• State Resolved
  • Locked Locked
  • Replies 15 replies
  • Answers 5 answers
  • Subscribers 31 subscribers
  • Views 2707 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPA3140D2PWPEVM: TPA3140D2PWPEVM

Fahad Ansari20
Intellectual 480 points
Part Number: TPA3140D2PWPEVM
Other Parts Discussed in Thread: TPA3140D2, LM4836

Hi,

I have TPA3140D2PWPEVM (Eval board for TPA3140D2) which I am using with a Bluetooth module. When I power up the Eval board, even if the Bluetooth module is turned off there is a continuous noise present.

This noise increase when I touch the J11 and J12 connectors.

Even after connecting Bluetooth with mobile, the noise remains. When playing audio, the audio level is much higher then the noise so noise is barely audible but when we pause audio, the noise is again heard.

I am using Bluetooth output and connecting it to male headers on eval board (L- and GND, R- and GND).

Load is 5W, 4ohm speaker, BTL mode.

LIMRATE is OFF, GAIN set is 32dB (I tried with different settings but could not see any change in audio levels), SS off 1 is selected.

Attached is waveform when there is no audio input (captured between L+ and GND)

  • 0
    TI__Guru 65760 points
    Fahad,

    If I read your scope correctly, you have a 640kHz square wave at 10V between the input pin and GND? There is also bad ringing on this square wave.

    Are you sure you are not injecting any signals into the board GND? Do you have anything connected to the L- or R-?

    Do you have a second EVM that is unmodified that you can measure without adding anything?

    Regards,

    -Adam
  • 0 in reply to Adam Sidelsky
    TI__Mastermind 24835 points
    Hi Adam,
    Thank you very much for the reply.
    Best regards,
    Shawn Zheng
  • 0
    TI__Mastermind 24835 points
    Hi Fahad,
    From the scope screen captured, you are using ferrite bead in the output filter, right? If it is, I think the output waveform looks expected. So you didn't change anything on the board, right?
    For the noise, could you please help to do one experiment? Disconnect the input module from the EVM board, just apply the power supply on the EVM board, and see if the noise is still there. If the noise disappears, it means that the noise is coupled from the front-end board. How long is the speaker wire, could you please try with short wires?
    The noise floor on the TPA3140D2 is every low, so the noise is around 65uV @ 20dB gain. So there shoudn't be obvious noise on the speaker.
    Best regards,
    Shawn Zheng
  • 0 in reply to Adam Sidelsky
    Intellectual 480 points

    If I read your scope correctly, you have a 640kHz square wave at 10V between the input pin and GND? There is also bad ringing on this square wave.

    Yes, you are right

    Are you sure you are not injecting any signals into the board GND? Do you have anything connected to the L- or R-?

    Yes, the input audio signals are connected to Bluetooth module, which is turned OFF. Even if we keep these pins open, this noise is heard. I have connected Bluetooth audio output to L- and R-.

    Do you have a second EVM that is unmodified that you can measure without adding anything?

    We do not have another EVM but this very EVM is not touched for any modifications. Just connected audio input and speakers for output

     

  • 0 in reply to Shawn Zheng
    Intellectual 480 points

    From the scope screen captured, you are using ferrite bead in the output filter, right?

    Yes, the EVM uses ferrite beads in the output filter.

    So you didn't change anything on the board, right?

    Yes, the EVM is untouched.

    Disconnect the input module from the EVM board, just apply the power supply on the EVM board, and see if the noise is still there.

    The noise is still there. Though it increases when we connect input to it. Even if input is not connected, if we touch the J11 and J12 (Lin and Rin) connectors, a very loud noise is heard.

    How long is the speaker wire, could you please try with short wires?

    Wire length is around 6,7 inches. I even tried with twisted wires and short wires for about 2 inches, there is no change in output

    Even if there is nothing connected at the input side, when I touch the input pins or input audio connector body, I can hear a very loud noise. I want to know, if this is noise picked at the input pins? when input is connected, is the wire picking some noise which is getting amplified?

    I have my layout ready for fab but seeing this issue on eval board, cannot proceed with PCB fab, please suggest what else can be tried to remove this noise

  • 0 in reply to Fahad Ansari20
    TI__Mastermind 24835 points
    Hi Fahad,
    Thanks for your reply. TPA3140D2 has been released to market as a inductor-free device for a long time, and we never heard of the noise issue on the output. I believe this is a setup issue or measurement issue. Could you please help to double-check the jumper setting on the board? L- and R- are connected to GND on jumper J1 and J7, right? Without input signal on the board, the left end of input capacitors C1,C5,C21 and C27 are actually connected to GND, right? Could you please help to take photo of the EVM under test and the whole test platform? Please also make sure the power supply on the EVM is good and the power supply cable is good and short.
    Best regards,
    Shawn Zheng
  • 0 in reply to Shawn Zheng
    Intellectual 480 points
    Hi Shawn Zheng,
    Below are the things I tried and current status.
    L- and R- are soldered to ground.
    There is no audio input connected to the board.
    Board is powered from a 5V,2A wall adaptor. I am still observing noise, though it is lower than the one I get when I power up with my 5V, 2A SMPS board.
    Here is the image for your reference and the audio noise I captured.
    Link for audio
  • 0 in reply to Fahad Ansari20
    TI__Mastermind 24835 points

    Hi Fahad,

    Thank you for the update. From the picture which you shown above, the INP input is flaoating, could you please short both P and N input to the GND. The LINP. LINN, RINP and RINN are shown in the red circles in the below picture.

