Using Timers in PIC18F Microcontrollers
5 likes6,663 views
TIMER0 is an 8/16-bit timer in PIC18F microcontrollers that can be used to track time. It uses an internal or external clock source, and its frequency can be divided with a prescaler. TIMER0's value is stored in registers and increments in hardware. When it overflows at the maximum value, an interrupt flag is set. Software can check this flag to trigger events at timed intervals without needing to continuously check the timer value. Using overflow mode reduces code complexity compared to clearing the timer value periodically.
Using Timers in PIC18F Microcontrollers
- 1. Using Timers of Microchip PIC18F Microcontrollers Corrado Santoro ARSLAB - Autonomous and Robotic Systems Laboratory Dipartimento di Matematica e Informatica - Universit`a di Catania, Italy santoro@dmi.unict.it L.A.P. 1 Course Corrado Santoro Using Timers in PIC18F MCUs
- 2. What is a “Timer”? It is a circuit to let a software have the “knowledge of flow of time” It is a composed of: A clock source; usually the system clock or an external signal; A programmable frequency divisor, called prescaler, to divide clock source frequency, if needed; Some SFRs which hold a 8-, 16- or 32-bit value that is incremented in hardware using the clock source. Some SFRs which give some state information, e.g overflow (zero crossing). PIC18F family has 7 timers, called TIMER0, TIMER1, ..., TIMER5, TIMER6 Each timer has different characteristics and may be used together with other peripherals. Corrado Santoro Using Timers in PIC18F MCUs
- 3. The TIMER0 of PIC18 TIMER0 is a 8/16 bit timer/counter (figure shows the 8bit mode); TMR0L is the SFR containing the value that is incremented; All the parts to the left are the clock source circuits. T0CON (Timer 0 Control) register is used to program the timer, and includes the bits shown in figure (T0CS, PSA, T0PS, etc.) Corrado Santoro Using Timers in PIC18F MCUs
- 4. The 16-bit version of TIMER0 In 16-bit mode, two SFR are used TMR0L and TMR0H; In write operations, TMR0H must be written before TMR0L; In read operations, TMR0L must be read before TMR0H; However, XC8 offers a single 16-bit variable TMR0 which includes both low and high part of TMR0. Corrado Santoro Using Timers in PIC18F MCUs
- 5. The 16-bit version of TIMER0 The T0CON (Timer 0 Control) SFR includes all the bits which control TIMER0 functioning. Corrado Santoro Using Timers in PIC18F MCUs
- 6. TIMER0: Selecting clock source Clock source can be internal or external and is controlled by bit T0CS: T0CS = 0; → clock source is internal and is taken from Fosc/4. T0CS = 1; → clock source is external and is taken from T0CKI pin; in this case T0SE controls the edge of the signal which triggers increment. Corrado Santoro Using Timers in PIC18F MCUs
- 7. TIMER0: Dividing clock frequency In some cases, the clock coming from the oscillator could be too fast for our applications: we can lower it by using the frequency prescaler. The prescaler is a circuit which divides the signal frequency by 2, 4, 8, 16, ..., 256. The prescaler is activated by bit PSA: PSA = 0; → prescaler is selected, frequency division is controlled by bits T0PS. PSA = 1; → prescaler is not selected. Corrado Santoro Using Timers in PIC18F MCUs
