I'm having trouble generating specific time for the STM32F103C8 (Blue Pill). Apparently, the AHB main clock is set to 72 MHz. However, regardless of whether the SysTick clock source is AHB or AHB/8, the time always turns out to be 10 times longer.
Clock config
void delay(){
SysTick->LOAD = 7199999;
SysTick->CTRL = 0x05;
while((SysTick->CTRL&(1<<16)) == 0);
SysTick->CTRL = 0x00;}
This delay should be 0.1 sec. But it always works in 1 sec. Other values are also 10 times higher, regardless of whether CLKSOURCE is AHB or AHB/8.
If anyone can help, I appreciate it.
Below is how I use the Systick timer:
void delaySysTicks(uint32_t msDelay) {
for(uint32_t c = 0; c < msDelay; c++) {
SysTick->CTRL = 0x0; // disable systick
SysTick->LOAD = (SystemCoreClock / 1000U); // count down reload - 1ms
SysTick->VAL = 0;
SysTick->CTRL |= SysTick_CTRL_CLKSOURCE_Msk;
SysTick->CTRL |= SysTick_CTRL_ENABLE_Msk;
while(!(SysTick->CTRL & SysTick_CTRL_COUNTFLAG_Msk)); // wait for timer rollover
}
}
delaySysTicks(1000); // busy wait 1 second
Related
Afternoon all
I'm looking for some assistance please with something that has been confusing me whilst trying to learn timer interrupts
You'd be best treating me as a novice. I have no specific goal here other than learning something that I think would be useful feather to add to my cap!
I have written the below sketch as a stab at a rigid framework for executing different fcns at different rates. I've done something similar using millis() and whilst that worked I found it inelegant that a) there was no obvious way to check for task overruns and backing-up the execution rate and b) the processor is bunged up by checking millis() every program cycle.*
Essentially what I think should be a 1ms timer interrupt on Timer2 (16MHz/64 prescaler /250 compare register =1000hz) is coming out around 0.5ms. I've been confused for hours on this but I'm prepared to accept it could be something fundamental/basic!
What's also throwing a spanner in the works is that using serial comms to try and debug the faster task rates seems to slow things down considerably, so I'm inferring the problem by counting up 1ms tasks to call 10,100 and 1000ms tasks and debugging at the slower level. I suppose chewing through a few characters at 9600baud probably is quite slow.**
I've pasted the code below. Any pointers highly appreciated. Be as harsh as you like :)
cheers
Al
*Whilst not what I'm confused about - any comments on my logic here also welcome
** Although I don't get how Serial.println manages to slow the program down. It's driven from interrupts so it should surely just drop the comms and perform the next ISR - effectively a task overrun. Any comments here also welcome
//NOTES
//https://www.robotshop.com/letsmakerobots/arduino-101-timers-and-interrupts
//https://sites.google.com/site/qeewiki/books/avr-guide/timers-on-the-atmega328
//http://ww1.microchip.com/downloads/en/DeviceDoc/Atmel-42735-8-bit-AVR-Microcontroller-ATmega328-328P_Datasheet.pdf
//
//INITIALISE EVERYTHING
const int ledPin = 13;
volatile int nStepTask1ms = 0; // init 0 - to be used for counting number of 1ms tasks run and hence calling 10ms task
volatile int nStepTask10ms = 0;
volatile int nStepTask100ms = 0;
volatile int nStepTask1000ms = 0;
volatile int LEDFlashState = 0; // init 0 - variable to flip when LED on
volatile int millisNew = 0; // to store up to date time
volatile int millisOld = 0; // to store prev time
volatile int millisDelta = 0; // to store deltas
//
void setup()
{
Serial.begin(9600); //set up serial comms back to PC
pinMode(ledPin,OUTPUT); //to flash the embedded LED
noInterrupts(); //turn off interrupts while we set the registers
//set up TIMER first
TCCR2A = 0; //sets TCCR1A byte to zero, bits to be later individually mod'd
TCCR2B = 0; //sets TCCR1B byte to zero, bits to be later individually mod'd
TCNT2 = 0; //ensures counter value starting from zero
Serial.println("Timer1 vars reset");
TCCR2B |= (1<<WGM12); // bitwise or between itself and WGM12. TCCR2B = TCCR2B | 00001000. Sets WGM12 high. (CTC mode so WGM12=1, WGM 13,11,10 all 0) https://stackoverflow.com/questions/141525/what-are-bitwise-shift-bit-shift-operators-and-how-do-they-work
Serial.println("Mode 4 CTC set");
Serial.println("TCCR2B=");
Serial.println(TCCR2B,BIN);
TCCR2B |= (1<<CS11); // sets CS11 high
TCCR2B |= (1<<CS10); // sets CS10 high (i.e. this and above give /64 prescaler)
Serial.println("Prescaler set to 64");
OCR2A = 250; //compare match register for timer2
Serial.println("Compare Register set");
Serial.println("OCR2A=");
Serial.println(OCR2A);
TIMSK2 |= (1 << OCIE2A); //enables interrupts - https://playground2014.wordpress.com/arduino/basics-timer-interrupts/
Serial.println("Interrupt Mask Register Set");
Serial.println("TIMSK2=");
Serial.println(TIMSK2);
interrupts(); //enable interrupts again - not sure if this is required given OCIE1A being set above?
