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.

TAS5825M: Audio line and signal problems

Part Number: TAS5825M

Tool/software:

Hi TI,

(1) When using the TAS5825 EVM with PVDD at 20V, BD mode, and 384kHz switching frequency, we observed that the speaker(2w/4Ω) output voltage after the LC filter is a sine wave centered around PVDD/2

1.Is the LC filter a low-pass filter? What is its cut-off frequency?

2.When subtracting the two PWM signals from OUT1+ and OUT1- of the Class-D amplifier, does the resulting waveform reconstruct the original audio signal?

 

(2)In our circuit with PVDD at 20V, BD mode, and 384kHz switching frequency, the speaker(2w/4Ω) output only has a ferrite bead (220 ohms @ 100kHz) and a small capacitor of 220 pF. WSPK1 is the speaker connector.

1.If the LC filter is a low-pass filter, but our circuit does not include such an LC filter at all, is this approach still feasible?

2.At OUT_A+ and OUT_A-, we measured the following signals. After subtracting the two waveforms (simulating differential probe measurement), the resulting signal still has a very large amplitude.

3.We’re very concerned that such a large voltage across the speaker(2w/4Ω) might damage it.

4.Although the SmartAmp protects the speaker, is such a high voltage really necessary? It seems to cause the temperature to rise more quickly.

5.In the past, we could only tune audio on your TAS5825 EVM, which was inconvenient. Is it possible to perform real-time tuning directly on the TAS5825M in our device? Please provide support.

6.Since we found that the TAS5825 EVM has an LC filter, which is different from our circuit, and we could only tune audio on the EVM in the past, what problems might occur if we directly apply the tuning settings to our device?

 

  • Hello,

    1: 

    A. Is the LC filter a low-pass filter? What is its cut-off frequency?

    Yes this is a low pass filter. The cutoff is 61kHz. This tool may be helpful: https://www.ti.com/tool/LCFILTER-CALC-TOOL 

    B. When subtracting the two PWM signals from OUT1+ and OUT1- of the Class-D amplifier, does the resulting waveform reconstruct the original audio signal?

    Yes it should reconstruct the input signal with some gain applied.

    2: 

    A. If the LC filter is a low-pass filter, but our circuit does not include such an LC filter at all, is this approach still feasible?

    If you do not want to use an LC filter it is recommended to at least use a ferrite bead. 

    B. At OUT_A+ and OUT_A-, we measured the following signals. After subtracting the two waveforms (simulating differential probe measurement), the resulting signal still has a very large amplitude.

    You are measuring the class-d switching frequency. To see the resulting audio signal you need to filter the audio signal as a speaker acts as a natural low pass filter. 

    C. We’re very concerned that such a large voltage across the speaker(2w/4Ω) might damage it.

    TAS5825M is quite over spec for a 2W 4 ohm application. Are you open to other device recommendations? If you want DSP TAS2653 may be a good option. 

    D. Although the SmartAmp protects the speaker, is such a high voltage really necessary? It seems to cause the temperature to rise more quickly.

    For a 2W application this high voltage is not necessary. You could  lower PVDD to a smaller value. 

    E. In the past, we could only tune audio on your TAS5825 EVM, which was inconvenient. Is it possible to perform real-time tuning directly on the TAS5825M in our device? Please provide support.

    Yes you can route the I2C pins from the EVM to your custom board. Then you can still use PPC3. 

    F. Since we found that the TAS5825 EVM has an LC filter, which is different from our circuit, and we could only tune audio on the EVM in the past, what problems might occur if we directly apply the tuning settings to our device?

    The LC filter is mainly for EMI considerations. There should not be a large difference. You could also remove the LC filter on the EVM. 

    Regards,
    Sydney Northcutt 

  • Hi Sydney,

    Based on the test shown in the image, is there any test report or documentation related to using the TAS5825 without an LC filter?
    Please help provide relevant reference materials.

    Regards,

    Wei Hsu

  • Hi Wei Hsu,

    All of our data is gathered with some sort of output filter. Apologies for the inconvenience. 

    Regards,
    Sydney Northcutt