Building a Roger Mayer Stone Fuzz

[jgenet99]

I've been researching this pedal and I've attempted to build myself, there's very little info on this pedal (not even that many sound clips really!) but some of the information is contradictory. In particular, the second transistor appears to be biased by two 10k resistors (confirmed from the few gutshots I can find) but a few different sources have indicated that transistor is biased at the traditional 4.5v. Every attempt I've made biases incredibly low (less than 2v), and I've seen references to a few individuals who have had similar experiences, so it would appear that something is off. Does anyone have an actual pedal that they can confirm the voltages? Is there a component that is missing from the schematic that's circulating? From the research I've done, it would appear that it is the same board as the classic fuzz (which definitely doesn't match the schematics circulating) with some components attached to the other side of the board or wired off board. Thanks!

[Kipper4]

Welcome.
It was s customary to post a schematic to which you refer. So we can see and compare.

Is this one of the drawings you refer to.

http://guitarpedalbuilders.blogspot.com/2012/06/roger-mayer-stone-fuzz.html

[jgenet99]

Yes, that's the correct schematic.

[PRR]

> second transistor appears to be biased by two 10k resistors

Not realy bias resistors. This is a 20k load resistor. Basically a FuzzFace with values 2X higher. In any case this circuit delivers MUCH more output than we can want. The Mayer taps the 20k half-way to get less output.

The "bias" could be understood as the collector voltage. You say 4.5V, that plan shows 4.3V, all the same for R&R. The actual bias here will depend on some Vbe and the hFE of Q1, but anywhere near there is about the same, wide differences make different fuzz, there's no "perfect".

[lcv]

If the schematic is  correct, the indicated 4.3 V is plain wrong .
Given the bias arrangement, the voltage on Q2's collector cannot be 4.3V if  Q2's emitter voltage is 0.74 V.
I think  0.74V (across the 2.2K pot)  is the correct one.
That gives about 340uA into Q2's  collector, in turn giving a voltage drop of 20K * 340uA= 6.7V , so you correctly see about 2V to ground.

[jgenet99]

That's what I was coming across myself, it seemed hard to believe that it could be biasing as high as the 4.5v range. I've seen it confirmed in another post on an actual unit, but I had to wonder if that was just being copied from the schematic that was circulating.

I adjusted the resistors to somewhere in the 6k range, and that brings the voltage to the right range and it sounds very good. I've only found two videos online of this unit, and it's hard to hear which sounds closer (although the first sounds underbiased, weird decay, very little sustain). It appears Mayer tweaked the 'classic fuzz', which also uses two 10k resistors, variable 47k on the first transistor, and two AC128s. There are some tweaks that happen off board (if the notes I've read are correct) which includes an offboard resistor (4.7k) and the capacitor on the first transistor. I've found some gutshots of the Stone Fuzz (and the classic fuzz, which uses the same layout on the board), and the components that I can see look to be set up correctly. I suppose a resistor could be added to the back in series to lower one of the values (which would strike me as very odd). At any rate I'd love to get any information I can if someone has an original unit.

There's also a discussion on the final capacitor, but 10n sounds really odd to me...the lowest string has almost no fundamental pitch, just the overtones. I'd be curious to know what the pedal sounds like if anyone has one.

[PRR]

> correctly see about 2V to ground.

Yes. (My abacus says 2.2V; either way, gross asymmetry.)

The 0.1u output cap on the linked plan should be full bass (to 50Hz). 10nFd (0.01uFd) will be severely short bass (to 500Hz).

The 2uFd input cap lets all bass in. This may fart. A smaller value here would shave bass and reduce intermodulation, but this is very much "to taste".

[bool]

Well to me it appears that the upper of the Q2 collector resistors (both 10k in the schem) should be in fact a 10k and a 1k at R2. I made a very quick sim in falstad and there it appears this to be the case (used hfe 300 at Q1 and 450 at Q2, should be representative for BC559B). Everything seems to bias more "correctly" from the get-go.

[jgenet99]

Thanks to everyone that's contributed, this is confirming what I've been seeing all along. I'm still hoping someone can look into a physical unit and piece together the puzzle. I've found two gut shots online, but I only have one I can find now, but you can see that the resistors are both 10k (I also suspected at first that one of them might have been 1k)


I've tried to think of some other way to make this work, but it looks correct.

[PRR]

> a 10k and a 1k at R2

Makes much sense. Signal here is BIG. A 2:1 reduction seems inadequate. FuzzFace often goes 20:1 (8.2k:470r). 10:1 is still "more!", and also a very "human error" when transposing the as-built to a drawing. (It is possible some Factory Builds got the red- and orange-stripe resistors wrong....)

[Rob Strand]

> a 10k and a 1k at R2

Makes much sense.

+1

I've been stung a few time tracing from pics.  For some reason on some pics orange and red look similar.

About 15 yrs ago I got some pics from a Colorsound pedal.   I *know* the bands are orange but they look red (opposite to the case here).  There's a big orange cap and it does look orange.   I've looked at it on 4 or more PCs and I struggle to convince myself the bands are orange!   