}
//set up ISR for Timer2 - timer structure called every interrump (1ms) that subsequently calls 1,10,100 and 1000msec task fcns
ISR(TIMER2_COMPA_vect)
{
TASK_1ms();
if (nStepTask1ms>9)
{
TASK_10ms();
if (nStepTask10ms>9)
{
TASK_100ms();
if (nStepTask100ms>9)
{
TASK_1000ms();
}
}
}
}
void TASK_1ms()
{
// 1ms tasks here
nStepTask1ms++;
}
void TASK_10ms()
{
//10ms tasks here
nStepTask1ms=0;
nStepTask10ms++;
}
void TASK_100ms()
{
//100ms tasks here
nStepTask10ms=0;
nStepTask100ms++;
//Serial.println(nStepTask100ms);
}
void TASK_1000ms()
{
//1000ms tasks here
nStepTask100ms=0;
//do something
changeLEDFlashState();
//check timing tick of this task
millisNew=millis();
millisDelta=millisNew-millisOld;
Serial.println(millisDelta);
millisOld=millisNew;
nStepTask1000ms++;
}
void changeLEDFlashState()
{
if(LEDFlashState==0)
{
digitalWrite(ledPin,HIGH);
LEDFlashState=1;
//Serial.println("LED Turned On");
}
else
{
digitalWrite(ledPin,LOW);
LEDFlashState=0;
//Serial.println("LED Turned Off");
}
}
void loop()
{
// empty
}
You have two lines here:
TCCR2B |= (1<<CS11); // sets CS11 high
TCCR2B |= (1<<CS10); // sets CS10 high (i.e. this and above give /64 prescaler)
These two lines set the lower three bits of TCCR2B to 011, which is a /32 prescaler.
Note that for Timer1 and Timer2, the prescaler settings are different than Timer0.
For Timer0, the settings above would give you a /64 prescaler.
I developed an application to generate the pulses(pwm) using timer of lpc1769. The period of pulse is 10milli seconds and pulse width can be varied as required. The pulse width is generated depending on the reference signal which is a square wave of 10milli seconds(with on period of 7.2ms and off period of 2.8ms). whenever there is rising edge of this signal PWM pulse should start. Now this is working fine. To detect the rising edge i used GPIO interrupt with external interrupt3 the ISR is as :
1) if rising edge on GPIO pin(P2.11)
2) Clear the rising edge status on the pin.
3) then set the timer for few milli seconds
and in timer ISR
1) Clear any pending Timer IRQs
2)Make the GPIO pin high on which generating PWM pulse(P2.6)
3) set the timer for few milli seconds and then clear the pin(P2.6)(same timer is used in both the ISRs)
4) Disabling the timer and re enabling the rising edge interrupt on GPIO pin(P2.11), so again serves the ISR of external interrupt3 on rising edge of reference signal and it continues as above.
Now i am getting the problem in the code in developing the application as,
1) if rising edge disable the interrupt and set timer to 8 milli seconds delay.
2) In timer ISR generate 10 pwm cycles after this re enable the external interrupt. (the same timer is used in both ISRs).
