ST92F120 8/16-BIT FLASH MCU FAMILY WITH RAM, EEPROM AND J1850 BLPD PRELIMINARY DATA ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ TQFP64 PQFP100 ■ ■ ■ ■ Full I2C multiple Master/Slave Interface supporting ACCESS BUS Rich Instruction Set with 14 Addressing Modes Division-by-zero trap generation Versatile Development Tools, including Assembler, Linker, C-Compiler, Archiver, Source Level Debugger, Hardware Emulators and Real Time Operating System DEVICE SUMMARY Device ST92F120R6T Package - TQFP 64 - 48 ST92F120JR6T 1 ST92F120V6Q - ST92F120JV6Q 1 ST92F120R9T - ST92F120JR9T 1 ST92F120V9Q - ST92F120JV9Q 1 ST92F120R1T - ST92F120JR1T 1 ST92F120V1Q - ST92F120JV1Q 1 I/Os SCI ■ EFT ■ Register oriented 8/16 bit CORE with RUN, WFI, SLOW, HALT and STOP modes 0 - 24 MHz Operation (internal Clock), 4.5 - 5.5 Volt voltage range PLL Clock Generator (3-5 MHz crystal) -40oC to 105oC or -40oC to 85oC temperature range Minimum instruction time: 83 ns (24 MHz internal clock) Internal Memory: Single Voltage FLASH up to 128 Kbytes, RAM 1.5 to 4 Kbytes, EEPROM 512 to 1K bytes 224 general purpose registers (register file) available as RAM, accumulators or index pointers TQFP64 or PQFP100 package DMA controller for reduced processor overhead 48 (77 on PQFP100 version) I/O pins 4 external fast interrupts + 1 NMI Up to 16 pins programmable as wake-up or additional external interrupt with multi-level interrupt handler 16-bit Timer with 8 bit Prescaler, able to be used as a Watchdog Timer with a large range of service time (HW/SW enabling through dedicated pin) 16-bit Standard Timer that can be used to generate a time base independent of PLL Clock Generator Two 16-bit independent Extended Function Timers (EFTs) with Prescaler, 2 Input Captures and two Output Compares (PQFP100 only) Two 16-bit Multifunction Timers, with Prescaler, 2 Input Captures and two Output Compares 8-bit Analog to Digital Converter allowing up to 8 input channels on TQFP64 or 16 input channels on PQFP100 One or two Serial Communications Interfaces with asynchronous and synchronous capabilities. Software Management and synchronous mode supported Serial Peripheral Interface (SPI) with Selectable Master/Slave mode J1850 Byte Level Protocol Decoder (JBLPD) (on some versions only) J1850 ■ 1 PQFP 100 2 77 TQFP 64 - 48 1 PQFP 100 2 77 2 TQFP 64 - 48 1 PQFP 100 2 77 2 Flash RAM E 36K 1.5K 512 60K 2K 512 128K 4K 1K Rev. 2.1 January 2000 1/320 This is preliminary information on a new product now in development or undergoing evaluation. Details are subject to change without notice. 9 Table of Contents 1 GENERAL DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.1.1 ST9+ Core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.1.2 External Memory Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.1.3 On-chip Peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.2 PIN DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.2.1 Electromagnetic Compatibility (EMC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.2.2 I/O Port Alternate Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.2.3 Termination of Unused Pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.2.4 Avoidance of Pin Damage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.3 I/O PORTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1.4 OPERATING MODES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2 DEVICE ARCHITECTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.1 CORE ARCHITECTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.2 MEMORY SPACES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.2.1 Register File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.2.2 Register Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.3 SYSTEM REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 2.3.1 Central Interrupt Control Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 2.3.2 Flag Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3.3 Register Pointing Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 2.3.4 Paged Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.3.5 Mode Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.3.6 Stack Pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 2.4 MEMORY ORGANIZATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.5 MEMORY MANAGEMENT UNIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 2.6 ADDRESS SPACE EXTENSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 2.6.1 Addressing 16-Kbyte Pages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 2.6.2 Addressing 64-Kbyte Segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.7 MMU REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.7.1 DPR[3:0]: Data Page Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.7.2 CSR: Code Segment Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2.7.3 ISR: Interrupt Segment Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2.7.4 DMASR: DMA Segment Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2.8 MMU USAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2.8.1 Normal Program Execution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2.8.2 Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2.8.3 DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3 SINGLE VOLTAGE FLASH & EEPROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.2 FUNCTIONAL DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 3.2.1 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 3.2.2 Software or Hardware EEPROM Emulation (Device dependent option) . . . . . . . . . 45 3.2.3 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 3.3 REGISTER DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3.3.1 Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .320 . . . 47 3.3.2 Status Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 3.4 WRITE OPERATION EXAMPLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 2/320 9 Table of Contents 3.5 EEPROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.1 Hardware EEPROM Emulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.2 EEPROM Update Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 PROTECTION STRATEGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.1 Non Volatile Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.2 Temporary Unprotection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7 FLASH IN-SYSTEM PROGRAMMING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7.1 First Programming of a virgin Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 REGISTER AND MEMORY MAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 MEMORY CONFIGURATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 ST92F120 REGISTER MAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 INTERRUPTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 INTERRUPT VECTORING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 Divide by Zero trap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.2 Segment Paging During Interrupt Routines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 INTERRUPT PRIORITY LEVELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 PRIORITY LEVEL ARBITRATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1 Priority level 7 (Lowest) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.2 Maximum depth of nesting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.3 Simultaneous Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.4 Dynamic Priority Level Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 ARBITRATION MODES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.1 Concurrent Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.2 Nested Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 EXTERNAL INTERRUPTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.1 Standard External Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7 TOP LEVEL INTERRUPT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.8 ON-CHIP PERIPHERAL INTERRUPTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.9 INTERRUPT RESPONSE TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.10 INTERRUPT REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.11 WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (WUIMU) . . . . . . . . . . . . . . . . . . 5.11.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.11.2Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.11.3Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.11.4Programming Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.11.5Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 ON-CHIP DIRECT MEMORY ACCESS (DMA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 DMA PRIORITY LEVELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 DMA TRANSACTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 DMA CYCLE TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 SWAP MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6 DMA REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 53 54 55 55 57 57 57 59 59 59 62 72 72 72 72 73 73 73 73 73 73 74 74 74 77 79 79 81 81 82 83 86 86 86 87 90 91 94 94 94 95 97 97 98 3/320 9 Table of Contents 7 RESET AND CLOCK CONTROL UNIT (RCCU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 7.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 7.2 CLOCK CONTROL UNIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 7.2.1 Clock Control Unit Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 7.3 CLOCK MANAGEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 7.3.1 PLL Clock Multiplier Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 7.3.2 CPU Clock Prescaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 7.3.3 Peripheral Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 7.3.4 Low Power Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 7.3.5 Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 7.4 CLOCK CONTROL REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 7.5 OSCILLATOR CHARACTERISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 7.6 RESET/STOP MANAGER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 7.6.1 Reset Pin Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 7.7 STOP MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 8 EXTERNAL MEMORY INTERFACE (EXTMI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 8.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 8.2 EXTERNAL MEMORY SIGNALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 8.2.1 AS: Address Strobe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 8.2.2 DS: Data Strobe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 8.2.3 DS2: Data Strobe 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 8.2.4 RW: Read/Write . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 8.2.5 BREQ, BACK: Bus Request, Bus Acknowledge . . . . . . . . . . . . . . . . . . . . . . . . . . 118 8.2.6 PORT 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 8.2.7 PORT 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 8.2.8 WAIT: External Memory Wait . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 8.3 REGISTER DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 9 I/O PORTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 9.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 9.2 SPECIFIC PORT CONFIGURATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 9.3 PORT CONTROL REGISTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 9.4 INPUT/OUTPUT BIT CONFIGURATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 9.5 ALTERNATE FUNCTION ARCHITECTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 9.5.1 Pin Declared as I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 9.5.2 Pin Declared as an Alternate Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 9.5.3 Pin Declared as an Alternate Function Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 9.6 I/O STATUS AFTER WFI, HALT AND RESET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 10 ON-CHIP PERIPHERALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 10.1 TIMER/WATCHDOG (WDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 10.1.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 10.1.2Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 10.1.3Watchdog Timer Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 10.1.4WDT Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 10.1.5Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 10.2 STANDARD TIMER (STIM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 10.2.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .320 . . 136 10.2.2Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 4/320 1 Table of Contents 10.2.3Interrupt Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2.4Register Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2.5Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3 EXTENDED FUNCTION TIMER (EFT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.2Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.3Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.4Interrupt Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.5Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4 MULTIFUNCTION TIMER (MFT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4.2Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4.3Input Pin Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4.4Output Pin Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4.5Interrupt and DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4.6Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5 SERIAL COMMUNICATIONS INTERFACE (SCI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.2Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.3SCI Operating Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.4Serial Frame Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.5Clocks And Serial Transmission Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.6SCI Initialization Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.7Input Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.8Output Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.9Interrupts and DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5.10Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6 SERIAL PERIPHERAL INTERFACE (SPI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.2Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.3General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.4Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.5Interrupt Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.6.6Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7 I2C BUS INTERFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7.2Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7.3Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7.4I2C State Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7.5Interrupt Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7.6DMA Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7.7Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.8 J1850 BYTE LEVEL PROTOCOL DECODER (JBLPD) . . . . . . . . . . . . . . . . . . . . . . . . . 10.8.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.8.2Main Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.8.3Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.8.4Peripheral Functional Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.8.5Interrupt Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.8.6DMA Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 138 139 140 140 140 140 150 152 160 160 162 165 169 171 173 184 184 185 186 189 192 192 194 194 195 198 209 209 209 209 211 218 219 221 221 221 222 224 229 230 232 243 243 243 245 256 257 259 5/320 1 Table of Contents 10.8.7Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.9 EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) . . . . . . . . . . . . . . . . . . . 10.9.1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.9.2Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.9.3Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.9.4Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 ELECTRICAL CHARACTERISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 GENERAL INFORMATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.1 PACKAGE MECHANICAL DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320 6/320 1 263 284 284 285 287 288 292 319 319 ST92F120 - GENERAL DESCRIPTION 1 GENERAL DESCRIPTION 1.1 INTRODUCTION The ST92F120 microcontroller is developed and manufactured by STMicroelectronics using a proprietary n-well HCMOS process. Its performance derives from the use of a flexible 256-register programming model for ultra-fast context switching and real-time event response. The intelligent onchip 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 new-generation ST9 MCU devices now also support low power consumption and low voltage operation for power-efficient and low-cost embedded systems. 1.1.1 ST9+ Core The advanced Core consists of the Central Processing Unit (CPU), the Register File, the Interrupt and DMA controller, and the Memory Management Unit. The MMU allows a single linear address space of up to 4 Mbytes. 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 makes the ST9 family devices highly efficient for accessing on and offchip 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. The powerful I/O capabilities demanded by microcontroller applications are fulfilled by the ST92F120 with 48 (TQFP64) or 77 (PQFP100) I/O lines dedicated to digital Input/Output. These lines are grouped into up to ten 8-bit I/O Ports and can be configured on a bit basis under software control to provide timing, status signals, an address/data bus for interfacing to the external memory, timer inputs and outputs, analog inputs, external interrupts and serial or parallel I/O. Two memory spaces are available to support this wide range of configurations: a combined Program/Data Memory Space and the internal Register File, which includes the control and status registers of the onchip peripherals. 1.1.2 External Memory Interface PQFP100 devices have a 16-bit external address bus allowing them to address up to 64K bytes of external memory. TQFP64 devices have an 11-bit external address bus for addressing up to 2K bytes. 1.1.3 On-chip Peripherals Two 16-bit MultiFunction Timers, each with an 8 bit Prescaler and 12 operating modes allow simple use for complex waveform generation and measurement, PWM functions and many other system timing functions by the usage of the two associated DMA channels for each timer. On PQFP100 devices, two Extended Function Timers provide further timing and signal generation capabilities. A Standard Timer can be used to generate a stable time base independent from the PLL. An I2C interface provides fast I2C and Access Bus support. The SPI is a synchronous serial interface for Master and Slave device communication. It supports single master and multimaster systems. A J1850 Byte Level Protocol Decoder is available (on some devices only) for communicating with a J1850 network. In addition, there is an 16 channel Analog to Digital Converters with integral sample and hold, fast conversion time and 8-bit resolution. In the TQFP64 version only 8 input channels are available. Completing the device are two or one full duplex Serial Communications Interfaces with an integral generator, asynchronous and synchronous capability (fully programmable format) and associated address/wake-up option, plus two DMA channels. Finally, a programmable PLL Clock Generator allows the usage of standard 3 to 5 MHz crystals to obtain a large range of internal frequencies up to 24MHz. Low power Run (SLOW), Wait For Interrupt, low power Wait For Interrupt, STOP and HALT modes are also available. 7/320 9 ST92F120 - GENERAL DESCRIPTION Figure 1. ST92F120JR: Architectural Block Diagram (TQFP64 version) Ext. MEM. FLASH 36/60/128 Kbytes ADDRES S DATA EEPROM 512 /1K bytes INT[6:0] WKUP[15:0] 256 bytes Register File MEMORY BUS AS DS WAIT NMI RW DS2 RAM 1.5/2/4 Kbytes Ext. MEM. ADDRESS STOUT J1850 JBLPD (optional) VPWI VPWO I2C BUS SDAI SDAO SCLI SCLO RCCU ST. TIMER TINPA0 TOUTA0 TINPB0 TOUTB0 MF TIMER 0 TINPA1 TOUTA1 TINPB1 TOUTB1 MF TIMER 1 REGISTER BUS OSCIN OSCOUT RESET CLOCK2/8 INTCLK I/Os 8/16 bits CPU ST9 CORE A[10:8] P0[7:0] P1[2:0] P2[7:0] P3[7:4] P4[7:4] P5[7:0] P6[5:2,0] P7[7:0] Fully Prog. Interrupt Management A[7:0] D[7:0] WATCHDOG SPI WDOUT HW0SW1 MISO MOSI SCK SS A/D CONV. 0 A0IN[7:0] EXTRG SCI 0 TXCLK0 RXCLK0 SIN0 DCD0 SOUT0 CLKOUT0 RTS0 All alternate functions (Italic characters ) are mapped on Port2, Port3, Port4, Port5, Port6,and Port7 8/320 9 ST92F120 - GENERAL DESCRIPTION Figure 2. ST92F120JV: Architectural Block Diagram (PQFP100 version) Ext. MEM. FLASH 36/60/128 Kbytes ADDRESS DATA EEPROM 512 /1K bytes INT[6:0] WKUP[15:0] 256 bytes Register File MEMORY BUS AS DS RW WAIT NMI RW DS2 RAM 1.5/2/4 Kbytes Ext. MEM. ADDRESS Fully Prog. I/Os 8/16 bits CPU J1850 JBLPD (optional) Interrupt Management ST9 CORE RCCU WATCHDOG STOUT ST. TIMER ICAPA0 OCMPA0 ICAPB0 OCMPB0 EXTCLK0 EF TIMER 0 ICAPA1 OCMPA1 ICAPB1 OCMPB1 EXTCLK1 REGISTER BUS OSCIN OSCOUT RESET CLOCK2/8 CLOCK2 INTCLK I2C BUS SPI A/D CONV. 0 A[7:0] D[7:0] A[15:8] P0[7:0] P1[7:0] P2[7:0] P3[7:1] P4[7:0] P5[7:0] P6[5:0] P7[7:0] P8[7:0] P9[7:0] VPWI VPWO SDAI SDAO SCLI SCLO WDOUT HW0SW1 MISO MOSI SCK SS A0IN[7:0] EXTRG A/D CONV. 1 SCI 0 TXCLK0 RXCLK0 SIN0 DCD0 SOUT0 CLKOUT0 RTS0 SCI 1 * TXCLK1 RXCLK1 SIN1 DCD1 SOUT1 CLKOUT1 RTS1 EF TIMER 1 TINPA0 TOUTA0 TINPB0 TOUTB0 MF TIMER 0 TINPA1 TOUTA1 TINPB1 TOUTB1 MF TIMER 1 A1IN[7:0] All alternate functions (Italic characters ) are mapped on Port2, Port3, Port4, Port5, Port6, Port7, Port8 and Port9 * Available on some versions only 9/320 9 ST92F120 - GENERAL DESCRIPTION 1.2 PIN DESCRIPTION AS. Address Strobe (output, active low, 3-state). Address Strobe is pulsed low once at the beginning of each memory cycle. The rising edge of AS indicates that address, Read/Write (RW), and Data signals are valid for memory transfers. DS. Data Strobe (output, active low, 3-state). Data Strobe provides the timing for data movement to or from Port 0 for each memory transfer. During a write cycle, data out is valid at the leading edge of DS. During a read cycle, Data In must be valid prior to the trailing edge of DS. When the ST9 accesses on-chip memory, DS is held high during the whole memory cycle. RESET. Reset (input, active low). The ST9 is initialised by the Reset signal. With the deactivation of RESET, program execution begins from the Program memory location pointed to by the vector contained in program memory locations 00h and 01h. RW. Read/Write (output, 3-state). Read/Write determines the direction of data transfer for external memory transactions. RW is low when writing to external memory, and high for all other transactions. OSCIN, OSCOUT. Oscillator (input and output). These pins connect a parallel-resonant crystal, or an external source to the on-chip clock oscillator and buffer. OSCIN is the input of the oscillator inverter and internal clock generator; OSCOUT is the output of the oscillator inverter. HW0SW1. When connected to VDD through a 1K pull-up resistor, the software watchdog option is selected. When connected to VSS through a 1K pull-down resistor, the hardware watchdog option is selected. VPWO. This pin is the output line of the J1850 peripheral (JBLPD). It is available only on some devices. On devices without JBLPD peripheral, this pin must not be connected. P0[7:0], P1[2:0] or P1[7:0](Input/Output, TTL or CMOS compatible). 11 lines (TQFP64 devices) or 16 lines (PQFP100 devices) providing the external memory interface for addressing 2K or 64 K bytes of external memory. P0[7:0], P1[2:0], P2[7:0], P3[7:4], P4.[7:4], P5[7:0], P6[5:2,0], P7[7:0] I/O Port Lines (Input/ Output, TTL or CMOS compatible). I/O lines grouped into I/O ports of 8 bits, bit programmable under software control as general purpose I/O or as alternate functions. 10/320 9 P1[7:3], P3[3:1], P4[3:0], P6.1, P8[7:0], P9[7:0] Additional I/O Port Lines available on PQFP100 versions only. AVDD. Analog VDD of the Analog to Digital Converter (common for A/D 0 and A/D 1). AVSS. Analog VSS of the Analog to Digital Converter (common for A/D 0 and A/D 1). VDD. Main Power Supply Voltage. Four pins are available on PQFP100 versions, two on TQFP64 versions. The pins are internally connected. VSS. Digital Circuit Ground. Four pins are available on PQFP100 versions, two on TQFP64 versions. The pins are internally connected. VPP. Power Supply Voltage for Flash test purposes. This pin is bonded and must be kept to 0 in user mode. VREG. 3V regulator output. 1.2.1 Electromagnetic Compatibility (EMC) To reduce the electromagnetic interference the following features have been implemented: – A low power oscillator is included with a controlled gain to reduce EMI and the power consumption in Halt mode. – Two or Four pairs of digital power supply pins (VDD, VSS) are located on each side of the PQFP100 package (2 pairs on TQFP64). – Digital and analog power supplies are completely separated. – Digital power supplies for internal logic and I/O ports are separated internally. – Internal decoupling capacitance is located between VDD and VSS. Note: Each pair of digital VDD/VSS pins should be externally connected by a 10 µF chemical pulling capacitor and a 100 nF ceramic chip capacitor. 1.2.2 I/O Port Alternate Functions Each pin of the I/O ports of the ST92F120 may assume software programmable Alternate Functions as shown in Section 1.3. 1.2.3 Termination of Unused Pins The ST9 device is implemented using CMOS technology; therefore unused pins must be properly terminated in order to avoid application reliability problems. In fact, as shown in Figure 3, the standard input circuitry is based on the CMOS inverter structure. ST92F120 - GENERAL DESCRIPTION Figure 3. CMOS basic inverter VDD P OUT IN N VSS When an input is kept at logic zero, the N-channel transistor is off, while the P-channel is on and can conduct. The opposite occurs when an input is kept at logic one. CMOS transistors are essentially linear devices with relatively broad switching points. During commutation, the input passes through midsupply, and there is a region of input voltage values where both P and N-channel transistors are on. Since normally the transitions are fast, there is a very short time in which a current can flow: once the switching is completed there is no longer current. This phenomenon explains why the overall current depends on the switching rate: the consumption is directly proportional to the number of transistors inside the device which are in the linear region during transitions, charging and discharging internal capacitances. In order to avoid extra power supply current, it is important to bias input pins properly when not used. In fact, if the input impedance is very high, pins can float, when not connected, either to a midsupply level or can oscillate (injecting noise in the device). Depending on the specific configuration of each I/O pin on different ST9 devices, it can be more or less critical to leave unused pins floating. For this reason, on most pins, the configuration after RESET enables an internal weak pull-up transistor in order to avoid floating conditions. For other pins this is intrinsically forbidden, like for the true opendrain pins. In any case, the application software must program the right state for unused pins to avoid conflicts with external circuitry (whichever it is: pull-up, pull-down, floating, etc.). The suggested method of terminating unused I/O is to connect an external individual pull-up or pulldown for each pin, even though initialization software can force outputs to a specified and defined value, during a particular phase of the RESET routine there could be an undetermined status at the input section. Usage of pull-ups and/or pull-downs is preferable in place of direct connection to VDD or VSS. If pullup or pull-down resistors are used, inputs can be forced for test purposes to a different value, and outputs can be programmed to both digital levels without generating high current drain due to the conflict. Anyway, during system verification flow, attention must be paid to reviewing the connection of each pin, in order to avoid potential problems. 1.2.4 Avoidance of Pin Damage Although integrated circuit data sheets provide the user with conservative limits and conditions in order to prevent damage, sometimes it is useful for the hardware system designer to know the internal failure mechanisms: the risk of exposure to illegal voltages and conditions can be reduced by smart protection design. It is not possible to classify and to predict all the possible damage resulting from violating maximum ratings and conditions, due to the large number of variables that come into play in defining the failures: in fact, when an overvoltage condition is applied, the effects on the device can vary significantly depending on lot-to-lot process variations, operating temperature, external interfacing of the ST9 with other devices, etc. In the following sections, background technical information is given in order to help system designers to reduce risk of damage to the ST9 device. 1.2.4.1 Electrostatic Discharge and Latchup CMOS integrated circuits are generally sensitive to exposure to high voltage static electricity, which can induce permanent damage to the device: a typical failure is the breakdown of thin oxides, which causes high leakage current and sometimes shorts. Latchup is another typical phenomenon occurring in integrated circuits: unwanted turning on of parasitic bipolar structures, or silicon-controlled rectifiers (SCR), may overheat and rapidly destroy the device. These unintentional structures are composed of P and N regions which work as emitters, bases and collectors of parasitic bipolar transistors: the bulk resistance of the silicon in the wells and substrate act as resistors on the SCR structure. Applying voltages below VSS or above VDD, and when the level of current is able to generate a 11/320 1 ST92F120 - GENERAL DESCRIPTION voltage drop across the SCR parasitic resistor, the SCR may be turned on; to turn off the SCR it is necessary to remove the power supply from the device. The present ST9 design implements layout and process solutions to decrease the effects of electrostatic discharges (ESD) and latchup. Of course it is not possible to test all devices, due to the destructive nature of the mechanism; in order to guarantee product reliability, destructive tests are carried out on groups of devices, according to STMicroelectronics internal Quality Assurance standards and recommendations. count, for those applications and systems where ST9 pins are exposed to illegal voltages and high current injections, the user is strongly recommended to implement hardware solutions which reduce the risk of damage to the microcontroller: low-pass filters and clamp diodes are usually sufficient in preventing stress conditions. The risk of having out-of-range voltages and currents is greater for those signals coming from outside the system, where noise effect or uncontrolled spikes could occur with higher probability than for the internal signals; it must be underlined that in some cases, adoption of filters or other dedicated interface circuitries might affect global microcontroller performance, inducing undesired timing delays, and impacting the global system speed. 1.2.4.2 Protective Interface Although ST9 input/output circuitry has been designed taking ESD and Latchup problems into acFigure 4. Digital Input/Output - Push-Pull I/OCIRCUITRY P OUTPUT BUFFER PIN N P P EN P INPUT N BUFFER ESDPROTECTION CIRCUITRY EN N PORTCIRCUITRY 12/320 1 ST92F120 - GENERAL DESCRIPTION 1.2.4.3 Internal Circuitry: Digital I/O pin In Figure 4 a schematic representation of an ST9 pin able to operate either as an input or as an output is shown. The circuitry implements a standard input buffer and a push-pull configuration for the output buffer. It is evident that although it is possible to disable the output buffer when the input section is used, the MOS transistors of the buffer itself can still affect the behaviour of the pin when exposed to illegal conditions. In fact, the P-channel transistor of the output buffer implements a direct diode to VDD (P-diffusion of the drain connected to the pin and N-well connected to VDD), while the Nchannel of the output buffer implements a diode to VSS (P-substrate connected to VSS and N-diffusion of the drain connected to the pin). In parallel to these diodes, dedicated circuitry is implemented to protect the logic from ESD events (MOS, diodes and input series resistor). The most important characteristic of these extra devices is that they must not disturb normal operating modes, while acting during exposure to over limit conditions, avoiding permanent damage to the logic circuitry. All I/O pins can generally be programmed to work also as open-drain outputs, by simply writing in the corresponding register of the I/O Port. The gate of the P-channel of the output buffer is disabled: it is important to highlight that physically the P-channel transistor is still present, so the diode to VDD works. In some applications it can occur that the voltage applied to the pin is higher than the VDD value (supposing the external line is kept high, while the ST9 power supply is turned off): this condition will inject current through the diode, risking permanent damages to the device. In any case, programming I/O pins as open-drain can help when several pins in the system are tied to the same point: of course software must pay attention to program only one of them as output at any time, to avoid output driver contentions; it is advisable to configure these pins as output opendrain in order to reduce the risk of current contentions. Figure 5. Digital Input/Output - True Open Drain Output I/OCIRCUITRY OUTPUT BUFFER PIN N P EN P INPUT N BUFFER ESDPROTECTION CIRCUITRY EN N PORT CIRCUITRY 13/320 9 ST92F120 - GENERAL DESCRIPTION In Figure 6 a true open-drain pin schematic is shown. In this case all paths to VDD are removed (P-channel driver, ESD protection diode, internal weak pull-up) in order to allow the system to turn off the power supply of the microcontroller and keep the voltage level at the pin high without injecting current in the device. This is a typical condition which can occur when several devices interface a serial bus: if one device is not involved in the communication, it can be disabled by turning off its power supply to reduce the system current consumption. When an illegal negative voltage level is applied to the ST9 I/O pins (both versions, push-pull and true open-drain output) the clamp diode is always present and active (see ESD protection circuitry and N-channel driver). 1.2.4.4 Internal Circuitry: Analog Input pin Figure 6 shows the internal circuitry used for analog input. It is substantially a digital I/O with an added analog multiplexer for the A/D Converter input signal selection. The presence of the multiplexer P-channel and Nchannel can affect the behaviour of the pin when exposed to illegal voltage conditions. These transistors are controlled by a low noise logic, biased through AVDD and AVSS including P-channel Nwell: it is important to always verify the input voltage value with respect to both analog power supply and digital power supply, in order to avoid unintended current injections which (if not limited) could destroy the device. Figure 6. Digital Input/Output - Push-Pull Output - Analog Multiplexer Input I/OCIRCUITRY P OUTPUT BUFFER PIN N P P EN INPUT N BUFFER ESDPROTECTION EN N CIRCUITRY N AVDD P PORTCIRCUITRY 14/320 9 P ST92F120 - GENERAL DESCRIPTION 1.2.4.5 Power Supply and Ground As already said for the I/O pins, in order to guarantee ST9 compliancy with respect to Quality Assurance recommendations concerning ESD and Latchup, dedicated circuits are added to the different power supply and ground pins (digital and analog). These structures create preferred paths for the high current injected during discharges, avoiding damage to active logic and circuitry. It is important for the system designer to take this added circuitry into account, which is not always transparent with respect to the relative level of voltages applied to the different power supply and ground pins. Figure 7 shows schematically the protection net implemented on ST9 devices, composed of diodes and other special structures. The clamp structure between the VDD and V SS pins is designed to be active during very fast tran- sitions (typical of electrostatic discharges). Other paths are implemented through diodes: they limit the possibility of positively differentiating AVDD and VDD (i.e. AVDD > VDD); similar considerations are valid for AVSS and VSS due to the back-toback diode structure implemented between the two pins. Anyway, it must be highlighted that, because VSS and AVSS are connected to the substrate of the silicon die (even though in different areas of the die itself), they represent the reference point from which all other voltages are measured, and it is recommended to never differentiate AVSS from VSS. Note: If more than one pair of pins for VSS and VDD is available on the device, they are connected internally and the protection net diagram remains the same as shown in Figure 7. Figure 7. Power Supply and Ground configuration N P VDD VSS AVDD AVSS P N VPP 15/320 9 ST92F120 - GENERAL DESCRIPTION HW0SW1 RESET OSCOUT OSCIN VDD VSS P7.7/A0IN7/WKUP13 P7.6/A0IN6/WKUP12 P7.5/A0IN5/WKUP11 P7.4/A0IN4/WKUP3 P7.3/A0IN3 P7.2/A0IN2 P7.1/A0IN1 P7.0/A0IN0 AVSS AVDD Figure 8. ST92F120: Pin configuration (top-view TQFP64) 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 1 47 2 46 3 45 4 44 5 43 6 42 7 41 8 40 9 39 10 38 11 37 12 36 13 35 14 34 15 33 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 ST92F120 VPWO** P6.5/WKUP10/INTCLK/VPWI P6.4/NMI P6.3/INT3/INT5 P6.2/INT2/INT4/DS2 P6.0/INT0/INT1/CLOCK2/8 P0.7/A7/D7 P0.6/A6/D6 P0.5/A5/D5 P0.4/A4/D4 P0.3/A3/D3 P0.2/A2/D2 P0.1/A1/D1 P0.0/A0/D0 AS DS RW TINPA0/P2.0 TINPB0/P2.1 TOUTA0/P2.2 TOUTB0/P2.3 TINPA1/P2.4 TINPB1/P2.5 TOUTA1/P2.6 TOUTB1/P2.7 VSS VDD VREG ***VPP A8/P1.0 A9/P1.1 A10/P1.2 WAIT/WKUP5/P5.0* WKUP6/WDOUT/P5.1* SIN0/WKUP2/P5.2 SOUT0/P5.3 TXCLK0/CLKOUT0/P5.4 RXCLK0/WKUP7/P5.5 DCD0/WKUP8/P5.6 WKUP9/RTS0/P5.7 EXTCLK1/WKUP4/P4.4 EXTRG/STOUT/P4.5 SDA/P4.6 WKUP1/SCL/P4.7 EXTCLK0/SS/P3.4 MISO/P3.5 MOSI/P3.6 SCK/WKUP0/P3.7 N.C. = Not connected (no physical bonding wire) * Alternate function for CAN interface, reserved for future use: P5.0/TXCAN0; P5.1/RXCAN0. ** On devices without JBPLD peripheral, this pin must not be connected. *** VPP must be kept low in standard operating mode. 16/320 9 ST92F120 - GENERAL DESCRIPTION P9.2/TXCLK1/CLKOUT1 P9.1/SOUT1 P9.0/SIN1 HW0SW1 RESET OSCOUT OSCIN VDD VSS P7.7/A0IN7/WKUP13 P7.6/A0IN6/WKUP12 P7.5/A0IN5/WKUP11 P7.4/A0IN4/WKUP3 P7.3/A0IN3 P7.2/A0IN2 P7.1/A0IN1 P7.0/A0IN0 AV SS AV DD P8.7/A1IN0 Figure 9. ST92F120: Pin Configuration (top-view PQFP100) 100 99 98 97 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81 80 1 2 79 3 78 4 77 5 76 6 7 75 74 8 9 10 73 72 71 70 69 11 12 13 14 15 16 17 18 19 ST92F120 20 21 68 67 66 65 64 63 62 61 22 23 24 60 59 58 57 25 26 27 56 55 54 28 53 29 52 51 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 30 P8.6/A1IN1 P8.5/A1IN2 P8.4/A1IN3 P8.3/A1IN4 P8.2/A1IN5 P8.1/A1IN6/WKUP15 P8.0/A1IN7/WKUP14 VPWO P6.5/WKUP10/INTCLK/VPWI P6.4/NMI P6.3/INT3/INT5 P6.2/INT2/INT4/DS2 P6.1/INT6/RW P6.0/INT0/INT1/CLOCK2/8 P0.7/A7/D7 V DD V SS P0.6/A6/D6 P0.5/A5/D5 P0.4/A4/D4 P0.3/A3/D3 P0.2/A2/D2 P0.1/A1/D1 P0.0/A0/D0 AS DS P1.7/A15 P1.6/A14 P1.5/A13 P1.4/A12 V REG RW TINPA0/P2.0 TINPB0/P2.1 TOUTA0/P2.2 TOUTB0/P2.3 TINPA1/P2.4 TINPB1/P2.5 TOUTA1/P2.6 TOUTB1/P2.7 VSS VDD V REG ***V PP A8/P1.0 A9/P1.1 A10/P1.2 A11/P1.3 **N.C. **N.C. RXCLK1/P9.3 DCD1/P9.4 RTS1/P9.5 CLOCK2/P9.6 P9.7 WAIT/WKUP5/P5.0* WKUP6/WDOUT/P5.1* SIN0/WKUP2/P5.2 SOUT0/P5.3 TXCLK0/CLKOUT0/P5.4 RXCLK0/WKUP7/P5.5 DCD0/WKUP8/P5.6 WKUP9/RTS0/P5.7 ICAPA1/P4.0 P4.1 OCMPA1/P4.2 VSS VDD ICAPB1/OCMPB1/P4.3 EXTCLK1/WKUP4/P4.4 EXTRG/STOUT /P4.5 SDA/P4.6 WKUP1/SCL/P4.7 ICAPB0/P3.1 ICAPA0/OCMPA0/P3.2 OCMPB0/P3.3 EXTCLK0/SS/P3.4 MISO/P3.5 MOSI/P3.6 SCK/WKUP0/P3.7 N.C. = Not connected (no physical bonding wire) * Alternate function for CAN interface, reserved for future use: P5.0/TXCAN0; P5.1/RXCAN0 ** Pin reserved for future use: 49 - RXCAN1; 50 - TXCAN1 *** VPP must be kept low in standard operating mode. 17/320 9 ST92F120 - GENERAL DESCRIPTION 18 42 65 93 17 41 64 92 Analog Circuit Supply Voltage 49 82 AVSS Analog Circuit Ground 50 83 VPP Must be kept low in standard operating mode 29 44 VREG 3V regulator output 28 31 43 VDD VSS AVDD Function Main Power Supply Voltage ( pins internally connected) Digital Circuit Ground (pins internally connected) Name Function QFP100 27 60 26 59 Name QFP64 QFP100 Table 2. ST92F120 Primary Function Pins QFP64 Table 1. ST92F120 Power Supply Pins AS DS RW Address Strobe Data Strobe Read/Write 34 33 17 56 55 32 OSCIN OSCOUT Oscillator Input Oscillator Output Reset to initialize the Microcontroller Watchdog HW/SW enabling selection J1850 JBLPD Output. On devices without JBPLD peripheral, this pin must not be connected. 61 62 94 95 63 96 64 97 48 73 RESET HW0SW1 VPWO Figure 10. Recommended connections for VREG QFP100 300 nF QFP 64 300 nF 600 nF Note : For future compatibility with shrinked versions, the VREG pins should be connected to a minimum of 600 nF (total). Special care should be taken to minimize the distance between the ST9 microcontroller and the capacitors. 18/320 9 ST92F120 - GENERAL DESCRIPTION 1.3 I/O PORTS Port 0 and Port 1 provide the external memory interface. 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). Internal Weak Pull-up As shown in Table 3, not all input sections implement a Weak Pull-up. This means that the pull-up must be connected externally when the pin is not used or programmed as bidirectional. 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”. Schmitt Trigger Input Two different kind of Schmitt Trigger circuitries are implemented: Standard and High Hysteresis. Standard Schmitt Trigger is widely used (see Table 3), while the High Hysteresis one is present on the NMI and VPWI input function pins mapped on Port 6 [5:4] (see Table 4). All inputs which can be used for detecting interrupt events have been configured with a “standard” Schmitt Trigger, apart from, as already said, the NMI pin which implements the “High Hysteresis” version. In this way, all interrupt lines are guaranteed as “level sensitive”. 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 Port Port Port Port Port Port Port Port Port 0[7:0] 1[7:0] 2[1:0] 2[3:2] 2[5:4] 2[7:6] 3[2:1] 3.3 3[7:4] 4.0, Port 4.4 4.1 4.2, Port 4.5 4.3 4[7:6] 5[2:0], Port [7:4] 5.3 6[3:0] 6[5:4] Port Port Port Port 7[7:0] 8[1:0] 8[7:2] 9[7:0] Input TTL/CMOS TTL/CMOS Schmitt trigger TTL/CMOS Schmitt trigger TTL/CMOS Schmitt trigger TTL/CMOS Schmitt trigger Schmitt trigger TTL/CMOS TTL/CMOS Schmitt trigger Schmitt trigger inside I/O cell Schmitt trigger TTL/CMOS Schmitt trigger High hysteresis Schmitt trigger inside I/O cell Schmitt trigger Schmitt trigger Schmitt trigger Schmitt trigger Outpu t Push-Pull/OD Push-Pull/OD Push-Pull/OD Pure OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Pure OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Weak Pull-Up No No Yes No Yes Yes Yes Yes Yes No Yes Yes Yes No No Yes Yes Yes (inside I/O cell) Reset State Bidirectional Bidirectional Input Input CMOS Input Input CMOS Input Input CMOS Input Input Bidirectional WPU Input CMOS Input Input Input Input CMOS Input Input Push-Pull/OD Push-Pull/OD Push-Pull/OD Push-Pull/OD Yes Yes Yes Yes Input Input Bidirectional WPU Bidirectional WPU Legend: WPU = Weak Pull-Up, OD = Open Drain 19/320 9 ST92F120 - GENERAL DESCRIPTION How to Configure the I/O ports To configure the I/O ports, use the information in Table 3, Table 4 and the Port Bit Configuration Table in the I/O Ports Chapter (See page 125). Input Note = the hardware characteristics fixed for each port line in Table 3. – If Input note = TTL/CMOS, either TTL or CMOS input level can be selected by software. – If Input note = Schmitt trigger, selecting CMOS or TTL input by software has no effect, the input will always be Schmitt Trigger. Alternate Functions (AF) = More than one AF cannot be assigned to an I/O pin at the same time: An alternate function can be selected as follows. AF Inputs: – AF is selected implicitly by enabling the corresponding peripheral. Exception to this are A/D inputs which must be explicitly selected as AF by software. AF Outputs or Bidirectional Lines: – In the case of Outputs or I/Os, AF is selected explicitly by software. 20/320 9 Example 1: SCI input AF: SIN0, Port: P5.2, Input note: Schmitt Trigger. Write the port configuration bits: P5C2.2=1 P5C1.2=0 P5C0.2 =1 Enable the SCI peripheral by software as described in the SCI chapter. Example 2: SCI output AF: SOUT0, Port: P5.3, Output note: Push-Pull/OD. Write the port configuration bits (for AF OUT PP): P5C2.3=0 P5C1.3=1 P5C0.3 =1 Example 3: External Memory I/O AF: A0/D0, Port : P0.0, Input Note: TTL/CMOS Write the port configuration bits: P0C2.0=1 P0C1.0=1 P0C0.0 =1 Example 4: Analog input AF: A0IN0, Port : 7.0, Input Note: does not apply to analog input Write the port configuration bits: P7C2.0=1 P7C1.0=1 P7C0.0 =1 ST92F120 - GENERAL DESCRIPTION Table 4. I/O Port Alternate Functions Port Name General Purpose I/O Pin No. Alternate Function s TQFP64 PQFP100 P0.0 35 57 A0/D0 I/O Address/Data bit 0 P0.1 36 58 A1/D1 I/O Address/Data bit 1 P0.2 37 59 A2/D2 I/O Address/Data bit 2 P0.3 38 60 A3/D3 I/O Address/Data bit 3 P0.4 39 61 A4/D4 I/O Address/Data bit 4 P0.5 40 62 A5/D5 I/O Address/Data bit 5 P0.6 41 63 A6/D6 I/O Address/Data bit 6 P0.7 42 66 A7/D7 I/O Address/Data bit 7 P1.0 30 45 A8 I/O Address bit 8 P1.1 31 46 A9 I/O Address bit 9 P1.2 32 47 A10 I/O Address bit 10 P1.3 - 48 A11 I/O Address bit 11 P1.4 - 51 A12 I/O Address bit 12 P1.5 - 52 A13 I/O Address bit 13 P1.6 - 53 A14 I/O Address bit 14 P1.7 - 54 A15 I/O Address bit 15 18 33 TINPA0 19 34 20 35 21 P2.4 P2.0 I Multifunction Timer 0 - Input A TINPB0 I Multifunction Timer 0 - Input B TOUTA0 O Multifunction Timer 0 - Output A 36 TOUTB0 O Multifunction Timer 0 - Output B 22 37 TINPA1 I Multifunction Timer 1 - Input A P2.5 23 38 TINPB1 I Multifunction Timer 1 - Input B P2.6 24 39 TOUTA1 O Multifunction Timer 1 - Output A P2.7 25 40 TOUTB1 O Multifunction Timer 1 - Output B P3.1 - 24 ICAPB0 I Ext. Timer 0 - Input Capture B P3.2 - 25 ICAPA0 I Ext. Timer 0 - Input Capture A OCMPA0 O Ext. Timer 0 - Output Compare A P3.3 - 26 OCMPB0 O Ext. Timer 0 - Output Compare B P3.4 13 27 EXTCLK0 I Ext. Timer 0 - Input Clock I SPI - Slave Select P3.5 14 28 MISO I/O SPI - Master Input/Slave Output Data P3.6 15 29 MOSI I/O SPI - Master Output/Slave Input Data P2.1 P2.2 P2.3 All ports useable for general purpose I/O (input, output or bidirectional) SS SCK P3.7 P4.0 16 - 30 14 I SPI - Serial Input Clock WKUP0 I Wake-up Line 0 SCK O SPI - Serial Output Clock ICAPA1 I Ext. Timer 1 - Input Capture A 21/320 9 ST92F120 - GENERAL DESCRIPTION Port Name General Purpose I/O Pin No. Alternate Function s TQFP64 PQFP100 P4.1 - 15 P4.2 - 16 P4.3 - 19 P4.4 9 20 P4.5 10 21 P4.6 11 22 P4.7 12 23 P5.0 1 6 P5.1 2 7 3 8 4 9 5 10 I/O OCMPA1 ICAPB1 I OCMPB1 O Ext. Timer 1 - Output Compare B EXTCLK1 I Ext. Timer 1 - Input Clock WKUP4 I Wake-up Line 4 EXTRG I A/D 0 and A/D 1 - Ext. Trigger STOUT O Standard Timer Output SDA I/O I2C Data WKUP1 P5.2 P5.3 P5.4 P5.5 6 11 P5.6 7 12 P5.7 P6.0 P6.1 P6.2 22/320 9 All ports useable for general purpose I/O (input, output or bidirectional) 8 43 - 44 13 67 68 69 O Ext. Timer 1 - Output Compare A SCL I Ext. Timer 1 - Input Capture B Wake-up Line 1 I/O I2C Clock WAIT I External Wait Request WKUP5 I Wake-up Line 5 WKUP6 I Wake-up Line 6 WDOUT O Watchdog Timer Output SIN0 I SCI0 - Serial Data Input WKUP2 I Wake-up Line 2 SOUT0 O SCI0 - Serial Data Output TXCLK0 I CLKOUT0 O SCI0 - Clock Output RXCLK0 I SCI0 - Receive Clock Input WKUP7 I Wake-up Line 7 DCD0 I SCI0 - Data Carrier Detect WKUP8 I Wake-up Line 8 WKUP9 I Wake-up Line 9 RTS0 O SCI0 - Request To Send INT0 I External Interrupt 0 INT1 I External Interrupt 1 CLOCK2/ 8 O CLOCK2 divided by 8 SCI0 - Transmit Clock Input INT6 I RW O Read/Write External Interrupt 6 INT2 I External Interrupt 2 INT4 I External Interrupt 4 DS2 O Data Strobe 2 ST92F120 - GENERAL DESCRIPTION Port Name General Purpose I/O Pin No. P6.3 45 70 P6.4 46 71 P6.5 Alternate Function s TQFP64 PQFP100 47 72 INT3 I External Interrupt 3 INT5 I External Interrupt 5 NMI I Non Maskable Interrupt WKUP10 I Wake-up Line 10 VPWI I JBLPD input INTCLK O Internal Main Clock P7.0 51 84 A0IN0 I A/D 0 - Analog Data Input 0 P7.1 52 85 A0IN1 I A/D 0 - Analog Data Input 1 P7.2 53 86 A0IN2 I A/D 0 - Analog Data Input 2 P7.3 54 87 A0IN3 I A/D 0 - Analog Data Input 3 P7.4 55 88 WKUP3 I Wake-up Line 3 A0IN4 I A/D 0 - Analog Data Input 4 P7.5 56 89 A0IN5 I A/D 0 - Analog Data Input 5 WKUP11 I Wake-up Line 11 P7.6 57 90 A0IN6 I A/D 0 - Analog Data Input 6 WKUP12 I Wake-up Line 12 58 91 A0IN7 I A/D 0 - Analog Data Input 7 WKUP13 I Wake-up Line 13 - 74 A1IN7 I A/D 1 - Analog Data Input 7 WKUP14 I Wake-up Line 14 P8.1 - 75 A1IN6 I A/D 1 - Analog Data Input 6 WKUP15 I Wake-up Line 15 P8.2 - 76 A1IN5 I A/D 1 - Analog Data Input 5 P8.3 - 77 A1IN4 I A/D 1 - Analog Data Input 4 P8.4 - 78 A1IN3 I A/D 1 - Analog Data Input 3 P8.5 - 79 A1IN2 I A/D 1 - Analog Data Input 2 P8.6 - 80 A1IN1 I A/D 1 - Analog Data Input 1 P8.7 - 81 A1IN0 I A/D 1 - Analog Data Input 0 P9.0 - 98 SIN1 I SCI1 - Serial Data Input P9.1 - 99 SOUT1 O SCI1 - Serial Data Output P9.2 - 100 P9.3 - P9.4 P7.7 P8.0 All ports useable for general purpose I/O (input, output or bidirectional) TXCLK1 I CLKOUT1 O SCI1 - Clock Input SCI1 - Transmit Clock input 1 RXCLK1 I SCI1 - Receive Clock Input - 2 DCD1 I SCI1 - Data Carrier Detect P9.5 - 3 RTS1 O SCI1 - Request To Send P9.6 - 4 CLOCK2 O CLOCK2 internal signal P9.7 - 5 I/O 23/320 9 ST92F120 - GENERAL DESCRIPTION 1.4 OPERATING MODES To optimize the performance versus the power consumption of the device, the ST92F120 supports different 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 depending on the CCU programming. LOW POWER WAIT FOR INTERRUPT MODE: Combining SLOW mode and Wait For Interrupt mode it is possible to reduce the power consumption by more than 80%. STOP MODE: When the STOP is requested by executing the STOP bit writing sequence (see dedicated section on Wake-up Management Unit paragraph), and if NMI is kept low, the CPU and the peripherals stop operating. Operations resume 24/320 9 after a wake-up line is activated (16 wake-up lines plus NMI pin). See the RCCU and Wake-up Management Unit paragraphs in the following for the details. The difference with the HALT mode consists in the way the CPU exits this state: when the STOP is executed, the status of the registers is recorded, and when the system exits from the STOP mode the CPU continues the execution with the same status, without a system reset. When the MCU enters STOP mode the Watchdog stops counting. After the MCU exits from STOP mode, the Watchdog resumes counting from where it left off. When the MCU exits from STOP mode, the oscillator, which was sleeping too, requires about 5 ms to restart working properly (at a 4 MHz oscillator frequency). An internal counter is present to guarantee that all operations after exiting STOP Mode, take place with the clock stabilised. The counter is active only when the oscillation has already taken place. This means that 1-2 ms must be added to take into account the first phase of the oscillator restart. 