface signals described may not be connected to external pins. For the list of WDT pins present on the ST9 device, refer to the device pinout description in the first section of the data sheet. 8.1.1 Introduction The Timer/Watchdog (WDT) peripheral consists of a programmable 16-bit timer and an 8-bit prescaler. It can be used, for example, to: – Generate periodic interrupts – Measure input signal pulse widths – Request an interrupt after a set number of events – Generate an output signal waveform – Act as a Watchdog timer to monitor system integrity The main WDT registers are: – Control register for the input, output and interrupt logic blocks (WDTCR) – 16-bit counter register pair (WDTHR, WDTLR) – Prescaler register (WDTPR) The hardware interface consists of up to five signals: – WDIN External clock input – WDOUT Square wave or PWM signal output – INT0 External interrupt input – NMI Non-Maskable Interrupt input – HW0SW1 Hardware/Software Watchdog enable. Figure 61. Timer/Watchdog Block Diagram INEN INMD1 INMD2 INPUT WDIN1 & CLOCK CONTROL LOGIC MUX WDT CLOCK WDTPR 8-BIT PRESCALER WDTRH , WDTRL 16-BIT DOWNCOU NTER END OF COUNT INTCLK/4 OUTMD WROUT OUTEN OUTPUT CONTROL LOGIC NMI 1 INT0 1 WDOUT1 HW0SW 11 MUX WDGEN INTERRUPT IAOS TLIS CONTROL LOGIC RESET TOP LEVEL INTERRUPT REQUES T 1Pin not present on some ST9 devices. 104/224 INTA0 REQUEST ST92163 - TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 8.1.2 Functional Description 8.1.2.1 External Signals The HW0SW1 pin can be used to permanently enable Watchdog mode. Refer to Section 0.1.3.1. The WDIN Input pin can be used in one of four modes: – Event Counter Mode – Gated External Input Mode – Triggerable Input Mode – Retriggerable Input Mode The WDOUT output pin can be used to generate a square wave or a Pulse Width Modulated signal. An interrupt, generated when the WDT is running as the 16-bit Timer/Counter, can be used as a Top Level Interrupt or as an interrupt source connected to channel A0 of the external interrupt structure (replacing the INT0 interrupt input). The counter can be driven either by an external clock, or internally by INTCLK divided by 4. 8.1.2.2 Initialisation The prescaler (WDTPR) and counter (WDTRL, WDTRH) registers must be loaded with initial values before starting the Timer/Counter. If this is not done, counting will start with reset values. 8.1.2.3 Start/Stop The ST_SP bit enables downcounting. When this bit is set, the Timer will start at the beginning of the following instruction. Resetting this bit stops the counter. If the counter is stopped and restarted, counting will resume from the last value unless a new constant has been entered in the Timer registers (WDTRL, WDTRH). A new constant can be written in the WDTRH, WDTRL, WDTPR registers while the counter is running. The new value of the WDTRH, WDTRL registers will be loaded at the next End of Count (EOC) condition while the new value of the WDTPR register will be effective immediately. End of Count is when the counter is 0. When Watchdog mode is enabled the state of the ST_SP bit is irrelevant. 8.1.2.4 Single/Continuous Mode The S_C bit allows selection of single or continuous mode.This Mode bit can be written with the Timer stopped or running. It is possible to toggle the S_C bit and start the counter with the same instruction. Single Mode On reaching the End Of Count condition, the Timer stops, reloads the constant, and resets the Start/ Stop bit. Software can check the current status by reading this bit. To restart the Timer, set the Start/ Stop bit. Note: If the Timer constant has been modified during the stop period, it is reloaded at start time. Continuous Mode On reaching the End Of Count condition, the counter automatically reloads the constant and restarts. It is stopped only if the Start/Stop bit is reset. 8.1.2.5 Input Section If the Timer/Counter input is enabled (INEN bit) it can count pulses input on the WDIN pin. Otherwise it counts the internal clock/4. For instance, when INTCLK = 20MHz, the End Of Count rate is: 3.35 seconds for Maximum Count (Timer Const. = FFFFh, Prescaler Const. = FFh) 200 ns for Minimum Count (Timer Const. = 0000h, Prescaler Const. = 00h) The Input pin can be used in one of four modes: – Event Counter Mode – Gated External Input Mode – Triggerable Input Mode – Retriggerable Input Mode The mode is configurable in the WDTCR. 8.1.2.6 Event Counter Mode In this mode the Timer is driven by the external clock applied to the input pin, thus operating as an event counter. The event is defined as a high to low transition of the input signal. Spacing between trailing edges should be at least 8 INTCLK periods (or 400ns with INTCLK = 20MHz). Counting starts at the next input event after the ST_SP bit is set and stops when the ST_SP bit is reset. 105/224 ST92163 - TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 8.1.2.7 Gated Input Mode This mode can be used for pulse width measurement. The Timer is clocked by INTCLK/4, and is started and stopped by means of the input pin and the ST_SP bit. When the input pin is high, the Timer counts. When it is low, counting stops. The maximum input pin frequency is equivalent to INTCLK/8. 8.1.2.8 Triggerable Input Mode The Timer (clocked internally by INTCLK/4) is started by the following sequence: – setting the Start-Stop bit, followed by – a High to Low transition on the input pin. To stop the Timer, reset the ST_SP bit. 8.1.2.9 Retriggerable Input Mode In this mode, the Timer (clocked internally by INTCLK/4) is started by setting the ST_SP bit. A High to Low transition on the input pin causes counting to restart from the initial value. When the Timer is stopped (ST_SP bit reset), a High to Low transition of the input pin has no effect. 8.1.2.10 Timer/Counter Output Modes Output modes are selected by means of the OUTEN (Output Enable) and OUTMD (Output Mode) bits of the WDTCR register. No Output Mode (OUTEN = “0”) The output is disabled and the corresponding pin is set high, in order to allow other alternate functions to use the I/O pin. Square Wave Output Mode (OUTEN = “1”, OUTMD = “0”) The Timer outputs a signal with a frequency equal to half the End of Count repetition rate on the WDOUT pin. With an INTCLK frequency of 20MHz, this allows a square wave signal to be generated whose period can range from 400ns to 6.7 seconds. Pulse Width Modulated Output Mode (OUTEN = “1”, OUTMD = “1”) The state of the WROUT bit is transferred to the output pin (WDOUT) at the End of Count, and is held until the next End of Count condition. The user can thus generate PWM signals by modifying the status of the WROUT pin between End of Count events, based on software counters decremented by the Timer Watchdog interrupt. 106/224 8.1.3 Watchdog Timer Operation This mode is used to detect the occurrence of a software fault, usually generated by external interference or by unforeseen logical conditions, which causes the application program to abandon its normal sequence of operation. The Watchdog, when enabled, resets the MCU, unless the program executes the correct write sequence before expiry of the programmed time period. The application program must be designed so as to correctly write to the WDTLR Watchdog register at regular intervals during all phases of normal operation. 8.1.3.1 Hardware Watchdog/Software Watchdog The HW0SW1 pin (when available) selects Hardware Watchdog or Software Watchdog. If HW0SW1 is held low: – The Watchdog is enabled by hardware immediately after an external reset. (Note: Software reset or Watchdog reset have no effect on the Watchdog enable status). – The initial counter value (FFFFh) cannot be modified, however software can change the prescaler value on the fly. – The WDGEN bit has no effect. (Note: it is not forced low). If HW0SW1 is held high, or is not present: – The Watchdog can be enabled by resetting the WDGEN bit. 8.1.3.2 Starting the Watchdog In Watchdog mode the Timer is clocked by INTCLK/4. If the Watchdog is software enabled, the time base must be written in the timer registers before entering Watchdog mode by resetting the WDGEN bit. Once reset, this bit cannot be changed by software. If the Watchdog is hardware enabled, the time base is fixed by the reset value of the registers. Resetting WDGEN causes the counter to start, regardless of the value of the Start-Stop bit. In Watchdog mode, only the Prescaler Constant may be modified. If the End of Count condition is reached a System Reset is generated. ST92163 - TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 8.1.3.3 Preventing Watchdog System Reset In order to prevent a system reset, the sequence AAh, 55h must be written to WDTLR (Watchdog Timer Low Register). Once 55h has been written, the Timer reloads the constant and counting restarts from the preset value. To reload the counter, the two writing operations must be performed sequentially without inserting other instructions that modify the value of the WDTLR register between the writing operations. The maximum allowed time between two reloads of the counter depends on the Watchdog timeout period. 8.1.3.4 Non-Stop Operation In Watchdog Mode, a Halt instruction is regarded as illegal. Execution of the Halt instruction stops further execution by the CPU and interrupt acknowledgment, but does not stop INTCLK, CPUCLK or the Watchdog Timer, which will cause a System Reset when the End of Count condition is reached. Furthermore, ST_SP, S_C and the Input Mode selection bits are ignored. Hence, regardless of their status, the counter always runs in Continuous Mode, driven by the internal clock. The Output mode should not be enabled, since in this context it is meaningless. Figure 62. Watchdog Timer Mode COUNT VALUE TIMER START COUNTING RESET WRITEWDTRH,WDTRL WDGEN=0 WRITE AAh,55h INTOWDTRL PRODUCE COUNT RELOAD SOFTWAREFAIL (E.G. INFINITELOOP) ORPERIPHERALFAIL VA00220 107/224 ST92163 - TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 8.1.4 WDT Interrupts The Timer/Watchdog issues an interrupt request at every End of Count, when this feature is enabled. A pair of control bits, IA0S (EIVR.1, Interrupt A0 selection bit) and TLIS (EIVR.2, Top Level Input Selection bit) allow the selection of 2 interrupt sources (Timer/Watchdog End of Count, or External Pin) handled in two different ways, as a Top Level Non Maskable Interrupt (Software Reset), or as a source for channel A0 of the external interrupt logic. A block diagram of the interrupt logic is given in Figure 3. Note: Software traps can be generated by setting the appropriate interrupt pending bit. Table 1 below, shows all the possible configurations of interrupt/reset sources which relate to the Timer/Watchdog. A reset caused by the watchdog will set bit 6, WDGRES of R242 - Page 55 (Clock Flag Register). See section CLOCK CONTROL REGISTERS. Figure 63. Interrupt Sources TIMER WATC HDOG RESET WDGEN (WCR.6) 0 MUX INT0 INTA0 REQUEST 1 IA0S (EIVR .1) 0 TOP LEVEL INTERRUP T REQUEST MUX NMI 1 TLIS (EIVR.2) VA00293 Table 19. Interrupt Configuration Control Bits Enabled Sources Operating Mode WDGEN IA0S TLIS Reset INTA0 Top Level 0 0 0 0 0 0 1 1 0 1 0 1 WDG/Ext Reset WDG/Ext Reset WDG/Ext Reset WDG/Ext Reset SW TRAP SW TRAP Ext Pin Ext Pin SW TRAP Ext Pin SW TRAP Ext Pin Watchdog Watchdog Watchdog Watchdog 1 1 1 1 0 0 1 1 0 1 0 1 Timer Timer Ext Pin Ext Pin Timer Ext Pin Timer Ext Pin Timer Timer Timer Timer Ext Ext Ext Ext Reset Reset Reset Reset Legend: WDG = Watchdog function SW TRAP = Software Trap Note: If IA0S and TLIS = 0 (enabling the Watchdog EOC as interrupt source for both Top Level and INTA0 interrupts), only the INTA0 interrupt is taken into account. 108/224 ST92163 - TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 8.1.5 Register Description The Timer/Watchdog is associated with 4 registers mapped into Group F, Page 0 of the Register File. WDTHR: Timer/Watchdog High Register WDTLR: Timer/Watchdog Low Register WDTPR: Timer/Watchdog Prescaler Register WDTCR: Timer/Watchdog Control Register Three additional control bits are mapped in the following registers on Page 0: Watchdog Mode Enable, (WCR.6) Top Level Interrupt Selection, (EIVR.2) Interrupt A0 Channel Selection, (EIVR.1) Note: The registers containing these bits also contain other functions. Only the bits relevant to the operation of the Timer/Watchdog are shown here. Counter Register This 16 bit register (WDTLR, WDTHR) is used to load the 16 bit counter value. The registers can be read or written “on the fly”. TIMER/WATCHDOG HIGH REGISTER (WDTHR) R248 - Read/Write Register Page: 0 Reset value: 1111 1111 (FFh) 7 R15 0 R14 R13 R12 R11 R10 R9 R8 Bit 7:0 = R[15:8] Counter Most Significant Bits. TIMER/WATCHDOG LOW REGISTER (WDTLR) R249 - Read/Write Register Page: 0 Reset value: 1111 1111b (FFh) 7 R7 TIMER/WATCHDOG PRESCALER REGISTER (WDTPR) R250 - Read/Write Register Page: 0 Reset value: 1111 1111 (FFh) 7 0 PR7 PR6 PR5 PR4 PR3 PR2 PR1 PR0 Bit 7:0 = PR[7:0] Prescaler value. A programmable value from 1 (00h) to 256 (FFh). Warning: In order to prevent incorrect operation of the Timer/Watchdog, the prescaler (WDTPR) and counter (WDTRL, WDTRH) registers must be initialised before starting the Timer/Watchdog. If this is not done, counting will start with the reset (un-initialised) values. WATCHDOG TIMER CONTROL REGISTER (WDTCR) R251- Read/Write Register Page: 0 Reset value: 0001 0010 (12h) 7 0 ST_SP S_C INMD1 INMD2 INEN OUTMD WROUT OUTEN Bit 7 = ST_SP: Start/Stop Bit. This bit is set and cleared by software. 0: Stop counting 1: Start counting (see Warning above) Bit 6 = S_C: Single/Continuous . This bit is set and cleared by software. 0: Continuous Mode 1: Single Mode 0 R6 R5 R4 R3 R2 R1 R0 Bit 7:0 = R[7:0] Counter Least Significant Bits. Bit 5:4 = INMD[1:2]: Input mode selection bits. These bits select the input mode: INMD1 INMD2 INPUT MODE 0 0 Event Counter 0 1 Gated Input (Reset value) 1 0 Triggerable Input 1 1 Retriggerable Input 109/224 ST92163 - TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) Bit 3 = INEN: Input Enable. This bit is set and cleared by software. 0: Disable input section 1: Enable input section Note: This bit is ignored if the Hardware Watchdog option is enabled by pin HW0SW1 (if available). Bit 2 = OUTMD: Output Mode. This bit is set and cleared by software. 0: The output is toggled at every End of Count 1: The value of the WROUT bit is transferred to the output pin on every End Of Count if OUTEN=1. Bit 1 = WROUT: Write Out. The status of this bit is transferred to the Output pin when OUTMD is set; it is user definable to allow PWM output (on Reset WROUT is set). Bit 0 = OUTEN: Output Enable bit. This bit is set and cleared by software. 0: Disable output 1: Enable output WAIT CONTROL REGISTER (WCR) R252 - Read/Write Register Page: 0 Reset value: 0111 1111 (7Fh) 7 x 0 WDGEN x x x x x x Bit 6 = WDGEN: Watchdog Enable (active low). Resetting this bit via software enters the Watchdog mode. Once reset, it cannot be set anymore by the user program. At System Reset, the Watchdog mode is disabled. 110/224 EXTERNAL INTERRUPT VECTOR REGISTER (EIVR) R246 - Read/Write Register Page: 0 Reset value: xxxx 0110 (x6h) 7 x 0 x x x x TLIS IA0S x Bit 2 = TLIS: Top Level Input Selection. This bit is set and cleared by software. 0: Watchdog End of Count is TL interrupt source 1: NMI is TL interrupt source Bit 1 = IA0S: Interrupt Channel A0 Selection. This bit is set and cleared by software. 0: Watchdog End of Count is INTA0 source 1: External Interrupt pin is INTA0 source Warning: To avoid spurious interrupt requests, the IA0S bit should be accessed only when the interrupt logic is disabled (i.e. after the DI instruction). It is also necessary to clear any possible interrupt pending requests on channel A0 before enabling this interrupt channel. A delay instruction (e.g. a NOP instruction) must be inserted between the reset of the interrupt pending bit and the IA0S write instruction. Other bits are described in the Interrupt section. ST92163 - MULTIFUNCTION TIMER (MFT) 8.2 MULTIFUNCTION TIMER (MFT) 8.2.1 Introduction The Multifunction Timer (MFT) peripheral offers powerful timing capabilities and features 12 operating modes, including automatic PWM generation and frequency measurement. The MFT comprises a 16-bit Up/Down counter driven by an 8-bit programmable prescaler. The input clock may be INTCLK/3 or an external source. The timer features two 16-bit Comparison Registers, and two 16-bit Capture/Load/Reload Registers. Two input pins and two alternate function output pins are available. Several functional configurations are possible, for instance: – 2 input captures on separate external lines, and 2 independent output compare functions with the counter in free-running mode, or 1 output compare at a fixed repetition rate. – 1 input capture, 1 counter reload and 2 independent output compares. – 2 alternate autoreloads and 2 independent output compares. – 2 alternate captures on the same external line and 2 independent output compares at a fixed repetition rate. When two timers are present in an ST9 device, a combined operating mode is available. An internal On-Chip Event signal can be used on some devices to control other on-chip peripherals. The two external inputs may be individually programmed to detect any of the following: – rising edges – falling edges – both rising and falling edges Figure 64. MFT Simplified Block Diagram 111/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) The configuration of each input is programmed in the Input Control Register. Each of the two output pins can be driven from any of three possible sources: – Compare Register 0 logic – Compare Register 1 logic – Overflow/Underflow logic Each of these three sources can cause one of the following four actions, independently, on each of the two outputs: – Nop, Set, Reset, Toggle In addition, an additional On-Chip Event signal can be generated by two of the three sources mentioned above, i.e. Over/Underflow event and Compare 0 event. This signal can be used internally to Figure 65. Detailed Block Diagram 112/224 synchronise another on-chip peripheral. Five maskable interrupt sources referring to an End Of Count condition, 2 input captures and 2 output compares, can generate 3 different interrupt requests (with hardware fixed priority), pointing to 3 interrupt routine vectors. Two independent DMA channels are available for rapid data transfer operations. Each DMA request (associated with a capture on the REG0R register, or with a compare on the CMP0R register) has priority over an interrupt request generated by the same source. A SWAP mode is also available to allow high speed continuous transfers (see Interrupt and DMA chapter). ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.2 Functional Description The MFT operating modes are selected by programming the Timer Control Register (TCR) and the Timer Mode Register (TMR). 8.2.2.1 Trigger Events A trigger event may be generated by software (by setting either the CP0 or the CP1 bits in the T_FLAGR register) or by an external source which may be programmed to respond to the rising edge, the falling edge or both by programming bits A0A1 and B0-B1 in the T_ICR register. This trigger event can be used to perform a capture or a load, depending on the Timer mode (configured using the bits in Table 4). An event on the TxINA input or setting the CP0 bit triggers a capture to, or a load from the REG0R register (except in Bicapture mode, see Section 0.1.2.11). An event on the TxINB input or setting the CP1 bit triggers a capture to, or a load from the REG1R register. In addition, in the special case of ”Load from REG0R and monitor on REG1R”, it is possible to use the TxINB input as a trigger for REG0R.” 8.2.2.2 One Shot Mode When the counter generates an overflow (in upcount mode), or an underflow (in down-count mode), that is to say when an End Of Count condition is reached, the counter stops and no counter reload occurs. The counter may only be restarted by an external trigger on TxINA or B or a by software trigger on CP0 only. One Shot Mode is entered by setting the CO bit in TMR. 8.2.2.3 Continuous Mode Whenever the counter reaches an End Of Count condition, the counting sequence is automatically restarted and the counter is reloaded from REG0R (or from REG1R, when selected in Biload Mode). Continuous Mode is entered by resetting the C0 bit in TMR. 8.2.2.4 Triggered And Retriggered Modes A triggered event may be generated by software (by setting either the CP0 or the CP1 bit in the T_FLAGR register), or by an external source which may be programmed to respond to the rising edge, the falling edge or both, by programming bits A0-A1 and B0-B1 in T_ICR. In One Shot and Triggered Mode, every trigger event arriving before an End Of Count, is masked. In One Shot and Retriggered Mode, every trigger received while the counter is running, automatically reloads the counter from REG0R. Triggered/Retriggered Mode is set by the REN bit in TMR. The TxINA input refers to REG0R and the TxINB input refers to REG1R. WARNING. If the Triggered Mode is selected when the counter is in Continuous Mode, every trigger is disabled, it is not therefore possible to synchronise the counting cycle by hardware or software. 8.2.2.5 Gated Mode In this mode, counting takes place only when the external gate input is at a logic low level. The selection of TxINA or TxINB as the gate input is made by programming the IN0-IN3 bits in T_ICR. 8.2.2.6 Capture Mode The REG0R and REG1R registers may be independently set in Capture Mode by setting RM0 or RM1 in TMR, so that a capture of the current count value can be performed either on REG0R or on REG1R, initiated by software (by setting CP0 or CP1 in the T_FLAGR register) or by an event on the external input pins. WARNING. Care should be taken when two software captures are to be performed on the same register. In this case, at least one instruction must be present between the first CP0/CP1 bit set and the subsequent CP0/CP1 bit reset instructions. 8.2.2.7 Up/Down Mode The counter can count up or down depending on the state of the UDC bit (Up/Down Count) in TCR, or on the configuration of the external input pins, which have priority over UDC (see Input pin assignment in T_ICR). The UDCS bit returns the counter up/down current status (see also the Up/ Down Autodiscrimination mode in the Input Pin Assignment Section). 113/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.2.8 Free Running Mode The timer counts continuously (in up or down mode) and the counter value simply overflows or underflows through FFFFh or zero; there is no End Of Count condition as such, and no reloading takes place. This mode is automatically selected either in Bicapture Mode or by setting REG0R for a capture function (Continuous Mode must also be set). In Autoclear Mode, free running operation can be had, with the possibility of choosing a maximum count value before overflow or underflow which is less than 216 (see Autoclear Mode). 8.2.2.9 Monitor Mode When the RM1 bit in TMR is reset, and the timer is not in Bivalue Mode, REG1R acts as a monitor, duplicating the current up or down counter contents, thus allowing the counter to be read “on the fly”. 8.2.2.10 Autoclear Mode A clear command forces the counter either to 0000h or to FFFFh, depending on whether upcounting or downcounting is selected. The counter reset may be obtained either directly, through the CCL bit in TCR, or by entering the Autoclear Mode, through the CCP0 and CCMP0 bits in TCR. Every capture performed on REG0R (if CCP0 is set), or every successful compare performed by CMP0R (if CCMP0 is set), clears the counter and reloads the prescaler. 114/224 The Clear On Capture mode allows direct measurement of delta time between successive captures on REG0R, while the Clear On Compare mode allows free running with the possibility of choosing a maximum count value before overflow or underflow which is less than 216 (see Free Running Mode). 8.2.2.11 Bivalue Mode Depending on the value of the RM0 bit in TMR, the Biload Mode (RM0 reset) or the Bicapture Mode (RM0 set) can be selected as illustrated in Figure 1 below: Table 20. Bivalue Modes RM0 0 1 TMR bits RM1 X X BM 1 1 Timer Operating Modes BiLoad mode BiCapture Mode A) Biload Mode The Biload Mode is entered by selecting the Bivalue Mode (BM set in TMR) and programming REG0R as a reload register (RM0 reset in TMR). At any End Of Count, counter reloading is performed alternately from REG0R and REG1R, (a low level for BM bit always sets REG0R as the current register, so that, after a Low to High transition of BM bit, the first reload is always from REG0R). ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) Every software or external trigger event on REG0R performs a reload from REG0R resetting the Biload cycle. In One Shot mode (reload initiated by software or by an external trigger), reloading is always from REG0R. B) Bicapture Mode The Bicapture Mode is entered by selecting the Bivalue Mode (the BM bit in TMR is set) and by programming REG0R as a capture register (the RM0 bit in TMR is set). Every capture event, software simulated (by setting the CP0 flag) or coming directly from the TxINA input line, captures the current counter value alternately into REG0R and REG1R. When the BM bit is reset, REG0R is the current register, so that the first capture, after resetting the BM bit, is always into REG0R. 8.2.2.12 Parallel Mode When two timers are present on an ST9 device, the parallel mode is entered when the ECK bit in the TMR register of Timer 1 is set. The Timer 1 prescaler input is internally connected to the Timer 0 prescaler output. Timer 0 prescaler input is connected to the system clock line. By loading the Prescaler Register of Timer 1 with the value 00h the two timers (Timer 0 and Timer 1) are driven by the same frequency in parallel mode. In this mode the clock frequency may be divided by a factor in the range from 1 to 216. 8.2.2.13 Autodiscriminator Mode The phase difference sign of two overlapping pulses (respectively on TxINB and TxINA) generates a one step up/down count, so that the up/down control and the counter clock are both external. The setting of the UDC bit in the TCR register has no effect in this configuration. Figure 66. Parallel Mode Description INTCLK/ 3 PRESCALER 0 TIMER 0 PRESCALER 1 TIMER 1 (See note) Note: Timer 1 is not available on all devices. Refer to the device block diagram and register map. 115/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.3 Input Pin Assignment The two external inputs (TxINA and TxINB) of the timer can be individually configured to catch a particular external event (i.e. rising edge, falling edge, or both rising and falling edges) by programming the two relevant bits (A0, A1 and B0, B1) for each input in the external Input Control Register (T_ICR). The 16 different functional modes of the two external inputs can be selected by programming bits IN0 - IN3 of the T_ICR, as illustrated in Figure 2 Table 21. Input Pin Function I C Reg. IN3-IN0 bits 0000 0001 0010 0011 0100 0101 0110 0111 1000 1001 1010 1011 1100 1101 1110 1111 TxINA Input Function not used not used Gate Gate not used Trigger Gate Trigger Clock Up Up/Down Trigger Up Up/Down Autodiscr. Trigger Ext. Clock Trigger TxINB Input Function not used Trigger not used Trigger Ext. Clock not used Ext. Clock Trigger Clock Down Ext. Clock Trigger Down not used Autodiscr. Ext. Clock Trigger Gate Some choices relating to the external input pin assignment are defined in conjunction with the RM0 and RM1 bits in TMR. For input pin assignment codes which use the input pins as Trigger Inputs (except for code 1010, Trigger Up:Trigger Down), the following conditions apply: 116/224 – a trigger signal on the TxINA input pin performs an U/D counter load if RM0 is reset, or an external capture if RM0 is set. – a trigger signal on the TxINB input pin always performs an external capture on REG1R. The TxINB input pin is disabled when the Bivalue Mode is set. Note: For proper operation of the External Input pins, the following must be observed: – the minimum external clock/trigger pulse width must not be less than the system clock (INTCLK) period if the input pin is programmed as rising or falling edge sensitive. – the minimum external clock/trigger pulse width must not be less than the prescaler clock period (INTCLK/3) if the input pin is programmed as rising and falling edge sensitive (valid also in Auto discrimination mode). – the minimum delay between two clock/trigger pulse active edges must be greater than the prescaler clock period (INTCLK/3), while the minimum delay between two consecutive clock/ trigger pulses must be greater than the system clock (INTCLK) period. – the minimum gate pulse width must be at least twice the prescaler clock period (INTCLK/3). – in Autodiscrimination mode, the minimum delay between the input pin A pulse edge and the edge of the input pin B pulse, must be at least equal to the system clock (INTCLK) period. – if a number, N, of external pulses must be counted using a Compare Register in External Clock mode, then the Compare Register must be loaded with the value [X +/- (N-1)], where X is the starting counter value and the sign is chosen depending on whether Up or Down count mode is selected. ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.3.1 TxINA = I/O - TxINB = I/O Input pins A and B are not used by the Timer. The counter clock is internally generated and the up/ down selection may be made only by software via the UDC (Software Up/Down) bit in the TCR register. 8.2.3.2 TxINA = I/O - TxINB = Trigger The signal applied to input pin B acts as a trigger signal on REG1R register. The prescaler clock is internally generated and the up/down selection may be made only by software via the UDC (Software Up/Down) bit in the TCR register. 8.2.3.3 TxINA = Gate - TxINB = I/O The signal applied to input pin A acts as a gate signal for the internal clock (i.e. the counter runs only when the gate signal is at a low level). The counter clock is internally generated and the up/down control may be made only by software via the UDC (Software Up/Down) bit in the TCR register. 8.2.3.4 TxINA = Gate - TxINB = Trigger Both input pins A and B are connected to the timer, with the resulting effect of combining the actions relating to the previously described configurations. 8.2.3.5 TxINA = I/O - TxINB = Ext. Clock The signal applied to input pin B is used as the external clock for the prescaler. The up/down selection may be made only by software via the UDC (Software Up/Down) bit in the TCR register. 8.2.3.6 TxINA = Trigger - TxINB = I/O The signal applied to input pin A acts as a trigger for REG0R, initiating the action for which the register was programmed (i.e. a reload or capture). The prescaler clock is internally generated and the up/down selection may be made only by software via the UDC (Software Up/Down) bit in the TCR register. (*) The timer is in One shot mode and REGOR in Reload mode 8.2.3.7 TxINA = Gate - TxINB = Ext. Clock The signal applied to input pin B, gated by the signal applied to input pin A, acts as external clock for the prescaler. The up/down control may be made only by software action through the UDC bit in the TCR register. 