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.

TPA3255: Restrict TPA3255 HF Response with PFFB

Part Number: TPA3255


We would link to restrict the HF end of the response curve of the TPA3255 so that it starts to roll off at about 10 KHz. We want the response to be a long way down before we hit the resonant frequency of the output filter at around 50KHz. We are not able to do this in the preceding gain stage. There is not sufficient information to calculate the PFFB components. Does anyone have experience of this?
  • Hello, we are looking into this and will get back to you later.
  • Hi Alex,

    would this document help?

    www.ti.com/lit/an/slaa788a/slaa788a.pdf

    Kai
  • Hi, Kai.

    Although the document is quite informative, it's real focus is reducing distortion due to the output filter. We have used and experimented with the suggested network which works well for the published purpose, but there is insufficient information to allow us to use it to substantially modify the frequency response. To do this we really need to know more, mainly about the phase response of the signal path within the TPA3255.

    But thanks for the response.

    Alex
  • Hi Alex,

    What values are you using for the LC filter on the output? Using a higher value inductor can cut down the corner frequency quite quickly. With a 33 uH inductor and a 4.7 uF capacitor, you could have a cutoff of around 13 kHz. At 50 kHz you would be about 24 dB down.

    Justin
  • Hi, Justin.

    Thanks for that.

    I will try it but I'm not sure that it would help. The TPA3255 is being used to drive constant current into a variety of inductive loads. To maintain the constant current, feedback is being provided by a Hall-effect current sensor (bandwith good to about 100KHz) whose output is then inverse-summed back into the balanced gain stage which feeds the TPA3255. Our issue seems to be coming from the phase shift caused by the output filter. With loads that have a low Land R the output filter becomes poorly damped and the phase shift near its resonant frequency results in positive feedback around the current control loop and consequential oscillation at a few tens of KHz. Lowering the cutoff frequency of the output loop will, I think, just shift the resonant frequency and therefore phase shift lower and I think it may just oscillate at a lower frequency.

    I hope that helps. Many thanks for the response.

    Alex
  • Hi, Justin.

    Just to correct my previous posting (it was a long day), The final sentence should read "Lowering the corner frequency of the output filter will, I think, just shift the resonant frequency and therefore phase shift lower and I think it may just oscillate at a lower frequency.".

    Thanks,

    Alex
  • Hi Alex,

    Yes you are correct, changing the filter for a lower cutoff would be moving the resonance lower. We do not have any guide on selecting the values for the PFFB circuits, but we have been working on  a TINA spice model of the part that will allow you to test circuits with the gain and phase of the part. This should be released soon.

    Justin

  • Thanks, Justin.

    It's probably a little unfair to ask, but can you give an expected release date for the Tina model, please?

    Thanks,

    Alex
  • Hi Alex,

    I hope to get them released before the end of May.

    Justin