Maestro FZ-1 Fuzz Box knockoff

[Bo Deadly]

There are two major groups of Germanium transistors. There are ones that tend to be leaky and vary greatly with gain, and those that tend to be relatively low leakage and more stable gain numbers. I used to think that this was due to grown versus alloy junction fabrication differences but I now think the grown ones where never really used that much and that there was some other kind of improvement in the alloy junction process that made much more consistent and reliable parts. My feeling is that the first transistor in a conventional fuzz should be about 100uA leakage and gain 80. Most of the newer Germaniums have much lower leakage. So you want some leakage IMO.

However, I think you can use the leakier transistors. Even the 300uA if they're high gain. As demonstrated by CJ's fz-1 circuit, the leaky ones will self bias. This might be exploited as a feature and not a fault. You just have to design the circuit around individual transistors. You can use a 1M (or higher) bias resistor instead of the 100k in a typical fuzz-face. That could yield different sounds depending on how much the emitter resistor is bypassed.

Create a rig for testing leaky self-biasing transistors. Use a 1K current limiting collector resistor and two trimpots to trim out the collector and emitter resistances. Play with the trimpots to see what sort of bias you can get. Consider that if the emitter is fully bypassed you might not want exactly V/2. For best results, use a signal generator, attenuator and scope to see how it clips.

 

[dmp]

Anyway to eliminate the harsh ' torn speaker' crackle on top of the smooth distortion? I have a rangemaster circuit on the protoboard trying different Ge transistors. The tone or character of the individual transistors are all different, some better some worse. Bias is straightforward. With collector at V/2, the tone is great but I don't like the harsh distortion that crackles out. Playing soft reduces it. Add resistor before the base?

 

[CJ]

Just depends on what song you are playing. Good to have a pedal for each, the underbiased rancid clipping sound, and the more saturated sound. I like it when you can get feedback at low volume levels ,

Some boxes like the Big Cheese añd the ehx cock fight have an underbiased setting but it does not sound like the Ge sound to me.

Gain/leakage for the 3 transistors I bought for the tonebender were:

Q1. Gain 43 Leakage 203 uA Q2 55/166 Q3 89/231

Temperature fluctuations are more of an issue with the ancient transistors. PRR has some semi recent (2021) posts over at the diy stompbox pages. Good to know he is ok.

 

[fazer]

Of the two under 300, I got gain 41 and 73.

Yes most Ge are in that range As long as the leakage is not bad, I find the Tonebender circuit to work well. It’s kind of like a rangemaster in front of a fuzz face. Plenty of gain and sustain. Also I like those old British GE’s tone but most mine are like you say gain at 50 and excessive leakage.

 

[dmp]

Good to have a pedal for each, the underbiased rancid clipping sound, and the more saturated sound.

Bringing the bias higher solved it and got it to sound great, using a AC128. Went all the way to the end of the trimmer to have Re=1k
"underbiased rancid clipping sound" would be an awesome name for a pedal - too bad it sounds bad
This is still the rangemaster circuit, moving on to the fuzz tone / tone bender circuits next.
For the rangemaster circuit, I calculated the operating point (Vc) with a sweep of the emitter resistor (1k-6k) to see how it behaves in theory.
Leakage kills the operating point as it gets too high - the collector voltage is due to leakage current.
Gain isn't super important for the DC bias, obviously important for AC response.
The problem with targeting a Vc for the class A bias setup, is the 'real' bias is the transistor current minus leakage

 

[soapfoot]

I hope it’s okay for me to ask a related question here:

If a Ge fuzz pedal (Mk. II tone bender circuit) sounds amazing but is a fair bit too noisy (hiss), is this a sign of an issue with the transistors, how they’re biased, something else, or impossible to say?

 

[Bo Deadly]

Gain isn't super important for the DC bias

Sure it is. Gain is collector current / base current and bias is the current at equilibrium. So more gain, current and bias are all directly related. This also applies to leakage. If you have 300uA leakage and the gain is 100, that means you're base is leaking 3uA.

, obviously important for AC response.

