As the title says, I'm having some issues with battery charger noise and an audio amplifier.
The PCB with the charger and amplifier is connected to a GSM device, and that's where the audio signal for the amplfiier comes from. The amplifier itself is a class D TPA2000D1. A USB connection charges the battery via the AAT3681A-1 charger IC. The battery voltage is used to power the rest of the board (i.e. not the 5V directly). The amplifier gets its power through a low noise, high PSSR LDO, the ADP150.
Now, the issue I have is that when the USB and the battery are used during a call, clirps / bursts of noise about a second apart can be heard on the speaker. If I remove the USB power and use only the battery as the power source, the amplifier is silent! This means, for me, that the noise is somehow caused by the USB power and presumably because of the switching nature of the charger IC.
So I tried to measure this: although I see ripples on the V_BATT line when the USB is connected, the TPA_VDD line is silent and looks exactly the same with and without the USB connection. Although I measure there about 30mV peak-to-peak noise, the oscilloscope alone shows about 25-28mV peak-to-peak noise when nothing is connected.
I also tried other 5V sources (external power supply) with the same results.
My questions are: what am I doing wrong? What can I do to reduce this noise (e.g. filters, other charger IC)? How can I measure this noise anyway? I hear it but I can't see it. Any ideas are highly appreciated!
EDIT Here is the layout. There is a ground plane over the board, which is not visible here, so all components share the same ground.