8.2.3.8 TxINA = Trigger - TxINB = Trigger The signal applied to input pin A (or B) acts as trigger signal for REG0R (or REG1R), initiating the action for which the register has been programmed. The counter clock is internally generated and the up/down selection may be made only by software via the UDC (Software Up/Down) bit in the TCR register. 117/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.3.9 TxINA = Clock Up - TxINB = Clock Down The edge received on input pin A (or B) performs a one step up (or down) count, so that the counter clock and the up/down control are external. Setting the UDC bit in the TCR register has no effect in this configuration, and input pin B has priority on input pin A. 8.2.3.10 TxINA = Up/Down - TxINB = Ext Clock An High (or Low) level applied to input pin A sets the counter in the up (or down) count mode, while the signal applied to input pin B is used as clock for the prescaler. Setting the UDC bit in the TCR register has no effect in this configuration. 118/224 8.2.3.11 TxINA = Trigger Up - TxINB = Trigger Down Up/down control is performed through both input pins A and B. A edge on input pin A sets the up count mode, while a edge on input pin B (which has priority on input pin A) sets the down count mode. The counter clock is internally generated, and setting the UDC bit in the TCR register has no effect in this configuration. 8.2.3.12 TxINA = Up/Down - TxINB = I/O An High (or Low) level of the signal applied on input pin A sets the counter in the up (or down) count mode. The counter clock is internally generated. Setting the UDC bit in the TCR register has no effect in this configuration. ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.3.13 Autodiscrimination Mode The phase between two pulses (respectively on input pin B and input pin A) generates a one step up (or down) count, so that the up/down control and the counter clock are both external. Thus, if the rising edge of TxINB arrives when TxINA is at a low level, the timer is incremented (no action if the rising edge of TxINB arrives when TxINA is at a high level). If the falling edge of TxINB arrives when TxINA is at a low level, the timer is decremented (no action if the falling edge of TxINB arrives when TxINA is at a high level). Setting the UDC bit in the TCR register has no effect in this configuration. 8.2.3.14 TxINA = Trigger - TxINB = Ext. Clock The signal applied to input pin A acts as a trigger signal on REG0R, initiating the action for which the register was programmed (i.e. a reload or cap- ture), while the signal applied to input pin B is used as the clock for the prescaler. (*) The timer is in One shot mode and REG0R in reload mode 8.2.3.15 TxINA = Ext. Clock - TxINB = Trigger The signal applied to input pin B acts as a trigger, performing a capture on REG1R, while the signal applied to input pin A is used as the clock for the prescaler. 8.2.3.16 TxINA = Trigger - TxINB = Gate The signal applied to input pin A acts as a trigger signal on REG0R, initiating the action for which the register was programmed (i.e. a reload or capture), while the signal applied to input pin B acts as a gate signal for the internal clock (i.e. the counter runs only when the gate signal is at a low level). 119/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.4 Output Pin Assignment Two external outputs are available when programmed as Alternate Function Outputs of the I/O pins. Two registers Output A Control Register (OACR) and Output B Control Register (OBCR) define the driver for the outputs and the actions to be performed. Each of the two output pins can be driven from any of the three possible sources: – Compare Register 0 event logic – Compare Register 1 event logic – Overflow/Underflow event logic. Each of these three sources can cause one of the following four actions on any of the two outputs: – Nop – Set – Reset – Toggle Furthermore an On Chip Event signal can be driven by two of the three sources: the Over/Underflow event and Compare 0 event by programming the CEV bit of the OACR register and the OEV bit of OBCR register respectively. This signal can be used internally to synchronise another on-chip peripheral. Output Waveforms Depending on the programming of OACR and OBCR, the following example waveforms can be generated on TxOUTA and TxOUTB pins. For a configuration where TxOUTA is driven by the Over/Underflow (OUF) and the Compare 0 event (CM0), and TxOUTB is driven by the Over/Underflow and Compare 1 event (CM1): 120/224 OACR is programmed with TxOUTA preset to “0”, OUF sets TxOUTA, CM0 resets TxOUTA and CM1 does not affect the output. OBCR is programmed with TxOUTB preset to “0”, OUF sets TxOUTB, CM1 resets TxOUTB while CM0 does not affect the output. OACR = [101100X0] OBCR = [111000X0] T0OUTA OUF COMP0 OUF COMP0 COMP1 COMP1 T0OUTB OUF OUF For a configuration where TxOUTA is driven by the Over/Underflow, by Compare 0 and by Compare 1; TxOUTB is driven by both Compare 0 and Compare 1. OACR is programmed with TxOUTA preset to “0”. OUF toggles Output 0, as do CM0 and CM1. OBCR is programmed with TxOUTB preset to “1”. OUF does not affect the output; CM0 resets TxOUTB and CM1 sets it. OACR = [010101X0] OBCR = [100011X1] COMP1 COMP1 T0OUTA OUF OUF COMP0 COMP0 COMP1 COMP1 T0OUTB COMP0 COMP0 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) For a configuration where TxOUTA is driven by the Over/Underflow and by Compare 0, and TxOUTB is driven by the Over/Underflow and by Compare 1. OACR is programmed with TxOUTA preset to “0”. OUF sets TxOUTA while CM0 resets it, and CM1 has no effect. OBCR is programmed with TxOUTB preset to “1”. OUF toggles TxOUTB, CM1 sets it and CM0 has no effect. Output Waveform Samples In Biload Mode TxOUTA is programmed to monitor the two time intervals, t1 and t2, of the Biload Mode, while TxOUTB is independent of the Over/Underflow and is driven by the different values of Compare 0 and Compare 1. OACR is programmed with TxOUTA preset to “0”. OUF toggles the output and CM0 and CM1 do not affect TxOUTA. OBCR is programmed with TxOUTB preset to “0”. OUF has no effect, while CM1 resets TxOUTB and CM0 sets it. Depending on the CM1/CM0 values, three different sample waveforms have been drawn based on the above mentioned configuration of OBCR. In the last case, with a different programmed value of OBCR, only Compare 0 drives TxOUTB, toggling the output. For a configuration where TxOUTA is driven by the Over/Underflow and by Compare 0, and TxOUTB is driven by Compare 0 and 1. OACR is programmed with TxOUTA preset to “0”. OUF sets TxOUTA, CM0 resets it and CM1 has no effect. OBCR is programmed with TxOUTB preset to “0”. OUF has no effect, CM0 sets TxOUTB and CM1 toggles it. OACR = [101100X0] OBCR = [000111X0] T0OUTA OUF COMP0 OUF COMP0 COMP1 COMP1 T0OUTB COMP0 COMP0 Note (*) Depending on the CMP1R/CMP0R values 121/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.5 Interrupt and DMA 8.2.5.1 Timer Interrupt The timer has 5 different Interrupt sources, belonging to 3 independent groups, which are assigned to the following Interrupt vectors: Table 22. Timer Interrupt Structure Interrupt Source COMP 0 COMP 1 CAPT 0 CAPT 1 Overflow/Underflow Vector Address xxxx x110 xxxx x100 xxxx x000 The three least significant bits of the vector pointer address represent the relative priority assigned to each group, where 000 represents the highest priority level. These relative priorities are fixed by hardware, according to the source which generates the interrupt request. The 5 most significant bits represent the general priority and are programmed by the user in the Interrupt Vector Register (T_IVR). Each source can be masked by a dedicated bit in the Interrupt/DMA Mask Register (IDMR) of each timer, as well as by a global mask enable bit (IDMR.7) which masks all interrupts. If an interrupt request (CM0 or CP0) is present before the corresponding pending bit is reset, an overrun condition occurs. This condition is flagged in two dedicated overrun bits, relating to the Comp0 and Capt0 sources, in the Timer Flag Register (T_FLAGR). 8.2.5.2 Timer DMA Two Independent DMA channels, associated with Comp0 and Capt0 respectively, allow DMA transfers from Register File or Memory to the Comp0 Register, and from the Capt0 Register to Register File or Memory). If DMA is enabled, the Capt0 and Comp0 interrupts are generated by the corresponding DMA End of Block event. Their priority is set by hardware as follows: – Compare 0 Destination — Lower Priority – Capture 0 Source — Higher Priority The two DMA request sources are independently maskable by the CP0D and CM0D DMA Mask bits in the IDMR register. 122/224 The two DMA End of Block interrupts are independently enabled by the CP0I and CM0I Interrupt mask bits in the IDMR register. 8.2.5.3 DMA Pointers The 6 programmable most significant bits of the DMA Counter Pointer Register (DCPR) and of the DMA Address Pointer Register (DAPR) are common to both channels (Comp0 and Capt0). The Comp0 and Capt0 Address Pointers are mapped as a pair in the Register File, as are the Comp0 and Capt0 DMA Counter pair. In order to specify either the Capt0 or the Comp0 pointers, according to the channel being serviced, the Timer resets address bit 1 for CAPT0 and sets it for COMP0, when the D0 bit in the DCPR register is equal to zero (Word address in Register File). In this case (transfers between peripheral registers and memory), the pointers are split into two groups of adjacent Address and Counter pairs respectively. For peripheral register to register transfers (selected by programming “1” into bit 0 of the DCPR register), only one pair of pointers is required, and the pointers are mapped into one group of adjacent positions. The DMA Address Pointer Register (DAPR) is not used in this case, but must be considered reserved. Figure 67. Pointer Mapping for Transfers between Registers and Memory Register File Address Pointers Comp0 16 bit Addr Pointer Capt0 16 bit Addr Pointer DMA Counters Comp0 DMA 16 bit Counter Capt0 DMA 16 bit Counter YYYYYY11(l) YYYYYY10(h) YYYYYY01(l) YYYYYY00(h) XXXXXX11(l) XXXXXX10(h) XXXXXX01(l) XXXXXX00(h) ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) Figure 68. Pointer Mapping for Register to Register Transfers Register File 8 bit Counter XXXXXX11 8 bit Addr Pointer XXXXXX10 8 bit Counter XXXXXX01 8 bit Addr Pointer XXXXXX00 Compare 0 Capture 0 8.2.5.4 DMA Transaction Priorities Each Timer DMA transaction is a 16-bit operation, therefore two bytes must be transferred sequentially, by means of two DMA transfers. In order to speed up each word transfer, the second byte transfer is executed by automatically forcing the peripheral priority to the highest level (000), regardless of the previously set level. It is then restored to its original value after executing the transfer. Thus, once a request is being serviced, its hardware priority is kept at the highest level regardless of the other Timer internal sources, i.e. once a Comp0 request is being serviced, it maintains a higher priority, even if a Capt0 request occurs between the two byte transfers. 8.2.5.5 DMA Swap Mode After a complete data table transfer, the transaction counter is reset and an End Of Block (EOB) condition occurs, the block transfer is completed. The End Of Block Interrupt routine must at this point reload both address and counter pointers of the channel referred to by the End Of Block interrupt source, if the application requires a continuous high speed data flow. This procedure causes speed limitations because of the time required for the reload routine. The SWAP feature overcomes this drawback, allowing high speed continuous transfers. Bit 2 of the DMA Counter Pointer Register (DCPR) and of the DMA Address Pointer Register (DAPR), toggles after every End Of Block condition, alternately providing odd and even address (D2-D7) for the pair of pointers, thus pointing to an updated pair, after a block has been completely transferred. This allows the User to update or read the first block and to update the pointer values while the second is being transferred. These two toggle bits are software writable and readable, mapped in DCPR bit 2 for the CM0 channel, and in DAPR bit 2 for the CP0 channel (though a DMA event on a channel, in Swap mode, modifies a field in DAPR and DCPR common to both channels, the DAPR/ DCPR content used in the transfer is always the bit related to the correct channel). SWAP mode can be enabled by the SWEN bit in the IDCR Register. WARNING: Enabling SWAP mode affects both channels (CM0 and CP0). 123/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 8.2.5.6 DMA End Of Block Interrupt Routine An interrupt request is generated after each block transfer (EOB) and its priority is the same as that assigned in the usual interrupt request, for the two channels. As a consequence, they will be serviced only when no DMA request occurs, and will be subject to a possible OUF Interrupt request, which has higher priority. The following is a typical EOB procedure (with swap mode enabled): – Test Toggle bit and Jump. – Reload Pointers (odd or even depending on toggle bit status). – Reset EOB bit: this bit must be reset only after the old pair of pointers has been restored, so that, if a new EOB condition occurs, the next pair of pointers is ready for swapping. – Verify the software protection condition (see Section 0.1.5.7). – Read the corresponding Overrun bit: this confirms that no DMA request has been lost in the meantime. – Reset the corresponding pending bit. – Reenable DMA with the corresponding DMA mask bit (must always be done after resetting the pending bit) 124/224 – Return. WARNING: The EOB bits are read/write only for test purposes. Writing a logical “1” by software (when the SWEN bit is set) will cause a spurious interrupt request. These bits are normally only reset by software. 8.2.5.7 DMA Software Protection A second EOB condition may occur before the first EOB routine is completed, this would cause a not yet updated pointer pair to be addressed, with consequent overwriting of memory. To prevent these errors, a protection mechanism is provided, such that the attempted setting of the EOB bit before it has been reset by software will cause the DMA mask on that channel to be reset (DMA disabled), thus blocking any further DMA operation. As shown above, this mask bit should always be checked in each EOB routine, to ensure that all DMA transfers are properly served. 8.2.6 Register Description Note: In the register description on the following pages, register and page numbers are given using the example of Timer 0. On devices with more than one timer, refer to the device register map for the adresses and page numbers. ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) CAPTURE LOAD 0 HIGH REGISTER (REG0HR) R240 - Read/Write Register Page: 10 Reset value: undefined 7 R15 R14 R13 R12 R11 R10 R9 0 7 R8 R15 This register is used to capture values from the Up/Down counter or load preset values (MSB). CAPTURE LOAD 0 LOW REGISTER (REG0LR) R241 - Read/Write Register Page: 10 Reset value: undefined 7 COMPARE 0 HIGH REGISTER (CMP0HR) R244 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 0 0 R14 R13 R12 R11 R10 R9 R8 This register is used to store the MSB of the 16-bit value to be compared to the Up/Down counter content. COMPARE 0 LOW REGISTER (CMP0LR) R245 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 R7 R6 R5 R4 R3 R2 R1 R7 This register is used to capture values from the Up/Down counter or load preset values (LSB). CAPTURE LOAD 1 HIGH REGISTER (REG1HR) R242 - Read/Write Register Page: 10 Reset value: undefined 7 R15 0 R14 R13 R12 R11 R10 R9 R8 This register is used to capture values from the Up/Down counter or load preset values (MSB). CAPTURE LOAD 1 LOW REGISTER (REG1LR) R243 - Read/Write Register Page: 10 Reset value: undefined 7 R7 0 R6 R5 R4 R3 R2 R1 0 R0 R0 R6 R5 R4 R3 R2 R1 This register is used to store the LSB of the 16-bit value to be compared to the Up/Down counter content. COMPARE 1 HIGH REGISTER (CMP1HR) R246 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 R15 0 R14 R13 R12 R11 R10 R9 R8 This register is used to store the MSB of the 16-bit value to be compared to the Up/Down counter content. COMPARE 1 LOW REGISTER (CMP1LR) R247 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 This register is used to capture values from the Up/Down counter or load preset values (LSB). R0 R7 0 R6 R5 R4 R3 R2 R1 R0 This register is used to store the LSB of the 16-bit value to be compared to the Up/Down counter content. 125/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) TIMER CONTROL REGISTER (TCR) R248 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 CEN 0 CCP CCMP CCL UDC 0 0 UDC S OF0 CS Bit 7 = CEN: Counter enable. This bit is ANDed with the Global Counter Enable bit (GCEN) in the CICR register (R230). The GCEN bit is set after the Reset cycle. 0: Stop the counter and prescaler 1: Start the counter and prescaler (without reload). Note: Even if CEN=0, capture and loading will take place on a trigger event. Bit 6 = CCP0: Clear on capture. 0: No effect 1: Clear the counter and reload the prescaler on a REG0R or REG1R capture event Bit 5 = CCMP0: Clear on Compare. 0: No effect 1: Clear the counter and reload the prescaler on a CMP0R compare event Bit 4 = CCL: Counter clear. This bit is reset by hardware after being set by software (this bit always returns “0” when read). 0: No effect 1: Clear the counter without generating an interrupt request 126/224 Bit 3 = UDC: Up/Down software selection. If the direction of the counter is not fixed by hardware (TxINA and/or TxINB pins, see par. 10.3) it can be controlled by software using the UDC bit. 0: Down counting 1: Up counting Bit 2 = UDCS: Up/Down count status. This bit is read only and indicates the direction of the counter. 0: Down counting 1: Up counting Bit 1 = OF0: OVF/UNF state. This bit is read only. 0: No overflow or underflow occurred 1: Overflow or underflow occurred during a Capture on Register 0 Bit 0 = CS Counter Status. This bit is read only and indicates the status of the counter. 0: Counter halted 1: Counter running ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) TIMER MODE REGISTER (TMR) R249 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) Bit 3 = RM0: REG0R mode. This bit works together with the BM and RM1 bits to select the timer operating mode. Refer to Table 4. 7 OE1 0 OE0 BM RM1 RM0 ECK REN C0 Bit 7 = OE1: Output 1 enable. 0: Disable the Output 1 (TxOUTB pin) and force it high. 1: Enable the Output 1 (TxOUTB pin) The relevant I/O bit must also be set to Alternate Function Bit 6 = OE0: Output 0 enable. 0: Disable the Output 0 (TxOUTA pin) and force it high 1: Enable the Output 0 (TxOUTA pin). The relevant I/O bit must also be set to Alternate Function Bit 5 = BM: Bivalue mode. This bit works together with the RM1 and RM0 bits to select the timer operating mode (see Table 4). 0: Disable bivalue mode 1: Enable bivalue mode Bit 4 = RM1: REG1R mode. This bit works together with the BM and RM0 bits to select the timer operating mode. Refer to Table 4. Note: This bit has no effect when the Bivalue Mode is enabled (BM=1). Table 23. Timer Operating Modes TMR Bits Timer Operating Modes BM RM1 RM0 1 x 0 Biload mode 1 x 1 Bicapture mode 0 0 0 0 1 0 0 0 1 0 1 1 Load from REG0R and Monitor on REG1R Load from REG0R and Capture on REG1R Capture on REG0R and Monitor on REG1R Capture on REG0R and REG1R Bit 2 = ECK Timer clock control. 0: The prescaler clock source is selected depending on the IN0 - IN3 bits in the T_ICR register 1: Enter Parallel mode (for Timer 1 and Timer 3 only, no effect for Timer 0 and 2). See Section 0.1.2.12. Bit 1 = REN: Retrigger mode. 0: Enable retriggerable mode 1: Disable retriggerable mode Bit 0 = CO: Continous/One shot mode. 0: Continuous mode (with autoreload on End of Count condition) 1: One shot mode 127/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) EXTERNAL INPUT CONTROL (T_ICR) R250 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) REGISTER 7 Bit 1:0 = B[0:1]: TxINB Pin event. These bits are set and cleared by software. B0 0 0 1 1 0 IN3 IN2 IN1 IN0 A0 A1 B0 IN[3:0] bits 0000 0001 0010 0011 0100 0101 0110 0111 1000 1001 1010 1011 1100 1101 1110 1111 TxINB Input Pin Function not used Trigger not used Trigger Ext. Clock not used Ext. Clock Trigger Clock Down Ext. Clock Trigger Down not used Autodiscr. Ext. Clock Trigger Gate Bit 3:2 = A[0:1]: TxINA Pin event. These bits are set and cleared by software. A0 0 0 1 1 128/224 A1 0 1 0 1 TxINA Pin Event No operation Falling edge sensitive Rising edge sensitive Rising and falling edges TxINB Pin Event No operation Falling edge sensitive Rising edge sensitive Rising and falling edges B1 Bit 7:4 = IN[3:0]: Input pin function. These bits are set and cleared by software. TxINA Pin Function not used not used Gate Gate not used Trigger Gate Trigger Clock Up Up/Down Trigger Up Up/Down Autodiscr. Trigger Ext. Clock Trigger B1 0 1 0 1 PRESCALER REGISTER (PRSR) R251 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 P7 0 P6 P5 P4 P3 P2 P1 P0 This register holds the preset value for the 8-bit prescaler. The PRSR content may be modified at any time, but it will be loaded into the prescaler at the following prescaler underflow, or as a consequence of a counter reload (either by software or upon external request). Following a RESET condition, the prescaler is automatically loaded with 00h, so that the prescaler divides by 1 and the maximum counter clock is generated (OSCIN frequency divided by 6 when MODER.5 = DIV2 bit is set). The binary value programmed in the PRSR register is equal to the divider value minus one. For example, loading PRSR with 24 causes the prescaler to divide by 25. ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) OUTPUT A CONTROL REGISTER (OACR) R252 - Read/Write Register Page: 10 Reset value: 0000 0000 7 0 C0E0 C0E1 C1E0 C1E1 OUE0 OUE1 CEV 0P Note: Whenever more than one event occurs simultaneously, the action taken will be the result of ANDing the event bits xxE1-xxE0. Bit 7:6 = C0E[0:1]: COMP0 event bits. These bits are set and cleared by software. Action on TxOUTA pin on a sucC0E0 C0E1 cessful compare of the CMP0R register 0 0 1 1 0 1 0 1 Set Toggle Reset NOP Bit 5:4 = C1E[0:1]: COMP1 event bits. These bits are set and cleared by software. C1E0 C1E1 0 0 1 1 0 1 0 1 Action on TxOUTA pin on a successful compare of the CMP1R register Set Toggle Reset NOP Bit 3:2 = OUE[0:1]: OVF/UNF event bits. These bits are set and cleared by software. OUE0 OUE1 0 0 1 1 0 1 0 1 Action on TxOUTA pin on an Overflow or Underflow on the U/D counter Set Toggle Reset NOP Note: Whenever more than one event occurs simultaneously, the action taken will be the result of ANDing the event xxE1-xxE0 bits. Bit 1 = CEV: On-Chip event on CMP0R. This bit is set and cleared by software. 0: No action 1: A successful compare on CMP0R activates the on-chip event signal (a single pulse is generated) Bit 0 = OP: TxOUTA preset value. This bit is set and cleared by software and by hardware. The value of this bit is the preset value of the TxOUTA pin. Reading this bit returns the current state of the TxOUTA pin (useful when it is selected in toggle mode). 129/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) OUTPUT B CONTROL REGISTER (OBCR) R253 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 OUE0 OUE1 0 0 1 1 0 1 0 1 0 C0E0 C0E1 C1E0 C1E1 OUE0 OUE1 OEV 0P Note: Whenever more than one event occurs simultaneously, the action taken will be the result of ANDing the event bits xxE1-xxE0. Bit 7:6 = C0E[0:1]: COMP0 event bits. These bits are set and cleared by software. Action on TxOUTB pin on a sucC0E0 C0E1 cessful compare of the CMP0R register 0 0 1 1 0 1 0 1 Set Toggle Reset NOP Bit 5:4 = C1E[0:1]: COMP1 event bits. These bits are set and cleared by software. C1E0 C1E1 0 0 1 1 0 1 0 1 130/224 Bit 3:2 = OUE[0:1]: OVF/UNF event bits. These bits are set and cleared by software. Action on TxOUTB pin on a successful compare of the CMP1R register Set Toggle Reset NOP Action on TxOUTB pin on an Overflow or Underflow on the U/D counter Set Toggle Reset NOP Bit 1 = OEV: On-Chip event on OVF/UNF. This bit is set and cleared by software. 0: No action 1: An underflow/overflow activates the on-chip event signal (a single pulse is generated) Bit 0 = OP: TxOUTB preset value. This bit is set and cleared by software and by hardware. The value of this bit is the preset value of the TxOUTB pin. Reading this bit returns the current state of the TxOUTB pin (useful when it is selected in toggle mode). ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) FLAG REGISTER (T_FLAGR) R254 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 CP0 GTIEN and CM1I bits in the IDMR register are set. The CM1 bit is cleared by software. 0: No Compare 1 event 1: Compare 1 event occurred 0 CP1 CM0 CM1 OUF OCP OCM 0 0 A0 Bit 7 = CP0: Capture 0 flag. This bit is set by hardware after a capture on REG0R register. An interrupt is generated depending on the value of the GTIEN, CP0I bits in the IDMR register and the A0 bit in the T_FLAGR register. The CP0 bit must be cleared by software. Setting by software acts as a software load/capture to/from the REG0R register. 0: No Capture 0 event 1: Capture 0 event occurred Bit 6 = CP1: Capture 1 flag . This bit is set by hardware after a capture on REG1R register. An interrupt is generated depending on the value of the GTIEN, CP0I bits in the IDMR register and the A0 bit in the T_FLAGR register. The CP1 bit must be cleared by software. Setting by software acts as a capture event on the REG1R register, except when in Bicapture mode. 0: No Capture 1 event 1: Capture 1 event occurred Bit 5 = CM0: Compare 0 flag. This bit is set by hardware after a successful compare on the CMP0R register. An interrupt is generated if the GTIEN and CM0I bits in the IDMR register are set. The CM0 bit is cleared by software. 0: No Compare 0 event 1: Compare 0 event occurred Bit 3 = OUF: Overflow/Underflow. This bit is set by hardware after a counter Over/ Underflow condition. An interrupt is generated if GTIEN and OUI=1 in the IDMR register. The OUF bit is cleared by software. 0: No counter overflow/underflow 1: Counter overflow/underflow Bit 2 = OCP0: Overrun on Capture 0. This bit is set by hardware when more than one INT/DMA requests occur before the CP0 flag is cleared by software or whenever a capture is simulated by setting the CP0 flag by software. The OCP0 flag is cleared by software. 0: No capture 0 overrun 1: Capture 0 overrun Bit 1 = OCM0: Overrun on compare 0. This bit is set by hardware when more than one INT/DMA requests occur before the CM0 flag is cleared by software.The OCM0 flag is cleared by software. 0: No compare 0 overrun 1: Compare 0 overrun Bit 0 = A0: Capture interrupt function. This bit is set and cleared by software. 0: Configure the capture interrupt as an OR function of REG0R/REG1R captures 1: Configure the capture interrupt as an AND function of REG0R/REG1R captures Bit 4 = CM1: Compare 1 flag. This bit is set after a successful compare on CMP1R register. An interrupt is generated if the 131/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) INTERRUPT/DMA MASK REGISTER (IDMR) R255 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 GTCP0D CP0I IEN 0 CP1I CM0 CM0I CM1I D OUI Bit 7 = GTIEN: Global timer interrupt enable. This bit is set and cleared by software. 0: Disable all Timer interrupts 1: Enable all timer Timer Interrupts from enabled sources Bit 6 = CP0D: Capture 0 DMA mask. This bit is set by software to enable a Capt0 DMA transfer and cleared by hardware at the end of the block transfer. 0: Disable capture on REG0R DMA 1: Enable capture on REG0R DMA Bit 5 = CP0I: Capture 0 interrupt mask. 0: Disable capture on REG0R interrupt 1: Enable capture on REG0R interrupt (or Capt0 DMA End of Block interrupt if CP0D=1) Bit 4 = CP1I: Capture 1 interrupt mask. This bit is set and cleared by software. 0: Disable capture on REG1R interrupt 1: Enable capture on REG1R interrupt Bit 3 = CM0D: Compare 0 DMA mask. This bit is set by software to enable a Comp0 DMA transfer and cleared by hardware at the end of the block transfer. 0: Disable compare on CMP0R DMA 1: Enable compare on CMP0R DMA Bit 2 = CM0I: Compare 0 Interrupt mask. This bit is set and cleared by software. 0: Disable compare on CMP0R interrupt 1: Enable compare on CMP0R interrupt (or Comp0 DMA End of Block interrupt if CM0D=1) 132/224 Bit 1 = CM1I: Compare 1 Interrupt mask. This bit is set and cleared by software. 0: Disable compare on CMP1R interrupt 1: Enable compare on CMP1R interrupt Bit 0 = OUI: Overflow/Underflow interrupt mask. This bit is set and cleared by software. 0: Disable Overflow/Underflow interrupt 1: Enable Overflow/Underflow interrupt DMA COUNTER POINTER REGISTER (DCPR) R240 - Read/Write Register Page: 9 Reset value: undefined 7 DCP7 DCP6 DCP5 DCP4 DCP3 DCP2 0 DMA REG/ SRCE MEM Bit 7:2 = DCP[7:2]: MSBs of DMA counter register address. These are the most significant bits of the DMA counter register address programmable by software. The DCP2 bit may also be toggled by hardware if the Timer DMA section for the Compare 0 channel is configured in Swap mode. Bit 1 = DMA-SRCE: DMA source selection. This bit is set and cleared by hardware. 0: DMA source is a Capture on REG0R register 1: DMA destination is a Compare on CMP0R register Bit 0 = REG/MEM: DMA area selection. This bit is set and cleared by software. It selects the source and destination of the DMA area 0: DMA from/to memory 1: DMA from/to Register File ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) DMA ADDRESS POINTER REGISTER (DAPR) R241 - Read/Write Register Page: 9 Reset value: undefined 7 DAP 7 0 DAP DMA PRG DAP5 DAP4 DAP3 DAP2 6 SRCE /DAT Bit 7:2 = DAP[7:2]: MSB of DMA address register location. These are the most significant bits of the DMA address register location programmable by software. The DAP2 bit may also be toggled by hardware if the Timer DMA section for the Compare 0 channel is configured in Swap mode. Note: During a DMA transfer with the Register File, the DAPR is not used; however, in Swap mode, DAPR(2) is used to point to the correct table. Bit 1 = DMA-SRCE: DMA source selection. This bit is fixed by hardware. 0: DMA source is a Capture on REG0R register 1: DMA destination is a Compare on the CMP0R register Bit 0 = PRG/DAT: DMA memory selection. This bit is set and cleared by software. It is only meaningful if DCPR.REG/MEM=0. 0: The ISR register is used to extend the address of data transferred by DMA (see MMU chapter). 