In weird cases I mess with colors on the pic, firstly with Gamma correction and Saturation, then more drastic measures.

From the Stone Fuzz pic at least you can see the two resistors aren't the exactly same.


Another funny thing is the pic seems to show a 47kB (linear) pot but the schematic shows a log pot!

[willienillie]

B taper isn't always linear.  I don't know who makes those, they look familiar, but try to find 2k2 or 47k by Alpha or CTS.  Maybe it's a British brand, they do lots of things oddly, don't even know which side of the road to drive on.

[Rob Strand]

B taper isn't always linear.  I don't know who makes those, they look familiar, but try to find 2k2 or 47k by Alpha or CTS.  Maybe it's a British brand, they do lots of things oddly, don't even know which side of the road to drive on.

Very true.  There's a whole heap of odd-ball markings out there.  It's so long since I've seen a pot not marked in the modern way.

EDIT:

So B taper it *is* in fact a log taper.

They are Omeg brand pots "Made in UK".
https://www.omeg.co.uk/wp-content/uploads/2019/07/OW20BU.pdf

[willienillie]

Yeah I think I've seen those in Marshalls.

[Rob Strand]

Interesting,  this one looks like it has 2x10k's as well,
https://reverb.com/item/5365495-roger-mayer-classic-rocket-fx-fuzz-the-blue-one-hendrix-tone

The circuit here looks identical to the Stone Fuzz except it uses Ge transistors.

Clearly the CB cap on Q1 on the Stone Fuzz must be under the board.

Weird thing is the Classic Fuzz pic doesn't match the schematic I have for the Classic Fuzz.
However, the Classic Fuzz schematic has 18K + 1k on the collector of Q2.
From a biasing point of view that aligns with the 10k + 10k parts.

Did Roger Mayer change to 2x10k's at some point?
---------------------------
Part 2

I can't find many pics for the Stone Fuzz.   From what I can see for the *Classic Fuzz*, the early silver units were 1k + 18k and the later Blue (one one Black) have 2x10k.    The other thing I noticed is the Classic Fuzz uses a 10nF output cap, and the later ones at least have a 47k output pot.  So now I'm not 100% convinced about the 100nF cap on the Stone Fuzz; maybe Gottfried couldn't see the part and guessed the 100nF value because of the 47k output pot.

[willienillie]

this one looks like it has 2x10k's as well,

I'm seeing a red stripe (1k) on the resistor in the foreground, the one behind it looks barely more orange.  I think it's down to individual computer monitors and the eyes looking at them.

For the output cap, maybe some had 10n+470k pot (or 500k like the original FFs), while others had 100n+47k pot.  The HPF cutoff frequency is the same, but I guess the output impedance difference would be audible.

[bool]

Seems that the 1k instinct was in fact true... And the 100n/47k combo would at the highest "output impedance" position (i.e. halfway) yield 12k-ish out. imped. within the passband (and 23/24k-ish below the passband, the highest out-of-passband imp. being 47k at "full" position).

Will sound more "present" compared to classic fuzzes.

[Rob Strand]

Well based on the pics I've seen my current bests guesses go against normal builds!

For the Classic Fuzz I'm seeing 47k pots with 10nF output caps, not much doubt about it.   
The Stone Fuzz definitely uses a 47k pot, schematic shows 100nF but there's no pics to back it up.
For Roger Mayer's "The Fuzz Face" I'm seeing a 470k pot, *maybe* 10n cap.

For the resistors. The resistor in the foreground, near the electrolytic,  goes to the power rail.  In the Stone Fuzz pic, to me that resistor looks more like 10k and the one to the collector looks more like 1k.  I suppose the fact that every pic seems to have ambiguous colors possibly means they are in fact the same value.

One thing that Roger Mayer does is increase the Drive pot from 1k to 2k (or 2k2).  So that means the Q2 collector resistor must scale up by a factor of two to keep the same ratio.   I know some of his earlier pedals used 1k + 18k for the output resistors.   So 1k + 18k = 19k   which is close to 10k + 10k = 20k.

So based on all that:
- No doubt about the Stone Fuzz output pot, 47k log.
- We don't really know the Stone Fuzz output cap.   But I would go with 10nF based on the Classic Fuzz.
- The Q2 collector resistor which goes to the power rail is still uncertain.  If I was to draw up a schematic I would not commit a value it could 1k or 10k.   My best guess would be 10k.

A real PITA all round.

[Rob Strand]

Another pic of the Stone Fuzz is here,
(copy and paste the link and don't add http or https)
smash-hits.jugem.jp/?month=201512

Still doesn't work

Try,
smash-hits.jugem.jp   
then navigate year 2015 then december

Use these,
http://img-cdn.jg.jugem.jp/97f/3138217/20151230_1367897.jpg
http://img-cdn.jg.jugem.jp/97f/3138217/20151230_1367898.jpg

[bool]

WTF??? I see a 5K6, 1K, 100K, 47K and the one that could be a 10K or a 1K ??

From the pics it appears that the RED stripe seems a bit wider/bolder than the orange stripe. It is also my impression that cyan-bodied resistors are a PITA to read compared to the beige or bluey-bodied ones ...