Tonebender Face Switch + Pot

[bonhampala]

Hi Folks!
So I've been experimenting with a Tonebender/Fuzz Face type of build.
A seemingly common mod to the Tonebender is to add a Switch to jump the Booster stage (Q1), since everything afterwards is basically "just" a Fuzz Face circuit. For those who don't know what I'm talking about:
This explains the mod: http://www.bigtonemusic.com/TBmods.html
For comparison:
Tonebender MKII: http://fuzzcentral.ssguitar.com/mkII.php
Arbiter Fuzz Face: http://fuzzcentral.ssguitar.com/fuzzface.php

First question: The cap in front of Q2 is not polarized in the Tonebender (Film cap), in the Fuzz Face this is the input cap and is polarized (electrolytic cap). By adding the switch and thus bypassing the booster stage in the Tonebender, the non-polarized cap becomes the input cap.
I've already tried this and it seems to work just fine, but somehow my brain is telling me that something's wrong.
The same thing goes for some pedal builders being proud of having replaced every electrolytic cap in the circuit with non-polarized ones to improve the sound. What exactly is up with that?

Second question: My idea is to incorporate said switch into the final version as a footswitch, so I can add/remove boost whenever I want to. But I need some sort of regulation, especially when the Fuzz control is somewhere before noon, the boost just makes it too loud. So I added a third pot right in front of the switch, same configuration as the Volume pot at the end of the circuit (see schematic below).
Any concerns with that? I figured since it's a simple voltage divider it should work, but I'm not too sure about the value. (Though I think with an audio taper the same value as at the end should be a good starting point?)

cheers
Chris

[Cortexturizer]

This interests me.

[raulduke]

First question: The cap in front of Q2 is not polarized in the Tonebender (Film cap), in the Fuzz Face this is the input cap and is polarized (electrolytic cap). By adding the switch and thus bypassing the booster stage in the Tonebender, the non-polarized cap becomes the input cap.
I've already tried this and it seems to work just fine, but somehow my brain is telling me that something's wrong.
The same thing goes for some pedal builders being proud of having replaced every electrolytic cap in the circuit with non-polarized ones to improve the sound. What exactly is up with that?

Second question: My idea is to incorporate said switch into the final version as a footswitch, so I can add/remove boost whenever I want to. But I need some sort of regulation, especially when the Fuzz control is somewhere before noon, the boost just makes it too loud. So I added a third pot right in front of the switch, same configuration as the Volume pot at the end of the circuit (see schematic below).
Any concerns with that? I figured since it's a simple voltage divider it should work, but I'm not too sure about the value. (Though I think with an audio taper the same value as at the end should be a good starting point?)

cheers
Chris

No nothing is wrong. The input cap is polarized mainly due to the value on a Fuzz Face (usually 2u2), rather than any other reason.

The cap effectively forms a high pass filter; higher cap value = lower cutoff frequency = more bass.

2u2 is a far more common (and cheaper) value with polarized capacitors (tantalum and electrolytic) than it is with film. This applied even more so in the 60's. I would imagine the electrolytic was originally chosen due to cost and availability rather than its polarized characteristic.

Your idea in question 2 sounds fine to me. Just remember: If in doubt, get the breadboard out  !

[midwayfair]

since everything afterwards is basically "just" a Fuzz Face circuit

This may be true from a topology perspective, but it's not even remotely true regarding biasing or transistor selection.

Here are just a handful of differences you'll find between a tone bender MKII's Q2 and Q3 vs. a Fuzz Face:

The bias for Q2 in the MKII settles at as little as .1-.2v. Compare that with the .5V on the first transistor in the Fuzz Face. LOTS more saturation.

The proper bias for Q3 in the MKII is ~8V, as measured on numerous vintage units by several builders (Electric Warrior, D*A*M, Luciferstrip, and others). The Fuzz Central article gives the wildly inappropriate bias voltage of 4.5V for this transistor. You might get a decent sounding fuzz by doing that, but it's not going to act or sound like a MKII.

The MKII's biasing does several things: First, it's more saturation than the Fuzz Face, so there's more distortion overall, and the final transistor is closer to cutoff. Second, there is a combination of slight gating, with a slightly "sticky" attack and a more explosive sound, along with some sagging that creates bloom. This comes as much from the biasing as it does from the third transistor stage.

If you make the first stage switchable, you'll end up with a boosted fuzz face. Not a MKII.

I haven't touched on transistor selection yet.

The Fuzz Face wants the least leaky transistors you can find.

The MKII is picky about transistors and requires some leakage in all three transistors to sound right (and for Q1 to work at all).