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. ST92F120 - 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 11. 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 12): 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 13. Figure 11. 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 25/320 9 ST92F120 - DEVICE ARCHITECTURE MEMORY SPACES (Cont’d) Figure 12. Register Groups Figure 13. 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 14. 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 26/320 9 ST92F120 - 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 14). 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 2.3.3 Register Pointing Techniques, and illustrated in Figure 15 and in Figure 16. 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 2.3 SYSTEM REGISTERS. 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 5. 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 27/320 9 ST92F120 - DEVICE ARCHITECTURE 2.3 SYSTEM REGISTERS The System registers are listed in Table 6. 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 6. 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. GCE TLIP N 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. 28/320 9 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. ST92F120 - 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. 29/320 9 ST92F120 - 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 30/320 9 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”. ST92F120 - 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. 31/320 9 ST92F120 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) Figure 15. Pointing to a single group of 16 registers REGIST ER GROUP BLOCK NUMBER REGISTER GROUP BLOCK NUMBER Figure 16. 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 32/320 9 r7 2 addressed by BLOCK 2 1 0 0 GROUP 1 addressed by BLOCK 2 ST92F120 - 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, 33/320 9 ST92F120 - 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 34/320 9 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 6. 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. ST92F120 - DEVICE ARCHITECTURE SYSTEM REGISTERS (Cont’d) USER STACK POINTER HIGH REGISTER (USPHR) R236 - Read/Write Register Group: E (System) Reset value: undefined 7 SYSTEM STACK POINTER HIGH REGISTER (SSPHR) R238 - Read/Write Register Group: E (System) Reset value: undefined 0 7 0 USP1 USP1 USP1 USP1 USP1 USP1 USP9 USP8 5 4 3 2 1 0 SSP1 SSP1 SSP1 SSP1 SSP1 SSP1 SSP9 SSP8 5 4 3 2 1 0 USER STACK POINTER LOW REGISTER (USPLR) R237 - Read/Write Register Group: E (System) Reset value: undefined SYSTEM STACK POINTER LOW REGISTER (SSPLR) R239 - Read/Write Register Group: E (System) Reset value: undefined 7 0 7 0 USP7 USP6 USP5 USP4 USP3 USP2 USP1 USP0 SSP7 SSP6 SSP5 SSP4 SSP3 SSP2 SSP1 SSP0 Figure 17. Internal Stack Mode Figure 18. External Stack Mode REGIST ER FILE REGISTER FILE STACK POINTER (LOW) points to: F 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 35/320 9 ST92F120 - 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. 36/320 9 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. ST92F120 - 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 19. 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 37/320 9 ST92F120 - 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 20. 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 20). 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 38/320 9 2M SB ST92F120 - 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 21). 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 19. Figure 21. 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 39/320 9 ST92F120 - 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 DPR0 _7 DPR0 _6 DPR0 _5 DPR0 _4 DPR0 _3 DPR0 DPR0 _2 _1 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. 0 7 DPR0 _0 DPR2 _7 0 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 DPR1 _7 DPR1 _6 DPR1 _5 DPR1 _4 DPR1 _3 DPR1 DPR1 _2 _1 0 7 DPR1 _0 DPR3 _7 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. 40/320 9 0 DPR3 _6 DPR3 _5 DPR3 _4 DPR3 _3 DPR3 _2 DPR3 _1 DPR3 _0 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. ST92F120 - 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_ CSR_ 2 1 CSR_ 0 Bit 7:6 = Reserved, keep in reset state. ISR and ENCSR bit (EMR2 register) are also described in the chapter relating to Interrupts, please refer to this description for further details. 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 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 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 0 0 0 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. 41/320 9 ST92F120 - DEVICE ARCHITECTURE MMU REGISTERS (Cont’d) Figure 22. 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 42/320 9 ISR 64K CSR 16K 64K 010000h 00C000h 000000h ST92F120 - 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). 43/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM 3 SINGLE VOLTAGE FLASH & EEPROM 3.1 INTRODUCTION The Flash circuitry contains one array divided in two main parts that can each be read independently. The first part contains the main Flash array for code storage, a reserved array (TestFlash) for system routines and a 128-byte area available as one time programmable memory (OTP). The second part contains the two dedicated Flash sectors used for EEPROM Hardware Emulation. The write operations of the two parts are managed by an embedded Program/Erase Controller, that uses a dedicated ROM (256 words of 12 bits each). Through a dedicated RAM buffer the Flash and the EEPROM can be written in blocks of 16 bytes. Figure 23. Flash Memory structure (Example for 128K Flash device) 8 sense + 8 program load Address 230000h 231F80h TestFlash 8 Kbytes User OTP and Protection registers 000000h Sector F0 64 Kbytes Data Register Interface RAM buffer 16 bytes 010000h Sector F1 48 Kbytes 01C000h 01E000h 228000h 22C000h Sector F2 8 Kbytes Sector F3 8 Kbytes Sector F4/E0 4 Kbytes Sector F5/E1 4 Kbytes 8 sense + 8 program load 44/320 9 Program / Erase Controller ST92F120 - SINGLE VOLTAGE FLASH & EEPROM 3.2 FUNCTIONAL DESCRIPTION 3.2.1 Structure The Flash memory is composed of three parts (see following table): – 1 reserved sector for system routines (TestFlash including user OTP area) – 4 main sectors for code – 2 sectors of the same size for EEPROM emulation The last 128 bytes of the TestFlash are available to the user as an OTP area. The user can program these bytes, but cannot erase them. The last 4 bytes of this OTP area (231FFCh to 231FFFh for 128K Flash device and 230FFCh to 230FFFh for 60K/36K Flash devices) are reserved for the NonVolatile Protection registers and cannot be used as a storage area (see Section 3.6 PROTECTION STRATEGY for more details). 3.2.2 Software or Hardware EEPROM Emulation (Device dependent option) The MCU can be factory-configured to allow the user to manage the EEPROM emulation by software (using for example the Intel algorithm), by directly addressing the two dedicated sectors F4 and F5. In this case the Hardware EEPROM emulation will not be available. Hardware EEPROM emulation By default, a hardware EEPROM emulation is implemented using special flash sectors F4/E0 and F5/E1 to emulate an EEPROM memory whose size is 1/4 of a sector (1 Kbytes max). This EEPROM can be directly addressed from 220000h to 2203FFh for 128K Flash device and 220000h to 2201FFh for 60K/36K Flash devices. In this case, Flash sectors F4 and F5 are not directly accessible. (see Section 3.5.1 Hardware EEPROM Emulation for more details). Table 7. Memory Structure for 128K Flash device Sector Addresses Max Size TestFlash (TF) (Reserved) 230000h to 231F7Fh 8064 bytes User OTP Area 231F80h to 231FFFh 128 bytes Flash 0 (F0) 000000h to 00FFFFh 64 Kbytes Flash 1 (F1) 010000h to 01BFFFh 48 Kbytes Flash 2 (F2) 01C000h to 01DFFFh 8 Kbytes Flash 3 (F3) 01E000h to 01FFFFh 8 Kbytes Flash 4 / EEPROM 0 (F4/E0) 228000h to 228FFFh 4 Kbytes Flash 5 / EEPROM 1 (F5/E1) 22C000h to 22CFFFh 4 Kbytes Emulated EEPROM 220000h to 2203FFh 1 Kbyte Table 8. Memory Structure for 60K Flash device Sector Addresses Max Size TestFlash (TF) (Reserved) 230000h to 230F7Fh 3968 bytes User OTP Area 230F80h to 230FFFh 128 bytes Flash 0 (F0) 000000h to 000FFFh 4 Kbytes Flash 1 (F1) 010000h to 017FFFh 32 Kbytes Flash 2 (F2) 018000h to 01BFFFh 16 Kbytes Flash 3 (F3) 01C000h to 01DFFFh 8 Kbytes Flash 4 / EEPROM 0 (F4/E0) 228000h to 2287FFh 2 Kbytes Flash 5 / EEPROM 1 (F5/E1) 22C000h to 22C7FFh 2 Kbytes Emulated EEPROM 220000h to 2201FFh 512 bytes 45/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM FUNCTIONAL DESCRIPTION (Cont’d) Table 9. Memory Structure for 36K Flash device Sector Addresses Max Size TestFlash (TF) (Reserved) 230000h to 230F7Fh 3968 bytes User OTP Area 231F80h to 230FFFh 128 bytes Flash 0 (F0) 000000h to 000FFFh 4 Kbytes Flash 1 (F1) 010000h to 013FFFh 16 Kbytes Flash 2 (F2) 014000h to 015FFFh 8 Kbytes Flash 3 (F3) 016000h to 017FFFh 8 Kbytes Flash 4 / EEPROM 0 (F4/E0) 228000h to 2287FFh 2 Kbytes Flash 5 / EEPROM 1 (F5/E1) 22C000h to 22C7FFh 2 Kbytes Emulated EEPROM 220000h to 2201FFh 512 bytes 3.2.3 Operation The memory has a register interface mapped in memory space (segment 22h). All operations are enabled through the FCR (Flash Control Register) ECR (EEPROM Control Register). All operations on the Flash must be executed from another memory (internal RAM, EEPROM, external memory). Flash (including TestFlash) and EEPROM have duplicated sense amplifiers, so that one can be read while the other is written. However simultaneous Flash and EEPROM write operations are forbidden. An interrupt can be generated at the end of a Flash or an EEPROM write operation: this interrupt is multiplexed with an external interrupt EXTINTx (device dependent) to generate an interrupt INTx. The status of a write operation inside the Flash and the EEPROM memories can be monitored through the FESR[1:0] registers. 46/320 9 Control and Status registers are mapped in memory (segment 22h), as shown in the following figure. Figure 24. Control and Status Register Map. Register Interface 224000h 224001h 224002h 224003h FCR ECR FESR0 FESR1 During a write operation, if the power supply drops or the RESET pin is activated, the write operation is immediately interrupted. In this case the user must repeat the last write operation following power on or reset. ST92F120 - SINGLE VOLTAGE FLASH & EEPROM 3.3 REGISTER DESCRIPTION 3.3.1 Control Registers FLASH CONTROL REGISTER (FCR) Address: 224000h - Read/Write Reset value: 0000 0000 (00h) 7 6 5 4 3 2 1 0 FWM FPAG FCHI FBYT FSEC FSUS FBUS PROT Y S E P E T P The Flash Control Register is used to enable all the operations for the Flash and the TestFlash memories, but also for the two dedicated EEPROM sectors F4/E0 and F5/E1 if they are addressed directly when using software EEPROM emulation (the FCR register must be used in this case only to select operations, while the ECR register must still be used to start the operation with the EWMS bit). The write access to the TestFlash is possible only in test mode, except the OTP area of the TestFlash that can be programmed in user mode (but not erased). Bit 7 = FWMS: Flash Write Mode Start (Read/ Write). This bit must be set to start every write/erase operation in Flash memory. At the end of the write/ erase operation or during a Sector Erase Suspend this bit is automatically reset. To resume a suspended Sector Erase operation, this bit must be set again. Resetting this bit by software does not stop the current write operation. 0: No effect 1: Start Flash write Bit 6 = FPAGE: Flash Page program (Read/Write). This bit must be set to select the Page Program operation in Flash memory. The Page Program operation allows to program “0”s in place of “1”s. From 1 to 16 bytes can be entered (in any order, no need for an ordered address sequence) before starting the execution by setting the FWMS bit . All the addresses must belong to the same page (only the 4 LSBs of address can change). Data to be programmed and addresses in which to program must be provided (through an LD instruction, for example). Data contained in page addresses that are not entered are left unchanged. This bit is automatically reset at the end of the Page Program operation. 0: Deselect page program 1: Select page program Bit 5 = FCHIP: Flash CHIP erase (Read/Write). This bit must be set to select the Chip Erase operation in Flash memory. The Chip Erase operation allows to erase all the Flash locations to FFh. The operation is limited to Flash code (sectors F0-F3; TestFlash and EEPROM sectors excluded). The execution starts by setting the FWMS bit. It is not necessary to pre-program the sectors to 00h, because this is done automatically. This bit is automatically reset at the end of the Chip Erase operation. 0: Deselect chip erase 1: Select chip erase Bit 4 = FBYTE: Flash byte program (Read/Write). This bit must be set to select the Byte Program operation in Flash memory. The Byte Program operation allows “0”s to be programmedin place of “1”s. Data to be programmed and an address in which to program must be provided (through an LD instruction, for example) before starting execution by setting bit FWMS. This bit is automatically reset at the end of the Byte Program operation. 0: Deselect byte program 1: Select byte program Bit 3 = FSECT: Flash sector erase (Read/Write). This bit must be set to select the Sector Erase operation in Flash memory. The Sector Erase operation erases all the Flash locations to FFh. From 1 to 4 sectors (F0, ..,F3) can be simultaneously erased, while TF, F4, F5 must be individually erased. Sectors to be simultaneously erased can be entered before starting the execution by setting the FWMS bit. An address located in the sector to erase must be provided (through an LD instruction, for example), while the data to be provided is don’t care. It is not necessary to pre-program the sectors to 00h, because this is done automatically. This bit is automatically reset at the end of the Sector Erase operation. 0: Deselect sector erase 1: Select sector erase Bit 2 = FSUSP: Flash sector erase suspend (Read/Write). This bit must be set to suspend the current Sector Erase operation in Flash memory in order to read data to or from program data to a sector not being erased. The Erase Suspend operation resets the Flash memory to normal read mode (automatically resetting bit FBUSY) in a maximum time of 15µs. 47/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM REGISTER DESCRIPTION (Cont’d) When in Erase Suspend the memory accepts only the following operations: Read, Erase Resume and Byte Program. Updating the EEPROM memory is not possible during a Flash Erase Suspend. The FSUSP bit must be reset (and FWMS must be set again) to resume a suspended Sector Erase operation. 0: Resume sector erase when FWMS is set again. 1: Suspend Sector erase Bit 1 = PROT: Set Protection (Read/Write). This bit must be set to select the Set Protection operation. The Set Protection operation allows “0”s in place of “1”s to be programmed in the four Non Volatile Protection registers. From 1 to 4 bytes can be entered (in any order, no need for an ordered address sequence) before starting the execution by setting the FWMS bit . Data to be programmed and addresses in which to program must be provided (through an LD instruction, for example). Protection contained in addresses that are not entered are left unchanged. This bit is automatically reset at the end of the Set Protection operation. 0: Deselect protection 1: Select protection Bit 0 = FBUSY: Flash Busy (Read Only). This bit is automatically set during Page Program, Byte Program, Sector Erase or Set Protection operations when the first address to be modified is latched in Flash memory, or during Chip Erase operation when bit FWMS is set. When this bit is set every read access to the Flash memory will output invalid data (FFh equivalent to a NOP instruction), while every write access to the Flash memory will be ignored. At the end of the write operations or during a Sector Erase Suspend this bit is automatically reset and the memory returns to read mode. After an Erase Resume this bit is automatically set again. If the two EEPROM sectors E0 and E1 are used instead of the embedded hardware emulation, FBUSY remains low during a modification in those sectors (while EBUSY rises), so that reading in Flash memory remains possible. The FBUSY bit remains high for a maximum of 10µs after PowerUp and when exiting Power-Down mode, meaning that the Flash memory is not yet ready to be accessed. 0: Flash not busy 1: Flash busy EEPROM CONTROL REGISTER (ECR) Address: 224001h - Read/Write Reset value: 000x x000 (xxh) 7 6 5 EWM EPAG ECHI S E P 9 3 2 WFIS 1 0 FEIE EBUS N Y The EEPROM Control Register is used to enable all the operations for the EEPROM memory in devices with EEPROM hardware emulation. The ECR also contains two bits (WFIS and FEIEN) that are related to both Flash and EEPROM memories. Bit 7 = EWMS: EEPROM Write Mode Start. This bit must be set to start every write/erase operation in the EEPROM memory. At the end of the write/erase operation this bit is automatically reset. Resetting by software this bit does not stop the current write operation. 0: No effect 1: Start EEPROM write Bit 6 = EPAGE: EEPROM page update. This bit must be set to select the Page Update operation in EEPROM memory. The Page Update operation allows to write a new content: both “0”s in place of “1”s and “1”s in place of “0”s. From 1 to 16 bytes can be entered (in any order, no need for an ordered address sequence) before starting the execution by setting bit EWMS. All the addresses must belong to the same page (only the 4 LSBs of address can change). Data to be programmed and addresses in which to program must be provided (through an LD instruction, for example). Data contained in page addresses that are not entered are left unchanged. This bit is automatically reset at the end of the Page Update operation. 0: Deselect page update 1: Select page update Bit 5 = ECHIP: EEPROM chip erase. This bit must be set to select the Chip Erase operation in the EEPROM memory. The Chip Erase operation allows to erase all the EEPROM locations to (E0 and E1 sectors) FFh. The execution starts by setting bit EWMS. This bit is automatically reset at the end of the Chip Erase operation. 0: Deselect chip erase 1: Select chip erase Bit 4:3 = Reserved. 48/320 4 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM REGISTER DESCRIPTION (Cont’d) Bit 2 = WFIS: Wait For Interrupt Status. If this bit is reset, the WFI instruction puts the Flash macrocell in Stand-by mode (immediate read possible, but higher consumption: 100 µA); if it is set, the WFI instruction puts the Flash macrocell in Power-Down mode (recovery time of 10µs needed before reading, but lower consumption: 10µA). The Stand-by mode or the Power-Down mode will be entered only at the end of any current Flash or EEPROM write operation. In the same way following an HALT or a STOP instruction, the Memory enters Power-Down mode only after the completion of any current write operation. 0: Flash in Standby mode on WFI 1: Flash in Power-Down mode on WFI Note: HALT or STOP mode can be exited without problems, but the user should take care when exiting WFI Power Down mode. If WFIS is set, the user code must reset the XT_DIV16 bit in the R242 register (page 55) before executing the WFI instruction. When exiting WFI mode, this gives the Flash enough time to wake up before the interrupt vector fetch. Bit 0 = EBUSY: EEPROM Busy (Read Only). This bit is automatically set during a Page Update operation when the first address to be modified is latched in the EEPROM memory, or during Chip Erase operation when bit EWMS is set. At the end of the write operation or during a Sector Erase Suspend this bit is automatically reset and the memory returns to read mode. When this bit is set every read access to the EEPROM memory will output invalid data (FFh equivalent to a NOP instruction), while every write access to the EEPROM memory will be ignored. At the end of the write operation this bit is automatically reset and the memory returns to read mode. EBUSY rises also if a write operation is started in one of the two EEPROM sectors, used as Flash sectors. Bit EBUSY remains high for a maximum of 10ms after Power-Up and when exiting Power-Down mode, meaning that the EEPROM memory is not yet ready to be accessed. 0: EEPROM not busy 1: EEPROM busy Bit 1 = FEIEN: Flash & EEPROM Interrupt enable. This bit selects the source of interrupt channel INTx between the external interrupt pin and the Flash/EEPROM End of Write interrupt. Refer to the Interrupt chapter for the channel number. 0: External interrupt enabled 1: Flash & EEPROM Interrupt enabled 49/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM REGISTER DESCRIPTION (Cont’d) 3.3.2 Status Registers During a Flash or an EEPROM write operation any attempt to read the memory under modification will output invalid data (FFh equivalent to a NOP instruction). This means that the Flash memory is not fetchable when a write operation is active: the write operation commands must be given from another memory (EEPROM, internal RAM, or external memory). Two Status Registers (FESR[1:0] are available to check the status of the current write operation in Flash and EEPROM memories. FLASH & EEPROM STATUS REGISTER 0 (FESR0) Address: 224002h -Read/Write Reset value: 0000 0000 (00h) 7 6 5 4 3 2 1 0 FEER FESS FESS FESS FESS FESS FESS FESS R 6 5 4 3 2 1 0 Bit 7 = FEERR: Flash or EEPROM write ERRor (Read/Write). This bit is set by hardware when an error occurs during a Flash or an EEPROM write operation. It must be cleared by software. 0: Write OK 1: Flash or EEPROM write error Bit 6:0 = FESS[6:0]. Flash and EEPROM Status Sector 6-0 (Read Only). These bits are set by hardware and give the status of the 7 Flash and EEPROM sectors (TF, F5, F4, F3, F2, F1, F0). The meaning of FESSx bit for sector x is given by the following table: Table 10. FESSx bit Values FEERR EBUSY FSUSP FESSx=1 meaning 1 - - Write Error in Sector x 0 1 - Write operation on-going in sector x 50/320 9 FBUSY Table 10. FESSx bit Values FBUSY FEERR EBUSY FSUSP FESSx=1 meaning 0 0 1 Sector Erase Suspended in sector x 0 0 0 Don’t care FLASH & EEPROM STATUS REGISTER 1 (FESR1) Address: 224003h -Read Only Reset value: 0000 0000 (00h) 7 6 ERER PGER 5 4 3 2 1 0 SWE R Bit 7 = ERER. Erase error (Read Only). This bit is set by hardware when an Erase error occurs during a Flash or an EEPROM write operation. This error is due to a real failure of a Flash cell, that can not be erased anymore. This kind of error is fatal and the sector where it occurred must be discarded (if it was in one of the EEPROM sectors, the hardware emulation can not be used anymore). This bit is automatically cleared when bit FEERR of the FESR0 register is cleared by software. 0: Erase OK 1: Erase error Bit 6 = PGER. Program error (Read Only). This bit is automatically set when a Program error occurs during a Flash or an EEPROM write operation. This error is due to a real failure of a Flash cell, that can not be programmed anymore. The byte where this error occurred must be discarded (if it was in the EEPROM memory, the byte must be reprogrammed to FFh and then discarded, to avoid the error occurring again when that byte is internally moved). This bit is automatically cleared when bit FEERR of the FESR0 register is cleared by software. 0: Program OK 1: Flash or EEPROM Programming error ST92F120 - SINGLE VOLTAGE FLASH & EEPROM REGISTER DESCRIPTION (Cont’d) Bit 5 = SWER. Swap or 1 over 0 Error (Read Only). This bit has two different meanings, depending on whether the current write operation is to Flash or EEPROM memory. In Flash memory this bit is automatically set when trying to program at 1 bits previously set at 0 (this does not happen when programming the Protection bits). This error is not due to a failure of the Flash cell, but only flags that the desired data has not been written. In the EEPROM memory this bit is automatically set when a Program error occurs during the swapping of the unselected pages to the new sector when the old sector is full (see Section 3.5.1 Hardware EEPROM Emulation for more details). This error is due to a real failure of a Flash cell, that can not be programmed anymore. When this error is detected, the embedded algorithm automatically exits the Page Update operation at the end of the Swap phase, without performing the Erase Phase 0 on the full sector. In this way the old data are kept, and through predefined routines in TestFlash (Find Wrong Pages = 230029h and Find Wrong Bytes = 23002Ch), the user can compare the old and the new data to find where the error occurred. Once the error has been discovered the user must take to end the stopped Erase Phase 0 on the old sector (through another predefined routine in TestFlash: Complete Swap = 23002Fh). The byte where the error occurred must be reprogrammed to FFh and then discarded, to avoid the error occurring again when that byte is internally moved. This bit is automatically cleared when bit FEERR of the FESR0 register is cleared by software. Bit 4:0 = Reserved. 51/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM 3.4 WRITE OPERATION EXAMPLE Each operation (both Flash and EEPROM) is activated by a sequence of instructions like the following: OR LD LD .. LD CR, #OPMASK ADD1, #DATA1 ADD2, #DATA2 ...., ...... ADDn, #DATAn OR CR, #80h ;Operation selection ;1st Add and Data ;2nd Add and Data ;nth Add and Data ;n range = (1 to 16) ;Operation start The first instruction is used to select the desired operation by setting its corresponding selection bit in the Control Register (FCR for Flash operations, ECR for EEPROM operations). The load instructions are used to set the addresses (in the Flash or in the EEPROM memory space) and the data to be modified. The last instruction is used to start the write operation, by setting the start bit (FWMS for Flash operations, EWMS for EEPROM operation) in the Control register. Once selected, but not yet started, one operation can be cancelled by resetting the operation selection bit. Any latched address and data will be reset. Warning: during the Flash Page Program or the EEPROM Page Update operation it is forbidden to change the page address: only the last page address is effectively kept and all programming will effect only that page. A summary of the available Flash and EEPROM write operations (including the Flash Write Operations on the EEPROM sectors when the EEPROM Hardware Emulation is not used) are shown in the following tables: Table 11. Flash Write Operations Operation Selection bit Addresses and Data Start bit Typical Duration Byte Program FBYTE 1 byte FWMS 10 µs Page Program FPAGE From 1 to 16 bytes FWMS 160 µs (16 bytes) Sector Erase FSECT From 1 to 4 sectors FWMS 1.5 s (1 sector) Sector Erase Suspend FSUSP None None 15 µs Chip Erase FCHIP None FWMS 3s Set Protection PROT From 1 to 4 bytes FWMS 40 µs (4 bytes) Table 12. EEPROM Write Operations Operation Selection bit Addresses and Data Start bit Typical Duration Page Update EPAGE From 1 to 16 bytes EWMS 30 ms Chip Erase ECHIP None EWMS 70 ms Table 13. Flash Write Operations on EEPROM sectors Operation Selection bit Addresses and Data Start bit Typical Duration Byte Program FBYTE 1 byte EWMS 10 µs Page Program FPAGE From 1 to 16 bytes EWMS 160 µs (16 bytes) Sector Erase FSECT 1 sector EWMS 70 ms Sector Erase Suspend FSUSP None None 15 µs 52/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM 3.5 EEPROM 3.5.1 Hardware EEPROM Emulation Note: This section provides general information only. Users do not have to be concerned with the hardware EEPROM emulation. The last 256 bytes of the two EEPROM dedicated sectors (229000h to 2290FFh for sector E0 and 22D000h to 22D0FFh for sector E1) are reserved for the Non Volatile pointers used for the hardware Emulation. When the EEPROM is directly addressed through the addresses 220000h to 2203FFh, a Hardware Emulation mechanism is automatically activated, so avoiding the user having to manage the Non Volatile pointers that are used to map the EEP- ROM inside the two dedicated Flash sectors E0 and E1. The structure of the hardware emulation is shown in Figure 25. Each one of the two EEPROM dedicated Flash sectors E0 and E1 is divided in 4 blocks of the same size of the EEPROM to emulate (1Kbyte max). Each one of the 4 blocks is then divided in up to 64 pages of 16 bytes, the size of the available RAM buffer. The RAM buffer is used internally to temporarily store the new content of the page to update, during the Page Program operation (both in Flash and in EEPROM). Figure 25. Segment 22h structure (Example for 128K Flash device). Flash/EEPROM sector F4/E0 Flash/EEPROM sector F5/E1 228000h 22C000h Page 0 - 16 byte Page 0 - 16 byte 229000h 22D000h HW emulated EEPROM 2203FFh Page 2 to 61 Page 62 - 16 byte Page 63 - 16 byte Page 0 - 16 byte Page 1 - 16 byte Page 2 to 61 Page 62 - 16 byte Page 63 - 16 byte Page 0 - 16 byte Page 1 - 16 byte Page 2 to 61 Page 62 - 16 byte Page 63 - 16 byte Page 0 - 16 byte Page 1 - 16 byte Page 2 to 61 Page 62 - 16 byte Page 63 - 16 byte Non Volatile Status 256 byte User Registers 220000h 1 Kbyte 1 Kbyte Page 62 - 16 byte Page 63 - 16 byte 64 pages 1 Kbyte 22CC00h Page 2 to 61 Non Volatile Status 256 byte 1 Kbyte 22C800h Page 2 to 61 Page 62 - 16 byte Page 63 - 16 byte Page 0 - 16 byte Page 1 - 16 byte 1 Kbyte 22C400h Page 2 to 61 Page 62 - 16 byte Page 63 - 16 byte Page 0 - 16 byte Page 1 - 16 byte Block 0 1 Kbyte Block 3 228C00h Page 62 - 16 byte Page 63 - 16 byte Page 0 - 16 byte Page 1 - 16 byte Block 1 1 Kbyte Block 2 228800h Page 2 to 61 Block 2 1 Kbyte Block 1 228400h Page 1 - 16 byte Block 3 1 Kbyte Block 0 Page 1 - 16 byte 224000h FCR, ECR, FESR1-0 - 4 byte RAM buffer Page buffer - 16 byte 53/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM EEPROM (Cont’d) 3.5.2 EEPROM Update Operation The update of the EEPROM content can be made by pages of 16 consecutive bytes. The Page Update operation allows up to 16 bytes to be loaded into the RAM buffer that replace the ones already contained in the specified address. Each time a Page Update operation is executed in the EEPROM, the RAM buffer content is programmed in the next free block relative to the specified page (the RAM buffer is previously automatically filled with old data for all the page addresses not selected for updating). If all the 4 blocks of the specified page in the current EEPROM sector are full, the page content is copied to the complementary sector, that becomes the new current one. After that the specified page has been copied to the next free block, one erase phase is executed on the complementary sector, if the 4 erase phases have not yet been executed. When the selected page is copied to the complementary sector, the remaining 63 pages are also copied to the first block of the new sector; then the first erase phase is executed on the previous full sector. All this is executed in a hidden manner, and the End Page Update Interrupt is generated only after the end of the complete operation. At Reset the two status pages are read in order to detect which is the sector that is currently mapping the EEPROM, and in which block each page is mapped. A system defined routine written in TestFlash is executed at reset, so that any previously aborted write operation is restarted and completed. Figure 26. Hardware Emulation Flow Emulation Flow Program selected Page from RAM buffer in next free block Reset Read Status Pages new sector ? Map EEPROM in current sector Yes No Write operation to complete ? No Yes Complete Write operation Update Status page Complementary sector erased ? No 1/4 erase of complementary sector Wait for Update commands Page Update Command Update Status Page End Page Update Interrupt (to Core) 54/320 9 Copy all other Pages into RAM buffer; then program them in next free block Yes ST92F120 - SINGLE VOLTAGE FLASH & EEPROM 3.6 PROTECTION STRATEGY The protection bits are stored in the last 4 locations of the TestFlash (from 231FFCh) (see Figure 27). All the available protections are forced active during reset, then in the initialisation phase they are read from the TestFlash. The protections are stored in 2 Non Volatile Registers. Other 2 Non Volatile Registers can be used as a password to re-enable test modes once they have been disabled. The protections can be programmed using the Set Protection operation (see Control Registers paragraph), that can be executed from all the internal or external memories except the Flash or TestFlash itself. Figure 27. Protection Map. Protection 231FFCh 231FFDh 231FFEh 231FFFh NVAPR NVWPR NVPWD0 NVPWD1 3.6.1 Non Volatile Registers The 4 Non Volatile Registers used to store the protection bits for the different protection features are one time programmable by the user, but they are erasable in test mode (if not disabled). Access to these registers is controlled by the protections related to the TestFlash where they are mapped. Since the code to program the Protection Registers cannot be fetched by the Flash or the TestFlash memories, this means that, once the APRO or APBR bits in the NVAPR register are programmed, it is no longer possible to modify any of the protection bits. For this reason the NV Password, if needed, must be set with the same Set Protection operation used to program these bits. For the same reason it is strongly advised to never program the WPBR bit in the NVWPR register, as this will prevent any further write access to the TestFlash, and consequently to the Protection Registers. NON VOLATILE ACCESS PROTECTION REGISTER (NVAPR) Address: 231FFCh - Read/Write Delivery value: 1111 1111 (FFh) 7 1 6 5 4 3 2 1 0 APRO APBR APEE APEX PWT2 PWT1 PWT0 Bit 7 = Reserved. Bit 6 = APRO: ROM access protection. This bit, if programmed at 0, disables any access (read/write) to operands mapped inside the Flash address space (EEPROM excluded), unless the current instruction is fetched from the TestFlash or from the Flash itself. 0: ROM protection on 1: ROM protection off Bit 5 = APBR: BootROM access protection. This bit, if programmed at 0, disables any access (read/write) to operands mapped inside the TestFlash address space, unless the current instruction is fetched from the TestFlash itself. 0: BootROM protection on 1: BootROM protection off Bit 4 = APEE: EEPROM access protection. This bit, if programmed at 0, disables any access (read/write) to operands mapped inside the EEPROM address space, unless the current instruction is fetched from the TestFlash or from the Flash, or from the EEPROM itself. 0: EEPROM protection on 1: EEPROM protection off Bit 3 = APEX: Access Protection from External memory. This bit, if programmed at 0, disables any access (read/write) to operands mapped inside the address space of one of the internal memories (TestFlash, Flash, EEPROM, RAM), if the current instruction is fetched from an external memory. 0: Protection from external memory on 1: Protection from external memory off 55/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM PROTECTION STRATEGY (Cont’d) Bit 2:0 = PWT[2:0]: Password Attempt 2-0. If the TMDIS bit in the NVWPR register (231FFDh) is programmed to 0, every time a Set Protection operation is executed with Program Addresses equal to NVPWD1-0 (231FFE-Fh), the two provided Program Data are compared with the NVPWD1-0 content; if there is not a match one of PWT2-0 bits is automatically programmed to 0: when these three bits are all programmed to 0 the test modes are disabled forever. In order to intentionally disable test modes forever, it is sufficient to set a random Password and then to make 3 wrong attempts to enter it. NON VOLATILE WRITE PROTECTION REGISTER (NVWPR) Address: 231FFDh - Read/Write Delivery value: 1111 1111 (FFh) 7 6 5 4 3 2 1 0 TMDI PWO WPB WPE WPRS WPRS WPRS WPRS S K R E 3 2 1 0 Bit 7 = TMDIS: Test mode disable (Read Only). This bit, if set to 1, allows to bypass all the protections in test mode. If programmed to 0, on the contrary, all the protections remain active also in test mode. The only way to enable the test modes if this bit is programmed to 0, is to execute the Set Protection operation with Program Addresses equal to NVPWD1-0 (231FFF-Eh) and Program Data matching with the content of NVPWD1-0. This bit is read only: it is automatically programmed to 0 when NVPWD1-0 are written. 0: Test mode disabled 1: Test mode enabled Bit 6 = PWOK: Password OK (Read Only). If the TMDIS bit is programmed to 0, when the Set Protection operation is executed with Program Addresses equal to NVPWD[1:0] and Program Data matching with NVPWD[1:0] content, the PWOK bit is automatically programmed to 0. When this bit is programmed to 0 TMDIS protection is bypassed and the test modes are enabled. 0: Password OK 1: Password not OK 56/320 9 Bit 5 = WPBR: BootROM Write Protection. This bit, if programmed at 0, disables any write access to the TestFlash address space. This protection cannot be temporarily disabled. 0: BootROM write protection on 1: BootROM write protection off Bit 4 = WPEE: EEPROM Write Protection . This bit, if programmed to 0, disables any write access to the EEPROM address space. This protection can be temporary disabled by executing the Set Protection operation and writing 1 into this bit. To restore the protection it needs to reset the micro or to execute another Set Protection operation and write 0 to this bit. 0: EEPROM write protection on 1: EEPROM write protection off Bit 3:0 = WPRS[3:0]: ROM Segments 3-0 Write Protection. These bits, if programmed to 0, disable any write access to the 4 Flash sectors address spaces. These protections can be temporary disabled by executing the Set Protection operation and writing 1 into these bits. To restore the protection it needs to reset the micro or to execute another Set Protection operation and write 0 into these bits. 0: ROM Segments 3-0 write protection on 1: ROM Segments 3-0 write protection off NON VOLATILE PASSWORD (NVPWD1-0) Address: 231FFF-231FFEh - Write Only Delivery value: 1111 1111 (FFh) 7 6 5 4 3 2 1 0 PWD PWD PWD PWD PWD PWD PWD PWD 7 6 5 4 3 2 1 0 Bit 7:0 = PWD[7:0]: Password bits 7:0 (Write Only). These bits must be programmed with the Non Volatile Password that must be provided with the Set Protection operation to reenable the test modes. These two registers can be accessed only in write mode and only once; when they are written (simultaneously with the same Set Protection operation), bit TMDIS of NVWPR (231FFDh) is simultaneously programmed and test modes are disabled. ST92F120 - SINGLE VOLTAGE FLASH & EEPROM 3.6.2 Temporary Unprotection On user request the memory can be configured so as to allow the temporary unprotection also of all access protections bits of NVAPR (write protection bits of NVWPR are always temporarily unprotectable). Bit APEX can be temporarily disabled by executing the Set Protection operation and writing 1 into this bit, but only if this write instruction is executed from an internal memory (Flash and Test Flash excluded). Bit APEE can be temporarily disabled by executing the Set Protection operation and writing 1 into this bit, but only if this write instruction is executed from the memory itself to unprotect (EEPROM). Bits APRO and APBR can be temporarily disabled through a direct write at NVAPR location, by overwriting at 1 these bits, but only if this write instruction is executed from the memory itself to unprotect. To restore the access protection bits it needs to reset the micro or to execute a Set Protection operation and write 0 into the desired bits. When an internal memory (Flash, TestFlash or EEPROM) is protected in access, also the data access through a DMA of a peripheral is forbidden (it returns FFh). To read data in DMA mode from a protected memory, first it is necessary to temporarily unprotect that memory. The temporary unprotection allows also to update a protected code. 3.7 FLASH IN-SYSTEM PROGRAMMING The Flash memory can be programmed in-system through a serial interface (SCI0). Exiting from reset, the ST9 executes the initialization from the BootROM code (written in TestFlash), where it checks the value of the SOUT0 pin. If it is at 0, this means that the user wishes to update the Flash code, otherwise normal execution continues. In this second case, the BootROM code reads the first two locations of the Flash memory (000000h-000001h) that represent the pointer to the start of the user code. If the Flash is virgin (read content is always FFh), its first two locations contain FFFFh. This will represent the last location of segment 0h, and it is interpreted by the BootROM code as a flag indicating that the Flash memory is virgin and needs to be programmed. If the value 1 is detected on the SOUT0 pin and the Flash is virgin, a HALT instruction is executed, waiting for a hardware Reset. 3.7.1 First Programming of a virgin Flash After checking that the SOUT0 pin is at 0, the Boot -ROM code enables the serial interface (typically an SCI) and writes in its Address Compare Register (ACR for SCI) a predefined address for recognition. The BootROM initializes the serial interface, including the interrupt vector table in the TestFlash itself, the Interrupt Vector Register for the serial interface (IVR for SCI), the mask bit to enable the address match interrupt (bit RXA of IMR for SCI). When the serial interface has received an address matching with the content of its Address Compare Register, an interrupt is generated and a predefined routine is executed located in TestFlash (Code Update Routine), that loads at a predefined address in the internal RAM a predefined number of bytes (the first datum sent) from the serial interface. These bytes must represent a routine (the in-system programming routine) which is called at the end of the transfer. This routine can, for example, load in the internal RAM (through the serial interface in DMA mode) a first table of data (256 bytes for example; depending on the available internal RAM size) to be programmed in Flash. Then the routine starts to program the Flash memory using the first table, while, in parallel, a second table of data are loaded in another location of the internal RAM, through the serial interface. When the slower of these two parallel operations is ended, a new cycle can start, till the whole Flash memory is programmed. At the end of Flash programming the execution returns to the Code Update Routine in TestFlash that puts the ST9 in HALT mode, waiting for a hardware reset. 57/320 9 ST92F120 - SINGLE VOLTAGE FLASH & EEPROM Figure 28. Flash in-system Programming. Internal RAM (User Code Example) TestFlash Code In-system prog routine Start Initialisation No Jump to Flash Main User Code SOUT0 =0? Address Match Interrupt (from SCI) Yes No Flash virgin ? Yes Erase sectors Enable Serial Interface WFI Test Flash Load 1st table of data in RAM through S.I. Code Update Routine Enable DMA Load in-system prog routine in internal RAM through SCI. Call in-system prog routine Prog 1st table of data from RAM in Flash Load 2nd table of data in RAM through SCI Inc. Address Last Address ? Yes RET HALT 58/320 9 No ST92F120 - REGISTER AND MEMORY MAP 4 REGISTER AND MEMORY MAP 4.1 INTRODUCTION The ST92F120 register map, memory map and peripheral options are documented in this section. Use this reference information to supplement the functional descriptions given elsewhere in this document. 4.2 MEMORY CONFIGURATION The Program memory space of the ST92F120 up to 128K bytes of directly addressable on-chip memory, is fully available to the user. The first 256 memory locations from address 0 to FFh hold the Reset Vector, the Top-Level (Pseudo Non-Maskable) interrupt, the Divide by Zero Trap Routine vector and, optionally, the interrupt vector table for use with the on-chip peripherals and the external interrupt sources. Apart from this case no other part of the Program memory has a predetermined function. Table 14. First 6 Bytes of Program Space 0 Address high of Power on Reset routine 1 2 3 4 5 Address Address Address Address Address low of Power on Reset routine high of Divide by zero trap Subroutine low of Divide by zero trap Subroutine high of Top Level Interrupt routine low of Top Level Interrupt routine 59/320 9 ST92F120 - REGISTER AND MEMORY MAP Figure 29. ST92F120JV1Q7/ST92F120V1Q7 User Memory Map (part 1) 01FFFFh SECTOR F3 8 Kbytes PAGE 7 - 16 Kbytes 01C000h 01BFFFh SEGMENT 1 64 Kbytes SECTOR F2 8 Kbytes PAGE 6 - 16 Kbytes 018000h 017FFFh PAGE 5 - 16 Kbytes 014000h 013FFFh PAGE 4 - 16 Kbytes 010000h 00FFFFh SECTOR F1 48 Kbytes PAGE 3 - 16 Kbytes 00C000h 00BFFFh SEGMENT 0 64 Kbytes SECTOR F0 64 Kbytes PAGE 2 - 16 Kbytes 008000h 007FFFh PAGE 1 - 16 Kbytes 004000h 003FFFh PAGE 0 - 16 Kbytes 000000h FLASH - 128 Kbytes 22FFFFh PAGE 91 - 16 Kbytes 22C000h 22BFFFh SEGMENT 22 64 Kbytes PAGE 90 - 16 Kbytes 228000h 227FFFh PAGE 89 - 16 Kbytes 224000h 223FFFh PAGE 88 - 16 Kbytes 220000h 2203FFh 1 Kbyte 220000h Emulated EEPROM - 1 Kbyte 22FFFFh PAGE 91 - 16 Kbytes 22C000h 22BFFFh SEGMENT 22 64 Kbytes PAGE 90 - 16 Kbytes 228000h 227FFFh PAGE 89 - 16 Kbytes 224000h 223FFFh PAGE 88 - 16 Kbytes 220000h 22CFFFh 4 Kbytes Sector F5/E1 22C000h 228FFFh 4 Kbytes Sector F4/E0 228000h FLASH / EEPROM - 4 + 4 Kbytes 60/320 9 Not Available ST92F120 - REGISTER AND MEMORY MAP Figure 30. ST92F120JV1Q7/ST92F120V1Q7 User Memory Map (part 2) 23FFFFh PAGE 95 - 16 Kbytes 23C0 00h 23BFFFh PAGE 94 - 16 Kbytes SEGMENT 23 64 Kbytes 23800 0h 237F FFh PAGE 93 - 16 Kbytes 23400 0h 233F FFh PAGE 92 - 16 Kbytes 23000 0h 231F FFh 8 Kbytes 23000 0h TESTFLASH - 8 Kbytes 231F FFh 128 bytes 231F 80h FLASH OTP - 128 bytes 231F FFh 4 bytes 231F FCh FLASH OTP Protection - 4 bytes 22FFFFh PAGE 91 - 16 Kbytes 22C0 00h 22BFFFh SEGMENT 22 64 Kbytes PAGE 90 - 16 Kbytes 22800 0h 227F FFh PAGE 89 - 16 Kbytes 22400 0h 223F FFh PAGE 88 - 16 Kbytes 22000 0h 224003h 4 bytes 22400 0h FLASH Registers - 4 bytes Not Available 61/320 9 ST92F120 - REGISTER AND MEMORY MAP Figure 31. ST92F120 User Memory Map (part 3) 20FFFFh PAGE 83 - 16 Kbytes 20C000h 20BFFFh PAGE 82 - 16 Kbytes SEGMENT 20 64 Kbytes 208000h 207FFFh PAGE 81 - 16 Kbytes 204000h 203FFFh PAGE 80 - 16 Kbytes 200000h 200FFFh 4 Kbytes 2 Kbytes 2007FFh 2005FFh 1.5 Kbytes 200000h Not Available RAM 4.3 ST92F120 REGISTER MAP Table 16 contain the map of the group F peripheral pages. The common registers used by each peripheral are listed in Table 15. 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 initialised. 