Troubleshooting a double overdrive built

[Groovenut]

Everything everyone has suggested is good design SOP.

I'm still sticking to my earlier assessment that it's happening at the A>B B>A switches.

Especially if they are DIP switches. Depending on which type of switch, the lead proximity could be a mere 1.27mm apart.
Easily close enough for coupling.

[jackwithoneye]

Everything everyone has suggested is good design SOP.

I'm still sticking to my earlier assessment that it's happening at the A>B B>A switches.

Especially if they are DIP switches. Depending on which type of switch, the lead proximity could be a mere 1.27mm apart.
Easily close enough for coupling.

Thanks, I'm considering this too

What about the PCB tracks proximity, coupling possibility too, right?

[jackwithoneye]

does the whistle persist if you put a buffered pedal in front? like a boss pedal in bypass mode, or a buffered tuner.

cheers

I've been mistaken, i've tried with an HX stomp before, and the pedal (in FXloop) and the whistle was there.
But as a matter of fact, when i put a real buffered pedal before, the whistle is not there anymore.
I don't know what i can conclude but your question was in the bulls eye.

[iainpunk]

does the whistle persist if you put a buffered pedal in front? like a boss pedal in bypass mode, or a buffered tuner.

cheers

I've been mistaken, i've tried with an HX stomp before, and the pedal (in FXloop) and the whistle was there.
But as a matter of fact, when i put a real buffered pedal before, the whistle is not there anymore.
I don't know what i can conclude but your question was in the bulls eye.

i think it has to do with coupling between signal paths internally.
when a good buffer provides almost no impedance, and the coupling does have impedance, in the form of capacitive coupling, the signal from the buffer will win out, even when its pushing silence.
when you use the FX send of the helix or your guitar, there is some output impedance, and the coupling, being capacitive gets the upper hand in the higher frequency range, and causes positive feedback, AKA oscillations.

the best option is to check if the in and out signal wires run close to each other. if they do, separate them, if simple separation isn't possible, use some aluminium foil which is grounded to ''shield'' the input cable.

cheers

[jackwithoneye]

does the whistle persist if you put a buffered pedal in front? like a boss pedal in bypass mode, or a buffered tuner.

cheers

I've been mistaken, i've tried with an HX stomp before, and the pedal (in FXloop) and the whistle was there.
But as a matter of fact, when i put a real buffered pedal before, the whistle is not there anymore.
I don't know what i can conclude but your question was in the bulls eye.

i think it has to do with coupling between signal paths internally.
when a good buffer provides almost no impedance, and the coupling does have impedance, in the form of capacitive coupling, the signal from the buffer will win out, even when its pushing silence.
when you use the FX send of the helix or your guitar, there is some output impedance, and the coupling, being capacitive gets the upper hand in the higher frequency range, and causes positive feedback, AKA oscillations.

the best option is to check if the in and out signal wires run close to each other. if they do, separate them, if simple separation isn't possible, use some aluminium foil which is grounded to ''shield'' the input cable.

cheers

Physically, the in and out signal are running close to each other in the PCB at the order dipswitchs spot (A>B and B>A), hard to avoid completely.
Maybe it's a dumb question, but as i tested out, on your recommandation, that a buffer before solve the whistle, could i add a buffer next to the input would make it even easier(as i quick drawed bellow)?  O maybe 2, one at the beginning of each circuit (A and B)?

[iainpunk]

you could add a buffer, yes, i think it has to go as soon in the pedal as possible, and an opamp buffer is better than a Jfet buffer, since the output impedance of the HX stomp is also on par with Jfet buffers, so lower impedance is needed here.
especially with dip switches, where cross-talk is quite common, isolation is harder to achieve, since the internal capacitance between the poles of a DIP switch is quite high, due to the small size.

cheers

[jackwithoneye]

Do you guys have any recommandations for the input buffer values and arrangement?
Here's what i drawed, does it feel right to you? value wise?

[composition4]

R20/21 are way too low. They combine for an input impedance of 5k, which means you'd need to drive it with something at least around 500R output impedance. If you want to plug a guitar straight into it, I'd be increasing them to at least 1M each... Or better yet, replacing them with a 1M resistor with the other end attached to your VREF.

Also C21 is far too large. Not technically a problem but depending on what follows it, you can get away with a smaller value. Or remove it altogether and direct couple it to the input of your next opamp. If you cobble what you have together with your existing circuit and post a schematic I'm happy to explain in more detail.

[iainpunk]

i'd replace the 10k resistors with 220k, since higher resistors are more noisy, which is especially detrimental in the beginning of a circuit.
C2 could be in the 100nf to 1uf range, and followed by a 1M resistor to ground to negate popping when switching.
the high resistance noise is less of a problem since the opamp will just soak it all up with its superior output power.

cheers

[jackwithoneye]

R20/21 are way too low. They combine for an input impedance of 5k, which means you'd need to drive it with something at least around 500R output impedance. If you want to plug a guitar straight into it, I'd be increasing them to at least 1M each... Or better yet, replacing them with a 1M resistor with the other end attached to your VREF.

Also C21 is far too large. Not technically a problem but depending on what follows it, you can get away with a smaller value. Or remove it altogether and direct couple it to the input of your next opamp. If you cobble what you have together with your existing circuit and post a schematic I'm happy to explain in more detail.

i'd replace the 10k resistors with 220k, since higher resistors are more noisy, which is especially detrimental in the beginning of a circuit.
C2 could be in the 100nf to 1uf range, and followed by a 1M resistor to ground to negate popping when switching.
the high resistance noise is less of a problem since the opamp will just soak it all up with its superior output power.

cheers

Thank you so much for your inputs guys for all these informations and wise advices, i've to test out now.

A small question for iainpunk : Does the 1M pulldown resistor, next to C2 you wrote about, is still needed as i already have a Pulldown res at the beginning of each section (A and B) knowing that the Footswitch 3pdt is wired as a true bypass 3PDT

Signal chain :

[iainpunk]

the idea is that both sides of each switch contact has a pull down, so that it won't build up a static charge when its not connected, which cause the most pops.

cheers

[jackwithoneye]

what is the best setup for an unused opamp of a dual opamp?
I presume that letting these pins open may result in bad results, as the opamp may run full power gain/noise and power consumption.

[composition4]

Inverting input connected to output, and noninverting input connected somewhere in the middle of your power supply pins... VREF is handy

[jackwithoneye]

Inverting input connected to output, and noninverting input connected somewhere in the middle of your power supply pins... VREF is handy

Thanks !