• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 83 subscribers
  • Views 304 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.

TMS320F28388D: Is there a maximum delay between synchronized PWM outputs

jimj2713
Expert 2395 points

Part Number: TMS320F28388D

We have a motor-controller product that requires synchronized PWM outputs, For the two-sets of 3-phase outputs, PMW2-6 are synchronized to PWM1. What I am seeing at the DSP-output, is everything is within ~2ns except PWM4, which has about 10-12ns delay from the PWM1 edges. I don't see a specified maximum time-delay in the datasheet. Is this 10-12ns delay between PWM1 and PWM4 reasonable?

Thanks,

Jim

  • 0
    TI__Guru 52200 points

    You should be able to perfectly synchronize your EPWMs in a given device.

  • 0 in reply to Nima Eskandari
    Expert 2395 points

    Nima,

    Thanks for the quick reply. So I measured <2ns between 4 of the 6 PWM outputs -- should the synchronization even be less than that? I would have assumed there would be some tolerance on propagation delays through all of the internal gates, so ~2ns felt possibly OK, and the question was -- is 10ns too much?

    Also, I have the deadtime registers identical between all 6 PWMs, all 6 TBPHS=0, and PWM2-6 EPWMSYNCINSEL= 1 (for PWM1). What else could be done to make this synchronization better?

    Thanks,

    Jim

  • 0 in reply to jimj2713
    TI__Guru 52200 points

    Okay you are seeing 2ns! That is very small. Are you using HRPWM?

    Can you also share the screenshot for this delay you are capturing so I can double check?

  • +1 in reply to Nima Eskandari
    Expert 2395 points

    Nina,

    You indicated 2ns is very small (good), so let me restate the original question -- is 12ns too large or unexpected?

    I don't have a picture of the oscilloscope measurement, but the delay from PWM4-AHI to PWM1-AHI was ~12ns, and the other PWM2, 3, 5, and 6 edge-to-edge measurements were ~2ns. Is this 12ns measurement within the possible gate-propagation-delays within the DSP, or is this 12ns too large and I may have a problem?

    Thanks,

    Jim

  • 0 in reply to jimj2713
    TI__Guru 52200 points

    You are running at 200MHz. Each coarse EPWM step is 5ns. If you are seeing numbers above 5ns it because your SW is not correct.

    For example if you sync two EPWM modules to each other, and set the TBPHS of the SYNC receiver to 0, that doesnt mean they will be in perfect sync. 

    if the EPWMCLK = TBCLK, you need TBPHS=2 for a perfect sync and if EPWMCLK>TBCLK (meaning you are using a non-1 divider for HSDIV/DIV for the TBCLK) then the TBPHS=1 would give you a sync.

    Now that means with the correct SW you can get the TBCTR values to be in perfect sync. With that said, you can still have up to 2.5ns skew across two EPWM modules.