• State TI Thinks Resolved
  • Locked Locked
  • Replies 22 replies
  • Answers 1 answer
  • Subscribers 29 subscribers
  • Views 562 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.

TPA2016D2: Irreversible change occurs

user3555915
Prodigy 230 points
Part Number: TPA2016D2
Other Parts Discussed in Thread: TPA2017D2, TPA2012D2, , TPA2012D2EVM, TPS61253,

I shorted P and N of differential output for speaker each other for 1 or 2 seconds.

As a result, speaker output sound remained alive, however consumption current of 5V power supply is much larger than usual (approx 1A) @ 8ohm load x 2.

It doesn't return to normal condition anymore.

I guess that this phenomenon is hard to be recognized as a breakdown of the device, because the function (sound) isn't lost.

Do you have any idea for the same or similar phenomenon?

P. S. 1) Actually same phenomenon has occured on TPA2017D2.

P. S. 2) Static discharge to differential output causes the same phenomenon.

Best regards,

Atsushi HANAZAWA (Sharp Corporation)

  • 0
    TI__Guru* 98965 points

    Hi Hanazawa-san,

    I will try to replicate your test conditions next monday.

    In the meantime I have one question:

    • the increased current consumption persists even after device reset?

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    TI__Guru* 98965 points

    Hanazawa-san,

    Do you have any comment on the question above?
    Additionally, are you using the EVM or a custom board for this test?

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    I apologize for my late reply. We took New Year's vacation last week.

    >the increased current consumption persists even after device reset?

    Yes, it does. The increased current consumption status isn't to be altered anymore.

    >Do you have any comment on the question above?
    >Additionally, are you using the EVM or a custom board for this test?

    We are using our custom board, and TPA2012D2RTJ (QFN package) and TPS61253AYFFR (5V power supply for TPA2012D2) are on the board.

    Best regards,

    Atsushi HANZAWA (SHARP corporation)

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hanazawa-san,

    I'll try to replicate your conditions on an EVM over the next week.
    However we haven't received similar questions so I don't think the EVM will show the same behavior as your custom board. Is there a chance you can test with an external source instead of TPS?

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    We've got TPA2016D2 output waveforms of both Normal and Abnormal conditions.

    Could you refer to the attached excel document?

    Best regards,

    Atsushi HANZAWA (SHARP Corporation)

    200114_TPA2016D2.xlsx

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hanazawa-san,

    I have tried with TPA2012D2EVM. Current consumption is lower after short at the output, similar to having the amplifier in shutdown mode.
    Device current consumption is normal after cycling SHUTDOWN pin to take the device back to active mode.

    As mentioned above:

    • Have you tried to test with an external source instead of TPS?

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    Is "source" which you have mentioned power supply?

    We have no chance to replace TPS61253 with the other DC-DC converters for TPA2016D power supply, because our mass-production will start soon.

    Therefore we haven't tried to replace the source and test.

    Could you teach me the reason why you recommend me to replace the source?

    TPA2016D2 on EVM and our PCB are different each other (the former is TPA2016D2YZH and the latter is TPA2016D2RTJ).

    Is there a possibility that the package difference causes the phenomenon I mentioned?

    Best regards,

    Atsushi HANZAWA (SHARP corporation)

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hanazawa-san,

    I ask to test with an external voltage source (power supply) just to isolate TPA and keep TPS out of the equation. If the device is working normally using external voltage source then the problem could be at the TPS or other part of the circuit.
    I don't think the package could make this kind of difference, but will check if there's a board I can use to test the QFN instead of DSBGA.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    I understand a little what you've mentioned. I guess that maximum load current from TPS61253 is too low to activate TPA2016D's current protector, but it is enough to destroy TPA2016D. We'll try to supply power for TPA2016D from DC stable power supply instead of TPS61253, but it is very hard for us to alternate our PCB to experiment.

    On the other hand, could you try to experiment with TPS61253 EVM?

    And I have one more question. Where did you short on your EVM? If your short path doesn't include LC filter, the short current will increase.

    Best regards,

    Atsushi HANZAWA (SHARP Corporation)

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hanazawa-san,

    I tried by shorting at J9-J6 and J11-J10, you mean I should short directly at the output pins of the device?

