CCS/MSP430F5529: To stop PWM

One possibility is to set a separate timer to control how long the PWM channel runs.
e,g PRF is 500kHz, you want 37 pulses. Set up a timer to run for 74 us. Start the timer and the clock to the PWM channel, kill the clock to the PWM channel when the timer expires.

Another approach: wrap the PWM output back around to a counter-timer channel and configure said channel to generate an ISR after the desired number of pulses.

Hi Jeff,

Thanks for your replay.  Can you please send me an example code for the second option that you have mentioned.  First option is not working out for me.

Regards,

Dipu.

void void_GenerateStepInput() // function to generate step input
{

	XTimerPeriod    = 400;
	void_EnableStepperMotor(XAxis);//Enable stepper motor along X Axis
	Timer_B_outputPWMParam GenerateStepInputXParam = {0};//Setting TA2 parameters for step input generation.
	GenerateStepInputXParam.clockSource 		   = TIMER_B_CLOCKSOURCE_SMCLK;
	GenerateStepInputXParam.clockSourceDivider     = TIMER_B_CLOCKSOURCE_DIVIDER_1;
	GenerateStepInputXParam.timerPeriod            = XTimerPeriod;
	GenerateStepInputXParam.compareRegister        = TIMER_B_CAPTURECOMPARE_REGISTER_2;
	GenerateStepInputXParam.compareOutputMode      = TIMER_B_OUTPUTMODE_TOGGLE;
	GenerateStepInputXParam.dutyCycle              = XTimerPeriod/2;
	Timer_B_outputPWM(TIMER_B0_BASE,&GenerateStepInputXParam); //step input generation 
}
void TIMER0_B1_SpecificDistance_ISR(void)
{
	if(++TimerCountX >= NumberofSteps) 
	{
		void_TerminateStepInput();
		Timer_B_disableCaptureCompareInterrupt(TIMER_B0_BASE,TIMER_B_CAPTURECOMPARE_REGISTER_2);
		TimerCountX = 0;
	}

	Timer_B_clearCaptureCompareInterrupt(TIMER_B0_BASE,TIMER_A_CAPTURECOMPARE_REGISTER_2);
}

Thanks for your replay.  My code is attached herewith.

Regards,

Dipu.

This seems to be missing void_TerminateStepInput, which was kind of the point. That said:

The quickest way to stop the PWM and leave the output at 0 (low) would be: "TB0CCTL2 &= ~(OUTMOD_7|OUT);//OUTMOD=0,OUT=0". [I don't know the DriverLib language for that.]

Doing that in the CCR2 interrupt (as you appear to be doing) would probably result in a very short "glitch" pulse, which may or may not be forgiven. I would suggest doing this in a TBCCR0 (or TBIFG) interrupt instead. By my reading, the signal at that moment would be 0, and will stay that way for 200 clocks, so you shouldn't see any glitch.

Is there any chance that what you're seeing is actually overshoot from the abrupt stoppage of the pulse train?

[Edit: Fixed goof in register name. Twice.]