• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 81 subscribers
  • Views 835 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

pwm low duty cycle

Felix Zhang
Intellectual 645 points

Hi,

I configure ePWM1A/1B and 2A/2B 107kHz complementary output, using F28035 @60MHz.

I'm facing a problem that when the register "EPwm1Regs.CMPA.half.CMPA" is writen by "1","2" or "3", there is no output PWM waveform, till I increase it to 4 or more than 4.

Is there any solution for this? thanks!

Regards,

Felix

  • 0
    TI__Expert 4760 points

    Hi Felix,

    After discussing this with some of the C2000 apps. engineers, this does not appear to be a documented problem with the PWM module.

    One possibility is that this could be a deadband issue.  Do you have any values loaded for deadband?  If so, set DBRED and DBFED registers to zero, and see if this fixes the problem.

    Other data which may help us debug the problem:

    1. Are you using Up-count or up-down count mode?
    2. What are the action qualifier settings?
    3. Is the PWM time base getting synchronized to an external/internal signal?
    4. Are you using dead-band to generate the complementary output? If yes, which mode Active-hi/Active-lo?

    … there may be more questions that will arise once we get answers to these questions.

    Thanks, 

    Dave

     

  • 0 in reply to Dave (Wisconsin) Wilson
    TI__Guru 55700 points

    Felix,

    It is possible that the pulse generated is small enough the you are not able to capture the pulse on oscilloscope because of sample taken / resolution issue.

    Did you try to triggerring for rising edge in your digital oscilloscope?

    Regards,

    Manoj

  • 0 in reply to Dave (Wisconsin) Wilson
    Intellectual 645 points

    Hi Dave,

    Yes, I loaded values for deadband, after I remove that values the problem didn't occure.

    But there must be a deadband in my application, how do I evade the confilict? Is there any solution for that, thanks!

    Regards,

    Felix

  • 0 in reply to Manoj Santha Mohan
    Intellectual 645 points

    Hi Manoj,

    After I clear for the deadband I got PWM output of CMPA=1, so it should not be oscilloscope issue, thanks a lot!

    Regards,

    Felix

  • 0 in reply to Felix Zhang
    TI__Expert 4760 points

    Hi Felix,

    Are you designing a power stage consisting of totem-pole transistors stacked in a top and bottom configuration?  If so, go ahead and put your deadband in, and you should still be OK.  In fact, this is the way all PWM modules I have ever worked with operate.  With deadband enabled, when the PWM value gets close to zero or to the upper modulo limit, the pulse width will collapse to a very small value, and then disappear altogether during the deadtime.  This is OK.  In fact, you don't want pulse transitions occuring during the deadtime interval, or you could violate the switching time requirements of your transistors, and cause shoot-thru current.

    -Dave

     

  • 0 in reply to Dave (Wisconsin) Wilson
    TI__Guru 55700 points

    Felix,

    Yes, it should not be oscillscope resolution issue.

    Regards,

    Manoj

  • 0 in reply to Dave (Wisconsin) Wilson
    Intellectual 645 points

    Hi Dave,

    Ok, thank you for your explanation, I tried to reduce the deadband and got more output when it gets close to zero.

    Regards,

    Felix