- 8. TIMER0: Dividing clock frequency When the prescaler is activated (PSA = 0), division is performed as: T0PS = 111, division 1:256 T0PS = 110, division 1:128 T0PS = 101, division 1:64 .... T0PS = 000, division 1:2 Corrado Santoro Using Timers in PIC18F MCUs
- 9. TIMER0: controlling depth and on/off Finally, T0CON includes these other two bits: TMR0ON, turns on/off the timer; T08BIT, selects 8 (value “1”) or 16 (value “0”) bit mode. Corrado Santoro Using Timers in PIC18F MCUs
- 10. A case-study: a timer to flash a LED We want to use the system clock, T0CS = 0; In our board, we have FOSC = 64MHz, therefore the basic frequency is FOSC/4 = 16MHz, the P = 62.5ns; Let’s use the prescaler and divide the frequency by 256, so PSA = 0; T0PS = 0b111; The timer increments using a period P = 62.5ns · 256 = 16µs. Corrado Santoro Using Timers in PIC18F MCUs
- 11. A case-study: a timer to flash a LED ... the timer increments using a period P = 62.5ns · 256 = 16µs. Let us suppose we want a period of half a second 500ms Therefore 500·10−3 16·10−6 = 31250 A delay of 500ms implies 31250 counts Corrado Santoro Using Timers in PIC18F MCUs
- 12. A case-study: a timer to flash a LED ✞ int main(void) { TRISBbits.TRISB0 = 0; // output T0CONbits.TMR0ON = 0; // stop the timer T0CONbits.T08BIT = 0; // timer configured as 16-bit T0CONbits.T0CS = 0; // use system clock T0CONbits.PSA = 0; // use prescaler T0CONbits.T0PS = 0b111; // prescaler 1:256 (’0b’ is a prefix for binary) TMR0 = 0; // clear timer value T0CONbits.TMR0ON = 1; // start the timer for (;;) { unsigned int t; t = TMR0; if (t >= 31250) { // equivalent of 500 ms TMR0 = 0; LATBbits.LATB0 = !LATBbits.LATB0; } } } ✡✝ ✆ Corrado Santoro Using Timers in PIC18F MCUs
- 13. Case-study 2: more LEDs flashing Let us suppose we want to: flash led in RB0 at a period of 500 ms flash led in RB1 at a period of 750 ms Do we need two timers?? NO! 1 compute the greatest common divisor, which is 250ms 2 use it as your “timer period” 3 toggle RB0 after two periods 4 toggle RB1 after three periods Corrado Santoro Using Timers in PIC18F MCUs
- 14. Case-study 2: more LEDs flashing Using the same set-up of the previous example, since our period is 250ms we have 250·10−3 16·10−6 = 15625 A delay of 250ms implies 15625 counts Corrado Santoro Using Timers in PIC18F MCUs
- 15. Case-study 2: more LEDs flashing ✞ int main(void) { char c0 = 0, c1 = 0; // why char? because they are 8 bits TRISBbits.TRISB0 = 0; // output TRISBbits.TRISB1 = 0; // output T0CONbits.TMR0ON = 0; // stop the timer T0CONbits.T08BIT = 0; // timer configured as 16-bit T0CONbits.T0CS = 0; // use system clock T0CONbits.PSA = 0; // use prescaler T0CONbits.T0PS = 0b111; // prescaler 1:256 (’0b’ is a prefix for binary) TMR0 = 0; // clear timer value T0CONbits.TMR0ON = 1; // start the timer for (;;) { unsigned int t; t = TMR0; if (t >= 15625) { // equivalent of 250 ms TMR0 = 0; ++c0; ++c1; if (c0 == 2) { // flash led 0 LATBbits.LATB0 = !LATBbits.LATB0; c0 = 0; } if (c1 == 3) { // flash led 1 LATBbits.LATB1 = !LATBbits.LATB1; c1 = 0; } } } } ✡✝ ✆ Corrado Santoro Using Timers in PIC18F MCUs
- 16. Timer Overflow In our examples, we check the timer value and, after reaching a certain maximum, we clear it However, what does it happen if we don’t modify TMR0? At a certain point, the TMR0 reaches its maximum possible value, which is 255 (0xff) at 8 bit and 65535 (0xffff) at 16 bit The next increment will overflow TMR0, which thus goes to zero This event is signalled by the hardware by setting a proper bit in a SFR The bit is called T0IF and belongs to register INTCON The bit set by the hardware and cleared by software Corrado Santoro Using Timers in PIC18F MCUs