    You can do this easily by short the right two headers on jumper J1 and J7. There shouldn't be any sound/noise without input signal.

    So when you use single-ended input mode, please connect the input signal to INP and short INN to GND for both right and left channels.

    Best regards,

    Shawn Zheng

  • 0 in reply to Shawn Zheng
    Intellectual 480 points

    We have shorted LINN and RINN to GND, by soldering them to ground, as shown in image below and used only INP pins for connection.

    With both inputs connected to GND we did not observe the any audio noise.

    When we are connecting audio input (to LINP and RINP), the noise is there even with no audio playing, what can be done? Right now we are using wires (teflon) for connection between Bluetooth module and EVM.

    I would also like to mention that we have been using similar setup with LM4836 but we never observed noise there. This LM4836 is not eval board but our on PCB.

  • 0 in reply to Fahad Ansari20
    TI__Mastermind 24835 points
    Hi Fahad,
    Thanks for your reply. By this experiment, we can see that the noise is not from the device itself or the EVM board. I think this noise should be from the input signal. Please make sure the GNDing between EVM and the front-end board is good. Did you check the input signal, is it clear enough? One thing you could do is to measure the output signal on the speaker by AP instrument, and perform FFT on it to see the noise frequency component. With this info, you could try to find where the noise is from. Another experiment that you can do is to use the input signal from AP to see if there is still noise on the speaker. The signal from AP should be clear, so there should be no noise in this case.

    Best regards,
    Shawn Zheng
  • 0 in reply to Shawn Zheng
    Intellectual 480 points
    Hi,
    From all the above excercises, one thing was clear to us that this noise is amplified version of input noise, which could be some pickup. We tried improving grounding between EVM board and our Bluetooth module but the noise was still there.
    When we observed the floating input on scope (without any connection with Bluetooth card), we found a frequency component of 1.73KHz present (in FFT mode). This was observed on both the channels. As soon as we connect the input pins to GND, this component vanishes and so does the noise. Remove the GND-LINP/RINP jumper, and this frequency is seen and noise heard.
    Then we connected a 0.1uF capacitor (Ceramic) between each of the input pin and GND. The frequency component was gone and so was the noise.
    Although this resolved our issue, I would like to know if this bypass capacitor at input is OK and if this should create any issue in audio. I heard a couple of songs and Audio sound good to me (there was no ill effect of bypass capacitor).
    Thanks a lot for all the guidance and support.
  • 0 in reply to Fahad Ansari20
    TI__Mastermind 24835 points
    Hi Fahad,
    Thank you for the update. I'm impressed by your strong ability in debugging and experiment. The 1.73kHz tone noise shouldn't be caused by the device, because there is no such frequency component from the AMP. Did you check the power supply, is it possibly from the power supply? Maybe you could try with the different power supply. The 0.1uF capacitor between input pin and GND can be used but not recommended actually, because this may cause the loss on the high frequency audio components. We need to find the source of the 1.73kHz noiseand the isolate it from the AMP.
    Best regards,
    Shawn Zheng
  • 0 in reply to Shawn Zheng
    Intellectual 480 points

    Hi Shawn Zheng,

    We tried powering up the system with a wall adapter, 5V - 2A. The noise is still there with Bluetooth connected and no audio input signal (without the capacitors at input). With capacitors at input, this also goes.

    I was now designing the PCB (assuming I can live with ill effects of this bypass capacitor on high frequency audio) and was referring to datasheet for layout but saw some contradiction in datasheet recommendations and EVM user guide. Please suggest which document should I refer to for layout designing. Also let me know if I can continue asking Layout related questions here or should I start another thread.

  • 0 in reply to Fahad Ansari20
    TI__Mastermind 24835 points

    Hi Fahad,

    Thank you for the update. Could you please let me know the controdiction between datasheet and EVM design? Sorry for the confusion. For the PCB design, please close this question and post a new question in the forum. Thank you very much!

    Best regards,

    Shawn Zheng  

  • 0 in reply to Shawn Zheng
    Intellectual 480 points

    Ok, Will start a separate thread and share details there.

    Thanks a lot Shawn, you were very helpful.