BE-OD Clone Oscillating - help!

[boga]

Hi!

I made a BE-OD clone according to this schematic:
https://1.bp.blogspot.com/-FFC5U2CNcKM/WcVUpTksN0I/AAAAAAAAAsY/U6cdl102b24YAOVcN5ppeBseXtux4QKxgCLcBGAs/s1600/Thermionic%2BDistortion%2BSchematic%2BHighlighted.png

They are also selling a PCB which is called Thermionic Overdrive (BE-OD Clone), the schematic is taken from the build document:
https://www.pedalpcb.com/docs/ThermionicDistortion.pdf

So I guess the schematic should be OK and working, otherwise they could not sell the PCBs.

In the schematic I did a small change. As I did not find 1MOhm pots I change the Gain pot to 100k and re-calculated and simulated the other components. So R9 and R10 are 2.2k and C8 is 1uF. According to the simulations should work fine (but who knows. Maybe some of you can check).

Therefore I believe that the problem should be in my layout design. I made it with SMD components and I have quite a lot of parallel wires. As I was checking the layout by myself I did not find non-inverting input and op-amp output in parallel. Probably it is something else.

Can somebody who is experienced in making high gain pedals please check my layout and give me advice? I have added the image to this post.

I am also interested in general guidlines, materials, forum topics etc. covering the principles of making high-gain pedal layouts (overdrives, distortions, fuzzes etc).

I can hear a high-pitch squealing when the guitar is not plugged in (or if it is then I just need to turn up the gain or volume and it starts the same):
Treble, Presence, Volume, Gain and Trim all make the squealing louder. The Tight pot changes the pitch of the squealing. 

Here is the Layout image link:
https://www.dropbox.com/s/c8757vsjuei9tch/Layout-pic.png?dl=0

The bottom layer is all ground.

[PRR]

> R9 and R10 are 2.2k and C8 is 1uF

C9 is in here also. Change that?

> squealing when the guitar is not plugged in

Don't play un-plugged.

We normally wire the Input jack so, when not plugged, a 3rd finger shorts the input. This plays "nothing" instead of "all the crap in the air".

> turn up the gain or volume and it starts

Output wires couple (through air!) to input wires. Same as loudspeaker to microphone, squeal. High-gain circuits need tight neat wiring.

Pictures?

Or wait... did you design a PCB so all wiring and jacks are on-board? Someone with more PCB-foo than me should look it over.

[PRR]

?

[dschwartz]

 Not even ground pour between the tracks....bad

[PRR]

Looks like ground-pour to me?

I would think that would take the edge off the sneakage. But those IN and OUT tracks maybe should not be that close together.

[dschwartz]

Yeah i meant no ground pour between the in and out tracks.
I'd cut one of the traces on both sides and jumper it it a shielded wire.

[boga]

Hey, guys!

@PPR

I did not change the C9 value. And I understood that as soon as I posted the question over here. My first logic was to increase it so that the R*C ratios would be the same (as I did with the other components). I also found some thread which was also about oscillation and I found out that usually these small caps between op-amp's invereted input and op-amp's output are added for stabilizing purposes.
So I put 470pF instead 47pF over there. That decreased the oscillation point, but did not remove it. So I put 6.8nF cap for C9 - now the oscillation is almost entirely gone. I turn all the controls way up (including the trim) and no oscillation. Although it seems that I lost some of the treble because of that. I probably have to find some value in-between.

And I have connected the 3PDT switch so that the input is grounded when the FX is bypassed. But if I just remove the guitar from the input and the FX is ON, then I get a real nasty oscillation and squealing - must be somehow related to the guitar impedance.

IN and OUT tracks are pretty close to each other, yes! I could try cutting one and moving further. There is a small ground fill area between them. And the bottom layer is all GND. I am thinking is this even correct way to do it? To make GND polygons? From the op amp point of view this is like -4.5V as it is biased from the non-inverting input with +4.5V (which is virtual ground).

[boga]

> R9 and R10 are 2.2k and C8 is 1uF

C9 is in here also. Change that?

No I did not change that in the first place. But take a look at my last answer. Found that this value makes a big difference. At 6.8nF my pedal is not squealing anymore.

Also found a good application note which talks about op-amps oscillation:
http://www.analog.com/media/en/technical-documentation/application-notes/an148fa.pdf

I have not read it through yet!

> R9 and R10 are 2.2k and C8 is 1uF

> squealing when the guitar is not plugged in

Don't play un-plugged.

We normally wire the Input jack so, when not plugged, a 3rd finger shorts the input. This plays "nothing" instead of "all the crap in the air".

Looked at your questions and post again PPR. Now I understand what you ment. OK, I need to use a stereo socket at the input and if I pull out the jack then it will already short the input. Nice trick! Thanks alot. I will do that.

> turn up the gain or volume and it starts

Output wires couple (through air!) to input wires. Same as loudspeaker to microphone, squeal. High-gain circuits need tight neat wiring.

Pictures?

Or wait... did you design a PCB so all wiring and jacks are on-board? Someone with more PCB-foo than me should look it over.

I did not make a PCB yet for the 3PDT switch. So yes, they are right now just air-wires and not shielded. So probably input and output wires are close together or even touching. Should I use shielded wires for the connections? Probably LED connections are whatever, but for input and output connections I should use shielded wires, probably?

Do you have any good recommendations where I can find more information about High-Gain pedal layout, shielding, design and such?