I agree but I wouldn't say it's obvious. Most germanium circuits, even at very high gain, are going to have a response that will extend above 20kHz. If you find a transistor / gain combo that limits frequency response to 10kHz, that's fantastic. Store that pedal in a secure location.

 

[dmp]

If you have 300uA leakage and the gain is 100, that means you're base is leaking 3uA.

The leakage is the Ic of the transistor without any external base current. Normal transistor behavior is Ic = hfe*Ib
Not much difference is needed in the bias resistor with difference in hfe, that is what I was observing, however a big difference is needed with leakage current. "false leakage gain" : REAL gain is important or you won't hear anything from the pedal.

Anyway this is just a simple inverting amplifier to understand what is going on with a leaky Ge transistor. The FZ-1 is more complicated and not in the 'simple' operating range

 

[Bo Deadly]

The leakage is the Ic of the transistor without any external base current. Normal transistor behavior is Ic = hfe*Ib

If the transistor is biased normally with a small forward Vbe, you will find that your equation holds true with and without leakage.

The way I compare the leakage of two transistors is to divide leakage by gain. The part with the lower ratio is generally going to perform better.

So gain needs to be considered when thinking about biasing leaky transistors.

 

[dmp]

Isn't full collector current:
Ic = hfe*Ib + I_leakage

According to the measurements in Keen it seems to be

 

[Bo Deadly]

Isn't full collector current:
Ic = hfe*Ib + I_leakage

According to the measurements in Keen it seems to be

I don't know. Maybe I'm misremembering all of this.

I'm trying to think of a way to measure if leakage is completely independent of gain but that's non-trivial I think.

Maybe by comparing the DC gain (Keen) to the AC gain of a simple fully bypassed common emitter amplifier.

 

[CJ]

Trying different supply voltages might be fun, 1.5, 3. 4.2 (vape Batt), 6, 9, 3 kv

 

[CJ]

did the cap A/B test ,

definately some difference,

one of the caps shorted out after a 1 hour burn in at 5 volts

so this test was for the first cap in the signal chain only.

 

[Bo Deadly]

Interesting. Leakage? Parasitic resistance?

 

[dmp]

Anyone know what the ballpark transistor voltages should be for FZ-1?

I am trying different PNP germanium with the 'Mister Fuzz' pcb which uses a DC converter to get -3v
I am using 10k for the Q2 collector resistor instead of 1.5k and 1uF caps between Q1/Q2 and Q2/Q3
Right now I have Q1_Ve= -0.63, Q2_Vc= -2.18, Q3_Vc= -1.71 (measured from ground)
These transistors have Q1 G=73, L=216uA, Q2 G=93, L=366 uA, Q3 G=111,L=441 using the RG Keen method

Fuzz is good on higher stings and 'Satisfaction' sounds right but playing a power chord and it sounds anemic, like it's running out of juice.
Almost like a heavy compressor and sustain.

 

[Rob Flinn]

I love these Maestro pedals. I had a stagephaser MP1 in for repair a few years back & liked it so much I made myself one on veroboard.

 

[fazer]

Fuzz is good on higher stings and 'Satisfaction' sounds right but playing a power chord and it sounds anemic, like it's running out of juice.

Chords never sounded good with the FZ-1. I got mine back in the 60s. Single notes sounded best. It was years later in the 80s that I bought a Roger Mayer fuzz (fuzz face) that a realized the power of the guitar volume setting’s importance on the sound with a fuzz face/ Tonebender circuit. They make soloing so easy with the sustain and touch sensitivity of the sound/guitar volume interaction. It’s a great deal of fun.

 

[dmp]

Good to know. Maybe it's close then

 

[dmp]

I ended up using a 3.6k on the Cathode resistor for Q2
I adjusted a 10k trimmer to find something that worked well with the range of the attack pot and the transistors
Has a cool sound for leads or picking. The compression/sustain of it.

 

[CJ]

here is a thread with some diy tweaking

http://tagboardeffects.blogspot.com/2014/08/maestro-fz-1-fuzz-tone.html