1: The DMASR register is used to extend the address of data transferred by DMA (see MMU chapter). REG/MEM PRG/DAT DMA Source/Destination 0 0 ISR register used to address memory 0 1 DMASR register used to address memory 1 0 Register file 1 1 Register file INTERRUPT VECTOR REGISTER (T_IVR) R242 - Read/Write Register Page: 9 Reset value: xxxx xxx0 7 V4 0 V3 V2 V1 V0 W1 W0 0 This register is used as a vector, pointing to the 16-bit interrupt vectors in memory which contain the starting addresses of the three interrupt subroutines managed by each timer. Only one Interrupt Vector Register is available for each timer, and it is able to manage three interrupt groups, because the 3 least significant bits are fixed by hardware depending on the group which generated the interrupt request. In order to determine which request generated the interrupt within a group, the T_FLAGR register can be used to check the relevant interrupt source. Bit 7:3 = V[4:0]: MSB of the vector address. These bits are user programmable and contain the five most significant bits of the Timer interrupt vector addresses in memory. In any case, an 8-bit address can be used to indicate the Timer interrupt vector locations, because they are within the first 256 memory locations (see Interrupt and DMA chapters). Bit 2:1 = W[1:0]: Vector address bits. These bits are equivalent to bit 1 and bit 2 of the Timer interrupt vector addresses in memory. They are fixed by hardware, depending on the group of sources which generated the interrupt request as follows:. W1 0 0 1 1 W0 0 1 0 1 Interrupt Source Overflow/Underflow even interrupt Not available Capture event interrupt Compare event interrupt Bit 0 = This bit is forced by hardware to 0. 133/224 ST92163 - MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) INTERRUPT/DMA CONTROL REGISTER (IDCR) R243 - Read/Write Register Page: 9 Reset value: 1100 0111 (C7h) 7 CPE Bit 3 = SWEN: Swap function enable. This bit is set and cleared by software. 0: Disable Swap mode 1: Enable Swap mode for both DMA channels. 0 CME DCTS DCT SWE D N PL2 PL1 PL0 Bit 7 = CPE: Capture 0 EOB. This bit is set by hardware when the End Of Block condition is reached during a Capture 0 DMA operation with the Swap mode enabled. When Swap mode is disabled (SWEN bit = “0”), the CPE bit is forced to 1 by hardware. 0: No end of block condition 1: Capture 0 End of block Bit 2:0 = PL[2:0]: Interrupt/DMA priority level. With these three bits it is possible to select the Interrupt and DMA priority level of each timer, as one of eight levels (see Interrupt/DMA chapter). I/O CONNECTION REGISTER (IOCR) R248 - Read/Write Register Page: 9 Reset value: 1111 1100 (FCh) 7 Bit 6 = CME: Compare 0 EOB. This bit is set by hardware when the End Of Block condition is reached during a Compare 0 DMA operation with the Swap mode enabled. When the Swap mode is disabled (SWEN bit = “0”), the CME bit is forced to 1 by hardware. 0: No end of block condition 1: Compare 0 End of block Bit 5 = DCTS: DMA capture transfer source. This bit is set and cleared by software. It selects the source of the DMA operation related to the channel associated with the Capture 0. Note: The I/O port source is available only on specific devices. 0: REG0R register 1: I/O port. Bit 4 = DCTD: DMA compare transfer destination. This bit is set and cleared by software. It selects the destination of the DMA operation related to the channel associated with Compare 0. Note: The I/O port destination is available only on specific devices. 0: CMP0R register 1: I/O port 134/224 0 SC1 SC0 Bit 7:2 = not used. Bit 1 = SC1: Select connection odd. This bit is set and cleared by software. It selects if the TxOUTA and TxINA pins for Timer 1 and Timer 3 are connected on-chip or not. 0: T1OUTA / T1INA and T3OUTA/ T3INA unconnected 1: T1OUTA connected internally to T1INA and T3OUTA connected internally to T3INA Bit 0 = SC0: Select connection even. This bit is set and cleared by software. It selects if the TxOUTA and TxINA pins for Timer 0 and Timer 2 are connected on-chip or not. 0: T0OUTA / T0INA and T2OUTA/ T2INA unconnected 1: T0OUTA connected internally to T0INA and T2OUTA connected internally to T2INA Note: Timer 1 and 2 are available only on some devices. Refer to the device block diagram and register map. ST92163 - USB PERIPHERAL (USB) 8.3 USB PERIPHERAL (USB) 8.3.1 Introduction The USB Peripheral provides a full-speed function interface between the USB bus and the ST9 microcontroller. 8.3.2 Main Features ■ USB Specification Version 1.1 Compliant ■ Supports 8 device addresses ■ 16 software configurable endpoints supporting: • All USB transfer types (control, interrupt, bulk , and isochronous) • Burst-DMA transfers to up to 1K bytes RAM buffers ■ ■ ■ ■ USB Suspend/Resume operations On-Chip 3.3V Regulator On-Chip USB Transceiver Special functions on alternate output pins: • USB upstream port Output Enable signal (USBOE) • Start-Of-Frame pulse (SOFP) Two bit-mirror registers 8.3.3 Functional Description The USB interface is composed of the following blocks (see Figure 69) ■ SIE (Serial Interface Engine): implementing USB protocol layer (for a detailed description of USB protocol refer to chapter 7, 8 of the “Universal Serial Bus Specification”). The functions of this block include: synchronisation ■ ■ ■ ■ ■ pattern recognition, bit-stuffing, CRC generation and checking, PID verification/generation and handshake evaluation. It must interface with USB bus on one side and with the ST9 core on the other side. ST9 interface: this block is connected between the SIE and the ST9 microcontroller. Its purpose is to handle specific endpoint registers and provide interrupt and DMA servicing to the SIE. The unit transfers data from or to the memory, and uses the Register File to save/load pointer and counter values without CPU intervention, taking control of the ST9 buses (Burst-DMA interface). The ST9 Interface unit uses interrupts to require attention from the microcontroller at the end of data transmission/ reception, and to flag error conditions. Port Interface: this block contains the logic related to the USB physical port. The port interface also handles the resume detection. Frame Timer: its function is to monitor the Startof-Frame (SOF) timing points. It detects a global suspend (from the host) when no USB traffic has been seen for 3 ms. It also generates the SOFP output for any external device requiring Start-of-Frame synchronization. Port transceiver: containing differential and single-ended receivers and slew rate controlled output buffers. Figure 69. Block Diagram DM0 Endpoint Logic and Registers Transceiver + 3.3 V regulator Full-Speed SIE Burst DMA Controller Interrupt Controller ST9 bus DP0 SOFP USBOE LP_SUSP Port Interface SOF Timer WKUP15 WUIMU 135/224 ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) 8.3.3.1 DMA transfer DMA descriptors for each endpoint, located in the ST9 register file, indicate where the related memory buffer is located in RAM, how large the allocated buffer is and how many bytes must be transmitted. When a data transfer takes place, the USB-FS buffering data loaded in an internal 8 byte long FIFO buffer, and performing Burst-DMA transfers as appropriate. Then, if needed, the proper handshake answer is generated or expected, according to the direction of the transfer. At the end of the transaction, an interrupt is generated: using status registers and different interrupt vectors, the microcontroller can determine which endpoint was served, which type of transaction took place, if errors occurred (bit stuffing, format, CRC, protocol, missing ACK, over/underrun, etc...). 8.3.3.2 Structure and usage of DMA buffers Each endpoint has two DMA buffers (one for transmission and the other for reception) whose size may be up to 1023 bytes each. They can be placed anywhere in memory (internally or externally). For each endpoint, eight Register File locations are used: ADDRn_TH and ADDRn_TL: These registers point to the starting address of the memory buffer containing the data to be transmitted by endpoint n at the next IN token addressed to it. 136/224 COUNTn_TL and COUNTn_TH: These registers contain the number of bytes to be transmitted by endpoint n at the next IN token addressed to it. ADDRn_RL and ADDRn_RH: These registers point to the starting address of the memory buffer which will contain the data received by endpoint n at the next OUT/SETUP token addressed to it. COUNTn_RL and COUNTn_RH: These registers contain the allocated buffer size for endpoint n reception, setting the maximum number of bytes the related endpoint can receive with the next OUT/ SETUP transaction. Other register locations related to unsupported transfer directions or unused endpoints, are available to the user. Isochronous endpoints have a special way of handling DMA buffers. The relationship between register file locations and memory buffer areas is depicted in Figure 70. Each DMA buffer is used starting from the bottom, either during reception or transmission. The USB interface never changes the contents of memory locations adjacent to the DMA memory buffers; even if a packet bigger than the allocated buffer length is received (buffer overrun condition) the data will be copied in memory only up to the last available location. ST92163 - USB PERIPHERAL (USB) Figure 70. DMA buffers and related register file locations Register File R17 ADDR2_TL R16 ADDR2_TH R15 COUNT1_RL R14 COUNT1_RH R13 ADDR1_RL R12 ADDR1_RH R11 COUNT1_TL R10 COUNT1_TH R9 ADDR1_TL R8 ADDR1_TH R7 COUNT0_RL R6 COUNT0_RH R5 ADDR0_RL R4 ADDR0_RH R3 COUNT0_TL R2 COUNT0_TH R1 ADDR0_TL R0 ADDR0_TH REGISTER FILE LOCATIONS 8.3.3.3 Interrupt modes Two interrupt modes are available: single vector, where the application software has to poll the USB registers to determine which event required its attention, and multi-vector, where a specific interrupt vector is used for each endpoint. Special support is offered to isochronous transfers, implementing a swapped DMA buffer usage. TRANSMISSION BUFFER FOR ENDPOINT 1 RECEPTION BUFFER FOR ENDPOINT 0 TRANSMISSION BUFFER FOR ENDPOINT 0 MEMORY AREA 8.3.3.4 Suspend mode The unit can be placed in low-power mode (SUSPEND mode), by writing in a control register. At this time, all static power dissipation is avoided, except for the generation of the 3.3V supply for the external pull-up resistor. The detection of activity at the USB inputs while in low-power mode wakes the device up asynchronously. A special interrupt source, ESUSP, is connected directly to a wakeup line so as to allow the ST9 to exit from HALT condition. 137/224 ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) 8.3.4 Register Description USB registers can be divided into three groups: – Common registers (page 15): interrupt registers and USB control registers. – Function and endpoint registers (pages 15, 4 and 5 depending on how many endpoints are implemented): USB function addresses and endpoint status/configurations. – Extra registers (page 60): device configuration. 8.3.4.1 Common registers These registers affect all endpoints in the USB interface. They are all mapped in the same ST9 register page (page number 15). The USB interface implements vectorized interrupts: through a vector table it is possible to automatically identify the starting address of each Interrupt Service Routine. The vector table contains the 16-bit addresses pointing to each of the interrupt service routines related to the CTR interrupt for each endpoint. Other two 8-bit locations are used to store the address of the service routine handling the interrupts described in the USBISTR register. When an interrupt request is acknowledged, the USBIVR register provides a vector pointing to the location in the vector table, containing the start address of the interrupt service routine related to the serviced interrupt. INTERRUPT VECTOR REGISTER (USBIVR) R248 - Read/Write Register page: 15 Reset Value: xxxx xxx0 (xxh) 7 A1 0 A0 CTRO V3 V2 V1 V0 0 This register may be used in two different ways depending on the value of the SDNAV bit in the CTRL register. – If SDNAV bit = 1, Bits 7:1 are user programmable (bit 0 is fixed to 0). The software writes the address of a vector pointing to a single interrupt routine. The application program has to select the routine related to the pending interrupts using the USBISTR and CTRINF registers. 138/224 – If SDNAV = 0, this register is used as a vector pointing to the 16-bit interrupt vectors in program memory containing the start addresses of the interrupt service routines related to the occurred interrupt. If several interrupts are simultaneously pending, hardware writes in this register the interrupt routine address related to the highest priority pending interrupt. In this case the meaning of each bit is: Bits 7:6 = A[1:0]: Vector table Address. These two bits are user programmable and they contain the two most significant bits of the interrupt vector table. This allows the user to define the interrupt vector table position inside the first 256 locations of program memory at 64 bytes boundary. Bit 5 = CTRO: Correct Transfer interrupt occurred. 0: CTR interrupt pending 1: One of the interrupt flags in the USBISTR register is pending Note: If several interrupts are simultaneously pending, hardware writes this bit according to their relative priorities as listed below starting from the highest priority one to the lowest priority one: – DMA Over/Underrun (see Table 24 and USBISTR register description) – Correct Transfer on isochronous endpoints (see EPnRA register description) – Correct Transfer on non-isochronous endpoints (see EPnRA register description) – Notification events (see Table 24 and USBISTR register description). Bits 4:1 = V[3:1]: Endpoint Vector. If CTRO = 1, these bits are written by hardware to specify the endpoint identifier which has generated the CTR interrupt request. If several CTR interrupts are pending, hardware writes the endpoint identifier related to the endpoint with the highest priority. Endpoint priority is defined according to the following rule: endpoint 0 has the highest priority, then endpoint 1 follows and so on up to the highest endpoint register pair (EP15) with the lowest priority. If CTRO = 0, these bits are fixed to 1. In this case only one interrupt vector is used for all the interrupts defined in the USBISTR register. ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) Bit 0 = Reserved. This bit is fixed by hardware at 0. INTERRUPT STATUS REGISTER (USBISTR) R249 - Read/Write Register page: 15 Reset Value: 0000 0000 (00h) 7 0 0 DOVR ERR ESUSP SUSP RESET SOF ESOF Each bit is set by hardware when the related event occurs. If one of these bits is set, the hardware clears the CTRO bit and sets the V[3:0] bits in the USBIVR register. In this way the USBIVR register will point to the interrupt vector containing the address of the service routine related to these interrupt sources. If several bits are set only a single interrupt will be generated. Note: to avoid spurious clearing of some bits, it is recommended to clear them with a load instruction where all bits which must not be altered are set to 1, and all bits to be cleared are set to 0. Readmodify-write instructions like AND, XOR,... are to be avoided: consider the case of clearing bit 0 of USBISTR with an AND instruction, when only bit 7 of USBISTR is at 1 and the others at 0. First the microcontroller reads the content of USBISTR (=10h), then it clears bit 7 and writes the result (=00h) in USBISTR. If between the read and the write operations another bit were set by hardware (e.g. bit 5), writing 00h would clear it before the microprocessor has the time to service the event. Bit 7 = Reserved. This bit is fixed by hardware at 0. Bit 6 = DOVR: DMA over/underrun. ST9 processor has not been able to answer a DMA request in time and the USB FIFO buffer is full or empty depending on the transfer direction (reception or transmission). The USB handles this event in the following way: during reception the ACK handshake packet is not sent, during transmission a bit-stuffing error is forced on the transmitted stream. In both cases the host will retry the transaction. The DOVR interrupt should never occur during normal operations. Bit 5 = ERR: Error. One of the errors listed below has occurred: – NANS: No answer. The timeout for a host response has expired. – CRC: CRC error. One of the received CRCs, either in the token or in the data, was wrong. – BST: Bit Stuffing error. A bit stuffing error was detected anywhere in the PID, data, and/or CRC. – FVIO: Framing format violation. A nonstandard frame was received (EOP not in the right place, wrong token sequence, etc.). – BUFOVR: Buffer overrun. A packet longer than the allocated DMA buffer has been received. Table 24. Classification of Interrupt Sources: Interrupt DOVR ERR ESUSP SUSP RESET SOF ESOF Class DMA Over/Underrun event Notification event Notification event Notification event Notification event Notification event Notification event 139/224 ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) Bit 4 = ESUSP: End Suspend mode. 0: No activity detected during Suspend mode 1: USB activity is detected that wakes up the USB interface during suspend mode. Note: This event asynchronously clears the LP_SUSP bit in the USBCTLR register and activates the WKUP15 internal wake-up line to notify the WUIMU (if STOP_CK_EN=1 in the DEVCONF1 register) Bit 3 = SUSP Suspend mode request. 0: No Suspend mode request 1: No USB traffic has been received for 3 ms. Note: The suspend condition check is enabled immediately after any USB reset and is disabled by hardware when suspend mode is active (LP_SUSP = 1) until the end of resume sequence. Bit 2 = RESET: USB Reset request. 0: No USB Reset received. 1: USB Reset received. Note: Device address and endpoint registers are reset by an USB reset. Bit 1 = SOF: Start Of Frame. 0: No SOF packet received. 1: SOF packet received. INTERRUPT MASK REGISTER (USBIMR) R250 - Read/Write Register page: 15 Reset Value: 0000 0000 (00h) 0 0 DOVR M ERR M ESUSP M SUSP M RESET M SOF M ESOF M This register contains mask bits for all interrupt condition bits included in the USBISTR register. 140/224 INTERRUPT PRIORITY REGISTER (USBIPR) R251- Read/Write Register page: 15 Reset Value: 0xxx 0xxx (xxh) 7 IEE 0 PIECE PIECE PIECE 2 1 0 NIEE PNIEN PNIEN PNIEN 2 1 0 Bit 7 = IEE: Isochronous Endpoint Enable. Set by software to enable Isochronous Endpoints. This also enables CTR interrupts related to isochronous endpoints and DMA overrun interrupts. 0: Isochronous endpoints disabled 1: Isochronous endpoints enabled Bits 6:4 = PIECE[2:0]: Priority level on Isochronous Endpoint and DMA Over/Underrun. Set by software to define the priority level of the isochronous endpoint CTR events (0 is the highest priority). Bit 3 = NIEE: Non-Isochronous Endpoint Enable. Set by software to enable Non-Isochronous Endpoints. This also enables CTR interrupts related to non-isochronous (bulk, control, interrupt) endpoints. 0: Non-Isochronous endpoints disabled 1: Non-Isochronous endpoints enabled Bit 0 = ESOF: Expected Start Of Frame. 0: No SOF packet missed 1: SOF packet is expected but not received. 7 Whenever one of the USBIMR bits is 1, if the corresponding USBISTR bit is 1, an interrupt request is generated. For an explanation of each bit, refer to the USBISTR register description. Bits 2:0 = PNIEN[2:0]: Priority level of Non Isochronous Endpoints and Notification. Set by software to define the priority level of nonisochronous endpoint (bulk, control, interrupt) CTR events. 0 is the highest priority. ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) Interrupt sources are classified according to the following table CONTROL REGISTER (USBCTLR) R252 - Read/Write Register page: 15 Reset Value: 0001 0101 (15h) 7 0 0 TIM_ 0 SUSP SDNAV RESUME PDWN LP_ SUSP FRES Bit 7 = Reserved. This bit is fixed by hardware at 0. Bit 6 = TIM_SUSP: Timed Suspend. Set by software when the SUSP interrupt is received to enter “timed-suspend” state. 0: Timed suspend inactive 1: Timed suspend active. The USB interface operates as in suspend mode but clocks and static power dissipation in the analog transceiver are not stopped. Bit 3 = RESUME: Resume request. Set by software to send a Resume signal to the host. 0: Resume signal not forced on USB data lines. 1: Resume signal forced on USB data lines. Bit 2 = PDWN: Power Down. Set by software to turn off the 3.3V on-chip voltage regulator that supplies the external pull-up resistor and the transceiver. Note: As a consequence the voltage on both USB root port signal lines will drift to 0V because of the pull-down resistors in the upstream USB host or hub, generating a disconnect indication. At least 2 µs are required until the 3.3V supply falls within specifications after this bit is cleared, depending on the value of the external bypass capacitor. Bit 5 = Reserved. Bit 1 = LP_SUSP: Low-power suspend. Set by software to put the USB interface in “lowpower suspend” state. This condition should be entered while in “timed suspend” state (TIM_SUSP=1). 0: Low-power suspend inactive 1: Low-power suspend active Bit 4 = SDNAV: Single/Multiple Vector Selection. This bit is set by software to select the single or multiple interrupt vector mode . 0: Multiple interrupt vector mode 1: Single interrupt vector mode See USBIVR register description for detailed information. Bit 0 = FRES: Force USB Reset. Set by software to force a reset of the USB interface, just like a RESET signal on the USB. The USB interface is held in RESET state until software clears this bit, and a “USB-RESET” interrupt is generated, if enabled. 0: Reset not forced 1: USB interface reset forced 141/224 ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) CTR INTERRUPT FLAGS (CTRINF) R253 - Read/Write Register page: 15 Reset Value: 00xx xxx0 (xxh) 7 0 0 0 INTO ENID3 ENID2 ENID1 ENID0 0 Note: This register is used only when the SDNAV bit is 1. Bit 7:6 = Reserved. These bits are fixed by hardware at 0. Bit 5 = INTO: Interrupt occurred. Set by hardware when SDNAV = 1 in the same way as the CTRO bit in USBIVR register is set when SDNAV = 0. Bit 4:1 = ENID[3:0]: Endpoint identifier. Set by hardware when SDNAV = 1 in the same way as V[3:0] bits in USBIVR register are set when SDNAV = 0. Bit 0 = Reserved. This bit is fixed by hardware at 0. FRAME NUMBER REGISTER HIGH (FNRH) R254 - Read-only Register page: 15 Reset Value: 0000 0xxx (0xh) 7 RXDP RXDM LSOF1 LSOF0 FN10 FN9 Bits 2:0 = FN[10:8]: Frame Number bits 10:8. These bits contain the most significant bits of the Frame number received with the last received SOF packet. These bits are updated when a SOF interrupt is generated. FRAME NUMBER REGISTER LOW (FNRL) R255 - Read-only Register page: 15 Reset Value: xxxx xxxx (xxh) FN7 FN8 Bit 7 = RXDP. Set/cleared by hardware to indicate D+ upstream port data line status. Bit 6 = RXDM. Set/cleared by hardware to indicate D- upstream port data line status. 142/224 Bits 4:3 = LSOF[1:0]: Lost SOF . Set by hardware when an ESOF interrupt is generated, counting the number of consecutive SOF packets lost. On reception of a SOF packet, these bits are cleared. 7 0 LCK Bit 5 = LCK: Locked. Set by hardware when at least two consecutive SOF packets have been received after the end of a USB reset condition or after the end of a USB resume sequence. 0: Frame timer not locked 1: Frame timer locked Note: Once locked, the frame timer remains in this state until a USB reset or USB suspend event occurs. 0 FN6 FN5 FN4 FN3 FN2 FN1 FN0 Bits 7:0 = FN[7:0]: Frame Number bits 7:0. Set by hardware, this register contains the least significant bits of the 11-bit frame number contained in the last received SOF packet. The 3 remaining most significant bits are stored in the FNRH register. This register is updated when a SOF interrupt is generated. 8.3.4.2 Device and Endpoint specific Registers For each function a DADDR register is available to store the device address and for each endpoint a pair of EPnR registers is available to store endpoint specific information. ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) DEVICE n ADDRESS (DADDRn) R240 to R247 - Read/Write Register page: 15 Reset Value: 0000 0000 (00h) 7 EF 0 ADD6 ADD5 ADD4 ADD3 ADD2 ADD1 ADD0 Note: This register is also reset when a USB reset is received from the USB bus or forced through bit FRES in the USBCTLR register. Bit 7 = EF: Enable Function. Set by software to enable the USB function whose address is contained in the following ADD[6:0] bits. 0: Function disabled 1: Function enabled Bits 6:0 = ADD[6:0]: Device Address. Software must write into these bits the USB device address assigned by the host PC during the enumeration. ENDPOINT n REGISTER A (EPnRA) (TRANSMISSION) R240-R254 (even) Read/Write Register pages: 4 & 5 Reset value: 0000 0000 (00h) 7 CTR 0 DTOG_ STAT_ STAT_ PIDR1 PIDR0 TX TX1 TX0 CEP ISO These registers are used for controlling data transmission. They are also reset when a USB reset is received or forced through the FRES bit in the USBCTLR register. Note: The CTR bits are not affected by a USB reset Each endpoint has its EPnRA register where n is the endpoint identifier in the range 0 to 15. Bit 7 = CTR: Correct Transfer. Set by hardware when a transaction is successfully completed on this endpoint; software can only clear this bit. 0: No correct transfer occured 1: Correct transfer occured Note: A transaction ended with a NAK or STALL handshake does not set this bit, since no data is actually transferred, as in the case of protocol errors or data toggle mismatches. The CTR bit is also used to perform flow control: while CTR=1, a valid endpoint answers NAK to every transaction addressed to it (except SETUP requests which are simply ignored) until the application software acknowledges the CTR event, resetting this bit. This does not apply to isochronous endpoints where no handshake phase is used. Bit 6 = DTOG_TX: Data Toggle, for transmission transfers. If the endpoint is non-isochronous, this bit contains the required value of the data toggle bit (0=DATA0, 1=DATA1) for the next data packet to be transmitted. Hardware toggles this bit when the ACK handshake is received from the USB host, following a data packet transmission. If the endpoint is defined as a control one, hardware sets this bit to 1 on reception of a SETUP PID addressed to this endpoint. If the endpoint is isochronous, this bit is used to support DMA buffer swapping since no data toggling is used for this sort of endpoint and only DATA0 packet are transmitted. Hardware toggles this bit just after the end of data packet transmission, since no handshake is used for isochronous transfers. Note: this bit can be also written by software to initialize it (mandatory when the endpoint is not a control endpoint) or to force specific data toggle/ DMA buffer usage. 143/224 ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) Bit 5:4 = STAT_TX [1:0] Status bits, for transmission transfers. These bits contain the information about the endpoint status, which are listed below: Table 25. Transmission status encoding STAT_TX [1:0] 00 01 10 11 Meaning DISABLED: all transmission requests addressed to this endpoint are ignored. STALL: the endpoint is stalled and all transmission requests result in a STALL handshake. NAK: the endpoint is NAKed and all transmission requests result in a NAK handshake. VALID: this endpoint is enabled for transmission. These bits are written by software, but hardware sets the STAT_TX bits to NAK when a correct transfer has occurred (CTR=1) related to a IN or SETUP (control only) transaction addressed to this endpoint, waiting for the software to prepare the next set of data to be transmitted. If the endpoint is defined as isochronous, its status can be only “VALID” or “DISABLED” so no hardware change of the endpoint status will take place after a successful transaction. Bits 3:2 = PIDR[1:0]: PID Received. These bits are read-only and contain the two most significant bits of the PID field of the last token PID addressed to this endpoint. These bits are kept frozen while CTR bit is at 1. The USB standard defines PIDR bits as in the following table: Notes: If a control endpoint is defined as NAK in the receive direction, the USB interface will not answer, when a SETUP transaction is received. If the control endpoint is defined as STALL in the receive direction, then the SETUP packet will be accepted anyway, transferring data and issuing the CTR interrupt. Bit 0 = ISO: Isochronous endpoint. Software must set this bit to configure this endpoint as an isochronous endpoint. 0: Not an isochronous endpoint 1:isochronous endpoint Note: Since isochronous transfer has no handshake phase, the only legal values for the STAT_RX/STAT_TX bit pairs are ‘00’ (Disabled) and ‘11’ (Valid), any other value will produce results not compliant to the USB standard. Isochronous endpoints implement double-buffering, using both ‘transmission’ and ‘reception’ memory areas to manage buffer swapping on each successful transaction. The memory buffer that is currently used by the USB interface is defined by the DTOG bit corresponding to the endpoint direction (DTOG_RX in EPnRB for ‘reception’ isochronous endpoints, DTOG_TX in EPnRA for ‘transmission’ isochronous endpoints) according to the following table: Table 27. Isochronous memory buffers usage 0 Table 26. PID encoding PIDR[1:0] 00 10 11 PID OUT IN SETUP Bit 1 = CEP: Control Endpoint . Software must set this bit to configure this endpoint as a control endpoint. 0: Non-control endpoint 1: Control endpoint 144/224 DMA buffer used by application software ADDRn_T / ADDRn_R / COUNTn_T register COUNTn_R register file locations. file locations. ADDRn_R / ADDRn_T / COUNTn_R register COUNTn_T register file locations. file locations. DTOG bit DMA buffer used by value USB Interface 1 Since the swapped buffer management requires the usage of all 8 Register File locations hosting the address pointer and the length of the allocated memory buffers, isochronous endpoints are forced to be unidirectional so it is not possible to enable an isochronous endpoint both for transmission and reception. ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) ENDPOINT n REGISTER B (EPnRB) (RECEPTION) R241-R255 (odd) - Read/Write Register pages: 4 & 5 Reset value: 0000 0000 (00h) 7 ST_OU DTOG_ STAT_ STAT_ EnA3 T RX RX1 RX0 0 EnA2 EnA1 EnA0 These registers are used for controlling data reception. They are also reset when a USB reset is received from the USB bus or forced through bit FRES in the USBCTLR register. Each endpoint has its EPnRB register where n is the endpoint identifier in the range 0 to 15. since no handshake is used for isochronous transfers. Note: this bit can be also written by software to initialize it (mandatory when the endpoint is not a control endpoint) or to force a specific data toggle/ DMA buffer usage. Bit 5:4 = STAT_RX [1:0] Status bits, for reception transfers. These bits contain the information about the endpoint status, as listed below: Table 28. Reception status encoding STAT_TX [1:0] 00 01 Bit 7 = ST_OUT Status out. This bit is set by software to indicate that a status out transaction is expected: in this case all OUT transactions containing more than zero data bytes are answered STALL instead of ACK. This bit may be used to improve the robustness of the application to protocol errors during control transfers and its usage is intended on control endpoints only. When ST_OUT is reset, OUT transactions can have any number of bytes, as needed. Bit 6 = DTOG_RX: Data Toggle, for reception transfers. If the endpoint is non-isochronous, this bit contains the expected value of the data toggle bit (0=DATA0, 1=DATA1) for the next data packet to be received. Hardware toggles this bit when the ACK handshake is sent to the USB host, following a data packet reception with a matching data PID value. If the endpoint is defined as a control endpoint, hardware resets this bit on reception of a SETUP PID addressed to this endpoint. If the endpoint is isochronous, this bit is used to support DMA buffer swapping since no data toggling is used for this sort of endpoint and only DATA0 packets are received. Hardware toggles this bit just after the end of data packet reception, 10 11 Meaning DISABLED: all reception requests addressed to this endpoint are ignored. STALL: the endpoint is stalled and all reception requests result in a STALL handshake. NAK: the endpoint is naked and all reception requests result in a NAK handshake. VALID: this endpoint is enabled for reception. These bits are written by software, but hardware sets the STAT_RX bits to NAK when a correct transfer has occurred (CTR=1) related to a OUT or SETUP (control only) transaction addressed to this endpoint, so software has time to interpret the received data before acknowledging a new transaction. If the endpoint is defined as isochronous, its status can be only “VALID” or “DISABLED” so no hardware change of the endpoint status will take place after a successful transaction. Bit 3:0 = EnA[3:0] Endpoint n Address, bits 3-0. Software must write in this field the 4-bit address used to identify the transactions directed to this endpoint. A value must be written before enabling the corresponding endpoint. 145/224 ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) 8.3.4.3 Miscellaneous Registers These registers contain device configuration parameters or optional functions of USB interface. Bits 3:0 = Reserved. These bits are fixed by hardware at 1. DEVICE CONFIGURATION 2 (DEVCONF2) R245 - Read/Write Register page: 60 Reset value: 0000 0000 (00h) DEVICE CONFIGURATION 1 (DEVCONF1) R244 - Read/Write Register page: 60 Reset value: 0000 1111 (0Fh) 7 7 0 0 0 - USBOE LS_ _EN DEVICE 1 1 1 Bit 6 = Reserved. Must be kept at 0. Bit 5 = USBOE_EN: USBOE Signal Enable. Set by software to enable the alternate output function connected to the upstream output enable signal. 1: USBOE signal output enabled 0: USBOE signal output disabled Bit 4 = Reserved. 0 0 0 0 0 LVD_ DISAB LE SOFP_ ENABLE 1 This register contains device configuration data that should be written by the software at the beginning of the application program and should never be changed during normal operations. Bit 7:6 = Reserved. This bit is fixed by hardware at 0. 146/224 0 This register contains device configuration data that should be written by the software at the beginning of the application program and should never be changed during normal operations. Bits 7:2 = Reserved. These bits must be always written to 0. Bit 1 = LVD_DISABLE: Low Voltage Detector Disable Set by the software to disable the generation of the device reset signal by the LVD. 1: LVD disabled 0: LVD enabled If this bit is at 0, an internal reset can be issued both by the external RESET pin and the LVD when it detects a low supply voltage; if this bit is at 1 only the external pad can start a device reset. Note: this bit can be only set to 1 by software, writing 0 has no effect and the LVD can be re-enabled only using a device reset. ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) Bit 0 = SOFP_ENABLE: Start-Of-Frame Pulse Enable. Set by software to enable the use of the USBSOF alternate output function. 1: USBSOF alternate function enabled 0: USBSOF output disabled USBSOF outputs a 333.33 ns long pulse at each frame start (actually 10 bits before the SOF packet expected start). This pulse is not generated until two SOF packets are received after any USB reset or USB resume (frame timer locking) and they stop as soon as the suspend condition is entered (TIM_SUSP bit in USBCTLR register). MIRROR REGISTER B (MIRRB) R247 - Read/Write Register page: 60 Reset value: xxxx xxxx (xxh) 7 0 MIRB7 MIRB6 MIRB5 MIRB4 MIRB3 MIRB2 MIRB1 MIRB0 Bit 7:0 = MIRB[7:0] Mirror register B, bits 7-0. This register acts as a bit mirroring location where software can write a byte and read it back with the LSB and MSB position swapped. MIRROR REGISTER A (MIRRA) R246 - Read/Write Register page: 60 Reset value: xxxx xxxx (xxh) 7 0 MIRA7 MIRA6 MIRA5 MIRA4 MIRA3 MIRA2 MIRA1 MIRA0 Bit 7:0 = MIRA[7:0] Mirror register A, bits 7-0. This register acts as a bit mirroring location where software can write a byte and read it back with the LSB and MSB position swapped. 147/224 ST92163 - USB PERIPHERAL (USB) USB INTERFACE (Cont’d) 8.3.5 Register pages summary The registers are located in different register file pages. To help finding the correct page for each register, the following tables show the page mapping of all USB registers. Table 29. USB Register Page Mapping Address Page 4 Page 5 Page 15 Page 60 240 (F0h) EP0RA EP8RA DADDR0 241 (F1h) EP0RB EP8RB DADDR1 242 (F2h) EP1RA EP9RA DADDR2 243 (F3h) EP1RB EP9RB DADDR3 244 (F4h) EP2RA EP10RA DADDR4 DEVCONF1 245 (F5h) EP2RB EP10RB DADDR5 DEVCONF2 246 (F6h) EP3RA EP11RA DADDR6 MIRRA 247 (F7h) EP3RB EP11RB DADDR7 MIRRB 248 (F8h) EP4RA EP12RA USBIVR 249 (F9h) EP4RB EP12RB USBISTR 250 (FAh) EP5RA EP13RA USBIMR 251 (FBh) EP5RB EP13RB USBIPR 252 (FCh) EP6RA EP14RA USBCTLR 253 (FDh) EP6RB EP14RB CTRINF 254 (FEh) EP7RA EP15RA FNRH 255 (FFh) EP7RB EP15RB FNRL Table 30. USB Register page 15 Map and Reset Values Reg. No. Register Name R240 DADDRn R247 Reset Value R248 R249 R250 R251 R252 148/224 USBIVR 7 6 5 4 3 2 1 0 EF ADD6 ADD5 ADD4 ADD3 ADD2 ADD1 ADD0 0 0 0 0 0 0 0 0 A1 A0 CTRO V3 V2 V1 V0 0 Reset Value x x x x x x x 0 USBISTR 0 DOVR DOVR ESUSP SUSP RESET SOF ESOF Reset Value 0 0 0 0 0 0 0 0 USBIMR 0 DOVRM DOVRM ESUSPM SUSPM RESETM SOFM ESOFM Reset Value 0 0 0 0 0 0 0 0 USBIPR IEE PIECE2 PIECE1 PIECE0 NIEE PNIEN2 PNIEN1 PNIEN0 Reset Value 0 x x x 0 x x x USBCTLR 0 TIM_SUSP 0 SDNAV RESUME PDWN LP_SUSP FRES Reset Value 0 0 0 1 0 1 0 1 ST92163 - USB PERIPHERAL (USB) Reg. No. R253 R254 R255 Register Name 7 6 5 4 3 2 1 0 CTRINF 0 0 INTO ENID3 ENID2 ENID1 ENID0 0 Reset Value 0 0 x x x x x 0 RXDP RXDM LCK LSOF1 LSOF0 FN10 FN9 FN8 0 0 0 0 0 x x x FN7 FN6 FN5 FN4 FN3 FN2 FN1 FN0 x x x x x x x x FNRH Reset Value FNRL Reset Value 149/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) 8.4 SERIAL COMMUNICATIONS INTERFACE (SCI) 8.4.1 Introduction The Serial Communications Interface (SCI) offers full-duplex serial data exchange with a wide range of external equipment. The SCI offers four operating modes: Asynchronous, Asynchronous with synchronous clock, Serial expansion and Synchronous. The SCI offers the following principal features: ■ Full duplex synchronous and asynchronous operation. ■ Transmit, receive, line status, and device address interrupt generation. ■ Integral Baud Rate Generator capable of dividing the input clock by any value from 2 to 216-1 (16 bit word) and generating the internal 16X data sampling clock for asynchronous operation or the 1X clock for synchronous operation. ■ Fully programmable serial interface: – 5, 6, 7, or 8 bit word length. – Even, odd, or no parity generation and detection. – 0, 1, 1.5, 2, 2.5, 3 stop bit generation. – Complete status reporting capabilities. – Line break generation and detection. ■ ■ ■ ■ Programmable address indication bit (wake-up bit) and user invisible compare logic to support multiple microcomputer networking. Optional character search function. Internal diagnostic capabilities: – Local loopback for communications link fault isolation. – Auto-echo for communications link fault isolation. Separate interrupt/DMA channels for transmit and receive. In addition, a Synchronous mode supports: – High speed communication – Possibility of hardware synchronization (RTS/ DCD signals). – Programmable polarity and stand-by level for data SIN/SOUT. – Programmable active edge and stand-by level for clocks CLKOUT/RXCL. – Programmable active levels of RTS/DCD signals. – Full Loop-Back and Auto-Echo modes for DATA, CLOCKs and CONTROLs. Figure 71. SCI Block Diagram ST9 CORE BUS DMA CONTROLLER TRANSMIT BUFFER REGISTER TRANSMIT SHIFT REGISTER DMA CONTROLLER ADDRESS COMPARE REGISTER RECEIVER BUFFER REGISTER Frame Control and STATUS RECEIVER SHIFT REGISTER CLOCK and BAUD RATE GENERATOR ALTERNATE FUNCTION SOUT RTS SDS TXCLK/CLKOUT RXCLK 150/224 DCD SIN VA00169A ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.2 Functional Description The SCI offers four operating modes: – Asynchronous mode – Asynchronous mode with synchronous clock – Serial expansion mode – Synchronous mode Asynchronous mode, Asynchronous mode with synchronous clock and Serial expansion mode output data with the same serial frame format. The differences lie in the data sampling clock rates (1X, 16X) and in the protocol used. Figure 72. SCI Functional Schematic RX buffer register XBRG RXclk RX shift register Baud rate generator 1 Divider by 16 LBEN 0 CD XRX INPEN (*) Sin OUTPL (*) 1 Divider by 16 stand by polarity OCKPL (*) 0 CD OCLK TX buffer register DCDEN (*) AEN (*) TX shift register stand by polarity polarity LBEN (*) polarity INPL (*) INTCLK Sout AEN OUTSB (*) Enveloper OCKSB (*) OCLK Polarity Polarity XTCLK AEN (*) RTSEN (*) VR02054 TXclk / CLKout DCD RTS The control signals marked with (*) are active only in synchronous mode (SMEN=1) Note: Some pins may not be available on some devices. Refer to the device Pinout Description. 151/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.3 SCI Operating Modes 8.4.3.1 Asynchronous Mode In this mode, data and clock can be asynchronous (the transmitter and receiver can use their own clocks to sample received data), each data bit is sampled 16 times per clock period. The baud rate clock should be set to the ÷16 Mode and the frequency of the input clock (from an external source or from the internal baud-rate generator output) is set to suit. 8.4.3.2 Asynchronous Mode with Synchronous Clock In this mode, data and clock are synchronous, each data bit is sampled once per clock period. For transmit operation, a general purpose I/O port pin can be programmed to output the CLKOUT signal from the baud rate generator. If the SCI is provided with an external transmission clock source, there will be a skew equivalent to two INTCLK periods between clock and data. Data will be transmitted on the falling edge of the transmit clock. Received data will be latched into the SCI on the rising edge of the receive clock. Figure 73. Sampling Times in Asynchronous Format SDIN rcvck 0 1 2 3 4 5 7 8 9 10 11 12 13 14 15 rxd rxclk VR001409 LEGEND: Serial Data Input line SIN: rcvck: Internal X16 Receiver Clock Internal Serial Data Input Line rxd: rxclk: Internal Receiver Shift Register Sampling Clock 152/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.3.3 Serial Expansion Mode This mode is used to communicate with an external synchronous peripheral. The transmitter only provides the clock waveform during the period that data is being transmitted on the CLKOUT pin (the Data Envelope). Data is latched on the rising edge of this clock. Whenever the SCI is to receive data in serial port expansion mode, the clock must be supplied externally, and be synchronous with the transmitted data. The SCI latches the incoming data on the rising edge of the received clock, which is input on the RXCLK pin. 8.4.3.4 Synchronous Mode This mode is used to access an external synchronous peripheral, dummy start/stop bits are not included in the data frame. Polarity, stand-by level and active edges of I/O signals are fully and separately programmable for both inputs and outputs. It’s necessary to set the SMEN bit of the Synchronous Input Control Register (SICR) to enable this mode and all the related extra features (otherwise disabled). The transmitter will provide the clock waveform only during the period when the data is being transmitted via the CLKOUT pin, which can be enabled by setting both the XTCLK and OCLK bits of the Clock Configuration Register. Whenever the SCI is to receive data in synchronous mode, the clock waveform must be supplied externally via the RXCLK pin and be synchronous with the data. For correct receiver operation, the XRX bit of the Clock Configuration Register must be set. Two external signals, Request-To-Send and DataCarrier-Detect (RTS/DCD), can be enabled to synchronise the data exchange between two serial units. The RTS output becomes active just before the first active edge of CLKOUT and indicates to the target device that the MCU is about to send a synchronous frame; it returns to its stand-by state following the last active edge of CLKOUT (MSB transmitted). The DCD input can be considered as a gate that filters RXCLK and informs the MCU that a transmitting device is transmitting a data frame. Polarity of RTS/DCD is individually programmable, as for clocks and data. The data word is programmable from 5 to 8 bits, as for the other modes; parity, address/9th, stop bits and break cannot be inserted into the transmitted frame. Programming of the related bits of the SCI control registers is irrelevant in Synchronous Mode: all the corresponding interrupt requests must, in any case, be masked in order to avoid incorrect operation during data reception. 153/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Figure 74. SCI Operating Modes I/O 16 PARITY STOP BIT DATA START BIT 16 16 CLOCK I/O DATA PARITY STOP BIT START BIT CLOCK VA00271 VA00272 Asynchronous Mode I/O START BIT (Dummy) Asynchronous Mode with Synchronous Clock DATA STOP BIT (Dummy) CLOCK stand-by DATA stand-by stand-by CLOCK stand-by stand-by RTS/DCD stand-by VA0273A Serial Expansion Mode VR02051 Synchronous Mode Note: In all operating modes, the Least Significant Bit is transmitted/received first. 154/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.4 Serial Frame Format Characters sent or received by the SCI can have some or all of the features in the following format, depending on the operating mode: START: the START bit indicates the beginning of a data frame in Asynchronous modes. The START condition is detected as a high to low transition. A dummy START bit is generated in Serial Expansion mode. The START bit is not generated in Synchronous mode. DATA: the DATA word length is programmable from 5 to 8 bits, for both Synchronous and Asynchronous modes. LSB are transmitted first. PARITY: The Parity Bit (not available in Serial Expansion mode and Synchronous mode) is optional, and can be used with any word length. It is used for error checking and is set so as to make the total number of high bits in DATA plus PARITY odd or even, depending on the number of “1”s in the DATA field. ADDRESS/9TH: The Address/9th Bit is optional and may be added to any word format. It is used in both Serial Expansion and Asynchronous modes to indicate that the data is an address (bit set). The ADDRESS/9TH bit is useful when several microcontrollers are exchanging data on the same serial bus. Individual microcontrollers can stay idle on the serial bus, waiting for a transmitted address. When a microcontroller recognizes its own address, it can begin Data Reception, likewise, on the transmit side, the microcontroller can transmit another address to begin communication with a different microcontroller. The ADDRESS/9TH bit can be used as an additional data bit or to mark control words (9th bit). STOP: Indicates the end of a data frame in Asynchronous modes. A dummy STOP bit is generated in Serial Expansion mode. The STOP bit can be programmed to be 1, 1.5, 2, 2.5 or 3 bits long, depending on the mode. It returns the SCI to the quiescent marking state (i.e., a constant high-state condition) which lasts until a new start bit indicates an incoming word. The STOP bit is not generated in Synchronous mode. Figure 75. SCI Character Formats # bits START(2) DATA (1) PARITY (3) ADDRESS (2) STOP (2) 1 5, 6, 7, 8 0, 1 0, 1 1, 1.5, 2, 2.5, 1, 2, 3 NONE ODD EVEN ON OFF states 16X 1X (1) LSB First Not available in Synchronous mode (3) Not available in Serial Expansion mode and Synchronous mode (2) 155/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.4.1 Data transfer Data to be transmitted by the SCI is first loaded by the program into the Transmitter Buffer Register. The SCI will transfer the data into the Transmitter Shift Register when the Shift Register becomes available (empty). The Transmitter Shift Register converts the parallel data into serial format for transmission via the SCI Alternate Function output, Serial Data Out. On completion of the transfer, the transmitter buffer register interrupt pending bit will be updated. If the selected word length is less than 8 bits, the unused most significant bits do not need to be defined. Incoming serial data from the Serial Data Input pin is converted into parallel format by the Receiver Shift Register. At the end of the input data frame, the valid data portion of the received word is transferred from the Receiver Shift Register into the Receiver Buffer Register. All Receiver interrupt conditions are updated at the time of transfer. If the selected character format is less than 8 bits, the unused most significant bits will be set. The Frame Control and Status block creates and checks the character configuration (Data length and number of Stop bits), as well as the source of the transmitter/receiver clock. The internal Baud Rate Generator contains a programmable divide by “N” counter which can be used to generate the clocks for the transmitter and/or receiver. The baud rate generator can use INTCLK or the Receiver clock input via RXCLK. The Address bit/D9 is optional and may be added to any word in Asynchronous and Serial Expansion modes. It is commonly used in network or machine control applications. When enabled (AB set), an address or ninth data bit can be added to a transmitted word by setting the Set Address bit (SA). This is then appended to the next word entered into the (empty) Transmitter Buffer Register and then cleared by hardware. On character input, a set Address Bit can indicate that the data preceding the bit is an address which may be compared in hardware with the value in the Address Compare Register (ACR) to generate an Address Match interrupt when equal. The Address bit and Address Comparison Register can also be combined to generate four different types of Address Interrupt to suit different protocols, based on the status of the Address Mode Enable bit (AMEN) and the Address Mode bit (AM) in the CHCR register. The character match Address Interrupt mode may be used as a powerful character search mode, generating an interrupt on reception of a predetermined character e.g. Carriage Return or End of Block codes (Character Match Interrupt). This is the only Address Interrupt Mode available in Synchronous mode. The Line Break condition is fully supported for both transmission and reception. Line Break is sent by setting the SB bit (IDPR). This causes the transmitter output to be held low (after all buffered data has been transmitted) for a minimum of one complete word length and until the SB bit is Reset. Break cannot be inserted into the transmitted frame for the Synchronous mode. Testing of the communications channel may be performed using the built-in facilities of the SCI peripheral. Auto-Echo mode and Loop-Back mode may be used individually or together. In Asynchronous, Asynchronous with Synchronous Clock and Serial Expansion modes they are available only on SIN/SOUT pins through the programming of AEN/ LBEN bits in CCR. In Synchronous mode (SMEN set) the above configurations are available on SIN/ SOUT, RXCLK/CLKOUT and DCD/RTS pins by programming the AEN/LBEN bits and independently of the programmed polarity. In the Synchronous mode case, when AEN is set, the transmitter outputs (data, clock and control) are disconnected from the I/O pins, which are driven directly by the receiver input pins (Auto-Echo mode: SOUT=SIN, CLKOUT=RXCLK and RTS=DCD, even if they act on the internal receiver with the programmed polarity/edge). When LBEN is set, the receiver inputs (data, clock and controls) are disconnected and the transmitter outputs are looped-back into the receiver section (Loop-Back mode: SIN=SOUT, RXCLK=CLKOUT, DCD=RTS. The output pins are locked to their programmed stand-by level and the status of the INPL, XCKPL, DCDPL, OUTPL, OCKPL and RTSPL bits in the SICR register are irrelevant). Refer to Figure 6, Figure 7, and Figure 8 for these different configurations. Table 31. Address Interrupt Modes If 9th Data Bit is set (1) If Character Match If Character Match and 9th Data Bit is set(1) If Character Match Immediately Follows BREAK (1) (1) 156/224 Not available in Synchronous mode ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Figure 76. Auto Echo Configuration DCD TRANSMITTE R SOUT TRANS MITTER RTS RXCLK RECEIVER SOUT RECEIVER SIN SIN CLKOUT VR00210A VR000210 All modes except Synchronous Synchronous mode (SMEN=1) Figure 77. Loop Back Configuration DCD TRANSMITT ER LOGICAL 1 SOUT RTS stand-by value TRANSMIT TER clock RXCLK RECEIVER SIN CLKOUT stand-by value SOUT data RECEIVER SIN stand-by value VR00211A VR000211 All modes except Synchronous Synchronous mode (SMEN=1) Figure 78. Auto Echo and Loop-Back Configuration DCD TRANSMI TTER SOUT TRANSMIT TER RTS clock RECEIVE R SIN RXCLK SOUT data RECEIVER SIN CLKOUT VR000212 All modes except Synchronous VR00212A Synchronous mode (SMEN=1) 157/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.5 Clocks And Serial Transmission Rates The communication bit rate of the SCI transmitter and receiver sections can be provided from the internal Baud Rate Generator or from external sources. The bit rate clock is divided by 16 in Asynchronous mode (CD in CCR reset), or undivided in the 3 other modes (CD set). With INTCLK running at 24MHz and no external Clock provided, a maximum bit rate of 3MBaud and 750KBaud is available in undivided and divide by-16-mode respectively. With INTCLK running at 24MHz and an external Clock provided through the RXCLK/TXCLK lines, a maximum bit rate of 3MBaud and 375KBaud is avaiable in undivided and divided by 16 mode respectively (see Figure 10 ”Receiver and Transmitter Clock Frequencies”)” External Clock Sources. The External Clock input pin TXCLK may be programmed by the XTCLK and OCLK bits in the CCR register as: the transmit clock input, Baud Rate Generator output (allowing an external divider circuit to provide the receive clock for split rate transmit and receive), or as CLKOUT output in Synchronous and Serial Expansion modes. The RXCLK Receive clock input is enabled by the XRX bit, this input should be set in accordance with the setting of the CD bit. Baud Rate Generator. The internal Baud Rate Generator consists of a 16-bit programmable divide by “N” counter which can be used to generate the transmitter and/or receiver clocks. The minimum baud rate divisor is 2 and the maximum divisor is 216-1. After initialising the baud rate generator, the divisor value is immediately loaded into the counter. This prevents potentially long random counts on the initial load. The Baud Rate generator frequency is equal to the Input Clock frequency divided by the Divisor value. WARNING: Programming the baud rate divider to 0 or 1 will stop the divider. The output of the Baud Rate generator has a precise 50% duty cycle. The Baud Rate generator can use INTCLK for the input clock source. In this case, INTCLK (and therefore the MCU Xtal) should be chosen to provide a suitable frequency for division by the Baud Rate Generator to give the required transmit and receive bit rates. Suitable INTCLK frequencies and the respective divider values for standard Baud rates are shown in Table 2. 8.4.6 SCI Initialization Procedure Writing to either of the two Baud Rate Generator Registers immediately disables and resets the SCI baud rate generator, as well as the transmitter and receiver circuitry. After writing to the second Baud Rate Generator Register, the transmitter and receiver circuits are enabled. The Baud Rate Generator will load the new value and start counting. To initialize the SCI, the user should first initialize the most significant byte of the Baud Rate Generator Register; this will reset all SCI circuitry. The user should then initialize all other SCI registers (SICR/SOCR included) for the desired operating mode and then, to enable the SCI, he should initialize the least significant byte Baud Rate Generator Register. ’On-the-Fly’ modifications of the control registers’ content during transmitter/receiver operations, although possible, can corrupt data and produce undesirable spikes on the I/O lines (data, clock and control). Furthermore, modifying the control registers’ content without reinitialising the SCI circuitry (during stand-by cycles, waiting to transmit or receive data) must be kept carefully under control by software to avoid spurious data being transmitted or received. Note: For synchronous receive operation, the data and receive clock must not exhibit significant skew between clock and data. The received data and clock are internally synchronized to INTCLK. Figure 79. SCI Baud Rate Generator Initialization Sequence MOST SIGNIFICANT BYTE INITIALIZATIO N SELECT SCI WORKING MODE LEAST SIGNIFICANT BYTE INITIALIZATIO N 158/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Table 32. SCI Baud Rate Generator Divider Values Example 1 INTCLK: 19660.800 KHz Baud Rate Clock Factor Desired Freq (kHz) Divisor Dec Hex Actual Baud Rate Actual Freq (kHz) Deviation 50.00 16 X 0.80000 24576 6000 50.00 0.80000 0.0000% 75.00 16 X 1.20000 16384 4000 75.00 1.20000 0.0000% 110.00 16 X 1.76000 11170 2BA2 110.01 1.76014 -0.00081% 300.00 16 X 4.80000 4096 1000 300.00 4.80000 0.0000% 600.00 16 X 9.60000 2048 800 600.00 9.60000 0.0000% 1200.00 16 X 19.20000 1024 400 1200.00 19.20000 0.0000% 2400.00 16 X 38.40000 512 200 2400.00 38.40000 0.0000% 4800.00 16 X 76.80000 256 100 4800.00 76.80000 0.0000% 9600.00 16 X 153.60000 128 80 9600.00 153.60000 0.0000% 19200.00 16 X 307.20000 64 40 19200.00 307.20000 0.0000% 38400.00 16 X 614.40000 32 20 38400.00 614.40000 0.0000% 76800.00 16 X 1228.80000 16 10 76800.00 1228.80000 0.0000% Table 33. SCI Baud Rate Generator Divider Values Example 2 INTCLK: 24576 KHz Baud Rate Clock Factor Desired Freq (kHz) Divisor Dec Hex Actual Baud Rate Actual Freq (kHz) Deviation 50.00 16 X 0.80000 30720 7800 50.00 0.80000 0.0000% 75.00 16 X 1.20000 20480 5000 75.00 1.20000 0.0000% 110.00 16 X 1.76000 13963 383B 110.01 1.76014 -0.00046% 300.00 16 X 4.80000 5120 1400 300.00 4.80000 0.0000% 600.00 16 X 9.60000 2560 A00 600.00 9.60000 0.0000% 1200.00 16 X 19.20000 1280 500 1200.00 19.20000 0.0000% 2400.00 16 X 38.40000 640 280 2400.00 38.40000 0.0000% 4800.00 16 X 76.80000 320 140 4800.00 76.80000 0.0000% 9600.00 16 X 153.60000 160 A0 9600.00 153.60000 0.0000% 19200.00 16 X 307.20000 80 50 19200.00 307.20000 0.0000% 38400.00 16 X 614.40000 40 28 38400.00 614.40000 0.0000% 76800.00 16 X 1228.80000 20 14 76800.00 1228.80000 0.0000% 159/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.7 Input Signals SIN: Serial Data Input. This pin is the serial data input to the SCI receiver shift register. TXCLK: External Transmitter Clock Input. This pin is the external input clock driving the SCI transmitter. The TXCLK frequency must be greater than or equal to 16 times the transmitter data rate (depending whether the X16 or the X1 clock have been selected). A 50% duty cycle is required for this input and must have a period of at least twice INTCLK. The use of the TXCLK pin is optional. RXCLK: External Receiver Clock Input. This input is the clock to the SCI receiver when using an external clock source connected to the baud rate generator. INTCLK is normally the clock source. A 50% duty cycle is required for this input and must have a period of at least twice INTCLK. Use of RXCLK is optional. DCD: Data Carrier Detect. This input is enabled only in Synchronous mode; it works as a gate for the RXCLK clock and informs the MCU that an emitting device is transmitting a synchronous frame. The active level can be programmed as 1 or 0 and must be provided at least one INTCLK period before the first active edge of the input clock. 8.4.8 Output Signals SOUT: Serial Data Output. This Alternate Function output signal is the serial data output for the SCI transmitter in all operating modes. CLKOUT: Clock Output. The alternate Function of this pin outputs either the data clock from the transmitter in Serial Expansion or Synchronous modes, or the clock output from the Baud Rate Generator. In Serial expansion mode it will clock only the data portion of the frame and its stand-by state is high: data is valid on the rising edge of the clock. Even in Synchronous mode CLKOUT will only clock the data portion of the frame, but the stand-by level and active edge polarity are programmable by the user. When Synchronous mode is disabled (SMEN in SICR is reset), the state of the XTCLK and OCLK bits in CCR determine the source of CLKOUT; ’11’ enables the Serial Expansion Mode. When the Synchronous mode is enabled (SMEN in SICR is set), the state of the XTCLK and OCLK bits in CCR determine the source of CLKOUT; ’00’ disables it for PLM applications. RTS: Request To Send. This output Alternate Function is only enabled in Synchronous mode; it becomes active when the Least Significant Bit of the data frame is sent to the Serial Output Pin (SOUT) and indicates to the target device that the MCU is about to send a synchronous frame; it returns to its stand-by value just after the last active edge of CLKOUT (MSB transmitted). The active level can be programmed high or low. SDS: Synchronous Data Strobe. This output Alternate function is only enabled in Synchronous mode; it becomes active high when the Least Significant Bit is sent to the Serial Output Pins (SOUT) and indicates to the target device that the MCU is about to send the first bit for each synchronous frame. It is active high on the first bit and it is low for all the rest of the frame. The active level can not be programmed. Figure 80. Receiver and Transmitter Clock Frequencies External RXCLK Receiver Clock Frequency Internal Receiver Clock External TXCLK Transmitter Clock Frequency Internal Transmitter Clock Note: The internal receiver and transmitter clocks are the ones applied to the Tx and Rx shift registers (see Figure 1). 