DC voltage getting through output coupling caps

[drchek]

I have a completed project where a small DC voltage is appearing at the output, somehow making it past the coupling caps and pulldown resistor. As you can see in the relevant part of the circuit above, I'm measuring the bias voltage before the caps, which I would expect, but there remains an ~800mV DC at the output. This is causing the pedal to pop when the switch (dis/)connects the output to the jack.

Any idea where this is coming from or how to eliminate it?

[Kipper4]

Is it a leaky cap?
Have you tried replacing them with a single 2u2 electrolytic?

[drchek]

I replaced both caps, but also checked with just the single non-polarized cap since electrolytic capacitors can sometimes be leaky. No difference.

[bluebunny]

What's G1 connected to?  What DC do you measure if R11 is disconnected?

[drchek]

Have you tried replacing them with a single 2u2 electrolytic?

Weird. Just for the hell of it I replaced the two caps with a single 2u2 EL and now the voltage does go down to zero. Now its got me wondering what the point of splitting the cap in the original circuit was? I've never seen that before. Most circuits have a single cap to couple to the output...

[GibsonGM]

Doesn't C8 || C9 = 2uF?   Yeah, showing it that way is weird.     

Weirder still is that .8V you measured, tho! 

[drchek]

Doesn't C8 || C9 = 2uF?   Yeah, showing it that way is weird.     

Weirder still is that .8V you measured, tho!

I'm guess there was some reasoning to have both a polarized and a non-polarized cap in there? Not sure what advantage that has though?

Oh well, my pedal is pop free now!

[GibsonGM]

If there was a reason, I can't find it, LOL!  I try to use NP where possible.  Glad it worked!

[R.G.]

I replaced both caps, but also checked with just the single non-polarized cap since electrolytic capacitors can sometimes be leaky. No difference.

Some questions:
1. What kind of cap is the 1uF non-electro?
2. Is it really a cap?
3. What happens to the DC levels if you replace both caps with 0.1pF/2KV capacitors - that is, remove them entirely? This tells you if the problem is the caps or the PCB, or some other thing you're confusing with leaky caps. More directly, it tells you if the leak is from G1 or the "output".
4. What happens to the DC level if you remove just the polarized cap? What you're describing is very much like what would happen if the non-electro cap was leaky. Rare, but it does happen. As Sherlock told Watson - eliminate the impossible, and whatever remains, no matter how unlikely, must be the truth.
5. Was the PCB dirty and/or contaminated with leftover flux or bar funk? That can be conductive too, and soldering can boil out the water and other contaminants that let it conduct.