I can get no satisfaction (from my Fuzz-Tone Clone)

[alderbody]

I built a Maestro Fuzz-Tone (FZ-1A) clone with original spec 2N2614 PNP Ge transistors (Small Bear).

This circuit runs with a single AA 1.5V battery.

C1 is a Nichiden Audio cap, C2 and C3 are axial lead electros and C4 is an Orange drop.
Pots are Alpha 24mm and resistors are all 1/2Watt carbon film except R1(C.comp).



The problem is that the sound i get is very strange.

The notes have an audible attack (distorted as it should be), but then there is a very fast decay and the signal drops to zero,
as if someone cuts off the power...

The battery doesn't seem to be dead or dying (1493mV), but the effect gets worse as time passes.
I mean, as soon as the power gets connected the sound is a bit better (still not good),
and the notes can be heard for a little longer.

Shortly after power up, the effect gets worse.
In idle, the output is zero and even the 60Hz hum cannot be heard (talking about noiseless!... )
The signal seems to "wake up" only when the guitar is played and then it falls back to the "dead zone"... 

Any ideas about what might be happening?

EDIT: The transistors are put on bases to avoid any accidental overheating during soldering...

[davebungo]

I have to say that I don't like the look of this circuit, but I don't know it so I may be a little out of line.  The biasing looks like it would never function.  Q2 and Q3 VBEs don't appear to be anything near correct.

Is the battery connected the right way around?  (All your voltages are shown positive rather than negative but you may have simply omitted the "-" sign).

There is no decoupling cap shown, should there be one?  I would never run a circuit without one.

I'm a little surprised about the circuit running off a 1.5V battery, but again, I don't know this circuit.  It does not give you much headroom.  What's wrong with a 9V battery?

[alderbody]

I have to say that I don't like the look of this circuit, but I don't know it so I may be a little out of line.  The biasing looks like it would never function.  Q2 and Q3 VBEs don't appear to be anything near correct.

Is the battery connected the right way around?  (All your voltages are shown positive rather than negative but you may have simply omitted the "-" sign)

. The battery is connected the right way (positive cround). The voltages are shown the way they appear on the DMM's display.

There is no decoupling cap shown, should there be one?  I would never run a circuit without one.

That's a copy of the original schematic from Maestro. It should work like that.

I'm a little surprised about the circuit running off a 1.5V battery, but again, I don't know this circuit.  It does not give you much headroom.  What's wrong with a 9V battery?

I guess you should ask Gibson about it... That's the way they designed it back in the 60's.

[alderbody]

The biasing looks like it would never function.  Q2 and Q3 VBEs don't appear to be anything near correct.

Do you think that C2 and C3 might be bad or something?

What if i replaced them with NP's or film caps?

[davebungo]

So how have you got your DMM connected?  Is it black lead to gnd?  All voltages should appear negative in this case.

[alderbody]

So how have you got your DMM connected?  Is it black lead to gnd?  All voltages should appear negative in this case.

yep!

[gez]

Have you read my notes on this circuit?

Search the archives for this circuit, I've posted in numerous threads on how to get the best from it.

I can't even remember the proper satisfaction riff, and I don't own a Les Paul, but even my Jazzmaster sounds similar thru this unit:

http://hometown.aol.co.uk/Gezpaton/Fuzz+Tone+II.MP3

http://hometown.aol.co.uk/Gezpaton/Fuzz+Tone+1.MP3

[davebungo]

Hi Gez,
could you point us to your notes as this circuit has me a little perplexed (I have done a quick search but there is a lot to trawl through)?  Is the biasing signal dependent?  What I mean is, does the circuit rely on a signal to provide bias?

[alderbody]

I think i found some of gez's notes (about leakage and stuff, right?)

If it reqiures a lot of transistor testing i think i'm doomed already. I just bought three and that's all...

Any other types of Ge or Si that work in there?

[davebungo]

Q3 can't be very leaky (unless that is actually what the circuit depends on!)  Only 3.5uA flowing into the emitter.  A circuit description is what we need here.

[smnm]

I don't know if this is helpful, but my original FZ-1A sounds pretty much as you describe with most of my cheap 60s gtrs, unless I use a higher output pickup (Hagstrom, Supro or Burns Tri Sonics work for me).
After much research I decided it's meant to be that way - remember it wasn't designed to make the Gary Hurst kind of '66 saturated fuzz...

I swapped out the caps, but it made no difference, so I put them back. Didn't test the tranny though.

Have you listened to Satisfaction lately? - IMO the part sounds pretty crappy anyway! Love it and everything, but still... (btw consensus seems to be it was a Tele, not LP, on this)

It sounded better with 3V (on the 1.5V circuit), but I couldn't fit both in the housing.

S

[alderbody]

Q3 can't be very leaky (unless that is actually what the circuit depends on!)  Only 3.5uA flowing into the emitter.  A circuit description is what we need here.

gez says in his notes that Q3 could be the one with the most leakage...



...very tricky circuit....

[alderbody]

I don't know if this is helpful, but my original FZ-1A sounds pretty much as you describe with most of my cheap 60s gtrs, unless I use a higher output pickup (Hagstrom, Supro or Burns Tri Sonics work for me).
After much research I decided it's meant to be that way - remember it wasn't designed to make the Gary Hurst kind of '66 saturated fuzz...

