In order to try to use a high quality headphones-amp TI TPA6111A2, I'm connecting the input to 0V (i.e. no audio input) to first see if I get a clean very-low-noise silence sound in the case no audio is flowing.
According to the datasheet linked before, it gives something like this:
simulate this circuit – Schematic created using CircuitLab
I payed attention to "Typical Application Circuit" (page 1) and page 14, 15, 16 for the choices of the resistors / capacitors.
Still, I get an audio buzz. It is clearly too much, as the TPA6111 is said to be really better than I have according to its specs.
How to debug such a buzz, even when no audio is flowing?
Best Answer
A starting point to troubleshoot an apparent 'BUZZ' with a shorted input would be to power your audio widget separately with a low-noise supply (like a linear bench lab supply).
If you don't have a supply handy, you can excessively filter the incoming power rail. The no-load but active quiescent current of the TPA6111A2 is 3 mA. So you can easily add a series resistor of 100 ohms and afford drop 300 mV. Add additional bypass capacitors to lower the low-pass filter even further.
Layout is also critical for performance, tie the ground leg of Ci as close to the op-amp ground pin as possible.