Buzzing noise with audio line input from cap sensors

Status
Not open for further replies.

Andy Belov

Well-known member
I have 3 Adafruit_MPR121 capacitive sensors and a touch screen with my Teensy 4.0.
When I touch or release the screen or the sensors, I hear a pretty loud buzzing sound from the Audio Board, but only when the line input is enabled.
There is no noise when I switch to the microphone.

I have LM358 op amp as a buffer for the guitar input connected to the Audio Board "line in". The preamp is not screened at all, bare bones.
The screen is Paul's ILI9341.
What can I do to reduce the noise? The

Thank you.
 
Sounds to me like a "ground-loop" caused by common currents in the power line connection. These can be real nasty to track down because they can be dependent on your wiring. I've had that problem from time to time, and now I use audio transformers in non critical cases to completely isolate the audio board input ground from the driving circuitry.

PJRC sells a audio ground loop isolator here for $6 (it's an in-line transformer) which may be something to try.
 
I see what you mean, Derek, but my guitar buffer (LM358) is powered by the same Teensy.
Everything is on the same board, powered by the same battery, connected to the same ground.
The guitar pickups are passive, plugged into the op amp.
 
Long shot, if you still have issue after checking for ground loops:

Try lowering the current to the electrodes to a minimum that still allows function.

Register 0x5C default is 0x20. Try 0x19, 0x18, etc. Steps are 1ua each.
 
Also... LM358 may not be your best choice powered by the teensy. Id try a TS922. Low power, rail to rail tolerant, very low noise.
 
Appreciate the advice. I'm new to guitar preamps and just followed SparkFun's proto pedal diagram.
I'll try that.

And I just found out that MPR121 (12 capacitive sensors in one chip) is no longer made and has been replaces with CAP1188.
Maybe the CAP1188 are not as "buzzy".
 
Just got the TS922 today: http://www.digikey.com/products/en?keywords=497-6035-1
Did my usual "wire mounted device" technique:
20191015_215951.jpg

It is much more quiet than the LM358 and I think it just sounds better, don't know.

But one thing for sure, it didn't change the fact that there is a pretty audible buzzing every time I touch the screen.
I'll try to "screen the screen" when I make the actual device. I'll try to shield it with a PCB.

If someone can suggest some kind of power line filtering for that, please do. I think the screen may be noisy just on the power.
 
Try the usual decoupling capacitor voodoo, large ground plane, diodes..EA5A5342-F446-4A71-ABE6-704033D91E4B.jpeg

Noise killing appears to me to be more alchemy than science.


This is from a cellphone project, the frequencies you're encountering may need different capacitors. If you have an oscilloscope you should be able to get the noise frequencies and use calculators online to choose capacitors.

I would expect the replacement device to cause as much noise seeing as how there's not much apparent difference in electrical characteristics from a glance at the datasheet. The upside is that what you do about noise for the MPR121 will likely help with the cap1188.
 
Thanks so much. This is what I was looking for. I'll try that, as I am in possession of a good analog oscilloscope.
But the screen and the sensors "sound" differently, so I may need 2 filters.
 
That circuit was copied from the sim800L datasheet ( cellphone chip) and is definitely an extreme measure... but it works.

I also use MG Chemicals' 841AR nickel conductive shield paint to good result for blocking rfi.
 
Status
Not open for further replies.
Back
Top