Red Llama build high pitch whine after gain increase mod?

[Bandwagonesque]

Hi everyone, hope this doesnt fall under TL;DR. I built a red llama clone based off this 1590bb layout seen below. It worked great stock, but decided to go ahead and do the old mod where the 100K feedback loop resistor is set on a switch VS a 10M resistor as used on the original tube sound layout to achieve a foot switchable second channel of sorts with increased gain. That worked great too...almost.





with the pedal set using the 10M resistor with higher gain, I am experiencing a high pitched oscillation when I have the volume dimed (and drive maxed), the pitch really starts right around 3 o'clock. Its quite annoying and while I reckon I won't be pushing the volume that high on this in a live context, this pedal would be just about perfect if I could get rid of the whine wherever the knobs are set, especially if I decide to dime things. There is also a weird chirp that happens whenever I kick the effect on when I have the 10M set. Its like a bypass pop but a chirp instead. Would a pull down resistor possibly help in remedying this particular problem and where would it go?

I tried the power filtering tricks (small disc cap from output to ground, cap to ground with 100ohm resistor in series to 9v, but haven't had any luck without big sonic detriment.

Upon trying to hack a solution I came across this post of interest that I figured id bring up. I put this into practice but turned out to be irrelevant as this layout seems like everything is in its right place.

Here's how you can check - Pins 3, 5, 7, 9, 11 and 14 of the 4049 chip are the amplifier inputs. Only two amplifiers are used, so four of these pins should read continuity between them. Once you've identified which four amps are tied together, measure that point (the junction of all the unused inputs) to ground and to V+. If they are all tied to V+, you're all good and that's not the source of your noise. If they are tied to ground, that's the issue and you'll need to lift that point from ground and tie it to V+. Let me know what you find!)

would love to be told if im wrong though and if the layout i worked off does have things wrong with the amplifiers on the chip

The build is clean but tight in a small enclosure, but I don't think that using a grounded shielded wire for the input is going to help much (the lead from the jack to switch is less than an inch). Ive shielded the enclosure to ground as the jacks im using are the all plastic low profile non grounding type and I am using an unbuffered UBE so i don't think either of those are the issue. I've taken the guts and moved the in/out leads to see if that helps but the high pitched whine still remains.  The problem only exists on the 10M setting after all. But does anyone know what I can do to solve this and get rid of the ringing? The layout seems correct but if anyone sees anything amiss I'm all ears. Thanks for any help folks and im open to all ideas and solutions (thinking out loud, but how does one 'kill off' the last 3/4ths of a pot so that the end of its throw stops right before the point at 3 o'clock where it starts to oscillate? figure if I don't give it the chance for the pot to be turned up high enough to start whining, then i can live with that), it'd be greatly appreciated if i can get this thing perfect and rid of the whistling.

[shooter_mi]

I've built a three of these pedals, two to Tube Sound Fuzz specs and one Red Llama. The most recent Tube Sound Fuzz I built did exactly what you're describing. When the 10M resistor is engaged, and volume is wide, it oscillates. For some people, it might be a cool sound—you could manipulate the knob with a foot to induce/reduce the oscillation. I'm more like you, though, and I don't really like it.

I found that added a low pass filter before the output stops the oscillation. I added a 25k pot to a .083 cap to ground just after the output cap. With the filter open, it still oscillates, but backed off even a little bit the highs are attentuated and it stops. You could probably achieve the same thing with a cap from the effect side lug of the volume pot soldered back to ground via the back of the pot. A value might be something like .022u or even smaller. You just want to reduce the high end enough to kill the oscillation.

I got the Anderton book back in 1990 and have a lot of affinity for those projects. But, I think the Red Llama mods do make it more usable. The Llama has larger caps across the inverters to reduce highs, so making that mod might fix the oscillation, too.

[anotherjim]

As it is, it shouldn't be unstable even if you do increase the gain. But, x2 inverters means the circuit is non-inverting and any feedback from output to input will be positive feedback and that can cause it to self-oscillate. When feedback only occurs with increased gain, I have to suspect there is a path somewhere between input and output.

Wiring is often the culprit - wires too long, tangled. All wires can provide an unintended path via capacitive or inductive coupling -  it's weak, but if you increase the gain?
In and out jacks adjacent instead of opposite sides.
Adjacent contacts in a stomp switch. If a 3PDT, put the LED switch contact in the middle row to separate in and out contacts.
How have you done the resistor switch mod? I would swap the 100k for 1M and drill some more holes in the pcb so the 100k can go next to it and connects to the chip output, then two wires to a switch connects the 100k to the input side at the other end. Only 2 wires to the switch which can be twisted together to reduce noise pickup effects. When switched on the 100k is in parallel with 1M which is slighty less than with just 100k, but not audibly so.

[LightSoundGeometry]

As it is, it shouldn't be unstable even if you do increase the gain. But, x2 inverters means the circuit is non-inverting and any feedback from output to input will be positive feedback and that can cause it to self-oscillate. When feedback only occurs with increased gain, I have to suspect there is a path somewhere between input and output.

Wiring is often the culprit - wires too long, tangled. All wires can provide an unintended path via capacitive or inductive coupling -  it's weak, but if you increase the gain?
In and out jacks adjacent instead of opposite sides.
Adjacent contacts in a stomp switch. If a 3PDT, put the LED switch contact in the middle row to separate in and out contacts.
How have you done the resistor switch mod? I would swap the 100k for 1M and drill some more holes in the pcb so the 100k can go next to it and connects to the chip output, then two wires to a switch connects the 100k to the input side at the other end. Only 2 wires to the switch which can be twisted together to reduce noise pickup effects. When switched on the 100k is in parallel with 1M which is slighty less than with just 100k, but not audibly so.

I found if the power wires run parallel or close to the audio wires it causes noise and whine

[Bandwagonesque]

thanks a ton you guys. wouldn't have been able to figure it out without your help! i ended up sticking a cap on the effect side of the volume pot to ground as suggested by shooter_mi and that was able to do the trick just great!

[Radical CJ]

 I've never tried the 'tube sound' specs, but my home made Llama clone is horrifically loud. I never take the volume past 9am. The last thing I'd think of is "you know what?.. this thing needs more gain!"

I've never compared it to another, so it could be an anomaly. But I wouldn't be all that surprised if increasing the gain on mine brought it to the point of self-oscillation.