I swapped out the caps, but it made no difference, so I put them back. Didn't test the tranny though.

Have you listened to Satisfaction lately? - IMO the part sounds pretty crappy anyway! Love it and everything, but still... (btw consensus seems to be it was a Tele, not LP, on this)

It sounded better with 3V (on the 1.5V circuit), but I couldn't fit both in the housing.

S

The 3Volt trick came to my mind but i didn't try it. Now i will.
I'll also try it with a humbucker loaded guitar.

thanks anyway

[Quackzed]

i would try lowering the values of r7 and/or r9. i built some fuzzez with a 100k bias pot to get misbiased sounds and the signal dieing quickly or gating out as i call it is usually a consequence of turning up the bias pot too high or having the bias resistor's resistance too high...  
maybee not but it would be easy enough to take a resistor of equal value of r7 and/or r9 and just touch it (paralell) to (around) r7 or r9 and see if it sounds better with half the resistance...ie 2 equal paralell resistors...
  btw, i always use a trimpot or pot to bias fuzzes, transistors are too inconsistant and fuzzes are too sensitive... it makes life much easier and you can just measure the pot and sub a resistor of close to that value... also there are lots of different tones and attack charicteristics to be had depending on how you bias so maybee try a trimmer or a pot and i bet you could get it to behave better if not perfectly..

[alderbody]

i would try lowering the values of r7 and/or r9. i built some fuzzez with a 100k bias pot to get misbiased sounds and the signal dieing quickly or gating out as i call it is usually a consequence of turning up the bias pot too high or having the bias resistor's resistance too high...  
maybee not but it would be easy enough to take a resistor of equal value of r7 and/or r9 and just touch it (paralell) to (around) r7 or r9 and see if it sounds better with half the resistance...ie 2 equal paralell resistors...
  btw, i always use a trimpot or pot to bias fuzzes, transistors are too inconsistant and fuzzes are too sensitive... it makes life much easier and you can just measure the pot and sub a resistor of close to that value... also there are lots of different tones and attack charicteristics to be had depending on how you bias so maybee try a trimmer or a pot and i bet you could get it to behave better if not perfectly..

thanks Quackzed, i'll try that too!

[smnm]

... take a resistor of equal value of r7 and/or r9 and just touch it (paralell) to (around) r7 or r9 and see if it sounds better with half the resistance...ie 2 equal paralell resistors...

I'll try this trick on my old one too! the resistors have probably drifted off spec, I read somewhere that carbon comp resistors increase in value a they age (?)

alderbody - let us know what happens with yours...

[gez]

From the archives:

"I've built a few of these, all 1a (the ones with the 1.5V supply). I tried out a lot of different transistors and it didn't seem to make much difference so long as they're Ge.

What does make a difference (more than gain in my opinion) is the leakage of each transistor, as it's leakage which biases each device into conduction. I didn't test specifically for collector-base leakage, instead I used RG's circuit as outlined in his fuzz face article to test for both gain and leakage.

For Q1 I'd recommend using a low leakage device (the ones giving best results measured between 58uA and 120uA in my tests). If the transistor is too leaky virtually all the supply appears across the emitter resistor, clipping the signal on +ve swings resulting in a constricted, muffled sound. If you're too lazy to do all that measuring just stick a trannie in there and if the emitter measures somewhere between 0.4 and 0.6V it should sound sweet.

For Q2 you need a slightly higher leakage (my results were between 100uA and 220uA). If leakage is too low the transistor biases towards cut-off, collector current is too low and gain is pitiful, resulting in hardly any fuzz. If leakage is too high it biases towards saturation and the fuzz sounds constricted again. There's a definite 'sweet spot' for collector voltages between 0.75 and 1V with the 'attack' pot set for max fuzz.

For Q3 you can get away with higher leakages (188uA - 367uA for my results). Best results were to be had with the collector measuring between 1.25 and 1.28V. This is the only stage where I found gain made any difference (though only slight) - the higher the better.

Wear rubber gloves when handling the trannies as finger heat will totally distort all measurements. It's best to wait a few minutes to allow the circuit to settle in before making measurements (keep a book on standby)."

I learnt how this circuit works courtesy of a book called 'Semiconductor Circuit Design' which happened to have a chapter covering leakage.  With Ge (won't work with Si) the collector base leakage is high enough to form a voltage across any largish base resistor to ground (think of the c-b junction as acting like a huge resistor which then forms a divider with the resistor from base to ground).  It's about the worse bias you can use in a circuit, which is why virtually every fuzztone has its own unique sound.

RG has an article (Zonk something or other?) covering this bias type, which I've subsequently read.  Had I known about it a few years ago I might have gone down that path instead. 

[gez]

http://www.geofex.com/Article_Folders/zonkmach/zonkbst.pdf

[gez]

Have you listened to Satisfaction lately? - IMO the part sounds pretty crappy anyway! Love it and everything, but still...

I know, it's bloody awful, but if you want a really nasty fuzz to make you sound 'edgy' (or plain crap) you could do worse! 

[davebungo]

http://www.geofex.com/Article_Folders/zonkmach/zonkbst.pdf

And that circuit has a decoupling cap.  It does make me wonder how variable the sound must have been when these were originally made.  Perhaps they had a transistor selection process in the factory or, then again, maybe they just didn't bother.  Who knows.