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 15. Common Registers Function or Peripheral SCI, MFT A/D SPI, WDT, STIM I/O PORTS EXTERNAL INTERRUPT RCCU 62/320 9 Common Registers CICR + NICR + DMA REGISTERS + I/O PORT REGISTERS CICR + NICR + I/O PORT REGISTERS CICR + NICR + EXTERNAL INTERRUPT REGISTERS + I/O PORT REGISTERS I/O PORT REGISTERS + MODER INTERRUPT REGISTERS + I/O PORT REGISTERS INTERRUPT REGISTERS + MODER ST92F120 - REGISTER AND MEMORY MAP Table 16. Group F Pages Register Map Resources available on the ST92F120 device: Reg. Page 0 10 11 20 21 23 24 25 28 29 43 55 57 61 63 WUIMU Res. Res. Res. Res. Res. STIM SPI Port 4 R241 Port 0 R242 MFT0 Res. Res. RCCU Res. Res. R243 A/D 0 A/D 1 Port 8 EFT1 EFT0 SCI1 SCI0 MMU MFT1 Port 5 Port 1 EXT INT R246 JBLPD R247 I2C MFT1 R248 Res. MFT0 Res. MFT0 MFT1 Port 6 Port 2 WDT Res R249 R244 9 Port 9 Port 3 R251 R245 8 WCR R253 R250 7 Res R254 Res. R252 3 Port 7 R255 2 R240 63/320 9 ST92F120 - REGISTER AND MEMORY MAP Table 17. Detailed Register Map Page (Dec) Block Core N/A I/O Port 0:5 INT 0 WDT 2 64/320 9 Reg. No. Register Name Description Reset Value Hex. Doc. Page R230 CICR Central Interrupt Control Register 87 28 R231 FLAGR Flag Register 00 29 R232 RP0 Pointer 0 Register xx 31 R233 RP1 Pointer 1 Register xx 31 R234 PPR Page Pointer Register xx 33 R235 MODER Mode Register E0 33 R236 USPHR User Stack Pointer High Register xx 35 R237 USPLR User Stack Pointer Low Register xx 35 R238 SSPHR System Stack Pointer High Reg. xx 35 R239 SSPLR System Stack Pointer Low Reg. xx 35 R224 P0DR Port 0 Data Register FF R225 P1DR Port 1 Data Register FF R226 P2DR Port 2 Data Register FF R227 P3DR Port 3 Data Register FF R228 P4DR Port 4 Data Register FF R229 P5DR Port 5 Data Register FF R242 EITR External Interrupt Trigger Register 00 123 83 R243 EIPR External Interrupt Pending Reg. 00 84 R244 EIMR External Interrupt Mask-bit Reg. 00 84 R245 EIPLR External Interrupt Priority Level Reg. FF 84 R246 EIVR External Interrupt Vector Register x6 135 R247 NICR Nested Interrupt Control 00 85 R248 WDTHR Watchdog Timer High Register FF 134 R249 WDTLR Watchdog Timer Low Register FF 134 R250 WDTPR Watchdog Timer Prescaler Reg. FF 134 R251 WDTCR Watchdog Timer Control Register 12 134 135 R252 WCR Wait Control Register 7F I/O R240 P0C0 Port 0 Configuration Register 0 00 Port R241 P0C1 Port 0 Configuration Register 1 00 0 R242 P0C2 Port 0 Configuration Register 2 00 I/O R244 P1C0 Port 1 Configuration Register 0 00 Port R245 P1C1 Port 1 Configuration Register 1 00 1 R246 P1C2 Port 1 Configuration Register 2 00 I/O R248 P2C0 Port 2 Configuration Register 0 FF Port R249 P2C1 Port 2 Configuration Register 1 00 2 R250 P2C2 Port 2 Configuration Register 2 00 I/O R252 P3C0 Port 3 Configuration Register 0 FE Port R253 P3C1 Port 3 Configuration Register 1 00 3 R254 P3C2 Port 3 Configuration Register 2 00 123 ST92F120 - REGISTER AND MEMORY MAP Page (Dec) 3 Block Reg. No. Register Name Description Reset Value Hex. I/O R240 P4C0 Port 4 Configuration Register 0 FD 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 FF 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 3F R249 P6C1 Port 6 Configuration Register 1 00 R250 P6C2 Port 6 Configuration Register 2 00 R251 P6DR Port 6 Data Register FF R252 P7C0 Port 7 Configuration Register 0 FF R253 P7C1 Port 7 Configuration Register 1 00 R254 P7C2 Port 7 Configuration Register 2 00 R255 P7DR Port 7 Data Register FF R240 SPDR0 SPI0 Data Register 00 R241 SPCR0 SPI0 Control Register 00 219 R242 SPSR0 SPI0 Status Register 00 220 R243 SPPR0 SPI0 Prescaler Register 00 220 I/O Port 6 I/O Port 7 7 Doc. Page SPI 123 219 65/320 9 ST92F120 - REGISTER AND MEMORY MAP Page (Dec) Block 8 MFT1 9 MFT0,1 MFT0 10 66/320 9 Reg. No. Register Name Description Reset Value Hex. Doc. Page R240 REG0HR1 Capture Load Register 0 High xx 174 R241 REG0LR1 Capture Load Register 0 Low xx 174 R242 REG1HR1 Capture Load Register 1 High xx 174 R243 REG1LR1 Capture Load Register 1 Low xx 174 R244 CMP0HR1 Compare 0 Register High 00 174 R245 CMP0LR1 Compare 0 Register Low 00 174 R246 CMP1HR1 Compare 1 Register High 00 174 R247 CMP1LR1 Compare 1 Register Low 00 174 R248 TCR1 Timer Control Register 00 175 R249 TMR1 Timer Mode Register 00 176 R250 T_ICR1 External Input Control Register 00 177 R251 PRSR1 Prescaler Register 00 177 R252 OACR1 Output A Control Register 00 178 R253 OBCR1 Output B Control Register 00 179 R254 T_FLAGR1 Flags Register 00 179 R255 IDMR1 Interrupt/DMA Mask Register 00 181 R244 DCPR1 DMA Counter Pointer Register xx 174 R245 DAPR1 DMA Address Pointer Register xx 174 R246 T_IVR1 Interrupt Vector Register xx 174 R247 IDCR1 Interrupt/DMA Control Register C7 174 R248 IOCR I/O Connection Register FC 183 R240 DCPR0 DMA Counter Pointer Register xx 181 R241 DAPR0 DMA Address Pointer Register xx 182 R242 T_IVR0 Interrupt Vector Register xx 182 R243 IDCR0 Interrupt/DMA Control Register C7 183 R240 REG0HR0 Capture Load Register 0 High xx 174 R241 REG0LR0 Capture Load Register 0 Low xx 174 R242 REG1HR0 Capture Load Register 1 High xx 174 R243 REG1LR0 Capture Load Register 1 Low xx 174 R244 CMP0HR0 Compare 0 Register High 00 174 R245 CMP0LR0 Compare 0 Register Low 00 174 R246 CMP1HR0 Compare 1 Register High 00 174 R247 CMP1LR0 Compare 1 Register Low 00 174 R248 TCR0 Timer Control Register 00 175 R249 TMR0 Timer Mode Register 00 176 R250 T_ICR0 External Input Control Register 00 177 R251 PRSR0 Prescaler Register 00 177 R252 OACR0 Output A Control Register 00 178 R253 OBCR0 Output B Control Register 00 179 R254 T_FLAGR0 Flags Register 00 179 R255 IDMR0 Interrupt/DMA Mask Register 00 181 ST92F120 - REGISTER AND MEMORY MAP Page (Dec) 11 20 Block STIM I2C MMU 21 EXTMI Reg. No. Register Name Description Reset Value Hex. Doc. Page R240 STH Counter High Byte Register FF 139 R241 STL Counter Low Byte Register FF 139 R242 STP Standard Timer Prescaler Register FF 139 R243 STC Standard Timer Control Register 14 139 R240 I2DCCR I2C Control Register 00 232 R241 I2CSR1 I2C Status Register 1 00 233 R242 I2CSR2 I2C Status Register 2 00 235 R243 I2CCCR I2C Clock Control Register 00 236 R244 I2COAR1 I2C Own Address Register 1 00 236 R245 I2COAR2 I2C Own Address Register 2 00 237 R246 I2CDR R247 I2CADR I 2C Data Register I2C General Call Address 2 00 237 A0 237 R248 I2CISR I C Interrupt Status Register xx 238 R249 I2CIVR I2C Interrupt Vector Register xx 239 R250 I2CRDAP Receiver DMA Source Addr. Pointer xx 239 R251 I2CRDC Receiver DMA Transaction Counter xx 239 R252 I2CTDAP Transmitter DMA Source Addr. Pointer xx 240 R253 I2CTDC Transmitter DMA Transaction Counter xx 240 R254 I2CECCR Extended Clock Control Register 00 240 R255 I2CIMR I2C Interrupt Mask Register x0 241 R240 DPR0 Data Page Register 0 xx 40 R241 DPR1 Data Page Register 1 xx 40 R242 DPR2 Data Page Register 2 xx 40 R243 DPR3 Data Page Register 3 xx 40 R244 CSR Code Segment Register 00 41 R248 ISR Interrupt Segment Register xx 41 R249 DMASR DMA Segment Register xx 41 R245 EMR1 External Memory Register 1 80 120 R246 EMR2 External Memory Register 2 1F 121 67/320 9 ST92F120 - REGISTER AND MEMORY MAP Page (Dec) 23 24 68/320 9 Block JBLPD SCI0 Reg. No. Register Name Description Reset Value Hex. Doc. Page R240 STATUS Status Register 40 264 R241 TXDATA Transmit Data Register xx 265 R242 RXDATA Receive Data Register xx 266 R243 TXOP Transmit Opcode Register 00 266 R244 CLKSEL System Frequency Selection Register 00 271 R245 CONTROL Control Register 40 271 R246 PADDR Physiscal Address Register xx 272 R247 ERROR Error Register 00 273 R248 IVR Interrupt Vector Register xx 275 R249 PRLR Priority Level Register 10 275 R250 IMR Interrupt Mask Register 00 275 R251 OPTIONS Options and Register Group Selection 00 277 R252 CREG0 Current Register 0 xx 279 R253 CREG1 Current Register 1 xx 279 R254 CREG2 Current Register 2 xx 279 R255 CREG3 Current Register 4 xx 279 R240 RDCPR0 Receiver DMA Transaction Counter Pointer xx 199 R241 RDAPR0 Receiver DMA Source Address Pointer xx 199 R242 TDCPR0 Transmitter DMA Transaction Counter Pointer xx 199 R243 TDAPR0 Transmitter DMA Destination Address Pointer xx 199 R244 S_IVR0 Interrupt Vector Register xx 201 R245 ACR0 Address/Data Compare Register xx 201 R246 IMR0 Interrupt Mask Register x0 201 R247 S_ISR0 Interrupt Status Register xx 201 R248 RXBR0 Receive Buffer Register xx 203 R248 TXBR0 Transmitter Buffer Register xx 203 R249 IDPR0 Interrupt/DMA Priority Register xx 204 R250 CHCR0 Character Configuration Register xx 205 R251 CCR0 Clock Configuration Register 00 206 R252 BRGHR0 Baud Rate Generator High Reg. xx 207 R253 BRGLR0 Baud Rate Generator Low Register xx 207 R254 SICR0 Synchronous Input Control 03 207 R255 SOCR0 Synchronous Output Control 01 208 ST92F120 - REGISTER AND MEMORY MAP Page (Dec) 25 28 Block SCI1 EFT0 Reg. No. Register Name Description Reset Value Hex. Doc. Page R240 RDCPR1 Receiver DMA Transaction Counter Pointer xx 199 R241 RDAPR1 Receiver DMA Source Address Pointer xx 199 R242 TDCPR1 Transmitter DMA Transaction Counter Pointer xx 199 R243 TDAPR1 Transmitter DMA Destination Address Pointer xx 199 R244 S_IVR1 Interrupt Vector Register xx 201 R245 ACR1 Address/Data Compare Register xx 201 R246 IMR1 Interrupt Mask Register x0 201 R247 S_ISR1 Interrupt Status Register xx 201 R248 RXBR1 Receive Buffer Register xx 203 R248 TXBR1 Transmitter Buffer Register xx 203 R249 IDPR1 Interrupt/DMA Priority Register xx 204 R250 CHCR1 Character Configuration Register xx 205 R251 CCR1 Clock Configuration Register 00 206 R252 BRGHR1 Baud Rate Generator High Reg. xx 207 R253 BRGLR1 Baud Rate Generator Low Register xx 207 R254 SICR1 Synchronous Input Control 03 207 R255 SOCR1 Synchronous Output Control 01 208 R240 IC1HR0 Input Capture 1 High Register xx 152 R241 IC1LR0 Input Capture 1 Low Register xx 152 R242 IC2HR0 Input Capture 2 High Register xx 152 R243 IC2LR0 Input Capture 2 Low Register xx 152 R244 CHR0 Counter High Register FF 153 R245 CLR0 Counter Low Register FC 153 R246 ACHR0 Alternate Counter High Register FF 153 R247 ACLR0 Alternate Counter Low Register FC 153 R248 OC1HR0 Output Compare 1 High Register 80 154 R249 OC1LR0 Output Compare 1 Low Register 00 154 R250 OC2HR0 Output Compare 2 High Register 80 154 R251 OC2LR0 Output Compare 2 Low Register 00 154 R252 CR1_0 Control Register 1 00 156 R253 CR2_0 Control Register 2 00 156 R254 SR0 Status Register 00 156 R255 CR3_0 Control Register 3 00 156 69/320 9 ST92F120 - REGISTER AND MEMORY MAP Page (Dec) 29 Block EFT1 I/O Port 8 43 I/O Port 9 55 57 70/320 9 RCCU WUIMU Reg. No. Register Name Description Reset Value Hex. Doc. Page R240 IC1HR1 Input Capture 1 High Register xx 152 R241 IC1LR1 Input Capture 1 Low Register xx 152 R242 IC2HR1 Input Capture 2 High Register xx 152 R243 IC2LR1 Input Capture 2 Low Register xx 152 R244 CHR1 Counter High Register FF 153 R245 CLR1 Counter Low Register FC 153 R246 ACHR1 Alternate Counter High Register FF 153 R247 ACLR1 Alternate Counter Low Register FC 153 R248 OC1HR1 Output Compare 1 High Register 80 154 R249 OC1LR1 Output Compare 1 Low Register 00 154 R250 OC2HR1 Output Compare 2 High Register 80 154 R251 OC2LR1 Output Compare 2 Low Register 00 154 R252 CR1_1 Control Register 1 00 156 R253 CR2_1 Control Register 2 00 156 R254 SR1 Status Register 00 156 R255 CR3_1 Control Register 3 00 156 R248 P8C0 Port 8 Configuration Register 0 03 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 00 106 R242 CLK_FLAG Clock Flag Register 48, 28 or 08 107 R246 PLLCONF PLL Configuration Register xx 107 R249 WUCTRL Wake-Up Control Register 00 91 R250 WUMRH Wake-Up Mask Register High 00 92 R251 WUMRL Wake-Up Mask Register Low 00 92 R252 WUTRH Wake-Up Trigger Register High 00 93 R253 WUTRL Wake-Up Trigger Register Low 00 93 R254 WUPRH Wake-Up Pending Register High 00 93 R255 WUPRL Wake-Up Pending Register Low 00 93 123 ST92F120 - REGISTER AND MEMORY MAP Page (Dec) 61 63 Block A/D 1 A/D 0 Reg. No. Register Name Description Reset Value Hex. Doc. Page R240 D0R1 Channel 0 Data Register xx 288 R241 D1R1 Channel 1 Data Register xx 288 R242 D2R1 Channel 2 Data Register xx 288 R243 D3R1 Channel 3 Data Register xx 288 R244 D4R1 Channel 4 Data Register xx 288 R245 D5R1 Channel 5 Data Register xx 288 R246 D6R1 Channel 6 Data Register xx 288 R247 D7R1 Channel 7 Data Register xx 288 R248 LT6R1 Channel 6 Lower Threshold Reg. xx 289 R249 LT7R1 Channel 7 Lower Threshold Reg. xx 289 R250 UT6R1 Channel 6 Upper Threshold Reg. xx 289 R251 UT7R1 Channel 7 Upper Threshold Reg. xx 289 R252 CRR1 Compare Result Register 0F 289 R253 CLR1 Control Logic Register 00 290 R254 AD_ICR1 Interrupt Control Register 0F 291 R255 AD_IVR1 Interrupt Vector Register x2 291 R240 D0R0 Channel 0 Data Register xx 288 R241 D1R0 Channel 1 Data Register xx 288 R242 D2R0 Channel 2 Data Register xx 288 R243 D3R0 Channel 3 Data Register xx 288 R244 D4R0 Channel 4 Data Register xx 288 R245 D5R0 Channel 5 Data Register xx 288 R246 D6R0 Channel 6 Data Register xx 288 R247 D7R0 Channel 7 Data Register xx 288 R248 LT6R0 Channel 6 Lower Threshold Reg. xx 289 R249 LT7R0 Channel 7 Lower Threshold Reg. xx 289 R250 UT6R0 Channel 6 Upper Threshold Reg. xx 289 R251 UT7R0 Channel 7 Upper Threshold Reg. xx 289 R252 CRR0 Compare Result Register 0F 289 R253 CLR0 Control Logic Register 00 290 R254 AD_ICR0 Interrupt Control Register 0F 291 R255 AD_IVR0 Interrupt Vector Register x2 291 Note: xx denotes a byte with an undefined value, however some of the bits may have defined values. Refer to register description for details. 71/320 9 ST92F120 - INTERRUPTS 5 INTERRUPTS 5.1 INTRODUCTION 5.2 INTERRUPT VECTORING 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. Interrupt service routines are addressed through a vector table mapped in Memory. 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. 5.2.1 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 32. Interrupt Response n NORMAL PROGRAM FLOW INTERRUPT INTERRUPT SERVICE ROUTINE CLEAR PENDING BIT IRET INSTRUCTION VR001833 72/320 9 ST92F120 - INTERRUPTS 5.2.2 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 0 1 Mode ST9 Compatible ST9+ Pushed/Popp ed PC, FLAGR, PC, FLAGR Registers CSR Max. Code Size 64KB No limit for interrupt Within 1 segment Across segments service routine 5.3 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). 5.4 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. 5.4.1 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. 5.4.2 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. 5.4.3 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 73/320 9 ST92F120 - INTERRUPTS with the highest position in the chain, as shown in Table 18 Table 18. Daisy Chain Priority Highest Position INTA0 / Watchdog Timer INTA1 / Standard Timer INTB0 / Extended Function Timer 0 INTB1 / Extended Function Timer 1 INTC0 / EEPROM/Flash INTC1 / SPI INTD0 / RCCU INTD1 / WKUP MGT Multifunction Timer 0 JBLPD I2C bus Interface A/D Converter 0 A/D Converter 1 Multifunction Timer 1 Serial Communication Interface 0 Lowest Position Serial Communication Interface 1 5.4.4 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 33. Figure 33. 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 5.5 ARBITRATION MODES The ST9 provides two interrupt arbitration modes: Concurrent mode and Nested mode. Concurrent mode is the standard interrupt arbitration mode. 74/320 9 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. 5.5.1 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 ST92F120 - INTERRUPTS ARBITRATION MODES (Cont’d) re-enabled, they will be acknowledged regardless of the interrupt service routine’s priority. This may cause undesirable interrupt response sequences. 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 34) 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 34. 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 MAIN CPL = 7 75/320 9 ST92F120 - INTERRUPTS ARBITRATION MODES (Cont’d) Example 2 In the second example, (more complex, Figure 35), 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 35. 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 76/320 9 MAIN CPL = 7 ST92F120 - INTERRUPTS ARBITRATION MODES (Cont’d) 5.5.2 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 36. 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 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 77/320 9 ST92F120 - 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 36 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 37 contains a more complex example showing how nested mode allows nested interrupt processing (enabled inside the interrupt service routinesi using the ei instruction) according to their priority level. Figure 37. 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 6 ei INT5 MAIN CPL is set to 7 78/320 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 5 9 INTERRUPT 4 HAS PRIORI TY LEVEL 4 ei 4 7 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 ST92F120 - INTERRUPTS 5.6 EXTERNAL INTERRUPTS 5.6.1 Standard External Interrupts The standard ST9 core contains 8 external interrupts sources grouped into four pairs. Table 19. External Interrupt Channel Grouping External Channel Interrupt WKUP[0:15] P8[1:0] P7[7:5] P6[7,5] P5[7:5, 2:0] P4[7,4] INTD1 INT6 INT5 INT4 INT3 INT2 INT1 INT0 I/O Port Pin INTD0 INTC1 INTC0 INTB1 INTB0 INTA1 INTA0 P6.1 P6.3 P6.2 P6.3 P6.2 P6.0 P6.0 Each source has 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,..,7 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.7,..,0). See Figure 39. Figure 38. Priority Level Examples PL2D PL1D PL2 C PL1C PL2B PL1B PL2A PL1 A 1 SOURCE PRIORITY 0 0 0 1 0 0 1 EIPLR SOURCE PRIORITY INT.D0: 100=4 INT.D1: 101=5 INT.A0: 010=2 INT.A1: 011=3 INT.C0: 000=0 INT.C1: 001=1 INT.B0: 100=4 INT.B1: 101=5 VR00015 1 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. Figure 38 shows an example of priority levels. Figure 39 gives an overview of the external interrupts and vectors. Channel Internal Interrupt Source External Interrupt Source INTA0 Timer/Watchdog INT0 INTA1 Standard Timer INT1 INTB0 Extended Function Timer 0 INT2 INTB1 Extended Function Timer 1 INT3 INTC0 EEPROM/Flash INT4 INTC1 SPI Interrupt INT5 INTD0 RCCU INT6 INTD1 Wake-up Management Unit – The source of interrupt channel A0 can be selected between the external pin INT0 or the Timer/Watchdog peripheral using the IA0S bit in the EIVR register (R246 Page 0). – The source of interrupt channel A1 can be selected between the external pin INT1 or the Standard Timer using the INTS bit in the STC register (R232 Page 11). – The source of the interrupt channel B0 can be selected between the external pin INT2 or the on-chip Extended Function Timer 0 using the EFTIS bit in the CR3 register (R255 Page 28). – The source of interrupt channel B1 can be selected between external pin INT3 or the on-chip Extended Function Timer 1 using the EFTIS bit in the CR3 register (R255 Page 29). – The source of the interrupt channel C0 can be selected between external pin INT4 or the Onchip EEPROM/Flash Memory using bit FEIEN in the ECR register (Address 224001h). – The source of interrupt channel C1 can be selected between external pin INT5 or the on-chip SPI using the SPIS bit in the SPCR0 register (R241 Page 7). – The source of interrupt channel D0 can be selected between external pin INT6 or the Reset and Clock Unit RCCU using the INT_SEL bit in the CLKCTL register (R240 Page 55). – The source of interrupt channel D1 selected between the NMI pin and the WUIMU Wakeup/Interrupt Lines using the ID1S bit in the WUCRTL register (R248 Page 9). Warning: When using external interrupt channels shared by both external interrupts and peripherals, special care must be taken to configure control registers both for peripheral and interrupts. 79/320 9 ST92F120 - INTERRUPTS EXTERNAL INTERRUPTS (Cont’d) Figure 39. 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 X X 0 “1” Mask bit IMA0 0 X INT A0 request Pending bit IPA0 INTS TEA1 STD Timer INT 1 pin* “0” V7 V6 V5 V4 0 0 VECTOR Priority level X X 1 “1” Mask bit IMA1 1 X INT A1 request Pending bit IPA1 EFTI S TEB0 EFT 0 Timer INT 2 pin* “1” V7 V6 V5 V4 0 1 VECTO R Priority level X X 0 “0” Mask bit IMB0 0 X INT B0 request Pending bit IPB0 EFTIS TEB1 EFT1 Timer “0” INT 3 pin* “1” V7 V6 V5 V4 0 1 VECTOR Priority level X X 1 Mask bit IMB1 1 X INT B1 request Pending bit IPB1 FEIEN TEC0 EEPROM/Flas h “1” INT 4 pin* “0” V7 V6 V5 V4 1 0 VECTO R Priority level X X 0 Mask bit IMC0 0 X INT C0 request Pending bit IPC0 SPIS TEC1 SPI Interrupt “1” INT 5 pin* “0” RCCU V7 V6 V5 V4 1 0 VECTOR Priority level X X 1 Mask bit IMC1 1 X INT C1 request Pending bit IPC1 INT_SEL TED0 INT 6 pin “1” V7 V6 V5 V4 1 1 VECTOR Priority level X X 0 “0” Mask bit IMD0 0 X INT D0 request Pending bit IPD0 ID1S NMI Wake-up Controller “1” V7 V6 V5 V4 1 1 VECTOR Priority level X X 1 “0” Mask bit IMD1 1 X Pending bit IPD1 WKUP (0:15) * Only four interrupt pins are available at the same time. Refer to Table 19 for I/O pin mapping. 80/320 9 INT D1 request ST92F120 - INTERRUPTS 5.7 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) must also be enabled in order to allow a Top Level Request. The second mask NICR.TLNM (R247.7) is a setonly 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. Furthermore the TLI never modifies the CPL bits and the NICR register. 5.8 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 40. Top Level Interrupt Structure n WATCHDOG ENABLE WDGEN CORE RESET TLIP WATCHDOG TIMER END OF COUNT MUX PENDING MASK NMI TOP LEVEL INTERRUPT REQUEST OR TLIS TLTEV TLNM TLI IEN VA00294 n 81/320 9 ST92F120 - INTERRUPTS 5.9 INTERRUPT RESPONSE TIME The interrupt arbitration protocol functions completely asynchronously from instruction flow and requires 5 clock cycles. One more CPUCLK cycle is required when an interrupt is acknowledged. 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 55 clock cycles (DIV instruction), 53 clock 82/320 9 cycles (DIVWS and MUL instructions) or 49 for other instructions. 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 51 clock cycles (DIV instruction), 49 clock cycles (DIVWS and MUL instructions) or 45 for other instructions. 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. ST92F120 - INTERRUPTS 5.10 INTERRUPT REGISTERS CENTRAL INTERRUPT CONTROL REGISTER (CICR) R230 - Read/Write Register Group: System Reset value: 1000 0111 (87h) 7 GCEN TLIP 0 TLI IEN IAM the IEN bit when interrupts are disabled or when no peripheral can generate interrupts. For example, 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. 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: A Top Level Interrupt is generared when TLIP is set, only if TLNM=1 in the NICR register (independently of the value of the IEN bit). 1: A Top Level Interrupt request is generated when IEN=1 and the TLIP bit are set. 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 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 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 83/320 9 ST92F120 - 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 84/320 9 LEVEL Hardware bit 0 1 0 1 0 1 0 1 Priority 0 (Highest) 1 2 3 4 5 6 7 (Lowest) ST92F120 - INTERRUPTS INTERRUPT REGISTERS (Cont’d) EXTERNAL INTERRUPT VECTOR REGISTER (EIVR) R246 - Read/Write Register Page: 0 Reset value: xxxx 0110b (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 39. 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. 85/320 9 ST92F120 - INTERRUPTS 5.11 WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (WUIMU) 5.11.1 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 (when available) 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 41). Figure 41. Wake-Up Lines / Interrupt Management NMI WKUP[ 7:0] 5.11.2 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 WUCTR L 1 STOP ID1S WKUP-INT Set Reset SW SETT ING1) 0 TO CPU INTD1 - External Interrupt Channel TO CPU TO RCCU - Stop Mode Control Note 1: The reset signal on the Stop bit is stronger than the set signal. 86/320 9 ST92F120 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) able the INT7 external interrupt source and 5.11.3 Functional Description enable the 16 wake-up lines as external inter5.11.3.1 Interrupt Mode rupt source lines. This is not mandatory if the To configure the 16 wake-up lines as interrupt wake-up event does not require an interrupt sources, use the following procedure: response. 1. Configure the mask bits of the 16 wake-up lines 7. Write the sequence 1,0,1 to the STOP bit of the (WUMRL, WUMRH) WUCTRL register with three consecutive write operations. This is the STOP bit setting 2. Configure the triggering edge registers of the sequence. wake-up lines (WUTRL, WUTRH) To detect if STOP Mode was entered or not, im3. Set bit 7 of EIMR (R244 Page 0) and EITR mediately after the STOP bit setting sequence, (R242 Page 0) registers of the CPU: so an poll the RCCU EX_STP bit (R242.7, Page 55) and interrupt coming from one of the 16 lines can be the STOP bit itself. correctly acknowledged 4. Reset the WKUP-INT bit in the WUCTRL register to disable Wake-up Mode 5.11.3.3 STOP Mode Entry Conditions 5. Set the ID1S bit in the WUCTRL register to disAssuming the ST9 is in Run mode: during the able the INT7 external interrupt source and STOP bit setting sequence the following cases enable the 16 wake-up lines as external intermay occur: rupt source lines. Case 1: NMI = 0, wrong STOP bit setting seTo return to standard mode (INT7 external interquence rupt source enabled and 16 wake-up lines disaThis can happen if an Interrupt/DMA request is acbled) it is sufficient to reset the ID1S bit. knowledged during the STOP bit setting sequence. In this case polling the STOP and EX_STP bits will give: 5.11.3.2 Wake-up Mode Selection STOP = 0, EX_STP = 0 To configure the 16 lines as wake-up sources, use the following procedure: This means that the ST9 did not enter STOP mode due to a bad STOP bit setting sequence: the user 1. Configure the mask bits of the 16 wake-up lines must retry the sequence. (WUMRL, WUMRH). Case 2: NMI = 0, correct STOP bit setting se2. Configure the triggering edge registers of the quence wake-up lines (WUTRL, WUTRH). In this case the ST9 enters STOP mode. There are 3. Set, as for Interrupt Mode selection, bit 7 of two ways to exit STOP mode: EIMR and EITR registers only if an interrupt routine is to be executed after a wake-up event. 1. A wake-up interrupt (not an NMI interrupt) is Otherwise, if the wake-up event only restarts acknowledged. That implies: the execution of the code from where it was STOP = 0, EX_STP = 1 stopped, the INTD1 interrupt channel must be masked or the external source must be This means that the ST9 entered and exited STOP selected by resetting the ID1S bit. mode due to an external wake-up line event. 4. Since the RCCU can generate an interrupt 2. A NMI rising edge woke up the ST9. This request when exiting from STOP mode, take implies: care to mask it even if the wake-up event is STOP = 1, EX_STP = 1 only to restart code execution. This means that the ST9 entered and exited STOP 5. Set the WKUP-INT bit in the WUCTRL register mode due to an NMI (rising edge) event. The user to select Wake-up Mode should clear the STOP bit via software. 6. Set the ID1S bit in the WUCTRL register to dis- 87/320 9 ST92F120 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) STOP = 0, EX_STP = 0 Case 3: NMI = 1 (NMI kept high during the 3rd write instruction of the sequence), bad STOP The application can determine why the ST9 did bit setting sequence not enter STOP mode by polling the pending The result is the same as Case 1: bits of the external lines (at least one must be at STOP = 0, EX_STP = 0 1). 2. Interrupt requests to CPU are enabled: in this This means that the ST9 did not enter STOP mode case the ST9 will not enter STOP mode and the due to a bad STOP bit setting sequence: the user interrupt service routine will be executed. The must retry the sequence. status of STOP and EX_STP bits will be again: Case 4: NMI = 1 (NMI kept high during the 3rd STOP = 0, EX_STP = 0 write instruction of the sequence), correct STOP bit setting sequence The interrupt service routine can determine why In this case: the ST9 did not enter STOP mode by polling the pending bits of the external lines (at least STOP = 1, EX_STP = 0 one must be at 1). This means that the ST9 did not enter STOP mode due to NMI being kept high. The user should clear If the MCU really exits from STOP Mode, the the STOP bit via software. RCCU EX_STP bit is still set and must be reset by Note: If NMI goes to 0 before resetting the STOP software. Otherwise, if NMI was high or an Interbit, the ST9 will not enter STOP mode. rupt/DMA request was acknowledged during the Case 5: A rising edge on the NMI pin occurs STOP bit setting sequence, the RCCU EX_STP bit during the STOP bit setting sequence. is reset. This means that the MCU has filtered the STOP Mode entry request. The NMI interrupt will be acknowledged and the ST9 will not enter STOP mode. This implies: The WKUP-INT bit can be used by an interrupt routine to detect and to distinguish events coming STOP = 0, EX_STP = 0 from Interrupt Mode or from Wake-up Mode, allowThis means that the ST9 did not enter STOP mode ing the code to execute different procedures. due to an NMI interrupt serviced during the STOP To exit STOP mode, it is sufficient that one of the bit setting sequence. At the end of NMI routine, the 16 wake-up lines (not masked) generates an user must re-enter the sequence: if NMI is still high event: the clock restarts after the delay needed for at the end of the sequence, the ST9 can not enter the oscillator to restart. STOP mode (See “NMI Pin Management” on page 89.). The same effect is obtained when a rising edge is detected on the NMI pin, which works as a 17th Case 6: A wake-up event on the external wakewake-up line. up lines occurs during the STOP bit setting sequence Note: After exiting from STOP Mode, the software can successfully reset the pending bits (edge senThere are two possible cases: sitive), even though the corresponding wake-up 1. Interrupt requests to the CPU are disabled: in line is still active (high or low, depending on the this case the ST9 will not enter STOP mode, no Trigger Event register programming); the user interrupt service routine will be executed and must poll the external pin status to detect and disthe program execution continues from the tinguish a short event from a long one (for example instruction following the STOP bit setting keyboard input with keystrokes of varying length). sequence. The status of STOP and EX_STP bits will be again: 88/320 9 ST92F120 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) 5.11.3.4 NMI Pin Management – If the ST9 is in Run mode and a rising edge occurs on the NMI pin: the NMI service routine is On the CPU side, if TLTEV=1 (Top Level Trigger executed and then the ST9 restarts the execuEvent, bit 3 of register R246, page 0) then a rising tion of the main program. Now, suppose that edge on the NMI pin will set the TLIP bit (Top Level the user wants to enter STOP mode with NMI Interrupt Pending bit, R230.6). At this point an instill at 1. The ST9 will not enter STOP mode terrupt request to the CPU is given either if TLand it will not execute an NMI routine beNM=1 (Top Level Not Maskable bit, R247.7 - once cause there were no transitions on the exterset it can only be cleared by RESET) or if TLI=1 nal NMI line. and IEN=1 (bits R230.5, R230.4). – If the ST9 is in run mode and a rising edge on Assuming that the application uses a non-maskaNMI pin occurs during the STOP bit setting seble Top Level Interrupt (TLNM=1): in this case, quence: the NMI interrupt will be acknowledged whenever a rising edge occurs on the NMI pin, the and the ST9 will not enter STOP mode. At the related service routine will be executed. To service end of the NMI routine, the user must re-enter further Top Level Interrupt Requests, it is necesthe sequence: if NMI is still high at the end of the sary to generate a new rising edge on the external sequence, the ST9 can not enter STOP mode NMI pin. (see previous case). The following summarizes some typical cases: – If the ST9 is in run mode and the NMI pin is high: – If the ST9 is in STOP mode and a rising edge on if NMI is forced low just before the third write inthe NMI pin occurs, the ST9 will exit STOP struction of the STOP bit setting sequence then mode and the NMI service routine will be exethe ST9 will enter STOP mode. cuted. 89/320 9 ST92F120 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) 9. Poll the wake-up pending bits to determine 5.11.4 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. 5.11.4.1 Procedure for Entering/Exiting STOP 5.11.4.2 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 87. 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. 90/320 9 ST92F120 - INTERRUPTS WAKE-UP / INTERRUPT LINES MANAGEMENT UNIT (Cont’d) low, the ST9 will enter STOP mode independently 5.11.5 Register Description of the status of the STOP bit. WAKE-UP CONTROL REGISTER (WUCTRL) WARNINGS: R249 - Read/Write Register Page: 57 – Writing the sequence 1,0,1 to the STOP bit will Reset Value: 0000 0000 (00h) enter STOP mode only if no other register write 7 0 instructions are executed during the sequence. If Interrupt or DMA requests (which always perform STOP ID1S WKUP- INT register write operations) are acknowledged during the sequence, the ST9 will not enter STOP mode: the user must re-enter the sequence to Bit 2 = STOP: Stop bit. set the STOP bit. To enter STOP Mode, write the sequence 1,0,1 to – Whenever a STOP request is issued to the MCU, this bit with three consecutive write operations. a few clock cycles are needed to enter STOP When a correct sequence is recognized, the mode (see RCCU chapter for further details). STOP bit is set and the RCCU puts the MCU in Hence the execution of the instruction following STOP Mode. The software sequence succeeds the STOP bit setting sequence might start before only if the following conditions are true: entering STOP mode: if such instruction per– The NMI pin is kept low, forms a register write operation, the ST9 will not enter in STOP mode. In order to avoid to execute – The WKUP-INT bit is 1, register write instructions after a correct STOP – All unmasked pending bits are reset bit setting sequence and before entering the – At least one mask bit is equal to 1 (at least one STOP mode, it is mandatory to execute 3 NOP external wake-up line is not masked). instructions after the STOP bit setting sequence. Otherwise the MCU cannot enter STOP mode, the program code continues executing and the STOP Bit 1 = ID1S: Interrupt Channel INTD1 Source. bit remains cleared. This bit is set and cleared by software. The bit is reset by hardware if, while the MCU is in 0: INT7 external interrupt source selected, excludSTOP mode, a wake-up interrupt comes from any ing wake-up line interrupt requests of the unmasked wake-up lines. The bit is kept 1: The 16 external wake-up lines enabled as interhigh if, during STOP mode, a rising edge on NMI rupt sources, replacing the INT7 external pin pin wakes up the ST9. In this case the user should function reset it by software. The STOP bit is at 1 in the four WARNING: To avoid spurious interrupt requests following cases (See “Wake-up Mode Selection” on the INTD1 channel due to changing the interon page 87. for details): rupt source, use this procedure to modify the ID1S – After the first write instruction of the sequence (a bit: 1 is written to the STOP bit) 1. Mask the INTD1 interrupt channel (bit 7 of reg– At the end of a successful sequence (i.e. after ister EIMR - R244, Page 0 - reset to 0). the third write instruction of the sequence) 2. Program the ID1S bit as needed. – The ST9 entered and exited STOP mode due to 3. Clear the IPD1 interrupt pending bit (bit 7 of a rising edge on the NMI pin. In this case the register EIPR - R243, Page 0) EX_STP bit in the CLK_FLAG is at 1 (see 4. Remove the mask on INTD1 (bit EIMR.7=1). RCCU chapter). – The ST9 did not enter STOP mode due to the NMI pin being kept high. In this case RCCU bit EX_STP is at 0 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 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 91/320 9 ST92F120 - 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). 92/320 9 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). ST92F120 - 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 WUTRLH 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. 93/320 9 ST92F120 - ON-CHIP DIRECT MEMORY ACCESS (DMA) 6 ON-CHIP DIRECT MEMORY ACCESS (DMA) 6.1 INTRODUCTION 6.2 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 42. 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 94/320 9 ST92F120 - ON-CHIP DIRECT MEMORY ACCESS (DMA) 6.3 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 43), 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 44). 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 43. 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 95/320 9 ST92F120 - 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 44. 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 96/320 9 MEMORY VECTOR TABLE ST92F120 - ON-CHIP DIRECT MEMORY ACCESS (DMA) DMA TRANSACTIONS (Cont’d) 6.4 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. 6.5 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 97/320 9 ST92F120 - ON-CHIP DIRECT MEMORY ACCESS (DMA) 6.6 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 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 98/320 9 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 6.2 DMA PRIORITY LEVELS for a description of priority levels. PRL2 0 0 0 0 1 1 1 1 PRL1 0 0 1 1 0 0 1 1 PRL0 0 1 0 1 0 1 0 1 Source Priority Level 0 Highest 1 2 3 4 5 6 7 Lowest DMA ADDRESS POINTER REGISTER (DAPR) Read/Write Address set by Peripheral Reset value: undefined 7 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). ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) 7 RESET AND CLOCK CONTROL UNIT (RCCU) 7.1 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. 7.2 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 51 and Figure 53). A low frequency external clock may be connected to the CK_AF pin, and this clock source may be selected when low power operation is required. 7.2.1 Clock Control Unit Overview As shown in Figure 45 a programmable divider can divide the CLOCK1 input clock signal by two. In practice, the divide-by-two is virtually always used in order to ensure a 50% duty cycle signal to the PLL multiplier circuit. 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 programmable 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. 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, a low frequency external clock connected to the CK_AF pin may be selected, whereupon the crystal controlled main oscillator may be turned off. 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). (See Figure 45) 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. 99/320 9 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) Figure 45. ST92F120 Clock Distribution Diagram 3-bit Prescaler 1...8 Baud Rate Generator 1/N SCK Master N=2,4,16,32 Conversion time 138 * INTCLK A/Dx LOGIC SPI Baud Rate Generator 1/N 16-bit Up Counter N=2,4,8 1/N N=2...( 216-1) EXTCLKx (Max INTCLK/4) EFTx SCIx 1...64 6-bit Prescaler 8-bit Prescaler MFTx 16-bit Up/Down Counter 1...2 56 1/3 TxINA/TxINB (Max INTCLK/3) J1850 Kernel SCK Slave (Max INTCLK/2) 1/2 JBLPD 16-bit Down Counter WDG 8-bit Prescaler 1...256 N=4,6, 8...258 STD 1/4 1/N FAST Fscl ≤100 kHz Fscl > 100 kHz Fscl ≤ 400 kHz N=6,9,12...387 1/N I 2C P6.5 1.. .8 8-bit Prescaler STIM 16-bit Down Counter 1... 256 1/4 CPUCLK 3-bit Prescaler CPU 1/16 P9.6 P6.0 CK_128 1/8 1/4 XT_DIV16 0 CSU_CKSEL Quartz Oscillator MX(1:0) 0 1/2 1 CLOCK2 PLL x 6/8/10/14 100/320 9 EPROM FLASH 0 1/ N DX(2:0) RCCU 1 RAM 1/16 DIV2 EMBEDDED MEMORY INTCLK 1 EEPROM ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) 7.3 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 46. Clock Control Unit Programming n XTSTOP DIV2 CSU_CKSEL CKAF_SEL (CLK_FLAG) (MODER) (CLK_FLAG) (CLKCTL) 1/16 1/4 CLK_128 0 0 Quartz oscillator CK_AF source 1/2 1 CLOCK2 0 PLL x 6/8/10/14 1/N MX(1:0) DX(2:0) 1 1 CLOCK1 0 INTCLK 1 to Peripherals CPU Clock Prescaler and P6.5 CK_AF (available only if mapped on ext. pin) (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. 101/320 9 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) CLOCK MANAGEMENT (Cont’d) 7.3.1 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. Since the input clock to the Clock Multiplier circuit requires a 50% duty cycle for correct PLL operation, the divide by two circuit should be enabled when a crystal oscillator is used, or when the external clock generator does not provide a 50% duty cycle. In practice, the divide-by-two is virtually always used in order to ensure a 50% duty cycle signal to the PLL multiplier circuit. 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. To select the multiplier clock, set the CSU_CKSEL bit in the CLK_FLAG Register after allowing a stabilisation period for the PLL. 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, as reported in Electrical Characteristics section. The ST9 being a static machine, there is no lower limit for INTCLK. However, some peripherals have their own minimum internal clock frequency limit below which the functionality is not guaranteed. 7.3.2 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. 102/320 9 The internal peripherals are not affected by the CPUCLK prescaler and continue to operate at the full INTCLK frequency. This is particularly useful 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 47, 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 47. CPU Clock Prescaling n INTCLK PRS VALUE 000 001 010 011 CPUCLK 100 101 110 111 VA00260 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) CLOCK MANAGEMENT (Cont’d) 7.3.3 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. 7.3.4 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 is stopped, thus suspending program execution, while the clock to the peripherals may be programmed as described in the following paragraphs. Two examples of Low Power operation in WFI are illustrated in Figure 48 and Figure 49. Providing that low power operation during Wait for Interrupt is enabled (by setting the LPOWFI bit 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 the external low frequency clock, CK_AF, if this has been selected by setting WFI_CKSEL, and providing CKAF_ST is set, thus indicating that the external clock is selected and actually present on the CK_AF pin. 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 crys- tal 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. 7.3.5 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. Note that this is the only case in the ST9 where an interrupt is generated with a high to low transition. Table 21. 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 1 103/320 9 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) Figure 48. Example of Low Power mode programming in WFI using CK_AF external clock n INTCLK FREQUENCY PROGRAM FLOW FXtal = 4 MHz, VDD = 4.5 V min Begin Reset State MX(1:0) ← 00 PLL multiply factor set to 10 DX2-0 ← 000 Divider factor set to 1, and PLL turned ON WAIT 2 MHz Wait for the PLL to lock TPLK* CSU_CKSEL ← 1 PLL is system clock source WFI_CKSEL ← 1 CK_AF clock selected in WFI state XTSTOP ← 1 Preselect Xtal stopped when CK_AF selected LPOWFI ← 1 Low Power Mode enabled in WFI state 20 MHz User’s Program WFI instruction Interrupt WFI status Interrupt Routine XTSTOP ← 0 WAIT CKAF_SEL ← 0 Wait For Interrupt activated CK_AF selected and Xtal stopped automatically No code is executed until an interrupt is requested Interrupt serviced while CK_AF is the System Clock and the Xtal restarts FCK_AF Wait for the Xtal to stabilise The System Clock switches to Xtal WAIT Wait for the PLL to lock CSU_CKSEL ← 1 PLL is System Clock source TSTUP** 2 MHz User’s Program Execution of user program resumes at full speed 20 MHz * TPLK = PLL lock-in time ** TSTUP = Quartz oscillator start-up time 104/320 9 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) Figure 49. Example of Low Power mode programming in WFI using CLOCK2/16 n INTCLK FREQUENCY PROGRAM FLOW FXtal = 4 MHz, VDD = 2.7 V min Begin Reset State MX(1:0) ← 01 PLL multiply factor set to 6 DX2-0 ← 000 Divider factor set to 1, and PLL turned ON WAIT Wait for the PLL to lock CSU_CKSEL ← 1 PLL is system clock source LPOWFI ← 1 Low Power Mode enabled in WFI state 2 MHz T PLK* User’s Program WFI instruction WFI status 12 MHz Wait For Interrupt activated CLOCK2/16 selected and PLL stopped automatically No code is executed until an interrupt is requested Interrupt 125 KHz Interrupt Routine Interrupt serviced PLL switched on CLOCK2 selected WAIT CSU_CKSEL ← 1 Wait for the PLL to lock T PLK* 2 MHz PLL is system clock source User’s Program Execution of user program resumes at full speed 12 MHz * TPLK = PLL lock-in time 105/320 9 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) 7.4 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_SE L 0 0 0 0 SRESEN CKAF_S WFI_CKS LPOWEL 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: The external interrupt channel input signal is selected (Reset state) 1: Select the internal RCCU interrupt as the source of the interrupt request 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 4:6 = Reserved for test purposes 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 clock is present on the CK_AF pin, 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 106/320 9 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) CLOCK CONTROL REGISTERS (Cont’d) CLOCK FLAG REGISTER (CLK_FLAG) R242 -Read/Write Register Page: 55 Reset Value: 01001000 after a Watchdog Reset Reset Value: 00101000 after a Software Reset Reset Value: 00001000 after a Power-On Reset WARNING: If this register is accessed with a logical instruction, such as AND or OR, some bits may not be set as expected. 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. 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 (TSTUP refer to Electrical Characteristics section). Bit 2 = CKAF_ST: (Read Only) If set, indicates that the alternate function clock has been selected. If no clock signal is present on the CK_AF pin, the selection will not occur. If reset, the PLL clock, CLOCK2 or CLOCK2/16 is selected (depending on bit 0). 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 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 7 EX_ STP WDGRES SOFTRES XTSTOP 0 XT_ CSU_ CKAF_ST LOCK DIV16 CKSEL 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). Bit 0 = CSU_CKSEL: CSU Clock Select. This bit is set and cleared by software. It is 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 (see Figure 46). 107/320 9 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) CLOCK CONTROL REGISTERS (Cont’d) PLL CONFIGURATION REGISTER (PLLCONF) R246 - Read/Write Register Page: 55 Reset Value: xx00x111 (xxh) 7 Table 22. PLL Multiplication Factors MX1 MX0 CLOCK2 x 1 0 1 0 0 0 1 1 14 10 8 6 0 - - MX1 MX0 - DX2 DX1 DX0 Table 23. PLL Divider Factors Bit 5:4 = MX[1:0]: PLL Multiplication Factor. Refer to Table 22 for multiplier settings. WARNING: After these bits are modified, take care that the PLL lock-in time has elapsed before setting the CSU_CKSEL bit in the CLK_FLAG register. Bit 2:0 = DX[2:0]: PLL output clock divider factor. Refer to Table 23 for divider settings. . DX2 DX1 DX0 CK 0 0 0 0 1 1 1 0 0 1 1 0 0 1 0 1 0 1 0 1 0 1 1 1 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 50. RCCU General Timing User program execution Boot ROM execution 20µs < 4µs PLL switched on by user PLL selected by user Reset phase RESET Internal Reset CLOCK2 PLL Multiplier Clock Internal Reset INTCLK TRSPH PLL Lock-in Time Exit from RESET 108/320 9 VR02113B ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) 7.5 OSCILLATOR CHARACTERISTICS The on-chip oscillator circuit uses an inverting gate circuit with tri-state output. OSCOUT must not be 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, having a a minimum duration of 20µs, as illustrated in Figure 56. 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. Figure 51. Crystal Oscillator Table 24. RS Crystal Specification C1 =C2 Freq. 5 MHz 4 MHz 3 MHz 33pF 22pF 80 ohm 130 ohm 250 ohm 130 ohm 200 ohm - Legend: C1, C2: Maximum Total Capacitances on pins OSCIN and OSCOUT (the value includes the external capacitance tied to the pin plus the parasitic capacitance of the board and of the device) Note: The tables are relative to the fundamental quartz crystal only (not ceramic resonator). Figure 52. Internal Oscillator Schematic HALT REF OSCOUT CRYSTA L CLOCK ST9 OSCIN CLOCK INPUT BUFFER OSCIN OSCOUT VR02086A Figure 53. External Clock n C2 C1 EXTERNAL CLOCK 1 MΩ * *Recommended for oscillator stability VR02116A 3.3 kΩ OSCOUT ST9 INPUT CLOCK OSCIN VR02116B 109/320 9 ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) CERAMIC RESONATORS Murata Electronics CERALOCK resonators have been tested with the ST92F120 at 3, 3.68, 4 and 5 MHz. Some resonators have built-in capacitors (see Table 25). The test circuit is shown in Figure 54. Figure 54. Test circuit ST92F120 VDD OSCIN VSS OSCOUT Rp CERALOCK V1 Rd V2 C1 C2 Table 25 shows the recommended conditions at different frequencies. Table 25. Obtained Results Freq. (MHz) 3 3.68 4 5 Parts Number C1 (PF) C2 (PF) Rp (Ohm) Rd (Ohm) CSA3.00MG CST3.00MGW CSA3.68MG CST3.68MGW CSTCC3.68MG CSA4.00MG CST4.00MGW CSTCC4.00MG CSA5.00MG CST5.00MGW CSTCC5.00MG 30 (30) 30 (30) (15) 30 (30) (15) 30 (30) (15) 30 (30) 30 (30) (15) 30 (30) (15) 30 (30) (15) Open Open Open Open Open Open Open Open Open Open Open 0 0 0 0 0 0 0 0 0 0 0 Advantages of using ceramic resonators: CST and CSTCC types have built-in loading capacitors (those with values shown in parentheses ()). Rp is always open in the previous table because there is no need for a parallel resistor with a resonator (it is needed only with a crystal). 110/320 9 Test conditions: The evaluation conditions are 4.5 to 5.5 V for the supply voltage and -40° to 105° C for the temperature range. Caution: The above circuit condition is for design reference only. Recommended C1, C2 value depends on the circuit board used. ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) 7.6 RESET/STOP MANAGER The Reset/Stop Manager resets the MCU when one of the three following events occurs: – A Hardware reset, initiated by a low level on the Reset pin. – A Software reset, initiated by a HALT instruction (when enabled). – A Watchdog end of count condition. The event which caused the last Reset is flagged in the CLK_FLAG register, by setting the SOF- TRES or the WDGRES bits respectively; a hardware initiated reset will leave both these bits reset. The hardware reset overrides all other conditions and forces the ST9 to the reset state. During Reset, the internal registers are set to their reset values (when these reset values are defined, otherwise the register content will remain unchanged), and the I/O pins are set to Bidirectional Weak-PullUp or High impedance input. See Section 1.3. Reset is asynchronous: as soon as the reset pin is driven low, a Reset cycle is initiated. Figure 55. Oscillator Start-up Sequence and Reset Timing n VDD MAX VDD MIN OSCIN TSTUP INTCLK RESET PIN VR02085A Note: For TSTUP value refer to Oscillator Electrical Characteristics 111/320 9 ST92F120 - 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.WDGEN 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 theReset pin goeshigh again, a deterministic number of oscillator clock cycles (CLOCK1) is counted (refer to TRSPH) 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 memory, 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 memory 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. 7.6.1 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 112/320 9 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 will elapse before the RCCU detects this rising front. From this event on, a defined number of CLOCK1 cycles (refer to TRSPH) is counted before exiting the Reset state (±1CLOCK1 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 56. Recommended Signal to be Applied on Reset Pin VRESETN VDD VIHRS VILRS 20µs Minimum ST92F120 - RESET AND CLOCK CONTROL UNIT (RCCU) 7.7 STOP MODE On ST9 devices provided with an external STOP pin, the Reset/Stop Manager can also stop all oscillators without resetting the device. To enter STOP Mode, the STOP pin must be tied low. When the STOP pin is tied high again, the MCU resumes execution of the program after a set number of CLOCK2 cycles (refer to TSTR in the Electrical Characteristics section), without losing the status. Note: If STOP Mode is entered, the clock is stopped: hence, also the watchdog counter is stopped. When the ST9 exits from STOP Mode, the watchdog counter restarts from where it was before STOP Mode was entered. 113/320 9 ST92F120 - EXTERNAL MEMORY INTERFACE (EXTMI) 8 EXTERNAL MEMORY INTERFACE (EXTMI) 8.1 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 57. 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 9 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 114/320 Relocation of P[3:0] and DPR[3:0] Registers 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 ST92F120 - EXTERNAL MEMORY INTERFACE (EXTMI) 8.2 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 59 8.2.1 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. 