This fuzz looks like a joke. But it's not.

[brett]

Hi
Maybe someone has thought of this before.
Just in case, here's the schematic:

This fuzz sounds like a cross between a fuzzface and a tube-sound fuzz.  It gets nicely asymetric like the fuzzface (see how the transistor on the left looks like Q1 in a fuzzface), but it also has softish clipping through the transistor on the right.

I rarely use vero, but in this case anything else seemed excessive:


My only complaint is that the bass is a bit flabby (like a fuzzface).  A smaller input cap might help solve that.  Also, like the fuzzface, this circuit is designed for low input impedance and works best if used as the first pedal in a chain. 

cheers

[Toney]

 

Thanks for sharing Brett!

Um, how does it work. Q2???

[brett]

Hi
Q2 is used for its two slightly dissimilar diodes.  (All  NPN transistors have two PN juctions in them).  The idea comes from the old FZ-1S fuzz.
cheers

[jmusser]

I have found that several of the simple concepts end up being some of the nicest sounding effects, such as Joe Davisson's Easy Drive and Antiquity Fuzz, and Pete Moore's Range Pig. I'll give this one a try. Thanks

[amz-fx]

Um, how does it work. Q2???

It is like having two diodes in parallel, as shown here:



Only one side of the signal is being clipped.   The transistor junction that has the lowest threshold will begin to conduct and the other junction will never get the chance to switch on.

regards, Jack

[Toney]

Ok got it.
So Q2 is being used purely as a diode.
Ah so... thanks!

[bwanasonic]

I'll have to breadboard this one. Brett, have you tried an additional ^ diode to ground? Thinking it might be worth putting one on a switch.

Kerry M

[disantlor]

Only one side of the signal is being clipped.   The transistor junction that has the lowest threshold will begin to conduct and the other junction will never get the chance to switch on.

so that means you can get away with just one?

[oldrocker]

What happens if you put two transisters to ground next to each other?  Would it then act as the two diodes?

[QSQCaito]

Aren't you missing the 1m resistor to avoid pops when bypassed at output? Is it needed?

Just my 0.00001 cent

bye bye

dac


PS: Is it also missing at input, right?

[Alex C]

Aren't you missing the 1m resistor to avoid pops when bypassed at output? Is it needed?

PS: Is it also missing at input, right?

I think the schematic is intended to show the functioning circuit, not necessarily a complete, "ready-to-build" circuit diagram.  In this case, a pulldown resistor at the input would be a good idea, but adding one at the output would be redundant, as the 100K pot on the output provides a path to ground.

-Alex

[mac]

And a 3rd PNP transistor in parallel with Q2?


mac

[Paul Marossy]

Interesting concept. Q2 might also sound cool with MOSFETs acting as clipping diodes. 

[jmusser]

Doesn't the Fat Gnat work somewhere along the lines? What are the clipping voltages were talking about for a BJT and a Mosfet, and how do they fit in the range of clipping voltages of germaniums, zeners and and Silicons?

[markm]

This looks like it could be interesting to mess around with for sure.

[petemoore]

  Might be a good candidate for 'Joke X 2', like the Bazz Fuss.
  Possibly one exception where a joke might improve when repeated.

[brett]

Hi
thanks for the interest.

I'll have to breadboard this one. Brett, have you tried an additional ^ diode to ground? Thinking it might be worth putting one on a switch.

My idea was that the first transistor clips one side of the signal, and the diode clips the other.  The transistors clips and compresses in three ways: 1. it can only swing a small amount before cut-off (because there's no emitter resistor, so the base and collector are quite close to the ground), 2.  the input impedance changes dynamically to reduce pickup output, and 3.  the current gain varies dynamically so that the approach to cutoff is smooth.

For this last reason, I suspect that modern high gain transistors (e.g. 2N5089) won't sound as good as the transistors suggested.  Transistors with big changes in hFE with Ic should sound good here as they do in the fuzzface (those hFE and Ic data or graphs are on most datasheets).  Good ones include the PN2369A, 2N2222A, and BD139.  Good ones seem to be mainly old devices, high speed switches and power devices.  It might have something to do with levels of doping or architecture that result in high transition frequencies or high power handling, but compromised hFE.

I *think*  the reason for the nice, fuzzy diode clipping in this circuit is the high output impedance from Q1 (about 100 kohms).  As a contrast, the MXR Distortion + has an output impedance of 10k, and the Proco Rat only 1k.  Those circuits must have lots more current flowing (and therefore clip over a wider range of voltage.)  Somebody might offer a comment about the characteristics of transistor junctions.  ??

I tried to find a schematic of the Fat Gnat, but couldn't.  Some discussion indicated that it is a version of the Rat, with jfet clippers.  If so, this is quite a different-sounding circuit.  More like a Fuzzface or maxxed out tube-sound fuzz or a vintage version of the EasyDrive.

cheers

[KerryF]

Here, I just created a veroboard layout.  Its unverified, but I figured it was easy enough...



I will breadboard it tomarrow and let you all know.  Maybe give some soundclips if it works good.

[Paul Marossy]

I tried to find a schematic of the Fat Gnat, but couldn't.

I've read about the Fat Gnat here before, but I've never seen the schematic for it. There is some talk of it at AMZ, though.

[brett]

Some photos of wave forms out of this critter.

I looked at sine waves with 3 input levels: 5, 20 and 100mV peak-to-peak.

Output from 5mV was barely distorted and 0.5 V p-p.  The interesting thing about this is that 500mV/5mV = a gain of only 100.  The hFE shown by my multimeter for this transistor was 250.  So I figure that the low operating point is working its magic and inhibiting the transistor's potential hFE.

Next, 20mV p-p.  The output range was +0.7V and -1.0V.  There is some solid clipping on the positive part of the curve due to the clipping transistor, and there is some nice assyemetry happening.   This photo shows the wave on a 0.5V grid (barely visible, clearer in next photo).


At 100mV p-p input, things get really interesting.  Both sides of the signal are clipped.  The top part at 0.8V, due to the clipping transitor, and the lower part, due to the amplifying transistor going into cutoff.  The output range of +0.8 to -1.5V is seen in this photo (0.5V grid).



The results indicate how much this circuit compresses the signal.  For a 500% increase in input (20mV to 100mV), the output only increases by 30% (from 1.7 V p-p to 2.2 V p-p).  But the shape changes from mild single-sided clipping to heavy double-sided clipping.