HRPWM outputs on 28035

Hi Peter,

Did you try with the HRPWM example code present in controlSuite?

Regards,

Gautam

Hi Gautam,

the code was initially using the controlSuite example. Now I want to use outB of the modlues as separate PWM outputs. In my first hardware version, the outputs was digital used, but now I want to dimm an LED on each outB.

I have 3 PWM outputs. On HRPWM2 and 3, the separate outputs works, but not on HRPWM1. The difference is only in SYNC unit (PWM1 as master, other as slave)

Regards

Peter

Hi all,

the problem is still not solved, is there any idea about this issue?

Peter

Hi Peter,

I have not seen an issue like this previously.

If PWM2 is working and PWM1 is not, how is PWM1's configuration different from PWM2?

One thing that I notice above is that PHSEN should be set to TB_DISABLE for the master.  This shouldn't cause an issue in your case, but it is a more correct implementation. 


Thank you,
Brett

Hi Brett,

thank you for your hint. My initialisation for PWM2 is nearly the same see below. Any ideas?

Regards, Peter

void HRPWM2_Config(period)
{
// ePWM2 register configuration with HRPWM
// ePWM2A toggle high/low with MEP control on falling edge
    EALLOW;
    EPwm2Regs.TBCTL.bit.PRDLD = TB_SHADOW;            // set Immediate load
    EPwm2Regs.TBCTL.bit.FREE_SOFT = 11;
    EPwm2Regs.TBPRD = period-1;                            // PWM frequency = 1 / period
    EPwm2Regs.CMPA.half.CMPA = period / 2;              // set duty 50% initially
    EPwm2Regs.CMPA.half.CMPAHR = (1 << 8);              // initialize HRPWM extension
    EPwm2Regs.TBPHS.all = 0;
    EPwm2Regs.TBCTR = 0;

    EPwm2Regs.TBCTL.bit.CTRMODE = TB_COUNT_UPDOWN;
    EPwm2Regs.TBCTL.bit.PHSEN = TB_ENABLE;                // EPwm3 is the Master
    EPwm2Regs.TBCTL.bit.SYNCOSEL = TB_SYNC_IN;
    EPwm2Regs.TBCTL.bit.HSPCLKDIV = TB_DIV1;
    EPwm2Regs.TBCTL.bit.CLKDIV = TB_DIV1;

    EPwm2Regs.CMPCTL.bit.LOADAMODE = CC_CTR_ZERO;
    EPwm2Regs.CMPCTL.bit.LOADBMODE = CC_CTR_ZERO;
    EPwm2Regs.CMPCTL.bit.SHDWAMODE = CC_SHADOW;
    EPwm2Regs.CMPCTL.bit.SHDWBMODE = CC_SHADOW;

    EPwm2Regs.AQCTLA.bit.CAU = AQ_CLEAR;                  // PWM toggle high/low
    EPwm2Regs.AQCTLA.bit.CAD = AQ_SET;
    EPwm2Regs.AQCTLB.bit.CBU = AQ_CLEAR;
    EPwm2Regs.AQCTLB.bit.CBD = AQ_SET;

    EPwm2Regs.ETSEL.bit.SOCAEN    = 1;                       //enable SOCA pulse generation
    EPwm2Regs.ETSEL.bit.SOCASEL    = ET_CTR_ZERO;        //pulse on counter equals zero
    EPwm2Regs.ETPS.bit.SOCAPRD     = ET_1ST;                //pulse on every event
    EPwm2Regs.ETSEL.bit.INTEN = 1;
    EPwm2Regs.ETSEL.bit.INTSEL = ET_CTR_ZERO;
    EPwm2Regs.ETPS.bit.INTPRD     = ET_1ST;

    EPwm2Regs.HRCNFG.all = 0x0;
    EPwm2Regs.HRCNFG.bit.EDGMODE = HR_BEP;               //MEP control both edges
    EPwm2Regs.HRCNFG.bit.SELOUTB = HR_NORM_B;            //Chan. B
    EPwm2Regs.HRCNFG.bit.CTLMODE = HR_CMP;
    EPwm2Regs.HRCNFG.bit.HRLOAD  = HR_CTR_ZERO_PRD;
    EPwm2Regs.HRCNFG.bit.AUTOCONV  = 1;                //Autoconversion ist an

    EPwm2Regs.HRPCTL.bit.TBPHSHRLOADE = 1;
    EPwm2Regs.HRPCTL.bit.HRPE = 1;

    // Configure EPWM-2 pins using GPIO regs
    GpioCtrlRegs.GPAPUD.bit.GPIO2 = 1;    // Disable pull-up on GPIO2 (EPWM2A)
    GpioCtrlRegs.GPAPUD.bit.GPIO3 = 1;    // Disable pull-up on GPIO3 (EPWM2B)

    GpioCtrlRegs.GPAMUX1.bit.GPIO2 = 1;   // Configure GPIO2 as EPWM2A
    GpioCtrlRegs.GPAMUX1.bit.GPIO3 = 1;   // Configure GPIO3 as EPWM2B

    EDIS;
}

Hi Peter,

Can you define what, "on HRPWM1 this does not work" means?  Does the B channel output of HRPWM1 stay high, low, or does it behave in an unexpected manner?

I have one thought:
In the HRPWM guide for the F28035, you will find a section on the proper configuration order for HRPWMs.  This includes disabling TBCLKSYNC prior to all PWM configuration and enabling TBCLKSYNC afterward.  I would recommend making sure that your configuration follows the order described in the documentation.

Outside of this, as you mentioned, the two PWM configurations look similar.  If you have a second MCU/board, does the same thing happen on it?  Have you ruled out your board hardware, leaving us with only the F28035 to blame?


Thank you,
Brett