How to make a distortion into a fuzz?

[ReeceAblaze]

Hello everyone,

Recently I was breadboarding and came up with a 3 transistor overdrive that I quite liked. Although I liked it as an overdrive I am not a big user of overdrive and tend to lean hard towards distortion and fuzz tones. So, I set out to change it into a distortion and I added in some diodes that hard clip to ground after stage 2 and now its just a distortion. What else could be done to make this a fuzz. Is it as simple as changing the filtering on the input to allow more bass?

Bonus Question: I would also like to add a voltage sag to the power rail for the pedal to get some spitty sounds. How could I do this?

[iainpunk]

hey, to answer this question, we need a schematic.

but basically, change the filters to allow for more bass and up the gain a little.
if you want spitty goodness, there are a few options, but most of those options are based on a potentiometer of some value placed between the V+ rail and the actual power source, but i think that you can probably get away with a single bias pot for one of the stages, like in this fuzz face, but only on one of the transistors.
https://www.electrosmash.com/images/tech/germanium-fuzz-face/Germanium-Fuzz-Face-Schematic.png

[ReeceAblaze]

hey, to answer this question, we need a schematic.

but basically, change the filters to allow for more bass and up the gain a little.
if you want spitty goodness, there are a few options, but most of those options are based on a potentiometer of some value placed between the V+ rail and the actual power source, but i think that you can probably get away with a single bias pot for one of the stages, like in this fuzz face, but only on one of the transistors.
https://www.electrosmash.com/images/tech/germanium-fuzz-face/Germanium-Fuzz-Face-Schematic.png

I will post a schematic once I have one. i need to update the original OD schematic with the clipping diodes and resistor changes.

And the bias control sounds like a good idea. I will give it a go and see what happens. Should there be a minimum value set to avoid the pedal just not working?

[Mark Hammer]

I take the somewhat unorthodox view that the boundaries between overdrive, distortion, and fuzz are largely a function of the period of the note's lifespan during which it reaches and exceeds some critical threshold that results in consistent (as opposed to sporadic) clipping.  That can be achieved by one or both of two strategies: upping the gain, such that the signal almost always exceeds the clipping threshold, and/or reducing the clipping threshold.  To a lesser extent, we equate "distortion" with more treble/harmonic content than is associated with overdrive (which we tend to think of as "warm").  Fuzz is sometimes every bit as stinging as distortion, but sometimes not.  Robert Fripp often used a fuzz for a round sustaining tone.  Again, that goes back to the principle of keeping the signal above a clipping threshold for as long as possible, such that the tone does not change substantially, while one is holding the note.  But in that case, it's not necessarily a stinging strident tone.

So, what does that suggest you do?  As Iain suggests, we probably need to see what the basic circuit is that you wish to convert.  But even without it, we can say that goosing the level of the input signal, will often nudge an overdrive or distortion further into fuzz territory.

[ReeceAblaze]

I take the somewhat unorthodox view that the boundaries between overdrive, distortion, and fuzz are largely a function of the period of the note's lifespan during which it reaches and exceeds some critical threshold that results in consistent (as opposed to sporadic) clipping.  That can be achieved by one or both of two strategies: upping the gain, such that the signal almost always exceeds the clipping threshold, and/or reducing the clipping threshold.  To a lesser extent, we equate "distortion" with more treble/harmonic content than is associated with overdrive (which we tend to think of as "warm").  Fuzz is sometimes every bit as stinging as distortion, but sometimes not.  Robert Fripp often used a fuzz for a round sustaining tone.  Again, that goes back to the principle of keeping the signal above a clipping threshold for as long as possible, such that the tone does not change substantially, while one is holding the note.  But in that case, it's not necessarily a stinging strident tone.

So, what does that suggest you do?  As Iain suggests, we probably need to see what the basic circuit is that you wish to convert.  But even without it, we can say that goosing the level of the input signal, will often nudge an overdrive or distortion further into fuzz territory.

I have a schematic in KiCad, sorry it is a bit messy atm. The reason for r4 being where it is is because there was a voltage divider gain control above it originally and 5k was to set the minimum signal that would pass through. The diodes are 1n4148's.

[ReeceAblaze]

we can say that goosing the level of the input signal, will often nudge an overdrive or distortion further into fuzz territory.