- 17. Timer Overflow We can exploit the overflow event as follows. Instead of clearing TMR0 and waiting for reaching our MAX (15625 in the example), we can: Set TMR0 to “65536 − MAX” (“65536 − 15625 = 49911” in our example) Wait for overflow by checking T0IF Clear T0IF Corrado Santoro Using Timers in PIC18F MCUs
- 18. Case-study 2: LED flashing with overflow ✞ int main(void) { char c0 = 0, c1 = 0; // why char? because they are 8 bits TRISBbits.TRISB0 = 0; // output TRISBbits.TRISB1 = 0; // output T0CONbits.TMR0ON = 0; // stop the timer T0CONbits.T08BIT = 0; // timer configured as 16-bit T0CONbits.T0CS = 0; // use system clock T0CONbits.PSA = 0; // use prescaler T0CONbits.T0PS = 0b111; // prescaler 1:256 (’0b’ is a prefix for binary) TMR0 = 49911; // initial timer value INTCONbits.T0IF = 0; // clear the overflow bit initially T0CONbits.TMR0ON = 1; // start the timer for (;;) { if (INTCONbits.T0IF == 1) { // overflow! TMR0 = 49911; // reload timer INTCONbits.T0IF = 0; // clear overflow ++c0; ++c1; if (c0 == 2) { // flash led 0 LATBbits.LATB0 = !LATBbits.LATB0; c0 = 0; } if (c1 == 3) { // flash led 1 LATBbits.LATB1 = !LATBbits.LATB1; c1 = 0; } } } } ✡✝ ✆ Corrado Santoro Using Timers in PIC18F MCUs
- 19. Timer Overflow Let’s consider the expression: “65536 − MAX”: We notice that 65536, in 16-bit arithmetic, does not exist and is equivalent to 0 therefore, “65536 − MAX = −MAX” Corrado Santoro Using Timers in PIC18F MCUs
- 20. Case-study 2: LED flashing with overflow ✞ int main(void) { char c0 = 0, c1 = 0; // why char? because they are 8 bits TRISBbits.TRISB0 = 0; // output TRISBbits.TRISB1 = 0; // output T0CONbits.TMR0ON = 0; // stop the timer T0CONbits.T08BIT = 0; // timer configured as 16-bit T0CONbits.T0CS = 0; // use system clock T0CONbits.PSA = 0; // use prescaler T0CONbits.T0PS = 0b111; // prescaler 1:256 (’0b’ is a prefix for binary) TMR0 = -15625; // initial timer value INTCONbits.T0IF = 0; // clear the overflow bit initially T0CONbits.TMR0ON = 1; // start the timer for (;;) { if (INTCONbits.T0IF == 1) { // overflow! TMR0 = -15625; // reload timer INTCONbits.T0IF = 0; // clear overflow ++c0; ++c1; if (c0 == 2) { // flash led 0 LATBbits.LATB0 = !LATBbits.LATB0; c0 = 0; } if (c1 == 3) { // flash led 1 LATBbits.LATB1 = !LATBbits.LATB1; c1 = 0; } } } } ✡✝ ✆ Corrado Santoro Using Timers in PIC18F MCUs
- 21. Comparing the techniques Let’s compare (1) ✞ unsigned int t; t = TMR0; if (t >= 15625) { // equivalent of 250 ms TMR0 = 0; ✡✝ ✆ to (2) ✞ if (INTCONbits.T0IF == 1) { // overflow! TMR0 = -15625; // reload timer INTCONbits.T0IF = 0; // clear overflow ✡✝ ✆ (1) uses a 16-bit comparison, (2) uses a single-bit comparson → less code since the CPU is 8-bit (2) uses polling but can be easily transformed into a interrupt-based code since overflows can be programmed to generate interrupts Corrado Santoro Using Timers in PIC18F MCUs
- 22. Using Timers of Microchip PIC18F Microcontrollers Corrado Santoro ARSLAB - Autonomous and Robotic Systems Laboratory Dipartimento di Matematica e Informatica - Universit`a di Catania, Italy santoro@dmi.unict.it L.A.P. 1 Course Corrado Santoro Using Timers in PIC18F MCUs