Rotary effect schematic(Korg CX3, schematic inside)

[Kevin Mitchell]

Hey man I have my doubts about the position of C7 & c23, shouldnt it be from the middle of the voltage divider to ground?

Yes of course. They were fine at the time when I said it. Since you've swapped the power rails rather than the resistor values now the polarity is wrong for those two capacitors.

[ferdinandstrat]

Hey man I have my doubts about the position of C7 & c23, shouldnt it be from the middle of the voltage divider to ground?

Yes of course. They were fine at the time when I said it. Since you've swapped the power rails rather than the resistor values now the polarity is wrong for those two capacitors.

I see, so I need to connect the cap from the point where the two resistors connect to ground yes?

[Kevin Mitchell]

Hey man I have my doubts about the position of C7 & c23, shouldnt it be from the middle of the voltage divider to ground?

Yes of course. They were fine at the time when I said it. Since you've swapped the power rails rather than the resistor values now the polarity is wrong for those two capacitors.

I see, so I need to connect the cap from the point where the two resistors connect to ground yes?

Yes. They are decoupling capacitors. The unit should work without them but it's good practice.

[ferdinandstrat]

Thank you kindly, I figured as much, it's basically a filter cap for a voltage divider

[ferdinandstrat]

Well I finished the PCB and it passes sound but I couldnt figure out if there's any chorusing going on since there's a very loud oscillation going on that drowns out the effect almost entirely, it's volume changes if I play with the AM gain trimpot

When I say oscillation it sounds like a damn car alarm "wee ooo wee ooo", oddly enough it doesnt seem to be affected by the freq trimmers

[Kevin Mitchell]

Crosstalk maybe? If your signal path is too close to high frequency paths it could cross over, effectively treating the pcb material as a capacitor.

Post your layout so we can pick at it. I'll take another look at your schematic as well.

[ferdinandstrat]

Crosstalk maybe? If your signal path is too close to high frequency paths it could cross over, effectively treating the pcb material as a capacitor.

Post your layout so we can pick at it. I'll take another look at your schematic as well.

[Kevin Mitchell]

Crosstalk maybe? If your signal path is too close to high frequency paths it could cross over, effectively treating the pcb material as a capacitor.

Post your layout so we can pick at it. I'll take another look at your schematic as well.

Need better resolution and more info. But first, have you traced the signal to find the point where oscillation gets into the signal?
Did you use a ground plane?
Did you manually route or use an autoroute feature?
Have you applied or been mindful of any design rules?
Any photos of your PCB?

[ferdinandstrat]

I havent traced it yet

I did use a ground plane, on both sides

I did the routing myself

Any specific design rules?

I will try to provide photos

[Kevin Mitchell]

Any specific design rules?

I had asked about design rules because if you're not mindful of the results of autorouting or the auto generated ground plane it could result in paths being too close together increasing the chances of crosstalk. If the ground is picking up any HF crosstalk it'll be nightmare to clean up without a redesign of the layout since you'd have to find where it's happening and gouge out the copper until the oscillation is no longer leaking over or just cut the trace and sub in a shielded wire.

I have seen this quite a few times during the development of BBD and synth designs where there are multiple HF tracks you must watch for. Usually these tracks are isolated safely between &/or underneath a ground route - again not too close.

Tom (ElectricDruid) had chimed in earlier. I recon he would have good advice to share about this as he's very familiar with such work. I myself haven't designed enough of these and routed the traces with every effort to avoid such problems. I'd call myself more fortunate than wise in that respect.

Audio probing should help pinpoint where it begins.

[ferdinandstrat]

You know, I dont think the issue is the pcb design for once simple reason: I already made a bbd based project, a chorus, and I designed the pcb all willy nilly and auto routed it, it has no issues in terms of noise

[Kevin Mitchell]

Okay. Trace it and report back.

[ferdinandstrat]

Okay. Trace it and report back.

I couldnt trace it lol, I started at the input and it is even present there

[Kevin Mitchell]

Okay sounds like oscillation getting into the power supply. Here's what I would do to try to pinpoint what's going on.

Let's get rid of the BBDs and their clock & LFO ICs. So remove U1, U2, U3, U4, & U6
Check the signal at the input. If the noise is gone take advantage of the moment and make sure the signal is finding it's way to the output with no noise as well.

If that cleans up the noise issue you'll know it's not the power supply directly but an error or poor routing related to the removed ICs. So now you should install the LFO chip (U6), check for the noise on the signal path again.
Install U1 & U3, check again.
Then finally install the BBDs.

It may be best to install each of the mentioned ICs one at a time and check each time. That way you can find what chip starts the trouble and also what chips make it worse - in case the cause of the problem is in multiple places.

[Steben]

Off topic while going on topic.
Wouldn't a very small fixed delay for the "dry" signal make an improvement?