Fuzz Face issue

[sonarchotic]

I've been bread boarding silicon Fuzz Faces with BC108's trying to dial in something decent. I finally found a pretty good sound and got rid of most of the noise/laser sounds/oscillation I had going on by messing with all of the resistor values but every version I've made just about has had this issue.. when I roll off the guitar volume the voltage at the collector of Q2 spikes. The voltage goes up with less than a quarter turn CCW on the volume knob to around 5.7v, resulting in a rapid volume drop instead of a smooth roll off. It then heads back down in voltage with continued CCW turning of the volume knob. At full CW the collector voltage is 4.5-ish. This happens with two different guitars that don't do this with the effect off.
I have the voltages at all the transistor pins about were they should be according to the silicon version voltages I've found online. I have found it necessary to bias both collector resistors and lower the value of the feedback resistor just to get to where I am. Not one of the versions I built although having similar part values worked without major tweaking. I'm using the lowest gain Bc108's I can find, around 170 hfe. I'm building this for a friend who specifically asked for Bc108's so I'm trying to stick with them. Any help would be great, thanks!

[sonarchotic]

Got it figured out! I lowered the feedback resistor too much. I raised it back up to 47k and now no volume issue even though I'm only getting .1v at q1 base. Weird thing is though if I go with the stock 100k I get no voltage back to the base of q1 and the volume issue with oscillation. I'm calling this one fixed.

[jkokura]

I was gonna say, the Fuzz Face is one of the most durable circuits I've played with. You can really mess with it for some fun. Check out the Fuzz Factory for example.

Jacob