Mystery Distortion/fuzz schematic

[VectorHorrific]

Howdy! Vector here, long-time lurker. I encountered a weird schematic for a distortion the other week after doing some deep dives on old GeoCities pages, defunct forums, wayback machine odysseys, etc. I can't remember exactly where I found it, but it was devoid of any context or explanation as to what the builder was trying to accomplish and it appears incomplete. It caught my eye ever since I first saw it and it has been mystifying me since! I have no idea if this topography is already in use in an existing pedal, but I looked and couldn't find anything matching this so I assume it was a unique project and not an attempt at cloning something else.

I thought it might be fun to share my find. Since I have no idea if this builder ever finished and tested it, I feel like it would be a fun diversion to try and finish it off for them and bring this weird little JPEG (and me) some closure.

Here it is:



I recognise bits and pieces of the structure, and it feels to me like like the builder was mashing parts of other circuits together. It also looks like they drew this in microsoft paint or something. In, out, voltage, and ground are all unlabelled but I assume 'in' is left, 'out' is right, power is top left, and everything else is ground. Bizarre. Since there's no part numbers on the schematic here I've included another image below, separating the different parts of the circuit in to chunks to make it easier to point things out (and also because I'm too lazy to annotate the circuit with numbers myself). Here's my edit:



Green:
This is what tipped me off that this was a Frankenstein's fuzz in the first place. The first and last parts of this circuit are unambiguously the input and gain recovery stages of a big muff. I've quite familiar with the circuit, but the values were unfamiliar to me (the MPSA18 in particular I'd never seen in a muff before). I did some googling and it looks like this is more or less the exact values used in the input and gain recovery stages of a V3 BMP. Also of note is that there is no potentiometer before what looks like the gain stage (red section), and that I think the level of distortion was supposed to be controlled by the purple section (more on that in a bit). Also there's some power filtering in the top right, there. That's nice.

Red:
Alright, this bit has me a little stumped. I've never seen two transistors configured like this before. It looks a bit like a Darlington pair, but with an NPN and a PNP sort of smashed together. I have tried googling this but was unable to find a name for it, and can only assume that (like a Darlington pair) it is supposed to get more gain out of two transistors. Since it looks like the muff was a touchstone for this project, perhaps the builder was trying to combine the two gain stages in to one in order to use less parts? If so, why not use a regular Darlington, rather than an NPN/PNP combo? I feel like I'm missing something here...

In any case, assuming those two transistors are intended to work as one, this looks more or less like a bazz fuss, with some gain limiting resistors to possibly soften the signal coming in to this stage in the absence of a proper potentiometer.

Purple:
This is silly. I'm fairly certain the builder took this from a fuzz face variant or something and was expecting to control the gain using a voltage feedback circuit. I feel like if they wanted to do this they should have removed the diode in the red section? They both produce asymmetrical clipping, so using both would probably sound mushy. I feel like only one would do, or maybe the builder was going for mushy? I'd try to stick with the voltage feedback biasing simply because I don't have the heart to remove what looks like the most ridiculous thing on this schematic.

Blue:
This is even more mystifying than the gain stage, and it's fully labelled with values! It's clearly a tonestack of some kind, but I can't find out which one. It's not a muff tonestack or any of its derivatives from what I can see, and it's certainly not a james/baxandall. I have a feeling this was also probably lifted directly from another circuit, but maybe I'm not giving the author enough credit here. I'm genuinely interested in learning more about the origins of this particular tone control.

--

I hope this didn't come across as too judgemental or snide. As bonkers as I found this image it has still captured my imagination and I hope it has brought you all at least some amusement, too. I am interested in finishing the circuit one way or another, and would love to hear some takes on it from others.

Thank you for looking,

[mac]

Red:
Sziklai pair
I used this pair to add a germ pnp to a Bazz Fuss.
Unlike mine, in the schematic the output is taken from the emitter of the PNP transistor.

mac

[antonis]

I think some grounds should be helpfull..

[ElectricDruid]

The tone control is close-but-not-exactly a Fender thing:

http://www.guitarscience.net/tsc/bassman5f6a.htm#RIN=1300&R1=56k&RT=250k&RB=1M&RM=25k&RL=1M&C1=250p&C2=20n&C3=20n&RB_pot=LogA&RM_pot=Linear&RT_pot=Linear

You could tweak the values to see what the one in the fuzz circuit would actually do.

Antonis is right though - the bottom of that lower pot should be grounded, along with most/all of those wandering lines going off towards the bottom of the picture.

[PRR]

Reading a book before the first draft is finished.

[cspar]

The lack of values around the Sziklai pair suggests to me that germanium is involved.

[amptramp]

It is interesting that the Sziklai pair in this case has the same schematic as a silicon controlled recitifier.  You have to keep the current low enough that it doesn't latch up but unless someone plays silly bugger with resistor values, that should not happen.

[mac]

The lack of values around the Sziklai pair suggests to me that germanium is involved.

I recovered this from Aron's new gallery:
The hybrid Si NPN- Ge PNP and the All Ge versions are still missing but he is adding more and more circuits from my old album. 



I'll try the diode from the PNP emitter.

mac

[antonis]

I'll try the diode from the PNP emitter.

You'll turn it into a Bazz Fuss..

[Rob Strand]

I'll try the diode from the PNP emitter.

