Makes Idea of differential input sense?

Status
Not open for further replies.

cebersp

Well-known member
Hi,
using the 16 bit audio shield and having some noise trouble. For example a serial line to a serial lcd can be heard.

Now I wonder if it would make sense for example to couple the second line input chanel with a resistor to ground and subtract (via mixer and negative gain) its input from the real input. What would be the best way to do this?

Suggestions welcome!
Thanks, Christof
 
Hm,Paul,at least I don't understand, why this should help here. I thought, that this would help against 50 or 100 Hz hum.
Best regards, Christof
 
Hm,Paul,at least I don't understand, why this should help here. I thought, that this would help against 50 or 100 Hz hum.
Best regards, Christof

Ground loops create a big antenna to pick up any stray interference that find their way in. While power line hum is a common symptom, anything to do with a computer is an EMI super-nova. Cell phones love to cause problems too when near a ground loop.

I agree with Paul, isolate ground loops first. The isolator above should work, but you can also rule out ground loops by:
1) program your teensy with USB, but then use a cell-phone charger to power the teensy (no connection to computer)
2) make sure the audio input is passive (microphone), or battery powered. I connect my mp3 player out to the line in, but you can also ground all the inputs to check the boards self-noise.
3) use actual headphones out, rather than plugging into some other device that is connected to the wall.

This way, there is no ground loop between the input, output and teensy power.
 
Could also be as simple as the LCD current taking an unexpected path through the analog ground your signal uses. Not all ground loop issues are AC mains hum.
 
Thanks for the answers!
Now I had some time to simply try the differential input and it does not help.
I had thought, that there was perhaps a path for the noise over the power lines to the adc. This is not the problem.
Printing (1000.0*rms1) of the input or after the differential input gives 0.18, when the inputs are just open.

Printing (1000.0*rms1) of the input or after the differential input gives 2..3 with a humbucker pickup, the guitar compressor (Orange Squeezer clone) active and 0.34 without the compressor active and without a guitar cable into the compressor. The compressor has a maximum gain of 20, so I assume that the main source of noise is before the compressor.
With Fender Telecaster Noisless Pickups N3 I get about 1..2.

What helped more against the lcd serial was to lower the input impedance by having a 3.3k resistor in parallel to the input. There is a voltage divider anyway before the line input.
It it is very helpful to have a possibility to measure and print the noise, because if you only listen, you tend to believe things....
Regards Christof
 
Status
Not open for further replies.
Back
Top