Bizarre problem: Ibanez CP9 Compressor - No sound when off

[clarisso11]

Hi everyone,

I'm working on repairing a 1981Ibanez CP9 compressor pedal, with a very strange problem: It works seemingly perfectly when turned on, but no sound passes through when off.

I started by checking all electrolytic capacitors with an ESR meter, and all were looking good. Next, I tried understanding the switching circuit, since that seemed part of the problem.

The pedal uses JFET bypass switching, and the audio signal always passes through the input and output buffer (schematic attached). With Q107 On, the signal passes through all effect stages, and this seems to be working well.

Now here is the bizarre thing: I traced the audio signal when the effect is off, and it goes all the way from input through the Q108 FET, but ends completely at C117. With the effect on, it suddenly passed C114! Voltages at CP8 and CP9 more or less check out (0.7V and 7.8V).

I have tried to annotate the schematic to show my observations (purple is the dry circuit, and green is part of the effect circuit).

I'm quite puzzled by this, and would really appreciate any help or explanation as to what is going on, I'm sure I'm missing something!

[Enzo]

Broken trace between here and there?

[clarisso11]

Hi Enzo,

No broken traces from what I can tell. The added wires and resistors turned out to be adjustments from factory, you can just about see them on the photo from the service manual.

I noticed that some of the traces had darkened, but confirmed continuity on all of those.

[Enzo]

So you have signal at both ends of Q108 but not at C114? Is there a zero ohm connection between Q107, Q108, and C114?

[clarisso11]

So you have signal at both ends of Q108 but not at C114? Is there a zero ohm connection between Q107, Q108, and C114?

I just double-checked, and I get audio signal across both ends of Q108 and reaching one end of C114, but not the other side of C114.

I'm not sure if I'm measuring the correct points, but resistance across both FETs is around 1MOhm when the effect is off, and drops to about 380 Ohm only on Q107, when the effect is on.

[Enzo]

Only one JFET should be on at a time. If they both go on or off at the same time, then look at the control signals CP8 and CP9.

I can't imagine why signal from one path would pass through C114 but the other signal wouldn't. Assuming both signals get to C114.

[clarisso11]

Only one JFET should be on at a time.

Would the drop in resistance for Q107 mean that it turns on, but Q108 remains closed since the resistance doesn't change?

I can't imagine why signal from one path would pass through C114 but the other signal wouldn't.

That's what's confusing me, but maybe the signal from the audio probe isn't giving the full picture..

[g1]

Would the drop in resistance for Q107 mean that it turns on, but Q108 remains closed since the resistance doesn't change?

Q108 is not remaining closed, it is open. Jfet should be low resistance when no power applied. Check drain to source resistance of both jfets with power disconnected. I think you will find Q107 is low resistance, but Q108 is open (bad).

[clarisso11]

Q108 is not remaining closed, it is open. Jfet should be low resistance when no power applied. Check drain to source resistance of both jfets with power disconnected. I think you will find Q107 is low resistance, but Q108 is open (bad).

Hey g1,
thanks for helping out! There seems to be similar drain to source resistance for both FETs, in-circuit, and power disconnected. Q108: 0.46 kOhm, Q107: 0.5 kOhm.

I think I will try desoldering both, and testing them out-of-circuit.

[clarisso11]

I removed Q108 and Q107, and both tested fine. In the end, I went with the nuclear option: Bypassing the JFET switching all together, by converting to true bypass. I works perfectly in this configuration, but it does bother me a little not to know what the issue was.

Thank you everyone for your help!