• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 2 answers
  • Subscribers 29 subscribers
  • Views 421 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TLV320AIC3120EVM-U: Unexpected Waveform in the speaker output

Ryotaro Fukui
Intellectual 2345 points
Part Number: TLV320AIC3120EVM-U
Other Parts Discussed in Thread: TLV320AIC3120

Hello team,

My customer has a question about TLV320AIC3120EVM audio test.

By using the tool in the link (www.ti.com/.../slaa233.pdf), my customer could play the audio file and change the volume by setting the following and changing the volume with Dvol.
・ Class-D Speaker Driver: 12dB
・ Analog Attenuation: -4dB

But when playing a 1kHz sound source file, instead of expected 1kHz sine wave, but a square wave like the graph in attached file appeared in the speaker output (A waveform like noise of about 400kHz is observed with a raised range)
Are there any possible factors?
The waveform is probed at the + terminal ⇔-terminal of the speaker.
The codec is planned to be used in conditions below, but are there any change in the register settings to solve the above problem?
(However, the sound source file is output correctly, so it would be helpful if you could give us some advice on possible factors other than settings of registers.)

・ The microphone is unused. Therefore, ADC is not used.
・ Does not use headphone output.
・ Sampling frequency: 16KHz or 8KHz
・ Quantization bit number: 16 bits fixed
・ Audio I / F: I2S interface (controlled by CPU)
・ BCLK input: Approximately 512 KHz (clock output from the CPU). Use the BCLK clock as the reference clock for the PLL
・ MCLK is not used

If there are additional information needed, please contact me.

I'm looking forward for hearing from you.

Best Regards,

Ryotaro Fukui

  • 0
    TI__Mastermind 22950 points

    Hello Fukui-san,

    Can you share an image of this noise? The SPKR outputs are PWM outputs and will not be sine waves. The amplifiers in for the speaker outputs are Class-D amplifiers. So I am wondering if what is being observed is expected. 

    Regards,

    Aaron Estrada

  • 0 in reply to Aaron Estrada
    TI__Intellectual 2345 points

    Hello Estrada-san,

    Sorry, forgot to attach the file. These are the waveform I received from my customer. Please feel free to ask if any other data are needed.

    Best Regards,

    Ryotaro Fukui

    TLV320AIC3120EVM-U_UnexpectedWaveform.pptx

  • 0 in reply to Ryotaro Fukui
    TI__Mastermind 22950 points

    Hello Fukui-san,

    Can you please share a schematic and register dump?

    Regards,

    Aaron Estrada

  • 0 in reply to Aaron Estrada
    TI__Intellectual 2345 points

    Hello Estrada-san,

    I received circuit and register setting information from the customer.

    Regarding the circuit, the evaluation board ACEV-1A EVM is used as it is.
    The waveform is probed from the speaker output (between SPKP / SPKM) of the J3 terminal block.

    Regarding the register settings, only the following are changed from the initial state, and the audio is played on the Windows media player.
    -Page 1 / Register 38 (0x26): Analog Volume to Class-D Output Driver
    (Analog Attenuation value setting)
    → 00000011
    -Page 1 / Register 42 (0x2A): Class-D Output Driver Driver
    (Class-D Speaker Driver value setting)
    → 00001000

    Is it necessary to set the DAC filter, etc.? (The audio is playing without any problems)
    It would be greatly appreciated if you could give us advice on the set values.

    Best Regards,

    Ryotaro Fukui

  • 0 in reply to Ryotaro Fukui
    TI__Mastermind 22950 points

    Hello Fukui-san,

    When I probed SPRP, I see the PWM signal shown below which is about 300kHz. This is expected as the outputs are Class-D amplifier outputs. 

    Can you probe one of the speaker outs? You should see something similar.

    The DAC filter is not necessary. It is if you would like to add any effects. 

    Regards,

    Aaron Estrada

  • 0 in reply to Aaron Estrada
    TI__Intellectual 2345 points

    Hello Estrada-san,

    I requested the customer if they could see the waveform you provided me. I will upload it once I received from them. I hope it will help you analyze the cause.

    Also I got additional question from the customer about the former question.

    • Is LPF required between TLV320AIC3120 and speakers? If necessary, what is the recommended capacitance for inductors and capacitors?

    Best Regards,

    Ryotaro Fukui

  • 0 in reply to Ryotaro Fukui
    TI__Mastermind 22950 points

    Hello Fukui-san,

    Thanks for the follow up. 

    To answer the next question, no a LPF is not required. If the customer wished to add one, please use the following app note for guidance on how to properly select the component values. 

    Regards,

    Aaron Estrada

  • 0 in reply to Aaron Estrada
    TI__Intellectual 2345 points

    Hello Estrada-san,

    Thank you for sharing your application notes.
    When I checked, it mentioned about EMI filters, but they also want to know about LC filters.

    It seems that class D amplifiers generally require an external LC filter. Is it possible to omit the external filter because the TLV320AIC3120 has a filterless modulation method configuration or some related functions?

    An LC filter was added and observed as a differential signal under the following test conditions.
    As a result, they were able to confirm conversion of a square wave ⇒ sine wave, but audibly, there was no difference with or without LPF.
    Is that possible?

    Test Details:
    * When LPF (22μH + 0.47μF) was added between the TLV320A IC3120 and the speaker and measured, it was confirmed that a 1kHz sine waveform was output when 1kHz.wav was played back (waveform attached).

    Oscilloscope waveform conditions
    1ch: SPK VDD current (evaluation board modified to 3.3V supply)
    2ch: speaker + pole (⇔GND)
    3ch: Speaker pole (⇔GND)
    4ch: 2ch-3ch (observed as a differential signal)

    Hope this will help in the cause analysis.

    Best Regards,

    Ryotaro Fukui

  • 0 in reply to Ryotaro Fukui
    TI__Mastermind 22950 points

    Hello Fukui-san,

    As stated before, the output of Class-D amplifiers are in a PWM format which is essentially a square wave with varying pulse widths. To convert this PWM signal to a sine wave, all that is needed is a Low-Pass filter. I believe this is what the customer added and it is shown above. 

    Adding a low pass filter at the output of the Class-D amplifiers is not necessary as the speaker itself will act as a low pass filter and convert this PWM signal to an analog signal. That is why there is no difference heard when the customer tested the audio with and without the filter. You can see in the typical application section (section 8.2) in the data sheet that a speaker can be directly connected to the speaker outputs. 

    I hope this helps. Please let me know if there are any more questions!

    Regards,

    Aaron Estrada