8.2.2 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 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. 8.2.3 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. 115/320 9 ST92F120 - EXTERNAL MEMORY INTERFACE (EXTMI) Figure 58. 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) 116/320 9 ST92F120 - EXTERNAL MEMORY INTERFACE (EXTMI) EXTERNAL MEMORY SIGNALS (Cont’d) Figure 59. 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) TAVWL WRITE TAVWH RW (MC=0) RW (MC=1) 117/320 9 ST92F120 - EXTERNAL MEMORY INTERFACE (EXTMI) EXTERNAL MEMORY SIGNALS (Cont’d) 8.2.4 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. 8.2.5 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 60. 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) RW (MC=1) 118/320 9 ADDRESS D.OUT ADD. DATA OUT WRITE MULTIPLEXED ST92F120 - 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. 8.2.6 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]. 8.2.7 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 61. 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. 8.2.8 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 60. 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 E A15-A8 119/320 9 ST92F120 - EXTERNAL MEMORY INTERFACE (EXTMI) 8.3 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 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 58 120/320 9 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. 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. ST92F120 - EXTERNAL MEMORY INTERFACE (EXTMI) EXTERNAL MEMORY INTERFACE REGISTERS (Cont’d) EXTERNAL MEMORY REGISTER 2 (EMR2) the contents of ISR. In this case, iret will also reR246 - Read/Write store CSR from the stack. This approach allows Register Page: 21 interrupt service routines to access the entire Reset value: 0001 1111 (1Fh) 4Mbytes of address space; the drawback is that the interrupt response time is slightly increased, 7 0 because of the need to also save CSR on the MEM stack. Full compatibility with the original ST9 is ENCSR DPRREM LAS1 LAS0 UAS1 UAS0 SEL 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 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 57 Bit 4 = MEMSEL: Memory Selection. Warning: Must be kept as it is set in the BootROM (Reset value is 1) . 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. 121/320 9 ST92F120 - EXTERNAL MEMORY INTERFACE (EXTMI) EXTERNAL MEMORY INTERFACE REGISTERS (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- 122/320 9 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. ST92F120 - I/O PORTS 9 I/O PORTS 9.1 INTRODUCTION 9.2 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. 9.3 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 62. 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 62. I/O Register Map GROUP E System Registers E5h E4h E3h E2h E1h E0h 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 123/320 9 ST92F120 - 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. 9.4 INPUT/OUTPUT BIT CONFIGURATION By programming the control bits PxC0.n and PxC1.n (see Figure 63) 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). 124/320 9 Each pin of an I/O port may assume software programmable Alternate Functions (refer to the device Pin Description and to Section 9.5 ALTERNATE FUNCTION ARCHITECTURE). 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 64. 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 65). – 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. ST92F120 - I/O PORTS INPUT/OUTPUT BIT CONFIGURATION (Cont’d) Figure 63. Control Bits Bit 7 Bit n Bit 0 PxC2 PxC27 PxC2n PxC20 PxC1 PxC17 PxC1n PxC10 PxC0 PxC07 PxC0n PxC00 n Table 26. 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) 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 For A/D Converter inputs. Legend: X = n = AF = BID = CMOS= HI-Z = IN = OD = OUT = PP = TTL = WP = Port Bit Alternate Function Bidirectional CMOS Standard Input Levels High Impedance Input Open Drain Output Push-Pull TTL Standard Input Levels Weak Pull-up 125/320 9 ST92F120 - I/O PORTS INPUT/OUTPUT BIT CONFIGURATION (Cont’d) Figure 64. 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 65. Input Configuration Figure 66. 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 INTERNAL DATA BUS n 126/320 9 INPUT LATCH ST92F120 - I/O PORTS INPUT/OUTPUT BIT CONFIGURATION (Cont’d) When Px.n is programmed as an Output: (Figure 66) – 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 67) – 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 68) – The Output Buffer is turned on in an Open-Drain or Push-Pull configuration. – 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 67. 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 68. 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 127/320 9 ST92F120 - I/O PORTS 9.5 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 9.5.1 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). 9.5.2 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. 9.5.3 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 128/320 9 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. 9.6 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). TIMER/WATCHDOG (WDT) 10 ON-CHIP PERIPHERALS 10.1 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. 10.1.1 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 69. 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. INTA0 REQUEST 129/320 9 TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 10.1.2 Functional Description 10.1.2.1 External Signals The HW0SW1 pin can be used to permanently enable Watchdog mode. Refer to section 10.1.3.1 on page 131. 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. 10.1.2.2 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. 10.1.2.3 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. 130/320 9 10.1.2.4 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. 10.1.2.5 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. 10.1.2.6 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. TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 10.1.2.7 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. 10.1.2.8 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. 10.1.2.9 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. 10.1.2.10 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. 10.1.3 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. 10.1.3.1 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. 10.1.3.2 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. 131/320 9 TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 10.1.3.3 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. 10.1.3.4 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 70. Watchdog Timer Mode COUNT VALUE TIMER START COUNTING RESET WRITEWDTRH,WDTRL WDGEN=0 WRITE AAh,55h INTOWDTRL PRODUCE COUNT RELOAD 132/320 9 SOFTWAREFAIL (E.G. INFINITELOOP) ORPERIPHERALFAIL VA00220 TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 10.1.4 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 71. Note: Software traps can be generated by setting the appropriate interrupt pending bit. Table 27 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 71. 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 27. 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. 133/320 9 TIMER/WATCHDOG (WDT) TIMER/WATCHDOG (Cont’d) 10.1.5 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 9 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. 134/320 TIMER/WATCHDOG PRESCALER REGISTER (WDTPR) R250 - Read/Write Register Page: 0 Reset value: 1111 1111 (FFh) 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 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 by the user program. At System Reset, the Watchdog mode is disabled. 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). WAIT CONTROL REGISTER (WCR) R252 - Read/Write Register Page: 0 Reset value: 0111 1111 (7Fh) 7 0 WDGEN x x x x 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 0 = OUTEN: Output Enable bit. This bit is set and cleared by software. 0: Disable output 1: Enable output x EXTERNAL INTERRUPT VECTOR REGISTER (EIVR) R246 - Read/Write Register Page: 0 Reset value: xxxx 0110 (x6h) 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 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. 135/320 9 STANDARD TIMER (STIM) 10.2 STANDARD TIMER (STIM) Important Note: This chapter is a generic description of the STIM peripheral. Depending on the ST9 device, some or all of the interface signals described may not be connected to external pins. For the list of STIM pins present on the particular ST9 device, refer to the pinout description in the first section of the data sheet. 10.2.1 Introduction The Standard Timer includes a programmable 16bit down counter and an associated 8-bit prescaler with Single and Continuous counting modes capability. The Standard Timer uses an input pin (STIN) and an output (STOUT) pin. These pins, when available, may be independent pins or connected as Alternate Functions of an I/O port bit. STIN can be used in one of four programmable input modes: – event counter, – gated external input mode, – triggerable input mode, – retriggerable input mode. STOUT can be used to generate a Square Wave or Pulse Width Modulated signal. The Standard Timer is composed of a 16-bit down counter with an 8-bit prescaler. The input clock to the prescaler can be driven either by an internal clock equal to INTCLK divided by 4, or by CLOCK2 derived directly from the external oscillator, divided by device dependent prescaler value, thus providing a stable time reference independent from the PLL programming or by an external clock connected to the STIN pin. The Standard Timer End Of Count condition is able to generate an interrupt which is connected to one of the external interrupt channels. The End of Count condition is defined as the Counter Underflow, whenever 00h is reached. Figure 72. Standard Timer Block Diagram n INEN INMD1 INMD2 STIN1 INPUT & (See Note 2) CLOCK CONTROL LOGIC INTCLK/4 STP 8-BIT PRESCALER MUX STAN DARD TIMER CLOCK STH,STL 16-BIT DOWNCOUNTER END OF COUNT CLOCK2/x OUTMD1 OUTMD2 STOUT1 OUTPU T CONTROL LOGIC EXTERN AL INTERRUPT 1 INTER RUPT INTS CONTROL LOGIC INTERRUPT REQUEST Note 1: Pin not present on all ST9 devices. Note 2: Depending on device, the source of the INPUT & CLOCK CONTROL LOGIC block may be permanently connected either to STIN or the RCCU signal CLOCK2/x. In devices without STIN and CLOCK2, the INEN bit must be held at 0. 136/320 9 STANDARD TIMER (STIM) STANDARD TIMER (Cont’d) 10.2.2 Functional Description 10.2.2.1 Timer/Counter control Start-stop Count. The ST-SP bit (STC.7) is used in order to start and stop counting. An instruction which sets this bit will cause the Standard Timer to start counting at the beginning of the next instruction. Resetting this bit will stop the counter. If the counter is stopped and restarted, counting will resume from the value held at the stop condition, unless a new constant has been entered in the Standard Timer registers during the stop period. In this case, the new constant will be loaded as soon as counting is restarted. A new constant can be written in STH, STL, STP registers while the counter is running. The new value of the STH and STL registers will be loaded at the next End of Count condition, while the new value of the STP register will be loaded immediately. WARNING: In order to prevent incorrect counting of the StandardTimer,the prescaler(STP) andcounter (STL, STH) registers must be initialised before the starting of the timer. If this is not done, counting will start with the reset values (STH=FFh, STL=FFh, STP=FFh). Single/Continuous Mode. The S-C bit (STC.6) selects between the Single or Continuous mode. SINGLE MODE: at the End of Count, the Standard Timer stops, reloads the constant and resets the Start/Stop bit (the user programmer can inspect the timer current status by reading this bit). Setting the Start/Stop bit will restart the counter. CONTINUOUS MODE: At the End of the Count, the counter automatically reloads the constant and restarts. Itis only stoppedbyresettingtheStart/Stop bit. The S-C bit can be written either with the timer stopped or running. It is possible to toggle the S-C bit and start the Standard Timer with the same instruction. 10.2.2.2 Standard Timer Input Modes (ST9 devices with Standard Timer Input STIN) Bits INMD2, INMD1 and INEN are used to select the input modes. The Input Enable (INEN) bit enables the input mode selected by the INMD2 and INMD1 bits. If the input is disabled (INEN=”0”), the values of INMD2 and INMD1 are not taken into account. In this case, this unit acts as a 16-bit timer (plus prescaler) directly driven by INTCLK/4 and transitions on the input pin have no effect. Event Counter Mode (INMD1 = ”0”, INMD2 = ”0”) The Standard Timer is driven by the signal applied to the input pin (STIN) which acts as an external clock. The unit works therefore as an event counter. The event is a high to low transition on STIN. Spacing between trailing edges should be at least the period of INTCLK multiplied by 8 (i.e. the maximum Standard Timer input frequency is 2.5 MHz with INTCLK = 20MHz). Gated Input Mode (INMD1 = ”0”, INMD2 = “1”) The Timer uses the internal clock (INTCLK divided by 4) and starts and stops the Timer according to the state of STIN pin. When the status of the STIN is High the Standard Timer count operation proceeds, and when Low, counting is stopped. Triggerable Input Mode(INMD1= “1”,INMD2=“0”) The Standard Timer is started by: a) setting the Start-Stop bit, AND b) a High to Low (low trigger) transition on STIN. In order to stop the Standard Timer in this mode, it is only necessary to reset the Start-Stop bit. Retriggerable Input Mode (INMD1 = “1”, INMD2 = “1”) In this mode, when the Standard Timer is running (with internal clock), a High to Low transition on STIN causes the counting to start from the last constant loaded into the STL/STH and STP registers. When the Standard Timer is stopped (ST-SP bit equal to zero), a High to Low transition on STIN has no effect. 10.2.2.3 Time Base Generator (ST9 devices without Standard Timer Input STIN) For devices where STIN is replaced by a connection to CLOCK2, the condition (INMD1 = “0”, INMD2 = “0”) will allow the Standard Timer to generate a stable time base independent from the PLL programming. 137/320 9 STANDARD TIMER (STIM) STANDARD TIMER (Cont’d) 10.2.2.4 Standard Timer Output Modes OUTPUT modes are selected using 2 bits of the STC register: OUTMD1 and OUTMD2. No Output Mode (OUTMD1 = “0”, OUTMD2 = “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 (OUTMD1 = “0”, OUTMD2 = “1”) The Standard Timer toggles the state of the STOUT pin on every End Of Count condition. With INTCLK = 12MHz, this allows generation of a square wave with a period ranging from 666ns to 11.18 seconds. PWM Output Mode (OUTMD1 = “1”) The value of the OUTMD2 bit is transferred to the STOUT output pin at the End Of Count. This allows the user to generate PWM signals, by modifying the status of OUTMD2 between End of Count events, based on software counters decremented on the Standard Timer interrupt. 10.2.3 Interrupt Selection The Standard Timer may generate an interrupt request at every End of Count. Bit 2 of the STC register (INTS) selects the interrupt source between the Standard Timer interrupt and the external interrupt pin. Thus the Standard Timer Interrupt uses the interrupt channel and takes the priority and vector of the external interrupt channel. If INTS is set to “1”, the Standard Timer interrupt is disabled; otherwise, an interrupt request is generated at every End of Count. Note: When enabling or disabling the Standard Timer Interrupt (writing INTS in the STC register) an edge may be generated on the interrupt channel, causing an unwanted interrupt. To avoid this spurious interrupt request, the INTS bit should be accessed only when the interrupt log- 138/320 9 ic is disabled (i.e. after the DI instruction). It is also necessary to clear any possible interrupt pending requests on the corresponding external interrupt channel before enabling it. A delay instruction (i.e. a NOP instruction) must be inserted between the reset of the interrupt pending bit and the INTS write instruction. 10.2.4 Register Mapping Depending on the ST9 device there may be up to 4 Standard Timers (refer to the block diagram in the first section of the data sheet). Each Standard Timer has 4 registers mapped into Page 11 in Group F of the Register File In the register description on the following page, register addresses refer to STIM0 only. STD Timer Register STIM0 STH0 STL0 STP0 STC0 STIM1 STH1 STL1 STP1 STC1 STIM2 STIM3 STH2 STL2 STP2 STC2 STH3 STL3 STP3 STC3 R240 R241 R242 R243 R244 R245 R246 R247 Register Address (F0h) (F1h) (F2h) (F3h) (F4h) (F5h) (F6h) (F7h) R248 R249 R250 R251 R252 R253 R254 R255 (F8h) (F9h) (FAh) (FBh) (FCh) (FDh) (FEh) (FFh) Note: The four standard timers are not implemented on all ST9 devices. Refer to the block diagram of the device for the number of timers. STANDARD TIMER (STIM) STANDARD TIMER (Cont’d) 10.2.5 Register Description STANDARD TIMER CONTROL (STC) R243 - Read/Write Register Page: 11 Reset value: 0001 0100 (14h) COUNTER HIGH BYTE REGISTER (STH) R240 - Read/Write Register Page: 11 Reset value: 1111 1111 (FFh) 7 0 ST.15 ST.14 ST.13 ST.12 ST.11 ST.10 ST.9 ST.8 COUNTER LOW BYTE REGISTER (STL) R241 - Read/Write Register Page: 11 Reset value: 1111 1111 (FFh) ST.7 0 ST.6 ST.5 ST.4 ST.3 ST.2 STSP 0 S-C INMD INMD OUTM INEN INTS 1 2 D1 OUTM D2 Bit 6 = S-C: Single-Continuous Mode Select. This bit is set and cleared by software. 0: Continuous Mode 1: Single Mode ST.1 ST.0 Bit 7:0 = ST.[7:0]: Counter Low Byte. Writing to the STH and STL registers allows the user to enter the Standard Timer constant, while reading it provides the counter’s current value. Thus it is possible to read the counter on-the-fly. STANDARD TIMER PRESCALER REGISTER (STP) R242 - Read/Write Register Page: 11 Reset value: 1111 1111 (FFh) 7 7 Bit 7 = ST-SP: Start-Stop Bit. This bit is set and cleared by software. 0: Stop counting 1: Start counting Bit 7:0 = ST.[15:8]: Counter High-Byte. 7 REGISTER 0 STP.7 STP.6 STP.5 STP.4 STP.3 STP.2 STP.1 STP.0 Bit 7:0 = STP.[7:0]: Prescaler. The Prescaler value for the Standard Timer is programmed into this register. When reading the STP register, the returned value corresponds to the programmed data instead of the current data. 00h: No prescaler 01h: Divide by 2 FFh: Divide by 256 Bit 5:4 = INMD[1:2]: Input Mode Selection. These bits select the Input functions as shown in Section 10.2.2.2, when enabled by INEN. INMD1 0 0 1 1 INMD2 0 1 0 1 Mode Event Counter mode Gated input mode Triggerable mode Retriggerable mode Bit 3 = INEN: Input Enable. This bit is set and cleared by software. If neither the STIN pin nor the CLOCK2 line are present, INEN must be 0. 0: Input section disabled 1: Input section enabled Bit 2 = INTS: Interrupt Selection. 0: Standard Timer interrupt enabled 1: Standard Timer interrupt is disabled and the external interrupt pin is enabled. Bit 1:0 = OUTMD[1:2]: Output Mode Selection. These bits select the output functions as described in Section 10.2.2.4. OUTMD1 0 0 1 OUTMD2 0 1 x Mode No output mode Square wave output mode PWM output mode 139/320 9 EXTENDED FUNCTION TIMER (EFT) 10.3 EXTENDED FUNCTION TIMER (EFT) 10.3.1 Introduction The timer consists of a 16-bit free-running counter driven by a programmable prescaler. It may be used for a variety of purposes, including pulse length measurement of up to two input signals (input capture) or generation of up to two output waveforms (output compare and PWM). Pulse lengths and waveform periods can be modulated from a few microseconds to several milliseconds using the timer prescaler and the INTCLK prescaler. 10.3.2 Main Features ■ Programmable prescaler: INTCLK divided by 2, 4 or 8. ■ Overflow status flag and maskable interrupts ■ External clock input (must be at least 4 times slower than the INTCLK clock speed) with the choice of active edge ■ Output compare functions with – 2 dedicated 16-bit registers – 2 dedicated programmable signals – 2 dedicated status flags – 1 dedicated maskable interrupt ■ Input capture functions with – 2 dedicated 16-bit registers – 2 dedicated active edge selection signals – 2 dedicated status flags – 1 dedicated maskable interrupt ■ Pulse width modulation mode (PWM) ■ One pulse mode ■ 5 alternate functions on I/O ports* ■ Up to 3 separate Timer interrupts or a global interrupt (depending on device) mapped on external interrupt channels: – ICI: Timer Input capture interrupt. – OCI: Timer Output compare interrupt. – TOI: Timer Overflow interrupt. – EFTI: Timer Global interrupt (replaces ICI, OCI and TOI). The Block Diagram is shown in Figure 73. 140/320 9 Table 28. EFT Pin Naming conventions Function Input Capture 1 ICAP1 Input Capture 2 ICAP2 Output Compare 1 OCMP1 Output Compare 2 OCMP2 EFT0 EFT1 EFTn ICAPA0 ICAPA1 ICAPAn ICAPB0 ICAPB1 ICAPBn OCMPA0 OCMPA1 OCMPAn OCMPB0 OCMPB1 OCMPBn *Note 1: Some external pins are not available on all devices. Refer to the device pin out description. *Note 2: Refer to the device interrupt description, to see if a single timer interrupt is used, or three separate interrupts. 10.3.3 Functional Description 10.3.3.1 Counter The principal block of the Programmable Timer is a 16-bit free running counter and its associated 16-bit registers: Counter Registers – Counter High Register (CHR) is the most significant byte (MSB). – Counter Low Register (CLR) is the least significant byte (LSB). Alternate Counter Registers – Alternate Counter High Register (ACHR) is the most significant byte (MSB). – Alternate Counter Low Register (ACLR) is the least significant byte (LSB). These two read-only 16-bit registers contain the same value but with the difference that reading the ACLR register does not clear the TOF bit (overflow flag), (see note page 142). Writing in the CLR register or ACLR register resets the free running counter to the FFFCh value. The timer clock depends on the clock control bits of the CR2 register, as illustrated in Table 29. The value in the counter register repeats every 131.072, 262.144 or 524.288 INTCLK cycles depending on the CC1 and CC0 bits. EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) Figure 73. Timer Block Diagram ST9 INTERNAL BUS INTCLK MCU-PERIPHERAL INTERFACE 8 low 8 8 low 8 low 8 high 8 high 8 low 8 high EXEDG 8 low 8-bit buffer high 8 high 16 1/2 1/4 1/8 16 BIT FREE RUNNING COUNTER OUTPUT COMPARE REGISTER 2 OUTPUT COMPARE REGISTER 1 INPUT CAPTURE REGISTER INPUT CAPTURE REGISTER 1 2 COUNTER ALTERNATE REGISTER 16 16 16 CC1 CC0 TIMER INTERNAL BUS 16 EXTCLK OVERFLOW DETECT CIRCUIT 16 OUTPUT COMPARE CIRCUIT 6 ICF1 OCF1 TOF ICF2 OCF2 0 0 EDGE DETECT CIRCUIT1 ICAP1 EDGE DETECT CIRCUIT2 ICAP2 LATCH1 OCMP1 LATCH2 OCMP2 0 SR ICIE OCIE TOIE FOLV2 FOLV1 OLVL2 IEDG1 OLVL1 OC1E OC2E OPM PWM CC1 CC0 IEDG2 EXEDG CR1 ETOI EOC CR2 EICI 0 0 0 0 TOIS OCIS ICIS EFTIS EEFTI 1 0 1 0 1 0 TOI OCI ICI CR3 1 0 EFTI 141/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) 16-bit read sequence: (from either the Counter Register or the Alternate Counter Register). Beginning of the sequence At t0 Read MSB LSB is buffered Other instructions Returns the buffered At t0 +Dt Read LSB LSB value at t0 Sequence completed The user must read the MSB first, then the LSB value is buffered automatically. This buffered value remains unchanged until the 16-bit read sequence is completed, even if the user reads the MSB several times. After a complete reading sequence, if only the CLR register or ACLR register are read, they return the LSB of the count value at the time of the read. An overflow occurs when the counter rolls over from FFFFh to 0000h then: – The TOF bit of the SR register is set. – A timer interrupt is generated if: – TOIE bit of the CR1 register is set – TOIS bit of the CR3 register is set (or EFTIS bit if only global interrupt is available). If one of these conditions is false, the interrupt remains pending to be issued as soon as they are both true. 142/320 9 Clearing the overflow interrupt request is done by: 1. Reading the SR register while the TOF bit is set. 2. An access (read or write) to the CLR register. Notes: The TOF bit is not cleared by accesses to ACLR register. This feature allows simultaneous use of the overflow function and reads of the free running counter at random times (for example, to measure elapsed time) without the risk of clearing the TOF bit erroneously. The timer is not affected by WAIT mode. In HALT mode, the counter stops counting until the mode is exited. Counting then resumes from the reset count (MCU awakened by a Reset). 10.3.3.2 External Clock The external clock (where available) is selected if CC0=1 and CC1=1 in CR2 register. The status of the EXEDG bit determines the type of level transition on the external clock pin EXTCLK that will trigger the free running counter. The counter is synchronised with the falling edge of INTCLK. At least four falling edges of the INTCLK must occur between two consecutive active edges of the external clock; thus the external clock frequency must be less than a quarter of the INTCLK frequency. EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) Figure 74. Counter Timing Diagram, INTCLK divided by 2 INTCLK INTERNAL RESET TIMER CLOCK FFFD FFFE FFFF 0000 COUNTER REGISTER 0001 0002 0003 OVERFLOW FLAG TOF Figure 75. Counter Timing Diagram, INTCLK divided by 4 INTCLK INTERNAL RESET TIMER CLOCK COUNTER REGISTER FFFC FFFD 0000 0001 OVERFLOW FLAG TOF Figure 76. Counter Timing Diagram, INTCLK divided by 8 INTCLK INTERNAL RESET TIMER CLOCK COUNTER REGISTER FFFC FFFD 0000 OVERFLOW FLAG TOF 143/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) 10.3.3.3 Input Capture In this section, the index, i, may be 1 or 2. The two input capture 16-bit registers (IC1R and IC2R) are used to latch the value of the free running counter after a transition detected by the ICAP i pin (see figure 5). ICiR MS Byte LS Byte ICiHR ICiLR ICi Rregister is a read-only register. The active transition is software programmable through the IEDGi bit of the Control Register (CRi). Timing resolution is one count of the free running counter: (INTCLK /CC[1:0]). Procedure To use the input capture function select the following in the CR2 register: – Select the timer clock (CC[1:0] (see Table 29). – Select the edge of the active transition on the ICAP2 pin with the IEDG2 bit. And select the following in the CR1 register: – Set the ICIE bit to generate an interrupt after an input capture. 144/320 9 – Select the edge of the active transition on the ICAP1 pin with the IEDG1 bit. When an input capture occurs: – ICFi bit is set. – The ICiR register contains the value of the free running counter on the active transition on the ICAPi pin (see Figure 78). – A timer interrupt is generated if the ICIE bit is set and the ICIS bit (or EFTIS bit if only global interrupt is available) is set. Otherwise, the interrupt remains pending until both conditions become true. Clearing the Input Capture interrupt request is done by: 1. Reading the SR register while the ICFi bit is set. 2. An access (read or write) to the ICiLR register. Note: After reading the ICiHR register, transfer of input capture data is inhibited until the ICiLR register is also read. The ICiR register always contains the free running counter value which corresponds to the most recent input capture. EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) Figure 77. Input Capture Block Diagram ICAP1 ICAP2 (Control Register 1) CR1 EDGE DETECT CIRCUIT2 EDGE DETECT CIRCUIT1 ICIE IEDG1 (Status Register) SR ICF1 IC1R IC2R ICF2 0 0 0 (Control Register 2) CR2 16-BIT 16-BIT FREE RUNNING CC1 CC0 IEDG2 COUNTER Figure 78. Input Capture Timing Diagram TIMER CLOCK COUNTER REGISTER FF01 FF02 FF03 ICAPi PIN ICAPi FLAG ICAPi REGISTER FF03 Note: Active edge is rising edge. 145/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) 10.3.3.4 Output Compare In this section, the index, i, may be 1 or 2. This function can be used to control an output waveform or indicating when a period of time has elapsed. When a match is found between the Output Compare register and the free running counter, the output compare function: – Assigns pins with a programmable value if the OCiE bit is set – Sets a flag in the status register – Generates an interrupt if enabled Two 16-bit registers Output Compare Register 1 (OC1R) and Output Compare Register 2 (OC2R) contain the value to be compared to the free running counter each timer clock cycle. OCiR MS Byte LS Byte OCiHR OCiLR These registers are readable and writable and are not affected by the timer hardware. A reset event changes the OCiR value to 8000h. Timing resolution is one count of the free running counter: (INTCLK /CC[1:0]). Procedure To use the output compare function, select the following in the CR2 register: – Set the OCiE bit if an output is needed then the OCMPi pin is dedicated to the output compare i function. – Select the timer clock CC[1:0] (see Table 29). And select the following in the CR1 register: – Select the OLVL i bit to applied to the OCMPi pins after the match occurs. – Set the OCIE and OCIS bits (or EFTIS bit if only global interrupt is available) to generate an interrupt if it is needed. When match is found: – OCFi bit is set. – The OCMPi pin takes OLVLi bit value (OCMPi pin latch is forced low during reset and stays low until valid compares change it to OLVLi level). – A timer interrupt is generated if the OCIE bit is set in the CR2 register and OCIS bit (or EFTIS bit 146/320 9 if only global interrupt is available) is set in the CR3 register. Clearing the output compare interrupt request is done by: 3. Reading the SR register while the OCFi bit is set. 4. An access (read or write) to the OCiLR register. Note: After a processor write cycle to the OCiHR register, the output compare function is inhibited until the OCiLR register is also written. If the OCiE bit is not set, the OCMPi pin is a general I/O port and the OLVLi bit will not appear when match is found but an interrupt could be generated if the OCIE bit is set. The value in the 16-bit OCiR register and the OLVLi bit should be changed after each successful comparison in order to control an output waveform or establish a new elapsed timeout. The OCiR register value required for a specific timing application can be calculated using the following formula: ∆ OCiR = Where: ∆t ∆t * INTCLK (CC1.CC0) = Desired output compare period (in seconds) INTCLK = Internal clock frequency CC1-CC0 = Timer clock prescaler The following procedure is recommended to prevent the OCFi bit from being set between the time it is read and the write to the OCiR register: – Write to the OCiHR register (further compares are inhibited). – Read the SR register (first step of the clearance of the OCFi bit, which may be already set). – Write to the OCiLR register (enables the output compare function and clears the OCFi bit). EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) Figure 79. Output Compare Block Diagram 16 BIT FREE RUNNING COUNTER OC1E OC2E CC1 CC0 (Control Register 2) CR2 16-bit (Control Register 1) CR1 OUTPUT COMPARE CIRCUIT 16-bit OC1R OCIE OLVL2 OLVL1 Latch 1 OCMP1 Latch 2 OCMP2 16-bit OC2R OCF1 OCF2 0 0 0 (Status Register) SR Figure 80. Output Compare Timing Diagram, Internal Clock Divided by 2 INTCLK TIMER CLOCK COUNTER OUTPUT COMPARE REGISTER FFFC FFFD FFFD FFFE FFFF 0000 CPU writes FFFF FFFF COMPARE REGISTER LATCH OCFi AND OCMPi PIN (OLVLi=1) 147/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) 10.3.3.5 Forced Compare Mode In this section i may represent 1 or 2. The following bits of the CR1 register are used: FOLV2 FOLV1 OLVL2 OLVL1 When the FOLVi bit is set, the OLVLi bit is copied to the OCMPi pin. The OLVLi bit has to be toggled in order to toggle the OCMPi pin when it is enabled (OC iE bit=1). The OCFi bit is not set, and thus no interrupt request is generated. 10.3.3.6 One Pulse Mode One Pulse mode enables the generation of a pulse when an external event occurs. This mode is selected via the OPM bit in the CR2 register. The one pulse mode uses the Input Capture1 function and the Output Compare1 function. Procedure To use one pulse mode, select the following in the the CR1 register: – Using the OLVL1 bit, select the level to be applied to the OCMP1 pin after the pulse. – Using the OLVL2 bit, select the level to be applied to the OCMP1 pin during the pulse. – Select the edge of the active transition on the ICAP1 pin with the IEDG1 bit. And select the following in the CR2 register: – Set the OC1E bit, the OCMP1 pin is then dedicated to the Output Compare 1 function. – Set the OPM bit. – Select the timer clock CC[1:0] (see Table 29). Load the OC1R register with the value corresponding to the length of the pulse (see the formula in Section 10.3.3.7). One pulse mode cycle When event occurs on ICAP1 Counter is initialized to FFFCh OCMP1 = OLVL2 When Counter = OC1R OCMP1 = OLVL1 Then, on a valid event on the ICAP1 pin, the counter is initialized to FFFCh and OLVL2 bit is loaded on the OCMP1 pin. When the value of the counter is equal to the value of the contents of the OC1R register, the OLVL1 bit is output on the OCMP1 pin, (See Figure 81). Note: The OCF1 bit cannot be set by hardware in one pulse mode but the OCF2 bit can generate an Output Compare interrupt. The ICF1 bit is set when an active edge occurs and can generate an interrupt if the ICIE bit is set. When the Pulse Width Modulation (PWM) and One Pulse Mode (OPM) bits are both set, the PWM mode is the only active one. Figure 81. One Pulse Mode Timing COUNTER .... FFFC FFFD FFFE 2ED0 2ED1 2ED2 FFFC FFFD 2ED3 ICAP1 OCMP1 OLVL2 OLVL1 compare1 Note: IEDG1=1, OC1R=2ED0h, OLVL1=0, OLVL2=1 148/320 9 OLVL2 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) 10.3.3.7 Pulse Width Modulation Mode Pulse Width Modulation mode enables the generation of a signal with a frequency and pulse length determined by the value of the OC1R and OC2R registers. The pulse width modulation mode uses the complete Output Compare 1 function plus the OC2R register. Procedure To use pulse width modulation mode select the following in the CR1 register: – Using the OLVL1 bit, select the level to be applied to the OCMP1 pin after a successful comparison with OC1R register. – Using the OLVL2 bit, select the level to be applied to the OCMP1 pin after a successful comparison with OC2R register. And select the following in the CR2 register: – Set OC1E bit: the OCMP1 pin is then dedicated to the output compare 1 function. – Set the PWM bit. – Select the timer clock CC[1:0] bits (see Table 29). Load the OC2R register with the value corresponding to the period of the signal. Load the OC1R register with the value corresponding to the length of the pulse if (OLVL1=0 and OLVL2=1). If OLVL1=1 and OLVL2=0 the length of the pulse is the difference between the OC2R and OC1R registers. The OCiR register value required for a specific timing application can be calculated using the following formula: OCiR Value = t* INTCLK Where: – t = Desired output compare period (seconds) – INTCLK = Internal clock frequency – CC1-CC0 = Timer clock prescaler The Output Compare 2 event causes the counter to be initialized to FFFCh (See Figure 82). Pulse Width Modulation cycle When Counter = OC1R When Counter = OC2R OCMP1 = OLVL1 OCMP1 = OLVL2 Counter is reset to FFFCh Note: After a write instruction to the OCi HR register, the output compare function is inhibited until the OCiLR register is also written. The OCF1 and OCF2 bits cannot be set by hardware in PWM mode therefore the Output Compare interrupt is inhibited. The Input Capture interrupts are available. When the Pulse Width Modulation (PWM) and One Pulse Mode (OPM) bits are both set, the PWM mode is the only active one. -5 CC[1:0] Figure 82. Pulse Width Modulation Mode Timing COUNTER 34E2 FFFC FFFD FFFE 2ED0 2ED1 2ED2 OLVL2 OCMP1 compare2 OLVL1 compare1 34E2 FFFC OLVL2 compare2 Note: OC1R=2ED0h, OC2R=34E2, OLVL1=0, OLVL2= 1 149/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) 10.3.4 Interrupt Management The interrupts of the Extended Function Timer are mapped on the eight external interrupt channels of the microcontroller (refer to the “Interrupts” chapter). Depending on device specification, one of the following configurations can occur: – The three interrupt sources are mapped on three different interrupt channels (to use this feature, the EFTIS bit must be reset) – The three interrupt sources are mapped on the same interrupt channel (to use this feature, the EFTIS bit must be set) Each External Interrupt Channel has: – A trigger control bit in the EITR register (R242 Page 0) – A pending bit in the EIPR register (R243 Page 0) – A mask bit in the EIMR register (R244 - Page 0) Program the interrupt priority level using the EIPLR register (R245 - Page 0). For a description of these registers refer to the “Interrupts” and “DMA” chapters. Use of three interrupt channels To use the interrupt features, for each interrupt channel used, perform the following sequence: – Set the priority level of the interrupt channel(s) used for the Extended Function Timer (EIPRL register) – Select the interrupt trigger edge(s) as rising edge (set the corresponding bit(s) in the EITR register) – Set the OCIS and/or ICIS and/or TOIS bit(s) of the CR3 register to select the peripheral interrupt source(s) – Set the OCIE and/or ICIE and/or TOIE bit(s) of the CR1 register to enable the peripheral to perform interrupt requests on the desiderate events – In the EIPR register, reset the pending bit(s) of the interrupt channels used by the peripheral interrupts to avoid any spurious interrupt requests being performed when the mask bit(s) is/are set – Set the mask bit(s) of the interrupt channel(s) used to enable the MCU to acknowledge the interrupt requests of the peripheral. 150/320 9 Use of one external interrupt channel for all the interrupts To use the interrupt features, perform the following sequence: – Set the priority level of the interrupt channel used (EIPRL register) – Select the interrupt trigger edge as rising edge (set the corresponding bit in the EITR register) – Set the EFTIS bit of the CR3 register to select the peripheral interrupt sources – Set the OCIE and/or ICIE and/or TOIE bit(s) of the CR1 register to enable the peripheral to perform interrupt requests on the wanted events – In the EIPR register, reset the pending bit of the interrupt channel used by the peripheral interrupts to avoid any spurious interrupt requests being performed when the mask bits is set – Set the mask bits of the interrupt channels used to enable the MCU to acknowledge the interrupt requests of the peripheral. Caution: Care should be taken when using only one of the input capture pins, as both capture interrupts are enabled by the ICIE bit in the CR1 register. If only ICAP1 is used (for example), an interrupt can still be generated by the ICAP2 pin when this pin toggles, even if it is configured as a standard output. If this case, the interrupt capture status bits in the SR register should handled in polling mode. Caution: 1. It is mandatory to clear all EFT interrupt flags simultaneously at least once before exiting an EFT timer interrupt routine (the SR register must = 00h at some point during the interrupt routine), otherwise no interrupts can be issued on that channel anymore. Refer to the following assembly code for an interrupt sequence example. 2. Since a loop statement is needed inside the IT routine, the user must avoid situations where an interrupt event period is narrower than the duration of the interrupt treatment. Otherwise nested interrupt mode must be used to serve higher priority requests. EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) Note: A single access (read/write) to the SR regisregisters must be accessed if the corresponding ter at the beginning of the interrupt routine is the flag is set. It is not necessary to access the SR first step needed to clear all the EFT interrupt register between these instructions, but it can flags. In a second step, the lower bytes of the data done. ; INTERRUPT ROUTINE EXAMPLE push R234 ; Save current page spp #28 ; Set EFT page L6: cp R254,#0 ; while E0_SR is not cleared jxz L7 tm R254,#128 ; Check Input Capture 1 flag jxz L2 ; else go to next test ld r1,R241 ; Dummy read to clear IC1LR ; Insert your code here L2: tm R254,#16 ; Check Input Capture 2 flag jxz L3 ; else go to next test ld r1,R243 ; Dummy read to clear IC2LR ; Insert your code here L3: tm R254,#64 ; Check Input Compare 1 flag jxz L4 ; else go to next test ld r1,R249 ; Dummy read to clear OC1LR ; Insert your code here L4: tm R254,#8 ; Check Input Compare 2 flag jxz L5 ; else go to next test ld r1,R251 ; Dummy read to clear OC1LR ; Insert your code here L5: tm R254,#32 ; Check Input Overflow flag jxz L6 ; else go to next test ld r1,R245 ; Dummy read to clear Overflow flag ; Insert your code here jx L6 L7: pop R234 ; Restore current page iret 151/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) 10.3.5 Register Description Each Timer is associated with three control and one status registers, and with six pairs of data registers (16-bit values) relating to the two input captures, the two output compares, the counter and the alternate counter. Notes: 1. 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. 2. To work correctly with register pairs, it is strongly recommended to use single byte instructions. Do not use word instructions to access any of the 16-bit registers. INPUT CAPTURE 1 HIGH REGISTER (IC1HR) R240 - Read Only Register Page: 28 Reset Value: Undefined This is an 8-bit read only register that contains the high part of the counter value (transferred by the input capture 1 event). 7 0 MSB LSB 152/320 9 INPUT CAPTURE 1 LOW REGISTER (IC1LR) R241 - Read Only Register Page: 28 Reset Value: Undefined This is an 8-bit read only register that contains the low part of the counter value (transferred by the input capture 1 event). 7 0 MSB LSB INPUT CAPTURE 2 HIGH REGISTER (IC2HR) R242 - Read Only Register Page: 28 Reset Value: Undefined This is an 8-bit read only register that contains the high part of the counter value (transferred by the Input Capture 2 event). 7 0 MSB LSB INPUT CAPTURE 2 LOW REGISTER (IC2LR) R243 - Read Only Register Page: 28 Reset Value: Undefined This is an 8-bit read only register that contains the low part of the counter value (transferred by the Input Capture 2 event). 7 0 MSB LSB EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) COUNTER HIGH REGISTER (CHR) R244 - Read Only Register Page: 28 Reset Value: 1111 1111 (FFh) This is an 8-bit register that contains the high part of the counter value. 7 0 MSB LSB COUNTER LOW REGISTER (CLR) R245 - Read/Write Register Page: 28 Reset Value: 1111 1100 (FCh) This is an 8-bit register that contains the low part of the counter value. A write to this register resets the counter. An access to this register after accessing the SR register clears the TOF bit. 7 0 MSB LSB ALTERNATE COUNTER HIGH REGISTER (ACHR) R246 - Read Only Register Page: 28 Reset Value: 1111 1111 (FFh) This is an 8-bit register that contains the high part of the counter value. 7 0 MSB LSB ALTERNATE COUNTER LOW REGISTER (ACLR) R247 - Read/Write Register Page: 28 Reset Value: 1111 1100 (FCh) This is an 8-bit register that contains the low part of the counter value. A write to this register resets the counter. An access to this register after an access to SR register does not clear the TOF bit in SR register. 7 0 MSB LSB 153/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) OUTPUT COMPARE 1 HIGH REGISTER (OC1HR) R248 - Read/Write Register Page: 28 Reset Value: 1000 0000 (80h) This is an 8-bit register that contains the high part of the value to be compared to the CHR register. OUTPUT COMPARE 2 HIGH REGISTER (OC2HR) R250 - Read/Write Register Page: 28 Reset Value: 1000 0000 (80h) This is an 8-bit register that contains the high part of the value to be compared to the CHR register. 7 0 7 0 MSB LSB MSB LSB OUTPUT COMPARE 1 LOW REGISTER (OC1LR) R249 - Read/Write Register Page: 28 Reset Value: 0000 0000 (00h) This is an 8-bit register that contains the low part of the value to be compared to the CLR register. OUTPUT COMPARE 2 LOW REGISTER (OC2LR) R251 - Read/Write Register Page: 28 Reset Value: 0000 0000 (00h) This is an 8-bit register that contains the low part of the value to be compared to the CLR register. 7 0 7 0 MSB LSB MSB LSB 154/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) CONTROL REGISTER 1 (CR1) R252 - Read/Write Register Page: 28 Reset Value: 0000 0000 (00h) 7 1: Forces the OLVL2 bit to be copied to the OCMP2 pin. 0 Bit 3 = FOLV1 Forced Output Compare 1. 0: No effect. 1: Forces OLVL1 to be copied to the OCMP1 pin. ICIE OCIE TOIE FOLV2 FOLV1 OLVL2 IEDG1 OLVL1 Bit 7 = ICIE Input Capture Interrupt Enable. 0: Interrupt is inhibited. 1: A timer interrupt is generated whenever the ICF1 or ICF2 bit of the SR register is set. Bit 6 = OCIE Output Compare Interrupt Enable. 0: Interrupt is inhibited. 1: A timer interrupt is generated whenever the OCF1 or OCF2 bit of the SR register is set. Bit 5 = TOIE Timer Overflow Interrupt Enable. 0: Interrupt is inhibited. 1: A timer interrupt is enabled whenever the TOF bit of the SR register is set. Bit 4 = FOLV2 Forced Output Compare 2. 0: No effect. Bit 2 = OLVL2 Output Level 2. This bit is copied to the OCMP2 pin whenever a successful comparison occurs with the OC2R register and OC2E is set in the CR2 register. This value is copied to the OCMP1 pin in One Pulse Mode and Pulse Width Modulation mode. Bit 1 = IEDG1 Input Edge 1. This bit determines which type of level transition on the ICAP1 pin will trigger the capture. 0: A falling edge triggers the capture. 1: A rising edge triggers the capture. Bit 0 = OLVL1 Output Level 1. The OLVL1 bit is copied to the OCMP1 pin whenever a successful comparison occurs with the OC1R register and the OC1E bit is set in the CR2 register. 155/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) CONTROL REGISTER 2 (CR2) R253 - Read/Write Register Page: 28 Reset Value: 0000 0000 (00h) 7 0 Bit 4 = PWM Pulse Width Modulation. 0: PWM mode is not active. 1: PWM mode is active, the OCMP1 pin outputs a programmable cyclic signal; the length of the pulse depends on the value of OC1R register; the period depends on the value of OC2R register. OC1E OC2E OPM PWM CC1 CC0 IEDG2 EXEDG Bit 7 = OC1E Output Compare 1 Enable. 0: Output Compare 1 function is enabled, but the OCMP1 pin is a general I/O. 1: Output Compare 1 function is enabled, the OCMP1 pin is dedicated to the Output Compare 1 capability of the timer. Bit 6 = OC2E Output Compare 2 Enable. 0: Output Compare 2 function is enabled, but the OCMP2 pin is a general I/O. 1: Output Compare 2 function is enabled, the OCMP2 pin is dedicated to the Output Compare 2 capability of the timer. Bit 5 = OPM One Pulse Mode. 0: One Pulse Mode is not active. 1: One Pulse Mode is active, the ICAP1 pin can be used to trigger one pulse on the OCMP1 pin; the active transition is given by the IEDG1 bit. The length of the generated pulse depends on the contents of the OC1R register. 156/320 9 Bit 3, 2 = CC[1:0] Clock Control. The value of the timer clock depends on these bits: Table 29. Clock Control Bits CC1 CC0 Timer Clock 0 0 0 1 1 0 INTCLK / 4 INTCLK / 2 INTCLK / 8 1 1 External Clock (where available) Bit 1 = IEDG2 Input Edge 2. This bit determines which type of level transition on the ICAP2 pin will trigger the capture. 0: A falling edge triggers the capture. 1: A rising edge triggers the capture. Bit 0 = EXEDG External Clock Edge. This bit determines which type of level transition on the external clock pin EXTCLK will trigger the free running counter. 0: A falling edge triggers the free running counter. 1: A rising edge triggers the free running counter. EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) STATUS REGISTER (SR) R254 - Read Only Register Page: 28 Reset Value: 0000 0000 (00h) The three least significant bits are not used. 7 ICF1 CONTROL REGISTER 3 (CR3) R255 - Read/Write Register Page: 28 Reset Value: 0000 0000 (00h) 0 OCF1 TOF ICF2 OCF2 0 0 7 0 0 0 Bit 7 = ICF1 Input Capture Flag 1. 0: No input capture (reset value). 1: An input capture has occurred. To clear this bit, first read the SR register, then read or write the low byte of the IC1R (IC1LR) register. Bit 6 = OCF1 Output Compare Flag 1. 0: No match (reset value). 1: The content of the free running counter has matched the content of the OC1R register. To clear this bit, first read the SR register, then read or write the low byte of the OC1R (OC1LR) register. Bit 5 = TOF Timer Overflow. 0: No timer overflow (reset value). 1: The free running counter rolled over from FFFFh to 0000h. To clear this bit, first read the SR register, then read or write the low byte of the CR (CLR) register. Note: Reading or writing the ACLR register does not clear TOF. 0 0 0 TOIS OCIS ICIS EFTIS Bit 7-4 = Unused Read as 0. Bit 3 = TOIS Timer Overflow Interrupt Selection. 0: Select External interrupt. 1: Select Timer Overflow Interrupt. Bit 2 = OCIS Output Compare Interrupt Selection. 0: Select External interrupt. 1: Select Timer Output Compare Interrupt. Bit 1 = ICIS Input Capture Interrupt Selection. 0: Select External interrupt. 1: Select Timer Input Capture Interrupt. Bit 0 = EFTIS Global Timer Interrupt Selection. 0: Select External interrupt. 1: Select Global Timer Interrupt. Bit 4 = ICF2 Input Capture Flag 2. 0: No input capture (reset value). 1: An input capture has occurred. To clear this bit, first read the SR register, then read or write the low byte of the IC2R (IC2LR) register. Bit 3 = OCF2 Output Compare Flag 2. 0: No match (reset value). 1: The content of the free running counter has matched the content of the OC2R register. To clear this bit, first read the SR register, then read or write the low byte of the OC2R (OC2LR) register. Bit 2-0 = Reserved, forced by hardware to 0. 