Muff Fuzz circuit questions

[eightsevenzero]

I recently built this Muff Fuzz which works perfectly but it doesn't hit unity gain until about 90% of the volume knob which makes having the knob just about pointless. I know the clipping diodes are cutting all the volume but I'm wondering if there's a way to get more gain out of this somehow or worst case a way to make the volume knob more meaningful in its control so that it hits unity lower in the adjustment? Thanks!

[amptramp]

1. You could add an extra gain stage to the output.
2. You could swap out the silicon diodes for LED's.  Visible LED's may have too high a voltage for your use so infrared LED's of the type used in optoisolators might be a better choice.  That does bring up an interesting circuit possibility - optoisolators that have the opto transistor output connected across the output to ground.  This might give you a harder clipping action.

[eightsevenzero]

I did happen to have a set of infrared 3mm led's but the clipping still wasn't enough to drive it into light fuzz territory. Without clipping diodes installed it hits unity at about 45% on the knob so it seems like everything is wired correctly. Would it be normal for a circuit of this type to have the volume knob fully cranked essentially at all times? Just seems odd to me.

[Mark Hammer]

1) The original diagram eightsevenzero shows lists germanium diodes.  The original used silicon 1N914-type diodes.  It will make a big difference to output level, roughly doubling it.

2) The Muff Fuzz is essentially a silicon Fuzz Face, with the gain fixed.  Stock, it is really more of an overdrive than fuzz, with the diodes simply adding a bit of "supplementary hair" when you hit the strings hard.  HOWEVER, the Fuzz Face is a classic "fuzz" design/engine that can be easily made fuzzier than the stock Muff Fuzz.  I did this to my own commercial Muff Fuzz, and it worked like a charm.

The 2K7 emitter resistor on Q2 is largely responsible for setting the gain.  Bypassing it with a variable resistance and electrolytic cap large enough to impact the passband of interest will increase the gain substantially.  I will recommend connecting a 100R fixed resistor to the wiper of a 10k pot, and connecting a 10uf-22uf cap from one of the outside lugs to ground.  As the total resistance of that pot and fixed resistor is reduced, gain will be increased.  The cap value sets how low in the spectrum that extra gain is applied.  Trust me, there is serious sizzle (and accompanying volume) available from the circuit by doing this.  You probably won't even need the diode pair any more, although I suppose they will keep the volume level sensible.

3) If you want some extra "tricks", stick a 150k resistor in series with the 100k that normally connects Q2 emitter with Q1 base, and use a SPST toggle to bridge it if you want to go back to stock.  It adds some glitchiness and "zipper" tones, and interacts with the gain setting.  The idea came from ZVex.  It's a very good idea.

[eightsevenzero]

Thanks Mark! One last question, it's my (limited) understanding that generally, hard clipping diodes are responsible for limiting the overall volume of a pedal like this. So in theory, if I were to make some of those changes to get more gain and kept the diodes in place, would I still be running into this issue with unity occurring towards the end of the volume knob? In other words, no matter what I set the gain at upstream, will the clipping diodes always clip the volume back down to the same level?

[antonis]

If recovery stage is placed after clipping configuration, Volume should be raised according to diode forward voltage drop x last stage gain..

[Mark Hammer]

Thanks Mark! One last question, it's my (limited) understanding that generally, hard clipping diodes are responsible for limiting the overall volume of a pedal like this. So in theory, if I were to make some of those changes to get more gain and kept the diodes in place, would I still be running into this issue with unity occurring towards the end of the volume knob? In other words, no matter what I set the gain at upstream, will the clipping diodes always clip the volume back down to the same level?

They will, but I think you'll find that using silicon diodes will raise your maximum possible output from around +/-300mv to around +/-550mv.  Trust me, that's a whole lotta signal.  I just brought mine up from the basement to see what's available, and at medium gain (which is a bit hotter than stock) I get equal bypass/effect volume with the volume knob set to 9:00.  That leaves from 9:00 to 5:00 to get even louder.  And since I installed a passive treble-cut between the diodes and volume pot, that eats up some of the signal, so without it, I'd probably get bypass/effect level-balance at around the 8:30 setting on the volume pot.

In short, do the suggested mods, and I seriously doubt you'd have any problems in getting a noticeable volume boost at pretty much ANY gain settings.

[pacealot]

It does seem worth noting that the output change between germanium and silicon diodes is in line with many of the "common" circuits, which is to say that many of the circuits which use germanium clippers also have a make-up gain stage afterwards (some to cope with a passive tone cut circuit as well, but several just to bring the overall output level back up to above unity).

You could conceivably go either way with it — if doing Mark's mods and switching to silicon diodes changes the tone too much from what you like, you could keep the Ge's in and add an LPB-type gain stage after instead...

[eightsevenzero]

Ended up changing diodes to 914's and adding cap from Q2 Emitter to ground and adjusting the 2K7 to around 2K which sounded best by ear. Didn't want an adjustable pot. This introduced "dolphin noise" oscillation so I started playing around a bit more and found if I added a 50K resistor between Q2 base and emitter went away completely without affecting much of the gain. I'm not exactly sure of the science behind what all I did but it's louder and has way more pleasing distortion/light fuzz. I think unity gain hits around 1 o'clock on the volume knob now.

[Mark Hammer]

I'm a little mystified by the emitter-base resistor, but hey, success is success.  And you have volume boost to spare.