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.

LMZ21701: Converter failure

Part Number: LMZ21701
Other Parts Discussed in Thread: BQ24765

We have observed some failures on the LMZ21701 converters we are using in our product. So far, on 100 boards assembled, we have seen 3 regulators fail into service. On all of these failed devices, the output of the regulator show a low impedance (in the 1 to 2 Ohms ballpark), a very low voltage at the output and PG pin stays low.
When we removed the regulator, we could confirm that the low impedance (short) comes from the regulator itself.

The issue is very similar to what was described here: https://e2e.ti.com/support/power-management/f/196/p/528460/1922132

About the application: Our device is used in a railway environment. The board use 5 of these converters, arranged in a row. All of these are supplied from a regulated 12V or a Li+ Battery supply, so the input voltage can be anywhere from 6 to 12V. a BQ24765 handles the power path management and the charging of the battery.

The power these converters have to deliver is quite low, between 0 and 110mA at either 1, 1.2, 2.5 or 3.3V.

The schematics and layout is below.

Its there any obvious issue with our design? What should I do to track down the issue?

  • Hi Charles,

    Can you provide the date code of for the LMZ21701? I'd like to run this through the quality team to see what they think. From an applications point of view, the PCB layout and schematic looks okay.

    Regards,

    Jimmy  

  • Hi Jimmy,

    I've asked our assembler to track down the info they have.

    But maybe you can know it from the marking on the device itself?

    On on of the faulty parts is looks like this:

    767 HYO

    1701 7405 EA

    Thanks

  • Hi Charles,

    If you still have the untampered failing units, would you be able to send one or two of them in for FA? I'll run the top mark by quality to get some more insight on this.

    Linked is the TI customer return landing page.

    Regards,

    Jimmy 

  • Hello Jimmy,

    Not sure what you mean by 'untampered' ? Should I send a failing unit I removed from a board, or a fresh ones from the same reel?

    Regards
    CH
  • Hi CH,

    If you have any failing units that you haven't removed from the board that will be considered untampered. Otherise you would need to bake the part to remove moisture before sending it in. This way the system’s failure event can be preserved and eliminates the possibility of external issues introduced by the unmounting process.

    Regards,
    Jimmy
  • Hi Jimmy,

    Here is what I did:

    1. Baked 2 assembled boards at 100°C for 12h (baked them as you advised).
    2. Once cooled (~40°C) I've check once again the resistance between VOUT and GND and the short disappeared (R > 100kOhms)
    3. Applied power to the board to check if the issue disappeared, but the converters failed once again, in the exact same way as before: there's a short between VOUT and GND (one board is 2.5Ohms, the other one 4.0Ohms).
    4. Removed both parts, labeled them, and placed them in a sealed bag with desiccant so it stays dry (not sure it's useful?).

    I filled a RMA request for these (CPR181021609). I did not send the other 2 failed parts we already had, because they were removed without the boards being baked.

    I also mentioned this thread in the RMA request so they have everything.

  • Another board failed today. But this time both Vin and VOUT shorted to GND, in a more spectacular way.

    Should I add it to the RMA?

  • Hi Charles,

    I just want to clarify, when you "Removed both parts, labeled them, and placed them in a sealed bag with desiccant so it stays dry" how did you remove them? As I've advised, it is recommended to bake them before removal for FA processing.

    It wouldn't hurt to add the additional board failure in for RMA.

    Regards,
    Jimmy