• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 574 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.

DRV2667: Failures on boost converter start-up

Eric Hazen
Prodigy 70 points
Part Number: DRV2667

We have designed a 4-layer PC board with (5) DRV2667 devices on it.  We closely followed the recommended PCB layout from the datasheet, but overlooked one feature-- we did not ground or provide thermal vias on the thermal pad.  Pins 4-6 are grounded to a copper area.

In addition we inadvertently used a 50V rated capacitor on the BST output (which is programmed for 100V).

Several devices have failed when the boost converter is switched on even with no piezo connected.  After the failure the parts exhibit a low (10's of ohms) resistance from BST pin to GND and the part draws excessive current.

Is the GND connection on the pad mandatory for correct no-load operation of the part?  Or is it only required when driving a load to cool the device?  We do plan to revise the PCB layout but I would like to be confident this is causing the failure.  Obviously we will also replace the under-rated capacitor.

Any insight into the type of over-stresses which can cause a catastrophic failure of this part would be helpful.

Thanks!

  • 0
    TI__Genius 10512 points
    Hi Eric,

    The datasheet does mention that grounding the thermal pad is a "must," and I have never tested a board without the pad being grounded. So it is a little difficult to know what will happens when the pad is not grounded.. I believe that the bulk of the BST power is being discharged through the thermal pad through a large trace in the IC. When that pin is not grounded, the current finds another path through the other ground pins, but those traces cannot handle the current. A few things to try:
    1. Try grounding the pad using some very flat wire/metal to one of the GND pins on your PCB?
    2. I'm not sure what your R_EXT value is, but increase it in order to limit the current on the boost converter. Maybe also change you FB resistors to target a lower boost voltage and see if there are any issues at this boost voltage.

    FYI, you will want a 250V rated capacitor. Check the derated capacitance at 100V and make sure it fits the spec.
    Also we can review your schematic and layout if you would like to post it on E2E. If not, we could review via email.
  • 0 in reply to Kelly Griffin
    Prodigy 70 points

    Hi Kelly-

    Thanks for the quick and helpful response!  The explanation makes sense, and would definitely explain the failure mode we've been seeing.  We will try some of your suggestions.

    We have a small lot (3 pcs) of boards out for assembly, and we've asked the assembly house to apply a bit of copper tape to connect the pad to GND before soldering; we are hoping to test out our prototype system without having re-spin the boards.  We have ordered 250V caps and will also reduce the boost current and voltage before we power up these boards.

    If this doesn't result in a working design we'll take you up on the offer to review.

    Best,
    Eric

  • 0 in reply to Eric Hazen
    TI__Genius 10512 points
    Ok, sounds good. Feel free to mark your question as solved unless you have more questions. You can always start a new thread if needed.