JPTR FX Kaleidoscope - Please explain this unusual filter

[SprinkleSpraycan]

Can someone please explain the filter containing C6,10 and R10,12,14,15? I did the trace from my own pedal. I have gone over it many times and am certain its correct. It sounds identical to the breadboard version.

It doesn't look familiar to me. There is a discussion started at  https://www.freestompboxes.org/viewtopic.php?t=34251 but no insights yet. I thought this might be a place for a more definitive ruling.
Is it even needed? I disconnected it on my breadboard to no audible change.

[antonis]

It's a composite LPF inside U2C NFB loop..

What puzzling me is the point IC2A output is connected..

[antonis]

Without the full schematic... you'll need to provide the complete schematic...

AI's are totally blind, don't they..??

@Stephen...!!!
Another customer for moderation here..

[ElectricDruid]

Can someone please explain the filter containing C6,10 and R10,12,14,15? I did the trace from my own pedal. I have gone over it many times and am certain its correct. It sounds identical to the breadboard version.

It doesn't look familiar to me. There is a discussion started at  https://www.freestompboxes.org/viewtopic.php?t=34251 but no insights yet. I thought this might be a place for a more definitive ruling.
Is it even needed? I disconnected it on my breadboard to no audible change.

Stuff like that screams "SIMPLIFY!" at me. I'd be very suspicous of it. Sometimes the design process is "well, I added this bit and it sounded better, so I left it, and then I added this other bit and it sounded more sagacious in the Oswald frequencies, and then I added the Zinger Mods from StompNonsense.com" etc etc. You just never know.

It almost looks like one of those "Generate an equivalent RC network for this" problems you get in textbooks. My process would be to do an LTSpice sim to see what the frequency response looks like and to see whether it really does anything at all, and then create a simpler replacement alongside and tweak it in the sim until I'd matched the frequency responses.

[ElectricDruid]

Ok, I've tried it in LTspice. Wired like that, it does basically nothing at all. I can completely believe that removing it made no change to the sound.

However, if the front of the network were connected to the -ve op-amp input (so that it's in parallel with the 100K feedback resistor) then it does something:



If that connection were moved, would that make more sense?

[antonis]

For low frequencies, there is a 172k (4 X 43k) set in parallel with 100k (R13) feedback resistor..
Fow high frequencies, a Δ-Y transformation is needed (and I'm very lazy for it..)

[Rob Strand]

Can someone please explain the filter containing C6,10 and R10,12,14,15? I did the trace from my own pedal. I have gone over it many times and am certain its correct. It sounds identical to the breadboard version.

It doesn't look familiar to me. There is a discussion started at  https://www.freestompboxes.org/viewtopic.php?t=34251 but no insights yet. I thought this might be a place for a more definitive ruling.
Is it even needed? I disconnected it on my breadboard to no audible change.

To me it looks like you have a subtle tracing problem.

If I had to make it into something, the feedback network looks like a tilt control.  It could perhaps be connected as two nested feedback networks but I've got some doubts about it actually working.

This might not be correct but it could inspire somethings to check:




A different type of correction would be C6 + R14 (input network) form a pre-emphasis network and C10 + R15 (feedback network) form a de-emphaisis network.   The main signal passes the first time with pre-emphasis + de-emphasis ie. flat.  However the repeats only see the de-emphasis.   That makes the high frequencies of the repeats attenuate quicker, like a room.

[bluebunny]

@Stephen...!!!
Another customer for moderation here..

The d**kh**d has been nuked. 

[ElectricDruid]

A different type of correction would be C6 + R14 (input network) form a pre-emphasis network and C10 + R15 (feedback network) form a de-emphaisis network.   The main signal passes the first time with pre-emphasis + de-emphasis ie. flat.  However the repeats only see the de-emphasis.   That makes the high frequencies of the repeats attenuate quicker, like a room.

I like this theory. That's good!

[duck_arse]

The d**kh**d has been nuked. 

who, me?

[SprinkleSpraycan]

To me it looks like you have a subtle tracing problem.

Checked again for sanity sake.

If I had to make it into something, the feedback network looks like a tilt control.  It could perhaps be connected as two nested feedback networks but I've got some doubts about it actually working.

This might not be correct but it could inspire somethings to check:

Played with this but it only served to dampen feedback in either configuration.

A different type of correction would be C6 + R14 (input network) form a pre-emphasis network and C10 + R15 (feedback network) form a de-emphaisis network.  The main signal passes the first time with pre-emphasis + de-emphasis ie. flat.  However the repeats only see the de-emphasis.  That makes the high frequencies of the repeats attenuate quicker, like a room.

This makes sense to me. To hear it I'm guessing I would need better ears and to play at much louder volumes...possibly damaging my hearing.......I see you God, well played.

Seems like Druid concurs too. Unless he was being sarcastic but that's probably because I read everything that way.

[ElectricDruid]

Seems like Druid concurs too. Unless he was being sarcastic but that's probably because I read everything that way.

No, no, I concur! I thought it was a good idea and it provides a viable explanation of what that whole mess is doing there in the first place, which is otherwise pretty obscure.

[bluebunny]

The d**kh**d has been nuked. 

who, me?

Other d**k.

[Rob Strand]

Other d**k.

Who is Dirk anyway?