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.

TLV320AIC3254: Same design, some boards work and some don't

Part Number: TLV320AIC3254
Other Parts Discussed in Thread: TLV320AIC3253


I have an ESP32 paired with the TLV320AIC3253 using I2C+I2S, with the ESP32 in Master mode, driving the BCLK.The design is inspired by the Audioboost board, with the codec acquiring 2 channel audio data from IN1L and IN1R channels, using its ADCs at 16bits, 48kHz. The codec is configured over I2C and provides the PCM encoded audio data over I2S to the ESP32. The codec also has a passthrough, where the audio signal is routed from the inputs to the headphone outputs, via the PGA. Basically, IN1L->L Mic PGA ->Mixer Amp->MAL->HPL and IN1R->R Mic PGA ->Mixer Amp->MAR->HPR in the diagram below.

Overall setup:

I got 10 PCBs assembled, out of which 6 work perfectly, i.e I get audio in the headphones and via I2S in the ESP32 which records to stereo WAV files. 4 don't work as expected. For the 4 non-working boards:

  1. I am receiving noise on I2S with no discernible audio content
  2. The audio passthrough works - I get clean audio in the headphones, which means that the analog audio path via the PGA is good and the I2C config is also good
  3. I disabled the audio passthrough and made a "loopback" via I2S - receive ADC data on I2S in the ESP32 and write it back via I2S to the codec DAC, which would then route it to the headphones.
    1. I got clean audio on the headphones on the 6 working boards - this verified that the loopback was correctly set up.
    2. I got noise in the headphones in the non-working boards
  4. Could it be a bad I2S transceiver on the ESP32 or the codec? Is there a way to route ADC to DAC directly in the codec?

What can I try next to find the root cause? I am stumped. Can it be bad codec chips? If that's the case, how can everything else except the I2S part be OK?



  • Hi Prince,

    Today is a TI Holiday and our experts are out of office. Please give us another day to look at and answer your question. For record keeping purposes please do not reply to this message unless 48 hours have passed without a response.

    Thank you for your patience,


  • Hi,

    Few debug tips:

    • For non-working boards, can you check from input to I2S out if it's recording correctly, basically checking each path (ADC and DAC) separately?
    • There's a register to do a loopback as shown below:


  • Thanks. I tried the ADC-DAC loopback. Works on the working boards - to validate that the registers were set correctly. Noise on the non-working boards. Does this mean that the clocks are wrong? How would that explain the working boards?

    Clock registers:


  • For the non-working, are you seeing the correct signal on ADC alone(record path) and if you play a tone on the DAC path alone are you seeing noise on the HP output? I guess is the issue in the input or output? What if you swap the unit between the good and bad boards? Looking at the clock setting above it looks fine for 48KHz sampling.