• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 3 answers
  • Subscribers 30 subscribers
  • Views 907 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.

TPA3123D2: Part overheats at power-On

Eric Conner
Prodigy 150 points
Part Number: TPA3123D2

Hello,

Our company has been using this amplifier on three product with total part volume of at least 100K.

Recently we have had some returns for amplifier failure.

After further inspection at our facility, we found that the amplifier will get HOT following power-ON cycle without any audio being passed through the part.

It seems to be in some unstable state with it consuming an extra 2-Watts of heat without audio.

When cooled down with freeze spray, it recovers and will operate normal until it is allowed to reach normal ambient temperature and then power is cycled again.

Any clue to the cause of this would be helpful. Attached is our audio portion of our schematic.

TI_TPA3123D2_in CD_Design.pdf

Thanks, Eric Conner

  • 0
    TI__Mastermind 24835 points

    Hi Eric,

    Could you please let me know how much is the failure rate of the issue boards? What if you replace the AMP device on the issue board, is the issue still there? Could you please try to do the following experiement? Disconnect the Codec and the Codec and AMP, then connect the left end of the AMP input caps(C99, C42) to GND. See if the HOT issue is still there. Try to set MUTE to High level or set SDZ to Low level, is the HOT issue still there? When the issue occurs, please measure the PWM waveform before and after the inductors in the LC fitler(L8, L9), see if it's normal. Please also check the board to see if there is any soldering issue or components failure.

    After reviewing your SCH, I found that there is no 470uF capacitors on both the AMP Left and right output(not on the GND connector). Then there is a PVCC/2 voltage on the speaker without input signal on AMP. Please refer to the figure 31 in the datasheet, which is a typical application.

    Best regards,

    Shawn Zheng         

  • 0 in reply to Shawn Zheng
    Prodigy 150 points
    Hi Shaun,
    So, far there have been only about 10-20 failures but we would like to understand what is going on.
    I will run the experiments that you suggest on next RMA units received.
    We placed the 470uF caps on the ground side so that the user could run the unit in BTL (mono) or SE (Stereo) mode at their option.
    I understand that this is different from your circuit but the capacitors still only allow AC to pass. I don't believe having 1/2PVCC should be a problem. We are seeing this overheating when there is no speaker attached also. Most customers are running the units in BTL mode per our operations manual.
    Recently, I have been able to repair the units by replacing C103/C115 and C266/C267 even though they don't seem to be damaged. And using capacitors with higher voltage rating (25 to 50V). Replacing the TPA3123 does not seem to be a catch all fix with varying results.
    I will update you on the requested experiments when able.
    Thanks, Eric
  • 0 in reply to Eric Conner
    TI__Mastermind 24835 points
    Hi Eric,
    Thanks for your update. If the device is used in BTL mode, the 470uF capacitors are not needed. But if it's used in single-ended mode, the capacitors on the output traces are necessary. From your experment, looks like there is no problem in the AMP devices. It could be components failure or soldering issue.
    Best regards,
    Shawn Zheng
  • 0 in reply to Shawn Zheng
    Prodigy 150 points

    Hi Shawn,

    I tried your suggested experiments you mentioned earlier on two new RMA failures.

    "Disconnect the Codec and the Codec and AMP, then connect the left end of the AMP input caps(C99, C42) to GND."

    Heating of the amplifier U6 still the same - HOT

    "Try to set MUTE to High level or set SDZ to Low level, is the HOT issue still there?"

    MUTE function has no change but the ShutDown pin turns off the amplifier and it cools down. As I enabled and disabled the SD, the heating comes and goes.

    "When the issue occurs, please measure the PWM waveform before and after the inductors in the LC fitler(L8, L9)"

    The PWM waveforms look the same to me as a good functioning board.

    "Please also check the board to see if there is any soldering issue or components failure."

    No noticeable poor soldering was seen on these failures.

    I am finding that 220nF Capacitors C25 and C36 seem to be failing while running in BTL mode. Then we are replacing them with 680nF. But in addition to replacing these two capacitor, it also seems that we need to replace C266/C267 and C103/C115. Changing these 6 capacitor seems to get the amplifier running normally.

    Regards, Eric

  • 0 in reply to Eric Conner
    TI__Mastermind 24835 points
    Hi Eric,
    Thank you for the update and I'm glad that you make it work normally. Is it possibly that the capacitors components are failed? In the SCH, 1uF capacitors are used for C103/C115, 0.22uF capacitors for C226/C267/C25/C36. I think it should be OK. Please also make sure the inductors L8/L9 are good.
    Best regards.
    Shawn Zheng