157/320 9 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) Table 30. Extended Function Timer Register Map Address (Dec.) R240 R241 R242 R243 R244 R245 R246 R247 R248 R249 R250 R251 R252 R253 R254 R255 158/320 9 Register Name IC1HR Reset Value IC1LR Reset Value IC2HR Reset Value IC2LR Reset Value CHR Reset Value CLR Reset Value ACHR Reset Value ACLR Reset Value OC1HR Reset Value OC1LR Reset Value OC2HR Reset Value OC2LR Reset Value CR1 Reset Value CR2 Reset Value SR 7 6 5 4 3 2 1 MSB x LSB x x x x x x x x x x x x x x x x x x x x x x x x 1 1 1 1 1 1 MSB x 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 LSB MSB 1 0 LSB MSB 0 1 LSB MSB 1 0 LSB MSB 1 1 LSB MSB 1 x LSB MSB 1 x LSB MSB 1 x LSB MSB x x LSB MSB x 0 0 LSB 0 0 0 0 0 0 0 0 0 0 0 0 0 0 OC1E OC2E OPM PWM CC1 CC0 IEDG2 EXEDG 0 0 0 0 0 0 0 0 ICIE OCIE TOIE FOLV2 FOLV1 OLVL2 IEDG1 OLVL1 0 0 0 0 0 0 0 0 MSB 0 LSB ICF1 OCF1 TOF ICF2 OCF2 0 0 0 Reset Value 0 0 0 0 0 0 0 0 CR3 0 0 0 0 TOIS OCIS ICIS EFTIS Reset Value 0 0 0 0 0 0 0 0 EXTENDED FUNCTION TIMER (EFT) EXTENDED FUNCTION TIMER (Cont’d) Table 31. Extended Function Timer Page Map Timer number Page (hex) EFT0 1C EFT1 1D 159/320 9 MULTIFUNCTION TIMER (MFT) 10.4 MULTIFUNCTION TIMER (MFT) 10.4.1 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. Figure 83. MFT Simplified Block Diagram 160/320 9 – 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 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 84. Detailed Block Diagram 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). 161/320 9 MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.2 Functional Description The MFT operating modes are selected by programming the Timer Control Register (TCR) and the Timer Mode Register (TMR). 10.4.2.1 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 35). 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 10.4.2.11). 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.” 10.4.2.2 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. 10.4.2.3 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. 10.4.2.4 Triggered And Retriggered Modes A triggered event may be generated by software (by setting either the CP0 or the CP1 bit in the 162/320 9 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. 10.4.2.5 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. 10.4.2.6 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. 10.4.2.7 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). MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.2.8 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). 10.4.2.9 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”. 10.4.2.10 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. 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). 10.4.2.11 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 32 below: Table 32. 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). 163/320 9 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. 10.4.2.12 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. 164/320 9 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. 10.4.2.13 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 85. 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. MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.3 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 33 Table 33. 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: – 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. 165/320 9 MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.3.1 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. 10.4.3.2 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. 10.4.3.3 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. 10.4.3.4 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. 10.4.3.5 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. 10.4.3.6 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). 166/320 9 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 10.4.3.7 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. 10.4.3.8 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. MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.3.9 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. 10.4.3.10 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. 10.4.3.11 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. 10.4.3.12 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. 167/320 9 MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.3.13 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. 10.4.3.14 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- 168/320 9 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 10.4.3.15 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. 10.4.3.16 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). MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.4 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): 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 169/320 9 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 170/320 9 MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.5 Interrupt and DMA 10.4.5.1 Timer Interrupt The timer has 5 different Interrupt sources, belonging to 3 independent groups, which are assigned to the following Interrupt vectors: Table 34. 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). 10.4.5.2 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. The two DMA End of Block interrupts are independently enabled by the CP0I and CM0I Interrupt mask bits in the IDMR register. 10.4.5.3 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 86. 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) 171/320 9 MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) Figure 87. 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 10.4.5.4 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. 172/320 9 10.4.5.5 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). MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) 10.4.5.6 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 10.4.5.7). – 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) – 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. 10.4.5.7 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. 10.4.6 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. 173/320 9 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 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 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 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 R7 7 R8 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. 174/320 9 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 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 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 175/320 9 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 35. 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 35). 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 35. Note: This bit has no effect when the Bivalue Mode is enabled (BM=1). 176/320 9 Table 35. 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 10.4.2.12. 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 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 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. 177/320 9 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 1 178/320 9 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). MULTIFUNCTION TIMER (MFT) MULTIFUNCTION TIMER (Cont’d) OUTPUT B CONTROL REGISTER (OBCR) R253 - Read/Write Register Page: 10 Reset value: 0000 0000 (00h) 7 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 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 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). 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 TxOUTB pin on a successful compare of the CMP1R register Set Toggle Reset NOP 179/320 9 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 4 = CM1: Compare 1 flag. This bit is set after a successful compare on CMP1R register. An interrupt is generated if the 180/320 9 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 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) 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 181/320 9 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 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. 182/320 9 0 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 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. 183/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) 10.5 SERIAL COMMUNICATIONS INTERFACE (SCI) 10.5.1 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 88. 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 184/320 9 DCD SIN VA00169A SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.2 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 89. 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. 185/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.3 SCI Operating Modes 10.5.3.1 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. 10.5.3.2 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 90. 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 186/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.3.3 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. 10.5.3.4 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. 187/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Figure 91. 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. 188/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.4 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 92. 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) 189/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.4.1 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 93, Figure 94, and Figure 95 for these different configurations. Table 36. 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) 190/320 9 Not available in Synchronous mode SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Figure 93. 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 94. 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 95. 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) 191/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.5 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 37. 10.5.6 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 96. SCI Baud Rate Generator Initialization Sequence MOST SIGNIFICANT BYTE INITIALIZATIO N SELECT SCI WORKING MODE LEAST SIGNIFICANT BYTE INITIALIZATIO N 192/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Table 37. 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 38. 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% 193/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.7 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. 10.5.8 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 97. 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 88). 194/320 9 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 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.9 Interrupts and DMA 10.5.9.1 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 98. 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 10.5.3 SCI Operating Modes for more details relating to Synchronous mode. Table 39. SCI Interrupt Internal Priority Receive DMA Request Highest Priority Transmit DMA Request Receive Interrupt Transmit Interrupt Lowest Priority 195/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) Table 40. 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 98. 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 196/320 9 D9=1 DATA DATA DATA DATA D9=1 DATA INTERRUPT DATA INTERRUPT DATA INTERRUPT INTERRUPT INTERRUPT VA00270 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.9.2 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. 197/320 9 SERIAL COMMUNICATIONS INTERFACE (SCI) SERIAL COMMUNICATIONS INTERFACE (Cont’d) 10.5.10 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 198/320 9 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 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. 199/320 9 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. 200/320 9 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. 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. 201/320 9 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. 202/320 9 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). 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. 203/320 9 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- 204/320 9 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 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 205/320 9 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). 206/320 9 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 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” 207/320 9 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” 208/320 9 SERIAL PERIPHERAL INTERFACE (SPI) 10.6 SERIAL PERIPHERAL INTERFACE (SPI) – MISO: Master In Slave Out pin – MOSI: Master Out Slave In pin – SCK: Serial Clock pin – SS: Slave select pin To use any of these alternate functions (input or output), the corresponding I/O port must be programmed as alternate function output. A basic example of interconnections between a single master and a single slave is illustrated on Figure 99. The MOSI pins are connected together as are MISO pins. In this way data is transferred serially between master and slave (most significant bit first). When the master device transmits data to a slave device via MOSI pin, the slave device responds by sending data to the master device via the MISO pin. This implies full duplex transmission with both data out and data in synchronized with the same clock signal (which is provided by the master device via the SCK pin). Thus, the byte transmitted is replaced by the byte received and eliminates the need for separate transmit-empty and receiver-full bits. A status flag is used to indicate that the I/O operation is complete. Four possible data/clock timing relationships may be chosen (see Figure 102) but master and slave must be programmed with the same timing mode. 10.6.1 Introduction The Serial Peripheral Interface (SPI) allows fullduplex, synchronous, serial communication with external devices. An SPI system may consist of a master and one or more slaves or a system in which devices may be either masters or slaves. The SPI is normally used for communication between the microcontroller and external peripherals or another Microcontroller. Refer to the Pin Description chapter for the devicespecific pin-out. 10.6.2 Main Features ■ Full duplex, three-wire synchronous transfers ■ Master or slave operation ■ Four master mode frequencies ■ Maximum slave mode frequency = INTCLK/2. ■ Fully programmable 3-bit prescaler for a wide range of baud rates, plus a programmable divider by 2 ■ Programmable clock polarity and phase ■ End of transfer interrupt flag ■ Write collision flag protection ■ Master mode fault protection capability. 10.6.3 General description The SPI is connected to external devices through 4 alternate pins: Figure 99. Serial Peripheral Interface Master/Slave SLAVE MASTER MSBit LSBit 8-BIT SHIFT REGISTER SPI CLOCK GENERATOR MSBit MISO MISO MOSI MOSI SCK SS LSBit 8-BIT SHIFT REGISTER SCK +5V SS 209/320 9 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) Figure 100. Serial Peripheral Interface Block Diagram Internal Bus Read DR 1 Read Buffer IT request 0 MOSI MISO Ext. INT SR 8-Bit Shift Register SPIF WCOL - MODF - - - - Write SPI STATE CONTROL SCK SS CR SPIE SPOE SPIS MSTR CPOL CPHA SPR1 SPR0 MASTER CONTROL SERIAL CLOCK GENERATOR PR DIV2 ST9 PERIPHERAL CLOCK (INTCLK) 210/320 9 1/2 0 1 PRESCALER /1 .. /8 PRS2 PRS1 PRS0 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) 10.6.4 Functional Description Figure 100 shows the serial peripheral interface (SPI) block diagram. This interface contains 4 dedicated registers: – A Control Register (CR) – A Prescaler Register (PR) – A Status Register (SR) – A Data Register (DR) Refer to the CR, PR, SR and DR registers in Section 10.6.6for the bit definitions. 10.6.4.1 Master Configuration In a master configuration, the serial clock is generated on the SCK pin. Procedure – Define the serial clock baud rate by setting/resetting the DIV2 bit of PR register, by writing a prescaler value in the PR register and programming the SPR0 & SPR1 bits in the CR register. – Select the CPOL and CPHA bits to define one of the four relationships between the data transfer and the serial clock (see Figure 102). – The SS pin must be connected to a high level signal during the complete byte transmit sequence. – The MSTR and SPOE bits must be set (they remain set only if the SS pin is connected to a high level signal). In this configuration the MOSI pin is a data output and to the MISO pin is a data input. Transmit sequence The transmit sequence begins when a byte is written the DR register. The data byte is parallel loaded into the 8-bit shift register (from the internal bus) during a write cycle and then shifted out serially to the MOSI pin most significant bit first. When data transfer is complete: – The SPIF bit is set by hardware – An interrupt is generated if the SPIS and SPIE bits are set. During the last clock cycle the SPIF bit is set, a copy of the data byte received in the shift register is moved to a buffer. When the DR register is read, the SPI peripheral returns this buffered value. Clearing the SPIF bit is performed by the following software sequence: 1. An access to the SR register while the SPIF bit is set 2. A write or a read of the DR register. Note: While the SPIF bit is set, all writes to the DR register are inhibited until the SR register is read. 211/320 9 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) 10.6.4.2 Slave Configuration In slave configuration, the serial clock is received on the SCK pin from the master device. The value of the PR register and SPR0 & SPR1 bits in the CR is not used for the data transfer. Procedure – For correct data transfer, the slave device must be in the same timing mode as the master device (CPOL and CPHA bits). See Figure 102. – The SS pin must be connected to a low level signal during the complete byte transmit sequence. – Clear the MSTR bit and set the SPOE bit to assign the pins to alternate function. In this configuration the MOSI pin is a data input and the MISO pin is a data output. Transmit Sequence The data byte is parallel loaded into the 8-bit shift register (from the internal bus) during a write cycle and then shifted out serially to the MISO pin most significant bit first. The transmit sequence begins when the slave device receives the clock signal and the most significant bit of the data on its MOSI pin. 212/320 9 When data transfer is complete: – The SPIF bit is set by hardware – An interrupt is generated if the SPIS and SPIE bits are set. During the last clock cycle the SPIF bit is set, a copy of the data byte received in the shift register is moved to a buffer. When the DR register is read, the SPI peripheral returns this buffered value. Clearing the SPIF bit is performed by the following software sequence: 1. An access to the SR register while the SPIF bit is set. 2. A write or a read of the DR register. Notes: While the SPIF bit is set, all writes to the DR register are inhibited until the SR register is read. The SPIF bit can be cleared during a second transmission; however, it must be cleared before the second SPIF bit in order to prevent an overrun condition (see Section 10.6.4.6). Depending on the CPHA bit, the SS pin has to be set to write to the DR register between each data byte transfer to avoid a write collision (see Section 10.6.4.4). SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) 10.6.4.3 Data Transfer Format During an SPI transfer, data is simultaneously transmitted (shifted out serially) and received (shifted in serially). The serial clock is used to synchronize the data transfer during a sequence of eight clock pulses. The SS pin allows individual selection of a slave device; the other slave devices that are not selected do not interfere with the SPI transfer. Clock Phase and Clock Polarity Four possible timing relationships may be chosen by software, using the CPOL and CPHA bits. The CPOL (clock polarity) bit controls the steady state value of the clock when no data is being transferred. This bit affects both master and slave modes. The combination between the CPOL and CPHA (clock phase) bits selects the data capture clock edge. Figure 102, shows an SPI transfer with the four combinations of the CPHA and CPOL bits. The diagram may be interpreted as a master or slave timing diagram where the SCK pin, the MISO pin, the MOSI pin are directly connected between the master and the slave device. The SS pin is the slave device select input and can be driven by the master device. The master device applies data to its MOSI pinclock edge before the capture clock edge. CPHA bit is set The second edge on the SCK pin (falling edge if the CPOL bit is reset, rising edge if the CPOL bit is set) is the MSBit capture strobe. Data is latched on the occurrence of the first clock transition. No write collision should occur even if the SS pin stays low during a transfer of several bytes (see Figure 101). CPHA bit is reset The first edge on the SCK pin (falling edge if CPOL bit is set, rising edge if CPOL bit is reset) is the MSBit capture strobe. Data is latched on the occurrence of the second clock transition. This pin must be toggled high and low between each byte transmitted (see Figure 101). To protect the transmission from a write collision a low value on the SS pin of a slave device freezes the data in its DR register and does not allow it to be altered. Therefore the SS pin must be high to write a new data byte in the DR without producing a write collision. Figure 101. CPHA / SS Timing Diagram MOSI/MISO Byte 1 Byte 2 Byte 3 Master SS Slave SS (CPHA=0) Slave SS (CPHA=1) 213/320 9 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) Figure 102. Data Clock Timing Diagram CPHA =1 CPOL = 1 CPOL = 0 MISO (from master) MOSI (from slave) MSBit Bit 6 Bit 5 Bit 4 Bit3 Bit 2 Bit 1 LSBit MSBit Bit 6 Bit 5 Bit 4 Bit3 Bit 2 Bit 1 LSBit SS (to slave) CAPTURE STROBE CPHA =0 CPOL = 1 CPOL = 0 MISO (from master) MOSI (from slave) MSBit MSBit Bit 6 Bit 5 Bit 4 Bit3 Bit 2 Bit 1 LSBit Bit 6 Bit 5 Bit 4 Bit3 Bit 2 Bit 1 LSBit SS (to slave) CAPTURE STROBE Note: This figure should not be used as a replacement for parametric information. Refer to the Control Timing chapter. 214/320 9 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) 10.6.4.4 Write Collision Error A write collision occurs when the software tries to write to the DR register while a data transfer is taking place with an external device. When this happens, the transfer continues uninterrupted and the software write will be unsuccessful. Write collisions can occur both in master and slave mode. Note: a ”read collision” will never occur since the received data byte is placed in a buffer in which access is always synchronous with the MCU operation. In Slave mode When the CPHA bit is set: The slave device will receive a clock (SCK) edge prior to the latch of the first data transfer. This first clock edge will freeze the data in the slave device DR register and output the MSBit on to the external MISO pin of the slave device. The SS pin low state enables the slave device but the output of the MSBit onto the MISO pin does not take place until the first data transfer clock edge. When the CPHA bit is reset: Data is latched on the occurrence of the first clock transition. The slave device does not have any way of knowing when that transition will occur; therefore, the slave device collision occurs when software attempts to write the DR register after its SS pin has been pulled low. For this reason, the SS pin must be high, between each data byte transfer, to allow the CPU to write in the DR register without generating a write collision. In Master mode Collision in the master device is defined as a write of the DR register while the internal serial clock (SCK) is in the process of transfer. The SS pin signal must be always high on the master device. WCOL bit The WCOL bit in the SR register is set if a write collision occurs. No SPI interrupt is generated when the WCOL bit is set (the WCOL bit is a status flag only). Clearing the WCOL bit is done through a software sequence (see Figure 103). Figure 103. Clearing the WCOL bit (Write Collision Flag) Software Sequence Clearing sequence after SPIF = 1 (end of a data byte transfer) 1st Step Read SR OR Read SR THEN THEN 2nd Step Read DR SPIF =0 WCOL=0 Write DR SPIF =0 WCOL=0 if no transfer has started WCOL=1 if a transfer has started before the 2nd step Clearing sequence before SPIF = 1 (during a data byte transfer) 1st Step Read SR THEN 2nd Step Read DR WCOL=0 Note: Writing in DR register instead of reading in it do not reset WCOL bit 215/320 9 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) 10.6.4.5 Master Mode Fault Master mode fault occurs when the master device has its SS pin pulled low, then the MODF bit is set. Master mode fault affects the SPI peripheral in the following ways: – The MODF bit is set and an SPI interrupt is generated if the SPIE bit is set. – The SPOE bit is reset. This blocks all output from the device and disables the SPI peripheral. – The MSTR bit is reset, thus forcing the device into slave mode. Clearing the MODF bit is done through a software sequence: 1. A read or write access to the SR register while the MODF bit is set. 2. A write to the CR register. Notes: To avoid any multiple slave conflicts in the case of a system comprising several MCUs, the SS pin must be pulled high during the clearing sequence of the MODF bit. The SPOE and MSTR 216/320 9 bits may be restored to their original state during or after this clearing sequence. Hardware does not allow the user to set the SPOE and MSTR bits while the MODF bit is set except in the MODF bit clearing sequence. In a slave device the MODF bit can not be set, but in a multi master configuration the device can be in slave mode with this MODF bit set. The MODF bit indicates that there might have been a multi-master conflict for system control and allows a proper exit from system operation to a reset or default system state using an interrupt routine. 10.6.4.6 Overrun Condition An overrun condition occurs, when the master device has sent several data bytes and the slave device has not cleared the SPIF bit issuing from the previous data byte transmitted. In this case, the receiver buffer contains the byte sent after the SPIF bit was last cleared. A read to the DR register returns this byte. All other bytes are lost. This condition is not detected by the SPI peripheral. SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) 10.6.4.7 Single Master and Multimaster Configurations There are two types of SPI systems: For more security, the slave device may respond to the master with the received data byte. Then the – Single Master System master will receive the previous byte back from the – Multimaster System slave device if all MISO and MOSI pins are connected and the slave has not written its DR register. Single Master System Other transmission security methods can use A typical single master system may be configured, ports for handshake lines or data bytes with comusing an MCU as the master and four MCUs as mand fields. slaves (see Figure 104). Multi-master System The master device selects the individual slave deA multi-master system may also be configured by vices by using four pins of a parallel port to control the user. Transfer of master control could be imthe four SS pins of the slave devices. plemented using a handshake method through the The SS pins are pulled high during reset since the I/O ports or by an exchange of code messages master device ports will be forced to be inputs at through the serial peripheral interface system. that time, thus disabling the slave devices. The multi-master system is principally handled by the MSTR bit in the CR register and the MODF bit Note: To prevent a bus conflict on the MISO line in the SR register. the master allows only one slave device during a transmission. Figure 104. Single Master Configuration SS SCK SS SS SCK Slave MCU Slave MCU MOSI MISO MOSI MISO SS SCK Slave MCU SCK Slave MCU MOSI MISO MOSI MISO SCK Master MCU 5V Ports MOSI MISO SS 217/320 9 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) 10.6.5 Interrupt Management The interrupt of the Serial Peripheral Interface is mapped on one of the eight External Interrupt Channels of the microcontroller (refer to the “Interrupts” chapter). Each External Interrupt Channel has: – A trigger control bit in the EITR register (R242 Page 0), – A pending bit in the EIPR register (R243 Page0), – A mask bit in the EIMR register (R244 - Page 0). Program the interrupt priority level using the EIPLR register (R245 - Page 0). For a description of these registers refer to the “Interrupts” and “DMA” chapters. To use the interrupt feature, perform the following sequence: – Set the priority level of the interrupt channel used for the SPI (EIPRL register) – Select the interrupt trigger edge as rising edge (set the corresponding bit in the EITR register) – Set the SPIS bit of the CR register to select the peripheral interrupt source – Set the SPIE bit of the CR register to enable the peripheral to perform interrupt requests – In the EIPR register, reset the pending bit of the interrupt channel used by the SPI interrupt to avoid any spurious interrupt requests being performed when the mask bit is set – Set the mask bit of the interrupt channel used to enable the MCU to acknowledge the interrupt requests of the peripheral. 218/320 9 Note: In the interrupt routine, reset the related pending bit to avoid the interrupt request that was just acknowledged being proposed again. Then, after resetting the pending bit and before the IRET instruction, check if the SPIF and MODF interrupt flags in the SR register) are reset; otherwise jump to the beginning of the routine. If, on return from an interrupt routine, the pending bit is reset while one of the interrupt flags is set, no interrupt is performed on that channel until the flags are set. A new interrupt request is performed only when a flag is set with the other not set. 10.6.5.1 Register Map Depending on the device, one or two Serial Peripheral interfaces can be present. The previous table summarizes the position of the registers of the two peripherals in the register map of the microcontroller. SPI0 SPI1 Address Page Name R240 (F0h) 7 DR0 R241 (F1h) 7 CR0 R242 (F2h) 7 SR0 R243 (F3h) 7 PR0 R248 (F8h) 7 DR1 R249 (F9h) 7 CR1 R250 (FAh) 7 SR1 R251 (FBh) 7 PR1 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) 10.6.6 Register Description DATA REGISTER (SPDR) R240 - Read/Write Register Page: 7 Reset Value: 0000 0000 (00h) 7 D7 Note: To use the MISO, MOSI and SCK alternate functions (input or output), the corresponding I/O port must be programmed as alternate function output. 0 D6 D5 D4 D3 D2 D1 D0 The DR register is used to transmit and receive data on the serial bus. In the master device only a write to this register will initiate transmission/reception of another byte. Notes: During the last clock cycle the SPIF bit is set, a copy of the received data byte in the shift register is moved to a buffer. When the user reads the serial peripheral data register, the buffer is actually being read. Warning: A write to the DR register places data directly into the shift register for transmission. A read to the DR register returns the value located in the buffer and not the content of the shift register (see Figure 100). CONTROL REGISTER (SPCR) R241 - Read/Write Register Page: 7 Reset Value: 0000 0000 (00h) 7 SPIE 0 SPOE SPIS MSTR CPOL CPHA SPR1 SPR0 Bit 7 = SPIE Serial peripheral interrupt enable. This bit is set and cleared by software. 0: Interrupt is inhibited 1: An SPI interrupt is generated whenever either SPIF or MODF are set in the SR register while the other flag is 0. Bit 5 = SPIS Interrupt Selection. This bit is set and cleared by software. 0: Interrupt source is external interrupt 1: Interrupt source is SPI Bit 4 = MSTR Master. This bit is set and cleared by software. It is also cleared by hardware when, in master mode, SS=0 (see Section 10.6.4.5 Master Mode Fault). 0: Slave mode is selected 1: Master mode is selected, the function of the SCK pin changes from an input to an output and the functions of the MISO and MOSI pins are reversed. Bit 3 = CPOL Clock polarity. This bit is set and cleared by software. This bit determines the steady state of the serial Clock. The CPOL bit affects both the master and slave modes. 0: The steady state is a low value at the SCK pin. 1: The steady state is a high value at the SCK pin. Bit 2 = CPHA Clock phase. This bit is set and cleared by software. 0: The first clock transition is the first data capture edge. 1: The second clock transition is the first capture edge. Bit 1:0 = SPR[1:0] Serial peripheral rate. These bits are set and cleared by software. They select one of four baud rates to be used as the serial clock when the device is a master. These 2 bits have no effect in slave mode. Table 41. Serial Peripheral Baud Rate Bit 6 = SPOE Serial peripheral output enable. This bit is set and cleared by software. It is also cleared by hardware when, in master mode, SS=0 (see Section 10.6.4.5 Master Mode Fault). 0: SPI alternate functions disabled (MISO, MOSI and SCK can only work as input) 1: SPI alternate functions enabled (MISO, MOSI and SCK can work as input or output depending on the value of MSTR) INTCLK Clock Divide 2 4 16 32 SPR1 0 0 1 1 SPR0 0 1 0 1 219/320 9 SERIAL PERIPHERAL INTERFACE (SPI) SERIAL PERIPHERAL INTERFACE (Cont’d) STATUS REGISTER (SPSR) R242 - Read Only Register Page: 7 Reset Value: 0000 0000 (00h) 7 SPIF 1: A fault in master mode has been detected Bits 3:0 = Unused. 0 WCOL - MODF - - - - Bit 7 = SPIF Serial Peripheral data transfer flag. This bit is set by hardware when a transfer has been completed. An interrupt is generated if SPIE=1 in the CR register. It is cleared by a software sequence (an access to the SR register followed by a read or write to the DR register). 0: Data transfer is in progress or has been approved by a clearing sequence. 1: Data transfer between the device and an external device has been completed. Note: While the SPIF bit is set, all writes to the DR register are inhibited. PRESCALER REGISTER (SPPR) R243 - Read/Write Register Page: 7 Reset Value: 0000 0000 (00h) 7 0 0 0 0 DIV2 0 PRS2 PRS1 PRS0 Bit 7:5 = Reserved, forced by hardware to 0. Bit 4 = DIV2 Divider enable. This bit is set and cleared by software. 0: Divider by 2 enabled. 1: Divider by 2 disabled. Bit 3 = Reserved. forced by hardware to 0. Bit 6 = WCOL Write Collision status. This bit is set by hardware when a write to the DR register is done during a transmit sequence. It is cleared by a software sequence (see Figure 103). 0: No write collision occurred 1: A write collision has been detected Bit 5 = Unused. Bit 2:0 = PRS[2:0] Prescaler Value. These bits are set and cleared by software. The baud rate generator is driven by INTCLK/(n1*n2*n3) where n1= PRS[2:0]+1, n2 is the value of the SPR[1:0] bits, n3 = 1 if DIV2=1 and n3= 2 if DIV2=0. Refer to Figure 100. These bits have no effect in slave mode. Table 42. Prescaler Baud Rate Bit 4 = MODF Mode Fault flag. This bit is set by hardware when the SS pin is pulled low in master mode (see Section 10.6.4.5 Master Mode Fault). An SPI interrupt can be generated if SPIE=1 in the CR register. This bit is cleared by a software sequence (An access to the SR register while MODF=1 followed by a write to the CR register). 0: No master mode fault detected 220/320 9 Prescaler Division Factor PRS2 PRS1 PRS0 1 (no division) 0 0 0 2 0 0 1 ... 8 1 1 1 I2C BUS INTERFACE 10.7 I2C BUS INTERFACE 10.7.1 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. 10.7.2 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. 221/320 9 I2C BUS INTERFACE I2C INTERFACE (Cont’d) Figure 105. 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 10.7.3 Functional Description Refer to the I2CCR, I2CSR1 and I2CSR2 registers in Section 10.7.7. 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. 222/320 9 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 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 106. 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. 10.7.3.1 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 223/320 9 I2C BUS INTERFACE I2C INTERFACE (Cont’d) Figure 106. 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. 10.7.4 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). 224/320 9 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. 10.7.4.1 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 107 Transfer sequencing EV1). 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 107 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 107 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. 10.7.4.2 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 107 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. 225/320 9 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 107 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 107 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. 226/320 9 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. I2C BUS INTERFACE I2C INTERFACE (Cont’d) Figure 107. 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. 227/320 9 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 108. 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 228/320 9 READY TO TRANSMIT or END OF BLOCK INTERRUPT REQUEST I2C BUS INTERFACE I2C INTERFACE (Cont’d) 10.7.5 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. 229/320 9 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. 10.7.6 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- 230/320 9 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. I2C BUS INTERFACE I2C INTERFACE (Cont’d) 10.7.6.1 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. 10.7.6.2 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. 10.7.6.3 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). 231/320 9 I2C BUS INTERFACE I2C INTERFACE (Cont’d) 10.7.7 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) 232/320 9 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. 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 107) 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 107) 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) 233/320 9 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). 234/320 9 – 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 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. 235/320 9 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. 236/320 9 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). 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. 237/320 9 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. 238/320 9 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). 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 10.7.6.1 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 10.7.6.1 and Section 10.7.6.2 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 239/320 9 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 10.7.6.2 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 10.7.6.1 and Section 10.7.6.2 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). 240/320 9 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 241/320 9 I2C BUS INTERFACE I2C INTERFACE (Cont’d) Table 43. 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 242/320 9 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 J1850 Byte Level Protocol Decoder (JBLPD) 10.8 J1850 Byte Level Protocol Decoder (JBLPD) 10.8.1 Introduction The JBLPD is used to exchange data between the ST9 microcontroller and an external J1850 transceiver I.C. The JBLPD transmits a string of variable pulse width (VPW) symbols to the transceiver. It also receives VPW encoded symbols from the transceiver, decodes them and places the data in a register. In-frame responses of type 0, 1, 2 and 3 are supported and the appropriate normalization bit is generated automatically. The JBLPD filters out any incoming messages which it does not care to receive. It also includes a programmable external loop delay. The JBLPD uses two signals to communicate with the transceiver: – VPWI (input) – VPWO (output) 10.8.2 Main Features ■ SAE J1850 compatible ■ Digital filter ■ In-Frame Responses of type 0, 1, 2, 3 supported with automatic normalization bit ■ Programmable External Loop Delay ■ Diagnostic 4x time mode ■ Diagnostic Local Loopback mode ■ Wide range of MCU internal frequencies allowed ■ Low power consumption mode (JBLPD suspended) ■ Very low power consumption mode (JBLPD disabled) ■ Don’t care message filter ■ Selectable VPWI input polarity ■ Selectable Normalization Bit symbol form ■ 6 maskable interrupts ■ DMA transmission and reception with End Of Block interrupts 243/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Figure 109. JBLPD Byte Level Protocol Decoder Block Diagram RXDATA VPW DIGITAL DECODER FILTER VPWI pin ERROR ARBITRATION CHECKER CONTROL I.D. Filter FREG[0:31] VPWI_LOOP STATUS JBLPD STATE MACHINE OPTIONS CRC LOOPBACK GENERATOR LOGIC TXOP CLOCK PRESCALER Prescaled Clock (Encoder/Decoder Clock) VPWO_LOOP CLKSEL CRC BYTE CRC\ BYTE MUX PADDR TXDATA Interrupt & DMA Logic and Registers 244/320 9 VPW ENCODER VPWO pin J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) nate any transmissions in progress, and disable 10.8.3 Functional Description receive transfers and RDRF flags until the VPW 10.8.3.1 J1850 protocol symbols decoder recognizes a valid EOF symbol from the J1850 symbols are defined as a duration (in microbus. seconds or clock cycles) and a state which can be The JBLPD’s state machine handles all the Tv either an active state (logic high level on VPWO) l.D.s in accordance with the SAE J1850 specificaor a passive state (logic low level on VPWO). tion. An idle J1850 bus is in a passive state. Note: Depending on the value of a control bit, the Any symbol begins by changing the state of the polarity of the VPWI input can be the same as the VPW line. The line is in this state for a specific duJ1850 bus or inverted with respect to it. ration depending on the symbol being transmitted. Durations, and hence symbols, are measured as time between successive state transitions. Each Table 44. J1850 Symbol definitions symbol has only one level transition of a specific duration. Symbol Definition Symbols for logic zero and one data bits can be eiPassive for Tv1 or ther a high or a low level, but all other symbols are Data Bit Zero Active for Tv2 defined at only one level. Each symbol is placed directly next to another. Therefore, every level transition means that another symbol has begun. Data bits of a logic zero are either a short duration if in a passive state or a long duration if in an active state. Data bits of a logic one are either a long duration if in a passive state or a short duration if in an active state. This ensures that data logic zeros predominate during bus arbitration. An eight bit data byte transmission will always have eight transitions. For all data byte and CRC byte transfers, the first bit is a passive state and the last bit is an active state. For the duration of the VPW, symbols are expressed in terms of Tv’s (or VPW mode timing values). J1850 symbols and Tv values are described in the SAE J1850 specification, in Table 44 and in Table 45. An ignored Tv I.D. occurs for level transitions which occur in less than the minimum time required for an invalid bit detect. The VPW encoder does not recognize these characters as they are filtered out by the digital filter. The VPW decoder does not resynchronize its counter with either edge of “ignored” pulses. Therefore, the counter which times symbols continues to time from the last transition which occurred after a valid symbol (including the invalid bit symbol) was recognized. A symbol recognized as an invalid bit will resynchronize the VPW decoder to the invalid bit edges. In the case of the reception of an invalid bit, the JBLPD peripheral will set the IBD bit in the ERROR register. The JBLPD peripheral shall termi- Start of Frame (SOF) Passive for Tv2 or Active for Tv1 Active for Tv3 End of Data (EOD) End of Frame (EOF) Passive for Tv3 Passive for Tv4 Data Bit One Inter Frame Separation Passive for Tv6 (IFS) IDLE Bus Condition (IDLE) Passive for > Tv6 Normalization Bit (NB) Break (BRK) Active for Tv1 or Tv2 Active for Tv5 Table 45. J1850 VPW Mode Timing Value (Tv) definitions (in clock cycles) Pulse Minimum Nominal Width or Tv Duration Duration I.D. Ignored 0 N/A <=7 Invalid Bit Tv1 >7 >34 N/A 64 <=34 <=96 Tv2 >96 128 <=163 Tv3 >163 200 <=239 Tv4 Tv5 >239 >239 280 300 N/A N/A Tv6 >280 300 N/A Maximum Duration 245/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) 10.8.3.2 Transmitting Messages chronize off the decoder output to time the VPWO symbol time. This section describes the general procedures used by the JBLPD to successfully transmit J1850 A detailed description of the JBLPD opcodes can frames of data out the VPWO pin. The first five be find in the description of the OP[2:0] bits in the sub-sections describe the procedures used for TXOP register. transmitting the specific transmit data types. The last section goes into the details of the transmitted Message Byte String Transmission (Type 0 symbol timing, synchronizing of symbols received IFR) from the external J1850 bus, and how data bit arbitration works. Message byte transmitting is the outputting of data bytes on the VPWO pin that occurs subsequent to The important concept to note for transmitting data a received bus idle condition. All message byte is: the activity sent over the VPWO line should be strings start with a SOF symbol transmission, then timed with respect to the levels and transitions one or more data bytes are transmitted. A CRC seen on the filtered VPWI line. byte is then transmitted followed by an EOD symThe J1850 bus is a multiplexed bus, and the bol (see Figure 110) to complete the transmission. VPWO & VPWI pins interface to this bus through a If transmission is queued while another frame is transceiver I.C. Therefore, the propagation delay being received, then the JBLPD will time an Interthrough the transceiver I.C. and external bus filterFrame Separation (IFS) time (Tv6) before coming must be taken into account when looking for mencing with the SOF character. transmitted edges to appear back at the receiver. The user program will decide at some point that it The external propagation delay for an edge sent wants to initiate a message byte string. The user out on the VPWO line, to be detected on the VPWI program writes the TXDATA register with the first line is denoted as Tp-ext and is programmable bemessage data byte to be transmitted. Next, the tween 0 and 31 µs nominal via the JDLY[4:0] bits TXOP register is written with the MSG opcode if in CONTROL register. more than one data byte is contained within the The transmitter VPW encoder sets the proper level message, or with MSG+CRC opcode if one data to be sent out the VPWO line. It then waits for the byte is to be transmitted. The action of writing the corresponding level transition to be reflected back TXOP register causes the TRDY bit to be cleared at the VPW decoder input. signifying that the TXDATA register is full and a Taking into account the external loop delay (Tp-ext) corresponding opcode has been queued. The and the digital filter delay, the encoder will time its JBLPD must wait for an EOF nominal time period output to remain at this level so that the received at which time data is transferred from the TXDATA symbol is at the correct nominal symbol time (refer register to the transmit shift register. The TRDY bit to “Transmit Opcode Queuing” section). If arbitrais again set since the TXDATA register is empty. tion is lost at any time during bit 0 or bit 1 transmisThe JBLPD should also begin transmission if ansion, then the VPWO line goes passive. At the end other device begins transmitting early. As long as of the symbol time on VPWO, the encoder changan EOF minimum time period elapses, the JBLPD es the state of VPWO if any more information is to should begin timing and asserting the SOF symbol be transmitted. It then times the new state change with the intention of arbitrating for the bus during from the receiver decoder output. the transmission of the first data byte. If a transmit Note that depending on the symbol (especially the is requested during an incoming SOF symbol, the SOF, NB0, NB1 symbols) the decoder output may JBLPD should be able to synchronize itself to the actually change to the desired state before the incoming SOF up to a time of Tv1 max. (96 µs) into transmit is attempted. It is important to still synthe SOF symbol before declaring that it was too late to arbitrate for this frame. 