• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 1453 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.

TAS5630 output noise

Alex Koop
Prodigy 40 points
Other Parts Discussed in Thread: TAS5630

Hi! :)

I have built the amp using TAS5630 chip. I have found in my prototype that the idle output noise level depends on the supply voltage (PVDD). The higher is the PVDD, the lower is the output noise. Is it typical for this chip? I am using PVDD=35V now, and the output noise is a little bit too high for me. I can hear it from the loudspeaker in the distance about 20-30 centimeters. I have developed my own PCB layout design, very similar to the reference design from TI, using a solid ground plane too. The noise seems to be not in the result of EMI, because I can touch any output switching node before LC filters, and the noise level stays the same. I have tried to insert some additional ceramic caps in various places with no any change also.

I know the clock generator in class-D amp can be the source of the noise too, so does it makes a sence to use some high quality low noise crystall oscillator together with TAS5630? I can run it in slave mode and connect the powered XTAL generator to the OSC+/- pins. Will this help? Or the source of the noise is something other?

Many thanks and best regards,

Alex

  • 0
    Prodigy 50 points

    Hello,

       The TAS5630 has a 100db SNR so you should expect a 3mw noise at the output which is audible , there is nothing you can do to remove this noise.

       If the noise is reducing when the supply voltage goes up, it may indicate that the noise is at the input stage.

     

    Mr Tester.

      

  • 0 in reply to Mr tester
    Prodigy 40 points

    Thanks for the answer!

    >>If the noise is reducing when the supply voltage goes up, it may indicate that the noise is at the input stage.

    I have nothing connected to the inputs yet, except 100pf capacitors to the AGND (ground plane), as shown in the reference schematic. I have also tried to connect IN+ and IN- together, but the noise stays unchanged.

    >>There is nothing you can do to remove this noise.

    Have you tried to do it already?

  • 0 in reply to Alex Koop
    Prodigy 50 points

    I never use that chip before but we can know just by reading the data sheet -> you can expect 3mw of audio noise, If you got input connected to ground

    this is the best it will ever be.

    Is the noise you are refering to is a "hiss" or a "hum", hiss isexpected, hum would mead that you power supply has ripple on it.

    Mr Tester.

     

  • 0 in reply to Mr tester
    Prodigy 40 points

    I have a "hiss" noise :(

  • 0 in reply to Alex Koop
    TI__Guru**** 189401 points

    Hi, Alex,

    We measured a slight correlation to lowering noise floor as PVDD is increased:

    This is A-weighted. In the d/s, we reference 270 uVrms of noise with a slightly different output filter than shown above (the d/s measurments used a 7 uH inductor).

    I listened to the noise, but I couldn't hear it at 20 cm, and the noise itself was white noise.

    Did you have the inputs terminated? Do you have any sort of measurement equipment that you can use to measure it to see if you are reasonably close to what we show in the d/s?

    -d2

    PS the plot above is for informational purposes only. TI does not guarantee noise performance of this amplifier, and your mileage may vary based on a wide range of factors.

  • 0 in reply to Don Dapkus
    Prodigy 40 points

    Hello Don!

    Many thanks for your answer and measurements!

    Yes, I have nothing connected to the inputs yet, except only ceramic 100pf caps to AGND. Input traces are only 4 millimeters of length. Unfortunately I do not have such precise equipment to measure noise in hundreds of microvolts :( I have only 8 bits 40 MS/s digital storage oscilloscope, usual multimeter and my ears :)

    Also, I am using TAS5630 in AD modulation mode (M1, M2, M3 connected to ground), but this shoult not affect S/N ratio too hard, though...

    Well, I will try to find the problems in my prototype then. Maybe I need to tie the PCB layout precisely to TI's reference layout.

     

    Best Regards,

    Alex