• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 814 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.

WEBENCH® Tools/TPS92691: Erratic Gate Drive

Brandon McLellan
Prodigy 40 points
Part Number: TPS92691

Tool/software: WEBENCH® Design Tools

I'm using a TPS92691 in a Buck-Boost application using the component values in Webench, and I'm finding the Gate Drive acting erratic.

The Gate output abruptly switches off (short duration) and then back on to complete the cycle (based on the Switching freq.).

The duration of the off time doesn't match the 35uS of the IS Fault, the IS line is staying under 525mV, and Soft Start is not indicating a Reset from a Fault.

The PWM pin is tied to Vcc and Iadj pin is tied to PWM through a 100K resistor.

The Vcc pin on the Driver has a very clean 7.5VDC.

Changing values of the Compensation components doesn't impact the timing of the problem.

What does help is increasing the output capacitance significantly beyond the Webench recommended value.

An additional odd thing to me in doing this, is that a Polymer Hybrid capacitor works better than multiple X5R ceramic capacitors in parallel.

I can't find anything in the datasheet that can explain why the driver would act this way.

  • 0
    TI__Mastermind 25500 points

    Hello Brandon,

    It could be a number of things. Could you share your schematic and BOM? I'd like to check some things there first (webench isn't always right on). It would also be good to get scope shots of what you are seeing (at least the gate, drain, LED current, and Vout if possible). If those look ok we can start considering the layout.

    Regards,

    Clint

  • 0 in reply to Clinton Jensen
    Prodigy 40 points

    webench_design_1040516_353_117815132.pdf

    The attached PDF is the Webench schematic of the circiut, mine is the same without the Level Shifting on OVP.

    Parts are:

    L1 = Coilcraft SER1360-602

    D1 = Vishay V12P10

    Q1 = ON Semi NTTFS5C673

    Cout = Nichicon PCR1V221MCL1GS (220uF, 35VDC, 21mOhm)

  • 0 in reply to Brandon McLellan
    TI__Mastermind 25500 points

    Hello Brandon,

    I don't see any issues with the components or schematic as-is. How exactly is your OVP circuit connected if you are not using the level shift BJT? What are the values for Rov1 and Rov2 (what is your OVP set level)?

    Are you able to get any of the scope shots I mentioned? That would tell us if it's noise related (usually layout) or something else.

    Thanks,

    Clint

  • 0 in reply to Clinton Jensen
    Prodigy 40 points

    The OVP is connected like a typical Boost (between Gnd and the LED output) to protect the Recirculating Diode, output caps, and MOSFET.

    The value of Rov1 is 5.11K and Rov2 is 169K, for a target of approximately 42VDC.

    I have previously shunted the OVP pin to ground, even though I didn't set a voltage high enough to trigger it, and no fault indications either (looking at the SS pin).

    I didn't have the software loaded and setup for my scope (new),, so I didn't have time to get good readings, I should have them to you by Monday.

  • 0 in reply to Brandon McLellan
    Prodigy 40 points

    7120.Waveforms.zip

    The four waveform you requested are attached.

  • 0 in reply to Brandon McLellan
    TI__Mastermind 25500 points

    Hello Brandon,

    Those waveforms don't look to point to a bad layout problem, but it still could be related. Do you have a long gate drive trace that also maybe runs by a noisy high current trace? It's also possible that parasitics play a part, for example if the diode recovery is pretty noisy it could couple through the Cdg of the FET.

    If you can provide the layout I'd be happy to look it over for issues. But the first thing I would try is a different diode that isn't oversized, then next if that doesn't do it I would try a different FET. But if it's neither it is likely layout, I have never seen that switching with a good layout. However, it could still possibly be a right half plane zero issue. That is easy to check, just cut the bandwidth by using a large COMP capacitor to ground and/or increase the output capacitance.

    Regards,

    Clint