0
\$\begingroup\$

I have a problem trying to split a drum trigger signal so that I can process it via an ATXMEGA64A3 microcontroller. The goal is to have the signal going from the trigger to both the existing sound module and also to my microcontroller for other processing.

The issue I am running into is that I need to isolate the orignal signal so that my microcontroller circuit doesn't affect the original signal going into the sound module.

How can I isolate my side of the signal process so that the original analog signal isn't affected by noise from my own circuit.

Attached is the schematic of the board I'm using to connect the trigger input, "thru" it out and a RJ45 connector that goes to the main microcontroller board. Please note that this RJ45 connector is receiving digital signal to control a bunch of ws2812b LED and this introduce noise in my analog signal (not sure how). On my side of the processing, I can live with a bit of noise, but this is the reason why I want to know how I can isolate the signal going to the sound module.

Schematic of the small trigger module

VCC is 5v provided by the mainboard through the network cable. GND is also provided by the main board through the network cable (using a ground plane on the main board PCB). WSLED input is the signal going to the ws2812b LEDS.

In case it might be of any help, here is the PCB design: enter image description here

When looking with the oscilloscope, between J1's pins, I see a bunch of noise probably due to my digital signals going on. I guess removing this noise is out of the scope of this question as it deals with mixed signals on the same board. I mainly want to protect J2 from receiving it.

Edit: Another important problem in this circuit is that if I measure the trigger output directly from a 1/4 cable, I get about +5v peak of signal. As soon as I connect my board, I get 500mv peak signal. Even with no LED, or RJ45 connected. Where is this signal loss occuring? I tried removing the capacitor in case it was filtering the signal, same loss of signal. I really don't get it. Can there be some inductance generated by the PCB which would kinda "absorb" the peaks of my AC signal?

\$\endgroup\$
0
\$\begingroup\$

A possible major source of noise is that you are connecting the (relatively) high-current LED power return and the analog signal ground together, so the voltage drop in the GND wire of the cable can introduce noise.

To reduce the effect, use two separate lines in the RJ45 connector/cable for the analog ground and the LED-power ground. At the microcontroller end, the LED-power ground should run directly to the power supply (or as close as you can manage), as opposed to sharing wiring with the microcontroller's ground (which is the reference for your ADC).

You will still get noise on the ground from the digital logic in the microcontroller, and crosstalk from the LED control signal; this is just a cheap improvement to your existing design.

\$\endgroup\$
  • \$\begingroup\$ Yes, I did that in the current test board I have right now. Didn't improve that much, maybe a little bit as you said. \$\endgroup\$ – Kévin Isabelle Oct 28 '17 at 1:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.