3) so my output should be with respect to rising edge of reference signal 8 milli second delay then 10 pwm cycles(all have time period of 10ms) again 8ms delay with respect to rising edge of reference signal and 10 pwm cycles.
But once after 10 cycles when i re enable the external interrupt irrespective of rising edge of reference signal a delay of 8ms is added.
whenever last 10th cycle is complete I am enabling the interrupt again, so from this point only it adding a 8ms delay. My question is if interrupt is enabled for rising edge of reference signal then it should serve ISR for rising edge only. But this is not happening in this case. I am not understanding this behaviour.
/////////////////////////////////////////////////////////////////////////////////////////////////////
This is my code, this is ISR for external interrupt3
void GPIO_IRQ_HANDLER(void)
{
uit32_t i;
if(((Chip_GPIOINT_GetStatusRising(LPC_GPIOINT, GPIO_INTERRUPT1_PORT) >> GPIO_INTERRUPT1_PIN) & 0x01) != 0) /*Group 1*/
{
count1 = 0;
start_timer1((HALF_WAVE_PERIOD - DELAY) * timerTick);
NVIC_ClearPendingIRQ(GPIO_INTERRUPT_NVIC_NAME);
NVIC_DisableIRQ(GPIO_INTERRUPT_NVIC_NAME);
}
}
Timer1 ISR is
void TIMER1_IRQHandler(void)
{
if (Chip_TIMER_MatchPending(LPC_TIMER1, 1) == 0)
{
return;
}
Chip_TIMER_ClearMatch(LPC_TIMER1, 1);
if(count1 > 10)
{
LPC_GPIO1->PIN = (LPC_GPIO1->PIN & ~(1 << 19));
stop_timer1();
}
else
{
switch(state1)
{
case 1:
state1 = 2;
LPC_GPIO1->PIN = (LPC_GPIO1->PIN | (1 << 19));
set_timer1(2000 * timerTick);
break;
case 2:
state1 = 1;
LPC_GPIO1->PIN = (LPC_GPIO1->PIN & ~(1 << 19));
count1++;
set_timer1((HALF_WAVE_PERIOD - DELAY) * timerTick);
break;
default:
break;
}
}
}
where,
HALF_WAVE_PERIOD = 10000 microseconds
DELAY = 2800 microseconds
GPIO_INTERRUPT1_PORT= PORT2
GPIO_INTERRUPT1_PIN = P2.12
GPIO_INTERRUPT_NVIC_NAME = EINT3_IRQn
GPIO_IRQ_HANDLER = EINT3_IRQHandler
and functions start_timer1 is used to initialize the timer1 and set the timer1 for given number of ticks and interrupt on match.
set_timer1 : set timer1 for given ticks and interrupt.
stop_timer1 : disable the timer1 and enables the external interrupt3.
programming on MSP430 in CCS
Using Timer_A, ACLK and his interrupt to blink LED(just blinking now- same long time torned off- same time turned on).
This code blink led with 2 sec delay. There is problem that register TA1CCR0 can be max 0xFFFF= 65535 (2 sec for ACLK). And for my application(blinking LED is only exapmle) I will need scale from 1 sec to 999 sec. (row 6-7 in code). How can I do that? Is it possible?
#include <msp430.h>
#include <msp430f6736.h>
void CfgTA(unsigned long delayCycles)
{
int t2=2; // must be variable from 1 to 999
t2=delayCycles*t2;
TA1CCTL0 |= CCIE; //Enable Interrupts on Timer
TA1CCR0 = t2-1; //Number of cycles in the timer
TA1CTL |= TASSEL_1 | MC_1; //ACLK , UP mode
}
void ledblink()
{
//LED config
P4DIR |= BIT6;
P4OUT &= ~BIT6;
CfgTA(32768); //Timer configuration to blink every 1 sec
while (1)
{
_bis_SR_register(LPM3_bits + GIE); //Enter Low Power Mode 3 with interrupts
}
}
#pragma vector=TIMER1_A0_VECTOR
__interrupt void Timer_A0(void)
{
P4OUT ^= BIT6; // Swapping on/off LED
}
int main(void) {
WDTCTL = WDTPW | WDTHOLD; // Stop watchdog timer
ledblink();
return 0;
}
how to count seconds on a 1 second interrupt.