160/224 Min 0 0 0 0 0 0 0 0 Max INTCLK/8 INTCLK/4 INTCLK/8 INTCLK/2 INTCLK/8 INTCLK/4 INTCLK/8 INTCLK/2 Conditions 1x mode 16x mode 1x mode 16x mode 1x mode 16x mode 1x mode 16x mode ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.9 Interrupts and DMA 8.4.9.1 Interrupts The SCI can generate interrupts as a result of several conditions. Receiver interrupts include data pending, receive errors (overrun, framing and parity), as well as address or break pending. Transmitter interrupts are software selectable for either Transmit Buffer Register Empty (BSN set) or for Transmit Shift Register Empty (BSN reset) conditions. Typical usage of the Interrupts generated by the SCI peripheral are illustrated in Figure 11. The SCI peripheral is able to generate interrupt requests as a result of a number of events, several of which share the same interrupt vector. It is therefore necessary to poll S_ISR, the Interrupt Status Register, in order to determine the active trigger. These bits should be reset by the programmer during the Interrupt Service routine. The four major levels of interrupt are encoded in hardware to provide two bits of the interrupt vector register, allowing the position of the block of pointer vectors to be resolved to an 8 byte block size. The SCI interrupts have an internal priority structure in order to resolve simultaneous events. Refer also to Section 0.1.3 for more details relating to Synchronous mode. Table 34. SCI Interrupt Internal Priority Receive DMA Request Highest Priority Transmit DMA Request Receive Interrupt Transmit Interrupt Lowest Priority 161/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Table 35. SCI Interrupt Vectors Interrupt Source Vector Address Transmitter Buffer or Shift Register Empty Transmit DMA end of Block xxx x110 Received Data Pending Receive DMA end of Block xxxx x100 Break Detector Address Word Match xxxx x010 Receiver Error xxxx x000 Figure 81. SCI Interrupts: Example of Typical Usage ADDRESS AFTER BREAK CONDITION DATA BREAK ADDRESS MATCH DATA DATA DATA DATA INTERRUPT BREAK INTERRUPT DATA INTERRUPT DATA INTERRUPT ADDRESS INTERRUPT BREAK ADDRESS NO MATCH DATA BREAK INTERRUPT ADDRESS WORD MARKED BY D9=1 DATA ADDRESS MATCH DATA ADDRESS INTERRUPT DATA DATA ADDRESS NO MATCH DATA DATA INTERRUPT DATA DATA INTERRUPT INTERRUPT CHARACTER SEARCH MODE DATA DATA DATA INTERRUPT MATCH DATA DATA DATA DATA CHAR MATCH INTERRUPT DATA INTERRUPT DATA DATA INTERRUPT INTERRUPT INTERRUPT D9 ACTING AS DATA CONTROL WITH SEPARATE INTERRUPT DATA DATA DATA INTERRUPT 162/224 D9=1 DATA DATA DATA DATA D9=1 DATA INTERRUPT DATA INTERRUPT DATA INTERRUPT INTERRUPT INTERRUPT VA00270 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.9.2 DMA Two DMA channels are associated with the SCI, for transmit and for receive. These follow the register scheme as described in the DMA chapter. DMA Reception To perform a DMA transfer in reception mode: 1. Initialize the DMA counter (RDCPR) and DMA address (RDAPR) registers 2. Enable DMA by setting the RXD bit in the IDPR register. 3. DMA transfer is started when data is received by the SCI. DMA Transmission To perform a DMA transfer in transmission mode: 1. Initialize the DMA counter (TDCPR) and DMA address (TDAPR) registers. 2. Enable DMA by setting the TXD bit in the IDPR register. 3. DMA transfer is started by writing a byte in the Transmitter Buffer register (TXBR). If this byte is the first data byte to be transmitted, the DMA counter and address registers must be initialized to begin DMA transmission at the second byte. Alternatively, DMA transfer can be started by writing a dummy byte in the TXBR register. DMA Interrupts When DMA is active, the Received Data Pending and the Transmitter Shift Register Empty interrupt sources are replaced by the DMA End Of Block receive and transmit interrupt sources. Note: To handle DMA transfer correctly in transmission, the BSN bit in the IMR register must be cleared. This selects the Transmitter Shift Register Empty event as the DMA interrupt source. The transfer of the last byte of a DMA data block will be followed by a DMA End Of Block transmit or receive interrupt, setting the TXEOB or RXEOB bit. A typical Transmission End Of Block interrupt routine will perform the following actions: 1. Restore the DMA counter register (TDCPR). 2. Restore the DMA address register (TDAPR). 3. Clear the Transmitter Shift Register Empty bit TXSEM in the S_ISR register to avoid spurious interrupts. 4. Clear the Transmitter End Of Block (TXEOB) pending bit in the IMR register. 5. Set the TXD bit in the IDPR register to enable DMA. 6. Load the Transmitter Buffer Register (TXBR) with the next byte to transmit. The above procedure handles the case where a further DMA transfer is to be performed. Error Interrupt Handling If an error interrupt occurs while DMA is enabled in reception mode, DMA transfer is stopped. To resume DMA transfer, the error interrupt handling routine must clear the corresponding error flag. In the case of an Overrun error, the routine must also read the RXBR register. Character Search Mode with DMA In Character Search Mode with DMA, when a character match occurs, this character is not transferred. DMA continues with the next received character. To avoid an Overrun error occurring, the Character Match interrupt service routine must read the RXBR register. 163/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 8.4.10 Register Description The SCI registers are located in the following pages in the ST9: SCI number 0: page 24 (18h) SCI number 1: page 25 (19h) (when present) The SCI is controlled by the following registers: Address 164/224 Register R240 (F0h) Receiver DMA Transaction Counter Pointer Register R241 (F1h) Receiver DMA Source Address Pointer Register R242 (F2h) Transmitter DMA Transaction Counter Pointer Register R243 (F3h) Transmitter DMA Destination Address Pointer Register R244 (F4h) Interrupt Vector Register R245 (F5h) Address Compare Register R246 (F6h) Interrupt Mask Register R247 (F7h) Interrupt Status Register R248 (F8h) Receive Buffer Register same Address as Transmitter Buffer Register (Read Only) R248 (F8h) Transmitter Buffer Register same Address as Receive Buffer Register (Write only) R249 (F9h) Interrupt/DMA Priority Register R250 (FAh) Character Configuration Register R251 (FBh) Clock Configuration Register R252 (FCh) Baud Rate Generator High Register R253 (FDh) Baud Rate Generator Low Register R254 (FEh) Synchronous Input Control Register R255 (FFh) Synchronous Output Control Register ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) RECEIVER DMA COUNTER POINTER (RDCPR) TRANSMITTER DMA COUNTER POINTER (TDCPR) R240 - Read/Write R242 - Read/Write Reset value: undefined Reset value: undefined 7 RC7 0 RC6 RC5 RC4 RC3 RC2 RC1 RR/M 7 TC7 Bit 7:1 = RC[7:1]: Receiver DMA Counter Pointer. These bits contain the address of the receiver DMA transaction counter in the Register File. Bit 0 = RR/M: Receiver Register File/Memory Selector. 0: Select Memory space as destination. 1: Select the Register File as destination. RECEIVER DMA ADDRESS POINTER (RDAPR) R241 - Read/Write Reset value: undefined 7 RA7 0 RA6 RA5 RA4 RA3 RA2 RA1 0 TC6 TC5 TC4 TC3 TC2 TC1 TR/M Bit 7:1 = TC[7:1]: Transmitter DMA Counter Pointer. These bits contain the address of the transmitter DMA transaction counter in the Register File. Bit 0 = TR/M: Transmitter Register File/Memory Selector. 0: Select Memory space as source. 1: Select the Register File as source. TRANSMITTER DMA ADDRESS POINTER (TDAPR) R243 - Read/Write Reset value: undefined RPS 7 Bit 7:1 = RA[7:1]: Receiver DMA Address Pointer. These bits contain the address of the pointer (in the Register File) of the receiver DMA data source. Bit 0 = RPS: Receiver DMA Memory Pointer Selector. This bit is only significant if memory has been selected for DMA transfers (RR/M = 0 in the RDCPR register). 0: Select ISR register for receiver DMA transfers address extension. 1: Select DMASR register for receiver DMA transfers address extension. TA7 0 TA6 TA5 TA4 TA3 TA2 TA1 TPS Bit 7:1 = TA[7:1]: Transmitter DMA Address Pointer. These bits contain the address of the pointer (in the Register File) of the transmitter DMA data source. Bit 0 = TPS: Transmitter DMA Memory Pointer Selector. This bit is only significant if memory has been selected for DMA transfers (TR/M = 0 in the TDCPR register). 0: Select ISR register for transmitter DMA transfers address extension. 1: Select DMASR register for transmitter DMA transfers address extension. 165/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) INTERRUPT VECTOR REGISTER (S_IVR) ADDRESS/DATA COMPARE REGISTER (ACR) R244 - Read/Write R245 - Read/Write Reset value: undefined Reset value: undefined 7 V7 V6 V5 V4 V3 EV2 EV1 0 7 0 AC7 Bit 7:3 = V[7:3]: SCI Interrupt Vector Base Address. User programmable interrupt vector bits for transmitter and receiver. Bit 2:1 = EV[2:1]: Encoded Interrupt Source. Both bits EV2 and EV1 are read only and set by hardware according to the interrupt source. EV2 EV1 Interrupt source 0 0 Receiver Error (Overrun, Framing, Parity) 0 1 Break Detect or Address Match 1 0 Received Data Pending/Receiver DMA End of Block 1 1 Transmitter buffer or shift register empty transmitter DMA End of Block Bit 0 = D0: This bit is forced by hardware to 0. 166/224 0 AC6 AC5 AC4 AC3 AC2 AC1 AC0 Bit 7:0 = AC[7:0]: Address/Compare Character. With either 9th bit address mode, address after break mode, or character search, the received address will be compared to the value stored in this register. When a valid address matches this register content, the Receiver Address Pending bit (RXAP in the S_ISR register) is set. After the RXAP bit is set in an addressed mode, all received data words will be transferred to the Receiver Buffer Register. ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Bit 4 = RXE: Receiver Error Mask. INTERRUPT MASK REGISTER (IMR) 0: Disable Receiver error interrupts (OE, PE, and R246 - Read/Write FE pending bits in the S_ISR register). Reset value: 0xx00000 1: Enable Receiver error interrupts. 7 BSN 0 RXEOB TXEOB RXE RXA RXB RXDI TXDI Bit 7 = BSN: Buffer or shift register empty interrupt. This bit selects the source of the transmitter register empty interrupt. 0: Select a Shift Register Empty as source of a Transmitter Register Empty interrupt. 1: Select a Buffer Register Empty as source of a Transmitter Register Empty interrupt. Bit 6 = RXEOB: Received End of Block. This bit is set by hardware only and must be reset by software. RXEOB is set after a receiver DMA cycle to mark the end of a data block. 0: Clear the interrupt request. 1: Mark the end of a received block of data. Bit 5 = TXEOB: Transmitter End of Block. This bit is set by hardware only and must be reset by software. TXEOB is set after a transmitter DMA cycle to mark the end of a data block. 0: Clear the interrupt request. 1: Mark the end of a transmitted block of data. Bit 3 = RXA: Receiver Address Mask. 0: Disable Receiver Address interrupt (RXAP pending bit in the S_ISR register). 1: Enable Receiver Address interrupt. Bit 2 = RXB: Receiver Break Mask. 0: Disable Receiver Break interrupt (RXBP pending bit in the S_ISR register). 1: Enable Receiver Break interrupt. Bit 1 = RXDI: Receiver Data Interrupt Mask. 0: Disable Receiver Data Pending and Receiver End of Block interrupts (RXDP and RXEOB pending bits in the S_ISR register). 1: Enable Receiver Data Pending and Receiver End of Block interrupts. Note: RXDI has no effect on DMA transfers. Bit 0 = TXDI: Transmitter Data Interrupt Mask. 0: Disable Transmitter Buffer Register Empty, Transmitter Shift Register Empty, or Transmitter End of Block interrupts (TXBEM, TXSEM, and TXEOB bits in the S_ISR register). 1: Enable Transmitter Buffer Register Empty, Transmitter Shift Register Empty, or Transmitter End of Block interrupts. Note: TXDI has no effect on DMA transfers. 167/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Note: The source of this interrupt is given by the INTERRUPT STATUS REGISTER (S_ISR) couple of bits (AMEN, AM) as detailed in the IDPR R247 - Read/Write register description. Reset value: undefined 7 OE 0 FE PE RXAP RXBP RXDP TXBE M TXSEM Bit 7 = OE: Overrun Error Pending. This bit is set by hardware if the data in the Receiver Buffer Register was not read by the CPU before the next character was transferred into the Receiver Buffer Register (the previous data is lost). 0: No Overrun Error. 1: Overrun Error occurred. Bit 6 = FE: Framing Error Pending bit. This bit is set by hardware if the received data word did not have a valid stop bit. 0: No Framing Error. 1: Framing Error occurred. Note: In the case where a framing error occurs when the SCI is programmed in address mode and is monitoring an address, the interrupt is asserted and the corrupted data element is transferred to the Receiver Buffer Register. Bit 5 = PE: Parity Error Pending. This bit is set by hardware if the received word did not have the correct even or odd parity bit. 0: No Parity Error. 1: Parity Error occurred. Bit 4 = RXAP: Receiver Address Pending. RXAP is set by hardware after an interrupt acknowledged in the address mode. 0: No interrupt in address mode. 1: Interrupt in address mode occurred. 168/224 Bit 3 = RXBP: Receiver Break Pending bit. This bit is set by hardware if the received data input is held low for the full word transmission time (start bit, data bits, parity bit, stop bit). 0: No break received. 1: Break event occurred. Bit 2 = RXDP: Receiver Data Pending bit. This bit is set by hardware when data is loaded into the Receiver Buffer Register. 0: No data received. 1: Data received in Receiver Buffer Register. Bit 1 = TXBEM: Transmitter Buffer Register Empty. This bit is set by hardware if the Buffer Register is empty. 0: No Buffer Register Empty event. 1: Buffer Register Empty. Bit 0 = TXSEM: Transmitter Shift Register Empty. This bit is set by hardware if the Shift Register has completed the transmission of the available data. 0: No Shift Register Empty event. 1: Shift Register Empty. Note: The Interrupt Status Register bits can be reset but cannot be set by the user. The interrupt source must be cleared by resetting the related bit when executing the interrupt service routine (naturally the other pending bits should not be reset). ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) TRANSMITTER BUFFER REGISTER (TXBR) RECEIVER BUFFER REGISTER (RXBR) R248 - Write only R248 - Read only Reset value: undefined Reset value: undefined 7 RD7 RD6 RD5 RD4 RD3 RD2 RD1 0 7 RD0 TD7 0 TD6 TD5 TD4 TD3 TD2 TD1 TD0 Bit 7:0 = RD[7:0]: Received Data. This register stores the data portion of the received word. The data will be transferred from the Receiver Shift Register into the Receiver Buffer Register at the end of the word. All receiver interrupt conditions will be updated at the time of transfer. If the selected character format is less than 8 bits, unused most significant bits will forced to “1”. Bit 7:0 = TD[7:0]: Transmit Data. The ST9 core will load the data for transmission into this register. The SCI will transfer the data from the buffer into the Shift Register when available. At the transfer, the Transmitter Buffer Register interrupt is updated. If the selected word format is less than 8 bits, the unused most significant bits are not significant. Note: RXBR and TXBR are two physically different registers located at the same address. Note: TXBR and RXBR are two physically different registers located at the same address. 169/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) INTERRUPT/DMA PRIORITY REGISTER (IDPR) R249 - Read/Write Reset value: undefined 7 AMEN 0 SB SA RXD TXD PRL2 PRL1 PRL0 Bit 7 = AMEN: Address Mode Enable. This bit, together with the AM bit (in the CHCR register), decodes the desired addressing/9th data bit/character match operation. In Address mode the SCI monitors the input serial data until its address is detected AMEN AM 0 0 Address interrupt if 9th data bit = 1 0 1 Address interrupt if character match 1 0 Address interrupt if character match and 9th data bit =1 1 1 Address interrupt if character match with word immediately following Break Note: Upon reception of address, the RXAP bit (in the Interrupt Status Register) is set and an interrupt cycle can begin. The address character will not be transferred into the Receiver Buffer Register but all data following the matched SCI address and preceding the next address word will be transferred to the Receiver Buffer Register and the proper interrupts updated. If the address does not match, all data following this unmatched address will not be transferred to the Receiver Buffer Register. In any of the cases the RXAP bit must be reset by software before the next word is transferred into the Buffer Register. When AMEN is reset and AM is set, a useful character search function is performed. This allows the SCI to generate an interrupt whenever a specific character is encountered (e.g. Carriage Return). Bit 6 = SB: Set Break. 0: Stop the break transmission after minimum break length. 1: Transmit a break following the transmission of all data in the Transmitter Shift Register and the Buffer Register. Note: The break will be a low level on the transmitter data output for at least one complete word for- 170/224 mat. If software does not reset SB before the minimum break length has finished, the break condition will continue until software resets SB. The SCI terminates the break condition with a high level on the transmitter data output for one transmission clock period. Bit 5 = SA: Set Address. If an address/9th data bit mode is selected, SA value will be loaded for transmission into the Shift Register. This bit is cleared by hardware after its load. 0: Indicate it is not an address word. 1: Indicate an address word. Note: Proper procedure would be, when the Transmitter Buffer Register is empty, to load the value of SA and then load the data into the Transmitter Buffer Register. Bit 4 = RXD: Receiver DMA Mask. This bit is reset by hardware when the transaction counter value decrements to zero. At that time a receiver End of Block interrupt can occur. 0: Disable Receiver DMA request (the RXDP bit in the S_ISR register can request an interrupt). 1: Enable Receiver DMA request (the RXDP bit in the S_ISR register can request a DMA transfer). Bit 3 = TXD: Transmitter DMA Mask. This bit is reset by hardware when the transaction counter value decrements to zero. At that time a transmitter End Of Block interrupt can occur. 0: Disable Transmitter DMA request (TXBEM or TXSEM bits in S_ISR can request an interrupt). 1: Enable Transmitter DMA request (TXBEM or TXSEM bits in S_ISR can request a DMA transfer). Bit 2:0 = PRL[2:0]: SCI Interrupt/DMA Priority bits. The priority for the SCI is encoded with (PRL2,PRL1,PRL0). Priority level 0 is the highest, while level 7 represents no priority. When the user has defined a priority level for the SCI, priorities within the SCI are hardware defined. These SCI internal priorities are: Receiver DMA request Transmitter DMA request Receiver interrupt Transmitter interrupt highest priority lowest priority ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) CHARACTER CONFIGURATION REGISTER (CHCR) Bit 4 = AB: Address/9th Bit. 0: No Address/9th bit. R250 - Read/Write 1: Address/9th bit included in the character format Reset value: undefined between the parity bit and the first stop bit. This 7 0 bit can be used to address the SCI or as a ninth data bit. AM EP PEN AB SB1 SB0 WL1 WL0 Bit 3:2 = SB[1:0]: Number of Stop Bits.. Bit 7 = AM: Address Mode. This bit, together with the AMEN bit (in the IDPR register), decodes the desired addressing/9th data bit/character match operation. Please refer to the table in the IDPR register description. Bit 6 = EP: Even Parity. 0: Select odd parity (when parity is enabled). 1: Select even parity (when parity is enabled). SB1 SB0 0 0 1 1 0 1 0 1 Number of stop bits in 16X mode in 1X mode 1 1 1.5 2 2 2 2.5 3 Bit 1:0 = WL[1:0]: Number of Data Bits Bit 5 = PEN: Parity Enable. 0: No parity bit. 1: Parity bit generated (transmit data) or checked (received data). Note: If the address/9th bit is enabled, the parity bit will precede the address/9th bit (the 9th bit is never included in the parity calculation). WL1 0 0 1 1 WL0 0 1 0 1 Data Length 5 bits 6 bits 7 bits 8 bits 171/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) CLOCK CONFIGURATION REGISTER (CCR) 0: Select 16X clock mode for both receiver and transmitter. R251 - Read/Write 1: Select 1X clock mode for both receiver and Reset value: 0000 0000 (00h) transmitter. 7 0 Note: In 1X clock mode, the transmitter will transmit data at one data bit per clock period. In 16X XTCLK OCLK XRX XBRG CD AEN LBEN STPE N mode each data bit period will be 16 clock periods long. Bit 7 = XTCLK This bit, together with the OCLK bit, selects the Bit 2 = AEN: Auto Echo Enable. source for the transmitter clock. The following ta0: No auto echo mode. ble shows the coding of XTCLK and OCLK. 1: Put the SCI in auto echo mode. Bit 6 = OCLK This bit, together with the XTCLK bit, selects the source for the transmitter clock. The following table shows the coding of XTCLK and OCLK. XTCLK OCLK 0 0 Pin is used as a general I/O 0 1 Pin = TXCLK (used as an input) 1 0 Pin = CLKOUT (outputs the Baud Rate Generator clock) 1 1 Pin = CLKOUT (outputs the Serial expansion and synchronous mode clock) Pin Function Bit 5 = XRX: External Receiver Clock Source. 0: External receiver clock source not used. 1: Select the external receiver clock source. Note: The external receiver clock frequency must be 16 times the data rate, or equal to the data rate, depending on the status of the CD bit. Bit 4 = XBRG: Baud Rate Generator Clock Source. 0: Select INTCLK for the baud rate generator. 1: Select the external receiver clock for the baud rate generator. Bit 3 = CD: Clock Divisor. The status of CD will determine the SCI configuration (synchronous/asynchronous). 172/224 Note: Auto Echo mode has the following effect: the SCI transmitter is disconnected from the dataout pin SOUT, which is driven directly by the receiver data-in pin, SIN. The receiver remains connected to SIN and is operational, unless loopback mode is also selected. Bit 1 = LBEN: Loopback Enable. 0: No loopback mode. 1: Put the SCI in loopback mode. Note: In this mode, the transmitter output is set to a high level, the receiver input is disconnected, and the output of the Transmitter Shift Register is looped back into the Receiver Shift Register input. All interrupt sources (transmitter and receiver) are operational. Bit 0 = STPEN: Stick Parity Enable. 0: The transmitter and the receiver will follow the parity of even parity bit EP in the CHCR register. 1: The transmitter and the receiver will use the opposite parity type selected by the even parity bit EP in the CHCR register. EP SPEN 0 (odd) 1 (even) 0 (odd) 1 (even) 0 0 1 1 Parity (Transmitter & Receiver) Odd Even Even Odd ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) BAUD RATE GENERATOR HIGH REGISTER 1: Select Synchronous mode with its programmed (BRGHR) I/O configuration. R252 - Read/Write Bit 6 = INPL: SIN Input Polarity. Reset value: undefined 0: Polarity not inverted. 1: Polarity inverted. 15 8 Note: INPL only affects received data. In AutoEcho mode SOUT = SIN even if INPL is set. In BG15 BG14 BG13 BG12 BG11 BG10 BG9 BG8 Loop-Back mode the state of the INPL bit is irrelevant. BAUD RATE GENERATOR LOW REGISTER (BRGLR) Bit 5 = XCKPL: Receiver Clock Polarity. R253 - Read/Write 0: RXCLK is active on the rising edge. Reset value: undefined 1: RXCLK is active on the falling edge. 7 BG7 0 BG6 BG5 BG4 BG3 BG2 BG1 BG0 Bit 15:0 = Baud Rate Generator MSB and LSB. The Baud Rate generator is a programmable divide by “N” counter which can be used to generate the clocks for the transmitter and/or receiver. This counter divides the clock input by the value in the Baud Rate Generator Register. The minimum baud rate divisor is 2 and the maximum divisor is 216-1. After initialization of the baud rate generator, the divisor value is immediately loaded into the counter. This prevents potentially long random counts on the initial load. If set to 0 or 1, the Baud Rate Generator is stopped. SYNCHRONOUS INPUT CONTROL (SICR) R254 - Read/Write Reset value: 0000 0011 (03h) 7 SMEN Note: XCKPL only affects the receiver clock. In Auto-Echo mode CLKOUT = RXCLK independently of the XCKPL status. In Loop-Back the state of the XCKPL bit is irrelevant. Bit 4 = DCDEN: DCD Input Enable. 0: Disable hardware synchronization. 1: Enable hardware synchronization. Note: When DCDEN is set, RXCLK drives the receiver section only during the active level of the DCD input (DCD works as a gate on RXCLK, informing the MCU that a transmitting device is sending a synchronous frame to it). Bit 3 = DCDPL: DCD Input Polarity. 0: The DCD input is active when LOW. 1: The DCD input is active when HIGH. Note: DCDPL only affects the gating activity of the receiver clock. In Auto-Echo mode RTS = DCD independently of DCDPL. In Loop-Back mode, the state of DCDPL is irrelevant. 0 INPL XCKPL DCDE DCDP INPEN N L X X Bit 7 = SMEN: Synchronous Mode Enable. 0: Disable all features relating to Synchronous mode (the contents of SICR and SOCR are ignored). Bit 2 = INPEN: All Input Disable. 0: Enable SIN/RXCLK/DCD inputs. 1: Disable SIN/RXCLK/DCD inputs. Bit 1:0 = “Don’t Care” 173/224 ST92163 - SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) SYNCHRONOUS OUTPUT CONTROL (SOCR) Bit 3 = RTSEN: RTS and SDS Output Enable. 0: Disable the RTS and SDS hardware synchroniR255 - Read/Write sation. Reset value: 0000 0001 (01h) 1: Enable the RTS and SDS hardware synchronisation. 7 0 Notes: – When RTSEN is set, the RTS output becomes OUTP OUTS OCKP OCKS RTSE RTS OUT X active just before the first active edge of CLKL B L B N PL DIS OUT and indicates to target device that the MCU is about to send a synchronous frame; it returns to its stand-by value just after the last active edge Bit 7 = OUTPL: SOUT Output Polarity. of CLKOUT (MSB transmitted). 0: Polarity not inverted. 1: Polarity inverted. – When RTSEN is set, the SDS output becomes active high and indicates to the target device that Note: OUTPL only affects the data sent by the the MCU is about to send the first bit of a syntransmitter section. In Auto-Echo mode SOUT = chronous frame on the Serial Output Pin SIN even if OUTPL=1. In Loop-Back mode, the (SOUT); it returns to low level as soon as the state of OUTPL is irrelevant. second bit is sent on the Serial Output Pin (SOUT). In this way a positive pulse is generated Bit 6 = OUTSB: SOUT Output Stand-By Level. each time that the first bit of a synchronous frame is present on the Serial Output Pin (SOUT). 0: SOUT stand-by level is HIGH. 1: SOUT stand-by level is LOW. Bit 5 = OCKPL: Transmitter Clock Polarity. 0: CLKOUT is active on the rising edge. 1: CLKOUT is active on the falling edge. Note: OCKPL only affects the transmitter clock. In Auto-Echo mode CLKOUT = RXCLK independently of the state of OCKPL. In Loop-Back mode the state of OCKPL is irrelevant. Bit 4 = OCKSB: Transmitter Clock Stand-By Level. 0: The CLKOUT stand-by level is HIGH. 1: The CLKOUT stand-by level is LOW. Bit 2 = RTSPL: RTS Output Polarity. 0: The RTS output is active when LOW. 1: The RTS output is active when HIGH. Note: RTSPL only affects the RTS activity on the output pin. In Auto-Echo mode RTS = DCD independently from the RTSPL value. In Loop-Back mode RTSPL value is ’Don’t Care’. Bit 1 = OUTDIS: Disable all outputs. This feature is available on specific devices only (see device pin-out description). When OUTDIS=1, all output pins (if configured in Alternate Function mode) will be put in High Impedance for networking. 0: SOUT/CLKOUT/enabled 1: SOUT/CLKOUT/RTS put in high impedance Bit 0 = “Don’t Care” 174/224 ST92163 - I2C BUS INTERFACE 8.5 I2C BUS INTERFACE 8.5.1 Introduction The I2C bus Interface serves as an interface between the microcontroller and the serial I2C bus. It provides both multimaster and slave functions with both 7-bit and 10-bit address modes; it controls all I2C bus-specific sequencing, protocol, arbitration, timing and supports both standard (100KHz) and fast I2C modes (400KHz). Using DMA, data can be transferred with minimum use of CPU time. The peripheral uses two external lines to perform the protocols: SDA, SCL. 8.5.2 Main Features 2 ■ Parallel-bus/I C protocol converter ■ Multi-master capability ■ 7-bit/10-bit Addressing 2 2 ■ Standard I C mode/Fast I C mode ■ Transmitter/Receiver flag ■ End-of-byte transmission flag ■ Transfer problem detection ■ Interrupt generation on error conditions ■ Interrupt generation on transfer request and on data received Interrupt Features: ■ Interrupt generation on error condition, on transmission request and on data received ■ Interrupt address vector for each interrupt source ■ Pending bit and mask bit for each interrupt source ■ Programmable interrupt priority respects the other peripherals of the microcontroller ■ Interrupt address vector programmable DMA Features: ■ DMA both in transmission and in reception with enabling bits ■ DMA from/toward both Register File and Memory ■ End Of Block interrupt sources with the related pending bits ■ Selection between DMA Suspended and DMA Not-Suspended mode if error condition occurs. I2C Master Features: ■ Start bit detection flag ■ Clock generation 2 ■ I C bus busy flag ■ Arbitration Lost flag ■ End of byte transmission flag ■ Transmitter/Receiver flag ■ Stop/Start generation I2C Slave Features: ■ Stop bit detection 2 ■ I C bus busy flag ■ Detection of misplaced start or stop condition 2 ■ Programmable I C Address detection (both 7bit and 10-bit mode) ■ General Call address programmable ■ Transfer problem detection ■ End of byte transmission flag ■ Transmitter/Receiver flag. 175/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) Figure 82. I2C Interface Block Diagram DATA BUS DATA REGISTE R DATA SHIFT REGISTER SDA DATA SDAI CONTROL COMPARA TOR OWN ADDRES S REGISTE R 1 OWN ADDRESS REGISTER 2 GENERAL CALL ADDRESS CLOCK CONTROL REGISTER SCL CLOCK SCLI CONTROL STATUS REGISTE R 1 STATUS REGISTE R 2 CONTROL REGISTE R LOGIC AND INTERRUPT/ DMA REGISTERS DMA CONTROL SIGNALS INTERRUPT VR02119A 8.5.3 Functional Description Refer to the I2CCR, I2CSR1 and I2CSR2 registers in Section 8.5.7. for the bit definitions. The I2C interface works as an I/O interface between the ST9 microcontroller and the I2C bus protocol. In addition to receiving and transmitting data, the interface converts data from serial to parallel format and vice versa using an interrupt or polled handshake. It operates in Multimaster/slave I2C mode. The selection of the operating mode is made by software. The I2C interface is connected to the I2C bus by a data pin (SDAI) and a clock pin (SCLI) which must be configured as open drain when the I2C cell is enabled by programming the I/O port bits and the PE bit in the I2CCR register. In this case, the value of the external pull-up resistance used depends on the application. When the I2C cell is disabled, the SDAI and SCLI ports revert to being standard I/ O port pins. 176/224 The I2C interface has sixteen internal registers. Six of them are used for initialization: – Own Address Registers I2COAR1, I2COAR2 – General Call Address Register I2CADR – Clock Control Registers I2CCCR, I2CECCR – Control register I2CCR The following four registers are used during data transmission/reception: – Data Register I2CDR – Control Register I2CCR – Status Register 1 I2CSR1 – Status Register 2 I2CSR2 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) The following seven registers are used to handle the interrupt and the DMA features: – Interrupt Status Register I2CISR – Interrupt Mask Register I2CIMR – Interrupt Vector Register I2CIVR – Receiver DMA Address Pointer Register I2CRDAP – Receiver DMA Transaction Counter Register I2CRDC – Transmitter DMA Address Pointer Register I2CTDAP – Transmitter DMA transaction Counter Register I2CTDC The interface can decode both addresses: – Software programmable 7-bit General Call address – I2C address stored by software in the I2COAR1 register in 7-bit address mode or stored in I2COAR1 and I2COAR2 registers in 10-bit address mode. After a reset, the interface is disabled. bit is set) the General Call address (stored in I2CADR register). It never recognizes the Start Byte (address byte 01h) whatever its own address is. Data and addresses are transferred in 8 bits, MSB first. The first byte(s) following the start condition contain the address (one byte in 7-bit mode, two bytes in 10-bit mode). The address is always transmitted in master mode. A 9th clock pulse follows the 8 clock cycles of a byte transfer, during which the receiver must send an acknowledge bit to the transmitter. Acknowledge is enabled and disabled by software. Refer to Figure 83. IMPORTANT : 1. To guarantee correct operation, before enabling the peripheral (while I2CCR.PE=0), configure bit7 and bit6 of the I2COAR2 register according to the internal clock INTCLK (for example 11xxxxxxb in the range 14 - 30 MHz). 