Ibanez SD-9 bypass problem revisited

[TheBigMan]

I swapped out the two 2SK30A switching FETs in the hopes of solving the lack of clean signal.  It worked for around 2 minutes before the clean sound faded out gradually and was replaced by bleed-through from the distortion stage.

Out with the audio probe, and I find that both stages seem to be working fine, but the FET which controls the clean signal doesn't flip when the other one flops (or is that vice-versa   ).  What I can't figure out is what might cause this.  Why should one half of the flip-flop work but not the other?  I can only think it's a bad component inbetween the flip-flop and the FET (biasing resistor or diode or something else) but any advice would be great cos this thing is driving me insane.  :evil:

[MartyMart]

Hey BigMan, i have the exact same problem ! I've fitted new SK30A's, new op amp new switch........ no dice, still have no "clean" signal just the horrible bleed through as you described !!.
Its in the cupboard until my mood improves about taking another look!
Could mean a BIG parts swop, any thoughts ??
Regards,
Marty.

[dolhop]

The FETs are driven by the flip-flop circuit which consists of the BJTs.  Check the biasing on each of the BJTs to see that they're doing their job.  It could be the CR combos that are troublesome and causing the trouble.  Don't start swapping until you have a good idea of what the problem is.

[Boofhead]

I think I see a problem.  As a quick test perhaps try putting 10k in parallel with the 56k's R22, R23.

No time to think to explain gotta go..

[TheBigMan]

OK, doing that puts the pedal in bypass mode permanently but with the LED on.  The footswitch has no effect, the pedal won't toggle.

[RDV]

Build a DIY version from GGG(with my mods to make it loud and obnoxious) and stuff it N the box with a 3PDT switch!

HAH! :twisted:

RDV

[Boofhead]

Did it fix the fizz problem though?

[TheBigMan]

I get clean signal all right, but the pedal won't toggle into active mode, so no distortion.

[Boofhead]

I should imagine an optimal solution will need quite a bit of care in redesign.

All I can suggest is perhaps using larger resistors (perhaps try 22k, 47k) and see if you can find a balance between working and no fizz - you can only hope.

Good luck.

[R.G.]

It worked for around 2 minutes before the clean sound faded out gradually and was replaced by bleed-through from the distortion stage.

Out with the audio probe, and I find that both stages seem to be working fine, but the FET which controls the clean signal doesn't flip when the other one flops (or is that vice-versa  ). What I can't figure out is what might cause this. Why should one half of the flip-flop work but not the other? I can only think it's a bad component inbetween the flip-flop and the FET (biasing resistor or diode or something else) but any advice would be great cos this thing is driving me insane.

Guys, Mother Nature is talking to us, but we aren't hearing it.

JFETs are reasonably simple devices. In a non-damaged JFET, letting the gate and source find their own voltage by opening the gate up will result in the device being as "on" as it can be. You turn it off by pulling the gate significantly below the source (negative for n-channels).

Mother Nature is whispering in your ear. You say

It worked for around 2 minutes before the clean sound faded out gradually and was replaced by bleed-through from the distortion stage.

That sounds a whole lot like there is a voltage somewhere that was correct, but then faded off to incorrect. There are only two voltages that matter here - the bias voltage and the gate control voltage. Either the bias voltage has drifted off to where the source (and presumably drain) is not being held up enough for the low-going gate control signal to turn the JFET off, or the low-going 'off' signal to the gate isn't going low enough. An oscilloscope or DMM ought to tell you which.

The other thing MN is whispering in your ear is that you may have inserted a component incorrectly. If the "worked for around 2 minutes" means it works for about 2 minutes every time power is applied, then you have a connection, pinout, or component placement problem. If that means that it worked only once for about two minutes, then you have a component that was damaged in about 2 minutes and now no longer works at all - like maybe an incorrectly polarized electro.

If it were me, I would use the DMM and find out whether the flipflop outputs - there are two, one for dry, one for wet - are both moving enough voltage to control the JFETs. Then I'd trace the control voltages to the R-C-diode that runs the JFET gate and find out whether the control voltage is sufficient there. That tracing should turn up where the control voltage is not working. If the whole chain is fine, I'd then turn to what is happening with the bias voltage.