Mossy Sloth Fuzz/OD -- Schematic, layout, and discussion

[midwayfair]

This is my new fuzz/OD hybrid design. Thanks to the forumites over on Madbean who voted on the name!

Background/goals:

I've been using a fuzz as my only dirt pedal most of the time for the past couple years. I really like the sound of certain fuzzes when they're pulled back a little, so I started brainstorming ideas for a fuzz that had more OD characteristics. But I also wanted it to be more than just a variation on the fuzz face or some other well-trodden vintage fuzz circuit. So after brainstorming and breadboarding various circuits for a few months, one day here on DIYSB, someone posted about emitter diodes to bias a transistor, and Jimi Photon linked to his Tone Blaster. This felt like the missing piece in the puzzle, so I went back to a few different ideas I'd been tossing around, and ended up with this:

Schematic


Here's my build!


Here's what it sounds like:

How It Works
This has a few unusual bits, so I'll go over the circuit briefly here. For a much, much more thorough breakdown, please read the build document:
https://docs.google.com/document/d/1U9ZENwZcjz2GvbG60vABIp15d1c4UQJ7asshD11qTyM

Q1 is an input buffer to give us a high input impedance (~2M).

Q1 is a drive stage very similar to the one in the Clipper Ship but with more gain. A volume control works as the gain control for the pedal, and then there's a MOSFET as the "hard" clipper at the end (ultimately based on the Electra Distortion). A MOSFET in this arrangement is an asymmetric clipping diode all on its own. One direction is a silicon body diode -- hard clipping -- while the other side has a very soft knee and works more like soft clipping. It's very efficient. The gain control also has a treble bypass to keep things from getting muddy when it's turned down.

Q3 is the "fuzz" stage, with a couple weird things. There's a MOSFET in the feedback loop similar to the Big Muff for a healthy dose of soft fuzzy clipping, and then the emitter has a MOSFET in place of a biasing resistor (this is the trick I picked up from Jimi Photo's Toneblaster). It's not actually clipping so to speak; it's actually more like an exciter effect. Small signals don't see the bypass; big signals see the bypass and get pushed way up. This makes for super touch sensitive distortion, and it also sounds very tube-y.

There's not a ton of output doing it this way, but in lieu of using a make-up stage I "figured out" (er ... IO plugged things into other things on the breadboard where they don't be long) that I could reference a bypass cap to the 9v supply and get quite a bit of extra gain.

Then there's another MOSFET clipper to ground, a bit of treble cut (R15 and C12), and a level control with a treble bypass like the gain control.

Finally, we reach our output buffer to ensure really low output impedance (a few hundred ohms).

Here's the layout:


And an etch image for all the etchers: https://dl.dropboxusercontent.com/u/9878279/Jon%20Patton%27s%20layouts/Mossy%20Sloth%20Fuzz/Mossy%20Sloth%20Fuzz%20etch%20transfer.PDF

I'm sure a PCB will emerge eventually.

[Ripdivot]

In the schematic it looks to me that for the bass control to work properly you would need to either move C6 to the pin 1 side of the bass pot or flip flop C4 and C5. Very cool looking circuit though, I will give it a go on the breadboard in the near future! 

[Ripdivot]

In the schematic it looks to me that for the bass control to work properly you would need to either move C6 to the pin 1 side of the bass pot or flip flop C4 and C5. Very cool looking circuit though, I will give it a go on the breadboard in the near future! 

Oops, sorry about that, I miss read the 10n cap in the bass control as 10uf. Disregard my statement above, all is well!

[WaveshapeIllusions]

Looks real nice. Sounds nice too.I especially like the amount of details, a lot of effort seemed
to go into it.

As to the circuit, a lot of things look especially clever. The bypass on Q3 is awesome, though a bit odd. Without the LED, the MOSFET would have little affect, as AC signal would be bypassed. With the LED, it should only affect the part of the signal above it's forward voltage. I'm pretty sure that's somewhere around 3V? So any signals 3V more negative than supply should break over the LED and get AC bypass. Since the Forward voltage of the MOSFET and LED are different it breaks over one then the other. Positive signals just get the MOSFET bypass boost.

The bias setup likely affects clipping in good ways too. Q1, Q3, and Q4 are biased nearly to supply, minus the voltage drop of the bias resistor. That sets the emitter at a higher voltage as well, which means a rather large current draw. What it also means is very assymetric clipping. On the buffers positive peaks will hit supply, or near, rather quickly; either way adds nice distortion. Negative waves have more space in the linear portion, so they are cleaner.

A very interesting setup. Thanks for putting it together.

[midwayfair]

Waveshapeillussions, thanks so much for the explanation on the q3 bypass!  It's good to know I can count on folks around here to tell me what I've done when something interesting happens on the breadboard.

[drolo]

That sounds nice
A pitty i always need to wait until i'm home to see your schematics, google docs are blocked at my work ....

The part in the demo with humbuckers sounded very Mick Ronson-ish :-)

[pinkjimiphoton]

awesome as usual jon!! can't wait to try it out here.
glad my toy helped you design this....it is a neat trick, it really makes the toneblaster touch sensitive. insanely touch sensitive.

[midwayfair]

glad my toy helped you design this

I think the fuzzy goodness bug is contagious!

[pinkjimiphoton]

i think so... the emitter trick with a diode is sick.
i'm thinking maybe it's an area worthy of more explanation. it definitely gives touch sensitivity... when it opens the valve, the emitter conducts and you get full gain.. you can ride that breakover point and go from clean to mean to scream in any order by dynamics alone.


i'd imagine that factor could be tweaked even more by adding a small value pot to set a minimum resistance to ground.. you could probably dial in the touch sensitivity that way i'd imagine.