2. Bit7 of the I2CCR register must be cleared. 8.5.3.1 Mode Selection In I2C mode, the interface can operate in the four following modes: – Master transmitter/receiver – Slave transmitter/receiver By default, it operates in slave mode. This interface automatically switches from slave to master after a start condition is generated on the bus and from master to slave in case of arbitration loss or stop condition generation. In Master mode, it initiates a data transfer and generates the clock signal. A serial data transfer always begins with a start condition and ends with a stop condition. Both start and stop conditions are generated in master mode by software. In Slave mode, it is able to recognize its own address (7 or 10-bit), as stored in the I2COAR1 and I2COAR2 registers and (when the I2CCR.ENGC 177/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) Figure 83. I2C BUS Protocol SDA ACK MSB SCL 1 2 START CONDITION 8 9 STOP CONDITION VR02119B Any transfer can be done using either the I2C registers directly or via the DMA. If the transfer is to be done directly on I2C, the interface waits (by holding the SCL line low) for software to write in the Data Register before transmission of a data byte, or to read the Data Register after a data byte is received. If the transfer is to be done via DMA, the interface sends a request for a DMA transfer. Then it waits for the DMA to complete. The transfer between the interface and the I 2C bus will begin on the next rising edge of the SCL clock. The SCL frequency (Fscl) generated in master mode is controlled by a programmable clock divider. The speed of the I2C interface may be selected between Standard (0-100KHz) and Fast (100400KHz) I2C modes. 8.5.4 I2C State Machine To enable the interface in I2C mode the I2CCR.PE bit must be set twice as the first write only activates the interface (only the PE bit is set); and the bit7 of I2CCR register must be cleared. The I2C interface always operates in slave mode (the M/SL bit is cleared) except when it initiates a transmission or a receipt sequencing (master mode). 178/224 The multimaster function is enabled with an automatic switch from master mode to slave mode when the interface loses the arbitration of the I2C bus. 8.5.4.1 I2C Slave Mode As soon as a start condition is detected, the address word is received from the SDA line and sent to the shift register; then it is compared with the address of the interface or the General Call address (if selected by software). Note: In 10-bit addressing mode, the comparison includes the header sequence (11110xx0) and the two most significant bits of the address. ■ Header (10-bit mode) or Address (both 10-bit and 7-bit modes) not matched: the state machine is reset and waits for another Start condition. ■ Header matched (10-bit mode only): the interface generates an acknowledge pulse if the ACK bit of the control register (I2CCR) is set. ■ Address matched: the I2CSR1.ADSL bit is set and an acknowledge bit is sent to the master if the I2CCR.ACK bit is set. An interrupt is sent to the microcontroller if the I2CCR.ITE bit is set.Then it waits for the microcontroller to read the I2CSR1 register by holding the SCL line low (see Figure 84 Transfer sequencing EV1). ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) Next, depending on the data direction bit (least significant bit of the address byte), and after the generation of an acknowledge, the slave must go in sending or receiving mode. In 10-bit mode, after receiving the address sequence the slave is always in receive mode. It will enter transmit mode on receiving a repeated Start condition followed by the header sequence with matching address bits and the least significant bit set (11110xx1). Slave Receiver Following the address reception and after I2CSR1 register has been read, the slave receives bytes from the SDA line into the Shift Register and sends them to the I2CDR register. After each byte it generates an acknowledge bit if the I2CCR.ACK bit is set. When the acknowledge bit is sent, the I2CSR1.BTF flag is set and an interrupt is generated if the I2CCR.ITE bit is set (see Figure 84 Transfer sequencing EV2). Then the interface waits for a read of the I2CSR1 register followed by a read of the I2CDR register, or waits for the DMA to complete; both holding the SCL line low. Slave Transmitter Following the address reception and after I2CSR1 register has been read, the slave sends bytes from the I2CDR register to the SDA line via the internal shift register. When the acknowledge bit is received, the I2CCR.BTF flag is set and an interrupt is generated if the I2CCR.ITE bit is set (see Figure 84 Transfer sequencing EV3). The slave waits for a read of the I2CSR1 register followed by a write in the I2CDR register or waits for the DMA to complete, both holding the SCL line low. Error Cases – BERR: Detection of a Stop or a Start condition during a byte transfer. The I2CSR2.BERR flag is set and an interrupt is generated if I2CCR.ITE bit is set. If it is a stop then the state machine is reset. If it is a start then the state machine is reset and it waits for the new slave address on the bus. – AF: Detection of a no-acknowledge bit. The I2CSR2.AF flag is set and an interrupt is generated if the I2CCR.ITE bit is set. Note: In both cases, SCL line is not stretched low; however, the SDA line, due to possible «0» bits transmitted last, can remain low. It is then necessary to release both lines by software. Other Events – ADSL: Detection of a Start condition after an acknowledge time-slot. The state machine is reset and starts a new process. The I2CSR1.ADSL flag bit is set and an interrupt is generated if the I2CCR.ITE bit is set. The SCL line is stretched low. – STOPF: Detection of a Stop condition after an acknowledge time-slot. The state machine is reset. Then the I2CSR2.STOPF flag is set and an interrupt is generated if the I2CCR.ITE bit is set. How to release the SDA / SCL lines Set and subsequently clear the I2CCR.STOP bit while the I2CSR1.BTF bit is set; then the SDA/ SCL lines are released immediately after the transfer of the current byte. This will also reset the state machine; any subsequent STOP bit (EV4) will not be detected. 8.5.4.2 I2C Master Mode To switch from default Slave mode to Master mode a Start condition generation is needed. Setting the I2CCR.START bit while the I2CSR1.BUSY bit is cleared causes the interface to generate a Start condition. Once the Start condition is generated, the peripheral is in master mode (I2CSR1.M/SL=1) and I2CSR1.SB (Start bit) flag is set and an interrupt is generated if the I2CCR.ITE bit is set (see Figure 84 Transfer sequencing EV5 event). The interface waits for a read of the I2CSR1 register followed by a write in the I2CDR register with the Slave address, holding the SCL line low. 179/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) Then the slave address is sent to the SDA line. In 7-bit addressing mode, one address byte is sent. In 10-bit addressing mode, sending the first byte including the header sequence causes the I2CSR1.EVF and I2CSR1.ADD10 bits to be set by hardware with interrupt generation if the I2CCR.ITE bit is set. Then the master waits for a read of the I2CSR1 register followed by a write in the I2CDR register, holding the SCL line low (see Figure 84 Transfer sequencing EV9). Then the second address byte is sent by the interface. After each address byte, an acknowledge clock pulse is sent to the SCL line if the I2CSR1.EVF and – I2CSR1.ADD10 bit (if first header) – I2CSR2.ADDTX bit (if address or second header) are set, and an interrupt is generated if the I2CCR.ITE bit is set. The peripheral waits for a read of the I2CSR1 register followed by a write into the Control Register (I2CCR) by holding the SCL line low (see Figure 84 Transfer sequencing EV6 event). If there was no acknowledge (I2CSR2.AF=1), the master must stop or restart the communication (set the I2CCR.START or I2CCR.STOP bits). If there was an acknowledge, the state machine enters a sending or receiving process according to the data direction bit (least significant bit of the address), the I2CSR1.BTF flag is set and an interrupt is generated if I2CCR.ITE bit is set (see Transfer sequencing EV7, EV8 events). If the master loses the arbitration of the bus there is no acknowledge, the I2CSR2.AF flag is set and the master must set the START or STOP bit in the control register (I2CCR).The I2CSR2.ARLO flag is set, the I2CSR1.M/SL flag is cleared and the process is reset. An interrupt is generated if I2CCR.ITE is set. Master Transmitter: The master waits for the microcontroller to write in the Data Register (I2CDR) or it waits for the DMA to complete both holding the SCL line low (see Transfer sequencing EV8). Then the byte is received into the shift register and sent to the SDA line. When the acknowledge bit is received, the I2CSR1.BTF flag is set and an interrupt is generated if the I2CCR.ITE bit is set or the DMA is requested. 180/224 Note: In 10-bit addressing mode, to switch the master to Receiver mode, software must generate a repeated Start condition and resend the header sequence with the least significant bit set (11110xx1). Master Receiver: The master receives a byte from the SDA line into the shift register and sends it to the I2CDR register. It generates an acknowledge bit if the I2CCR.ACK bit is set and an interrupt if the I2CCR.ITE bit is set or a DMA is requested (see Transfer sequencing EV7 event). Then it waits for the microcontroller to read the Data Register (I2CDR) or waits for the DMA to complete both holding SCL line low. Error Cases ■ BERR: Detection of a Stop or a Start condition during a byte transfer. The I2CSR2.BERR flag is set and an interrupt is generated if I2CCR.ITE is set. ■ AF: Detection of a no acknowledge bit The I2CSR2.AF flag is set and an interrupt is generated if I2CCR.ITE is set. ■ ARLO: Arbitration Lost The I2CSR2.ARLO flag is set, the I2CSR1.M/SL flag is cleared and the process is reset. An interrupt is generated if the I2CCR.ITE bit is set. Note: In all cases, to resume communications, set the I2CCR.START or I2CCR.STOP bits. Events generated by the I2C interface ■ STOP condition When the I2CCR.STOP bit is set, a Stop condition is generated after the transfer of the current byte, the I2CSR1.M/SL flag is cleared and the state machine is reset. No interrupt is generated in master mode at the detection of the stop condition. ■ START condition When the I2CCR.START bit is set, a start condition is generated as soon as the I2C bus is free. The I2CSR1.SB flag is set and an interrupt is generated if the I2CCR.ITE bit is set. ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) Figure 84. Transfer Sequencing 7-bit Slave receiver: S Address A Data1 A Data2 EV1 A EV2 ..... EV2 DataN A P EV2 EV4 7-bit Slave transmitter: S Address A Data1 A Data2 EV1 EV3 A EV3 EV3 .... . DataN NA P EV3-1 EV4 7-bit Master receiver: S Address A EV5 Data1 A Data2 EV6 A EV7 EV7 DataN .... . NA P EV7 7-bit Master transmitter: S Address A EV5 Data1 A EV6 EV8 Data2 A EV8 EV8 DataN .... . A P EV8 10-bit Slave receiver: S Header A Address A Data1 A EV1 ..... EV2 DataN A P EV2 EV4 10-bit Slave transmitter: Sr Header A Data1 A .... DataN EV3 . EV1 EV3 A P EV3-1 EV4 10-bit Master transmitter S Header EV5 A Address EV9 A Data1 A EV6 EV8 EV8 DataN .... . A P EV8 10-bit Master receiver: Header Sr EV5 A Data1 EV6 A EV7 .... . DataN A P EV7 Legend: S=Start, Sr = Repeated Start, P=Stop, A=Acknowledge, NA=Non-acknowledge, EVx=Event (with interrupt if ITE=1) EV1: EVF=1, ADSL=1, cleared by reading SR1 register. EV2: EVF=1, BTF=1, cleared by reading SR1 register followed by reading DR register or when DMA is complete. EV3: EVF=1, BTF=1, cleared by reading SR1 register followed by writing DR register or when DMA is complete. EV3-1: EVF=1, AF=1, BTF=1; AF is cleared by reading SR1 register, BTF is cleared by releasing the lines (STOP=1, STOP=0) or writing DR register (for example DR=FFh). Note: If lines are released by STOP=1, STOP=0 the subsequent EV4 is not seen. EV4: EVF=1, STOPF=1, cleared by reading SR2 register. 181/224 ST92163 - I2C BUS INTERFACE EV5: EVF=1, SB=1, cleared by reading SR1 register followed by writing DR register. EV6: EVF=1, ADDTX=1, cleared by reading SR1 register followed by writing CR register (for example PE=1). EV7: EVF=1, BTF=1, cleared by reading SR1 register followed by reading DR register or when DMA is complete. EV8: EVF=1, BTF=1, cleared by reading SR1 register followed by writing DR register or when DMA is complete. EV9: EVF=1, ADD10=1, cleared by reading SR1 register followed by writing DR register. Figure 85. Event Flags and Interrupt Generation ADSL SB AF STOPF ARLO BERR ADD10 ADDTX IERRM IERRP ERROR INTERRUPT REQUEST ITE IRXM BTF=1 & TRA=0 IRXP ITE DATA RECEIVED or END OF BLOCK INTERRUPT REQUEST REOBP Receiving DMA End Of Block ITXM BTF=1 & TRA=1 ITXP ITE TEOBP Transmitting DMA End Of Block I2CSR1.EVF 182/224 READY TO TRANSMIT or END OF BLOCK INTERRUPT REQUEST ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) 8.5.5 Interrupt Features The I2Cbus interface has three interrupt sources related to “Error Condition”, “Peripheral Ready to Transmit” and “Data Received”. The peripheral uses the ST9+ interrupt internal protocol without requiring the use of the external interrupt channel. Dedicated registers of the peripheral should be loaded with appropriate values to set the interrupt vector (see the description of the I2CIVR register), the interrupt mask bits (see the description of the I2CIMR register) and the interrupt priority and pending bits (see the description of the I2CISR register). The peripheral also has a global interrupt enable (the I2CCR.ITE bit) that must be set to enable the interrupt features. Moreover there is a global interrupt flag that is set when one of the interrupt events occurs (except the End Of Block interrupts - see the DMA Features section). The “Data Received” interrupt source occurs after the acknowledge of a received data byte is performed. It is generated when the I2CSR1.BTF flag is set and the I2CSR1.TRA flag is zero. If the DMA feature is enabled in receiver mode, this interrupt is not generated and the same interrupt vector is used to send a Receiving End Of Block interrupt (See the DMA feature section). The “Peripheral Ready To Transmit” interrupt source occurs as soon as a data byte can be transmitted by the peripheral. It is generated when the I2CSR1.BTF and the I2CSR1.TRA flags are set. If the DMA feature is enabled in transmitter mode, this interrupt is not generated and the same interrupt vector is used to send a Transmitting End Of Block interrupt (See the DMA feature section). The “Error condition” interrupt source occurs when one of the following condition occurs: – Address matched in Slave mode while I2CCR.ACK=1 (I2CSR1.ADSL and I2CSR1.EVF flags = 1) – Start condition generated (I2CSR1.SB and I2CSR1.EVF flags = 1) – No acknowledge received after byte transmission (I2CSR2.AF and I2CSR1.EVF flags = 1) – Stop detected in Slave mode (I2CSR2.STOPF and I2CSR1.EVF flags = 1) – Arbitration lost in Master mode (I2CSR2.ARLO and I2CSR1.EVF flags = 1) – Bus error, Start or Stop condition detected during data transfer (I2CSR2.BERR and I2CSR1.EVF flags = 1) – Master has sent the header byte (I2CSR1.ADD10 and I2CSR1.EVF flags = 1) – Address byte successfully transmitted in Master mode. (I2CSR1.EVF = 1 and I2CSR2.ADDTX=1) Note: Depending on the value of I2CISR.DMASTOP bit, the pending bit related to the “error condition” interrupt source is able to suspend or not suspend DMA transfers. Each interrupt source has a dedicated interrupt address pointer vector stored in the I2CIVR register. The five more significant bits of the vector address are programmable by the customer, whereas the three less significant bits are set by hardware depending on the interrupt source: – 010: error condition detected – 100: data received – 110: peripheral ready to transmit The priority with respect to the other peripherals is programmable by setting the PRL[2:0] bits in the I2CISR register. The lowest interrupt priority is obtained by setting all the bits (this priority level is never acknowledged by the CPU and is equivalent to disabling the interrupts of the peripheral); the highest interrupt priority is programmed by resetting all the bits. See the Interrupt and DMA chapters for more details. The internal priority of the interrupt sources of the peripheral is fixed by hardware with the following order: “Error Condition” (highest priority), “Data Received”, “Peripheral Ready to Transmit”. Note: The DMA has the highest priority over the interrupts; moreover the “Transmitting End Of Block” interrupt has the same priority as the “Peripheral Ready to Transmit” interrupt and the “Receiving End Of Block” interrupt has the same priority as the “Data received” interrupt. Each of these three interrupt sources has a pending bit (IERRP, IRXP, ITXP) in the I2CISR register that is set by hardware when the corresponding interrupt event occurs. An interrupt request is performed only if the corresponding mask bit is set (IERRM, IRXM, ITXM) in the I2CIMR register and the peripheral has a proper priority level. The pending bit has to be reset by software. 183/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) Note: Until the pending bit is reset (while the corresponding mask bit is set), the peripheral processes an interrupt request. So, if at the end of an interrupt routine the pending bit is not reset, another interrupt request is performed. Note: Before the end of the transmission and reception interrupt routines, the I2CSR1.BTF flag bit should be checked, to acknowledge any interrupt requests that occurred during the interrupt routine and to avoid masking subsequent interrupt requests. Note: The “Error” event interrupt pending bit (I2CISR.IERRP) is forced high until the error event flags are set (ADD10, ADSL and SB flags of the I2CSR1 register; SCLF, ADDTX, AF, STOPF, ARLO and BERR flags of the I2CSR2 register). Note: If the I2CISR.DMASTOP bit is reset, then the DMA has the highest priority with respect to the interrupts; if the bit is set (as after the MCU reset) and the “Error event” pending bit is set (I2CISR.IERRP), then the DMA is suspended until the pending bit is reset by software. In the second case, the “Error” interrupt sources have higher priority, followed by DMA, “Data received” and “Receiving End Of Block” interrupts, “Peripheral Ready to Transmit” and “Transmitting End Of Block”. Moreover the Transmitting End Of Block interrupt has the same priority as the “Peripheral Ready to Transmit” interrupt and the Receiving End Of Block interrupt has the same priority as the “Data received” interrupt. 8.5.6 DMA Features The peripheral can use the ST9+ on-chip Direct Memory Access (DMA) channels to provide highspeed data transaction between the peripheral and contiguous locations of Register File, and Memory. The transactions can occur from and toward the peripheral. The maximum number of transactions that each DMA channel can perform is 222 if the register file is selected or 65536 if memory is selected. The control of the DMA features is performed using registers placed in the peripheral register page (I2CISR, I2CIMR, I2CRDAP, I2CRDC, I2CTDAP, I2CTDC). Each DMA transfer consists of three operations: – A load from/to the peripheral data register (I2CDR) to/from a location of Register File/Mem- 184/224 ory addressed through the DMA Address Register (or Register pair) – A post-increment of the DMA Address Register (or Register pair) – A post-decrement of the DMA transaction counter, which contains the number of transactions that have still to be performed. Depending on the value of the DDCISR.DMASTOP bit the DMA feature can be suspended or not (both in transmission and in reception) until the pending bit related to the “Error event” interrupt request is set. The priority level of the DMA features of the I2C interface with respect to the other peripherals and the CPU is the same as programmed in the I2CISR register for the interrupt sources. In the internal priority level order of the peripheral, if DDCISR.DMASTOP=0, DMA has a higher priority with respect to the interrupt sources. Otherwise (if DDCISR.DMASTOP=1), the DMA has a priority lower than “error” event interrupt sources but greater than reception and transmission interrupt sources. Refer to the Interrupt and DMA chapters for details on the priority levels. The DMA features are enabled by setting the corresponding enabling bits (RXDM, TXDM) in the I2CIMR register. It is possible to select also the direction of the DMA transactions. Once the DMA transfer is completed (the transaction counter reaches 0 value), an interrupt request to the CPU is generated. This kind of interrupt is called “End Of Block”. The peripheral sends two different “End Of Block” interrupts depending on the direction of the DMA (Receiving End Of Block Transmitting End Of Block). These interrupt sources have dedicated interrupt pending bits in the I2CIMR register (REOBP, TEOBP) and they are mapped on the same interrupt vectors as respectively “Data Received” and “Peripheral Ready to Transmit” interrupt sources. The same correspondence exists about the internal priority between interrupts. Note: The I2CCR.ITE bit has no effect on the End Of Block interrupts. Moreover, the I2CSR1.EVF flag is not set by the End Of Block interrupts. ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) 8.5.6.1 DMA between Peripheral and Register File If the DMA transaction is made between the peripheral and the Register File, one register is required to hold the DMA Address and one to hold the DMA transaction counter. These two registers must be located in the Register File: – the DMA Address Register in the even addressed register, – the DMA Transaction Counter in the following register (odd address). They are pointed to by the DMA Transaction Counter Pointer Register (I2CRDC register in receiving, I2CTDC register in transmitting) located in the peripheral register page. In order to select the DMA transaction with the Register File, the control bit I2CRDC.RF/MEM in receiving mode or I2CTDC.RF/MEM in transmitting mode must be set. The transaction Counter Register must be initialized with the number of DMA transfers to perform and will be decremented after each transaction. The DMA Address Register must be initialized with the starting address of the DMA table in the Register File, and it is increased after each transaction. These two registers must be located between addresses 00h and DFh of the Register File. When the DMA occurs between Peripheral and Register File, the I2CTDAP register (in transmission) and the I2CRDAP one (in reception) are not used. 8.5.6.2 DMA between Peripheral and Memory Space If the DMA transaction is made between the peripheral and Memory, a register pair is required to hold the DMA Address and another register pair to hold the DMA Transaction counter. These two pairs of registers must be located in the Register File. The DMA Address pair is pointed to by the DMA Address Pointer Register (I2CRDAP register in reception, I2CTDAP register in transmission) located in the peripheral register page; the DMA Transaction Counter pair is pointed to by the DMA Transaction Counter Pointer Register (I2CRDC register in reception, I2CTDC register in transmission) located in the peripheral register page. In order to select the DMA transaction with the Memory Space, the control bit I2CRDC.RF/MEM in receiving mode or I2CTDC.RF/MEM in transmitting mode must be reset. The Transaction Counter registers pair must be initialized with the number of DMA transfers to perform and will be decremented after each transaction. The DMA Address register pair must be initialized with the starting address of the DMA table in the Memory Space, and it is increased after each transaction. These two register pairs must be located between addresses 00h and DFh of the Register File. 8.5.6.3 DMA in Master Receive To correctly manage the reception of the last byte when the DMA in Master Receive mode is used, the following sequence of operations must be performed: 1. The number of data bytes to be received must be set to the effective number of bytes minus one byte. 2. When the Receiving End Of Block condition occurs, the I2CCR.STOP bit must be set and the I2CCR.ACK bit must be reset. The last byte of the reception sequence can be received either using interrupts/polling or using DMA. If the user wants to receive the last byte using DMA, the number of bytes to be received must be set to 1, and the DMA in reception must be reenabled (IMR.RXDM bit set) to receive the last byte. Moreover the Receiving End Of Block interrupt service routine must be designed to recognize and manage the two different End Of Block situations (after the first sequence of data bytes and after the last data byte). 185/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) 8.5.7 Register Description IMPORTANT : 1. To guarantee correct operation, before enabling the peripheral (while I2CCR.PE=0), configure bit7 and bit6 of the I2COAR2 register according to the internal clock INTCLK (for example 11xxxxxxb in the range 14 - 30 MHz). 2. Bit7 of the I2CCR register must be cleared. I2C CONTROL REGISTER (I2CCR) R240 - Read / Write Register Page: 20 Reset Value: 0000 0000 (00h) 7 0 0 0 PE ENGC START ACK STOP ITE Bit 7:6 = Reserved Must be cleared Bit 5 = PE Peripheral Enable. This bit is set and cleared by software. 0: Peripheral disabled (reset value) 1: Master/Slave capability Notes: – When I2CCR.PE=0, all the bits of the I2CCR register and the I2CSR1-I2CSR2 registers except the STOP bit are reset. All outputs will be released while I2CCR.PE=0 – When I2CCR.PE=1, the corresponding I/O pins are selected by hardware as alternate functions (open drain). – To enable the I2C interface, write the I2CCR register TWICE with I2CCR.PE=1 as the first write only activates the interface (only I2CCR.PE is set). – When PE=1, the FREQ[2:0] and EN10BIT bits in the I2COAR2 and I2CADR registers cannot be written. The value of these bits can be changed only when PE=0. Bit 4 = ENGC General Call address enable. Setting this bit the peripheral works as a slave and the value stored in the I2CADR register is recognized as device address. This bit is set and cleared by software. It is also cleared by hardware when the interface is disabled (I2CCR.PE=0). 0: The address stored in the I2CADR register is ignored (reset value) 186/224 1: The General Call address stored in the I2CADR register will be acknowledged Note: The correct value (usually 00h) must be written in the I2CADR register before enabling the General Call feature. Bit 3 = START Generation of a Start condition. This bit is set and cleared by software. It is also cleared by hardware when the interface is disabled (I2CCR.PE=0) or when the Start condition is sent (with interrupt generation if ITE=1). – In master mode: 0: No start generation 1: Repeated start generation – In slave mode: 0: No start generation (reset value) 1: Start generation when the bus is free Bit 2 = ACK Acknowledge enable. This bit is set and cleared by software. It is also cleared by hardware when the interface is disabled (I2CCR.PE=0). 0: No acknowledge returned (reset value) 1: Acknowledge returned after an address byte or a data byte is received Bit 1 = STOP Generation of a Stop condition. This bit is set and cleared by software. It is also cleared by hardware in master mode. It is not cleared when the interface is disabled (I2CCR.PE=0). In slave mode, this bit must be set only when I2CSR1.BTF=1. – In master mode: 0: No stop generation 1: Stop generation after the current byte transfer or after the current Start condition is sent. The STOP bit is cleared by hardware when the Stop condition is sent. – In slave mode: 0: No stop generation (reset value) 1: Release SCL and SDA lines after the current byte transfer (I2CSR1.BTF=1). In this mode the STOP bit has to be cleared by software. ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) Bit 0 = ITE Interrupt Enable. The ITE bit enables the generation of interrupts. This bit is set and cleared by software and cleared by hardware when the interface is disabled (I2CCR.PE=0). 0: Interrupts disabled (reset value) 1: Interrupts enabled after any of the following conditions: – Byte received or to be transmitted (I2CSR1.BTF and I2CSR1.EVF flags = 1) – Address matched in Slave mode while I2CCR.ACK=1 (I2CSR1.ADSL and I2CSR1.EVF flags = 1) – Start condition generated (I2CSR1.SB and I2CSR1.EVF flags = 1) – No acknowledge received after byte transmission (I2CSR2.AF and I2CSR1.EVF flags = 1) – Stop detected in Slave mode (I2CSR2.STOPF and I2CSR1.EVF flags = 1) – Arbitration lost in Master mode (I2CSR2.ARLO and I2CSR1.EVF flags = 1) – Bus error, Start or Stop condition detected during data transfer (I2CSR2.BERR and I2CSR1.EVF flags = 1) – Master has sent header byte (I2CSR1.ADD10 and I2CSR1.EVF flags = 1) – Address byte successfully transmitted in Master mode. (I2CSR1.EVF = 1 and I2CSR2.ADDTX = 1) SCL is held low when the ADDTX flag of the I2CSR2 register or the ADD10, SB, BTF or ADSL flags of I2CSR1 register are set (See Figure 84) or when the DMA is not complete. The transfer is suspended in all cases except when the BTF bit is set and the DMA is enabled. In this case the event routine must suspend the DMA transfer if it is required. I2C STATUS REGISTER 1 (I2CSR1) R241 - Read Only Register Page: 20 Reset Value: 0000 0000 (00h) 7 EVF 0 ADD10 TRA BUSY BTF ADSL M/SL SB Note: Some bits of this register are reset by a read operation of the register. Care must be taken when using instructions that work on single bit. Some of them perform a read of all the bits of the register before modifying or testing the wanted bit. So other bits of the register could be affected by the operation. In the same way, the test/compare operations perform a read operation. Moreover, if some interrupt events occur while the register is read, the corresponding flags are set, and correctly read, but if the read operation resets the flags, no interrupt request occurs. Bit 7 = EVF Event Flag. This bit is set by hardware as soon as an event ( listed below or described in Figure 84) occurs. It is cleared by software when all event conditions that set the flag are cleared. It is also cleared by hardware when the interface is disabled (I2CCR.PE=0). 