What else? Oh right, tone shaping. The cutoff frequency of the output on the MKII is TWICE that of a Fuzz Face. (The Fuzz Face's cutoff frequency is 72Hz.) This is actually a huge part of the Tone Bender's sound, since it helps the TB cut through the mix better and tightens up the bass despite the greater amount of distortion compared with a fuzz face.

This really leaves you with a couple results:
-A great sounding fuzz face mode and a mediocre MKII mode
-A great sounding MKII and a mediocre sounding/misbiased (and possibly disfunctional) Fuzz Face mode
-Split the difference on biasing and have two mediocre modes
-Use several switches to swap out all the parts required to make both modes right.

Or ... you could just build two great sounding fuzzes in one enclosure. The circuits are tiny, you could easily get one of each in a 1590B.

I hope it doesn't sound like I'm discouraging you from trying this. But you need to consider ALL aspects of the circuitry when you do something like this, not just focus on a couple superficial changes like value of the input cap.

[Cortexturizer]

You might try doing what guys at Rambe FX done with their Twin Bender, check it out. There's a MK 1.5 and MKII switch, and the difference in tones is big, 1.5 acts like a FF, whereas MKII is much more distorted, as you would expect. I haven't looked into what they had done but I suspect it is something like what you want to get essentially.

[bonhampala]

@midwayfair
First of all that's exactly what I meant by Fuzz Face circuit. Throughout the decades many versions using different transistors and cap/resistor values/types have been put on the market and even more people have adapted it for their own creations. I did not want this to come across as a reference to the original first Fuzz Face, not at all.
Thank you anyway on your input, dissertations like these are always greatly appreciated!
(And by the way this does not at all discourage me, this is food for my brains!)

On your input towards the bias compromise:
The result of course depends greatly on the choice of transistors. For example I've built my first Tonebender MKII "clone" with NPN Germaniums, since I didn't know about voltage inverters back then, later I added the mentioned switch and it works just fine. The PCB I used was from Poodle and he designed it to house two trim pots for Q2 and Q3, which gives additional freedom. This would also be the way for me to go.

By now I've realized that I probably should've given more information on what I'm trying to achieve with this.
I don't want this design to be able to recreate an "original" Fuzz Face or Tonebender Sound. I'd like more low-end from both of them. Not woolly mammoth, but think Supa Fuzz or MJM London Fuzz, more that area.
Talking Gain: I've heard many people refer to the MKII gain characteristic as "starting where a Fuzz Face leaves". This might be true to some extent, but the fact remains that the sound of a full blown Fuzz Face and a lower Set MKII aren't the same at all.
And that's why I got the idea for the pot. The approach would then be to design the Q2-Q3 stage according to plan and then add as much boost as wanted. Using separate trim pots to bias both Q2 and Q3 should give enough freedom to make both sounds work.
If you call this a boosted Fuzz Face or a boosted MK1.5 (which would be a MKII) is not my concern.

At the end the result should be a fuzz with an additional boost footswitch.

Talking transistors: Until now I thought I'd use AC128 for all stages, but to be honest I'm not so sure about that anymore...
What I want is to keep the responsiveness to the guitar volume without the boost, which is why I chose to go with AC128 in the first place and use a 50K trimpot to bias Q2 and a 10K trimpot to bias Q3. (Which would be Q1 and Q2 in a Face layout).
Additional question: Since Q1 (boost stage) is only going to be used as boost and not being saturated, the choice here should not have much of an influence, right?

[bonhampala]

Ok so it's been some time but I finally got something together.
Ordered some beautiful germanium TO-5 can transistors and sorted out some low leakage (all below 100uA) ones with the following gains:
Q1: hfe 88, Q2: hfe 95, Q3: hfe 151, all measured with the Geofex method. (So this is "real" hfe)
Bias pots for Q2 and Q3 were set to 30K and 8K2, which sounded very nice to begin with.

So the circuit after the switch works nice and well, but when I switch on the Q1 stage, things do not work as they should. A lot more background noise and altough the signal itself is more distorted, there's less output.

Now there are two things to consider:
- Bias setting for Q2&Q3, which we already discussed. I haven't tried other settings yet, but maybe there are some that'll work better with the Q1 stage.
- I suspect that another reason for this behaviour is the pot before the switch. Even when it's set on 10, there's still a 250K resistance going to ground. (I used a 250KA for this try)

So my issue is that I don't fully understand this level pot circuit. It just regulates how much signal goes through and how much goes to ground, right? But doesn't signal going to ground cause noise to begin with?
Should I just try a higher value pot or even put an offset resistor between the pot and ground? (There's no use for the zero setting anyway)

Thoughts?