246/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) register except during DMA transfers (see Section If the J1850 bus was IDLE at the time the first data 10.8.6.4 DMA Management in Transmission byte and opcode are written, the transmitter will Mode). immediately transfer data from the TXDATA register to the transmit shift register. The TRDY bit will once again be set signifying the readiness to acTransmitting a Type 1 IFR cept a new data byte. The second data byte can then be written followed by the respective opcode. The user program will decide to transmit an IFR In the case of the last data byte, the TXOP register type 1 byte in response to a message which is curshould be written with the MSG+CRC opcode. The rently being received (See Figure 111). It does so transmitter will transmit the internally generated by writing the IFR1 opcode to the TXOP register. CRC after the last bit of the data byte. Once the Transmitting IFR data type 1 requires only a single TRDY bit is set signifying the acceptance of the write of the TXOP register with the IFR1 opcode last data byte, the first byte of the next message set. The MLC[3:0] bits should be set to the proper can be queued by writing the TXDATA register fol“byte-received-count-required-before-IFR’ing” vallowed by a TXOP register write. The block will wait ue. If no error conditions (IBD, IFD, TRA, RBRK or until the current data and the CRC data byte are CRCE) exist to prevent transmission, the JBLPD sent out and a new IFS has expired before transperipheral will then transmit out the contents of the mitting the new data. This is the case even if IFR PADDR register at the next EOD nominal time pedata reception takes place in the interim. riod or at a time greater than the EOD minimum time period if a falling edge is detected on filtered Lost arbitration any time during the transfer of type J1850 bus line signifying another transmitter is be0 data will be honoured by immediately relinquishginning early. The NB1 symbol precedes the PADing control to the higher priority message. The TLA DR register value and is followed with an EOF debit in the STATUS register is set accordingly and limiter. The TRDY flag is cleared on the write of the an interrupt will be generated assuming the TXOP register. The TRDY bit is set once the NB1 TLA_M bit in the IMR register is set. It is responsibegins transmitting. bility of the user program to re-send the message beginning with the first byte if desired. This may be Although the JBLPD should never lose arbitration done at any time by rewriting only the TXOP regisfor data in the IFR portion of a type 1 frame, higher ter if the TXDATA contents have not changed. priority messages are always honoured under the rules of arbitration. If arbitration is lost then the Any transmitted data and CRC bytes during the VPWO line is set to the passive state. The TLA bit transmit frame will also be received and transin the STATUS register is set accordingly and an ferred to the RXDATA register if the corresponding interrupt will be generated if enabled. The IFR1 is message filter bit is set in the FREG[0:31] regisnot retried. It is lost if the JBLPD peripheral loses ters. If the corresponding bit is not set in arbitration. Also, the data that made it out on the FREG[0:31], then the transmitted data is also not bus will be received in the RXDATA register if not transferred to RXDATA. Also, the RDRF will not put into sleep mode. Note that for the transmitter to get set during frame and receive events such as synchronize to the incoming signals of a frame, an RDOF & EODM. IFR should be queued before an EODM is reNOTE: The correct procedure for transmitting is to ceived for the present frame. write first the TXDATA register and then the TXOP 247/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) is currently being received (See Figure 113). It Transmitting a Type 2 IFR does so by writing the IFR3 or IFR3+CRC opcode The user program will decide to transmit an IFR to the TXOP register. Transmitting IFR data type 3 type 2 byte in response to a message which is curis similar to transmitting a message, in that the TXrently being received (See Figure 112). It does so DATA register is written with the first data byte folby writing the IFR2 opcode to the TXOP register. lowed by a TXOP register write. For a single data Transmitting IFR data type 2 requires only a single byte IFR3 transmission, the TXOP register would write of the TXOP register with the IFR2 opcode be written with IFR3+CRC opcode set. The set. The MLC[3:0] bits can also be set to check for MLC[3:0] bits can also be set to a proper value to message length errors. If no error conditions (IBD, check for message length errors before enabling IFD, TRA, RBRK or CRCE) exist to prevent transthe IFR transmit. mission, the JBLPD will transmit out the contents If no error conditions (IBD, IFD, TRA, RBRK or of the PADDR register at the next EOD nominal CRCE) exist to prevent transmission, the JBLPD time period or after an EOD minimum time period if will wait for an EOD nominal time period on the fila rising edge is detected on the filtered VPWI line tered VPWI line (or for at least an EOD minimum signifying another transmitter beginning early. The time followed by a rising edge signifying another NB1 symbol precedes the PADDR register value transmitter beginning early) at which time data is and is followed with an EOF delimiter. The TRDY transferred from the TXDATA register to the transflag will be cleared on the write of the TXOP regismit shift register. The TRDY bit is set since the TXter. The TRDY bit is set once the NB1 begins DATA register is empty. A NB0 symbol is output transmitting. on the VPWO line followed by the data byte and Lost arbitration for this case is a normal occurpossibly the CRC byte if a IFR3+CRC opcode was rence since type 2 IFR data is made up of single set. Once the first IFR3 byte has been successfully bytes from multiple responders. If arbitration is lost transmitted, successive IFR3 bytes are sent with the VPWO line is released and the JBLPD waits TXDATA/TXOP write sequences where the until the byte on the VPWI line is completed. Note MLC[3:O] bits are don’t cares. The final byte in the that the IFR that did make it out on the bus will be IFR3 string must be transmitted with the received in the RXDATA register if it is not put into IFR3+CRC opcode to trigger the JBLPD to apsleep mode. Then, the JBLPD re-attempts to send pend the CRC byte to the string. The user program its physical address immediately after the end of may queue up the next message opcode sethe last byte. The TLA bit is not set if arbitration is quence once the TRDY bit has been set. lost and the user program does not need to reAlthough arbitration should never be lost for data queue data or an opcode. The JBLPD will re-atin the IFR portion of a type 3 frame, higher priority tempt to send its PADDR register contents until it messages are always honoured under the rules of successfully does so or the 12-byte frame maxiarbitration. If arbitration is lost then the block mum is reached if NFL=0. If NFL=1, then re-atshould relinquish the bus by taking the VPWO line tempts to send an lFR2 are executed until canto the passive state. In this case the TLA bit in the celled by the CANCEL opcode or a JBLPD disaSTATUS register is set, and an interrupt will be ble. Note that for the transmitter to synchronize to generated if enabled. Note also, that the IFR data the incoming signals of a frame, an IFR should be that did make it out on the bus will be received in queued before an EODM is received for the the RXDATA register if not in sleep mode. Note present frame. that for the transmitter to synchronize to the incoming signals of a frame, an IFR should be Transmitting a Type 3 lFR Data String queued before an EODM is received for the current frame. The user program will decide to transmit an IFR type 3 byte string in response to a message which 248/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Figure 110. J1850 String Transmission Type 0 Frame Message SOF I.D. Byte Data byte(s) (if any) CRC EOF Figure 111. J1850 String Transmission Type 1 Frame IFR to be sent Message Rx’d from Another Node SOF I.D. Byte Data byte(s) (if any) CRC EOD NB1 IFR Byte EOF Figure 112. J1850 String Transmission Type 2 Frame IFR to be sent Message Rx’d from Another Node SOF I.D. Byte Data byte(s) (if any) CRC EOD NB1 IFR Byte ... ... IFR Byte EOF Figure 113. J1850 String Transmission Type 3 Frame IFR to be sent Message Rx’d from Another Node SOF I.D. Byte Data byte(s) (if any) CRC CRC EOD NB0 IFR Data Byte(s) Byte EOF 249/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) JBLPD has a receiver pin which tells the transmitTransmit Opcode Queuing ter about bus activity. Due to characteristics of the The JBLPD has the capability of queuing opcode J1850 bus and the eight-clock digital filter, the sigtransmits written to the TXOP register until J1850 nals presented to the VPW symbol decoder are bus conditions are in a correct state for the transdelayed a certain amount of time behind the actual mit to occur. For example, a MSGx opcode can be J1850 bus. Also, due to wave shaping and other queued when the JBLPD is presently receiving a signal conditioning of the transceiver I.C. the acframe (or transmitting a MSG+CRC opcode) or an tions of the VPWO pin on the transmitter take time IFRx opcode can be queued when currently reto appear on the bus itself. The total external ceiving or transmitting the message portion of a J1850 bus delays are defined in the SAE J1850 frame. standard as nominally 16 µs. The nominal 16 µs Queuing a MSG or MSG+CRC opcode for the next loop delay will actually vary between different frame can occur while another frame is in transceiver I.C’s. The JBLPD peripheral thus inprogress. A MSGx opcode is written to the TXOP cludes a programmability of the external loop deregister when the present frame is past the point lay in the bit positions JDLY[4:0]. This assures where arbitration for control of the bus for this only nominal transmit symbols are placed on the frame can occur. The JBLPD will wait for a nomibus by the JBLPD. nal IFS symbol (or EOFmin if another node begins The method of transmitting for the JBLPD includes early) to appear on the VPWI line before cominteraction between the transmitter and the receivmencing to transmit this queued opcode. The er. The transmitter starts a symbol by placing the TRDY bit for the queued opcode will remain clear proper level (active or passive) on its VPWO pin. until the EOFmin is detected on the VPWI line The transmitter then waits for the corresponding where it will then get set. Queued MSGx transmits pin transition (inverted, of course) at the VPW defor the next frame do not get cancelled for TLA, coder input. Note that the level may actually apIBD, IFD or CRCE errors that occur in the present pear at the input before the transmitter places the frame. An RBRK error will cancel a queued opvalue on the VPWO pin. Timing of the remainder code for the next frame. of the symbol starts when the transition is detectQueuing an IFRx opcode for the present frame ed. Refer to Figure 115, Case 1. The symbol timecan occur at any time after the detection of the beout value is defined as: ginning of an SOF character from the VPWI line. SymbolTimeout = NominalSymbolTime - ExternalLoopThe queued IFR will wait for a nominal EOD symDelay - 8 µs bol (or EODmin if another node begins early) before commencing to transmit the IFR. A queued NominalSymbolTime = Tv Symbol time ExternalLoopDelay = defined via JDLY[4:0] IFR transmit will be cancelled on IBD, lFD, CRCE, 8 µs = Digital Filter RBRK errors as well as on a correct message Bit-by-bit arbitration must be used to settle the length check error or frame length limit violation if conflicts that occur when multiple nodes attempt to these checks are enabled. transmit frames simultaneously. Arbitration is applied to each data bit symbol transmitted starting Transmit Bus Timing, Arbitration, and Synafter the SOF or NBx symbol and continuing until chronization the EOD symbol. During simultaneous transmissions of active and passive states on the bus, the The external J1850 bus on the other side of the resultant state on the bus is the active state. If the transceiver I.C. is a single wire multiplex bus with JBLPD detects a received symbol from the bus multiple nodes transmitting a number of different that is different from the symbol being transmitted, types of message frames. Each node can transmit then the JBLPD will discontinue its transmit operaat any time and synchronization and arbitration is tion prior to the start of the next bit. Once arbitraused to determine who wins control of the transtion has been lost, the VPWO pin must go passive mit. It is the obligation of the JBLPD transmitter within one period of the prescaled clock of the pesection to synchronize off of symbols on the bus, ripheral. Figure 114 shows 3 nodes attempting to and to place only nominal symbol times onto the arbitrate for the bus with Node B eventually winbus within the accuracy of the peripheral (+/- 1 µs). ning with the highest priority data. To transmit proper symbols the JBLPD must know what is going on out on the bus. Fortunately, the 250/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Figure 114. J1850 Arbitration Example Transmitting Node A Active Passive SOF Transmitting Node B Active Passive SOF Transmitting Node C Active Passive SOF Signal on Bus Active Passive SOF 0 0 1 1 0 0 0 1 0 0 1 1 0 0 0 0 0 0 1 1 0 1 0 0 1 1 0 0 0 0 Figure 115. J1850 Received Symbol Timing 178 µs VPWO Case 1 VPWI VPW Decoder 178 µs VPWO TX2 Case 2 VPWI VPW Decoder 178 µs VPWO TX2 Case 3 VPWI VPW Decoder 0 -6 14 8 22 200 214 192 208 222 251/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) 10.8.3.3 Receiving Messages Use of symbol and bit synchronization is an integral part of the J1850 bus scheme. Therefore, tight Data is received from the external analog transcoupling of the encoder and decoder functions is ceiver on the VPWI pin. VPWI data is immediately required to maintain synchronization during transpassed through a digital filter that ignores all pulsmits. Transmitted symbols and bits are initiated by es that are less than 7µs. Pulses greater than or the encoder and are timed through the decoder to equal to 7µs and less than 34µs are flagged as realize synchronization. Figure 115 exemplifies invalid bits (IBD) in the ERROR register. synchronization with 3 examples for an SOF symOnce data passes through the filter, all delimiters bol and JDLY[4:0] = 01110b. are stripped from the data stream and data bits are Case 1 shows a single transmitter arbitrating for shifted into the receive shift register by the decodthe bus. The VPWO pin is asserted, and 14µs later er logic. The first byte received after a valid SOF the bus transitions to an active state. The 14µs decharacter is compared with the flags contained in lay is due to the nominal delay through the exterFREG[0:31]. If the compare indicates that this nal transceiver chip. The signal is echoed back to message should be received, then the receive the transceiver through the VPWI pin, and proshift register contents are moved to the receive ceeds through the digital filter. The digital filter has data register (RXDATA) for the user program to a loop delay of 8 clock cycles with the signal finally access. The Receive Data Register Full bit presented to the decoder 22 µs after the VPWO (RDRF) is set to indicate that a complete byte has pin was asserted. The decoder waits 178 µs bebeen received. For each byte that is to be received fore issuing a signal to the encoder signifying the in a frame, once an entire byte has been received, end of the symbol. The VPWO pin is de-asserted the receive shift register contents are moved to the producing the nominal SOF bit timing (22 µs + receive data register (RXDATA). All data bits re178µs = 200 µs). ceived, including CRC bits, are transferred to the RXDATA register. The Receive Data Register Full Case 2 shows a condition where 2 transmitters atbit (RDRF) is set to indicate that a complete byte tempt to arbitrate for the bus at nearly the same has been received. time with a second transmitter, TX2, beginning slightly earlier than the VPWO pin. Since the If the first byte after a valid SOF indicates non-reJBLPD always times symbols from its receiver ception of this frame, then the current byte in the perspective, 178µs after the decoder sees the risreceive shift register is inhibited from being transing edge it issues a signal to the encoder to signify ferred to the RXDATA register and the RDRF flag the end of the SOF. Nominal SOF timings are remains clear (see the “Received Message Filtermaintained and the JBLPD re-synchronizes to ing” section). Also, no flags associated with receivTX2. ing a message (RDOF, CRCE, IFD, IBD) are set. Case 3 again shows an example of 2 transmitters A CRC check is kept on all bytes that are transattempting to arbitrate for the bus at nearly the ferred to the RXDATA register during message same time with the VPWO pin starting earlier than byte reception (succeeding an SOF symbol) and TX2. In this case TX2 is required to re-synchronize IFR3 reception (succeeding an NB0 symbol). The to VPWO. CRC is initialized on receipt of the first byte that follows an SOF symbol or an NB0 symbol. The All 3 examples exemplify how bus timings are drivCRC check concludes on receipt of an EODM en from the receiver perspective. Once the receivsymbol. The CRC error bit (CRCE), therefore, gets er detects an active bus, the transmitter symbol set after the EODM symbol has been recognized. timings are timed minus the transceiver and digital Refer to the “SAE Recommended Practice filter delays (i.e. SOF = 200 µs - 14µs - 8µs = J1850” manual for more information on CRCs. 178µs). This synchronization and timing off of the VPWI pin occurs for every symbol while transmitting. This ensures true arbitration during data byte transmissions. 252/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Received Message Filtering user program. All receiver flags and interrupts function normally. The FREG[0:31] registers can be considered an array of 256 bits (the FREG[0].0 bit is bit 0 of the Note that a break symbol received during a filtered array and the FREG[31].7 bit is bit 255). The I.D. out message will still be received. Note also that byte of a message frame is used as a pointer to the filter comparison occurs after reception of the the array (See Figure 116). first byte. So, any receive errors that occur before the message filter comparison (i.e. IBD, IFD) will Upon the start of a frame, the first data byte rebe active at least until the filter comparison. ceived after the SOF symbol determines the I.D. of the message frame. This I.D. byte addresses the I.D. byte flags stored in registers FREG[0:31]. This Transmitted Message Filtering operation is accomplished before the transfer of When transmitting a message, the corresponding the I.D. byte into the RXDATA register and before FREG[0:31] I.D. filter bit may be set or cleared. If the RDRF bit is set. set, then the JBLPD will receive all data informaIf the corresponding bit in the message filter array, tion transferred during the frame, unless sleep FREG[0:31], is set to zero (0), then the I.D. byte is mode is invoked. Everything the JBLPD transmits not transferred to the RXDATA register and the will be reflected in the RXDATA register. RDRF bit is not set. Also, the remainder of the Because the JBLPD has invalid bit detect (IBD), message frame is ignored until reception of an invalid frame detect (IFD), transmitter lost arbitraEOFmin symbol. A received EOFmin symbol tertion (TRA), and Cyclic Redundancy Check Error minates the operation of the message filter and (CRCE) it is not necessary for the transmitter to lisenables the receiver for the next message. None ten to the bytes that it is transmitting. The user of the flags related to the receiver, other than may wish to filter out the transmitted messages IDLE, are set. The EODM flag does not get set from the receiver. This can reduce interrupt burduring a filtered frame. No error flags other than den. When a transmitted I.D. byte is filtered by the RBRK can get set. receiver section of the block, then RDRF, RDOF, If the corresponding bit in the message filter array, EODM flags are inhibited and no RXDATA transFREG[0:31], is set to a one (1), then the I.D. byte fers occur. The other flags associated normally is transferred to the RXDATA register and the with receiving - RBRK, CRCE, IFD, and IBD - are RDRF is set. Also, the remainder of the message not inhibited, and they can be used to ascertain is received unless sleep mode is invoked by the the condition of the message transmit. Figure 116. I.D. Byte and Message Filter Array use Bit 0 = FREG[0].0 Bit 1 = FREG[0].1 Bit 2 = FREG[0].2 Bit 3 = FREG[0].3 Bit 4 = FREG[0].4 I.D. byte value = n Bit n-1 Bit n Bit n+1 Bit 254 = FREG[31].6 Bit 255 = FREG[31].7 253/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) ing the TRDY, TLA, TTO, TDUF, TRA, IBD, IFD, 10.8.3.4 Sleep Mode and CRCE bits to be set if required. This mode alSleep mode allows the user program to ignore the lows the user to not have to listen while talking. remainder of a message. Normally, the user program can recognise if the message is of interest from the header bytes at the beginning of the mes10.8.3.5 Normalization Bit symbol selection sage. If the user program is not interested in the The form of the NB0/NB1 symbol changes demessage it simply writes the SLP bit in the PRLR pending on the industry standard followed. A bit register. This causes all additional data on the bus (NBSYMS) in the OPTIONS register selects the to be ignored until an EOF minimum occurs. No symbol timings used. Refer to Table 46. additional flags (but not the EOFM flag) and, therefore, interrupts are generated for the remainder of the message. The single exception to this is a re10.8.3.6 VPWI input line management ceived break symbol while in sleep mode. Break The JBLPD is able to work with J1850 transceiver symbols always take precedence and will set the chips that have both inverted and not inverted RX RBRK bit in the ERROR register and generate an signal. A dedicated bit (INPOL) of the OPTIONS interrupt if the ERR_M bit in IMR is set. Sleep register must be programmed with the correct valmode and the SLP bit gets cleared on reception of ue depending on the polarity of the VPWI input an EOF or Break symbol. with respect to the J1850 bus line. Refer to the INWrites to the SLP bit will be ignored if: POL bit description for more details. 1) A valid EOFM symbol was the last valid symbol detected, 10.8.3.7 Loopback mode AND The JBLPD is able to work in loopback mode. This 2) The J1850 bus line (after the filter) is passive. mode, enabled setting the LOOPB bit of the OPTherefore, sleep mode can only be invoked after TIONS register, internally connects the output sigthe SOF symbol and subsequent data has been nal (VPWO) of the JBLPD to the input (VPWI) received, but before a valid EOF is detected. If without polarity inversion. The external VPWO pin sleep mode is invoked within this time window, of the MCU is forced in its passive state and the then any queued IFR transmit is aborted. If a MSG external VPWI pin is ignored (Refer to Figure 117). type is queued and sleep mode is invoked, then Note: When the LOOPB bit is set or reset, edges the MSG type will remain queued and an attempt could be detected by the J1850 decoder on the into transmit will occur after the EOF period has ternal VPWI line. These edges could be managed elapsed as usual. by the JBLPD as J1850 protocol errors. It is sugIf SLP mode is invoked while the JBLPD is currentgested to enable/disable LOOPB when the JBLPD ly transmitting, then the JBLPD effectively inhibits is suspended (CONTROL.JE=0, CONthe RDRF, RDT, EODM, & RDOF flags from being TROL.JDIS=0) or when the JBLPD is disabled set, and disallows RXDATA transfers. But, it other(CONTROL.JDIS=1). wise functions normally as a transmitter, still allowTable 46. Normalization Bit configurations IFR with CRC IFR without CRC 254/320 9 Symbol NBSYMS=0 NBSYMS=1 NB0 NB1 active Tv2 (active long) active Tv1 (active short) active Tv1 (active short) active Tv2 (active long) J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Figure 117. Local Loopback structure MCU JBLPD peripheral Passive state MCU VPWO pin VPWO from the peripheral logic VPWI toward the J1850 decoder Polarity manager OPTIONS.INPOL MCU VPWI pin OPTIONS.LOOPB 10.8.3.8 Peripheral clock management To work correctly, the encoder and decoder sections of the peripheral need an internal clock at 1MHz. This clock is used to evaluate the protocol symbols timings in transmission and in reception. The prescaled clock is obtained by dividing the MCU internal clock frequency. The CLKSEL register allows the selection of the right prescaling factor. The six least significant bits of the register (FREQ[5:0]) must be programmed with a value using the following formula: MCU Internal Freq. = 1MHz * (FREQ[5:0] + 1). Note: If the MCU internal clock frequency is lower than 1MHz, the JBLPD is not able to work correctly. If a frequency lower than 1MHz is used, the user program must disable the JBLPD. Note: When the MCU internal clock frequency or the clock prescaler factor are changed, the JBLPD could lose synchronization with the J1850 bus. 255/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) tion when the JBLPD is not used, even if the de10.8.4 Peripheral Functional Modes coder is able to follow the bus traffic. So, at any The JBLPD can be programmed in 3 modes, detime the JBLPD is enabled, it is immediately synpending on the value of the JE and JDIS bits in the chronized with the J1850 bus. CONTROL register, as shown in Table 47. Note: While the JBLPD is suspended, the STATable 47. JBLPD functional modes TUS register, the ERROR register and the SLP bit of the PRLR register are forced into their reset valJE JDIS mode ue. 0 1 JBLPD Disabled 0 0 JBLPD Suspended 10.8.4.3 JBLPD Disabled (Very Low Power Mode) 1 0 JBLPD Enabled Setting the JDIS bit in the CONTROL register, the JBLPD is stopped until the bit is reset by software. Depending on the mode selected, the JBLPD is Also the J1850 decoder is stopped, so the JBLPD able or unable to transmit or receive messages. is no longer synchronized with the bus. When the Moreover the power consumption of the peripheral bit is reset, the JBLPD will wait for a new idle state is affected. on the J1850 bus. This mode can be used to minimize power consumption when the JBLPD is not Note: The configuration with both JE and JDIS set used. is forbidden. Note: While the JDIS bit is set, the STATUS register, the ERROR register, the IMR register and the 10.8.4.1 JBLPD Enabled SLP, TEOBP and REOBP bits of the PRLR regisWhen the JBLPD is enabled (CONTROL.JE=1), it ter are forced to their reset value. is able to transmit and receive messages. Every Note: In order that the JDIS bit is able to reset the feature is available and every register can be writIMR register and the TEOBP and REOBP bits, the ten. JDIS bit must be left at 1 at least for 6 MCU clock cycles (3 NOPs). 10.8.4.2 JBLPD Suspended (Low Power Mode) Note: The JE bit of CONTROL register cannot be set with the same instruction that reset the JDIS When the JBLPD is suspended (CONTROL.JE=0 bit. It can be set only after the JDIS bit is reset. and CONTROL.JDIS=0), all the logic of the JBLPD is stopped except the decoder logic. This feature allows a reduction of power consump- 256/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) the RXDATA register (see also the RDRF bit de10.8.5 Interrupt Features scription of the STATUS register). The JBLPD has six interrupt sources that it han– The REOB (Receive End Of Block) interrupt is dles using the internal interrupts protocol. Other generated when receiving using DMA and the two interrupt sources (REOB and TEOB) are relatlast byte of a sequence of data is read from the ed to the DMA feature (See Section 10.8.6 DMA JBLPD. Features). No external interrupt channel is used by the – The TRDY interrupt is generated by two condiJBLPD. tions: when the TXOP register is ready to accept a new opcode for transmission; when the transThe dedicated registers of the JBLPD should be mit state machine accepts the opcode for transloaded with appropriate values to set the interrupt mission (a more detailed description of this vector (see the description of the IVR register), the condition is given in the TRDY bit description of interrupt mask bits (see the description of the IMR the STATUS register). register) and the interrupt pending bits (see the description of the STATUS and PRLR registers). – The TEOB (Transmit End Of Block) interrupt is generated when transmitting using DMA and the The interrupt sources are as follows: last byte of a sequence of data is written to the – The ERROR interrupt is generated when the ERJBLPD. ROR bit of the STATUS register is set. This bit is set when the following events occur: Transmitter Timeout, Transmitter Data Underflow, Receiver 10.8.5.1 Interrupt Management Data Overflow, Transmit Request Aborted, ReTo use the interrupt features the user has to follow ceived Break Symbol, Cyclic Redundancy Check these steps: Error, Invalid Frame Detect, Invalid Bit Detect (a more detailed description of these events is giv– Set the correct priority level of the JBLPD en in the description of the ERROR register). – Set the correct interrupt vector – The TLA interrupt is generated when the trans– Reset the Pending bits mitter loses the arbitration (a more detailed de– Enable the required interrupt source scription of this condition is given in the TLA bit description of the STATUS register). Note: It is strongly recommended to reset the pending bits before un-masking the related inter– The EODM interrupt is generated when the rupt sources to avoid spurious interrupt requests. JBLPD detects a passive level on the VPWI line longer than the minimum time accepted by the The priority with respect the other ST9 peripherals standard for the End Of Data symbol (a more deis programmable by the user setting the three tailed description of this condition is given in the most significant bits of the Interrupt Priority Level EODM bit description of the STATUS register). register (PRLR). The lowest interrupt priority is obtained by setting all the bits (this priority level is – The EOFM interrupt is generated when the never acknowledged by the CPU and is equivalent JBLPD detects a passive level on the VPWI line to disabling the interrupts of the JBLPD); the highlonger than the minimum time accepted by the est interrupt priority is programmed resetting the standard for the End Of Frame symbol (a more bits. See the Interrupt and DMA chapters of the detailed description of this condition is given in datasheet for more details. the EOFM bit description of the STATUS regisWhen the JBLPD interrupt priority is set, the priorter). ity between the internal interrupt sources is fixed – The RDRF interrupt is generated when a comby hardware as shown in Table 48. plete data byte has been received and placed in 257/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Note: After an MCU reset, the DMA requests of the JBLPD have a higher priority than the interrupt Each interrupt source has a pending bit in the requests. STATUS register, except the DMA interrupt sourcIf the DMASUSP bit of the OPTIONS register is es that have the interrupt pending bits located in set, while the ERROR and TLA flags are set, no the PRLR register. DMA transfer will be performed, allowing the reThese bits are set by hardware when the correlavent interrupt routines to manage each condition sponding interrupt event occurs. An interrupt reand, if necessary, disable the DMA transfer (Refer quest is performed only if the related mask bits are to Section 10.8.6 DMA Features). set in the IMR register and the JBLPD has priority. The pending bits have to be reset by the user softTable 48. JBLPD internal priority levels ware. Note that until the pending bits are set (while the corresponding mask bits are set), the JBLPD Priority Level Interrupt Source processes interrupt requests. So, if at the end of Higher ERROR, TLA an interrupt routine the related pending bit is not reset, another interrupt request is performed. EODM, EOFM To reset the pending bits, different actions have to RDRF, REOB be done, depending on each bit: see the descripLower TRDY, TEOB tion of the STATUS and PRLR registers. The user can program the most significant bits of the interrupt vectors by writing the V[7:3] bits of the IVR register. Starting from the value stored by the user, the JBLPD sets the three least significant bits of the IVR register to produce four interrupt vectors that are associated with interrupt sources as shown in Table 49. Table 49. JBLPD interrupt vectors Interrupt Vector Interrupt Source V[7:3] 000b V[7:3] 010b ERROR, TLA EODM, EOFM V[7:3] 100b RDRF, REOB V[7:3] 110b TRDY, TEOB 258/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) (odd address). They are pointed to by the DMA 10.8.6 DMA Features Transaction Counter Pointer Register (RDCPR The JBLPD can use the ST9 on-chip Direct Memregister in receiving, TDCPR register in transmitory Access (DMA) channels to provide high-speed ting) located in the JBLPD register page. data transactions between the JBLPD and contigTo select DMA transactions with the Register File, uous locations of Register File and Memory. The the control bits RDCPR.RF/MEM in receiving transactions can occur from and toward the mode or TDCPR.RF/MEM in transmitting mode JBLPD. The maximum number of transactions that must be set. each DMA channel can perform is 222 with Register File or 65536 with Memory. Control of the DMA The transaction Counter Register must be initialfeatures is performed using registers located in the ized with the number of DMA transfers to perform JBLPD register page (IVR, PRLR, IMR, RDAPR, and it will be decremented after each transaction. RDCPR, TDAPR, TDCPR). The DMA Address Register must be initialized with the starting address of the DMA table in the RegisThe priority level of the DMA features of the ter File, and it is incremented after each transacJBLPD with respect to the other ST9 peripherals tion. These two registers must be located between and the CPU is the same as programmed in the addresses 00h and DFh of the Register File. PRLR register for the interrupt sources. In the internal priority level order of the JBLPD, depending When the DMA occurs between JBLPD and Regon the value of the DMASUSP bit in the OPTIONS ister File, the TDAPR register (in transmission) register, the DMA may or may not have a higher and the RDAPR register (in reception) are not priority than the interrupt sources. used. Refer to the Interrupt and DMA chapters of the datasheet for details on priority levels. 10.8.6.2 DMA between JBLPD and Memory The DMA features are enabled by setting the apSpace propriate enabling bits (RXD_M, TXD_M) in the IMR register. It is also possible to select the direcIf the DMA transaction is made between the tion of the DMA transactions. JBLPD and Memory, a register pair is required to hold the DMA Address and another register pair to Once the DMA table is completed (the transaction hold the DMA Transaction counter. These two counter reaches 0 value), an interrupt request to pairs of registers must be located in the Register the CPU is generated if the related mask bit is set File. The DMA Address pair is pointed to by the (RDRF_M bit in reception, TRDY_M bit in transDMA Address Pointer Registers (RDAPR register mission). This kind of interrupt is called “End Of in reception, TDAPR register in transmission) loBlock”. The peripheral sends two different “End Of cated in the JBLPD register page; the DMA TransBlock” interrupts depending on the direction of the action Counter pair is pointed to by the DMA DMA (Receiving End Of Block (REOB) - TransmitTransaction Counter Pointer Registers (RDCPR ting End Of Block (TEOB)). These interrupt sourcregister in reception, TDCPR register in transmises have dedicated interrupt pending bits in the sion) located in the JBLPD register page. PRLR register (REOBP, TEOBP) and they are mapped to the same interrupt vectors: “Receive To select DMA transactions with Memory Space, Data Register Full (RDRF)” and “Transmit Ready the control bits RDCPR.RF/MEM in receiving (TRDY)” respectively. The same correspondence mode or TDCPR.RF/MEM in transmitting mode exists for the internal priority between interrupts must be reset. and interrupt vectors. The Transaction Counter register pair must be initialized with the number of DMA transfers to perform and it will be decremented after each transac10.8.6.1 DMA between JBLPD and Register File tion. The DMA Address register pair must be iniIf the DMA transaction is made between the tialized with the starting address of the DMA table JBLPD and the Register File, one register is rein Memory Space, and it is incremented after each quired to hold the DMA Address and one to hold transaction. These two register pairs must be lothe DMA transaction counter. These two registers cated between addresses 00h and DFh of the must be located in the Register File: the DMA AdRegister File. dress Register in an even addressed register, the DMA Transaction Counter in the following register 259/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) 10.8.6.3 DMA Management in Reception Mode through the DMA Address Register (or Register pair); The DMA in reception is performed when the RDRF bit of the STATUS register is set (by hard– A post-increment of the DMA Address Register ware). The RDRF bit is reset as soon as the DMA (or Register pair); cycle is finished. – A post-decrement of the DMA transaction counTo enable the DMA feature, the RXD_M bit of the ter, which contains the number of transactions IMR register must be set (by software). that have still to be performed. Each DMA request performs the transfer of a sinNote: When the REOBP pending bit is set (at the gle byte from the RXDATA register of the peripherend of the last DMA transfer), the reception DMA al toward Register File or Memory Space (Figure enable bit (RXD_M) is automatically reset by hard118). ware. However, the DMA can be disabled by softEach DMA transfer consists of three operations ware resetting the RXD_M bit. that are performed with minimum use of CPU time: Note: The DMA request acknowledge could de– A load from the JBLPD data register (RXDATA) pend on the priority level stored in the PRLR registo a location of Register File/Memory addressed ter. Figure 118. DMA in Reception Mode Register File or Memory space Previous data Data received RXDATA JBLPD peripheral 260/320 9 Current Address Pointer J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Register pair); it is the next location in the TXDA10.8.6.4 DMA Management in Transmission TA transfer cycle; Mode – A post-increment of the DMA Address Register DMA in transmission is performed when the TRDY (or Register pair); bit of the STATUS register is set (by hardware). The TRDY bit is reset as soon as the DMA cycle is – A post-decrement of the DMA transaction counfinished. ter, which contains the number of transactions To enable the DMA feature, the TXD_M bit in the that have still to be performed. IMR register must be set (by software). Note: When the TEOBP pending bit is set (at the Compared to reception, in transmission each DMA end of the last DMA transfer), the transmission request performs the transfer of either a single DMA enable bit (TXD_M) is automatically reset by byte or a couple of bytes depending on the value hardware. However, the DMA can be disabled by of the Transmit Opcode bits (TXOP.OP[2:0]) writsoftware resetting the TXD_M bit. ten during the DMA transfer. Note: When using DMA, the TXOP byte is written The table of values managed by the DMA must be before the TXDATA register. This order is accepta sequence of opcode bytes (that will be written in ed by the JBLPD only when the DMA in transmisthe TXOP register by the DMA) each one followed sion is enabled. by a data byte (that will be written in the TXDATA register by the DMA) if the opcode needs it (see Note: The DMA request acknowledge could deFigure 119). pend on the priority level stored in the PRLR register. In the same way, some time can occur beEach DMA cycle consists of the following transfers tween the transfer of the first byte and the transfer for a total of three/six operations that are perof the second one if another interrupt or DMA reformed with minimum use of CPU time: quest with higher priority occurs. – A load to the JBLPD Transmit Opcode register (TXOP) from a location of Register File/Memory addressed through the DMA Address Register 10.8.6.5 DMA Suspend mode (or Register pair); In the JBLPD it is possible to suspend or not to – A post-increment of the DMA Address Register suspend the DMA transfer while some J1850 pro(or Register pair); tocol events occur. The selection between the two modes is done by programming the DMASUSP bit – A post-decrement of the DMA transaction counof the OPTIONS register. ter, which contains the number of transactions If the DMASUSP bit is set (DMA suspended that have still to be performed; mode), while the ERROR or TLA flag is set, the and if the Transmit Opcode placed in TXOP reDMA transfers are suspended, to allow the user quires a datum: program to handle the event condition. If the DMASUSP bit is reset (DMA not suspended – A load to the peripheral data register (TXDATA) mode), the previous flags have no effect on the from a location of Register File/Memory adDMA transfers. dressed through the DMA Address Register (or 261/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Figure 119. DMA in Transmission Mode Register File or Memory space Previous Opcode sent (data not required) Previous Opcode sent (data required) Previous Data sent 1st byte Data sent TXOP 2nd byte TXDATA JBLPD peripheral 262/320 9 Opcode sent (data required) Opcode (data not required) Opcode (data required) Data J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) 10.8.7 Register Description The JBLPD peripheral uses 48 registers that are register (OPTIONS) are used to select the current mapped in a single page of the ST9 register file. sub-page. See Section 10.8.7.2 Stacked Registers section for a detailed description of these regTwelve registers are mapped from R240 (F0h) to isters. R251 (FBh): these registers are usually used to control the JBLPD. See Section 10.8.7.1 UnThe ST9 Register File page used is 23 (17h). Stacked Registers for a detailed description of these registers. NOTE: Bits marked as “Reserved” should be left at Thirty-six registers are mapped from R252 (FCh) their reset value to guarantee software compatibilto R255 (FFh). This is obtained by creating 9 subity with future versions of the JBLPD. pages, each containing 4 registers, mapped in the same register addresses; 4 bits (RSEL[3:0]) of a Figure 120. JBLPD Register Map R240 (F0h) R241 (F1h) R242 (F2h) R243 (F3h) R244 (F4h) R245 (F5h) R246 (F6h) R247 (F7h) R248 (F8h) R249 (F9h) R250 (FAh) R251 (FBh) STATUS TXDATA RXDATA TXOP CLKSEL CONTROL PADDR ERROR IVR PRLR IMR OPTIONS R252 (FCh) R253 (FDh) R254 (FEh) R255 (FFh) CREG0 CREG1 CREG2 CREG3 RDAPR RDCPR TDAPR TDCPR FREG28 FREG24FREG29 FREG20FREG25FREG30 FREG16FREG21FREG26FREG31 FREG12FREG17FREG22FREG27 FREG8 FREG13FREG18FREG23 FREG4 FREG9 FREG14FREG19 FREG0 FREG5 FREG10FREG15 FREG1 FREG6 FREG11 FREG2 FREG7 FREG3 263/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) 10.8.7.1 Un-Stacked Registers frame will not be cancelled for these errors, so TRDY would not get set. STATUS REGISTER (STATUS) – An RBRK error condition cancels all transmits for R240 - Read/Write this frame or any successive frames, so the Register Page: 23 TRDY bit will always be immediately set on an Reset Value: 0100 0000 (40h) RBRK condition. 7 0 TRDY is set on reset or while CONTROL.JE is reERR TRDY RDRF TLA RDT EODM EOFM IDLE set, or while the CONTROL.JDIS bit is set. If the TRDY_M bit of the IMR register is set, when this bit is set an interrupt request occurs. The bits of this register indicate the status of the 0: TXOP register not ready to receive a new opJBLPD peripheral. code This register is forced to its reset value after the 1: TXOP register ready to receive a new opcode MCU reset and while the CONTROL.JDIS bit is set. While the CONTROL.JE bit is reset, all bits except IDLE are forced to their reset values. Bit 5 = RDRF Receive Data Register Full Flag. RDRF is set when a complete data byte has been received and transferred from the serial shift regisBit 7 = ERR Error Flag. ter to the RXDATA register. The ERR bit indicates that one or more bits in the RDRF is cleared when the RXDATA register is ERROR register have been set. As long as any bit read (by software or by DMA). RDRF is also in the ERROR register remains set, the ERR bit recleared on reset or while CONTROL.JE is reset, or mains set. When all the bits in the ERROR register while CONTROL.JDIS bit is set. are cleared, then the ERR bit is reset by hardware. If the RDRF_M bit of the IMR register is set, when The ERR bit is also cleared on reset or while the this bit is set an interrupt request occurs. CONTROL.JE bit is reset, or while the CON0: RXDATA register doesn’t contain a new data TROL.JDIS bit is set. 1: RXDATA register contains a new data If the ERR_M bit of the IMR register is set, when this bit is set an interrupt request occurs. 0: No error Bit 4 = TLA Transmitter Lost Arbitration. 1: One or more errors have occurred The TLA bit gets set when the transmitter loses arbitration while transmitting messages or type 1 and 3 IFRs. Lost arbitration for a type 2 IFR does Bit 6 = TRDY Transmit Ready Flag. not set the TLA bit. (Type 2 messages require reThe TRDY bit indicates that the TXOP register is tries of the physical address if the arbitration is lost ready to accept another opcode for transmission. until the frame length is reached (if NFL=0)). The The TRDY bit is set when the TXOP register is TLA bit gets set when, while transmitting a MSG, empty and it is cleared whenever the TXOP regisMSG+CRC, IFR1, IFR3, or IFR3+CRC, the decodter is written (by software or by DMA). TRDY will ed VPWI data bit symbol received does not match be set again when the transmit state machine acthe VPWO data bit symbol that the JBLPD is atcepts the opcode for transmission. tempting to send out. If arbitration is lost, the When attempting to transmit a data byte without VPWO line is switched to its passive state and using DMA, two writes are required: first a write to nothing further is transmitted until an end-of-data TXDATA, then a write to the TXOP. (EOD) symbol is detected on the VPWI line. Also, – If a byte is written into the TXOP which results in any queued transmit opcode scheduled for transTRA getting set, then the TRDY bit will immedimission during this frame is cancelled (but the ately be set. TRA bit is not set). The TLA bit can be cleared by software writing a – If a TLA occurs and the opcode for which TRDY logic “zero” in the TLA position. TLA is also cleared is low is scheduled for this frame, then TRDY will on reset or while CONTROL.JE is reset, or while go high, if the opcode is scheduled for the next CONTROL.JDIS bit is set. frame, then TRDY will stay low. If the TLA_M bit of the IMR register is set, when – If an IBD, IFD or CRCE error condition occurs, this bit is set an interrupt request occurs. then TRDY will be set and any queued transmit 0: The JBLPD doesn’t lose arbitration opcode scheduled to transmit in the present 1: The JBLPD loses arbitration frame will be cancelled by the JBLPD peripheral. A MSGx opcode scheduled to be sent in the next 264/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Bit 3 = RDT Receive Data Type. The RDT bit indicates the type of data which is in Bit 0 = IDLE Idle Bus Flag the RXDATA register: message byte or IFR byte. IDLE is set when the JBLPD decoded VPWI pin Any byte received after an SOF but before an recognized an IFS symbol. That is, an idle bus is EODM is considered a message byte type. Any when the bus has been in a passive state for longbyte received after an SOF, EODM and NBx is an er that the Tv6 symbol time. The IDLE flag will reIFR type. main set as long as the decoded VPWI pin is pasRDT gets set or cleared at the same time that sive. IDLE is cleared when the decoded VPWI pin RDRF gets set. transitions to an active state. RDT is cleared on reset or while CONTROL.JE is Note that if the VPWI pin remains in a passive reset, or while CONTROL.JDIS bit is set. state after JE is set, then the IDLE bit may go high 0: Last RXDATA byte was a message type byte sometime before a Tv6 symbol is timed on VPWI 1: Last RXDATA byte was a IRF type byte (since VPWI timers may be active when JE is clear). IDLE is cleared on reset or while the CONBit 2 = EODM End of Data Minimum Flag. TROL.JDIS bit is set. The EODM flag is set when the JBLPD decoded 0: J1850 bus not in idle state VPWI pin has been in a passive state for longer 1: J1850 bus in idle state that the minimum Tv3 symbol time unless the EODM is inhibited by a sleep, filter or CRCE, IBD, IFD or RBRK error condition during a frame. JBLPD TRANSMIT DATA REGISTER (TXDATA) EODM bit does not get set when in the sleep mode R241- Read/Write or when a message is filtered. Register Page: 23 The EODM bit can be cleared by software writing a Reset Value: xxxx xxxx (xxh) logic “zero” in the EODM position. EODM is 7 0 cleared on reset, while CONTROL.JE is reset or while CONTROL.JDIS bit is set. TXD7 TXD6 TXD5 TXD4 TXD3 TXD2 TXD1 TXD0 If the EODM_M bit of the IMR register is set, when this bit is set an interrupt request occurs. 0: No EOD symbol detected The TXDATA register is an eight bits read/write 1: EOD symbol detected register in which the data to be transmitted must Note: The EODM bit is not an error flag. It means be placed. A write to TXDATA merely enters a that the minimum time related to the passive Tv3 byte into the register. To initiate an attempt to symbol is passed. transmit the data, the TXOP register must also be written. When the TXOP write occurs, the TRDY flag is cleared. While the TRDY bit is clear, the Bit 1 = EOFM End of Frame Minimum Flag. data is still in the TXDATA register, so writes to the The EOFM flag is set when the JBLPD decoded TXDATA register with TRDY clear will overwrite VPWI pin has been in a passive state for longer existing TXDATA. When the TXDATA is transthat the minimum Tv4 symbol time. EOFM will still ferred to the shift register, the TRDY bit is set get set at the end of filtered frames or frames again. where sleep mode was invoked. Consequently, Reads of the TXDATA register will always return multiple EOFM flags may be encountered bethe last byte written. tween frames of interest. TXDATA contents are undefined after a reset. The EOFM bit can be cleared by software writing a Note: The correct sequence to transmit is to write logic “zero” in the EOFM position. EOFM is first the TXDATA register (if datum is needed) and cleared on reset, while CONTROL.JE is reset or then the TXOP one. while CONTROL.JDIS bit is set. Only using the DMA, the correct sequence of writIf the EOFM_M bit of the IMR register is set, when ing operations is first the TXOP register and then this bit is set an interrupt request occurs. the TXDATA one (if needed). 0: No EOF symbol detected 1: EOF symbol detected Note: The EOFM bit is not an error flag. It means that the minimum time related to the passive Tv4 symbol is passed. 265/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) a byte. A write to the TXOP triggers the state maJBLPD RECEIVE DATA REGISTER (RXDATA) chine to initialize an attempt to serially transmit a R242- Read only byte out on the VPWO pin. An opcode which trigRegister Page: 23 gers a message byte or IFR type 3 to be sent will Reset Value: xxxx xxxx (xxh) transfer the TXDATA register contents to the 7 0 transmit serial shift register. An opcode which triggers a message byte or IFR type 3 to be sent with RXD7 RXD6 RXD5 RXD4 RXD3 RXD2 RXD1 RXD0 a CRC appended will transfer the TXDATA register contents to the transmit serial shift register and subsequently the computed CRC byte. An opcode The RXDATA register is an 8-bit read only register which triggers an IFR type 1 or 2 to be sent will in which the data received from VPWI is stored. transfer the PADDR register contents to the transVPWI data is transferred from the input VPW demit serial shift register. If a TXOP opcode is written coder to a serial shift register unless it is inhibited which is invalid for the bus conditions at the time by sleep mode, filter mode or an error condition (e.g. 12 byte frame or IFR3ing an IFR2), then no (IBD, IFD, CRCE, RBRK) during a frame. When transmit attempt is tried and the TRA bit in the ERthe shift register is full, this data is transferred to ROR register is set. the RXDATA register, and the RDRF flag gets set. Transmission of a string of data bytes requires All received data bytes are transferred to RXDATA multiple TXDATA/TXOP write sequences. Each including CRC bytes. A read of the RXDATA regwrite combination should be accomplished while ister will clear the RDRF flag. the TRDY flag is set. However, writes to the TXOP Note that care must be taken when reading RXDAwhen TRDY is not set will be accepted by the state TA subsequent to an RDRF flag. Multiple reads of machine, but it may override the previous data and RXDATA after an RDRF should only be attempted opcode. if the user can be sure that another RDRF will not Under normal message transmission conditions occur by the time the read takes place. the MSG opcode is written. If the last data byte of RXDATA content is undefined after a reset. a string is to be sent, then the MSG+CRC opcode will be written. An IFRx opcode is written if a reJBLPD TRANSMIT OPCODE REGISTER sponse byte or bytes to a received message (i.e. (TXOP) bytes received in RXDATA with RDT=0) is wanted R243 - Read/Write to transmit. The Message Length Count bits Register Page: 23 (MLC[3:0]) may be used to require that the IFR be Reset Value: 0000 0000 (00h) enabled only if the correct number of message bytes has been received. 7 0 NOTE: The correct sequence to transmit is to write MLC3 MLC2 MLC1 MLC0 OP2 OP1 OP0 first the TXDATA register and then the TXOP one. Only using the DMA, the correct sequence of writing operations is first the TXOP register and then TXOP is an 8-bit read/write register which contains the TXDATA one (if needed). the instructions required by the JBLPD to transmit 266/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Bit 7:4 = MLC[3:0] Message Length Count. MSG, Message Byte Opcode. The Message byte opcode is set when the user Message Length Count bits 3 to 0 are written when the program writes one of the IFR opcodes. Upon program wants to initiate or continue transmitting the body of a message out the VPWO pin. detection of the EOD symbol which delineates the body of a frame from the IFR portion of the frame, The body of a message is the string of data bytes the received byte counter is compared against the following an SOF symbol, but before the first EOD count contained in MLC[3:0]. If they match, then symbol in a frame. If the J1850 bus is in an idle condition when the opcode is written, an SOF the IFR will be transmitted. If they do not match, symbol is transmitted out the VPWO pin immedithen the TRA bit in the ERROR register is set and ately before it transmits the data contained in TXno transmit attempt occurs. DATA. If the JBLPD is not in idle and the J1850 – While NFL=0, an MCL[3:0] decimal value betransmitter has not been locked out by loss of arbitween 1 and 11 is considered valid. MCL[3:0] valtration, then the TXDATA byte is transferred to the ues of 12, 13, 14, 15 are considered invalid and serial output shift register for transmission immediwill set the Transmit Request Aborted (TRA) bit ately on completion of any previously transmitted in the ERROR register. data. The final byte of a message string is not – While NFL=1, an MCL[3:0] value between 1 and transmitted using the MSG opcode (use the 15 is considered valid. MSG+CRC opcode). – For NFL=1 or 0, MCL[3:0] bits are don’t care durSpecial Conditions for MSG Transmit: ing a MSG or MSG+CRC opcode write. – 1) A MSG cannot be queued on top of an execut– If writing an IFR opcode and MCL[3:0]=0000, ing IFR3 opcode. If so, then TRA is set, and then the message length count check is ignored TDUF will get set because the transmit state ma(i.e. MLC=Count is disabled), and the IFR is enchine will be expecting more data, then the inabled only on a correct CRC and a valid EOD verted CRC is appended to this frame. Also, no symbol assuming no other error conditions (IFD, message byte will be sent on the next frame. IBD, RBRK) appear. – 2) If NFL = 0 and an MSG queued without CRC on Received Byte Count for this frame=10 will trigger the TRA to get set, and TDUF will get set Bit 3 = Reserved. because the state machine will be expecting more data and the transmit machine will send the Bit 2:0 = OP[2:0] Transmit Opcode Select Bits. inverted CRC after the byte which is presently The bits OP[2:0] form the code that the transmitter transmitting. Also, no message byte will be sent uses to perform a transmit sequence. The codes on the next frame. are listed in Table 50. Caution should be taken when TRA gets set in Table 50. Opcode definitions these cases because the TDUF error sequence may engage before the user program has a OP[2:0] Transmit opcode Abbreviation chance to rewrite the TXOP register with the correct opcode. If a TDUF error occurs, a subsequent No operation or 000 CANCEL MSG write to the TXOP register will be used as the Cancel first byte of the next frame. 001 Send Break Symbol SBRK 010 Message Byte MSG 011 100 101 110 111 Message Byte then append CRC In-Frame Response Type 1 In-Frame Response Type 2 In-Frame Response Type 3 IFR Type 3 then append CRC MSG+CRC IFR1 IFR2 IFR3 IFR3+CRC 267/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) chance to rewrite the TXOP register with the corMSG+CRC, Message byte then append CRC oprect opcode. If a TDUF error occurs, a subsequent code. MSG+CRC write to the TXOP register will be used The ‘Message byte with CRC’ opcode is set when as the first byte of the next frame. the user program wants to transmit a single byte message followed by a CRC byte, or transmit the IFR1, In-Frame Response Type 1 opcode. final byte of a message string followed by a CRC The In-frame Response Type 1 (IFR 1) opcode is byte. written if the user program wants to transmit a A single byte message is basically an SOF symbol physical address byte (contained in the PADDR followed by a single data byte retrieved from TXregister) in response to a message that is currently DATA register followed by the computed CRC being received. byte followed by an EOD symbol. If the J1850 bus The user program decides to set up an IFR1 upon is in idle condition when the opcode is written, an receiving a certain portion of the data byte string of SOF symbol is immediately transmitted out the an incoming message. No write of the TXDATA VPWO pin. It then transmits the byte contained in register is required. The IFR1 gets its data byte the TXDATA register, then the computed CRC from the PADDR register. byte is transmitted. VPWO is then set to a passive The JBLPD block will enable the transmission of state. If the J1850 bus is not idle and the J1850 the IFR1 on these conditions: transmitter has not been locked out by loss of arbi– 1) The CRC check is valid (otherwise the CRCE tration, then the TXDATA byte is transferred to the is set) serial output shift register for transmission immediately on completion of any previously transmitted – 2) The received message length is valid if enadata. After completion of the TXDATA byte the bled (otherwise the TRA is set) computed CRC byte is transferred out the VPWO – 3) A valid EOD minimum symbol is received (othpin and then the VPWO pin is set passive to time erwise the IFD may eventually get set due to byte an EOD symbol. synchronization errors) Special Conditions for MSG+CRC Transmit: – 4) If NFL = 0 & Received Byte Count for this – 1) A MSG+CRC opcode cannot be queued on frame <=11 (otherwise TRA is set) top of an executing IFR3 opcode. If so, then TRA – 5) If not presently executing an MSG, IFR3, opis set, and TDUF will get set because the transcode (otherwise TRA is set, and TDUF will get mit state machine will be expecting more data, set because the transmit state machine will be then the inverted CRC is appended to this frame. expecting more data, so the inverted CRC will be Also, no message byte will be sent on the next appended to this frame) frame. – 6) If not presently executing an IFR1, IFR2, or – 2) If NFL=0, a MSG+CRC can only be queued if IFR3+CRC opcode otherwise TRA is set (but no Received Byte Count for this frame <=10 otherTDUF) wise the TRA will get set, and TDUF will get set because the state machine will be expecting – 7) If not presently receiving an IFR portion of a more data, so the transmit machine will send the frame, otherwise TRA is set. inverted CRC after the byte which is presently The IFR1 byte is then attempted according to the transmitting. Also, no message byte will be sent procedure described in section “Transmitting a on the next frame. type 1 IFR”. Note that if an IFR1 opcode is written, Caution should be taken when TRA gets set in a queued MSG or MSG+CRC is overridden by the these cases because the TDUF error sequence IFR1. may engage before the user program has a 268/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) ceived. IFR2, In-Frame Response Type 2 opcode. The IFR3 uses the contents of the TXDATA regisThe In-frame Response Type 2 (IFR2) opcode is ter for data. The user program decides to set up an set if the user program wants to transmit a physical IFR3 upon receiving a certain portion of the data address byte (contained in the PADDR register) in byte string of an incoming message. A previous response to a message that is currently being rewrite of the TXDATA register should have occeived. curred. The user program decides to set up an IFR2 upon The JBLPD block will enable the transmission of receiving a certain portion of the data byte string of the first byte of an IFR3 string on these conditions: an incoming message. No write of the TXDATA register is required. The IFR gets its data byte from – 1) The CRC check is valid (otherwise the CRCE the PADDR register. is set) The JBLPD block will enable the transmission of – 2) The received message length is valid if enathe IFR2 on these conditions: bled (otherwise the TRA is set) – 1) The CRC check is valid (otherwise the CRCE – 3) A valid EOD minimum symbol is received (othis set) erwise the IFD may eventually get set due to byte – 2) The received message length is valid if enasynchronization errors) bled (otherwise the TRA is set) – 4) If NFL = 0 & Received Byte Count for this – 3) A valid EOD minimum symbol is received (othframe <=9 (otherwise TRA is set and inverted erwise the IFD may eventually get set due to byte CRC is transmitted due to TDUF) synchronization errors) – 5) If not presently executing an MSG opcode – 4) If NFL = 0 & Received Byte Count for this (otherwise TRA is set, and TDUF will get set beframe <=11 (otherwise TRA is set) cause the transmit state machine will be expecting more data and the inverted CRC will be – 5) If not presently executing an MSG, IFR3, opappended to this frame) code (otherwise TRA is set, and TDUF will get set because the transmit state machine will be – 6) If not presently executing an IFR1, IFR2, or expecting more data, so the inverted CRC will be IFR3+CRC opcode, otherwise TRA is set (but no appended to this frame) TDUF) – 6) If not presently executing an IFR1, IFR2, or – 7) If not presently receiving an IFR portion of a IFR3+CRC opcodes, otherwise TRA is set (but frame, otherwise TRA is set. no TDUF) The IFR3 byte string is then attempted according – 7) If not presently receiving an IFR portion of a to the procedure described in section “Transmitframe, otherwise TRA is set. ting a type 3 IFR”. Note that if an IFR3 opcode is written, a queued MSG or MSG+CRC is overridThe IFR byte is then attempted according to the den by the IFR3. procedure described in section “Transmitting a type 2 IFR”. Note that if an IFR opcode is written, a The next byte(s) in the IFR3 data string shall also queued MSG or MSG+CRC is overridden by the be written with the IFR3 opcode except for the last IFR2. byte in the string which shall be written with the IFR3+CRC opcode. Each IFR3 data byte transIFR3, In-Frame Response Type 3 opcode. mission is accomplished with a TXDATA/TXOP The In-Frame Response Type 3 (IFR3) opcode is write sequence. The succeeding IFR3 transmit reset if the user program wants to initiate to transmit quests will be enabled on conditions 4 and 5 listed or continue to transmit a string of data bytes in reabove. sponse to a message that is currently being re- 269/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) The IFR3 byte is attempted according to the proIFR3+CRC, In-Frame Response Type 3 then apcedure described in section “Transmitting a type 3 pend CRC opcode. IFR”. The CRC byte is transmitted out on compleThe In-frame Response Type 3 then append CRC tion of the transmit of the IFR3 byte. opcode (IFR3+CRC) is set if the user program If this opcode sets up the last byte in an IFR3 data wants to either initiate to transmit a single data byte IFR3 followed by a CRC, or transmit the last string, then the TXDATA register contents shall be transmitted out immediately upon completion of data byte of an IFR3 string followed by the CRC the previous IFR3 data byte followed by the transbyte in response to a message that is currently bemit of the CRC byte. In this case the IFR3+CRC is ing received. The IFR3+CRC opcode transmits the contents of enabled on conditions 4 and 5 listed above. Note that if an IFR3+CRC opcode is written, a queued the TXDATA register followed by the computed CRC byte. The user program decides to set up an MSG or MSG+CRC is overridden by the IFR3+CRC. IFR3 upon receiving a certain portion of the data byte string of an incoming message. A previous SBRK, Send Break Symbol. write of the TXDATA register should have ocThe SBRK opcode is written to transmit a nominal curred. break (BRK) symbol out the VPWO pin. A Break symbol can be initiated at any time. Once the The J1850 block will enable the transmission of SBRK opcode is written a BRK symbol of the nomthe first byte of an IFR3 string on these conditions: inal Tv5 duration will be transmitted out the VPWO – 1) The CRC check is valid (otherwise the CRCE pin immediately. To terminate the transmission of is set) an in-progress break symbol the JE bit should be – 2) The received message length is valid if enaset to a logic zero. An SBRK command is nonbled (otherwise the TRA is set) maskable, it will override any present transmit operation, and it does not wait for the present trans– 3) A valid EOD minimum symbol is received (othmit to complete. Note that in the 4X mode a SBRK erwise the IFD may eventually get set due to byte will send a break character for the nominal Tv5 synchronization errors) time times four (4 x Tv5) so that all nodes on the – 4) If NFL = 0 & Received Byte Count for this bus will recognize the break. A CANCEL opcode frame <=10 (otherwise TRA is set and inverted does not override a SBRK command. CRC is transmitted) CANCEL, No Operation or Cancel Pending Trans– 5) If not presently executing an MSG opcode mit. (otherwise TRA is set, and TDUF will get set beThe Cancel opcode is used by the user program to cause the transmit state machine will be expecttell the J1850 transmitter that a previously queued ing more data and the inverted CRC will be opcode should not be transmitted. The Cancel opappended to this frame) code will set the TRDY bit. If the JBLPD peripheral – 6) If not presently executing an IFR1, IFR2 or is presently not transmitting, the Cancel command IFR3+CRC opcodes, otherwise TRA is set (but effectively cancels a pending MSGx or IFRx opno TDUF) code if one was queued, or it does nothing if no opcode was queued. If the JBLPD peripheral is – 7) If not presently receiving an IFR portion of a presently transmitting, then a queued MSGx or frame, otherwise TRA is set. IFRx opcode is aborted and the TDUF circuit may take affect. 270/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) JBLPD SYSTEM FREQUENCY SELECTION rect value must be written in the register. So an internal frequency less than 1MHz is not allowed. REGISTER (CLKSEL) R244- Read/Write Note: If the MCU internal clock frequency is lower Register Page: 23 than 1MHz, the peripheral is not able to work corReset Value: 0000 0000 (00h) rectly. If a frequency lower than 1MHz is used, the 7 0 user program must disable the peripheral. Note: When the clock prescaler factor or the MCU 4X FREQ5 FREQ4 FREQ3 FREQ2 FREQ1 FREQ0 internal frequency is changed, the peripheral could lose the synchronization with the J1850 bus. Bit 7 = 4X Diagnostic Four Times Mode. This bit is set when the J1850 clock rate is chosen JBLPD CONTROL REGISTER (CONTROL) four times faster than the standard requests, to R245- Read/Write force the BREAK symbol (nominally 300 µs long) Register Page: 23 and the Transmitter Timeout Time (nominally 1 Reset Value: 0100 0000 (40h) ms) at their nominal durations. 7 0 When the user want to use a 4 times faster J1850 clock rate, the new prescaler factor should be JE JDIS NFL JDLY4 JDLY3 JDLY2 JDLY1 JDLY0 stored in the FREQ[5:0] bits and the 4X bit must be set with the same instruction. In the same way, to exit from the mode, FREQ[5:0] and 4X bits must The CONTROL register is an eight bit read/write be placed at the previous value with the same inregister which contains JBLPD control information. struction. Reads of this register return the last written data. 0: Diagnostic Four Times Mode disabled 1: Diagnostic Four Times Mode enabled Bit 7 = JE JBLPD Enable. Note: Setting this bit, the prescaler factor is not auThe JBLPD block enable bit (JE) enables and distomatically divided by four. The user must adapt ables the transmitter and receiver to the VPWO the value stored in FREQ[5:0] bits by software. and VPWI pins respectively. When the JBLPD peNote: The customer should take care using this ripheral is disabled the VPWO pin is in its passive mode when the MCU internal frequency is less state and information coming in the VPWI pin is igthan 4MHz. nored. When the JBLPD block is enabled, the transmitter and receiver function normally. Note that queued transmits are aborted when JE is Bit 6 = Reserved. cleared. JE is cleared on reset, by software and setting the JDIS bit. 0: The peripheral is disabled Bit 5:0 = FREQ[5:0] Internal Frequency Selectors. 1: The peripheral is enabled These 6 bits must be programmed depending on the internal frequency of the device. The formula Note: It is not possible to reset the JDIS bit and to that must be used is the following one: set the JE bit with the same instruction. The corMCU Int. Freq.= 1MHz * (FREQ[5:0] + 1). rect sequence is to first reset the JDIS bit and then set the JE bit with another instruction. Note: To obtain a correct operation of the peripheral, the internal frequency of the MCU (INTCLK) must be an integer multiple of 1MHz and the cor- 271/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Bit 4:0 = JDLY[4:0] JBLPD Transceiver External Bit 6 = JDIS Peripheral clock frozen. When this bit is set by software, the peripheral is Loop Delay Selector. stopped and the bus is not decoded anymore. A These five bits are used to select the nominal exreset of the bit restarts the internal state machines ternal loop time delay which normally occurs when as after a MCU reset. The bit is reset by microconthe peripheral is connected and transmitting in a troller reset or by software. J1850 bus system. The external loop delay is deThe the JDIS bit is set on MCU reset. fined as the time between when the VPWO is set 0: The peripheral clock is running to a certain level to when the VPWI recognizes the 1: The peripheral clock is stopped corresponding (inverted) edge on its input. Refer to “Transmit Opcode Queuing” section and the Note: When the JDIS bit is set, the STATUS regSAE-J1850 standard for information on how the ister, the ERROR register, the IMR register and external loop delay is used in timing transmitted the TEOBP and REOBP bits of the PRLR register symbols. are forced into their reset value. The allowed values are integer values between 0 Note: It is not possible to reset the JDIS bit and to µs and 31 µs. set the JE bit with the same instruction. The correct sequence is to first reset the JDIS bit and then set the JE bit with another instruction. JBLPD PHYSICAL ADDRESS REGISTER (PADDR) R246- Read/Write Bit 5 = NFL No Frame Length Check Register Page: 23 The NFL bit is used to enable/disable the J1850 Reset Value: xxxx xxxx (xxh) requirement of 12 bytes maximum per frame limit. 7 0 The SAE J1850 standard states that a maximum of 12 bytes (including CRCs and IFRs) can be on ADR7 ADR6 ADR5 ADR4 ADR3 ADR2 ADR1 ADR0 the J1850 between a start of frame symbol (SOF) and an end of frame symbol (EOF). If this condition is violated, then the JBLPD peripheral gets an The PADDR is an eight bit read/write register Invalid Frame Detect (IFD) and the sleep mode which contains the physical address of the JBLPD ensues until a valid EOFM is detected. If the valid peripheral. During initialization the user program frame check is disabled (NFL=1), then no limits will write the PADDR register with its physical adare imposed on the number of data bytes which dress. The Physical Address is used during incan be sent or received on the bus between an frame response types 1 and 2 to acknowledge the SOF and an EOF. The default upon reset is for the receipt of a message. The JBLPD peripheral will frame checking to be enabled. transmit the contents of the PADDR register for The NFL bit is cleared on reset type 1 or 2 IFRs as defined by the TXOP register. 0: Twelve bytes frame length check enabled This register is undefined on reset. 1: Twelve bytes frame length check disabled 272/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) JBLPD ERROR REGISTER (ERROR) is set, then the TTO will timeout at 4000 prescaled clock cycles. When the TTO flag is set then the diR247- Read only agnostic circuit will disable the VPWO signal, and Register Page: 23 disable the JBLPD peripheral. The user program Reset Value: 0000 0000 (00h) must then clear the JE bit to remove the TTO error. 7 0 It can then retry the block by setting the JE bit again. TTO TDUF RDOF TRA RBRK CRCE IFD IBD The TTO bit can be used to determine if the external J1850 bus is shorted low. Since the transmitter looks for proper edges returned at the VPWI pin ERROR is an eight bit read only register indicating for its timing, a lack of edges seen at VPWI when error conditions that may arise on the VPWO and trying to transmit (assuming the RBRK does not VPWI pins. A read of the ERROR register clears get set) would indicate a constant low condition. all bits (except for TTO and possibly the RBRK bit) The user program can take appropriate actions to which were set at the time of the read. The register test the J1850 bus circuit when a TTO occurs. is cleared after the MCU reset, while the CONNote that a transmit attempt must occur to detect a TROL.JE bit is reset, or while the CONTROL.JDIS bus shorted low condition. bit is set. The TTO bit is cleared while the CONTROL.JE bit All error conditions that can be read in the ERROR is reset or while the CONTROL.JDIS bit is set. register need to have redundant ERROR indicator TTO is cleared on reset. flags because: 0: VPWO line at 1 for less than 1 ms – With JE set, the TDUF, RDOF, TRA, CRCE, IFD, 1: VPWO line at 1 for longer than 1 ms & IBD bits in the ERROR register can only be cleared by reading the register. Bit 6 = TDUF Transmitter Data Underflow. – The TTO bit can only be cleared by clearing the The TDUF will be set to a logic one if the transmitJE bit. ter expects more information to be transmitted, but – The RBRK bit can only be cleared by reading the a TXOP write has not occurred in time (by the end ERROR register after the break condition has of transmission of the last bit). disappeared. The transmitter knows to expect more information from the user program when transmitting messagError condition indicator flags associated with the es or type 3 IFRs only. If an opcode is written to error condition are cleared when the error condiTXOP that does not include appending a CRC tion ends. Since error conditions may alter the acbyte, then the JBLPD peripheral assumes more tions of the transmitter and receiver, the error condata is to be written. When the JBLPD peripheral dition indicators must remain set throughout the has shifted out the data byte it must have the next error condition. All error conditions, including the data byte in time to place it directly next to it. If the RBRK condition, are events that get set during a user program does not place new data in the TXparticular clock cycle of the prescaled clock of the DATA register and write the TXOP register with a peripheral. The IFD, IBD, RBRK, and CRCE error proper opcode, then the CRC byte which is being conditions are then cleared when a valid EOF kept tabulated by the transmitter is logically invertsymbol is detected from the VPWI pin. The TRA ed and transmitted out the VPWO pin. This will enerror condition is a singular event that sets the corsure that listeners will detect this message as an responding ERROR register bit, but this error itself error. In this case the TDUF bit is set to a logic causes no other actions. one. TDUF is cleared by reading the ERROR register Bit 7 = TTO Transmitter Timeout Flag with TDUF set. TDUF is also cleared on reset, The TTO bit is set when the VPWO pin has been in while the CONTROL.JE bit is reset or while the a logic one (or active) state for longer than 1 ms. CONTROL.JDIS bit is set. This flag is the output of a diagnostic circuit based 0: No transmitter data underflow condition ocon the prescaled system clock input. If the 4X bit is curred not set, the TTO will trip if the VPWO is constantly 1: Transmitter data underflow condition occurred active for 1000 prescaled clock cycles. If the 4X bit 273/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Bit 5 = RDOF Receiver Data Overflow The RDOF gets set to a logic one if the data in the 0: No valid Break symbol received RXDATA register has not been read and new data 1: Valid Break symbol received is ready to be transferred to the RXDATA register. The old RXDATA information is lost since it is Bit 2 = CRCE Cyclic Redundancy Check Error overwritten with new data. The receiver section always keeps a running tab of RDOF is cleared by reading the ERROR register the CRC of all data bytes received from the VPWl with RDOF set, while the CONTROL.JE bit is reset since the last EOD symbol. The CRC check is peror while the CONTROL.JDIS bit is set, or on reset. formed when a valid EOD symbol is received both 0: No receiver data overflow condition occurred after a message string (subsequent to an SOF 1: Receiver data overflow condition occurred symbol) and after an IFR3 string (subsequent to an NB0 symbol). If the received CRC check fails, then the CRCE bit is set to a logic one. CRC errors Bit 4 = TRA Transmit Request Aborted are inhibited if the JBLPD peripheral is in the The TRA gets set to a logic one if a transmit op“sleep or filter and NOT presently transmitting” code is aborted by the JBLPD state machine. mode. A CRC error occurs once for a frame. AfterMany conditions may cause a TRA. They are exwards, the receiver is disabled until an EOFM plained in the transmit opcode section. If the TRA symbol is received and queued transmits for the bit gets set after a TXOP write, then a transmit is present frame are cancelled (but the TRA bit is not not attempted, and the TRDY bit is not cleared. set). CRCE is cleared when ERROR is read. It is If a TRA error condition occurs, then the requested also cleared while the CONTROL.JE bit is reset or transmit is aborted, and the JBLPD peripheral while the CONTROL.JDIS bit is set, or on reset. takes appropriate measures as described under 0: No CRC error detected the TXOP register section. 1: CRC error detected TRA is cleared on reset, while the CONTROL.JE bit is reset or while the CONTROL.JDIS bit is set. 0: No transmission request aborted Bit 1 = IFD Invalid Frame Detect 1: Transmission request aborted The IFD bit gets set when the following conditions are detected from the filtered VPWI pin: – An SOF symbol is received after an EOD miniBit 3 = RBRK Received Break Symbol Flag The RBRK gets set to a logic one if a valid break mum, but before an EOF minimum. (BRK) symbol is detected from the filtered VPWI – An SOF symbol is received when expecting data pin. A Break received from the J1850 bus will canbits. cel queued transmits of all types. The RBRK bit re– If NFL = 0 and a message frame greater than 12 mains set as long as the break character is detectbytes (i.e. 12 bytes plus one bit) has been reed from the VPWI. Reads of the ERROR register ceived in one frame. will not clear the RBRK bit as long as a break character is being received. Once the break character – An EOD minimum time has elapsed when data is gone, a final read of the ERROR register clears bits are expected. this bit. – A logic 0 or 1 symbol is received (active for Tv1 An RBRK error occurs once for a frame if it is reor Tv2) when an SOF was expected. ceived during a frame. Afterwards, the receiver is – The second EODM symbol received in a frame disabled from receiving information (other than the is NOT followed directly by an EOFM symbol. break) until an EOFM symbol is received. RBRK bit is cleared on reset, while the CONIFD errors are inhibited if the JBLPD peripheral is TROL.JE bit is reset or while the CONTROL.JDIS in the “sleep or filter and NOT presently transmitbit is set. ting” mode. An IFD error occurs once for a frame. The RBRK bit can be used to detect J1850 bus Afterwards, the receiver is disabled until an EOFM shorted high conditions. If RBRK is read as a logic symbol is received, and queued transmits for the high multiple times before an EOFM occurs, then a present frame are cancelled (but the TRA bit is not possible bus shorted high condition exists. The set). IFD is cleared when ERROR is read. It is also user program can take appropriate measures to cleared while the CONTROL.JE bit is reset or test the bus if this condition occurs. Note that this while the CONTROL.JDIS bit is set or on reset. bit does not necessarily clear when ERROR is 0: No invalid frame detected read. 1: Invalid frame detected 274/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Bit 0 = IBD Invalid Bit Detect. Bit 0 = Reserved. The IBD bit gets set whenever the receiver detects that the filtered VPWI pin was not fixed in a state JBLPD PRIORITY LEVEL REGISTER (PRLR) long enough to reach the minimum valid symbol R249- Read/Write time of Tv1 (or 35 µs). Any timing event less than Register Page: 23 35 µs (and, of course, > 7 µs since the VPWI digitReset Value: 0001 0000 (10h) al filter will not allow pulses less than this through its filter) is considered as noise and sets the IBD 7 0 accordingly. At this point the JBLPD peripheral will PRL2 PRL1 PRL0 SLP REOBP TEOBP cease transmitting and receiving any information until a valid EOF symbol is received. IBD errors are inhibited if the JBLPD peripheral is Bit 7:5 = PRL[2:0] Priority level bits in the “sleep or filter and NOT presently transmitThe priority with respect to the other peripherals ting” mode. An IBD error occurs once for a frame. and the CPU is encoded with these three bits. The Afterwards, the receiver is disabled until an EOFM value of “0” has the highest priority, the value “7” symbol is received, and queued transmits for the has no priority. After the setting of this priority levpresent frame are cancelled (but the TRA bit is not el, the priorities between the different Interrupt set). sources and DMA of the JBLPD peripheral is hardIBD is cleared when ERROR is read. Note that if ware defined (refer to the “Status register” bits dean invalid bit is detected during a bus idle condiscription, the “Interrupts Management” and the tion, the IBD flag gets set and a new EOFmin must section about the explanation of the meaning of be seen after the invalid bit before commencing to the interrupt sources). receive again. IBD is also cleared while the CONDepending on the value of the OPTROL.JE bit is reset or while the CONTROL.JDIS TIONS.DMASUSP bit, the DMA transfers can or bit is set and on reset. cannot be suspended by an ERROR or TLA event. 0: No invalid bit detected Refer to the description of DMASUSP bit. 1: Invalid bit detected JBLPD INTERRUPT VECTOR REGISTER (IVR) R248- Read/Write (except bits 2:1) Register Page: 23 Reset Value: xxxx xxx0 (xxh) 7 Table 52. Internal Interrupt and DMA Priorities without DMA suspend mode Priority Level Higher Priority RX-DMA 0 V7 V6 V5 V4 V3 EV2 EV1 ERROR, TLA - Bit 7:3 = V[7:3] Interrupt Vector Base Address. User programmable interrupt vector bits. Bit 2:1 = EV[2:1] Encoded Interrupt Source (Read Only). EV2 and EV1 are set by hardware according to the interrupt source, given in Table 51 (refer to the Status register bits description about the explanation of the meaning of the interrupt sources) EODM, EOFM RDRF, REOB Lower Priority EV1 Interrupt Sources 0 0 0 1 ERROR, TLA EODM, EOFM 1 0 RDRF, REOB 1 1 TRDY, TEOB TRDY, TEOB Table 53. Internal Interrupt and DMA Priorities with DMA suspend mode Priority Level Higher Priority Table 51. Interrupt Sources EV2 Event Sources TX-DMA Event Sources ERROR, TLA TX-DMA RX-DMA EODM, EOFM RDRF, REOB Lower Priority TRDY, TEOB 275/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) the end of a block of data. An interrupt request is Bit 4 = SLP Receiver Sleep Mode. The SLP bit is written to one when the user properformed if the TRDY_M bit of the IMR register is gram does not want to receive any data from the set. TEOBP should be reset by software in order to avoid undesired interrupt routines, especially in inJBLPD VPWI pin until an EOFM symbol occurs. This mode is usually set when a message is reitialisation routine (after reset) and after entering the End Of Block interrupt routine. ceived that the user does not require - including Writing “0” in this bit will cancel the interrupt remessages that the JBLPD is transmitting. quest. If the JBLPD is not transmitting and is in Sleep This bit is reset when the CONTROL.JDIS bit is mode, no data is transferred to the RXDATA regisset at least for 6 MCU clock cycles (3 NOPs). ter, the RDRF flag does not get set, and errors associated with received data (RDOF, CRCE, IFD, Note: When the TEOBP flag is set, the TXD_M bit IBD) do not get set. Also, the EODM flag will not is reset by hardware. get set. Note: TEOBP can only be written to “0”. If the JBLPD peripheral is transmitting and is in sleep mode, no data is transferred to the RXDATA register, the RDRF flag does not get set and the JBLPD INTERRUPT MASK REGISTER (IMR) RDOF error flag is inhibited. The CRCE, IFD, and R250 - Read/Write IBD flags, however, will NOT be inhibited while Register Page: 23 transmitting in sleep mode. Reset Value: 0000 0000 (00h) The SLP bit cannot be written to zero by the user 7 0 program. The SLP bit is set on reset or TTO getting set, and it will stay set upon JE getting set until ERR TRDY RDRF TLA RXD EODM EOFM TXD an EOFM symbol is received. _M _M _M _M _M _M _M _M The SLP gets cleared on reception of an EOF or a Break symbol. SLP is set while CONTROL.JE is reset and while CONTROL.JDIS is set. To enable an interrupt source to produce an inter0: The JBLPD is not in Sleep Mode rupt request, the related mask bit must be set. 1: The JBLPD is in Sleep Mode When these bits are reset, the related Interrupt Pending bit can not generate an interrupt. Note: This register is forced to its reset value if the Bit 3:2 = Reserved. CONTROL.JDIS bit is set at least for 6 clock cycles (3 NOPs). If the JDIS bit is set for a shorter Bit 1 = REOP Receiver DMA End Of Block Pendtime, the bits could be reset or not reset. ing . This bit is set after a receiver DMA cycle to mark Bit 7 = ERR_M Error Interrupt Mask bit. the end of a block of data. An interrupt request is This bit enables the “error” interrupt source to genperformed if the RDRF_M bit of the IMR register is erate an interrupt request. set. REOBP should be reset by software in order This bit is reset if the CONTROL.JDIS bit is set at to avoid undesired interrupt routines, especially in least for 6 clock cycles (3 NOPs). initialisation routine (after reset) and after entering 0: Error interrupt source masked the End Of Block interrupt routine. 1: Error interrupt source un-masked Writing “0” in this bit will cancel the interrupt request. This bit is reset when the CONTROL.JDIS bit is Bit 6 = TRDY_M Transmit Ready Interrupt Mask set at least for 6 MCU clock cycles (3 NOPs). bit. Note: When the REOBP flag is set, the RXD_M bit This bit enables the “transmit ready” interrupt is reset by hardware. source to generate an interrupt request. This bit is reset if the CONTROL.JDIS bit is set at Note: REOBP can only be written to “0”. least for 6 clock cycles (3 NOPs). 0: TRDY interrupt source masked Bit 0 = TEOP Transmitter DMA End Of Block 1: TRDY interrupt source un-masked Pending. This bit is set after a transmitter DMA cycle to mark 276/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Bit 5 = RDRF_M Receive Data Register Full Interrupt Mask bit. Bit 0 = TXD_M Transmitter DMA Mask bit. This bit enables the “receive data register full” inIf this bit is “0” no transmitter DMA request will be terrupt source to generate an interrupt request. generated, and the TRDY bit, in the Status RegisThis bit is reset if the CONTROL.JDIS bit is set at ter (STATUS), can request an interrupt. If TXD_M least for 6 clock cycles (3 NOPs). bit is set to “1” then the TRDY bit can request a 0: RDRF interrupt source masked DMA transfer. TXD_M is reset by hardware when 1: RDRF interrupt source un-masked the transaction counter value decrements to zero, that is when a Transmitter End Of Block condition Bit 4 = TLA_M Transmitter Lost Arbitration Interoccurs (TEOBP flag set). rupt Mask bit. This bit is reset if the CONTROL.JDIS bit is set at This bit enables the “transmitter lost arbitration” inleast for 6 clock cycles (3 NOPs). terrupt source to generate an interrupt request. 0: Transmitter DMA disabled This bit is reset if the CONTROL.JDIS bit is set at 1: Transmitter DMA enabled least for 6 clock cycles (3 NOPs). 0: TLA interrupt source masked 1: TLA interrupt source un-masked JBLPD OPTIONS AND REGISTER GROUPS SELECTION REGISTER (OPTIONS) R251- Read/Write Bit 3 = RXD_M Receiver DMA Mask bit. Register Page: 23 If this bit is “0” no receiver DMA request will be Reset Value: 0000 0000 (00h) generated, and the RDRF bit, in the Status Register (STATUS), can request an interrupt. If RXD_M 7 0 bit is set to “1” then the RDRF bit can request a INPOL NBSYMS DMASUSP LOOPB RSEL3 RSEL2 RSEL1 RSEL0 DMA transfer. RXD_M is reset by hardware when the transaction counter value decrements to zero, that is when a Receiver End Of Block condition ocBit 7 = INPOL VPWI Input Polarity Selector. curs (REOBP flag set). This bit allows the selection of the polarity of the This bit is reset if the CONTROL.JDIS bit is set at RX signal coming from the transceivers. Dependleast for 6 clock cycles (3 NOPs). ing on the specific transceiver, the RX signal is in0: Receiver DMA disabled verted or not inverted respect the VPWO and the 1: Receiver DMA enabled J1850 bus line. 0: VPWI input is inverted by the transceiver with Bit 2 = EODM_M End of Data Minimum Interrupt respect to the J1850 line. Mask bit. 1: VPWI input is not inverted by the transceiver This bit enables the “end of data minimum” interwith respect to the J1850 line. rupt source to generate an interrupt request. This bit is reset if the CONTROL.JDIS bit is set at Bit 6 = NBSYMS NB Symbol Form Selector. least for 6 clock cycles (3 NOPs). This bit allows the selection of the form of the Nor0: EODM interrupt source mask malization Bits (NB0/NB1). 1: EODM interrupt source un-masked 0: NB0 active long symbol (Tv2), NB1 active short symbol (Tv1) Bit 1 = EOFM_M End of Frame Minimum Interrupt 1: NB0 active short symbol (Tv1), NB1 active long Mask bit. symbol (Tv2) This bit enables the “end of frame minimum” interrupt source to generate an interrupt request. This bit is reset if the CONTROL.JDIS bit is set at least for 6 clock cycles (3 NOPs). 0: EOFM interrupt source masked 1: EOFM interrupt source un-masked 277/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Bit 5 = DMASUSP DMA Suspended Selector. Note: When the LOOPB bit is set, also the INPOL If this bit is “0”, JBLPD DMA has higher priority bit must be set to obtain the correct management with respect to the Interrupts of the peripheral. of the polarity. DMA is performed even if an interrupt request is already scheduled or if the relative interrupt rouBit 3:0 = RSEL[3:0] Registers Group Selection tine is in execution. bits. If the bit is “1”, while the ERROR or TLA flag of the These four bits are used to select one of the 9 STATUS register are set, the DMA transfers are groups of registers, each one composed of four suspended. As soon as the flags are reset, the registers that are stacked at the addresses from DMA transfers can be performed. R252 (FCh) to R255 (FFh) of this register page 0: DMA not suspended (23). Unless the wanted registers group is already 1: DMA suspended selected, to address a specific registers group, Note: This bit has effect only on the priorities of these bits must be correctly written. the JBLPD peripheral. This feature allows that 36 registers (4 DMA registers - RDADR, RDCPR, TDAPR, TDCPR - and 32 Message Filtering Registers - FREG[0:31]) are Bit 4 = LOOPB Local Loopback Selector. mapped using only 4 registers (here called Current This bit allows the Local Loopback mode. When Registers - CREG[3:0]). this mode is enabled (LOOPB=1), the VPWO outSince the Message Filtering Registers put of the peripheral is sent to the VPWI input with(FREG[0:31]) are seldom read or written, it is sugout inversions whereas the VPWO output line of gested to always reset the RSEL[3:0] bits after acthe MCU is placed in the passive state. Moreover cessing the FREG[0:31] registers. In this way the the VPWI input of the MCU is ignored by the peDMA registers are the current registers. ripheral. (Refer to Figure 117). 0: Local Loopback disabled 1: Local Loopback enabled 278/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) JBLPD CURRENT REGISTER 2 (CREG2) JBLPD CURRENT REGISTER 0 (CREG0) R254- Read/Write R252- Read/Write Register Page: 23 Register Page: 23 Reset Value: xxxx xxxx (xxh) Reset Value: xxxx xxxx (xxh) 7 b7 b6 b5 b4 b3 b2 b1 0 7 b0 b7 0 b6 b5 b4 b3 b2 b1 b0 Depending on the RSEL[3:0] value of the OPTIONS register, this register is one of the following stacked registers: RDAPR, FREG0, FREG4, FREG8, FREG12, FREG16, FREG20, FREG24, FREG28. Depending on the RSEL[3:0] value of the OPTIONS register, this register is one of the following stacked registers: TDAPR, FREG2, FREG6, FREG10, FREG14, FREG18, FREG22, FREG26, FREG30. JBLPD CURRENT REGISTER 1 (CREG1) R253 - Read/Write Register Page: 23 Reset Value: xxxx xxxx (xxh) JBLPD CURRENT REGISTER 3 (CREG3) R255- Read/Write Register Page: 23 Reset Value: xxxx xxxx (xxh) 7 b7 b6 b5 b4 b3 b2 b1 0 7 b0 b7 Depending on the RSEL[3:0] value of the OPTIONS register, this register is one of the following stacked registers: RDCPR, FREG1, FREG5, FREG9, FREG13, FREG17, FREG21, FREG25, FREG29. 0 b6 b5 b4 b3 b2 b1 b0 Depending on the RSEL[3:0] value of the OPTIONS register, this register is one of the following stacked registers: TDCPR, FREG3, FREG7, FREG11, FREG15, FREG19, FREG23, FREG27, FREG31. Table 54. Stacked registers map RSEL[3:0] Current Registers CREG0 0000b 1000b 1001b 1010b 1011b 1100b 1101b 1110b 1111b RDAPR FREG0 FREG4 FREG8 FREG12 FREG16 FREG20 FREG24 FREG28 CREG1 RDCPR FREG1 FREG5 FREG9 FREG13 FREG17 FREG21 FREG25 FREG29 CREG2 TDAPR FREG2 FREG6 FREG10 FREG14 FREG18 FREG22 FREG26 FREG30 CREG3 TDCPR FREG3 FREG7 FREG11 FREG15 FREG19 FREG23 FREG27 FREG31 279/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) 10.8.7.2 Stacked Registers Register File) of the DMA receiver transaction counter when the DMA between Peripheral and See the description of the OPTIONS register to Memory Space is selected. Otherwise, if the DMA obtain more information on the map of the regisbetween Peripheral and Register File is selected, ters of this section. this register points to a pair of registers that are used as DMA Address register and DMA Transaction Counter. JBLPD RECEIVER DMA ADDRESS POINTER See Section 10.8.6.1and Section 10.8.6.2 for REGISTER (RDAPR) more details on the use of this register. R252 - RSEL[3:0]=0000b Register Page: 23 Reset Value: xxxx xxxx (xxh) Bit 0 = RF/MEM Receiver Register File/Memory 7 0 Selector. If this bit is set to “1”, then the Register File will be RA7 RA6 RA5 RA4 RA3 RA2 RA1 PS selected as Destination, otherwise the Memory space will be used. 0: Receiver DMA with Memory space To select this register, the RSEL[3:0] bits of the 1: Receiver DMA with Register File OPTIONS register must be reset Bit 7:1 = RA[7:1] Receiver DMA Address Pointer. RDAPR contains the address of the pointer (in the Register File) of the Receiver DMA data source when the DMA between the peripheral and the Memory Space is selected. Otherwise, when the DMA between the peripheral and Register File is selected, this register has no meaning. See Section 10.8.6.2 for more details on the use of this register. Bit 0 = PS Memory Segment Pointer Selector. This bit is set and cleared by software. It is only meaningful if RDCPR.RF/MEM = 1. 0: The ISR register is used to extend the address of data received by DMA (see MMU chapter) 1: The DMASR register is used to extend the address of data received by DMA (see MMU chapter) JBLPD RECEIVER DMA TRANSACTION COUNTER REGISTER (RDCPR) R253 - RSEL[3:0]=0000b Register Page: 23 Reset Value: xxxx xxxx (xxh) 7 RC7 0 RC6 RC5 RC4 RC3 RC2 RC1 RF/MEM To select this register, the RSEL[3:0] bits of the OPTIONS register must be reset Bit 7:1 = RC[7:1] Receiver DMA Counter Pointer. RDCPR contains the address of the pointer (in the 280/320 9 JBLPD TRANSMITTER DMA ADDRESS POINTER REGISTER (TDAPR) R254 - RSEL[3:0]=0000b Register Page: 23 Reset Value: xxxx xxxx (xxh) 7 TA7 0 TA6 TA5 TA4 TA3 TA2 TA1 PS To select this register, the RSEL[3:0] bits of the OPTIONS register must be reset Bit 7:1 = TA[7:1] Transmitter DMA Address Pointer. TDAPR contains the address of the pointer (in the Register File) of the Transmitter DMA data source when the DMA between the Memory Space and the peripheral is selected. Otherwise, when the DMA between Register File and the peripheral is selected, this register has no meaning. See Section 10.8.6.2 for more details on the use of this register. Bit 0 = PS Memory Segment Pointer Selector. This bit is set and cleared by software. It is only meaningful if TDCPR.RF/MEM = 1. 