    I'll try with TPS EVM, getting the EVM in the lab may take some time.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    >you mean I should short directly at the output pins of the device?

    Only I wanted to do is to confirm that your short path was same as mine or not, and I confirmed that it was same.

    >I'll try with TPS EVM, getting the EVM in the lab may take some time.

    Yes, please do so.

    Best regards,

    Atsushi HANZAWA (SHARP Corporation)

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hanazawa-san,

    I should get the TPS board this week, will test and provide results once I get it.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    TI__Guru* 98965 points

    Hanazawa-san,

    I just tried along with both TPS and TPA EVMs.

    The current consumption is the same before and after trying the short at the output.
    Some questions I'd like to add:

    • What is your input voltage to TPS device?
    • Have you tried to replace TPS device after performing short test on your system? If current consumption is back to normal it would mean something has happened to TPS device

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    [Answer for 1st question]

    We supply TPS61253 3.3 to 4.2V with the battery. If the battery is full, it can supply 4.2V, and if the battery is almost empty, it supplies 3.3V or less.

    [Answer for 2nd question]

    We haven't replaced TPS61253 because TPS61253 has not broken.

    We have replaced only TPA2016D after the irreversible change has occured, and current consumption has been back to normal condition.

    It says that TPA2016D has broken and TPS61253 hasn't broken.

    Best regards,

    Atsushi HANZAWA (SHARP corporation)

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hanazawa-san,

    So it seems the damage is happening only on custom boards, as using EVMs does not show damages.

    Does your board have filter at the output of the TPA amplifier? If you're trying SC test with filter installed, could you remove capacitors are the output? Perhaps the capacitors at the output are causing high current spikes.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    >Does your board have filter at the output of the TPA amplifier?

    Yes. Our board has chip ferrite bead and 20pF capacitor on every output line. Unfortunately we have no more breakable board and we cannot try short test without the capacitors.

    I am terribly sorry for the frequent favors, but could you try short test with the capacitors on your EVM?

    I guess that the capacitors are extraneous to this issue, because TPA2016D2's datasheet recommends to use 1nF capacitors for countermeasure to EMI.

    Best regards,

    Atsushi HANZAWA (SHARP corporation)

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hanazawa-san,

    TPA2016D2 have some filters already, will check if can use the same component values just to confirm.

    Additionally, can you provide custom board schematic and layout for review? You can send private message or email if don't want to post it: ivan.salazar@ti.com

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    200310_BD.xlsx200310_SD.pdf

    Could you refer to the attachments?

    I can release only a part of our PCB diagram.

    Best regards,

    Atsushi HANZAWA (SHARP Corporation) 

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hanazawa-san,

    After review of the design files I have a the comment below:

    • Decoupling capacitors should be on the same layer as the IC, and as close as possible to the pins. I've noticed some of the supply connections are actually directly after via.

    Our EVM has different component values for the ferrite/capacitor filter than your design, but I'll test this anyway and provide results early next week.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    Prodigy 230 points

    Could you let me know the result if it has been already clarified?

    I guess it is hard for you to advance the experiment because of the virus hazard.

    Best regards,

    Atsushi HANZAWA (SHARP Corporation)

  • 0 in reply to user3555915
    TI__Guru* 98965 points

    Hi Hanazawa-san,

    I'll be testing the TPA2016D2EVM board on Monday afternoon or Tuesday at the latest. I'll give you an update as soon as I finish testing.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators

  • 0 in reply to Ivan Salazar
    TI__Guru* 98965 points

    Hi Hanazawa-san,

    I tested TPA2016D2EVM for short circuit. The conclusion is that current consumption before and after the test is the same:

    • Current consumption during idle operation (Powered up but nothing played) is 6mA
    • Current consumption while playing 1kHz tone is 116mA
    • Current consumption after short circuit test, and before re-enabling the amplifier is 2mA

    These results show that device can withstand a short at the outputs, even with filter connected, with no impact on the current consumption. Although unfortunately this is not conclusive on the root cause of the observations on the customer board.

    Both TPS and TPA devices seem to operate correctly after the short circuit test. Perhaps board design is a factor to consider, it would be good to inspect supply voltages during the test to look for any high peak transients that could point out where and what is being damaged.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators