Help/Thoughts on Original Boost/Overdrive

[masinyourface]

I'm having somewhat of an issue with a gain stage I designed for an original overdrive circuit. Let me explain my thought process in designing it, then I'll get to the issue itself.

After reading the Bazz Fuss Fun article on the homewrecker site (http://www.home-wrecker.com/bazz.html), I thought it would be interesting to use a diode in the same manner for an overdrive circuit in order to bias the transistor and shape the tone/grit. Now, I've been experimenting alot with MOSFETS and JFETS lately, so of course the next logical step was to use the diode with a MOSFET. I ended up using a blue/white LED combo in order to get a larger voltage drop and in turn more volume/less grit. I then tried to figure out a way to control the gain of the circuit. A simple attenuating pot slapped on the front of the circuit would have probably worked fine, but having built a MOS-Face/Silicon Fuzz Face hybrid about a year ago that turned out amazing, I thought it would be interesting to use a pot/cap combo like that in the Fuzz Face going from my transistor's source to ground. Here's my current schematic:



Now for my problem....when the gain knob is at its minimum, the circuit works great. Everything sounds great. When it's turned up though, it starts to work pretty funky. Single notes and soft strums on my single-coil strat work, but anything louder, especially when coming from the humbucker on my semi-hollow, makes the whole pedal go haywire. Everything becomes gated and starts sounding hideous. One thing I've noticed is that when this happens, the drain voltage drops noticeably. When working right, Vd is about 5.6V. When it goes haywire, it drops to about 2.4V. After some experimenting, I found that the only 'remedy' is turning off the power and shorting out the leads on the transistor. Any ideas on what's causing this? Any ideas on how to fix this?

[PRR]

Bazz Fuss uses BJT which draws significant Base current, which causes voltage-drop in the Base bias diodes.

Your MOSFET draws _NO_ Gate current. I have no idea what the Gate voltage "should be". I suspect it "works" only by stray leakage. I don't see how it can be bi-stable (gets stuck low); maybe something breaks-down.

Put 100K from Gate to ground and see what happens. However I think it is an unpredictable and perhaps irreproducable design.