• State TI Thinks Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 1 answer
  • Subscribers 82 subscribers
  • Views 707 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TMS320F28069: HRPWM using SYNCI - Can there be a "Dead Band"?

John Westmoreland43
Guru 13875 points
Part Number: TMS320F28069

Hello,

I'm using one of the  HRPWM signal outputs to generate a 'clock' - call that "X" MHz.  I am using the EPWMSYNCI signal which consists of a few signals that are ANDed together; and that SYNCI signal has a duty cycle of ~50%; and the longest period it could be deasserted is 500mS.  I was using the EPWM7A output; but I have moved this to EPWM2A and am using both of those outputs now; not that this affects anything regarding to what I think is like a 'dead band' for lack of a better description right now.

When the EPWMSYNCI signal is valid (the AND is '1') - the output of 'X'MHz is fine; but when the signal is invalid (the AND is '0') - I'm seeing a sort of 'drift' of the 'X'MHz signal.  The signal on EPWM2A (configured of course as HRPWM) - tends to drift somewhat.  I wanted to know if this is something expected or maybe I have a configuration bit not set correctly; but once the SYNCI signal is valid again; all is well.  I just didn't think this could drift a bit for the time the EPWMSYNCI wasn't valid.  In addition to looking at this on a scope; I have a calibrated frequency counter on the output of the EPWM2A signal and the output frequency does change slightly during the period EPWMSYNCI isn't valid.

I also don't currently require the SYNCO to be used; I suppose it's OK to not have this configured even though I'm using the SYNCI signal; but it would be nice to have TI confirm this.

Thanks,
John W.

  • 0
    TI__Mastermind 42195 points

    John,

    1. Does it drift during extended 'high' periods as well as extended 'low' periods on EPWMSYNCI?
    2. Are you seeing drift in one direction, where it keeps getting further from the ideal? Or are you seeing more of a jitter behavior where the output is sometimes longer and sometimes shorter than expected?
    3. Are you running the SFO?
    4. Is it possible that your systems voltage or temperature is drifting?

    Finally, there is no need to use SYNCO, this only propagates the signal further into the sync chain. If you don't need to sync more PWMs/CAPs then disabling SYNCO is the correct thing to do.

    Regards,
    Cody 

  • 0 in reply to Cody Watkins
    Guru 13875 points
    Cody,

    I will answer what I can right now - further answers will require me to do some more testing and set up -

    1 - It appears to be OK during the 'high' periods but not in the 'low' periods.

    2 - I don't think it's jitter - but to make sure which direction the drift is actually in will require me to do some more testing - I'll let you know
    when I have an answer to that.

    3 - I have looked at the SFO but I am not running that right now. Can I run the SFO with a SYSCLK of 90MHz?

    4 - Voltage and temp should be OK - the board does get a little warm but it stays at a pretty constant temp once running. Note I have some ECL on this and that adds some heat; but I wouldn't think this is an issue at this time.

    OK - good to hear about SYNCO.

    Thanks,
    John
  • 0 in reply to John Westmoreland43
    Guru 13875 points

    Cody,

    Also - I am leaving all of the dead band register settings to their default settings right now.

    Thanks,
    John

  • 0 in reply to John Westmoreland43
    TI__Mastermind 42195 points
    John,
    if your board was slowly heating up I would expect the HRPWM's MEP size to slowly drift if you aren't running the SFO. Additionally I don't know of any restrictions on device frequency when running the SFO cal. library.

    Regards,
    Cody
  • 0 in reply to Cody Watkins
    Guru 13875 points

    Cody,

    OK - I'll add the SFO cal. library - maybe this could help with the 'drift' issue too.

    Thanks,
    John 

  • 0 in reply to John Westmoreland43
    TI__Mastermind 42195 points
    John,
    OK! It shouldn't need to be ran too often, only frequently enough to remove errors due to voltage and temperature fluctuations(which is largely system dependent so you will need to do a bit of investigation to fine tune the frequency).


    Regards,
    Cody
  • 0 in reply to Cody Watkins
    TI__Mastermind 42195 points

    John,

    i'm going to close this thread due to inactivity, feel free to post back if you have anymore information or mark helpful posts.

    Regards,
    Cody 

  • 0 in reply to Cody Watkins
    Guru 13875 points
    Hello Cody,

    I'm definitely still working on the PWM stuff but right now I'm adding a PID to the PWM control I have - I will get back to this but if you need to close the thread then it's OK - I still haven't completely rolled the SFO cal lib in.

    Thanks,
    John
  • 0 in reply to John Westmoreland43
    TI__Mastermind 42195 points

    John, 

    no problem, we can leave it open for the time being but the thread will automatically lock after 30 days on inactivity.

    Regards,
    Cody 

  • 0 in reply to Cody Watkins
    Guru 13875 points
    Cody,

    OK - understood.

    Thanks,
    John