ESP8266 Community Forum

- Tue Jul 14, 2015 8:47 pm #23288

Any ideas to fix this without using two separate adapters to power Atmega32U4 and ESP8266 ?

Mic preamps can be very sensitive to interference.

You need to determine exactly where the interference from the ESP8266 is getting into the audio. One thing you didn't mention - did you try powering the mic preamp with a separate supply, and the Atmega and ESP with the same supply? I guess that the problem is either the ESP interfering with the mic signal at the preamp, or the momentary current draw of the ESP when transmitting is messing up the voltage reference for the A/D.

More stuff to try:
- a bigger better 5v supply
- more big capacitors on the 5v supply, right near the Atmega

If the problem is centered on the mic preamp:
- choke and bigger cap on the mic preamp voltage in
- shield the mic cable, shield the preamp

- Wed Jul 15, 2015 1:25 am #23302

Just wondering if you use the traditional capacitors on each regulator?

I tried many values of SMD ceramic caps and lowESR electrolytic caps. They help a little but not that noticeable.

Mic preamps can be very sensitive to interference.

You need to determine exactly where the interference from the ESP8266 is getting into the audio. One thing you didn't mention - did you try powering the mic preamp with a separate supply, and the Atmega and ESP with the same supply? I guess that the problem is either the ESP interfering with the mic signal at the preamp, or the momentary current draw of the ESP when transmitting is messing up the voltage reference for the A/D.

More stuff to try:
- a bigger better 5v supply
- more big capacitors on the 5v supply, right near the Atmega

If the problem is centered on the mic preamp:
- choke and bigger cap on the mic preamp voltage in
- shield the mic cable, shield the preamp

I only tried powering ESP8266 from separate power supply and it works flawlessly.
When powering everything together I can notice that it influences analog inputs even when nothing is connected to them. Values jumps a little but when I connect OP-AMP values jump 50x more.
Also tried powering earphones directly from OP-AMP to hear noise and I can easily hear that noise. So... conclusion is that EPS8266 affects everything in this circuit including ATMEGA chip.

Tried better power supply but with no differences.
As I mentioned I tried using lot of capacitors close to everything and it helps but not a lot. Also tried choke in series with OP-AMP, and nothing changes.
That is very strange, there is lot of decoupling but even that cannot help.

Only thing that noticeably helps is using additional linear voltage regulator to step down voltage from 5V to 3.3V and use it for powering OP-AMP.
I tried LM1117 so far but maybe low-noise, low voltage dropout regulators like LP2992 can be even better.
But this increases cost of this device. Hmmm, using these modules is always challenging.