CCS/TMS320F28335: HRPWM looks exactly like EPWM

David,

In addition what Elizabeth suggested, can you zoom in on the falling edge? It looks like there is something bad going on there (why is the waveform so thick)?

Generally the way I test for HR movement is to setup another PWM identically and in sync without HRPWM enabled. Put these both on the scope at the same time and play with the CMPAHR/CMPBHR values on the PWM with HR enabled to see the movement.

Regards,

Kris

Thanks.

I have zoomed the falling edge, find below. It truly looks weird. I will also place the two side by side for better comparison. But in the mean time, what do you think about this?

yeah.
I am setting the CMPAHR and CMPA based on the output of adc conversion. Where dutycalc is the output of the adc conversion between 0 and 1. The method is as follows:

__interrupt void  epwm_isr(void)
{
  fract = modf(dutycalc*EPwm4Regs.TBPRD,&intduty);				//extract integer and fractional part
  EPwm4Regs.CMPA.half.CMPAHR = (fract*MEP_ScaleFactor+1.5)*256; //mutiply mep and add 1.5 to keep it btw 1 and 255. Multiply with 256 is same as <<8 //1st 8bits are unused
  EPwm4Regs.CMPA.half.CMPA = intduty;							//integer part of new duty cycle
  //EPwm4Regs.CMPA.half.CMPA = dutycalc*EPwm4Regs.TBPRD;
  EPwm4Regs.ETCLR.bit.INT = 1;       // Clear INT SEQ1 bit
  PieCtrlRegs.PIEACK.all = PIEACK_GROUP3;   // Acknowledge interrupt to PIE
  return;
}

Also, I successfully ran the Example_2833xHRPWM. I learnt from it but it wasn't completely applicable to my case. I feel if I get good result with this then I can move on to Scale Factor Optimization........

David

I agree with you. The falling edge is messed up.
I am not driving anything at all. I am just plotting the pwm output on a scope. The hrpwm initialization is as follows:

void Setup_HRPWM4A(void)
{
	EPwm4Regs.TBCTL.bit.CLKDIV = 0;			//SysClkOut/1 ie default
	EPwm4Regs.TBCTL.bit.HSPCLKDIV = 0;		//SysClkOut/1 but /2 is default //TBCLK = SysClkOut = 150MHz
	EPwm4Regs.TBCTL.bit.CTRMODE = 0;		//UP COUNT
	EPwm4Regs.TBCTL.bit.PRDLD = 0;			//TBPRD period shadow, load on ctr = 0
	EPwm4Regs.TBCTL.bit.SYNCOSEL = 3;		//disable sync output select
	EPwm4Regs.TBPRD = 4989/11;				//330kHz
	EPwm4Regs.CMPA.half.CMPA = 0;			//initial CMPA = 0

	EPwm4Regs.CMPA.half.CMPAHR = (1 << 8);       // initialize HRPWM extension
	EPwm4Regs.CMPCTL.bit.LOADAMODE = 0;
	EPwm4Regs.CMPCTL.bit.SHDWAMODE = 0;

	EPwm4Regs.AQCTLA.all = 0x0012;			//ctr = 0 (set), ctr = cmpa on up count (clear)
	EPwm4Regs.ETSEL.bit.SOCAEN = 1;        // Enable SOC on A group
	EPwm4Regs.ETSEL.bit.SOCASEL = 1;       // Select SOC from CTR=0
	EPwm4Regs.ETPS.bit.SOCAPRD = 1;        // Generate pulse on 1st event
	EPwm4Regs.ETSEL.bit.INTEN = 1;        // enable pwm interrupt
	EPwm4Regs.ETSEL.bit.INTSEL = 2;        // enable interrupt on ctr = PRD
	EPwm4Regs.ETPS.bit.INTPRD = 1;        // interrupt on 1st event

	EALLOW;
	EPwm4Regs.HRCNFG.all = 0x0;
	EPwm4Regs.HRCNFG.bit.EDGMODE = 2;		 // MEP control on falling edge
	EPwm4Regs.HRCNFG.bit.CTLMODE = 0;		//CMPAHR control
	EPwm4Regs.HRCNFG.bit.HRLOAD  = 0;		//load on ctr = 0
	EDIS;
}

Yes. Line 23-28 as well as 11 are commented out when I am not using HRPWM. I learnt of the 3cycle restriction (or delay before hrpwm kicks in), but I wasn't too bothered because I think I only need it for the Falling Edge. This is because I am using up-count.

Do you think that can be an issue?

David.

Hello Kris,

I have placed the two pwms together. Yellow - No hrpwm. Blue - with hrpwm



There is really no difference. What do you think?

Thanks.

David.

Hello,

This is a gentle reminder about my query.

Thanks.

David.

Thanks for your response.

I have tried debugging the pwms. I tried lower frequencies and I got good results. I think the falling edge blur may be because of the frequency I am working with and the fact that the adc output is never constant even at constant analogue input.

Ordinarily, this is what I am getting:

I only start seeing the blur whenever I zoom the falling edge. Is there a way to keep the output of the ADC constant as it seems its fluctuation is responsible for the blur?

Or what other thing can I try?

By the way, since I am working with duty cycles, the material suggested by Elizabeth advised working on CMPAHR and not TBPHSHR so I am not sure how I can avoid using CTR = 0. If I try using up-down, them I'd likely have to worry about both edges leading to TBPHSHR.

Hope to hear from you soon.

Kind regards.

David.

Thanks Kris.

Hope to hear from you soon.

David.