2SB405 Fuzz Ideas?

[neddyboy]

I recently acquired a matched pair of Sanyo 2SB405s, and I'd like to build a fuzz around them. I had heard they were the ones in the Shin Ei germanium fuzz, but the schematic calls for two different transistors. I've already built a Fuzzface with NOS AC128s, and a silicon Fuzzface. I'd like to build something different from both, but if I'm not mistaken, the Sanyos are a version of the 128. Any suggestions?

[Guitar Poppa]

Hello

You experienced the most difference is between silicon and germanium...

Regarding germanium, my experience is that there are families of transistors, which have some global sonic characteristics... Using transistors of this or that family will give a kind of tonal response, depending on bandwith, and a kind of texture, depending on leakage currents and internal comportement.

To be very synthetic : there are 4 families of germanium transistors, each with a specific sonic comportment ...
--- Audio preamplifiers like OC75, 76 etc., AC125 /126, SFT320-323, russian MP41A, MP42b etc. produce a rather bright or hi-mid sound, as well when clear or saturated... Some older of them, like OC71, MP40A are less trebly, and produce a more lo-mid body...
--- Audio amplifiers (500mW to 1w) like OC81, AC128 etc., Russian GT402s, and equivalents may produce a more creamy texture, with particular dynamics... Your 2SB405s belong to this family.
--- Radio Frequency preamplifiers like OC44, AF124, GT308b and GT313b etc. are very reliable and plain, but almost too good to give its dirty juice to a fuzz.
--- Computer and switching transistors like OC43, OC140, 2N404A and many other give a very homogeneous saturation, but warmer and more living than RF trannies...

These are global markers, globaly reliable, but not absolute.
In each category, each type will have some particularity : AC128 are the most creamy inaudio amplifiers, AC125 and OC76 among the most trebly, OC75 has a very special combiantion of mids and texture etc... Find and try different types in classical circuits : Fuzz Face, Tone Bender MkIII...

The main problem with germanium is that in a family, in a type, each individuate part can be different in leakage current.
Concerning FFs, the first transistor shoul have a leakage inferior than 250µA to work well... BUT : if it's too good, the compression won't be as onctuous you'll expect.
The 2nd transistor must have some leakage (>400µA) to give a classic smooth tone, up to 900µA... BUT sometimes to high leakage lowers bandwith, and level.

An adjustable biasing (trimpot in place of the collector resistor of 2nd transistors, or of the first) his more than usefull : necessary.
It will make the circuit work good if the trannies are correct...
But it won't change their personality. They are like they are, in their family and in their individuality.

Perhaps you expected more precise indications, and maybe even be safe stuff ...
They is no precision in the germanium world, and not much more in analog electronics. The art of analog is to do specific things with imprecise components. We, handymen have all the time to try and experiment: you have to have some benchmarks, to dare choices, and multiply testing, adjusting each circuit. Industrial manufacturers do not have this patience !

Regards !

Se my website for more :
http://guitarpoppa.com/category/techno-en/
http://guitarpoppa.com/category/products/

[mac]

Those matched 405s are 1A low power transistors, similar to AC128.
I'd build a Deacy amp.

mac

[garcho]

measure gains and leakages and that will guide your decision making

[neddyboy]

Hi guys,

I ended up doing a Sola Sound ToneBender Professional MKII and using three AC128s. It sounds great, a nice complement to the Fuzz Face.

I have one question: the pedal produces an unpleasant sound when I let a note trail off to silence. Kind of sputtery almost, sometimes cutting the sound off abruptly. This is a very small complaint. This problem only shows up an VERY low sound levels, something an audience would never hear. I'm mostly just curious if this is a "thing" in a 3-transistor fuzz. I don't get it in my Fuzz Face.

I have to admit that only two of the three transistors were matched. I had matched a bunch for a couple Fuzz Face pedals, but lost my tester since. I also did not use the 20k trim pot in place of the 8K2 resistor. Would either of those cause this odd decay behavior?

[PRR]

> unpleasant sound when I let a note trail off to silence

Post voltages.

"Matched" is totally bogus when not using transistors in push-pull (and you are not doing that).

[neddyboy]

Hi Paul. At what points should I check voltages? Sorry, up until now I've been a "build it and done" kind of guy. I assume you want voltages at the transistors, but which legs? Thanks!

[bluebunny]

. . . but which legs?

All of them.   

[neddyboy]

OK, voltages.

Q1: B=69mv; C=7v; E=0
Q2: B=52mv; C=86mv; E=0
Q3: B=85mv; C=7.75v; E=46mv

[neddyboy]

I've been looking at old posts describing problems in fuzz pedals, and I saw a better description of the "unpleasant sound" I hear: gating. Currently the fuzz tone gates inconsistently at low volumes. It does the same when I turn down the volume knob on my guitar.

[garcho]

shouldn't this be a positive-ground circuit?

Q3 should be ≈ -4.5V

[neddyboy]

Sorry Garcho, yeah. All those numbers should have minuses on them. I assume you're referring to the collector on Q3 being -4.5v? Would raising the value of the 8k2 resistor lower that voltage? Can you suggest a value? I was planning to put in a 20k pot in there, but I probably have the right value resistor in my tool box.

[Guitar Poppa]

OK, voltages.

Q1: B=69mv; C=7v; E=0
Q2: B=52mv; C=86mv; E=0
Q3: B=85mv; C=7.75v; E=46mv

Hello,

It should be possible to adjust the biasing of the three transistors, and the transmission will be more transparent at low levels...

>>> Q1's collector voltage is a bit too high (and the base too low)...
there won't be unlinear effects if this voltage is between 3V and 6V...
--- You can fix it using a transistor with more leakage. This trick is well known and works well. Just check if the transistor is not too noisy : leakage and noise often go together...
--- Another way is to increase the resistor between base and ground... If the value is inferior to 100k, you'll have to encrease it. 47k is a mid value ; the classic 100k value (Tone Bender III) is better. Higher values won't be usefull. You can try to remove the resistor...

>>> Q2 and Q3 work together, coupled on the one hand by a direct link from Q2 to Q3, and a negative feedback from Q3 to Q2. The potentials on the two collectors vary inversely ... Here, the collector of Q2 is a bit low, and that of Q3 is pretty high.
--- The common adjustment is to increase the resistor on the collector of Q3. You should have a result between 10 and 22K...
--- A more vigorous adjustment is to keep 8k2 in place, and decrease the resistor value on Q2's collector. It should work with a value between 22k and 68k.
Using trimpots is particularly easy and efficient in these cases.
--- As for transistors, as I wrote above, Q2 should have a leakage of 100 to 250μV, and Q3 250 to 800μA ...

Good trials !
Guitar Poppa

PS I have read the voltages as if they were negative...

[neddyboy]

Thanks Poppa!

[mac]

Members had posted Q3 C in original units near Vcc.
Given the bias resistors, and with a bit of leakage, is unlikely that Q3 C can sit near 4.5v.
However, the extra stage driving the fuzz face section *masks* Q3 high collector voltage.
But a high Q3 C and a low Q2 C can cause that sputtery sound at low signals levels.

mac