You could also try two diodes in series with that connection.  One diode really pushes the bias of the first transistor - might still sound OK through.

[mac]

You'll turn it into a Bazz Fuss..

You could also try two diodes in series with that connection.  One diode really pushes the bias of the first transistor - might still sound OK through.

It's from 2007. Most likely I did that but for some reason I chose the other way ... ¿?

mac

[VectorHorrific]

thank you everyone for your responses, especially with providing me with a name for the Sziklai Pair and the tonestack. I've been reading up on both and it gives me an idea of maybe what the original builder was going for here, even if it doesn't make a lot of sense to me.

I agree that if a sziklai is being used then there was probably as pair of low gain germaniums involved. I don't see the sense in pairing two silicons like this unless they wanted a ridiculous boost in volume. From what I've read, pairing germanium transistors in this configuration can add some stability and resistance to temperature changes outside of just increasing the gain, but texture-wise there's probably not much being added. Perhaps I am wrong, or perhaps we would need to know what the original parts intended to be used here were to judge any further.

Also, thank you Tom for the heads up on that tonestack simulator. I never would have guessed it was a bassman! I keyed in the parts from the schematic and got less dramatic results that the default. All pots turned up got a pretty flat response across the spectrum so maybe they were using this to just cut frequencies rather than boost them?

There's still a bit of an identity issue with regards to the first half of the circuit, I think. The bazz fuss-esque nature of the "red" section is fine but usually I see stages set up like with an attenuator after the gain stage (a la big muff). I don't know what the benefits of using a fuzz face style voltage feedback would be as the circuit is now. My feeling it to either swap it out for the attenuator or, if I'm keeping the feedback control, remove the diode.

This will of course turn it in to either a muff-style circuit with a single bazz fuss in the place of two gain stages, or a fuzz face style circuit with a tone control slapped on the end.

I'll have a play around this weekend and see if I can come up with a more coherent draft. Probably with some grounds, too 

[Ben N]

It's a Sziklai variant on a Whisker Biscuit, with some other weirdness like that feedback loop and TMB tonestack.

[VectorHorrific]

looke more in to the tonestack and found this page

https://robrobinette.com/How_The_TMB_Tone_Stack_Works.htm

Values here are the same as they are in our circuit. The extremes in the capacitor values looks to be in service of recreating a characteristic of old blackface amps caused by the slight delay between signals going through the bass and treble caps. I don't know how well that will translate with a distorted signal going through it but it's an interesting read.

[mac]

I don't see the sense in pairing two silicons like this unless they wanted a ridiculous boost in volume.

The difference between a single Si and darlington BF is that the darlington version sounds "in focus". This is because the BF, as old Arsenio Novo's circuit, sounds better when Vc=Vb, that is, there is no idle current through the diode. A single Si needs some base current coming from the diode unless the collector resistor is very, very high. A darlington needs a tiny current to work, so the diode current is almost zero. A 10k works just fine.
In both versions the diode limits the output volume to almost the same level.
To get a volume boost you can try a LED, or two series Si.

... but texture-wise there's probably not much being added

An all Ge BF sounds way better IMHO. But temperature and leakage are a problem. I recovered this one:



mac

[VectorHorrific]

I've gone away and been chipping away at what I think the author of the original circuit was trying to achieve. I'm still part ways through the first draft but I thought i'd just show my current progress and show my working



I decided to boil down the original idea and treat it like a hybrid tone bender: a single high gain silicon amplifier stage, followed by a saturation stage made up of two low gain germanium transistors in the sziklai configuration.

I took a lot of queues from Mac's Eight Up Diet Fuzz for the saturation stage; those circuits were very interesting and really helpful in understanding what I was working with and also unlocking more ideas for the Bazz Fuss! I'm unsure about including the soft clipping usually in the bazz fuss as I feel like we'll get enough distortion without it and I don't want to get too much of a compressed sound.

I also swapped the tone control out for something a little simpler, and placed it before the saturation stage. This placement is personal preference as I like the way you can get some weird peaks and valleys in the way the gain and tone controls interact.

I'm still playing around with values and will check in when I have a satisfying prototype. If anyone has any notes or ideas in the mean time, feel free to add them.

Thanks!

[antonis]

I'm still playing around with values and will check in when I have a satisfying prototype. If anyone has any notes or ideas in the mean time, feel free to add them.

I presume you plagiarize items values from similar circuits..

It should be a good idea for Q2/Q3 Sziklai pair to place a resistor between Q2 Collector and Q3 Emitter for Q2 to work in acceptable current gain range..
(as it is, its Collector current is only the one flowing out of Q3 Base..)
Its value should be set by 300mV (for  Q3 Ge BJT) divided by Q2 desirable current..
(any value between 1k and 4k7 should be OK for start..)

[VectorHorrific]

I presume you plagiarize items values from similar circuits..

I prefer the term "iterate" 

The topography is certainly lifted from other designs, but the values will all need to be changed to work together as a cohesive whole. I can't just slap three or four different circuits together like the original layout and expect it to sound good! 

But thank you for the tip! I'll incorporate it in and see how I get on!

[antonis]

The topography is certainly lifted from other designs, but the values will all need to be changed to work together as a cohesive whole. I can't just slap three or four different circuits together like the original layout and expect it to sound good!

As long as each individual building block works OK, "ensemble" of them should need minor modifications..

[duck_arse]

I think your T1 needs a series stop resistor, just in case.