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.

TMS320F280049: Trip zone configuration to prevent EPWMxA and EPWMxB signals overlaping

Part Number: TMS320F280049

Hi E2E Community,

Currently I am setting up EPWM sub-module to output two PWM output signals to drive a half bridge inverter. These two PWM signals (EPWMxA and EPWMxB) are interleaved with 180 degree phase delay. In the past, we used external logic circuit to prevent the PWM outputs overlapping (If EPWMxA and EPWMxB are both high, it will cause shoot-through in the inverter bridge).  The Dead_band sub-module seems to work with complementary PWM signals to prevent overlapping,  but not with interleaved PWM signals (Please correct me if I am wrong).

My question is here: Is the trip-zone logic able to prevent the two PWM signals overlapping in interleaved PWM signals?

Thanks for your support,

Qinghua

  • Qinghua,

    The tripzone submodule is really more of a pin level check to react to a condition as opposed to "preventing" a condition. With that being said, if your application can handle (rough guess) 10-30 ns of overlap before the output is shut off then there may be an option.

    On this device, the Input X-Bar could route your PWM outputs (at the pin level) back to the same PWM module generating them. You can then use the Digital Compare submodule to look for a both high condition and generate a trip event.

    Regards,

    Kris

  • Qinghua,

    What Kris has said is correct, the Trip-Zone is a reactive module not a preventative module. Stated another way, the trip-zone can correct overlap in the even of a software bug, but it cannot prevent overlap before it happens.

    The Dead-Band sub-module can apply 1 rising(RED) and 1 falling edge delay(FED) to both ePWMxA and ePWMxB. Yes, you can use the Dead-Band with interleaved signals. Furthermore, you can apply the RED and FED to any non-complementary signals. However, this will not guarantee that your signals are not overlapping. 

    There are two ways to ensure that you signals do not overlap:

    1. Hardware, as you have stated.
    2. Software, this will require you to error check the compare values as well as the RED and FED before they are written to their respective registers. 

    Regards,
    Cody 

  • Thank Kris and Cody for your help on my issue. My post title may be misleading a little bit. Actually I am looking for the trip_zone logic to detect the overlapping of two interleaved PWMs in case of software fault,  and then shut down the ePWM output to protect the hardware. I will take a careful look at Kris's suggestion on Digital Comparing sub-module and Cody's suggestion on Dead_Band configuration. I will come back to post my update later.

  • Hi Kris,

    I have went through many times with the technical reference of EPWM Digital Compare Sub-module, but still not figure out how to properly use it. Could you please share more details (or an example) how I can use Digital Compare Sub-module to detect the overlapping of EPWMxA and EPWMxB and trigger One-Shot Tripping?

    Thanks a lot,

    Qinghua
  • Quinghua,

    1. Place both outputs on the input X-Bar
    2. Assign one on DCAH and the other on DCAL
    3. Setup DCxEVTy to disable the PWM when DCAH and DCAL are both logic high

    You may need to invert one or both of the DCAx signals.

    Regards,
    Cody 

  • Hi Cody,

    Very appreciate your help. I can understand the steps you listed above, but not sure how to invert DCAH or DCAL signals. Shall I do it in Input XBAR or DC Sub-module? Thanks.

    Qinghua

  • Qinghua,
    It shouldn't matter where you invert it.

    Regards,
    Cody
  • Cody,

    Could I ask one more question: which register shall be used to invert DCAH or DCAL signals?

    Thanks,

    Qinghua
  • Qinghua,

    you can invert the signal in the ePWM X-Bar before the signal reaches DCAH or DCAL using the "TRIPOUTINV" register.

    Hope it helps!
    Cody 

  • Thank you, Cody. This shall solve my issue to detect overlapping PWM signals.
  • Qinghua,
    it is common practice to click the green "Resolved" button on only the posts that contain the answer to your questions. In the future this makes it easier for people looking at the thread to find the answers.

    Thanks,
    Cody