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.

TAS5822M: How to reduce the radiation?

Part Number: TAS5822M

Dears,

   My product use the TAS5822M for class-D amplifier, the radiation test exceed standard.

    radiation was traced to cables and speakers, tried to replace the cable and speakers, reduced power did not significantly improve.

   Do you have some advice to reduce it? the test report as attachment.

TEST REPORT.pdf

  • Hi 

    according to your test result, the peak frequency is so much high, it is difficult to believe they are caused by amplifier.

    can you set tas5822 to hiz, retest the radiation again. if the peak still exist, then it should not caused by amplifier.

    can you describe more about how you trace the radiation caused by the cable and speakers?

    thanks.

    jesse

  • Hi Jesse,

        We try to set Shutdown mode, the radiation has gone.

        Remove the cable from PCB board, the radiation disappears;

        Only remove the speaker from cable, there's still radiation. so we think the radiation is coming from the cable.

  • Hi

       May I ask how long is the cable? Except for the output, is there any other cable in your system? We could firstly try to remove all other cables to isolate the problem.

       For our device, you could try put 10nF MLCC cap at each of the output trace, place the cap very close to the PCB output port, will help reduce high frequency radiation. For your PCB, try to put all the output trace in the middle layer, will also help.

  • Hi Shadow,

        The cable length is about 50cm, and there are no other cables.

        We tried adding capacitance to the output port, but it didn't work.

        the sch and pcb as attchment, could you help to review, thanks!

       

      5822-pcb.pdf

  • Hi

    you have 2 pcs 5822 on your board. am i understanding correct?

    device has 2 pvdd, the MLCC cap for one pvdd is on the top, the other one, they are all on the bottom.

    this might be not very good for radiation.

    it would be better to put 0.1uF, 10uFx2 on the top, and put 390uF on the bottom.

    and put 0.1uF and 10uFx2 as close as to the output pin.

    thanks.

    jesse

  • Hi Jesse,

      The customer thinks it has nothing to do with the power supply, they have another product that uses only one TAS5822, the power supply is a different PCB layout, and the radiation is the same frequency point.

  • Hi 

    can you help perform below action to confirm whether the radiation gone or not?

    set 2 devices to hiz mode, to see whether the radiation gone or not?

    by this way, we can know that whether it is because the pwm at the output lead to cable generating radiation.

    we can also set one device to hiz to see whether it is just caused by 1 device or both.

    if yes, maybe adjusting pwm frequency lower might help.

    50cm cable is such long, can you take picture how the cable been placed when performing the test.

    thanks.

    jesse

  • Hi Jesse,

    All of the above methods have been tested and confirmed to lead from the output port;

    Our current PWM frequency is 756K, we will try to set it to 385K

  • Hi 

    ok, have you measured the output pwm waveform?

    does it have serious ripple on it?

    thanks.

    jesse

  • Hi Jesse,

        

  • Hi 

    how about the result changing to lower Fsw?

    you already using 22uH inductor and it still seems having very large ripple. it is rare.

    can you take a try using other inductor with larger Isat? or some better inductor?

    thanks.

    jesse

  • Hi Jesse,

        The customer adjusted the inductance from 10uH to 22uH, which was effective, but did not solve the problem. If the inductance was increased again, the heating would be serious.

    Is there a corresponding formula that can guide customers to adjust inductance and capacitance to match PWM frequency to reduce radiation?If blind to adjust, the efficiency is too low, the customer can not accept the high cost of testing institutions

  • Hi 

    8.2.2.2.2 Inductor Selections in the datasheet told us how to select inductor. it is just a reference.

    actually, we still need to adjust the value according to some actual test for final board.

    in my understanding, usually 22uH should be enough.

    according to customer's sch, the pvdd is 24v, what is the load resistance?

    usually, we choose the inductor with Isat > pvdd/Rload.

    also, customer doesn' t need go to perform the EMI test every time after changing the inductor.

    they can perform the measurement for the PWM to confirm whether the ripple been improved or not, it no, then there is no need go to perform EMI test.

    thanks.

    jesse