In my schematic how would I goose the level of the input signal?

[Mark Hammer]

Stick a 10uf cap between the emitter of Q2 and ground.  Increasing the value of C3 ought to help as well.

[ReeceAblaze]

Stick a 10uf cap between the emitter of Q2 and ground.  Increasing the value of C3 ought to help as well.

At the time I was using caps that I had. Once I have these values available I will make the changes. Could you explain to me what the 10uF cap would be doing? Also where would I be able to implement a gain control? Would from the emitter of Q2 be a good place like on the fuzz face?

[Karlsbad_kontradiction]

To a lesser extent, we equate "distortion" with more treble/harmonic content than is associated with overdrive (which we tend to think of as "warm").

Somehow I've got it into my head that fuzz is "technically" when the clipping happens as a part of the feedback network of an opamp or a transistor. I have no idea where I got this from, but in ReeceAblaze's schematic, you'd have the diode pair running from the collector of Q2 to the base. That would make it a soft-clipping "fuzz" as opposed to the hard-clipping "distortion." Or with an opamp in the proco rat-esque non-inverting topology you'd run them from the 741's output back to the (-) input. This ends up with a "softer" clipping that doesn't get quite so brutally aggressive when you crank it up. I would assume there's also a difference in the filtering too (though I haven't fully wrapped my head around how it works in the dynamic of said feedback network). Would this be a false distinction in your view?

EDIT: changed "emitter of Q2" to "collector"

[ReeceAblaze]

To a lesser extent, we equate "distortion" with more treble/harmonic content than is associated with overdrive (which we tend to think of as "warm").

Somehow I've got it into my head that fuzz is "technically" when the clipping happens as a part of the feedback network of an opamp or a transistor. I have no idea where I got this from, but in ReeceAblaze's schematic, you'd have the diode pair running from the emitter of Q2 to the base. That would make it a soft-clipping "fuzz" as opposed to the hard-clipping "distortion." Or with an opamp in the proco rat-esque non-inverting topology you'd run them from the 741's output back to the (-) input. This ends up with a "softer" clipping that doesn't get quite so brutally aggressive when you crank it up. I would assume there's also a difference in the filtering too (though I haven't fully wrapped my head around how it works in the dynamic of said feedback network). Would this be a false distinction in your view?

I assume fuzz can be made with hard clipping to ground also? It might be harsher but i guess it depends on the fuzz sound you are after. I am yet to make this into a fuzz but when I get a chance I'll see if I can get it to respond like a fuzz rather than a distortion.

[anotherjim]

Your scheme needs some resistance from Q2 collector in series to the following stages. As it is, when the clipping diodes conduct, you have a short circuit to ground. Similarly, when the tone pot is minimum resistance, high frequencies see a short to ground via C5. The additional resistor should be at least x5 the collector resistor (can't read the value) but x10 is preferable. R8 can probably be replaced by a wire.

[ReeceAblaze]

Your scheme needs some resistance from Q2 collector in series to the following stages. As it is, when the clipping diodes conduct, you have a short circuit to ground. Similarly, when the tone pot is minimum resistance, high frequencies see a short to ground via C5. The additional resistor should be at least x5 the collector resistor (can't read the value) but x10 is preferable. R8 can probably be replaced by a wire.

I will replace r8 with a wire. I did that in the first stage which originally had the same value emitter resistor. Do you know where I could add a gain control?

Edit: Would the resistor from Q2 to the next stage be before or after the capacitor?

[Mark Hammer]

Stick a 10uf cap between the emitter of Q2 and ground.  Increasing the value of C3 ought to help as well.

At the time I was using caps that I had. Once I have these values available I will make the changes. Could you explain to me what the 10uF cap would be doing? Also where would I be able to implement a gain control? Would from the emitter of Q2 be a good place like on the fuzz face?

Providing a path for AC from the emitter will raise the gain.  The cap I suggested will do that.  Just where in the AC spectrum that increased gain will start, depends on the value of the cap.  The smaller the value, the higher the frequency where the boost begins.  The suggested 10uf value is an educated guess about what will work nicely, based on many other gain circuits.  Connecting the cap directly to the emitter would max the gain, within the context of its current biasing components.  Inserting a variable resistance between the emitter and that cap would vary the gain from that maximum to whatever the gain is without the added cap.