1) initialize interrupt to occur once each second and reload its' timer/counter register
2) set global variable to number of seconds to delay:
int delaySeconds = 10;
3) inside the interrupt function
static int count =0;
count++;
if( count >= delaySeconds )
{
count = 0;
P4OUT ^= BIT6; // Swapping on/off LED
}
I think the interrupt function, before exiting, also needs to clear the time1 interrupt pending flag
On your MSP430, you can slow down ACLK with the DIVA field in the UCSCTL5 register, and you can further divide down the timer's clock input with the ID and IDEX fields in the TAxCTL and TAxEX0 registers.
With the timer input divided down to 16 Hz, you would be able to count for up to 4096 seconds.
I'm probably just having a can't-see-the-forest-for-the-trees moment with one of these registers, but I can't get the PWM working on the dsPIC33FJ32MC102 microcontroller (warning: big PDF) I'm playing around with. I've followed both the datasheet and further application note (warning: another PDF) and even code samples and I can't see what I'm doing wrong, though on my testbench I'm getting Vcc on the high output and Ground on the low output. I've tied the fault pins both to Vcc and disabled the register keycode so my changes should see some effect. What am I doing wrong?
#define FOSC (3686400ULL)
#define FCY (FOSC/2)
#include <xc.h>
#include <libpic30.h>
...
#pragma config PWMPIN = ON // Motor Control PWM Module Pin Mode bit (PWM module pins controlled by PORT register at device Reset)
#pragma config PWMLOCK = OFF
...
void main(void){
...
//setup PWM
//Clear faults
IFS3bits.PWM1IF = 0;
IFS3bits.FLTA1IF = 0;
IFS4bits.FLTB1IF = 0;
//Setup dead times
P1DTCON1bits.DTAPS = 0b00; //Dead time tick is 1 TCY
P1DTCON1bits.DTBPS = 0b00;
P1DTCON1bits.DTA = 10; //Dead time is 10TCY ~= 3uS
P1DTCON1bits.DTB = 10;
P1DTCON2bits.DTS1A = 0; //Active and Inactive transition dead times
P1DTCON2bits.DTS2A = 0; //0 takes A dead time
P1DTCON2bits.DTS3A = 0; //1 takes B dead time
P1DTCON2bits.DTS1I = 1;
P1DTCON2bits.DTS2I = 1;
P1DTCON2bits.DTS3I = 1;
P1TCONbits.PTOPS = 0b0000; //1 CPU tick = 1 PWM tick
P1TCONbits.PTCKPS = 0b00;
P1TCONbits.PTMOD = 0b00;
P1TCONbits.PTSIDL = 0; //Run when CPU idles
// no longer necessary since I disabled register write lock:
// __builtin_write_PWMSFR(&P1FLTACON,0x0000,&PWM1KEY);
// __builtin_write_PWMSFR(&P1FLTBCON,0x0000,&PWM1KEY);
// __builtin_write_PWMSFR(&PWM1CON1 ,0x0077,&PWM1KEY);
PWM1CON1 = 0x0077; //Enable all 3 channels
P1FLTACON = 0x0000; //Disable faults
P1FLTBCON = 0x0000;
//Setup Wave freq/duty
//Fosc = 7.3728 MHz -> Fcy = 3.6864MHz
//Desire a PWM of 20250Hz (smaller scalar error than 20kHz)
//P1TPER = [Fcy/(Fpwm*Scalar)] - 1
//Therefore P1TPER = [3.6864M/(20250*1)] - 1 = 181;
P1TPER = 181;
P1DC1 = 0x7FFF; // 0x7FFF for 50%
P1DC2 = 0x7FFF;
P1DC3 = 0x7FFF;
P1OVDCON = 0x3F00; //Disable override; override disables PWM
PWM1CON2 = 0x0000;
P1TCONbits.PTEN = 1; //Turn on
...
while(1);
}
I believe you have set up the timer to count from 0 to 181 (P1TPER) and then reset and repeat. But you have set the duty cycle registers to 0x7FFF, which is greater than 181. So I believe the duty cycle value will never be less than the timer value and therefore the output will never change. Try setting the duty cycle registers to 181/2 = 90 to get a duty cycle of 50%.
