ToneBender MkIII biasing questions

[cpk313]

Could someone please explain to me the relationship between the amount of bias voltage and the gain and leakage of a germanium transistors for this circuit? And how the diode interacts and what those values need to be to get it to work well?

I have a Aion Phobos board and a number of sets of transistors. 2 from SmallBear, some from pedal hacker and I have some AC128s coming from a gentleman in France. My DMM has a spot to measure hfe but I get crazy numbers with it on every transistor I have tried to measure on it which may just be the DMM. I read that DMM typically are expecting the transistor in test to have 0 leakage but some of the one from pedal hacker were supposed to according to that site's purveyor. The ones I got from small bear, a set of OC76 which give what sounds like distortion (not in a good way) along with the fuzz no matter how I bias them. Also have a set of Amperex made 2N281 which in hindsight sounded OK but I couldn't let it be and had to switch things up. These sets both have thinner leads which don't fit as snugly in the sockets. I also read somewhere that they really need to be soldered but that is a bit of a leap of faith at this time because I am not particularly happy with either set. Ones I got from pedal hacker seem to have highish leakage, given as .19 - .30 for a set of AC125 tungsrams or 0 leakage for some Russian made which give much higher bias voltage readings with the trimmer turned all the way down. The instructions for the pedal give a suggested bias voltage of -2.0 for the C of Q3 and -3.5 for the C of Q1 & 2. When I have done this even at the lowest fuzz setting it is a lot of fuzz. There is not a lot of versatility. I seem to have to push the Q1&2 into the -6.0-.5 range to get the pedal to clean up when I turn down the volume. 

So I am not sure what I am doing wrong, I was hoping for a pedal that went from some fuzz to crazy but cleaned up with a volume knob. Any help, guidance would be much appreciated. Thanks!!

[cpk313]

UPDATE:  I got it to work, what I think is well. I ended up using Russian GT2308s for Q1 & Q2 which ended up being at -6.47v at the C. The have gain of 71 & 79 respectively and very low leakage, I was told none but I can't verify that. Q3 ended up being a Tungsram AC125 that has a gain of 80 and a leakage of .25 and biased at -4.6v. Anytime I tried transistors that were what was recommended in the build instructions, while the pedal would make noise, I didn't think it was very good, whether I biased them where recommended or anywhere else. I could never get the set of OC76 from Smallbear to sound good at all. Part of that I think is they have thinner leads that do not make the best contact in the sockets. But to me it sounded like something was being overloaded and there was a persistent layer of distortion (not in a good way) along with the fuzz.

[gregtoolson]

a persistent layer of distortion (not in a good way) along with the fuzz.

I just finished my build using a AC116 in Q3 (hFE 95). Q1 and 2 are silicon (apparently, unlike the Mk I & II TBs, with the MkIII circuit it's only Q3 that really needs to be Germanium). Mine sounds really cool and that "persistent layer of distortion" is there too but I don't think I mind it, however compared to my other TBs it's quite a bit noisier, like a Big Muff with single coils kind of noisy. Maybe that's part of the "persistent layer". Anyway, I didn't aim for any specific bias on Q3, just used the 18k resistor and my collector reads about 2v. Q1 and 2 are up around 7v.

Was it getting to 4.5v on Q3 that cut the noise for yours? And are you saying rolling back your guitar volume cleans up the fuzz,
à la Fuzz Face? 'Cuz mine actually does sorta the opposite and gets more muddied. Really the only other fuzz that I found to do that clean up plus treble lift thing besides a FF is the EHX Muff Fuzz.

[mac]

The Fuzz pot does not work well.
Besides the first stage distorts a bit unless you lower the guitar volume.

You can wire the Fuzz pot as follows:
leg 1 to Base, leg 2 to 220n capacitor, leg 3 to Gnd (or 1 to gnd and 3 to base  )
This way the Fuzz pot will work as an "inverted" volume pot and won't change Q3 bias.
Of course you can wire it as a std volume pot and have the bias changed as you lower the pot.

mac