Op amp OD whining

[alderbody]

I recently built a Son Of Screamer /Eternity kind of Overdrive and although it sounds great i couldn't defeat a whinig noise at circuit (true) bypass, when the gain pot is at the higher levels.
When the circuit is engaged the noise is not audible, but it surely is annoying when bypassed so i would like to make it go away.

There is definitely some op amp oscillation happening, probably because i did not add the 100n power decoupling ceramic, but there also must be some ground loop issue. There are a lot of offboard connections in there (many switches), plus the power circuit (Vcc, Vref) is on a separate board.

I tried to touch the power pins (4 and 8 ) with the cap leads to see if it would stop the oscillation, but it didn't seem to kill it.

I read somewhere that grounding the jacks through the enclosure helps eliminate ground loops.

What are your de-noise_ing methods for such circuits?
The common practice for the power decoupling 100n is to place it as close to the chip as possible. Is soldering it across pins 4 and 8 (on the base leads) the correct way to go?

I will post a gutpic soon, if that would help getting a clearer picture of my setup.

For the time, here's some pics of the pedal.

[deadastronaut]

use 'grounded input' bypass switching.......scroll down to dino's comment.

https://www.diystompboxes.com/smfforum/index.php?topic=122000.0

and use sheilded input wiring.

[alderbody]

thanks! i'll read it asap.

I already have shielded wiring all around.

You'll get to see it in the (soon to be posted) gut pics.

Thanks again!

[alderbody]

Here's the gutpic...

Yes, that's a big input cap...
I used some parts that have once served my twin reverb, so this was one of them...

btw...

Grounding the input is a cool way to go for quiet bypass operation, indeed.
But i would like to be able to cope with the oscillation, because in some (of the many...) available settings the pedal squeals.

Maybe i'll re-build the power board and try the known anti-gndloop procedures.

I tried it with one of my strats which is intentionally rather microphonic and i almost couldn't find a setting that did not squeal.

...but with this strat (with which i made all the R&D) is sounds amazing

[amptramp]

You have two separate problems here, grounding and oscillation.  As well as the capacitors to the power leads, make sure the feedback capacitor (usually a small value on the order of tens of pF) is connected because otherwise you will get a phase shift that leads to high-frequency oscillation.

The grounding problem would show up as a hum or buzz when the pedal is switched in.  Please tell us how you are powering the circuit because if you are using a separate supply, it may be the culprit and you need to have a star ground system where all currents return to one point and this point is decoupled properly.  If you are using a battery and you still get hum or buzz, this is a ground loop problem and star grounding is good but it is not possible to run a star ground over the whole system of guitar, pedal and amplifier.  Grounding the input to output through the chassis works until you get corrosion on the chassis.  Then you need to add a wired direct connection between the grounds of the input and output jacks.

[alderbody]

You have two separate problems here, grounding and oscillation.  As well as the capacitors to the power leads, make sure the feedback capacitor (usually a small value on the order of tens of pF) is connected because otherwise you will get a phase shift that leads to high-frequency oscillation.

I tried a small cap in the feedback loop during the R&D phase of the project and i recall that it reduced if not eliminated the oscillation, but it altered the tone as well. In the Eternity schematic there is no such cap, though...  Maybe i'll do it after all.

The grounding problem would show up as a hum or buzz when the pedal is switched in.  Please tell us how you are powering the circuit because if you are using a separate supply, it may be the culprit and you need to have a star ground system where all currents return to one point and this point is decoupled properly.  If you are using a battery and you still get hum or buzz, this is a ground loop problem and star grounding is good but it is not possible to run a star ground over the whole system of guitar, pedal and amplifier.  Grounding the input to output through the chassis works until you get corrosion on the chassis.  Then you need to add a wired direct connection between the grounds of the input and output jacks.

The whining/oscillation happens with either a battery or an adapter. I only have that kind of noise. There's no hum whatsoever.

If i get to add a 100n ceramic to the power pins of the chip, could i just solder the cap leads directly on pins 4 and 8 (ok, on the chip base leads).
I mean, pin 4 is at ground potential and we need to place that cap as close to the chip as possible, according to theory.
If i add the 100n across the filter cap (100μF in my build) as i see in a lot of schematics, wouldn't that be far enough?

I don't know. i'm a bit confused about how to use that decoupler...

There's something in my circuit that causes oscillation.
The chip i'm using (LM1458) was the quietest of all i tried, and i tried a lot. I remember LF38something was excellent on the breadboard, but on the actual build it squealed like a pig at almost every setting. Only with the gain at 0_ish i got it to stop...
And that happened with the most chips i tried.
So there IS something in there...

[amptramp]

Power supply pin bypass caps are used to avoid exactly the problem you have.  In fact, there was a company that made DIP sockets with a thin Kapton film capacitor that fit under a DIP device.  It was often used for digital logic as some edge frequencies go very high - 74FXX series TTL has an edge rate corresponding to 175 MHz.  The capacitors would be much smaller, on the order of 1 nF because even with the capacitor being made part of the socket, the inductance at 175 MHz could be appreciable.

In your case, anything from 10 nF to 100 nF would be suitable.  The bypass has to operate at the frequency of the oscillation but it never hurts to stabilize the power supply in the audio region.

[alderbody]

Thanks!

I'll take my chances with those caps.

I already breadboarded a new power supply with 50μF electros (i have 100's) and i'll test some bypass ceramics to see if i get any better results.
I want to try a decoupler for the Vref branch as well.

But i'll definitely try the ceramics directly on the base after all.

I will get back as soon as i have results.

Thanks again!

[alderbody]

Found some time to deal with the pedal this weekend and i tried all the known remedies...

Nothing seemed to work at all, trying to kill the oscillation itself.
I breadboarded a new power supply, hardwired caps here and there, disconnected the jack ground wire, to no avail.
I even tried to lower the value of the gain pot (parallel resistor) and it only lowered the pitch of the whining...
I wanted to parallel a 500K fixed resistor in order to almost halve the value, but i used a 47K by mistake which gave a total of about 47K.
That would be a very low Gain pot value if i had to replace it, but it still oscillated even at that value.
So, no-go for that either...

The only thing that finally worked was the input grounding trick.
It may deal with the result and not the cause, but it definitely works.

...or is this the cause after all?

I mean, the noise happens only at circuit bypass. As the gain pot turns past about 1/3 of the travel, the oscillation starts and its frequency increases rapidly, then drops a bit and then rises more gradually till the max setting....
Using a croc tipped cable, as soon as the input is grounded i hear the pitch dropping rapidly and then vanishing to a dead quiet bypass...

I guess it's time to rewire that 3PDT...

Thanks for the tips, guys!

[abram]

if it's only whining in bypass mode, it's possible that the effect circuit starts oscillating once the input is left floating, and then that oscillation is able to couple into the bypass circuit. in that case, it very well could be the culprit. either way, glad to hear you found a solution!

[MrStab]

i once thought it would be a good idea to make a distortion pedal with a gain of over 1,000. it was a bit silly in hindsight, but i did manage to tame the squeal. consistently, over multiple builds.

a few thoughts that might help: what 3PDT wiring scheme are you using? i find it's better to put the signal on the outer lugs, so the ground/LED switching is on the middle poles. this helps keep distance between them.

also, if nothing else works, what happens if you turn Gain up all the way and lift the input & output pulldown resistors? where are they in the ground path, in relation to the op-amp and power filter grounds?

i see your output wire is shielded, but it does seem quite close to the wire for that pot. does moving wires around do anything?

[alderbody]

what 3PDT wiring scheme are you using? i find it's better to put the signal on the outer lugs, so the ground/LED switching is on the middle poles. this helps keep distance between them.

Yes, i always keep them (in/out) as far as possible.

also, if nothing else works, what happens if you turn Gain up all the way and lift the input & output pulldown resistors? where are they in the ground path, in relation to the op-amp and power filter grounds?

Haven't tried lifting the p/d resistor. If i remember well there's only one at the input side and it's rather far away from the chip.

i see your output wire is shielded, but it does seem quite close to the wire for that pot. does moving wires around do anything?

I have tried moving, crossing, paralleling, whatever_ing the wires. Nothing seems to alter anything.
Only when the circuit was on the breadboard i had noticed such effect with moving wires. Once placed in the board/box nothing changes.

I believe it must be what abram says above. The circuit possibly oscillates when the input is left floating.

I just need to find some time to work on the pedal to fix it right... 

I hope to be able to post some clips sometime...

[amptramp]

Don't leave the input floating in an overdrive when you are in bypass mode.  There is a different problem if you are designing a delay - if you ground the input, there is no signal at the output until the signal has passed through the delay, but that is a different problem.

[MrStab]

ah right. i think the "left floating" thing may actually be quite common with dirt pedals, and it's often masked by the input-jack-power-on system. i expect a lot of commercial pedals may even do that.

[alderbody]

I actually rewired it today and it's perfect.
Problem fixed!

Thanks to all!!!