I have an atmega168a chip. I use Counter 0 to toggle PORTC by using ISR(TIMER0_COMPA_vect) and ISR(TIMERB_COMPA_vect) interrupt sub-routines. I would like to activate the 16-bit timer when if condition is true. So, I use TIMSK1 = (1<<OCIE1A), but this line calls ISR(TIMER1_COMPA_vect) interrupt instantly where I want 16 bit timer to be interrupted only when the counter reaches to OCR1A value. How can I activate the 16-bit timer on the run-time without causing an instant interrupt?
here is my code:
#define F_CPU 8000000
#include <avr/io.h>
#include <avr/interrupt.h>
#include <stdint.h>
volatile uint8_t counter;
int main (void){
DDRC = 0xFF; //sets PORTC as output
PORTC = 0xFF; //initial output value
/*COUNTER0 settings*/
TIMSK0 = ((1<<OCIE0A) | (1<<OCIE0B)); // Enable Interrupt TimerCounter0 Compare Match A & B
TCCR0A = (1<<WGM01); // Mode = CTC
TCCR0B = (1<<CS01) | (1<<CS00); // Clock/64, 1/(8000000/64)= 0.000008 seconds per tick
OCR0A = 200; // 0.000008 *230 = 1.6 ms
OCR0B = 100; // 0.8 ms
/*16bit timer - counter1 settings*/
TIMSK1 &= ~(1<<OCIE1A); // Disable Interrupt Counter 1, output compare A (TIMER1_CMPA_vect)
TCCR1B = ((1<<CS12) | (1<<CS10) | (1<<WGM12)); // Clock/1024, 1/(8000000/1024) = 0.000128 seconds per tick, Mode=CTC
OCR1A = 40; // 0.000128*40 ~= 5.12 milliseconds
sei(); //interrupts are globally switched on
counter =0;
while(1){
if(counter >= 4){
TCNT1 = 0; // clear the counter 1
TIMSK1 = (1<<OCIE1A);// Enables the interrupt for Counter 1,(TIMER1_CMPA_vect)
TIMSK0 &= ~((1<<OCIE0A) | (1<<OCIE0B)); //disables the Counter 0's interrupts
counter = 0;
}
}
return 0;
}
ISR(TIMER0_COMPA_vect){ //1.6ms
PORTC = 0xFF;
counter++;
}
ISR(TIMER0_COMPB_vect){ //0.8 ms
PORTC = ~PORTC;
}
ISR(TIMER1_COMPA_vect){ // 5.2 milisecond interrupt
PORTC = 0x00;
TCNT0 = 0; //clear the counter of counter0
// TIMSK0 = ((1<<OCIE0A) | (1<<OCIE0B)); //Enable the Counter 0 interrupts
// TIMSK1 &= ~(1<<OCIE1A);// Disable Interrupt Counter 1, output compare A (TIMER1_CMPA_vect)
}
here is an oscilloscope output that shows why I don't want the interrupt to be set instantly, because it set the signal 0 instantly.
I think the problem might be that in CTC mode, interrupt is generated when OCF1A flag is set (in TIFR). Since your timer is always running, just not generating interrupts, it sets OCF1A flag, which never gets cleared. On page 142 in the datasheet it says:
OCF1B is automatically cleared when the Output Compare Match B
Interrupt Vector is executed. Alternatively, OCF1B can be cleared by
writing a logic one to its bit location.
This means that when you set up timer 1, you also need to clear OCF1A:
TIFR1 &= ~(1<<OCF1A)
However, I think you can do better. You could just stop the timer when not needed, and start it when you do, instead of twiddling the TIMSK and having timer 1 run always. If you set TCCR1B to zero, that clears CS12, CS11, and CS10, which, according to the datasheet means "Timer stopped." Then, when your counter reaches 4 you can turn on timer1 as you have it above:
TCCR1B = ((1<<CS12) | (1<<CS10) | (1<<WGM12));
If you do this, you shouldn't need to turn timer 1 interrupts on and off: just leave them on, and only turn the counting on when you need it.
Also I am wondering if it is actually necessary to fire off two interrupts to toggle pins on PORTC? Are you not using PWM for that because it doesn't give you the pulse lengths precisely enough?