Soundcard Fuzz - schematic & funny story

[midwayfair]

Some "fuzzy badness" for you guys.

A guy over on Madbean was looking for a way to recreate the sound of recording in the red direct to his soundcard. I gave him some suggestions and he told me he didn't solder. So I offered to breadboard something to see how close I could get and this is pretty much what I came up with. He said that would be great, so I filmed a quick video of a couple variations. He told me I got really really close and it was what he was looking for, then an hour later posted that he bought a Basic Audio Futureman instead, but it was okay and he would still consider buying one of these for "double the cost of parts" so I could have "some funds" to build one the next guy who wants it. 

Fortunately Madbean's forum is down for maintenance so I can't post a reply in that thread. (I'm not mad that he didn't want a build or that he bought something else because he was in a hurry, but I am a bit insulted that he thought I'd design him something almost from scratch to duplicate a sound he needed for like $40-50 when he was willing to impulse buy a $150 pedal.)

Technically that's Karma for the time I picked John Lyons's brain about building me a custom fuzz (this was before I could have built my own) and after he patiently explained a few ways to do it and improve on the idea, and quoted me a reasonable price, I went and bought something from someone else instead anyway. (Sorry, Mr. Lyons! But the good news is the fuzz I got instead is still on my board and has been my main dirt for years.)

Anyway, I thought I'd share this simple but fairly horrible sounding thing with everyone. It's simultaneously nasty and only medium gain. At full gain it does a bit of stuttering from too much bass overloading Q2, the biasing on which is just awful.

Primitive "How It Works"
Q1 is a simplified FETzer valve with an input cap. (I can never get a FETzer valve not to pop.) Q2 is a screwed up transistor stage followed by hard clipping diodes (D2 & D3) to ground, a semi-mysterious diode (D1) pointing toward the output, and a treble cut that is really just excessive.

D1 creates a very slight octave effect but also gates. (I'm pretty sure I've seen it somewhere, or at least something like it, I mean other than the Octavia and the Green Ringer, but I can't remember where, so if anyone knows what I should give credit to, please let me know and I'll add that accordingly.) D4 (using that emitter diode bias again here) creates a very tiny bit of gating sound as well. If you ground the emitter of Q2, it smooths things out a bit and sounds more normal. The gating was important because the direct recording doesn't seem to add any sustain or much compression. Revolution for instance is completely fuzzed out but that guitar sound certainly doesn't "sing". This doesn't get as dirty as Revolution, but that was just a common touchstone for people to know what I'm talking about here.

The tone control is something of an afterthought, and 33nF is probably excessive even for the sound I was going for. I thought it sounded "best" (purely subjective term here ... basically just the chunkiest sound) as just a 15nF to ground, so it could easily be a two knob job, or I suppose that cutting the tone cap value is an option. But it does have a kind of interesting thing going on where the 10K prevents the "octave" output from ever having as much treble cut as the main output, and the octave sounds a bit more prominent on higher tone settings.

As a final note, just like in the Electra and a few other circuits, you can sometimes get a mild down octave out of a 2N3565, so that's the recommended transistor. Does anyone know why that transistor seems to do that?



If anyone actually wants to hear it, here ya go:

[John Lyons]

It's kind of a bummer when stuff like that happens but people
can be and often are fickle and impulsive.

What is the pedal you have been using for so long?
Thanks for sharing the circuit, sounds interesting.

[deadastronaut]

That's human beings for you....don't have anything to do with them..

[Kipper4]

Intresting octave effect. Curious now about those transistors.

[midwayfair]

What is the pedal you have been using for so long?

Just a hybrid fuzz face with a ton of knobs and passive boost on a second footswitch.

[amptramp]

The second harmonic comes from the square-law transconductance curve of the FET.  From the trig identity:

cos²(u) = 0.5 + 0.5*cos(2u)

If we replace the angle u by 2*pi*f*t where f is the frequency and t is time, trig identities all have frequency identities corresponding to anglular identities.