Is this oscillation?

[living sounds]

I use a CAPI 2-ACA in my console master:

http://www.capi-gear.com/catalog/images/gallery/PCBs/2-ACA-Bo/CAPI-2-ACA-Bo-Rev-B.1-schematic.pdf

It's been adapated to fit my console gain structure, which sums through 22k resistors, so R13 and R10 are 13k now.

The console is very quit, but I've been doing tests with square waves. These look really solid through the four original group outputs, but on the master the square wave "waves" a little like in the video:

https://dl.dropboxusercontent.com/u/15815132/VID_20170226_133811.mp4

I've tried higher feeback cap values in real time both in the ACA and in the booster section, which had the predictable effect (visibly with way too high values like 1nf) of shaving away from the start of the square wave (low pass filter). But it did not stop the movement along the square wave maxima.

Changing DOAs (also tried hybrids) didn't have an effect either.

Is this even a problem? Thanks!

 

[RuudNL]

Looks like (mains frequency?) interference to me.

 

[living sounds]

Thanks. Mains (50 H)z noise is really low. It should not be visible this way. The pattern overlaying the square wave moves from left to right and back at a lot less than 50 Hz, too.

I made a comparison of a bus output and the master output using sinewaves (attached -bus 1 on the left, master out left on the right ). The master output has higher distortion (to be expected due to the two CAPI iron core transformers). But there also appears to be a bump of the noise floor around the harmonics. Any idea what might cause this and how it can be mitigated? Noise floor without signal is practically the same.

All the channels and buses have their own decoupling grounds that go straight to a star ground point. The 2-ACA has no explicit "dirt ground", though. Could this have something to do with it? The two summing inputs each have floating ground connected to ground by a 100k resistor paralleld with a 10n cap on the PCB. The PSU ground connection goes to the same star ground as the other decoupling ground lines. I've tried to connect it to audio ground, maybe caused a little higher 50 Hz hum but otherwise no difference...

 

[JohnRoberts]

It looks like a higher frequency signal (or noise) superimposed on top of the square wave. If you stop the square wave input, is the output dead silent?

JR

 

[living sounds]

Thanks guys, I was finally able to solve it. It was the Dimension D (BBD based chorus). The aux path was leaking a little, so a small dose of modulated signal got mixed into the master bus.  :