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.

TPA3118D2: poor manufacturing yield

Part Number: TPA3118D2

Hi, over the last 5 years or so, I have had manufactured for me about 1000 boards in different batches using TI part TPA3118D2DAP. Consistently, batch after batch of 105 pieces, about 5% have to be rejected due to a non functional TPA3118D2DAP part. The parts have power, are enabled and have an input signal, but provide no output signal. The last batch we made (in our ISO 9002 certified assembly plant) had a reject rate of about 20%! After paying extra money to have those chips reworked and replaced, the products worked correctly.

We have always purchased these parts from reputable North American suppliers such as Digi-Key and Mouser.com.

Why is such a high percentage of bad chips getting through TI's testing and the supply chain to its distributors?

On another note, I see that a newer part - TPA3118D2-Q1 is now available. What is the difference between the older part and the newer one?

Thanks.

  • David,

    Sorry to hear about your issue. Shawn will get back to you soon.

    Regards,

    -Adam
  • Hi David,
    I'm sorry for hear that. TPA3118D2 is very popular in the market and we never heard of this kind of issue on it before(the failure is pretty high). Is it possibly an application issue instead of the device issue? If we can see your SCH, we could help to review it.
    Basically the two devices are very similar. TPA3118D2-Q1 is qualified for automotive application, while TPA3118D2 is for consumer electronics application.
    Best regards,
    Shawn Zheng
  • Shawn, you are welcome to look at my schematic, inserted below, but it is text book from the datasheet in PBTL mode.

    Anyway it is clearly not a schematic issue because the other 95% of chips work fine right off the bat and, when replaced with good chips, the boards also work correctly.

    No, it is clearly a manufacturing yield or testing issue at TI or something happens to them at the distributors, i.e. Digi-Key or Mouser.

    The assembly process is fully automatic - they are not touched by hand. The parts are loaded directly in their tubes, as delivered, into the pick and place machines.

  • Hi David,

    Thanks for the explanation. Basically I can't find obvious application issue in the SCH. One thing I want to confirm is that 1uF input capacitors are used on the INN and INP but on the other page of the SCH, is my understanding correct? Because the device can't work normally without these capacitors.

    Did you check the FAULTZ pin when the issue occurs? If the pin is in low voltage, it means that a fault is triggered. Then the fault could be DC fault, pleaes try with lower value capacitors(e.g. 0.22uF, 0.1uF) for the input capacitors on input pins.

    If this doesn't work and you're sure that these devices are failed ones, please contact with the TI FAE. They will help you to submit a requirement for the failure analysis to the TI Quality/Reliability team. 

    Best regards,

    Shawn Zheng 

     

  • Yes, there are 1uF input capacitors, see other page of the schematic below.

    But all these questions fail to recognize that when the fault parts are replaced, the boards work OK, so it must have been a faulty part to begin with.

    I didn't check the FAULTZPin.

    Unfortunately I have not kept the faulty parts, but I will be sure to do so in the future so I can submit them to a TI FAE.

    Where is there a TI FAE in Ottawa, Ontario, Canada?

  • Hi David,

    Thanks for the info. As I mentioned in the above reply, you could try to use lower value caps for C51 and C52 if a FAULT is found on the failed device.

    If this doesn't fix the issue, you can contact Digi-Key or Mouser sales. I'm very sorry that I don't know the contact of the TI Ottawa FAE. I believe the distributor could help you to request the failture test.

    Best regards,

    Shawn Zheng