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.

TPA3129D2: Getting weird behavior from the chip

Part Number: TPA3129D2
Other Parts Discussed in Thread: OPA1604

Hello Team,

My customer is having a Problem measuring the output of the TPA3129D2. When they try to measure one of the Outputs with an oscilloscope, they get weird behaviour from the chip. Attached is the schematic and a simplified block diagram.

They are looking to have a problem with the filter-free output. The output with LC filter is very easy to measure with a differential connection from the oscilloscope

attachment.zip

Regards,

Renan

  • Hi Renan,

    Can you send us a scope shot of the issues they are having with the output of the amplifier? Can you share more details on what exactly is happening when running the TPA3129D2 filter free? 

    Thanks,

    Robert Clifton

  • Hello Robert,

    Below are the additional information from the customer.

    When he connect the built in Wavegen of the Oscilloscope everything works great and the speaker which he isconnected to the Amplifier with the Filterless Output reproduces everything just fine. But as soon as he connect a Probe to the Filterless Output the Amp outputs a terrribele noise and goes into shutdown. If he dont immediatly remove the Probe the chip gets destroyed.What is happening here?

    He read in one of TIs PDFs that when no Differential Signal is sent into the Inputs of an Amplifier a Groundloop can occur. He tried converting his single ended Signal into a Differential with the OPA1604. But he couldn't figure it out how to do this so that oscilloscope and Amplifier have the same Ground.
    The next thing he did was buying a Differential Amplifier IC which basically converts a Single Ended Signal into a Differential one.

    This has more or less solved the Problem but he is fighting with EMI there. He'd rather prefer a simple OPV circuit.

    He is looking forward to every critic, opinion and advice you have for them.

    Thanks!

    Art

  • Hi Art,

    Where is he placing the oscilloscope probe that is causing these issues? Perhaps a schematic of the circuit would help.

    Thanks,

    Robert Clifton

  • Hello Robert,


    Good day. 

    We apologize for this late response, my customer provided a late response to us with the following files and hoping you can provide feedback for us.

    TPA3129D2 Testboard.zip

    Thank you for your support.

    Regards,

    Renan

  • Hi Renan,

    Thanks for sharing with us! After reviewing the layout there doesn't seem to be anything that would suggest that it's the cause of this issue. 

    Does this problem only occur when there's a speaker connected to the output? 

    What settings is the oscilloscope? 

    Where on the board does the customer put the oscilloscope's ground test point? 

    Thanks,

    Robert Clifton 

  • Hello Robert,

    Please see the below the response from our customer :

    Does this problem only occur when there's a speaker connected to the output?

    >>The problem only occurs when there is a Probe connected to the output. It doesn't depend if there is a speaker on the output or not.

    What settings is the oscilloscope?

    >>AC Coupling, x10 Probe (10 Mega Ohm), HF Lowpass

    Where on the board does the customer put the oscilloscope's ground test point?

    >> I would connect 2 Probs to the Amplifiers output. and let the oscilloscope calculate the difference between the two probs. One probe is connected to OutA- and one to OutA+. The Grounds of the Probes are connected through a low pass on the output like in the picture i'll post as the next message.

    Regards,

    Renan

  • Hi Renan,

    I am Derek, responsible for TPA31xx devices.

    Does customer separate AMP ground from board system ground?

    What's the ground connection of Probe and board ground?

    We can discuss the issue offline. My e-mail is: derek-gong@ti.com

    You can send e-mail to me.

    Thanks!

    Regards,

    Derek