0: No event 1: One of the following events has occurred: – Byte received or to be transmitted (I2CSR1.BTF and I2CSR1.EVF flags = 1) – Address matched in Slave mode while I2CCR.ACK=1 (I2CSR1.ADSL and I2CSR1.EVF flags = 1) – Start condition generated (I2CSR1.SB and I2CSR1.EVF flags = 1) – No acknowledge received after byte transmission (I2CSR2.AF and I2CSR1.EVF flags = 1) – Stop detected in Slave mode (I2CSR2.STOPF and I2CSR1.EVF flags = 1) – Arbitration lost in Master mode (I2CSR2.ARLO and I2CSR1.EVF flags = 1) – Bus error, Start or Stop condition detected during data transfer (I2CSR2.BERR and I2CSR1.EVF flags = 1) – Master has sent header byte (I2CSR1.ADD10 and I2CSR1.EVF flags = 1) 187/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) – Address byte successfully transmitted in Master mode. (I2CSR1.EVF = 1 and I2CSR2.ADDTX=1) Bit 6 = ADD10 10-bit addressing in Master mode. This bit is set when the master has sent the first byte in 10-bit address mode. An interrupt is generated if ITE=1. It is cleared by software reading I2CSR1 register followed by a write in the I2CDR register of the second address byte. It is also cleared by hardware when peripheral is disabled (I2CCR.PE=0) or when the STOPF bit is set. 0: No ADD10 event occurred. 1: Master has sent first address byte (header). Bit 5 = TRA Transmitter/ Receiver. When BTF flag of this register is set and also TRA=1, then a data byte has to be transmitted. It is cleared automatically when BTF is cleared. It is also cleared by hardware after the STOPF flag of I2CSR2 register is set, loss of bus arbitration (ARLO flag of I2CSR2 register is set) or when the interface is disabled (I2CCR.PE=0). 0: A data byte is received (if I2CSR1.BTF=1) 1: A data byte can be transmitted (if I2CSR1.BTF=1) Bit 4 = BUSY Bus Busy. It indicates a communication in progress on the bus. The detection of the communications is always active (even if the peripheral is disabled). This bit is set by hardware on detection of a Start condition and cleared by hardware on detection of a Stop condition. This information is still updated when the interface is disabled (I2CCR.PE=0). 0: No communication on the bus 1: Communication ongoing on the bus Bit 3 = BTF Byte Transfer Finished. This bit is set by hardware as soon as a byte is correctly received or before the transmission of a data byte with interrupt generation if ITE=1. It is cleared by software reading I2CSR1 register followed by a read or write of I2CDR register or when DMA is complete. It is also cleared by hardware when the interface is disabled (I2CCR.PE=0). 188/224 – Following a byte transmission, this bit is set after reception of the acknowledge clock pulse. BTF is cleared by reading I2CSR1 register followed by writing the next byte in I2CDR register or when DMA is complete. – Following a byte reception, this bit is set after transmission of the acknowledge clock pulse if ACK=1. BTF is cleared by reading I2CSR1 register followed by reading the byte from I2CDR register or when DMA is complete. The SCL line is held low while I2CSR1.BTF=1. 0: Byte transfer not done 1: Byte transfer succeeded Bit 2 = ADSL Address matched (Slave mode). This bit is set by hardware if the received slave address matches the I2COAR1/I2COAR2 register content or a General Call address. An interrupt is generated if ITE=1. It is cleared by software reading I2CSR1 register or by hardware when the interface is disabled (I2CCR.PE=0). The SCL line is held low while ADSL=1. 0: Address mismatched or not received 1: Received address matched Bit 1 = M/SL Master/Slave. This bit is set by hardware as soon as the interface is in Master mode (Start condition generated on the lines after the I2CCR.START bit is set). It is cleared by hardware after detecting a Stop condition on the bus or a loss of arbitration (ARLO=1). It is also cleared when the interface is disabled (I2CCR.PE=0). 0: Slave mode 1: Master mode Bit 0 = SB Start Bit (Master mode). This bit is set by hardware as soon as the Start condition is generated (following a write of START=1 if the bus is free). An interrupt is generated if ITE=1. It is cleared by software reading I2CSR1 register followed by writing the address byte in I2CDR register. It is also cleared by hardware when the interface is disabled (I2CCR.PE=0). The SCL line is held low while SB=1. 0: No Start condition 1: Start condition generated ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) I2C STATUS REGISTER 2 (I2CSR2) R242 - Read Only Register Page: 20 Reset Value: 0000 0000 (00h) 7 0 0 0 ADDTX AF STOPF ARLO BERR GCAL Note: Some bits of this register are reset by a read operation of the register. Care must be taken when using instructions that work on single bit. Some of them perform a read of all the bits of the register before modifying or testing the wanted bit. So other bits of the register could be affected by the operation. In the same way, the test/compare operations perform a read operation. Moreover, if some interrupt events occur while the register is read, the corresponding flags are set, and correctly read, but if the read operation resets the flags, no interrupt request occurs. Bit 7:6 = Reserved. Forced to 0 by hardware. Bit 5 = ADDTX Address or 2nd header transmitted in Master mode. This bit is set by hardware when the peripheral, enabled in Master mode, has received the acknowledge relative to: – Address byte in 7-bit mode – Address or 2nd header byte in 10-bit mode. 0: No address or 2nd header byte transmitted 1: Address or 2nd header byte transmitted. Bit 4 = AF Acknowledge Failure. This bit is set by hardware when no acknowledge is returned. An interrupt is generated if ITE=1. It is cleared by software reading I2CSR2 register after the falling edge of the acknowledge SCL pulse, or by hardware when the interface is disabled (I2CCR.PE=0). The SCL line is not held low while AF=1. 0: No acknowledge failure detected 1: A data or address byte was not acknowledged Bit 3 = STOPF Stop Detection (Slave mode). This bit is set by hardware when a Stop condition is detected on the bus after an acknowledge. An interrupt is generated if ITE=1. It is cleared by software reading I2CSR2 register or by hardware when the interface is disabled (I2CCR.PE=0). The SCL line is not held low while STOPF=1. 0: No Stop condition detected 1: Stop condition detected (while slave receiver) Bit 2 = ARLO Arbitration Lost. This bit is set by hardware when the interface (in master mode) loses the arbitration of the bus to another master. An interrupt is generated if ITE=1. It is cleared by software reading I2CSR2 register or by hardware when the interface is disabled (I2CCR.PE=0). After an ARLO event the interface switches back automatically to Slave mode (M/SL=0). The SCL line is not held low while ARLO=1. 0: No arbitration lost detected 1: Arbitration lost detected Bit 1 = BERR Bus Error. This bit is set by hardware when the interface detects a Start or Stop condition during a byte transfer. An interrupt is generated if ITE=1. It is cleared by software reading I2CSR2 register or by hardware when the interface is disabled (I2CCR.PE=0). The SCL line is not held low while BERR=1. Note: If a misplaced start condition is detected, also the ARLO flag is set; moreover, if a misplaced stop condition is placed on the acknowledge SCL pulse, also the AF flag is set. 0: No Start or Stop condition detected during byte transfer 1: Start or Stop condition detected during byte transfer Bit 0 = GCAL General Call address matched. This bit is set by hardware after an address matches with the value stored in the I2CADR register while ENGC=1. In the I2CADR the General Call address must be placed before enabling the peripheral. It is cleared by hardware after the detection of a Stop condition, or when the peripheral is disabled (I2CCR.PE=0). 0: No match 1: General Call address matched. 189/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) I2C CLOCK CONTROL REGISTER (I2CCCR) R243 - Read / Write Register Page: 20 Reset Value: 0000 0000 (00h) 7 FM/SM I2C OWN ADDRESS REGISTER 1 (I2COAR1) R244 - Read / Write Register Page:20 Reset Value: 0000 0000 (00h) 0 CC6 CC5 CC4 CC3 7 0 CC2 CC1 CC0 ADD7 ADD6 ADD5 ADD4 ADD3 ADD2 ADD1 ADD0 I2C mode. Bit 7 = FM/SM Fast/Standard This bit is used to select between fast and standard mode. See the description of the following bits. It is set and cleared by software. It is not cleared when the peripheral is disabled (I2CCR.PE=0) Bit 6:0 = CC[6:0] 9-bit divider programming Implementation of a programmable clock divider. These bits and the CC[8:7] bits of the I2CECCR register select the speed of the bus (FSCL) depending on the I2C mode. They are not cleared when the interface is disabled (I2CCR.PE=0). – Standard mode (FM/SM=0): FSCL <= 100kHz FSCL = INTCLK/(2x([CC8..CC0]+2)) – Fast mode (FM/SM=1): FSCL > 100kHz FSCL = INTCLK/(3x([CC8..CC0]+2)) Note: The programmed frequency is available with no load on SCL and SDA pins. 190/224 7-bit Addressing Mode Bit 7:1 = ADD[7:1] Interface address. These bits define the I2C bus address of the interface. They are not cleared when the interface is disabled (I2CCR.PE=0). Bit 0 = ADD0 Address direction bit. This bit is don’t care; the interface acknowledges either 0 or 1. It is not cleared when the interface is disabled (I2CCR.PE=0). Note: Address 01h is always ignored. 10-bit Addressing Mode Bit 7:0 = ADD[7:0] Interface address. These are the least significant bits of the I2Cbus address of the interface. They are not cleared when the interface is disabled (I2CCR.PE=0). ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) I2C OWN ADDRESS REGISTER 2 (I2COAR2) R245 - Read / Write Register Page: 20 Reset Value: 0000 0000 (00h) 7 0 FREQ1 FREQ0 EN10BIT FREQ2 0 ADD9 ADD8 0 Bit 7:6,4 = FREQ[2:0] Frequency bits. IMPORTANT: To guarantee correct operation, set these bits before enabling the interface (while I2CCR.PE=0). These bits can be set only when the interface is disabled (I2CCR.PE=0). To configure the interface to I2C specified delays, select the value corresponding to the microcontroller internal frequency INTCLK. INTCLK Range (MHz) 2.5 - 6 6- 10 10- 14 14 - 30 30 - 50 FREQ2 FREQ1 FREQ0 0 0 0 0 1 0 0 1 1 0 0 1 0 1 0 Note: If an incorrect value, with respect to the MCU internal frequency, is written in these bits, the timings of the peripheral will not meet the I2C bus standard requirements. Note: The FREQ[2:0] = 101, 110, 111 configurations must not be used. Bit 5 = EN10BIT Enable 10-bit I2Cbus mode. When this bit is set, the 10-bit I2Cbus mode is enabled. This bit can be written only when the peripheral is disabled (I2CCR.PE=0). 0: 7-bit mode selected 1: 10-bit mode selected Bit 4:3 = Reserved. Bit 2:1 = ADD[9:8] Interface address. These are the most significant bits of the I2Cbus address of the interface (10-bit mode only). They are not cleared when the interface is disabled (I2CCR.PE=0). Bit 0 = Reserved. I2C DATA REGISTER (I2CDR) R246 - Read / Write Register Page: 20 Reset Value: 0000 0000 (00h) 7 DR7 0 DR6 DR5 DR4 DR3 DR2 DR1 DR0 Bit 7:0 = DR[7:0] I2C Data. – In transmitter mode: I2CDR contains the next byte of data to be transmitted. The byte transmission begins after the microcontroller has written in I2CDR or on the next rising edge of the clock if DMA is complete. – In receiver mode: I2CDR contains the last byte of data received. The next byte receipt begins after the I2CDR read by the microcontroller or on the next rising edge of the clock if DMA is complete. GENERAL CALL ADDRESS (I2CADR) R247 - Read / Write Register Page: 20 Reset Value: 1010 0000 (A0h) 7 0 ADR7 ADR6 ADR5 ADR4 ADR3 ADR2 ADR1 ADR0 Bit 7:0 = ADR[7:0] Interface address. These bits define the I2Cbus General Call address of the interface. It must be written with the correct value depending on the use of the peripheral.If the peripheral is used in I2C bus mode, the 00h value must be loaded as General Call address. The customer could load the register with other values. The bits can be written only when the peripheral is disabled (I2CCR.PE=0) The ADR0 bit is don’t care; the interface acknowledges either 0 or 1. Note: Address 01h is always ignored. 191/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) INTERRUPT STATUS REGISTER (I2CISR) R248 - Read / Write Register Page: 20 Reset Value: 1xxx xxxx (xxh) 7 DMASTOP PRL2 PRL1 PRL0 0 0 IERRP IRXP ITXP Bit 7 = DMASTOP DMA suspended mode. This bit selects between DMA suspended mode and DMA not suspended mode. In DMA Suspended mode, if the error interrupt pending bit (I2CISR.IERRP) is set, no DMA request is performed. DMA requests are performed only when IERRP=0. Moreover the “Error Condition” interrupt source has a higher priority than the DMA. In DMA Not-Suspended mode, the status of IERRP bit has no effect on DMA requests. Moreover the DMA has higher priority with respect to other interrupt sources. 0: DMA Not-Suspended mode 1: DMA Suspended mode Bit 6:4 = PRL[2:0] Interrupt/DMA Priority Bits. The priority is encoded with these three bits. The value of “0” has the highest priority, the value “7” has no priority. After the setting of this priority level, the priorities between the different Interrupt/ DMA sources is hardware defined according with the following scheme: – Error condition Interrupt (If DMASTOP=1) (Highest priority) – Receiver DMA request – Transmitter DMA request – Error Condition Interrupt (If DMASTOP=0 – Data Received/Receiver End Of Block – Peripheral Ready To Transmit/Transmitter End Of Block (Lowest priority) Bit 3 = Reserved. Must be cleared. 192/224 Bit 2 = IERRP Error Condition pending bit 0: No error 1: Error event detected (if ITE=1) Note: Depending on the status of the I2CISR.DMASTOP bit, this flag can suspend or not suspend the DMA requests. Note: The Interrupt pending bits can be reset by writing a “0” but is not possible to write a “1”. It is mandatory to clear the interrupt source by writing a “0” in the pending bit when executing the interrupt service routine. When serving an interrupt routine, the user should reset ONLY the pending bit related to the served interrupt routine (and not reset the other pending bits). To detect the specific error condition that occurred, the flag bits of the I2CSR1 and I2CSR2 register should be checked. Note: The IERRP pending bit is forced high until the error event flags are set (ADSL and SB flags in the I2CSR1 register, SCLF, ADDTX, AF, STOPF, ARLO and BERR flags in the I2CSR2 register). If at least one flag is set, it is not possible to reset the IERRP bit. Bit 1 = IRXP Data Received pending bit 0: No data received 1: data received (if ITE=1). Bit 0 = ITXP Peripheral Ready To Transmit pending bit 0: Peripheral not ready to transmit 1: Peripheral ready to transmit a data byte (if ITE=1). ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) INTERRUPT VECTOR REGISTER (I2CIVR) R249 - Read / Write Register Page: 20 Reset Value: Undefined 7 V7 (DMA between peripheral and Register file), this register has no meaning. See Section 8.5.6.1 for more details on the use of this register. 0 V6 V5 V4 V3 EV2 EV1 0 Bit 7:3 = V[7:3] Interrupt Vector Base Address. User programmable interrupt vector bits. These are the five more significant bits of the interrupt vector base address. They must be set before enabling the interrupt features. Bit 2:1 = EV[2:1] Encoded Interrupt Source. These Read-Only bits are set by hardware according to the interrupt source: – 01: error condition detected – 10: data received – 11: peripheral ready to transmit Bit 0 = RPS Receiver DMA Memory Pointer Selector. If memory has been selected for DMA transfer (DDCRDC.RF/MEM = 0) then: 0: Select ISR register for Receiver DMA transfer address extension. 1: Select DMASR register for Receiver DMA transfer address extension. RECEIVER DMA TRANSACTION COUNTER REGISTER (I2CRDC) R251 - Read / Write Register Page: 20 Reset Value: Undefined 7 0 RC7 RC6 RC5 RC4 RC3 RC2 RC1 RF/MEM Bit 0 = Reserved. Forced by hardware to 0. RECEIVER DMA SOURCE ADDRESS POINTER REGISTER (I2CRDAP) R250 - Read / Write Register Page: 20 Reset Value: Undefined 7 RA7 RA6 RA5 RA4 RA3 RA2 RA1 0 RPS Bit 7:1 = RA[7:1] Receiver DMA Address Pointer . I2CRDAP contains the address of the pointer (in the Register File) of the Receiver DMA data source when the DMA is selected between the peripheral and the Memory Space. Otherwise, Bit 7:1 = RC[7:1] Receiver DMA Counter Pointer. I2CRDC contains the address of the pointer (in the Register File) of the DMA receiver transaction counter when the DMA between Peripheral and Memory Space is selected. Otherwise (DMA between Peripheral and Register File), this register points to a pair of registers that are used as DMA Address register and DMA Transaction Counter. See Section 8.5.6.1 and Section 8.5.6.2 for more details on the use of this register. Bit 0 = RF/MEM Receiver Register File/ Memory Selector. 0: DMA towards Memory 1: DMA towards Register file 193/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) TRANSMITTER DMA SOURCE POINTER REGISTER (I2CTDAP) R252 - Read / Write Register Page: 20 Reset Value: Undefined 7 TA7 ADDRESS 0 TA6 TA5 TA4 TA3 TA2 TA1 TPS Bit 7:1= TA[7:1] Transmit DMA Address Pointer. I2CTDAP contains the address of the pointer (in the Register File) of the Transmitter DMA data source when the DMA between the peripheral and the Memory Space is selected. Otherwise (DMA between the peripheral and Register file), this register has no meaning. See Section 8.5.6.2 for more details on the use of this register. Bit 0 = TPS Transmitter DMA Memory Pointer Selector. If memory has been selected for DMA transfer (DDCTDC.RF/MEM = 0) then: 0: Select ISR register for transmitter DMA transfer address extension. 1: Select DMASR register for transmitter DMA transfer address extension. TRANSMITTER DMA TRANSACTION COUNTER REGISTER (I2CTDC) R253 - Read / Write Register Page: 20 Reset Value: Undefined 7 0 TC7 TC6 TC5 TC4 TC3 TC2 TC1 RF/MEM Bit 7:1 = TC[7:1] Transmit DMA Counter Pointer. I2CTDC contains the address of the pointer (in the Register File) of the DMA transmitter transaction counter when the DMA between Peripheral and Memory Space is selected. Otherwise, if the DMA between Peripheral and Register File is selected, this register points to a pair of registers that are used as DMA Address register and DMA Transaction Counter. See Section 8.5.6.1 and Section 8.5.6.2 for more details on the use of this register. Bit 0 = RF/MEM Transmitter Register File/ Memory Selector. 0: DMA from Memory 1: DMA from Register file EXTENDED CLOCK CONTROL REGISTER (I2CECCR) R254 - Read / Write Register Page: 20 Reset Value: 0000 0000 (00h) 7 0 0 0 0 0 0 0 CC8 CC7 Bit 7:2 = Reserved. Must always be cleared. Bit 1:0 = CC[8:7] 9-bit divider programming Implementation of a programmable clock divider. These bits and the CC[6:0] bits of the I2CCCR register select the speed of the bus (FSCL). For a description of the use of these bits, see the I2CCCR register. They are not cleared when the interface is disabled (I2CCCR.PE=0). 194/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) INTERRUPT MASK REGISTER (I2CIMR) R255 - Read / Write Register Page: 20 Reset Value: 00xx 0000 (x0h) 7 RXD TXD REOBP TEOBP M M interrupt request. Note: TEOBP can only be written to “0”. 0: End of block not reached 1: End of data block in DMA transmitter detected. 0 0 IERR IRX M M ITX M Bit 7 = RXDM Receiver DMA Mask. 0: DMA reception disable. 1: DMA reception enable RXDM is reset by hardware when the transaction counter value decrements to zero, that is when a Receiver End Of Block interrupt is issued. Bit 6 = TXDM Transmitter DMA Mask. 0: DMA transmission disable. 1: DMA transmission enable. TXDM is reset by hardware when the transaction counter value decrements to zero, that is when a Transmitter End Of Block interrupt is issued. Bit 5 = REOBP Receiver DMA End Of Block Flag. REOBP should be reset by software in order to avoid undesired interrupt routines, especially in initialization routine (after reset) and after entering the End Of Block interrupt routine.Writing “0” in this bit will cancel the interrupt request Note: REOBP can only be written to “0”. 0: End of block not reached. 1: End of data block in DMA receiver detected Bit 3 = Reserved. This bit must be cleared. Bit 2 = IERRM Error Condition interrupt mask bit. This bit enables/ disables the Error interrupt. 0: Error interrupt disabled. 1: Error Interrupt enabled. Bit 1 = IRXM Data Received interrupt mask bit. This bit enables/ disables the Data Received and Receive DMA End of Block interrupts. 0: Interrupts disabled 1: Interrupts enabled Note: This bit has no effect on DMA transfer Bit 0 = ITXM Peripheral Ready To Transmit interrupt mask bit. This bit enables/ disables the Peripheral Ready To Transmit and Transmit DMA End of Block interrupts. 0: Interrupts disabled 1: Interrupts enabled Note: This bit has no effect on DMA transfer. Bit 4 = TEOBP Transmitter DMA End Of Block TEOBP should be reset by software in order to avoid undesired interrupt routines, especially in initialization routine (after reset) and after entering the End Of Block interrupt routine.Writing “0” will cancel the 195/224 ST92163 - I2C BUS INTERFACE I2C INTERFACE (Cont’d) Table 36. I2C BUS Register Map and Reset Values Address (Hex.) F0h F1h F2h F3h F4h F5h F6h F7h F8h F9h FAh FBh FCh FDh FEh FFh 196/224 Register Name 7 6 5 4 3 2 1 0 I2CCR - - PE ENGC START ACK STOP ITE Reset Value 0 0 0 0 0 0 0 0 I2CSR1 EVF ADD10 TRA BUSY BTF ADSL M/SL SB Reset Value 0 0 0 0 0 0 0 0 I2CSR2 - 0 ADDTX AF STOPF ARLO BERR GCAL Reset Value 0 0 0 0 0 0 0 0 I2CCCR FM/SM CC6 CC5 CC4 CC3 CC2 CC1 CC0 Reset Value 0 0 0 0 0 0 0 0 I2COAR1 ADD7 ADD6 ADD5 ADD4 ADD3 ADD2 ADD1 ADD0 Reset Value 0 0 0 0 0 0 0 0 I2COAR2 FREQ1 FREQ0 EN10BIT FREQ2 0 ADD9 ADD8 0 Reset Value 0 0 0 0 0 0 0 0 I2CDR DR7 DR6 DR5 DR4 DR3 DR2 DR1 DR0 Reset Value 0 0 0 0 0 0 0 0 I2CADR ADR7 ADR6 ADR5 ADR4 ADR3 ADR2 ADR1 ADR0 Reset Value 1 0 1 0 0 0 0 0 I2CISR DMASTO P PRL2 PRL1 PRL0 IERRP IRXP ITXP Reset Value 1 X X X X X X X I2CIVR V7 V6 V5 V4 V3 EV2 EV1 0 Reset Value X X X X X X X 0 I2CRDAP RA7 RA6 RA5 RA4 RA3 RA2 RA1 RPS Reset Value X X X X X X X X I2CRDC RC7 RC6 RC5 RC4 RC3 RC2 RC1 RF/MEM Reset Value X X X X X X X X I2CTDAP TA7 TA6 TA5 TA4 TA3 TA2 TA1 TPS Reset Value X X X X X X X X I2CTDC TC7 TC6 TC5 TC4 TC3 TC2 TC1 RF/MEM Reset Value X X X X X X X X 0 0 0 0 0 0 CC8 CC7 0 0 0 0 0 0 0 0 I2CIMR RXDM TXDM REOBP TEOBP IERRM IRXM ITXM Reset Value 0 0 X X 0 0 0 I2CECCR 0 ST92163 - A/D CONVERTER (A/D) 8.6 A/D CONVERTER (A/D) 8.6.1 Introduction The 8 bit Analog to Digital Converter uses a fully differential analog configuration for the best noise immunity and precision performance. The analog voltage references of the converter are connected to the internal AVDD & AVSS analog supply pins of the chip if they are available, otherwise to the ordinary VDD and VSS supply pins of the chip. The guaranteed accuracy depends on the device (see Electrical Characteristics). A fast Sample/Hold allows quick signal sampling for minimum warping effect and conversion error. 8.6.2 Main Features 8-bit resolution A/D Converter ■ Single Conversion Time (including Sampling Time): – 138 internal system clock periods in slow mode (~5.6 µs @25Mhz internal system clock); – 78 INTCLK periods in fast mode (~6.5 µs @ 12MHZ internal system clock) ■ Sample/Hold: Tsample= – 84 INTCLK periods in slow mode (~3.4 µs @25Mhz internal system clock) – 48 INTCLK periods in fast mode (~4 µs @12Mhz internal system clock) ■ Up to 8 Analog Inputs (the number of inputs is device dependent, see device pinout) ■ ■ ■ ■ ■ ■ Single/Continuous Conversion Mode External/Internal source Trigger (Alternate synchronization) Power Down mode (Zero Power Consumption) 1 Control Logic Register 1 Data Register 8.6.3 General Description Depending on the device, up to 8 analog inputs can be selected by software. Different conversion modes are provided: single, continuous, or triggered. The continuous mode performs a continuous conversion flow of the selected channel, while in the single mode the selected channel is converted once and then the logic waits for a new hardware or software restart. A data register (ADDTR) is available, mapped in page 62, allowing data storage (in single or continuous mode). The start conversion event can be managed by software, writing the START/STOP bit of the Control Logic Register or by hardware using either: – An external signal on the EXTRG triggered input (negative edge sensitive) connected as an Alternate Function to an I/O port bit – An On Chip Event generated by another peripheral, such as the MFT (Multifunction Timer). Figure 86. A/D Converter Block Diagram n SUCCE SSIVE APPROXI MATION REGISTER ST9 BUS Ain0 S/H DATA REGIST ER ANALOG MUX Ain1 Ainx EXTRG CONTROL LOGIC INTRG (On Chip Event) 197/224 ST92163 - A/D CONVERTER (A/D) A/D CONVERTER (Cont’d) The conversion technique used is successive approximation, with AC coupled analog fully differential comparators blocks plus a Sample and Hold logic and a reference generator. The internal reference (DAC) is based on the use of a binary-ratioed capacitor array. This technique allows the specified monotonicity (using the same ratioed capacitors as sampling capacitor). A Power Down programmable bit sets the A/D converter analog section to a zero consumption idle status. 8.6.3.1 Operating Modes The two main operating modes, single and continuous, can be selected by writing 0 (reset value) or 1 into the CONT bit of the Control Logic Register. Single Mode In single mode (CONT=0 in ADCLR) the STR bit is forced to ’0’ after the end of channel i-th conversion; then the A/D waits for a new start event. This mode is useful when a set of signals must be sampled at a fixed frequency imposed by a Timer unit or an external generator (through the alternate synchronization feature). A simple software routine monitoring the STR bit can be used to save the current value before a new conversion ends (so to create a signal samples table within the internal memory or the Register File). Furthermore, if the R242.0 bit (register AD-INT, bit 0) is set, at the end of conversion a negative edge on the connected external interrupt channel (see Interrupts Chapter) is generated to allow the reading of the converted data by means of an interrupt routine. Continuous Mode In continuous mode (CONT=1 in ADCLR) a continuous conversion flow is entered by a start event on the selected channel until the STR bit is reset by software. At the end of each conversion, the Data Register (ADCDR) content is updated with the last conversion result, while the former value is lost. When the conversion flow is stopped, an interrupt request is generated with the same modality previously described. 8.6.3.2 Alternate Synchronization This feature is available in both single/continuous modes. The negative edge of external EXTRG signal or the occurrence of an on-chip event generated by another peripheral can be used to synchronize the conversion start with a trigger pulse. 198/224 These events can be enabled or masked by programming the TRG bit in the ADCLR Register. The effect of alternate synchronization is to set the STR bit, which is cleared by hardware at the end of each conversion in single mode. In continuous mode any trigger pulse following the first one will be ignored. The synchronization source must provide a pulse (1.5 internal system clock, 125ns @ 12 MHz internal clock) of minimum width, and a period greater (in single mode) than the conversion time (~6.5us @ 12 MHz internal clock). If a trigger occurs when the STR bit is still ’1’ (conversions still in progress), it is ignored (see Electrical Characteristics). WARNING: If the EXTRG or INTRG signals are already active when TRG bit is set, the conversion starts immediately. 8.6.3.3 Power-Up Operations Before enabling any A/D operation mode, set the POW bit of the ADCLR Register at least 60 µs before the first conversion starts to enable the biasing circuits inside the analog section of the converter. Clearing the POW bit is useful when the A/D is not used so reducing the total chip power consumption. This state is also the reset configuration and it is forced by hardware when the core is in HALT state (after a HALT instruction execution). 8.6.3.4 Register Mapping It is possible to have two independent A/D converters in the same device. In this case they are named A/D 0 and A/D 1. If the device has one A/D converter it uses the register addresses of A/D 0. The register map is the following: Register Address ADn Page 62 (3Eh) F0h A/D 0 ADDTR0 F1h A/D 0 ADCLR0 F2h A/D 0 ADINT0 F3-F7h A/D 0 Reserved F8h A/D 1 ADDTR1 F9h A/D 1 ADCLR1 FAh A/D 1 ADINT1 FB-FFh A/D 1 Reserved If two A/D converters are present, the registers are renamed, adding the suffix 0 to the A/D 0 registers and 1 to the A/D 1 registers. ST92163 - A/D CONVERTER (A/D) A/D CONVERTER (Cont’d) 8.6.4 Register Description A/D CONTROL LOGIC REGISTER (ADCLR) R241 - Read/Write Register Page: 62 Reset value: 0000 0000 (00h) 7 C2 0 C1 C0 FS TRG POW CONT STR This 8-bit register manages the A/D logic operations. Any write operation to it will cause the current conversion to be aborted and the logic to be re-initialized to the starting configuration. Bit 7:5 = C[2:0]: Channel Address. These bits are set and cleared by software. They select channel i conversion as follows: C2 0 0 0 0 1 1 1 1 C1 0 0 1 1 0 0 1 1 C0 0 1 0 1 0 1 0 1 Note: Triggering by on chip event is available on devices with the multifunction timer (MFT) peripheral. Bit 2 = POW: Power Enable. This bit is set and cleared by software. 0: Disables all power consuming logic. 1: Enables the A/D logic and analog circuitry. Channel Enabled Channel 0 Channel 1 Channel 2 Channel 3 Channel 4 Channel 5 Channel 6 Channel 7 Bit 4 = FS: Fast/Slow. This bit is set and cleared by software. 0: Fast mode. Single conversion time: 78 x INTCLK (5.75µs at INTCLK = 12 MHz) 1: Slow mode. Single conversion time: 138 x INTCLK (11.5µs at INTCLK = 12 MHz) Note: Fast conversion mode is only allowed for internal speeds which do not exceed 12 MHz. Bit 3 = TRG: External/Internal Trigger Enable. This bit is set and cleared by software. 0: External/Internal Trigger disabled. 1: Either a negative (falling) edge on the EXTRG pin or an On Chip Event writes a “1” into the STR bit, enabling start of conversion. Bit 1 = CONT: Continuous/Single Mode Select. This bit it set and cleared by software. 0: Single mode: after the current conversion ends, the STR bit is reset by hardware and the converter logic is put in a wait status. To start another conversion, the STR bit has to be set by software or hardware. 1: Select Continuous Mode, a continuous flow of A/D conversions on the selected channel, starting when the STR bit is set. Bit 0 = STR: Start/Stop. This bit is set and cleared by software. It is also set by hardware when the A/D is synchronized with an external/internal trigger. 0: Stop conversion on channel i. An interrupt is generated if the STR was previously set and the AD-INT bit is set. 1: Start conversion on channel i WARNING: When accessing this register, it is recommended to keep the related A/D interrupt channel masked or disabled to avoid spurious interrupt requests. A/D CHANNEL i DATA REGISTER (ADDTR) R240 - Read/Write Register Page: 62 Reset value: undefined 7 R.7 0 R.6 R.5 R.4 R.3 R.2 R.1 R.0 The result of the conversion of the selected channel is stored in the 8-bit ADDTR, which is reloaded with a new value every time a conversion ends. 199/224 ST92163 - A/D CONVERTER (A/D) A/D CONVERTER (Cont’d) A/D INTERRUPT REGISTER (ADINT) Register Page: 62 R242 - Read/write Reset value: 0000 0001 (01h) 7 0 0 0 0 0 0 0 0 AD-INT Bit 7:1 = Reserved. Bit 0 = AD-INT: AD Converter Interrupt Enable . This bit is set and cleared by software. It allows the interrupt source to be switched between the A/D Converter and an external interrupt pin (See Interrupts chapter). 0: A/D Interrupt disabled. External pin selected as interrupt source. 1: A/D Interrupt enabled 200/224 ST92163 - ELECTRICAL CHARACTERISTICS 9 ELECTRICAL CHARACTERISTICS The ST92163 device contains circuitry to protect the inputs against damage due to high static voltage or electric field. Nevertheless it is advised to take normal precautions and to avoid applying to this high impedance voltage circuit any voltage higher than the maximum rated voltages. It is recommended for proper operation that VIN and VOUT be constrained to the range: VSS ≤ (VIN or VOUT) ≤ VDD To enhance reliability of operation, it is recommended to connect unused inputs to an appropriate logic voltage level such as VSS or VDD. All the voltages in the following table, are referenced to VSS. ABSOLUTE MAXIMUM RATINGS Symbol VDD VI Parameter Input Voltage VAI Analog Input Voltage (A/D Converter) VO Output Voltage TSTG IINJ Value Unit – 0.3 to +7.0 V – 0.3 to VDD +0.3 V Supply Voltage VSS - 0.3 to VDD +0.3 AVSS - 0.3 to AVDD +0.3 V – 0.3 to VDD +0.3 V – 55 to + 150 °C Storage Temperature Pin Injection Current Digital and Analog Input Maximum Accumulated Pin injection Current in the device -5 to +5 mA -50 to +50 mA AVDD A/D Converter Analog Reference VDD -0.3 to VDD +0.3 V AV SS A/D Converter VSS VSS -0.3 to VSS +0.3 V Note: Stresses above those listed as “absolute maximum ratings“ may cause permanent damage to the device. This is a stress rating only and functional operation of the device at these conditions is not implied. Exposure to maximum rating conditions for extended periods may affect device reliability. All voltages are referenced to VSS RECOMMENDED OPERATING CONDITIONS (Normal Voltage Mode, all devices) Symbol TA V DD fINTCLK Value Parameter Operating Temperature Operating Supply Voltage Internal Clock Frequency @ 4.0V - 5.5V Unit Min. Max. 0 70 4.0 0 (1) 5.5 V 24 MHz °C Note 1. 1MHz when A/D is used RECOMMENDED OPERATING CONDITIONS (Low Voltage Modes, devices with suffix L or V)) Symbol TA V DD fINTCLK Parameter Max. Unit 0 70 °C 3.0 4.0 V 8-MHz Low Voltage devices (devices with suffix L) 0 (1) 8 MHz 16-MHz Low Voltage devices (devices with suffix V) 0 (1) 16 MHz Operating Temperature Operating Supply Voltage Internal Clock Frequency @ 3.0V - 4.0V Value Min. Note 1. 1MHz when A/D is used 201/224 ST92163 - ELECTRICAL CHARACTERISTICS DC ELECTRICAL CHARACTERISTICS (VDD = 4.0 - 5.5V, TA = 0°C + 70°C, unless otherwise specified) Symbol Parameter Test Conditio ns Value Min. Typ. Max. Unit VIHCK * Clock Input High Level External Clock 0.7 VDD VDD + 0.3 V VILCK * Clock Input Low Level External Clock – 0.3 0.3 V DD V 2.0 VDD + 0.3 V 0.7 VDD VDD + 0.3 V VIH * Input High Level VIL * Input Low Level TTL CMOS TTL – 0.3 0.8 V CMOS – 0.3 0.3 V DD V VIHRS * RESET Input High Level 0.7 VDD VDD + 0.3 V VILRS * RESET Input Low Level –0.3 0.3 V DD V VHYRS * RESET Input Hysteresis 0.3 1.5 V V OH * Output High Level Push Pull, Iload = – 0.8mA Output Low Level Push Pull or Open Drain, Iload = 1.6mA Output Low Level high sink pins (Port 6) Iload = 10mA IWPU Weak Pull-up Current Bidirectional Weak Pull-up, VOL = 0V – 40 IAPU Active Pull-up Current, for INT0 and INT7 only VIN < 0.8V, under Reset – 80 ILKIO * I/O Pin Input Leakage Input/Tri-State, 0V < VIN < VDD 0V < VIN < VDD VOL I LKRS * RESET Pin Input Leakage ILKA/D * A/D Conv. Input Leakage ILKAP * Active Pull-up Input Leakage 0V < VIN < 0.8V OSCIN Pin Input Leakage 0V < VIN < VDD ILKOS * VDD – 0.8 V 0.4 V 1 V – 200 – 420 µA – 200 – 420 µA – 10 + 10 µA – 30 + 30 µA –3 +3 µA – 10 + 10 µA ±3 µA *For devices with suffix L or V, these characteristics are valid for VDD = 3.0 - 5.5V Note: All I/O Ports are configured in bidirectional weak pull-up mode with no DC load external clock pin (OSCIN ) is driven by square wave external clock. No peripheral working. 202/224 ST92163 - ELECTRICAL CHARACTERISTICS AC ELECTRICAL CHARACTERISTICS (VDD = 4.0 - 5.5V, TA = 0°C + 70°C, unless otherwise specified) Symbol IDD VDD MAX Parameter Run Mode Current @ 24 MHz IWFI WFI Mode @ 24 MHz IHALT HALT Mode Current USB Suspend mode current I SUSP 85 (1) mA (1) mA 20 (2) Unit 100 µA 450 µA (VDD = 3.0 - 3.3 V, TA = 0°C + 70°C) V DD Symbol Unit Parameter TYP MAX IDD Run Mode Current 1.2 1.4 mA/MHz IWFI WFI Mode 0.3 0.4 mA/MHz IHALT HALT Mode Current 100 µA Note: All I/O Ports are configured in bidirectional weak pull-up mode with no DC load, external clock pin (OSCIN) is driven by square wave external clock (1) Preliminary value (2) External pull-up (1.5 Kohms connected to USBV ); Operating conditions: V CC DD = 4.0 - 5.25V ; TA = 25°C AC TEST CONDITIONS TTL INPUT FORCING CONDITI ON 2.4V 0.45V 0.8VDD CMOS INPUT FORCING CONDITI ON 0.2VDD PUSH-PULL OUTPUT TEST CONDITION 3.0V WEAK PULL-UP OUTPUT TEST CONDIT ION 3.0V 0.8V 0.8V 1 SOURCE CURRENT = -0.8mA 0 SINK CURRENT = -1.6mA 1 SOURCE CURRENT = 0 SINK CURRENT = 1.6mA 203/224 ST92163 - ELECTRICAL CHARACTERISTICS ELECTRICAL CHARACTERISTICS (Cont’d) Low Voltage Detector Reset Electrical Specifications* Symbol Parameter Conditions Min Typ Max Unit 3.5 3.7 3.9 V 3.3 3.5 3.7 V 150 200 250 mV LV Reset Trigger V LVDUP VDD rising edge INTCLK = 16MHz LV Reset Trigger VLVDDOWN VDD Falling edge VLVDHyst Hysteresis ** * Not available on devices with suffix L, V or E **Guaranteed by design I2C Interface Electrical specifications Parameter Symbol Standard mode I2C Min Fast mode I2C Max Min Max Unit Low level input voltage: V IL -0.5 1.5 -0.5 1.5 -0.5 0.3 VDD -0.5 0.3 VDD 0.7 VDD VDD+0.5 0.7 VDD VDD+0.5 N/A N/A 0.2 N/A N/A 0.05 VDD T SP N/A N/A 0 ns 50 ns at 3 mA sink current V OL1 0 0.4 0 0.4 at 6 mA sink current V OL2 N/A N/A 0 0.6 250 20+0.1Cb 250 N/A N/A 20+0.1Cb 250 - 10 10 -10 10 µA 10 pF fixed input levels VDD-related input levels V High level input voltage: VDD-related input levels V Hysteresis of Schmitt trigger inputs fixed input levels V HYS VDD-related input levels Pulse width of spikes which must be suppressed by the input filter V Low level output voltage (open drain and open collector) Output fall time from VIH min to VIL max with a bus capacitance from 10 pF to 400 pF with up to 3 mA sink current at VOL1 T OF with up to 6 mA sink current at VOL2 Input current each I/O pin with an input voltage between 0.4V and 0.9 VDD max I Capacitance for each I/O pin C N/A = not applicable 204/224 10 ns V ns ST92163 - ELECTRICAL CHARACTERISTICS ELECTRICAL CHARACTERISTICS (Cont’d) I2C Bus Timings Parameter Bus free time between a STOP and START condition Symbol Standard I2C Min Fast I2C Max Min Max Unit TBUF 4.7 1.3 ms THD:STA 4.0 0.6 µs LOW period of the SCL clock TLOW 4.7 1.3 µs HIGH period of the SCL clock THIGH 4.0 0.6 µs Set-up time for a repeated START condition TSU:STA 4.7 0.6 Data hold time THD:DAT 0 (1) 0 (1) Data set-up time TSU:DAT 250 Rise time of both SDA and SCL signals TR Fall time of both SDA and SCL signals TF Set-up time for STOP condition TSU :STO Capacitive load for each bus line Cb Hold time START condition. After this period, the first clock pulse is generated µs 0.9(2) ns 20+0.1Cb 300 ns 20+0.1Cb 300 100 1000 300 4.0 ns 0.6 400 ns ns 400 pF 1)The device must internally provide a hold time of at least 300 ns for the SDA signal in order to bridge the undefined region of the falling edge of SCL 2)The maximum hold time of the START condition has only to be met if the interface does not stretch the low period of SCL signal Cb = total capacitance of one bus line in pF 205/224 ST92163 - ELECTRICAL CHARACTERISTICS USB INTERFACE: DC CHARACTERISTICS (VDD = 4.0 - 5.5V, TA = 0°C + 70°C, unless otherwise specified) USB Interface: DC Electrical Characteristics Parameter Symbol Condition s Min. Differential Input Sensitivity VDI I(D+, D-) 0.2 Differential Common Mode Range VCM Includes V DI range Single Ended Receiver Threshold VSE Max. Unit Inputs Levels: V 0.8 2.5 V 0.8 2.0 V Output Levels Static Output Low VOL RL of 1.5K ohms to 3.6v 0.3 V Static Output High V OH RL of 15K ohm to GND 2.8 3.6 V V DD=5V 2.97 3.63 V -10 +10 µA USBVCC: voltage level Hi-Z State Data Line Leakage USB V ILO 0 V < Vin < 3.3 V (Regulator ON) RL is the load connected on the USB drivers. All voltages are measured from the local ground potential (USBGND). 206/224 ST92163 - ELECTRICAL CHARACTERISTICS ELECTRICAL CHARACTERISTICS (Cont’d) Figure 87. USB Interface: Data signal rise and fall time Differential Data Lines Crossover points VCRS USBGND tr tf USB INTERFACE: FULL SPEED CHARACTERISTICS (VDD = 4.0 - 5.5V, T A = 0°C + 70°C, unless otherwise specified) USB Interface: Full speed electrical characteristics Parameter Symbol Condition s Min Max Unit tr CL=50 pF1) 4 20 ns tf 1) Driver characteristics: Rise time Fall Time Rise/ Fall Time matching trfm Output signal Crossover Voltage VCRS 1) CL=50 pF tr/tf 4 20 ns 90 110 % 1.3 2.0 V Measured from 10% to 90% of the data signal 207/224 ST92163 - ELECTRICAL CHARACTERISTICS EXTERNAL INTERRUPT TIMING TABLE (VDD = 3.0 - 5.5V (1), T A = 0°C + 70°C, C Load = 50pF, fINTCLK = 24MHz, unless otherwise specified) Parameter Value (Note) N° Symbol 1 TwINTLR Low Level Pulse Width in Rising Edge Mode ≥ Tck+10 50 ns 2 TwINTHR High Level Pulse Width in Rising Edge Mode ≥ Tck+10 50 ns 3 TwINTHF High Level Pulse Width in Falling Edge Mode ≥ Tck+10 50 ns 4 TwINTLF Low Level Pulse Width in Falling Edge Mode ≥ Tck+10 50 ns (2) Formula Min Max Unit Note: The value in the left hand column shows the formula used to calculate the timing minimum or maximum from the oscillator clock period. The value in the right hand two columns show the timing minimum and maximum for an internal clock at 24MHz (INTCLK). (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) (2) Formula guaranteed by design. Legend: Tck = INTCLK period = OSCIN period when OSCIN is not divided by 2; 2 x OSCIN period when OSCIN is divided by 2; OSCIN period x PLL factor when the PLL is enabled. EXTERNAL INTERRUPT TIMING 208/224 ST92163 - ELECTRICAL CHARACTERISTICS WAKE-UP MANAGEMENT TIMING TABLE (VDD = 3.0 - 5.5V (1), TA = 0°C + 70°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Parameter Value (Note) N° Symbol 1 TwWKPLR Low Level Pulse Width in Rising Edge Mode ≥ Tck+10 50 ns 2 TwWKPHR High Level Pulse Width in Rising Edge Mode ≥ Tck+10 50 ns 3 TwWKPHF High Level Pulse Width in Falling Edge Mode ≥ Tck+10 50 ns 4 TwWKPLF Low Level Pulse Width in Falling Edge Mode ≥ Tck+10 50 ns (2) Formula Min Max Unit Note: The value in the left hand column shows the formula used to calculate the timing minimum or maximum from the oscillator clock period. The value in the right hand two columns show the timing minimum and maximum for an internal clock at 24MHz (INTCLK). The given data are related to Wake-up Management Unit used in External Interrupt mode. (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) (2) Formula guaranteed by design. Legend: Tck = INTCLK period = OSCIN period when OSCIN is not divided by 2; 2 x OSCIN period when OSCIN is divided by 2; OSCIN period x PLL factor when the PLL is enabled. WAKE-UP MANAGEMENT TIMING WKUPn n=0 – 15 209/224 ST92163 - ELECTRICAL CHARACTERISTICS RCCU CHARACTERISTICS (VDD = 3.0 - 5.5V (1), T A = 0°C + 70°C, C Load = 50pF, fINTCLK = 24MHz, unless otherwise specified) Symbol Parameter Comment Value (Note) Min Typ Max Unit VIHRS RESET Input High Level 0.7 x VDD VDD + 0.3 V VILRS RESET Input Low Level – 0.3 V V HYRS RESET Input Hysteresis 0.3 x VDD 1.5 ILKRS RESET Pin Input Leakage +1 µA 0.3 0V < VIN < VDD 0.9 –1 V (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) RCCU TIMING TABLE (VDD = 3.0 - 5.5V (1), TA = 0°C + 70°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Symbol TFRS TRSPH(2) TSTR Parameter Comment Value (Note) Min Max Unit 50 RESET Input Filtered Pulse ns µs 20 RESET Phase duration STOP Restart duration Typ DIV2 = 0 DIV2 = 1 20400 x Tosc µs 10200 x Tosc 20400 x Tosc µs (1) 3.0 -4.0V voltage range is only available on devices with suffix L or V, with different frequency limitations (L: 8 MHz, V:16 MHz) (2) Depending on the delay between rising edge of RESETN pin and the first rising edge of CLOCK1, the value can differ from the typical value for +/- 1 CLOCK1 cycle. Legend: Tosc = OSCIN clock cycles PLL CHARACTERISTICS (VDD = 3.0 - 5.5V (1), T A = 0°C + 70°C, C Load = 50pF, fINTCLK = 24MHz, unless otherwise specified) Parameter FXTL Crystal Reference Frequency 3 FVCO VCO Operating Frequency 9 TPLK Lock-in Time PLL Jitter Comment Value (Note) Symbol Min 0 Typ Max Unit 8 MHz 48 MHz 1000 x Tosc µs 850 (2) ps (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) (2) Measured at 24MHz (INTC LK). Guaranteed by Design Characterisation (not tested). Legend: Tosc = OSCIN clock cycles 210/224 ST92163 - ELECTRICAL CHARACTERISTICS OSCILLATOR CHARACTERISTICS (VDD = 3.0 - 5.5V (1), TA = 0°C + 70°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Symbol FOSC gm Parameter Crystal Frequency Oscillator Comment Fundamental mode crystal only VDD = 4.0 - 5.5 V VDD = 3.0 - 4.0 V Value (Note) Min Unit Typ Max 8 MHz 0.77 1.5 2.4 mA/V 0.5 0.73 3 1.47 mA/V VIHCK Clock Input High Level External Clock 0.8 x VDD VDD + 0.3 V VILCK Clock Input Low Level External Clock – 0.3 0.2 x VDD V ILKOS OSCIN/OSCOUT Pin Input Leakage 0V < VIN < VDD (HALT/STOP) –1 +1 µA TSTUP Oscillator Start-up Time VDD = 4.0 - 5.5 V 5 ms VDD = 3.0 - 4.0 V 20 ms (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) 211/224 ST92163 - ELECTRICAL CHARACTERISTICS EXTERNAL BUS TIMING TABLE (VDD = 3.0 - 5.5V (1), T A = 0°C + 70°C, C Load = 50pF, fINTCLK = 24MHz, unless otherwise specified) N° Symbol Value (Note) Parameter Formula Min Max Unit 1 TsA (AS) Address Set-up Time before AS ↑ Tck x Wa+TckH-9 12 ns 2 ThAS (A) Address Hold Time after AS ↑ TckL-4 17 ns 3 TdAS (DR) AS ↑ to Data Available (read) Tck x (Wd+1)+3 4 TwAS AS Low Pulse Width Tck x Wa+TckH-5 5 TdAz (DS) Address Float to DS ↓ 6 TwDS DS Low Pulse Width 45 ns 16 ns 0 0 ns Tck x Wd+TckH-5 16 ns 7 TdDSR (DR) DS ↓ to Data Valid Delay (read) Tck x Wd+TckH+4 8 ThDR (DS) Data to DS ↑ Hold Time (read) 7 7 ns 9 TdDS (A) DS ↑ to Address Active Delay TckL+11 32 ns 25 ns 10 TdDS (AS) DS ↑ to AS ↓ Delay TckL-4 17 ns 11 TsR/W (AS) RW Set-up Time before ASN ↑ Tck x Wa+TckH-17 4 ns 12 TdDSR (R/W) DS ↑ to RW and Address Not Valid Delay TckL-1 20 ns 13 TdDW (DSW) Write Data Valid to DS ↓ Delay -16 -16 ns 14 TsD(DSW) Write Data Set-up before DS ↑ Tck x Wd+TckH-16 5 ns 15 ThDS (DW) Data Hold Time after DS ↑ (write) TckL-3 18 ns 16 TdA (DR) Address Valid to Data Valid Delay (read) Tck x (Wa+Wd+1)+TckH-7 17 TdAs (DS) AS ↑ to DS ↓ Delay TckL-6 55 15 ns ns Note: The value in the left hand column shows the formula used to calculate the timing minimum or maximum from the oscillator clock period, prescaler value and number of wait cycles inserted. The values in the right hand two columns show the timing minimum and maximum for an external clock at 24MHz, prescaler value of zero and zero wait states. (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) Legend: Tck = INTCLK period = OSCIN period when OSCIN is not divided by 2; 2 x OSCIN period when OSCIN is divided by 2; OSCIN period x PLL factor when the PLL is enabled. TckH = INTCLK high pulse width (normally = Tck/2, except when INTCLK = OSCIN, in which case it is OSCIN high pulse width) TckL = INTCLK low pulse width (normally = Tck/2, except when INTCLK = OSCIN, in which case it is OSCIN low pulse width) P = clock prescaling value (=PRS; division factor = 1+P) Wa = wait cycles on AS; = max (P, programmed wait cycles in EMR2, requested wait cycles with WAIT) Wd = wait cycles on DS; = max (P, programmed wait cycles in WCR, requested wait cycles with WAIT) 212/224 ST92163 - ELECTRICAL CHARACTERISTICS EXTERNAL BUS TIMING CPUCLK RW PORT1 PORT0 (READ) AS DS PORT0 (WRITE) 213/224 ST92163 - ELECTRICAL CHARACTERISTICS WATCHDOG TIMING TABLE (VDD = 3.0 - 5.5V (1) , TA = 0°C + 70°C, CLoad = 50pF, fINTCLK = 24MHz, Push-pull output configuration, unless otherwise specified) N° Symbol Parameter Value (Note) Formula 4 x (Psc+1) x (Cnt+1) x Tck 1 TwWDOL WDOUT Low Pulse Width (Psc+1) x (Cnt+1) x TWDIN with TWDIN ≥ 8 x Tck 4 x (Psc+1) x (Cnt+1) x Tck 2 TwWDOH WDOUT High Pulse Width (Psc+1) x (Cnt+1) x TWDIN 3 TwWDIL WDIN High Pulse Width with TWDIN ≥ 8 x Tck ≥ 4 x Tck 4 TwWDIH WDIN Low Pulse Width ≥ 4 x Tck Min Max 167 Unit ns 2.8 333 s ns 167 ns 2.8 s 333 ns 167 ns 167 ns Note: The value in the left hand column shows the formula used to calculate the timing minimum or maximum from the oscillator clock period, watchdog prescaler and counter programmed values. The value in the right hand two columns show the timing minimum and maximum for an internal clock (INTCLK) at 24MHz, with minimum and maximum prescaler value and minimum and maximum counter value. (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) Legend: Tck = INTCLK period = OSCIN period when OSCIN is not divided by 2; 2 x OSCIN period when OSCIN is divided by 2; OSCIN period x PLL factor when the PLL is enabled. Psc = Watchdog Prescaler Register content (WDTPR): from 0 to 255 Cnt = Watchdog Couter Registers content (WDTRH,WD TRL): from 0 to 65535 TWDIN = Watchdog Input signal period (WDIN) WATCHDOG TIMING 214/224 ST92163 - ELECTRICAL CHARACTERISTICS MULTIFUNCTION TIMER EXTERNAL TIMING TABLE (VDD = 3.0 - 5.5V (1), T A = 0°C + 70°C, C Load = 50pF, fINTCLK = 24MHz, unless otherwise specified) N° Symbol 1 Tw CTW 2 Parameter Value Unit Note - ns (2) n x 42 - ns (2) Formula Min Max External clock/trigger pulse width n x Tck n x 42 TwCTD External clock/trigger pulse distance n x Tck 3 TwAED Distance between two active edges 3 x Tck 125 - ns 4 TwGW Gate pulse width 6 x Tck 250 - ns 5 TwLBA Distance between TINB pulse edge and the following TINA pulse edge Tck 42 - ns (3) 6 TwLAB Distance between TINA pulse edge and the following TINB pulse edge 0 - ns (3) 7 Tw AD Distance between two TxINA pulses 0 - ns (3) 8 TwOWD Minimum output pulse width/distance 125 - ns 3 x Tck Note: The value in the left hand column shows the formula used to calculate the timing minimum or maximum from the oscillator clock period, standard timer prescaler and counter programmed values. The value in the right hand two columns show the timing minimum and maximum for an internal clock (INTCLK) at 24MHz. (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) (2) n = 1 if the input is rising OR falling edge sensitive n = 3 if the input is rising AND falling edge sensitive (3) In Autodiscrimination mode Legend: Tck = INTCLK period = OSCIN period when OSCIN is not divided by 2; 2 x OSCIN period when OSCIN is divided by 2; OSCIN period x PLL factor when the PLL is enabled. MULTIFUNCTION TIMER EXTERNAL TIMING 215/224 ST92163 - ELECTRICAL CHARACTERISTICS SCI TIMING TABLE (VDD = 3.0 - 5.5V (1), T A = 0°C + 70°C, C Load = 50pF, fINTCLK = 24MHz, unless otherwise specified) N° Symbol Parameter FRxCKIN Frequency of RxCKIN TwRxCKIN RxCKIN shortest pulse FTxCKIN Frequency of TxCKIN TwTxCKIN TxCKIN shortest pulse 1 TsDS 2 3 Conditio n Value Min Max Unit 1x mode fINTCLK / 8 MHz 16x mode fINTCLK / 4 MHz 1x mode 4 x Tck s 16x mode 2 x Tck s 1x mode fINTCLK / 8 MHz 16x mode fINTCLK / 4 MHz 1x mode 4 x Tck s 16x mode 2 x Tck s DS (Data Stable) before rising edge of RxCKIN 1x mode reception with RxCKIN Tck / 2 ns TdD1 TxCKIN to Data out delay Time 1x mode transmission with external clock C Load < 50pF TdD2 CLKOUT to Data out delay Time 1x mode transmission with CLKOUT 2.5 x Tck TBD (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) Legend: Tck = INTCLK period = OSCIN period when OSCIN is not divided by 2; 2 x OSCIN period when OSCIN is divided by 2; OSCIN period x PLL factor when the PLL is enabled. SCI TIMING 216/224 ns ns ST92163 - ELECTRICAL CHARACTERISTICS A/D CONVERTER, EXTERNAL TRIGGER TIMING TABLE (VDD= 3.0 - 5.5V (1); T A= 0 to 70°C; unless otherwise specified) Symbol Parameter Tlow Pulse Width Thigh Pulse Distance OSCIN divide by 2;min/max OSCIN divide by 1; min/max Value min Unit max 1.5 INTCLK ns ns 78+1 INTCLK µs Text Period/fast Mode Tstr Start Conversion Delay Tlow Pulse Width ns Thigh Pulse Distance ns Text Period/fast Mode µs Tstr Start Conversion Delay ns 0.5 1.5 INTCLK Core Clock issued by Timing Controller (1) 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitatio ns (L: 8 MHz, V: 16 MHz) A/D CONVERTER. ANALOG PARAMETERS TABLE (VDD= 3.0 - 5.5V (1); T A= 0 to 70°C; unless otherwise specified) Parameter Value typ (*) Analog Input Range Conversion Time Fast/Slow Sample Time Fast/Slow Power-up Time Resolution Unit min max (**) VSS VDD V 78/138 INTCLK 51.5/87.5 INTCLK Note (2) µs 60 8 bits Differential Non Linearity 0.5 1.5 LSBs (4) Integral Non Linearity 0.5 1.5 LSBs (4) Offset Error 0.5 1 Gain Error 0.5 1.5 1 2 LSBs (4) Input Resistance 1.5 Kohm (3) Hold Capacitance 1.92 pF Absolute Accuracy Notes: (*) (**) (1) (2) (3) (4) The values are expected at 25 Celsius degrees with VDD= 5V ’LSBs’ , as used here, as a value of VDD/256 3.0 - 4.0V voltage range is only available on devices with suffix L or V, with different frequency limitations (L: 8 MHz, V: 16 MHz) including Sample time it must be considered as the on-chip series resistance before the sampling capacitor DNL ERROR= max {[V(i) -V(i-1)] / LSB-1} INL ERROR= max {[V(i) -V(0)] / LSB-i} ABSOLUTE ACCURACY= overall max conversion error 217/224 ST92163 - GENERAL INFORMATION 10 GENERAL INFORMATION 10.1 EPROM/OTP PROGRAMMING The 20 Kbytes of EPROM/OTP of the ST92E163/ ST92T163 may be programmed using the EPROM programming boards available from STMicroelectronics. EPROM Erasing The EPROM of the windowed package of the ST92E163 can be erased by exposure to Ultra-Violet light. The erasure characteristic of the ST92E163 is such that erasure begins when the memory is exposed to light with wave lengths shorter than approximately 4000Å. It should be noted that sunlight and some types of fluorescent lamps have wavelengths in the range 3000-4000Å. It is recom- 218/224 mended to cover the window of the ST92E163 packages by an opaque label to prevent unintentional erasure problems when testing the application in such an environment. The recommended erasure procedure of the EPROM is the exposure to short wave ultraviolet light which have a wave-length 2537Å. The integrated dose (i.e. U.V. intensity x exposure time) for erasure should be a minimum of 15W-sec/cm2. The erasure time with this dosage is approximately 30 minutes using an ultraviolet lamp with a 12000 mW/cm2 power rating. The device should be placed within 2.5 cm (1 inch) of the lamp tubes during erasure. ST92163 - GENERAL INFORMATION 10.2 PACKAGE DESCRIPTION Figure 88. 56-Pin Shrink Plastic Dual In Line Package, 600-mil Width Dim. mm Min Typ A Min Typ 6.35 A1 0.38 A2 3.18 b 0.015 4.95 0.125 0.89 0.035 0.20 0.38 0.008 D 50.29 53.21 1.980 E1 15.01 12.32 0.195 0.016 C E 0.015 2.095 0.591 14.73 0.485 0.580 e 1.78 0.070 eA 15.24 0.600 eB L Max 0.250 0.41 b2 PDIP56S inches Max 2.92 17.78 0.700 5.08 0.115 0.200 Number of Pins N 56 Figure 89. 64-Pin Thin Quad Flat Package 0.10mm .004 seating plane Dim mm Min Typ A Min Typ Max 1.60 0.063 0.15 0.002 0.006 A1 0.05 A2 1.35 1.40 1.45 0.053 0.055 0.057 B 0.30 0.37 0.45 0.012 0.015 0.018 C 0.09 0.20 0.004 0.008 D 16.00 0.630 D1 14.00 0.551 D3 12.00 0.472 E 16.00 0.630 E1 14.00 0.551 E3 12.00 0.472 e 0.80 0.031 K L L1 inches Max 0° 3.5° 7° 0.45 0.60 0.75 0.018 0.024 0.030 L1 1.00 L 0.039 Number of Pins N 64 ND 16 NE 16 K 219/224 ST92163 - GENERAL INFORMATION PACKAGE DESCRIPTION (Cont’d) Figure 90. 56-Pin Shrink Ceramic Dual In-Line Package, 600-mil Width Dim. mm Min Typ A Min Typ 4.17 Max 0.164 A1 0.76 B 0.38 0.46 0.56 0.015 0.018 0.022 B1 0.76 0.89 1.02 0.030 0.035 0.040 C 0.23 0.25 0.38 0.009 0.010 0.015 D 50.04 50.80 51.56 1.970 2.000 2.030 D1 E1 0.030 48.01 1.890 14.48 14.99 15.49 0.570 0.590 0.610 e 1.78 0.070 G 14.12 14.38 14.63 0.556 0.566 0.576 G1 18.69 18.95 19.20 0.736 0.746 0.756 G2 CDIP56SW inches Max 1.14 0.045 G3 11.05 11.30 11.56 0.435 0.445 0.455 G4 15.11 15.37 15.62 0.595 0.605 0.615 L 2.92 S 5.08 0.115 1.40 0.200 0.055 Number of Pins N 56 Figure 91. 64-Pin Ceramic Quad Flat Package Dim mm Min Typ A A1 B inches Max Min Typ 3.27 0.50 0.020 0.30 0.35 0.45 0.012 0.014 0.018 C 0.13 0.15 0.23 0.005 0.006 0.009 D 16.65 17.20 17.75 0.656 0.677 0.699 D1 13.57 13.97 14.37 0.534 0.550 0.566 D3 12.00 0.472 e 0.80 0.031 G 12.70 0.500 G2 0.96 0.038 L 0.35 0.80 0.014 0.031 0 8.31 0.327 Number of Pins CQFP064W 220/224 Max 0.129 N 64 ST92163 - GENERAL INFORMATION 10.3 ORDERING INFORMATION The following section deals with the procedure for transfer of customer codes to STMicroelectronics. generated by the development tool. All unused bytes must be set to FFh. The customer code should be communicated to STMicroelectronics with the correctly completed OPTION LIST appended. The STMicroelectronics Sales Organization will be pleased to provide detailed information on contractual points. 10.4 Transfer of Customer Code Customer code is made up of the ROM contents. The ROM contents are to be sent on diskette, or by electronic means, with the hexadecimal file Figure 92. Sales Type Coding Rules Family (92163, 92E163, 92T163) Number of pins ROM size Package Temperature Range Device Characteristics ROM Code (three letters) ST 92163 R 4 B 1 L / xxx L = 8-MHz Low Voltage Version 1= Standard (0 to +70°C) V = 16-MHz Low Voltage Version 0= 25°C E = Without LVD function (for Normal Voltage Versions) No letter = With LVD function (for Normal Voltage Versions) B = Plastic DIP D = Ceramic DIP G = Ceramic QFP 4 = 20K N = 56 pins R = 64 pins Table 37. Development Tools Development Tool Real time emulator1 Sales Type Remarks ST92163-EMU2 ST92E163-EPB/EU 220V Power Supply ST92E163-EPB/US ST92E16x-GP/DIP56 110V Power Supply SDIP56 package ST92E16x-GP/QFP ST9P-SWC/V4 TQFP64 package for PC EPROM Programming Board Gang Programmer C Hiware Compiler and Debugger Note 1: The emulator does not support Low Voltage Modes 221/224 ST92163 - GENERAL INFORMATION Table 38. Ordering Information Program Memory Sales Type 1) (bytes) ST92163N4B1/xxx 2) ST92163R4T1/xxx 2) ST92E163N4D0 Mode Operating Voltage Package 20K ROM 20K EPROM ST92E163R4G0 ST92T163N4B1 20K OTP ST92T163R4T1 ST92163N4B1E/xxx 2) ST92163R4T1E/xxx 2) ST92E163N4D1E ST92E163R4G1E ST92T163N4B1E ST92T163R4T1E ST92163R4T1L/xxx 2) Normal 20K ROM 20K EPROM 20K OTP 20K ROM ST92E163R4G1L 20K EPROM ST92E163N4D1L ST92T163R4T1L 20K OTP 20K ROM PSDIP56 TQFP64 4.0-5.5 V CSDIP56 4.0-5.5 V CQFP64 4.0-5.5 V PSDIP56 4.0-5.5 V TQFP64 4.0-5.5 V PSDIP56 4.0-5.5 V TQFP64 4.0-5.5 V CSDIP56 4.0-5.5 V CQFP64 4.0-5.5 V PSDIP56 4.0-5.5 V TQFP64 8-MHz LVM 3) 3.0-4.0 V (@8Mhz) NM 3) 4.0-5.5 V 8-MHz LVM 3) Normal NM 3) and 3) 8-MHz Low Voltage 8-MHz LVM 3) NM 3.0-4.0 V (@8Mhz) 8-MHz LVM 3) 3.0-4.0 V (@8Mhz) NM 3) 4.0-5.5 V 16-MHz LVM ST92163R4T1V/xxx 2) 4.0-5.5 V 4.0-5.5 V 3) 4.0-5.5 V 3.0-4.0 V (@8Mhz) 4.0-5.5 V 3.0V-4.0 (@16Mhz) 8-MHz LVM 3) 3.0V-4.0 (@8Mhz) NM 3) 4.0-5.5V TQFP64 CQFP64 CSDIP56 TQFP64 16-MHz LVM 3) 3.0V-4.0 (@16Mhz) ST92E163R4T1V 2) 20K EPROM ST92E163N4D1V 2) 8-MHz LVM 3) Normal, 3) 8-MHz Low Voltage NM and 16-MHz LVM 3) 16-MHz Low voltage 8-MHz LVM 3) NM 3) 3.0V-4.0 (@8Mhz) CQFP64 4.0-5.5V 3.0V-4.0 (@16Mhz) 3.0V-4.0 (@8Mhz) CSDIP56 4.0-5.5V 16-MHz LVM 3) 3.0V-4.0 (@16Mhz) ST92T163R4T1V 2) 20K OTP 8-MHz LVM 3) NM 3) 3.0V-4.0 (@8Mhz) 4.0-5.5V Note 1: xxx stands for the ROM code name assigned by STMicroelectronics Note 2: Contact sales office for availability Note 3: LVM = Low Voltage Mode and NM = Normal Mode 222/224 TQFP64 ST92163 - GENERAL INFORMATION STMicroelectronics OPTION LIST ST92163 MICROCONTROLLER FAMILY (ROM DEVICE) Customer: Address: . . .. .. .. . .. .. . ... .. .. . . .... . . . .. .. .. . .. .. . ... .. .. . . .... . . .. .. .. . .. ... . .. .... . ... .. .. Contact: . . .. .. .. . .. .. . ... .. .. . . .... . Phone No: . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference/ROM Code* : . . . . . . . . . . . . . . . . . . *The ROM code name is assigned by STMicroelectronics. STMicroelectronics reference: Device (PSDIP56): [ ] ST92163N4B1/xxx* [ ] ST92163N4B1E/xxx* Device (TQFP64): [ ] ST92163R4T1/xxx* [ ] ST92163R4T1E/xxx* [ ] ST92163R4T1L/xxx* [ ] ST92163R4T1V/xxx* *xxx = ROM code name Software Development: [ ] STMicroelectronics [ ] Customer [ ] External laboratory Special Marking: [ ] No [ ] Yes ”_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _” For marking, one line is possible with maximum 13 characters. Authorized characters are letters, digits, ’.’, ’-’, ’/’ and spaces only. We have checked the ROM code verification file returned to us by STMicroelectronics. It conforms exactly with the ROM code file orginally supplied. We therefore authorize STMicroelectronics to proceed with device manufacture. Signature . . .. .. .. . .. .. . ... .. .. . . .... . Date . . .. .. .. . .. .. . ... .. .. . . .... . 223/224 ST92163 - GENERAL INFORMATION Notes: Information furnished is believed to be accurate and reliable. However, STMicroelectronics assumes no responsibility for the consequences of use of such information nor for any infringement of patents or other rights of third parties which may result from its use. No license is granted by implication or otherwise under any patent or patent rights of STMicroelectronics. Specifications mentioned in this publication are subject to change without notice. This publication supersedes and replaces all information previously supplied. STMicroelectronics products are not authorized for use as critical components in life support devices or systems without the express written approval of STMicroelectronics. The ST logo is a registered trademark of STMicroelectronics 2000 STMicroelectronics - All Rights Reserved. Purchase of I2 C Components by STMicroelectronics conveys a license under the Philips I2C Patent. Rights to use these components in an I2C system is granted provided that the system conforms to the I2C Standard Specification as defined by Philips. STMicroelectronics Group of Companies Australia - Brazil - China - Finland - France - Germany - Hong Kong - India - Italy - Japan - Malaysia - Malta - Morocco - Singapore - Spain Sweden - Switzerland - United Kingdom - U.S.A. http:// www.st.com 224/224