• Locked Locked
  • Replies 5 replies
  • Subscribers 32 subscribers
  • Views 562 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.

TAS5602 Noise Problem

Christopher Ekbom
Prodigy 60 points

Other Parts Discussed in Thread: TAS5602

Hi,

We have designed a board utilizing three TAS5602 chips. Problem is, the amplifiers have a very high background noise.

 

Some info:

We use the stages in BTL mode

Decoupling is better than in the reference design

We don't have a real output filter, we only have ferrite beads and followed by a small capacitor to GND. The noise does not disappear if we mount a proper output filter.

The PWM lines feeding the amps are about 8 cm long, potential problem?

The amount of noise differs between the channels. But each of the three chips have one "good" and one "really bad" channel. If we connect the PWM signal from a good channel to a bad one, the bad channel is still noisy, so it's probably not the PWM that is bad.

The 24V line, the one powering the chips, seems OK when measured with oscilloscope, same for 3.3V.

A FPGA generates the PWM

The noise is mostly white

 

Anyone have any ideas or tips?

//Chris

  • TI__Mastermind 18510 points

    Chris, if you shut down all but one of the TAS5602, does the problem remain?

    Since swapping inputs does not seem to fix the problem, it may not be on the input side.  But I can't make further suggestions without seeing your schematic and PCB layout.

    Please send me those in PDF and I will try to help with this.  (I hope your schematic and PCB layout are close to the schematic and layout of the EVM.  If they are very different the problem could be there.)

    I will also need specs for decoupling components and output filter components (ferrite bead and cap).

    Regards,

    Steve.

  • in reply to SteveCrump
    Prodigy 60 points

    Thank you for your answer Steve.

    We have tried mounting only one stage of three, and it's still noisy. I have ordered the EVM, in order to have something to compare our design with, but seems like the delivery time is long.

    Do you have an email I can send the files to?


    BR

    //Chris

  • in reply to Christopher Ekbom
    TI__Mastermind 18510 points

    Chris, I received your files and have reviewed them, and I have a question:

    - Is the noise on C and D outputs?

    - What are the noise levels on your 24V and 3.3V buses?  Is it possible to provide scope shots and even FFTs?

    Regards,

    Steve.

  • in reply to SteveCrump
    TI__Mastermind 18510 points

    Chris, is there any news in this case?  Have you been able to identify the problem?

    Regards,

    Steve.

  • in reply to SteveCrump
    Prodigy 60 points

    Hi Steve!

    Sorry for the lack of updates, we have been working hard on the problem and made some interesting discoveries. But the problem is not solved.

    First, we ordered a batch of PCBs with alsmost exactly the same layout as the EVM. Result: Same problem, but even worse

    Today, we finally got the EVM. Result: Exactly the same problem, also bit worse than our design.

     

    We tried with a PWM generator from cirrus logic, and with that one, the channels had equal noise. So the problem seems to be at least partly, in our PWM generation. The strange thing is that the noise is much higher on the AB channels than in the CD channels, even if we connect the inputs of them to exactly the same PWM. The CD channels work OK with our design, but AB is way too noisy to be used in our products.

    We will now try to compare the PWM fom the "real" generator VS our, i suspect it has to do with the rising/falling edges of the PWM. But it seems like the channels have different tolerance of PWM quality on the TAS5602 chip?

    We need to borrow a more powerful oscilloscope to get the noise levels on the 24V and 3.3V lines measured more exactly, so it will take a few days. And the noise is on the AB outputs.

    Thank you for your help!

    Regards

    /Christopher