0: The ISR register is used to extend the address of data transmitted by DMA (see MMU chapter) 1: The DMASR register is used to extend the address of data transmitted by DMA (see MMU chapter) J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) JBLPD MESSAGE FILTERING REGISTERS JBLPD TRANSMITTER DMA TRANSACTION (FREG[0:31]) COUNTER REGISTER (TDCPR) R252/R253/R254/R255 - RSEL[3]=1 R255 - RSEL[3:0]=0000b Register Page: 23 Register Page: 23 Reset Value: xxxx xxxx (xxh) Reset Value: xxxx xxxx (xxh) 7 TC7 0 TC6 TC5 TC4 TC3 TC2 TC1 RF/MEM Register 7 FREG0 F_07 F_06 F_05 0 F_04 F_03 F_02 F_01 F_00 FREG1 F_0F F_0E F_0D F_0C F_0B F_0A F_09 F_08 To select this register, the RSEL[3:0] bits of the OPTIONS register must be reset Bit 7:1 = TC[7:1] Transmitter DMA Counter Pointer. RDCPR contains the address of the pointer (in the Register File) of the DMA transmitter transaction counter when the DMA between Memory Space and peripheral is selected. Otherwise, if the DMA between Register File and peripheral is selected, this register points to a pair of registers that are used as DMA Address register and DMA Transaction Counter. See Section 10.8.6.1and Section 10.8.6.2 for more details on the use of this register. FREG2 F_17 F_16 F_15 FREG3 F_1F F_1E F_1D F_1C F_1B F_1A F_19 F_18 FREG4 F_27 F_26 F_25 F_24 F_23 F_22 F_21 F_20 FREG5 F_2F F_2E F_2D F_2C F_2B F_2A F_29 F_28 FREG6 F_37 F_36 F_35 F_34 F_33 F_32 F_31 F_30 FREG7 F_3F F_3E F_3D F_3C F_3B F_3A F_39 F_38 FREG8 F_47 F_46 F_45 F_44 F_43 F_42 F_41 F_40 FREG9 F_4F F_4E F_4D F_4C F_4B F_4A F_49 F_48 FREG10 F_57 F_56 F_55 F_54 F_53 F_52 F_51 F_50 FREG11 F_5F F_5E F_5D F_5C F_5B F_5A F_59 F_58 FREG12 F_67 F_66 F_65 F_64 F_63 F_62 F_61 F_60 FREG13 F_6F F_6E F_6D F_6C F_6B F_6A F_69 F_68 FREG14 F_77 F_76 F_75 Bit 0 = RF/MEM Transmitter Register File/Memory Selector. If this bit is set to “1”, then the Register File will be selected as Destination, otherwise the Memory space will be used. 0: Transmitter DMA with Memory space 1: Transmitter DMA with Register File F_14 F_13 F_12 F_11 F_10 F_74 F_73 F_72 F_71 F_70 FREG15 F_7F F_7E F_7D F_7C F_7B F_7A F_79 F_78 FREG16 F_87 F_86 F_85 F_84 F_83 F_82 F_81 F_80 FREG17 F_8F F_8E F_8D F_8C F_8B F_8A F_89 F_88 FREG18 F_97 F_96 F_95 F_94 F_93 F_92 F_91 F_90 FREG19 F_9F F_9E F_9D F_9C F_9B F_9A F_99 F_98 FREG20 F_A7 F_A6 F_A5 F_A4 F_A3 F_A2 F_A1 F_A0 FREG21 F_AF F_AE F_AD F_AC F_AB F_AA F_A9 F_A8 FREG22 F_B7 F_B6 F_B5 F_B4 F_B3 F_B2 F_B1 F_B0 FREG23 F_BF F_BE F_BD F_BC F_BB F_BA F_B9 F_B8 FREG24 F_C7 F_C6 F_C5 F_C4 F_C3 F_C2 F_C1 F_C0 FREG25 F_CF F_CE F_CD F_CC F_CB F_CA F_C9 F_C8 FREG26 F_D7 F_D6 F_D5 F_D4 F_D3 F_D2 F_D1 F_D0 FREG27 F_DF F_DE F_DD F_DC F_DB F_DA F_D9 F_D8 FREG28 F_E7 F_E6 F_E5 F_E4 F_E3 F_E2 F_E1 F_E0 FREG29 F_EF F_EE F_ED F_EC F_EB F_EA F_E9 F_E8 FREG30 F_F7 F_F6 F_F5 F_F4 F_F3 F_F2 F_F1 F_F0 FREG31 F_FF F_FE F_FD F_FC F_FB F_FA F_F9 F_F8 281/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) These registers are structured in eight groups of register and the RDRF flag is set. Also, every other four registers. The user can gain access to these data byte received in this frame is transferred to registers programming the RSEL[2:0] bits of the the RXDATA register unless the JBLPD peripheral OPTIONS register while the RSEL[3] bit of the is put into sleep mode setting the SLP bit. same register must be placed at 1. In this way the If the bit of the array correspondent to the I.D. byte user can select the group where the registers that is clear, then the transfer of this byte as well as any he/she wants to use are placed. See the descripbyte for the balance of this frame is inhibited, and tion of OPTIONS register for the correspondence the RDRF bit remains cleared. between registers and the values of RSEL[2:0] bits The bit 0 of the FREG[0] register (FREG[0].0 (See Table 54). marked as F_00 in the previous table) correFrom the functional point of view, the FREG[0]sponds to the I.D. byte equal to 00h while the bit 7 FREG[31] registers can be seen as an array of of the FREG[31] register (FREG[31].7 - marked as 256 bits involved in the J1850 received message F_FF in the previous table) corresponds to the I.D. filtering system. byte equal to FFh. The first byte received in a frame (following a valid Note: The FREG registers are undefined upon rereceived SOF character) is an Identifier (I.D.) byte. set. Because of this, it is strongly recommended It is used by the JBLPD peripheral as the address that the contents of these registers has to be deof the 256 bits array. fined before JE is set for the first time after reset. If the bit of the array correspondent to the I.D. byte Otherwise, unpredictable results may occur. is set, then the byte is transferred to the RXDATA 282/320 9 J1850 Byte Level Protocol Decoder (JBLPD) J1850 BYTE LEVEL PROTOCOL DECODER (Cont’d) Register Address 7 0 STATUS reset value F0h ERR 0 TRDY 1 RDRF 0 TLA 0 RDT 0 EODM 0 EOFM 0 IDLE 0 TXDATA reset value F1h TXD7 x TXD6 x TXD5 x TXD4 x TXD3 x TXD2 x TXD1 x TXD0 x RXDATA reset value F2h RXD7 x RXD6 x RXD5 x RXD4 x RXD3 x RXD2 x RXD1 x RXD0 x TXOP reset value F3h MLC3 0 MLC2 0 MLC1 0 MLC0 0 0 OP2 0 OP1 0 OP0 0 CLKSEL reset value F4h 4X 0 0 FREQ5 0 FREQ4 0 FREQ3 0 FREQ2 0 FREQ1 0 FREQ0 0 CONTROL reset value F5h JE 0 JDIS 1 NFL 0 JDLY4 0 JDLY3 0 JDLY2 0 JDLY1 0 JDLY0 0 PADDR reset value F6h ADR7 x ADR6 x ADR5 x ADR4 x ADR3 x ADR2 x ADR1 x ADR0 x ERROR reset value F7h TTO 0 TDUF 0 RDOF 0 TRA 0 RBRK 0 CRCE 0 IFD 0 IBD 0 IVR reset value F8h V7 x V6 x V5 x V4 x V3 x EV2 x EV1 x 0 PRLR reset value F9h PRL2 0 PRL1 0 PRL0 0 SLP 1 0 0 REOBP 0 TEOBP 0 IMR reset value FAh ERR_M 0 TRDY_M 0 RDRF_M 0 TLA_M 0 RXD_M 0 EODM_M 0 EOFM_M 0 TXD_M 0 OPTIONS reset value FBh INPOL 0 NBSYMS 0 DMASUSP 0 LOOPB 0 RSEL3 0 RSEL2 0 RSEL1 0 RSEL0 0 CREG0 reset value FCh b7 x b6 x b5 x b4 x b3 x b2 x b1 x b0 x CREG1 reset value FDh b7 x b6 x b5 x b4 x b3 x b2 x b1 x b0 x CREG2 reset value FEh b7 x b6 x b5 x b4 x b3 x b2 x b1 x b0 x CREG3 reset value FFh b7 x b6 x b5 x b4 x b3 x b2 x b1 x b0 x 283/320 9 EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) 10.9 EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) 10.9.1 Introduction The 8-Channel Analog to Digital Converter (A/D) comprises an input multiplex channel selector feeding a successive approximation converter. Conversion requires 138 INTCLK cycles (of which 84 are required for sampling), conversion time is thus a function of the INTCLK frequency; for instance, for a 20MHz clock rate, conversion of the selected channel requires 6.9µs. This time includes the 4.2µs required by the built-in Sample and Hold circuitry, which minimizes the need for external components and allows quick sampling of the signal to minimise warping and conversion error. Conversion resolution is 8 bits, with ±1 LSB maximum error in the input range between VSS and the analog VDD reference. The converter uses a fully differential analog input configuration for the best noise immunity and precision performance. Two separate supply references are provided to ensure the best possible supply noise rejection. In fact, the converted digital value, is referred to the analog reference voltage which determines the full scale converted value. Naturally, Analog and Digital VSS MUST be common. If analog supplies are not present, input reference voltages are referred to the digital ground and supply. Up to 8 multiplexed Analog Inputs are available, depending on the specific device type. A group of signals can be converted sequentially by simply programming the starting address of the first analog channel to be converted and with the AUTOSCAN feature. Two Analog Watchdogs are provided, allowing continuous hardware monitoring of two input channels. An Interrupt request is generated whenever the converted value of either of these two analog inputs is outside the upper or lower programmed threshold values. The comparison result is stored in a dedicated register. Figure 121. Block Diagram n INTERRUPT UNIT COMPARE LOGIC INTERNAL TRIGGER INT. VECTOR POINTER INT. CONTROL REGISTER COMPARE RESULT REGISTER THRESHOLD REGISTER 7U THRESHOLD REGISTER 7L THRESHOLD REGISTER 6U THRESHOLD REGISTER 6L CONTROL LOGIC EXTERNAL TRIGGER CONTROL REG. DATA REGISTER 7 DATA REGISTER 6 DATA REGISTER 5 DATA REGISTER 4 DATA REGISTER 3 DATA REGISTER 2 DATA REGISTER 1 DATA REGISTER 0 CONVERSION RESULT ANALOG MUX SUCCESSIVE APPROXIMATION A/D CONVERTER AIN 7 AIN 6 AIN 5 AIN 4 AIN 3 AIN 2 AIN 1 AIN 0 AUTOSCAN LOGIC VA00223 284/320 9 EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) ANALOG TO DIGITAL CONVERTER (Cont’d) Single and continuous conversion modes are available. Conversion may be triggered by an external signal or, internally, by the Multifunction Timer. A Power-Down programmable bit allows the A/D to be set in low-power idle mode. The A/D’s Interrupt Unit provides two maskable channels (Analog Watchdog and End of Conversion) with hardware fixed priority, and up to 7 programmable priority levels. CAUTION: A/D INPUT PIN CONFIGURATION The input Analog channel is selected by using the I/O pin Alternate Function setting (PXC2, PXC1, PXC0 = 1,1,1) as described in the I/O ports section. The I/O pin configuration of the port connected to the A/D converter is modified in order to prevent the analog voltage present on the I/O pin from causing high power dissipation across the input buffer. Deselected analog channels should also be maintained in Alternate function configuration for the same reason. 10.9.2 Functional Description 10.9.2.1 Operating Modes Two operating modes are available: Continuous Mode and Single Mode. To enter one of these modes it is necessary to program the CONT bit of the Control Logic Register. Continuous Mode is selected when CONT is set, while Single Mode is selected when CONT is reset. Both modes operate in AUTOSCAN configuration, allowing sequential conversion of the input channels. The number of analog inputs to be converted may be set by software, by setting the number of the first channel to be converted into the Control Register (SC2, SC1, SC0 bits). As each conversion is completed, the channel number is automatically incremented, up to channel 7. For example, if SC2, SC1, SC0 are set to 0,1,1, conversion will proceed from channel 3 to channel 7, whereas, if SC2, SC1, SC0 are set to 1,1,1, only channel 7 will be converted. When the ST bit of the Control Logic Register is set, either by software or by hardware (by an internal or external synchronisation trigger signal), the analog inputs are sequentially converted (from the first selected channel up to channel 7) and the results are stored in the relevant Data Registers. In Single Mode (CONT = “0”), the ST bit is reset by hardware following conversion of channel 7; an End of Conversion (ECV) interrupt request is issued and the A/D waits for a new start event. In Continuous Mode (CONT = “1”), a continuous conversion flow is initiated by the start event. When conversion of channel 7 is complete, conversion of channel ’s’ is initiated (where ’s’ is specified by the setting of the SC2, SC1 and SC0 bits); this will continue until the ST bit is reset by software. In all cases, an ECV interrupt is issued each time channel 7 conversion ends. When channel ’i’ is converted (’s’ <’i’ <7), the related Data Register is reloaded with the new conversion result and the previous value is lost. The End of Conversion (ECV) interrupt service routine can be used to save the current values before a new conversion sequence (so as to create signal sample tables in the Register File or in Memory). 10.9.2.2 Triggering and Synchronisation In both modes, conversion may be triggered by internal or external conditions; externally this may be tied to EXTRG, as an Alternate Function input on an I/O port pin, and internally, it may be tied to INTRG, generated by a Multifunction Timer peripheral. Both external and internal events can be separately masked by programming the EXTG/ INTG bits of the Control Logic Register (CLR). The events are internally ORed, thus avoiding potential hardware conflicts. However, the correct procedure is to enable only one alternate synchronisation condition at any time. The effect either of these synchronisation modes is to set the ST bit by hardware. This bit is reset, in Single Mode only, at the end of each group of conversions. In Continuous Mode, all trigger pulses after the first are ignored. The synchronisation sources must be at a logic low level for at least the duration of one INTCLK cycle and, in Single Mode, the period between trigger pulses must be greater than the total time required for a group of conversions. If a trigger occurs when the ST bit is still set, i.e. when conversion is still in progress, it will be ignored. On devices where two A/D Converters are present they can be triggered from the same source. Converter External Trigger On Chip Event (Internal trigger) A/D 0 A/D 1 EXTRG pin MFT 0 10.9.2.3 Analog Watchdogs Two internal Analog Watchdogs are available for highly flexible automatic threshold monitoring of external analog signal levels. 285/320 9 EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) ANALOG TO DIGITAL CONVERTER (Cont’d) Analog channels 6 and 7 monitor an acceptable voltage level window for the converted analog inputs. The external voltages applied to inputs 6 and 7 are considered normal while they remain below their respective Upper thresholds, and above or at their respective Lower thresholds. When the external signal voltage level is greater than, or equal to, the upper programmed voltage limit, or when it is less than the lower programmed voltage limit, a maskable interrupt request is generated and the Compare Results Register is updated in order to flag the threshold (Upper or Lower) and channel (6 or 7) responsible for the interrupt. The four threshold voltages are user programmable in dedicated registers (08h to 0Bh) of the A/D register page. Only the 4 MSBs of the Compare Results Register are used as flags (the 4 LSBs always return “1” if read), each of the four MSBs being associated with a threshold condition. Following a hardware reset, these flags are reset. During normal A/D operation, the CRR bits are set, in order to flag an out of range condition and are automatically reset by hardware after a software reset of the Analog Watchdog Request flag in the AD_ICR Register. Figure 122. A/D Trigger Source n 286/320 9 10.9.2.4 Power Down Mode Before enabling an A/D conversion, the POW bit of the Control Logic Register must be set; this must be done at least 60µs before the first conversion start, in order to correctly bias the analog section of the converter circuitry. When the A/D is not required, the POW bit may be reset in order to reduce the total power consumption. This is the reset configuration, and this state is also selected automatically when the ST9 is placed in Halt Mode (following the execution of the halt instruction). Analog Voltage Upper threshold Normal Area (Window Guarded) Lower threshold EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) ANALOG TO DIGITAL CONVERTER (Cont’d) Figure 123. Application Example: Analog Watchdog used in Motorspeed Control n 10.9.3 Interrupts The A/D provides two interrupt sources: – End of Conversion – Analog Watchdog Request The A/D Interrupt Vector Register (AD_IVR) provides hardware generated flags which indicate the interrupt source, thus allowing automatic selection of the correct interrupt service routine. Analog Watchdog Request End of Conv. Request 7 X 0 X X X X X 0 7 X 0 0 X X X X X 1 0 Lower Word Address Upper Word Address The A/D Interrupt vector should be programmed by the User to point to the first memory location in the Interrupt Vector table containing the base address of the four byte area of the interrupt vector table in which the address of the A/D interrupt service routines are stored. The Analog Watchdog Interrupt Pending bit (AWD, AD_ICR.6), is automatically set by hardware whenever any of the two guarded analog inputs go out of range. The Compare Result Register (CRR) tracks the analog inputs which exceed their programmed thresholds. When two requests occur simultaneously, the Analog Watchdog Request has priority over the End of Conversion request, which is held pending. The Analog Watchdog Request requires the user to poll the Compare Result Register (CRR) to determine which of the four thresholds has been exceeded. The threshold status bits are set to flag an out of range condition, and are automatically reset by hardware after a software reset of the Analog Watchdog Request flag in the AD_ICR Register. The interrupt pending flags, ECV and AWD, should be reset by the user within the interrupt service routine. Setting either of these two bits by software will cause an interrupt request to be generated. 10.9.3.1 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 pages are the following: A/Dn Register Page A/D 0 63 A/D 1 61 287/320 9 EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) ANALOG TO DIGITAL CONVERTER (Cont’d) 10.9.4 Register Description DATA REGISTERS (DiR) The conversion results for the 8 available channels are loaded into the 8 Data registers following conversion of the corresponding analog input. CHANNEL 0 DATA REGISTER (D0R) R240 - Read/Write Register Page: 63 Reset Value: undefined 7 D0.7 D0.5 D0.4 D0.3 D0.2 D0.1 D3.7 0 D3.6 D3.5 D3.4 D3.3 D3.2 D3.1 D0.0 D3.0 CHANNEL 4 DATA REGISTER (D4R) R244 - Read/Write Register Page: 63 Reset Value: undefined 7 0 D0.6 7 D4.7 0 D4.6 D4.5 D4.4 D4.3 D4.2 D4.1 D4.0 Bit 7:0 = D4.[7:0]: Channel 4 Data Bit 7:0 = D0.[7:0]: Channel 0 Data. CHANNEL 5 DATA REGISTER (D5R) R245 - Read/Write Register Page: 63 Reset Value: undefined CHANNEL 1 DATA REGISTER (D1R) R241 - Read/Write Register Page: 63 Reset Value: undefined 7 D1.7 7 0 D1.6 D1.5 D1.4 D1.3 D1.2 D1.1 D1.0 D5.7 0 D5.6 D5.5 D5.4 D5.3 D5.2 D5.1 D5.0 Bit 7:0 = D5.[7:0]: Channel 5 Data. Bit 7:0 = D1.[7:0]: Channel 1 Data. CHANNEL 6 DATA REGISTER (D6R) R246 - Read/Write Register Page: 63 Reset Value: undefined CHANNEL 2 DATA REGISTER (D2R) R242 - Read/Write Register Page: 63 Reset Value: undefined 7 D2.7 7 0 D2.6 D2.5 D2.4 D2.3 D2.2 D2.1 D2.0 D6.7 0 D6.6 D6.5 D6.4 D6.3 D6.2 D6.1 D6.0 Bit 7:0 = D6.[7:0]: Channel 6 Data Bit 7:0 = D2.[7:0]: Channel 2 Data. CHANNEL 3 DATA REGISTER (D3R) R243 - Read/Write Register Page: 63 Reset Value: undefined Bit 7:0 = D3.[7:0]: Channel 3 Data. 288/320 9 CHANNEL 7 DATA REGISTER (D7R) R247 - Read/Write Register Page: 63 Reset Value: undefined 7 D7.7 0 D7.6 D7.5 D7.4 D7.3 D7.2 D7.1 D7.0 EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) ANALOG TO DIGITAL CONVERTER (Cont’d) CHANNEL 6 LOWER THRESHOLD REGISTER (LT6R) R248 - Read/Write Register Page: 63 Reset Value: undefined 7 0 LT6.7 LT6.6 LT6.5 LT6.4 LT6.3 LT6.2 LT6.1 LT6.0 Bit 7:0 = LT6.[7:0]: Channel 6 Lower Threshold User-defined lower threshold value for Channel 6, to be compared with the conversion results. CHANNEL 7 LOWER THRESHOLD REGISTER (LT7R) R249 - Read/Write Register Page: 63 Reset Value: undefined 7 0 LT7.7 LT7.6 LT7.5 LT7.4 LT7.3 LT7.2 LT7.1 LT7.0 Bit 7:0 = LT7.[7:0]: Channel 7 Lower Threshold. User-defined lower threshold value for Channel 7, to be compared with the conversion results. CHANNEL 6 UPPER THRESHOLD REGISTER (UT6R) R250 - Read/Write Register Page: 63 Reset Value: undefined 7 UT6. 7 0 UT6. UT6. 6 5 UT6. 4 UT6. UT6. 3 2 UT6. UT6. 1 0 Bit 7:0 = UT6.[7:0]: Channel 6 Upper Threshold value. User-defined upper threshold value for Channel 6, to be compared with the conversion results. CHANNEL 7 UPPER THRESHOLD REGISTER (UT7R) R251 - Read/Write Register Page: 63 Reset Value: undefined 7 UT7. 7 0 UT7. 6 UT7. UT7. 5 4 UT7. UT7. 3 2 UT7. 1 UT7. 0 Bit 7:0 = UT7.[7:0]: Channel 7 Upper Threshold value User-defined upper threshold value for Channel 7, to be compared with the conversion results. COMPARE RESULT REGISTER (CRR) R252 - Read/Write Register Page: 63 Reset Value: 0000 1111 (0Fh) 7 C7U 0 C6U C7L C6L 1 1 1 1 These bits are set by hardware and cleared by software. Bit 7 = C7U: Compare Reg 7 Upper threshold 0: Threshold not reached 1: Channel 7 converted data is greater than or equal to UT7R threshold register value. Bit 6 = C6U: Compare Reg 6Upper threshold 0: Threshold not reached 1: Channel 6 converted data is greater than or equal to UT6R threshold register value. Bit 5 = C7L: Compare Reg 7 Lower threshold 0: Threshold not reached 1: Channel 7 converted data is less than the LT7R threshold register value. Bit 4 = C6L: Compare Reg 6 Lower threshold 0: Threshold not reached 1: Channel 6 converted data is less than the LT6R threshold register value. Bit 3:0 = Reserved, returns “1” when read. Note: Any software reset request generated by writing to the AD_ICR, will also cause all the compare status bits to be cleared. 289/320 9 EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) ANALOG TO DIGITAL CONVERTER (Cont’d) CONTROL LOGIC REGISTER (CLR) The Control Logic Register (CLR) manages the A/D converter logic. Writing to this register will cause the current conversion to be aborted and the autoscan logic to be re-initialized. CONTROL LOGIC REGISTER (CLR) R253 - Read/Write Register Page: 63 Reset Value: 0000 0000 (00h) 7 SC2 0 SC1 SC0 EXT CON INTG POW G T ST Bit 7:5 = SC[2:0]: Start Conversion Address. These 3 bits define the starting analog input channel (Autoscan mode). The first channel addressed by SC[2:0] is converted, then the channel number is incremented for the successive conversion, until channel 7 (111) is converted. When SC2, SC1 and SC0 are all set, only channel 7 will be converted. Bit 4 = EXTG: External Trigger Enable. This bit is set and cleared by software. 0: External trigger disabled. 1: External trigger enabled. Allows a conversion sequence to be started on the subsequent edge of the external signal applied to the EXTRG pin (when enabled as an Alternate Function). Bit 3 = INTG: Internal Trigger Enable. This bit is set and cleared by software. 0: Internal trigger disabled. 1: Internal trigger enabled. Allows a conversion sequence to be started, synchronized by an internal signal (On-chip Event signal) from a Multifunction Timer peripheral. Both External and Internal Trigger inputs are internally ORed, thus avoiding Hardware conflicts; 290/320 9 however, the correct procedure is to enable only one alternate synchronization input at a time. Note: The effect of either synchronization mode is to set the START/STOP bit, which is reset by hardware when in SINGLE mode, at the end of each sequence of conversions. Requirements: The External Synchronisation Input must receive a low level pulse longer than an INTCLK period and, for both External and On-Chip Event synchronisation, the repetition period must be greater than the time required for the selected sequence of conversions. Bit 2 = POW: Power Up/Power Down. This bit is set and cleared by software. 0: Power down mode: all power-consuming logic is disabled, thus selecting a low power idle mode. 1: Power up mode: the A/D converter logic and analog circuitry is enabled. Bit 1 = CONT: Continuous/Single. 0: Single Mode: a single sequence of conversions is initiated whenever an external (or internal) trigger occurs, or when the ST bit is set by software. 1: Continuous Mode: the first sequence of conversions is started, either by software (by setting the ST bit), or by hardware (on an internal or external trigger, depending on the setting of the INTG and EXTG bits); a continuous conversion sequence is then initiated. Bit 0 = ST: Start/Stop. 0: Stop conversion. When the A/D converter is running in Single Mode, this bit is hardware reset at the end of a sequence of conversions. 1: Start a sequence of conversions. EIGHT-CHANNEL ANALOG TO DIGITAL CONVERTER (A/D) ANALOG TO DIGITAL CONVERTER (Cont’d) INTERRUPT CONTROL REGISTER (AD_ICR) R254 - Read/Write Register Page: 63 Reset Value: 0000 1111 (0Fh) 7 ECV 0 AWD ECI AWDI X PL2 PL1 PL0 Bit 7 = ECV: End of Conversion. This bit is set by hardware after a group of conversions is completed. It must be reset by the user, before returning from the Interrupt Service Routine. Setting this bit by software will cause a software interrupt request to be generated. 0: No End of Conversion event occurred 1: An End of Conversion event occurred Bit 6 = AWD: Analog Watchdog. This is automatically set by hardware whenever either of the two monitored analog inputs goes out of bounds. The threshold values are stored in registers F8h and FAh for channel 6, and in registers F9h and FBh for channel 7 respectively. The Compare Result Register (CRR) keeps track of the analog inputs exceeding the thresholds. The AWD bit must be reset by the user, before returning from the Interrupt Service Routine. Setting this bit by software will cause a software interrupt request to be generated. 0: No Analog Watchdog event occurred 1: An Analog Watchdog event occurred Bit 5 = ECI: End of Conversion Interrupt Enable. This bit masks the End of Conversion interrupt request. 0: Mask End of Conversion interrupts 1: Enable End of Conversion interrupts Bit 4 = AWDI: Analog Watchdog Interrupt Enable. This bit masks or enables the Analog Watchdog interrupt request. 0: Mask Analog Watchdog interrupts 1: Enable Analog Watchdog interrupts Bit 3 = Reserved. Bit 2:0 = PL[2:0]: A/D Interrupt Priority Level. These three bits allow selection of the Interrupt priority level for the A/D. INTERRUPT VECTOR REGISTER (AD_IVR) R255 - Read/Write Register Page: 63 Reset Value: xxxx xx10 (x2h) 7 V7 0 V6 V5 V4 V3 V2 W1 0 Bit 7:2 = V[7:2]: A/D Interrupt Vector. This vector should be programmed by the User to point to the first memory location in the Interrupt Vector table containing the starting addresses of the A/D interrupt service routines. Bit 1 = W1: Word Select. This bit is set and cleared by hardware, according to the A/D interrupt source. 0: Interrupt source is the Analog Watchdog, pointing to the lower word of the A/D interrupt service block (defined by V[7:2]). 1:Interrupt source is the End of Conversion interrupt, thus pointing to the upper word. Note: When two requests occur simultaneously, the Analog Watchdog Request has priority over the End of Conversion request, which is held pending. Bit 0 = Reserved. Forced by hardware to 0. 291/320 9 ST92F120 - ELECTRICAL CHARACTERISTICS 11 ELECTRICAL CHARACTERISTICS This product contains devices to protect the inputs against damage due to high static voltages, however it is advisable to take normal precautions to avoid application of any voltage higher than the specified maximum rated voltages. For proper operation it is recommended that VIN and VO be higher than VSS and lower than V DD. Reliability is enhanced if unused inputs are connected to an appropriate logic voltage level (VDD or VSS). Power Considerations. The average chip-junction temperature, TJ, in Celsius can be obtained from: TJ = TA + PD x RthJA Where: TA = Ambient Temperature. RthJA = Package thermal resistance (junction-to ambient). PD = PINT + PPORT. PINT = IDD x VDD (chip internal power). PPORT =Port power dissipation (determined by the user) ABSOLUTE MAXIMUM RATINGS Symbol VDD Parameter A/D Converter Analog Reference AVSS A/D Converter VSS VIN Unit – 0.3 to 6.5 AVDD VINOD Value Supply Voltage V (1) V – 0.3 to VDD + 0.3 V – 0.3 to 7 V up to VDD + 0.3 VSS Input Voltage (standard I/O pins) Input Voltage (open drain I/O pins) VAIN Analog Input Voltage (A/D Converter) AVSS to AVDD V TSTG Storage Temperature – 55 to +150 °C Pin Injection Current - Digital and Analog Input ±10 mA Maximum Accumulated Pin injection Current in the device ±100 mA ESD Susceptibility 2000 V IINJ ESD 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=0. (1) AVDD can be shut down while the A/D Converter is not in use. THERMAL CHARACTERISTICS Symbol RthJA Package Value TQFP64 45 PQFP100 40 Unit °C/W RECOMMENDED OPERATING CONDITIONS Symbol TA VDD Parameter Max Operating Temperature –40 105 °C Operating Supply Voltage 4.5 5.5 V 0 VDD + 0.3 V 0(1) 24 MHz AVDD Analog Supply Voltage Internal Clock Frequency @ 4.5V - 5.5V Note: (1) 1MHz when A/D or JBLPD is used, 2.6MHz when I2C is used. 1 Unit Min fINTCLK 292/320 Value ST92F120 - ELECTRICAL CHARACTERISTICS DC ELECTRICAL CHARACTERISTICS (VDD = 5V ± 10%, TA = –40°C to +105°C, unless otherwise specified) Symbol Parameter Input High Level TTL P0[7:0]-P1[7.0]-P2[7:6] P3.3 P4.2-P4.5-P5.3 CMOS Input High Level Pure open-drain I/O P2[3:2] Input High Level Standard Schmitt Trigger VIH Comment P2[5:4]-P2[1:0]-P3[7:4] P3[2:1]-P4[4:3]-P4[1:0] P5[7:4]-P5[2:0]-P6[3:0] P7[7:0]-P8[7:0]-P9[7:0] Input High Level Pure open-drain I/O Special Schmitt Trigger TTL Value Min Typ (1) 2.0 2.0 CMOS Input Threshold Max VDD + 0.3 V VDD + 0.3 V 7.0 V 7.0 V 2.5 V VDD + 0.3 Input Voltage range Input Threshold 2.9 Input Voltage range Unit V V 7.0 V P4[7:6] Input High Level High Hyst. Schmitt Trigger P6[5:4] 2.9 TTL – 0.3 P0[7:0]-P1[7:0]-P2[7:6] P2[3:2]-P3.3-P4.2-P4.5-P5.3 CMOS – 0.3 P2[5:4]-P2[1:0]-P3[7:4] P3[2:1]-P4[4:3]-P4[1:0] P5[7:4]-P5[2:0]-P6[3:0] P7[7:0]-P8[7:0]-P9[7:0] Input Low Level Special Schmitt Trigger P4[7:6] Input Low Level High Hyst. Schmitt Trigger P6[5:4] Input Threshold Input Voltage range 0.3 x VDD – 0.3 V V V 2.1 V 2.0 V 600 mV 800 mV 900 mV – 0.3 Input Threshold Input Voltage range V V 1.5 Input Threshold Input Voltage range V VDD + 0.3 Input Voltage range Input Low Level Input Low Level Standard Schmitt Trigger VIL Input Threshold – 0.3 Input Hysteresis (2) Standard Schmitt Trigger P2[5:4]-P2[1:0]-P3[7:4] P3[2:1]-P4[4:3]-P4[1:0] P5[7:4]-P5[2:0]-P6[3:0] P7[7:0]-P8[7:0]-P9[7:0] V HYS Input Hysteresis (2) Special Schmitt Trigger P4[7:6] Input Hysteresis (2) High Hyst. Schmitt Trigger P6[5:4] 293/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS Symbol Parameter Output High Level P0[7:0]-P6[5:4] AS-DS-RW VOH Output High Level P1[7:0]-P2[7:4]-P2[1:0] P3[7:1]-P4[5:0]-P5[7:0] P6[3:0]-P7[7:0]-P8[7:0] P9[7:0]-VPWO Output Low Level P0[7:0]-P2[3:2]-P4[7:6] P6[5:4]-AS-DS-RW VOL Output Low Level P1[7:0]-P2[7:4]-P2[1:0] P3[7:1]-P4[5:0]-P5[7:0] P6[3:0]-P7[7:0]-P8[7:0] P9[7:0]-VPWO Comment Value Min Typ (1) Max Unit Push Pull, IOH= – 2mA EMR1 Register - BSZ bit = 0 (3) V DD – 0.8 V Push Pull, IOH= – 8mA EMR1 Register - BSZ bit = 1 (3) V DD – 0.8 V Push Pull, IOH= – 2mA EMR1 Register - BSZ bit = 0 (3) V DD – 0.8 V Push Pull, IOH= – 4mA EMR1 Register - BSZ bit = 1 (3) V DD – 0.8 V Push Pull / Open Drain, IOL=2mA EMR1 Register - BSZ bit = 0 (3) 0.4 V Push Pull / Open Drain, IOL=8mA, EMR1 Register - BSZ bit = 1 (3) 0.4 V Push Pull / Open Drain, IOL=2mA, EMR1 Register - BSZ bit = 0 (3) 0.4 V Push Pull / Open Drain, IOL=4mA, EMR1 Register - BSZ bit = 1 (3) 0.4 V Weak Pull-up Current R WPU P2[7:4]-P2[1:0]-P3[7:1] P4.5-P4[3:1]-P5.3-P6[3:0] P7[7:0] P8[7:0]-P9[7:0] Bidirectional Weak Pull-up VOL = 0V 30 100 400 µA Weak Pull-up Current Bidirectional Weak Pull-up VOL = 0V 100 200 450 µA P6[5:4]-AS-DS ILKIO I/O Pin Input Leakage Input/Tri-State, 0V < VIN < VDD –1 +1 µA ILKIOD I/O Pin Open Drain Input Leakage Input/Tri-State, 0V < VIN < VDD –1 +1 µA –1 +1 µA Overload Current (4) 5 mA SRR Slew Rate Rise (5) 70 mA/ns SR F Slew Rate Fall (5) 70 mA/ns ILKA/D IOV A/D Conv. Input Leakage Note: (1) Unless otherwise stated, typical data are based on TA= 25°C and VDD= 5V. They are only reported for design guide lines not tested in production. Hysteresis voltage between switching levels: characterization results - not tested. (2) Hysteresis voltage between switching levels: characterization results - not tested. (3) For a description of the EMR1 Register - BSZ bit refer to the External Memory Interface Chapter. (4) Not 100% tested, guaranteed by design characterisation. The absolute sum of input overload currents on all port pins may not exceed 100 mA. (5) Indicative values extracted from Design simulation. 294/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS AC ELECTRICAL CHARACTERISTICS (VDD = 5V ± 10%, TA = –40°C to +105°C, unless otherwise specified) Symbol Parameter INTCLK IDDRUN Run Mode Current (2) 24 MHz IDDWFI WFI Mode Current 24 MHz IDDLPWFI Low Power WFI Mode Current (3) IDDHALT HALT Mode Current IDDTR Input Transient IDD Current 4MHz / 32 Typ (1) Max Unit 45 60 2.5 mA mA/MHz 14 22 0.9 mA mA/MHz 400 600 µA 10 µA (4) - 500 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) Unless otherwise stated, typical data are based on TA = 25°C and VDD= 5V. They are only reported for design guide lines not tested in production. (2) CPU running with memory access, all peripherals switched off. (3) FLASH/EEPROM in Power-Down Mode. (4) Measured in HALT Mode, forcing a slow ramp voltage on one I/O pin, configured in input. 295/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS FLASH / EEPROM SPECIFICATIONS (VDD = 5V ± 10%, TA = –40°C to +105°C, unless otherwise specified) (1) Parameter MAIN FLASH EEPROM RELIABILITY Min Byte Program 128 kbytes Flash Program 64 kbytes Flash Sector Erase 128 kbytes Flash Chip Erase Erase Suspend Latency 16 bytes Page Update (1k EEPROM) 16 bytes Page Update (0.5k EEPROM) EEPROM Chip Erase Flash Endurance 25°C Flash Endurance -40°C +105°C EEPROM Endurance Data Retention 1 Typ 10 1.3 1.5 3 Max 1200 15 Unit µs s s s µs 30 0.16 30 100 ms 0.16 15 50 ms 70 ms 10000 3000 100000 (2) 15 cycles cycles / sector Years Note: (1) The full range of characteristics will be available after final product characterisation. (2) Relationship computation between EEPROM sector cycling and single byte cycling is provided in a dedicated STMicroelectronics Application Note (ref. AN1102/98). FLASH / EEPROM DC & AC CHARACTERISTICS (VDD = 5V ± 10%, TA = –40°C to +105°C, unless otherwise specified) Symbol VCWL Parameter Write Lock Supply Voltage (1) VCRL Read Lock Supply Voltage (2) IDD1 Supply Current (Read) IDD2 Supply Current (Write) IDD3 IDD4 TPD Supply Current (Stand-by) Supply Current (Power-Down) Recovery from Power-Down Test Conditions VDD = 5.5 V, TA = –40°C, fINTCLK = 24 MHz VDD = 5.5 V, TA = –40°C, fINTCLK = 24 MHz VDD = 5.5 V, TA = –40°C VDD = 5.5 V, TA = 105°C VDD = 4.5 V, TA = 105°C Min 3 Max 4 Unit V 1.5 2.5 V 60 mA 60 mA 100 10 10 µA µA µs Note: (1) Below the min value the FLASH / EEPROM can never be written; above the max value FLASH/EEPROM can always be written. (2) Below the min value the FLASH / EEPROM can never be read; above the max value FLASH/EEPRO M can always be read. 296/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS EXTERNAL INTERRUPT TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) N° Symbol 1 TwINTLR 2 Parameter Value Unit Formula (1) Min Low Level Minimum Pulse Width in Rising Edge Mode ≥Tck+10 50 ns TwINTHR High Level Minimum Pulse Width in Rising Edge Mode ≥Tck+10 50 ns 3 TwINTHF High Level Minimum Pulse Width in Falling Edge Mode ≥Tck+10 50 ns 4 TwINTLF Low Level Minimum Pulse Width in Falling Edge Mode ≥Tck+10 50 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. The value in the right hand two columns shows the timing minimum and maximum for an internal clock at 24MHz (INTCLK) . Measurement points are V IH for positive pulses and VIL for negative pulses. (1) 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 297/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS WAKE-UP MANAGEMENT TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, , unless otherwise specified) N° Symbol Parameter 1 TwWKPLR 2 Value Unit Formula(1) Min Low Level Minimum Pulse Width in Rising Edge Mode ≥Tck+10 ≥ 50 ns TwWKPHR High Level Minimum Pulse Width in Rising Edge Mode ≥Tck+10 ≥ 50 ns 3 TwWKPHF High Level Minimum Pulse Width in Falling Edge Mode ≥Tck+10 50 ns 4 TwWKPLF Low Level Minimum Pulse Width in Falling Edge Mode ≥Tck+10 50 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. 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. Measurement points are V IH for positive pulses and VIL for negative pulses. (1) 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 – 15 298/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS RCCU CHARACTERISTICS (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Symbol Parameter VIHRS RESET Input High Level VILRS RESET Input Low Level V HYRS RESET Input Hysteresis ILKRS RESET Pin Input Leakage Comment Value Min Input Threshold Typ (1) 3.2 Input Threshold 2.4 V V – 0.3 800 0V < VIN < VDD Unit V VDD + 0.3 Input Voltage Range Input Voltage Range Max –1 mV +1 µA Note: (1) Unless otherwise stated, typical data are based on TA = 25°C and VDD= 5V. They are only reported for design guide lines not tested in production. RCCU TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Symbol Parameter TFRS RESET Input Filtered Pulse TNFR RESET Input Non Filtered Pulse TRSPH(2) RESET Phase duration TSTR STOP Restart duration Comment Min Value Typ (1) Max 50 ns µs 20 µs 20400 x Tosc 10200 x Tosc 20400 x Tosc DIV2 = 0 DIV2 = 1 Unit µs Note: (1) Unless otherwise stated, typical data are based on TA = 25°C and VDD= 5V. They are only reported for design guide lines not tested in production. (2) Depending on the delay between rising edge of RESET pin and the first rising edge of CLOCK1, the value can differ from the typical value for +/- 1 CLOCK1 cycle. Legend: Tosc = OSCIN clock periods. PLL CHARACTERISTICS (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Symbol Parameter Comment Value Min Typ (1) Max Unit FXTL Crystal Reference Frequency 2 5 MHz FVCO VCO Operating Frequency 6 24 MHz TPLK Lock-in Time PLL Jitter 350 x Tosc 0 1000 x Tosc 1200 (2) µs ps Note: (1) Unless otherwise stated, typical data are based on TA = 25°C and VDD= 5V. They are only reported for design guide lines not tested in production. (2) Measured at 24MHz (INTCLK). Guaranteed by Design Characterisation (not tested). Legend: Tosc = OSCIN clock periods. 299/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS OSCILLATOR CHARACTERISTICS (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Symbol FOSC gm Parameter Crystal Frequency Comment Fundamental mode crystal only Oscillator Value Min Typ (1) 2 1.77 2.0 Max Unit 5 MHz 3.8 mA/V VIHCK Clock Input High Level External Clock 1.2 VDD + 0.3 V VILCK Clock Input Low Level External Clock 0.4 0.5 V ILKOS OSCIN/OSCOUT Pins Input Leakage 0V < VIN < VDD (HALT/STOP) –1 +1 µA TSTUP Oscillator Start-up Time 5 ms Note: (1) Unless otherwise stated, typical data are based on TA = 25°C and VDD= 5V. They are only reported for design guide lines not tested in production. 300/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS EXTERNAL BUS TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 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 10 TdDS (AS) DS ↑ to AS ↓ Delay TckL-4 17 ns 11 TsR/W (AS) RW Set-up Time before AS ↑ Tck x Wa+TckH-17 4 ns 25 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. 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) 301/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS EXTERNAL BUS TIMING 302/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS WATCHDOG TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, Push-pull output configuration, unless otherwise specified) N° Symbol Parameter Value Formula (1) 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. Measurement points are V OH or VIH for positive pulses and VOL or V IL for negative pulses. (1) 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. 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 303/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS STANDARD TIMER TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, Push-pull output configuration, unless otherwise specified) N° Symbol Value Parameter Formula (1) 4 x (Psc+1) x (Cnt+1) x Tck 1 TwSTOL STOUT Low Pulse Width (Psc+1) x (Cnt+1) x TSTIN with TSTIN ≥ 8 x Tck 4 x (Psc+1) x (Cnt+1) x Tck 2 TwSTOH STOUT High Pulse Width (Psc+1) x (Cnt+1) x TSTIN Min Max 167 (2) Unit ns 2.8 s (2) ns 167 ns 2.8 s with TSTIN ≥ 8 x Tck (2) (2) ns 3 TwSTIL STIN High Pulse Width ≥ 4 x Tck (2) (2) ns 4 TwSTIH STIN Low Pulse Width ≥ 4 x Tck (2) (2) 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, 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, with minimum and maximum prescaler value and minimum and maximum counter value. Measurement points are V OH or VIH for positive pulses and VOL or V IL for negative pulses. (1) Formula guaranteed by design. (2) On this product STIN is not available as Alternate Function but it is internally connected to a precise clock source directly derived from OSCIN. Refer to RCCU chapter for details about clock distribution. 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 = Standard Timer Prescaler Register content (STP): from 0 to 255 Cnt = Standard Timer Couter Registers content (STH,STL ): from 0 to 65535 TSTIN = Standard Timer Input signal period (STIN ). STANDARD TIMER TIMING 304/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS EXTENDED FUNCTION TIMER EXTERNAL TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) N° Symbol Value Parameter Formula (1) Min Unit 1 TwPEWL External Clock low pulse width (EXTCLK) - 2 x Tck + 10 ns 2 TwPEWH External Clock high pulse width (EXTCLK) - 2 x Tck + 10 ns 3 Tw PIWL Input Capture low pulse width (ICAPx) - 2 x Tck + 10 ns 4 TwPIWH Input Capture high pulse width (ICAPx) - 2 x Tck + 10 ns 5 Tw ECKD Distance between two active edges on EXTCLK ≥ 4 x Tck + 10 177 ns 6 TwEICD Distance between two active edges on ICAPx 2 x Tck x Prsc +10 177 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, 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, and minimum prescaler factor (=2). Measurement points are V IH for positive pulses and VIL for negative pulses. (1) 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. Prsc = Precsaler factor defined by Extended Function Timer Clock Control bits (CC1,CC0) on control register CR2 (values: 2,4,8). EXTENDED FUNCTION TIMER EXTERNAL TIMING 1 2 EXTCLK 5 3 4 ICAPA ICAPB 6 305/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS MULTIFUNCTION TIMER EXTERNAL TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) N° Symbol 1 Tw CTW 2 Parameter Value Unit Note - ns (1) n x 42 - ns (1) 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 (2) 6 TwLAB Distance between TINA pulse edge and the following TINB pulse edge 0 - ns (2) 7 Tw AD Distance between two TxINA pulses 0 - ns (2) 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) n = 1 if the input is rising OR falling edge sensitive n = 3 if the input is rising AND falling edge sensitive (2) 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 306/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS SCI TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 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 350 ns ns 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 307/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS SPI TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Value (1) N° Symbol Parameter Condition Unit Min Max fINTCLK / 4 fINTCLK / 2 fSPI SPI frequency Master Slave fINTCLK / 128 0 1 tSPI SPI clock period Master Slave 4 x Tck 2 x Tck ns MHz 2 tLead Enable lead time Slave 40 ns 3 tLag Enable lag time Slave 40 ns 4 tSPI_H Clock (SCK) high time Master Slave 80 90 ns 5 tSPI_L Clock (SCK) low time Master Slave 80 90 ns 6 tSU Data set-up time Master Slave 40 40 ns 7 tH Data hold time (inputs) Master Slave 40 40 ns 8 tA Access time (time to data active from high impedance state) 9 tDis 10 tV 11 Disable time (hold time to high impedance state) 0 120 ns 240 ns 120 ns ns Slave Data valid Master (before capture edge) Slave (after enable edge) Tck / 4 tHold Data hold time (outputs) Master (before capture edge) Slave (after enable edge) Tck / 4 0 12 tRise Outputs: SCK,MOSI,MISO Rise time (20% VDD to 70% VDD, CL = 200pF) Inputs: SCK,MOSI,MISO,SS 100 100 ns µs 13 tFall Outputs: SCK,MOSI,MISO Fall time (70% VDD to 20% VDD, CL = 200pF) Inputs: SCK,MOSI,MISO,SS 100 100 ns µs Note: Measurement points are V OL, VOH, VIL and VIH in the SPI Timing Diagram. (1) Values guaranteed bu 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. 308/320 1 ns ns ST92F120 - ELECTRICAL CHARACTERISTICS SPI Master Timing Diagram CPHA=0, CPOL=0 SS (INPUT) 1 13 12 SCK (OUTPUT) 4 MISO (INPUT) 6 MOSI (OUTPUT) 10 5 D7-IN 7 D6-IN D7-OUT 11 D0-IN D6-OUT D0-OUT VR000109 SPI Master Timing Diagram CPHA=0, CPOL=1 SS (INPUT) 1 13 SCK (OUTPUT) 5 MISO (INPUT) 6 MOSI (OUTPU T) 10 12 4 D7-IN 7 D6-IN D7-OUT 11 D0-IN D6-OUT D0-OUT VR000110 SPI Master Timing Diagram CPHA=1, CPOL=0 SS (INPUT) 1 13 SCK (OUTPUT) 4 MISO (INPUT) 5 D7-OUT 6 MOSI (OUTPU T) 12 10 D6-OUT D0-OUT 7 D6-IN D7-IN 11 D0-IN VR000107 SPI Master Timing Diagram CPHA=1, CPOL=1 SS (INPUT) 1 12 SCK (OUTPUT) MISO (INPUT) MOSI (OUTPUT) 5 13 4 6 10 D7-IN 7 D7-OUT 11 D6-IN D6-OUT D0-IN D0-OUT VR000108 309/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS SPI Slave Timing Diagram CPHA=0, CPOL=0 SS (INPUT) 1 2 12 13 SCK (INPUT) 4 MISO HIGH-Z (OUTPU T) 8 MOSI (INPUT) 3 5 D7-OUT D6-OUT 10 D0-OUT 11 D7-IN 9 D6-IN D0-IN 7 6 VR000113 SPI Slave Timing Diagram CPHA=0, CPOL=1 SS (INPUT) 1 2 SCK (INPUT) 5 MISO HIGH-Z (OUTPU T) 8 MOSI (INPUT) 13 12 3 4 D7-OUT D6-OUT 10 D0-OUT 11 D7-IN 9 D6-IN D0-IN 7 6 VR000114 SPI Slave Timing Diagram CPHA=1, CPOL=0 SS (INPUT) 2 SCK (INPUT) HIGH-Z MISO (OUTPUT) 1 4 13 3 5 D7-OUT D6-OUT 8 10 MOSI (INPUT) 12 D7-IN D0-OUT 11 9 D6-IN D0-IN 7 6 VR000111 SPI Slave Timing Diagram CPHA=1, CPOL=1 SS (INPUT) 2 1 12 SCK (INPUT) HIGH-Z MISO (OUTPUT) 5 8 D6-OUT 10 D7-IN 6 3 4 D7-OUT MOSI (INPUT) 13 D0-OUT 11 D6-IN 9 D0-IN 7 VR000112 310/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS I2C/DDC-BUS ELECTRICAL SPECIFICATIONS (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Symbol Parameter Standard mode I2C Fast mode I2C Min Max Min Max fixed input levels N/A N/A 0.2 VDD-related input levels N/A N/A 0.05 VDD Pulse width of spikes which must be suppressed by the input filter N/A N/A 0 50 Unit Hysteresis of Schmitt trigger inputs VHYS TSP V ns Low level output voltage (open drain and open collector) VOL1 at 3 mA sink current 0 0.4 0 0.4 VOL2 at 6 mA sink current N/A N/A 0 0.6 250 20+0.1Cb 250 TOF 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 V ns with up to 6 mA sink current at VOL2 N/A N/A 20+0.1Cb 250 I Input current each I/O pin with an input voltage between 0.4V and 0.9 VDD max -10 10 -10 10 µA C Capacitance for each I/O pin 10 pF 10 Legend: N/A = not applicable Cb = capacitance of one bus in pF 311/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS I2C/DDC-BUS TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Protocol Specifications Symbol Parameter Formula fINTCLK Internal Frequency (Slave Mode) fSCL SCL clock frequency TBUF Bus free time between a STOP and START condition THIGH SCL clock high period TLOW SCL clock low period THD:STA Hold time START condition. After this period, the first clock pulse is generated = 000 = 001 = 010 = 011 Data set-up time (Without SCL stretching) TSU:DAT Data set-up time (With SCL stretching) Min Max Min Max 0 100 0 400 THIGH – 3 x Tck kHz 4.7 1.3 µs 4.0 0.6 µs 4.7 µs 1.3 TLOW + Tck 4.0 0.6 µs TLOW + T HIGH – THD:STA 4.7 0.6 µs 3 x Tck 4 x Tck 4 x Tck 10 x Tck 0 (1) 0 (1) 250 100 0.9 (2) ns TLOW – THD:DAT FREQ[2:0] FREQ[2:0] FREQ[2:0] FREQ[2:0] = 000 = 001 = 010 = 011 7 x Tck 15 x Tck 15 x Tck 31 x Tck TR Rise time of both SDA and SCL signals 1000 TF Fall time of both SDA and SCL signals 300 TSU:STO Set-up time for STOP condition Cb Unit MHz 2 x (THIGH – 3 x Tck) Fast Mode TSU:STA Set-up time for a repeated START condition THD:DAT Data hold time Fast I2C 2.5 Standard Mode FREQ[2:0] FREQ[2:0] FREQ[2:0] FREQ[2:0] Standard I2C Capacitive load for each bus line TLOW + T HIGH – THD:STA 4.0 ns 20+ ns 0.1Cb 20+ ns 0.1Cb 0.6 400 ns 400 pF Note: (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 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. Cb = total capacitance of one bus line in pF FREQ[ 2:0] = Frequency bits value of I2C Own Address Register 2 (I2COAR2) 312/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS I2C TIMING SDA t BUF t LOW tR tF t HD:STA t SP SCL t HD:STA P S t SU:STO t HD:DAT t HIGH t SU:DAT t SU:STA Sr P 313/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS J1850 BYTE LEVEL PROTOCOL DECODER TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) Value Symbol Parameter Receive Mode Transmission Mode Unit Note - µs (1)(2) Min Max Nominal 0 ≤7 TF Symbols Filtered TIB Invalid Bit Detected >7 ≤ 34 - µs (1)(2) TP0 Passive Data Bit “0” > 34 ≤ 96 64 µs (1)(2)(3) TA0 Active Data Bit “0” > 96 ≤ 163 128 µs (1)(2)(3) TP1 Passive Data Bit “1” > 96 ≤ 163 128 µs (1)(2)(3) TA1 Active Data Bit “1” > 34 ≤ 96 64 µs (1)(2)(3) TNBS Short Normalization Bit > 34 ≤ 96 64 µs (1)(2)(3) TNBL Long Normalization Bit > 96 ≤ 163 128 µs (1)(2)(3) TSOF Start Of Frame Symbol > 163 ≤ 239 200 µs (1)(2)(3) TEOD End Of Data Symbol > 163 ≤ 239 200 µs (1)(2)(3) TEOF End Of Frame Symbol > 239 - 280 µs (1)(2)(3) TBRK Break Symbol > 239 - 300 µs (1)(2)(3) TIDLE Idle Symbol > 280 - 300 µs (1)(2)(3) Note: (1) Values obtained with internal frequency at 24 MHz (INTC LK), with CLKSEL Register set to 23. (2) In Transmission Mode, symbol durations are compliant to nominal values defined by the J1850 Protocol Specifications. (3) All values are reported with a precision of ±1 µs. J1850 PROTOCOL TIMING T SOF T P0 T A0 T P1 T A1 T EOD T NB S T ID LE T EOF EOD T A0 T P1 T A1 T EOD T NB L EOF / IDLE “1” SHORT T P0 NB SHORT “1” LONG SOF “0” LONG T “0” SHORT SOF VPWO T ID LE T EOF 314/320 1 EOF / IDLE NB LONG EOD “1” SHORT “1” LONG “0” LONG “0” SHORT SOF VPWO ST92F120 - ELECTRICAL CHARACTERISTICS A/D EXTERNAL TRIGGER TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) N° Symbol Parameter Value (Note) Formula Min. Max. Unit 1 TwLOW External trigger pulse width 1.5 x Tck 62.5 - ns 2 TwHIGH External trigger pulse distance 1.5 x Tck 62.5 - ns 3 Tw EXT External trigger active edges distance 138 x n x Tck n x 5.75 - µs 0.5 x Tck 1.5 x Tck 20.8 62.5 ns 4 TdSTR EXTRG falling edge and first conversion start 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. Legend: Tck = INTCLK period = OSCIN period when OSCIN is not divided by 2; 2*OSCIN period when OSCIN is divided by 2; OSCIN period / PLL factor when the PLL is enabled. n = number of autoscanned channels (1 ≤ n ≤ 8) A/D EXTERNAL TRIGGER TIMING 315/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS A/D CHANNEL ENABLE TIMING TABLE (VDD = 5V ± 10%, TA = –40°C to +105°C, CLoad = 50pF, fINTCLK = 24MHz, unless otherwise specified) N° Symbol 1 TwEXT Parameter CEn Pulse width Value (Note) Formula Min. Max. 138 x n x Tck n x 5.75 - Unit µs 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. Legend: Tck = INTCLK period = OSCIN period when OSCIN is not divided by 2; 2*OSCIN period when OSCIN is divided by 2; OSCIN period / PLL factor when the PLL is enabled. n = number of autoscanned channels (1 ≤ n ≤ 8) A/D CHANNEL ENABLE TIMING 316/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS A/D ANALOG SPECIFICATIONS (VDD = 5V ± 10%, TA = –40°C to +105°C, fINTCLK = 24MHz, unless otherwise specified) Parameter Typical Minimum Maximum Units (1) Notes Conversion time 138 INTCLK (2)(6) Sample time 85 INTCLK (6) Power-up time 60 µs (6) Resolution 8 8 Monotonicity bits GUARANTEED No missing codes GUARANTEED Zero input reading 00 Full scale reading Hex FF (6) Hex (6) 0.5 LSBs (1)(4)(6) Gain error 0.6 LSBs (4)(6) DLE (Diff. Non Linearity error) 0.6 LSBs (4)(6) ILE (Int. Non Linearity error) 1.0 LSBs (4)(6) 1.0 LSBs (4)(6) kΩ (3)(5)(6) pF (5)(6) µA (6) Offset error 0.3 TUE (Absolute Accuracy) –1.0 Input Resistance 1.3 Hold Capacitance 1.4 Input Leakage 0.8 2.7 ±1 Note: “1LSBideal” has a value of AVDD/256 Including sample time This is the internal series resistance before the sampling capacitor This is a typical expected value, but not a tested production parameter. If V(i) is the value of the i-th transition level (0 ≤ i ≤ 254), the performance of the A/D converter has been evaluated as follows: OFFS ET ERROR= deviation between the actual V(0) and the ideal V(0) (=1/2 LSB) GAIN ERROR= deviation between the actual V(254) and the ideal V(254) - V(0) (ideal V(254)=AVDD-3/2 LSB) DNL ERROR= max {[V(i) - V(i-1)]/LSB - 1} INL ERROR= max {[V(i) - V(0)]/LSB - i} ABS. ACCURA CY= overall max conversion error (5) Simulated value, to be confirmed by characterisation. (6) The specified values are guaranteed only if an overload condition occurs on a maximum of 2 non-selected analog input pins and the absolute sum of input overload currents on all analog input pins does not exceed ±10 mA. (1) (2) (3) (4) 317/320 1 ST92F120 - ELECTRICAL CHARACTERISTICS Figure 124. A/D Conversion Characteristics Offset Error OSE Gain Error GE 255 254 253 252 251 250 ( 2) code out 7 ( 1) 6 5 (5) 4 (4) 3 (1) Example of an actual transfe rcurve (2) The ideal transf er curve (3) Differe ntial no n-linearity error (DNL) (4) Integral non-lineari tyerror (INL) (5) Center of a ste p of the actual transfer curve (3) 2 1 1 LSB (ideal) 0 1 2 3 Offset Error OSE 318/320 1 4 5 6 7 250 251 252 253 254 255 256 Vin(A) (LSBideal ) VR02133A ST92F120 - GENERAL INFORMATION 12 GENERAL INFORMATION 12.1 PACKAGE MECHANICAL DATA Figure 125. 64-Pin Thin Quad Flat Package 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 K L L1 inches Max 0° 3.5° 0.031 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 Max Min NE 16 K Figure 126. 100-Pin Plastic Quad Flat Package 0.10mm .004 seating plane Dim mm Min Typ A inches Typ Max 3.40 0.134 A1 0.25 A2 2.55 2.80 3.05 0.100 0.110 0.120 B 0.22 0.38 0.009 0.015 C 0.13 0.23 0.005 0.009 D 22.95 23.20 23.45 0.904 0.913 0.923 D1 19.90 20.00 20.10 0.783 0.787 0.791 D3 0.010 18.85 0.742 E 16.95 17.20 17.45 0.667 0.677 0.687 E1 13.90 14.00 14.10 0.547 0.551 0.555 E3 12.35 e 0.65 K L 0° 0.486 0.026 7° 0.65 0.80 0.95 0.026 0.031 0.037 L1 1.60 0.063 Number of Pins PQFP100 N 100 ND 30 NE 20 319/320 1 ST92F120 - 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 320/320 1