[ReeceAblaze]

Stick a 10uf cap between the emitter of Q2 and ground.  Increasing the value of C3 ought to help as well.

At the time I was using caps that I had. Once I have these values available I will make the changes. Could you explain to me what the 10uF cap would be doing? Also where would I be able to implement a gain control? Would from the emitter of Q2 be a good place like on the fuzz face?

Providing a path for AC from the emitter will raise the gain.  The cap I suggested will do that.  Just where in the AC spectrum that increased gain will start, depends on the value of the cap.  The smaller the value, the higher the frequency where the boost begins.  The suggested 10uf value is an educated guess about what will work nicely, based on many other gain circuits.  Connecting the cap directly to the emitter would max the gain, within the context of its current biasing components.  Inserting a variable resistance between the emitter and that cap would vary the gain from that maximum to whatever the gain is without the added cap.

Does the capacitor connect to the middle pin of the pot then like on the fuzz face?

[Mark Hammer]

Nah.  You can just wire it up like Jack Orman does here:

[Steben]

To a lesser extent, we equate "distortion" with more treble/harmonic content than is associated with overdrive (which we tend to think of as "warm").

Somehow I've got it into my head that fuzz is "technically" when the clipping happens as a part of the feedback network of an opamp or a transistor. I have no idea where I got this from, but in ReeceAblaze's schematic, you'd have the diode pair running from the emitter of Q2 to the base. That would make it a soft-clipping "fuzz" as opposed to the hard-clipping "distortion." Or with an opamp in the proco rat-esque non-inverting topology you'd run them from the 741's output back to the (-) input. This ends up with a "softer" clipping that doesn't get quite so brutally aggressive when you crank it up. I would assume there's also a difference in the filtering too (though I haven't fully wrapped my head around how it works in the dynamic of said feedback network). Would this be a false distinction in your view?

If one compares all the "fuzz" labeled units there are a lot of differences in circuit. A fuzz face is nowhere near a maestro or a univox.
fuzz to me is very much treble cut before clipping. Just as marshall-esque drive is bass cut before clipping.
Do not underestimate tone and filtering in the character of any drive tone. I once placed a bass cut control in front of a Fuzz Face for someone and he said "damn it is a metal distortion now"
I've thought about a big muff style pre-gain tone filter a long time ago but somehow I've never done it. 
This filter should have less the mid notch as the original BF tone control has.
Duncan's tone stack calculator gives this with the correct values:
LEFT = fuzz
MID = FLAT "blue" tone
RIGHT = (plexi) distortion "brown" tone

[teemuk]

IME, the archetypal "fuzz" tone is result of low frequency intermodulation distortion. If you examine pre and post (distortion) emphasis of typical fuzz effects you commonly encounter two trends: 1. Very little high-pass filtering introduced pre distortion, and 2. some low-pass filtering introduced post distortion. This applies to, for example, Fuzz-Tone, Fuzzface and Fuzzrite designs. How the signal gets distorted (soft vs. hard or symmetric vs. asymmetric clipping, circuit topology, etc.), in my experience, has very little overall importance in this regard.

This is different voicing tecnique in comparison to archetypal "distortion" effect voicing, which introduces plenty of high-pass filtering pre distortion (to reduce low frequency IMD) and some low-pass filtering post distortion (to tame the harshest high frequency upper order harmonics of distortion).

[Karlsbad_kontradiction]

If one compares all the "fuzz" labeled units there are a lot of differences in circuit.

I think I'm actually going to do this, put together a pdf with the distortion stages of all of the major pedals side-by-side so you can actually see the differences between them. It would be (and would have been) super useful to have when learning about these circuits... Or is that a thing that someone has already done? There's already one for amplifier tone-stacks: https://monster.partyhat.co/article/amplifier-tone-stacks/. Is there one for distortions/fuzzes/overdrives?

[anotherjim]

Would the resistor from Q2 to the next stage be before or after the capacitor?

It actually doesn't matter what order the series components go in.

In this circuit, the 741 is driving the clipping diodes. Note the 10k series resistor. There is a fixed tone circuit simply by the added .001uF cap at the end. The 10k is doing 2 jobs, limiting the current through the diodes and making a low pass filter with the capacitor.