Anyone ever made the Messenger "Tone Messer" fuzz circuit?

[Mark Hammer]

Someone on another forum posted an old Youtube of Grand Funk Railroad, where Mark Farner breaks into a solo and engages the built-in fuzz in his Messenger guitar that the company called  the "Tone Messer".  Naturally, I had to look for the circuit.  I can't vouch for the accuracy of this drawing, but its reliance on 1.5V suggests it is authentic, or close to it.  Any familiarity with it, or simply comments?

[Transistor-Transistor]

Wow this is a really interesting little circuit! I've never seen an inductor tied there in a fuzz before! I wonder how it would affect the sound if you put a cap in parallel to it like a resonant circuit?

[GibsonGM]

Following. I'd love to hear what it sounds like! 

[Mark Hammer]

Well I was prompted to look for it by the Farner solo around the 4:30 mark in this video of "Inside looking out".  We don't get to see anything related to him engaging it, so I have no idea if what we hear is everything dimed or what.  But it's a nice sounding fuzz.

[Transistor-Transistor]

Thats a really really nice sound.

[GibsonGM]

Splatty, a touch octavey - what's not to like?   

[Transistor-Transistor]

Yeah there's so much gain with such little voltage. Makes me wonder how much farther it can be pushed 

[Rob Strand]

I can't vouch for the accuracy of this drawing, but its reliance on 1.5V suggests it is authentic, or close to it.  Any familiarity with it, or simply comments?

That schematic has been messed with.

There was a thread some years back.

I pieced together this bundle.  Looks close but I may have more notes elsewhere.
The inductors vs diode is still unresolved (I favor the inductor as the part in the pic has sharp edges which you don't get on copper colored diodes).



Here's some pics as evidence of the original:







These 1.5V circuits are very fickle in terms of biasing.  Perhaps tweak R2 and R5 to suite the transistors.  As a starting point try 0.75V to 1V on the collectors and take it from there.  Perhaps running at 3V is a little less troublesome but set the collector bias higher.
A better pic showing the inductor/RF choke more clearly,

[R.G.]

These 1.5V circuits are very fickle in terms of biasing.   Perhaps tweak R2 and R5 to suite the transistors.  As a starting point try 0.75V to 1V on the collectors and take it from there.   Perhaps running at 3V is a little less troublesome but set the collector bias higher.

At this close to saturation, the bipolars could well have very odd nonlinearities. In a circuit like this, I might have to re-consider my mantra that any NPN will do.

If the sound is good enough to worry about the circuit always just working, I might be tempted to add on some servo opamps to keep it where I liked the sound.

[Transistor-Transistor]

Yeah I feel like the BC108's pictured in Mark's schematic would be perfect because of how well they work with low voltage.

[Rob Strand]

Yeah I feel like the BC108's pictured in Mark's schematic would be perfect because of how well they work with low voltage.

Given the vintage of the pedal I suspect the original transitors were low-ish gains, (as far as modern parts go...) maybe 2N3904's and maybe even 2N2222's.

The circuit bias current are quite low and that will reduce the gain of the transistor, especially a good proportion of vintage transistors.

The biasing of The first stage seems to be quite robust to hFE variations.   The bias voltage seems on high side though maybe 1.2V to 1.3V.  A higher gain transistor could drop that down.

The second stage seems to need a bit more care.  The bias perhaps favoring low gain devices.   High gain devices seem to bias the collector too low.    Whatever the case it seems the Q2 collector voltage will be lower than the Q1 collector voltage.

[amptramp]

This reminds me of the Heathkit fuzz that ran off 1.5 volts from a penlight cell.  Are low voltage circuits coming back into popularity?

[Mark Hammer]

I can't vouch for the accuracy of this drawing, but its reliance on 1.5V suggests it is authentic, or close to it.  Any familiarity with it, or simply comments?

That schematic has been messed with.

There was a thread some years back.

I pieced together this bundle.  Looks close but I may have more notes elsewhere.
The inductors vs diode is still unresolved (I favor the inductor as the part in the pic has sharp edges which you don't get on copper colored diodes).



Here's some pics as evidence of the original:







These 1.5V circuits are very fickle in terms of biasing.  Perhaps tweak R2 and R5 to suite the transistors.  As a starting point try 0.75V to 1V on the collectors and take it from there.  Perhaps running at 3V is a little less troublesome but set the collector bias higher.
A better pic showing the inductor/RF choke more clearly,

Many thanks for this Rob.  Always good to have photos of the actual board.
I started browsing through my directory of non-commercial distortions and fuzzes to see if there was anything resembling it, but gave up after a few hundred.  I hadn't realized that the directory had over 1500 files that folks have posted here over the years, many from folks we miss.  A fair share are op-amp based, but there are literally hundreds that are 2 and 3-transistor circuits.

[Transistor-Transistor]

Are low voltage circuits coming back into popularity?

I hope so! My favorite fuzz is the 3 volt selmer buzztone. You can get such fun sounds from low voltage stuff in my opinion

[Mark Hammer]

And let us be clear, this is NOT the same as a "dying battery".  If anything a single AA cell has more juice at 1.5V than a great many multi-cell 9V batteries would.

[amz-fx]

It doesn't seem like the small inductor is going to do very much. Its reactance is only a few ohms in the guitar frequency range. I would put a 4.7 or 10 ohm resistor in its place and try that.

regards, Jack

[Rob Strand]

Many thanks for this Rob.  Always good to have photos of the actual board.
I started browsing through my directory of non-commercial distortions and fuzzes to see if there was anything resembling it, but gave up after a few hundred.  I hadn't realized that the directory had over 1500 files that folks have posted here over the years, many from folks we miss.  A fair share are op-amp based, but there are literally hundreds that are 2 and 3-transistor circuits.

No problem.

I've got a list of 1.5V and 3V pedals on my old machine.  There's a few.

I remembered this old thread with the JOSTY unit.   The base resistors make the biasing a little less predictable than the Messenger.   The collector resistors are smaller which biases the transistors at a bit more favorable current but when you do that you need lower base resistors and then there's a trade-off loading the pickups.

https://www.diystompboxes.com/smfforum/index.php?topic=126307.0

It doesn't seem like the small inductor is going to do very much. Its reactance is only a few ohms in the guitar frequency range. I would put a 4.7 or 10 ohm resistor in its place and try that.

Agreed. It's definitely there as a precaution for RF rejection and not for audio.  The collector current is low so the transistor's 're' is about 1k ohm and is *much* higher than 10 ohm so you might as well just use a short.

[Rob Strand]

I've collated a bit more info:



The sensitivity of the bias voltages to various circuit parameters is pretty bad.
Vintage data:

[PRR]

The sensitivity of the bias voltages to various circuit parameters is pretty bad.

A nifty tool which seems to be forgotten is the Shea Stability Factor. (Google returns much crap not related to Mr Shea or to transistors.) Mis-stated: The base resistor divided by the emitter resistor, compared to hFE.

Here at Q1 we have 30,000 to say 10, which is WAY more than hFE. At Q2, 60,000 to 47, 1,276:1, out of site.

It's not "wrong" for amateurs in 1967, but it is wildly biased.

Q1 may be more critical. If Q1 is anywhere near a good bias, its output will be so big that Q2 will fuzz like a wet cat.
----------
Try "shea stability factor transistor circuit"

[Rob Strand]

A nifty tool which seems to be forgotten is the Shea Stability Factor. (Google returns much crap not related to Mr Shea or to transistors.) Mis-stated: The base resistor divided by the emitter resistor, compared to hFE.

Stability/Sensitivity for/to hFE changes,
https://www.gpangul.org/pdf/oc/Notes_of_Analog_Electronics/CHAPTER_5.pdf

The circuit is a good example where blind rules of thumb cause trouble.

A common rule of thumb is to pass much more current down the base voltage divider than the base current and that reduces the (direct) sensitivity to hFE, which it does.   However without an *emitter resistor* it makes the circuit very sensitive to the supply voltage, the base resistors and the
specific VBE of the transistor.  That circuit does pass a lot more current down the base dividers.  The emitter resistor on Q2 is too small to help biasing.

If we look at Q1, then assuming Vcc=1.5V, the Thevenin equivalent circuit of the base divider is, 
Vt1 = Vcc*R3/(R3+R2) = 1.5*47k / (82k+47k) = 0.547V and Rt1 = 82k // 47k = 30k

https://www.youtube.com/watch?v=1CG1lIJddA4

The base current is Ib1 = (Vt1 - Vbe1) / Rt1.   We can see already assuming an approximate VBE=0.6V will produce a negative base current.  We need a precise value for the Vbe.

Suppose for the nominal case we work backwards:   Lets assume the collector is biased to Vc1=1.1V.   The Q1 collector current is then IC1 = (1.5-1.1)/22k = 18.2uA.  If hFE1 = 100 at low currents then IB1 = 18.2uA / 100 = 182nA.
So we require Vbe1 = Vt1 - Ib1 * Rt1 = 0.547 - 182nA * 30k = 0.547 - 0.0055V = 547mV - 5.5mV = 541.5mV.

The thing to note is by using low valued base resistors it helps reduce variation due hFE, ie. the rule of thumb, however, the base small current is determine by a very small voltage drop (5.5mV) across the Thevenin resistance.   It also depends on the transistor VBE.  The VBE varies from unit to unit, +/-20mV wouldn't be a bad estimate, which means the base current could vary by a factor of 5 due to VBE.

If the supply voltage was 1.55V  the Thevenin voltage feeding the base is 1.55*47k/(82k + 47k) = 0.565V, ie. 565mV-547mV =  18mV higher than the nominal design voltage.   In this case the base current varies by a factor of 4.

We also have problems choosing the base resistors precisely enough.   Suppose R2 was changed to 68k, the next E12 size down from 82k, the Thevenin equivalent voltage is 1.5 * 47 / (68 + 47) = 0.613V ie.  613mV-547mV = 66mV.  So one step in resistor value causes 66/5.5 = 12 times the base current.   In practice we would consider the tolerance of R2 but there is also an issue choosing the right nominal value to start with.

Virtually every aspect of the circuit can upset the biasing by a significant factor.

If we add an emitter resistor to Q1 say 2.2k and a bypass cap say 47uF to 100uF in parallel with it, then it does help quite a bit.  Although it doesn't quite help the supply voltage dependency enough to flatten the battery.  (Similarly for Q2 we can use a bypassed emitter resistor of 1k.)   Making the emitter resistors larger helps but we only have a 1.5V supply so we don't want too much drop across the emitter resistor.   There's other approaches.