Germanium Muff bias issues

nbabmf · March 10, 2010, 02:34:39 AM

I can't seem to get Q2 and Q4 to bias like the others, and the result is a very weak effect - well below unity. I'm using CV7355 (mil spec) 2N1307. The gains are in the high 200's and leakage below 100uA by my measurements using this method: http://www.smallbearelec.com/HowTos/FuzzFaceFAQ/FFFAQ.htm I thought it might be the collector resistors (Q1 and Q3 have 22k, Q2 and Q4 have 12k), but bumping the one on Q2 to 22k didn't do anything at all so it's back to 12k for now. The emitter resistors were all replaced with 1N4001 diodes, anode side to ground.

Q1
C 4.32
B 0.63
E 0.52

Q2
C 0.71
B 0.71
E 0.51

Q3
C 4.26
B 0.61
E 0.51

Q4
C 0.64
B 0.72
E 0.57

I'm freaking baffled. This is on a GGG board with triangle specs aside from the emitter resistor change mentioned above.

http://www.generalguitargadgets.com/pdf/ggg_bmp_tri_sc.pdf - schematic
http://www.generalguitargadgets.com/pdf/ggg_bmp_tri_lo.pdf - layout

Jarno · March 10, 2010, 03:58:12 AM

I built one with the same transistors. I first tried the resistors as calculated by the big muff bias calculator, but that didn't work. So I just stuck in trimpots, adjusted, desoldered them, measured them and put in appropriate metal film resistors.

Excellent transistors imho, built quite nice FuzzFaces with them too (there were 80-125Hfe ones in my batch as well). I built two, one with these germanium ones, the other with silicon transistors. The texture is remarkebly similar, they respond differently. And the gain ranges are different, because of the difference in Hfe, I used 220-280 for the germanium, and around 350-400 for the silicon one.

trad3mark · March 10, 2010, 04:34:56 AM

i remember breadboarding a silicon muff before and testing out a tonne of different silicon transistors. I'd love to make a germanium muff, it's kinda on the to do list, but i've only just ordered some GE transistors. Some of them have a pretty high hfe, so they might be worth a go. What i'm more interested in though is the GE diodes i've ordered. they should give some smoother clipping for sure.

Jarno · March 10, 2010, 06:34:27 AM

By the way those CV7355's are PNP and not NPN, you did remember to reverse the polarity of the electrolytics and powersupply?

mac · March 10, 2010, 08:57:55 AM

C2 and B2 seems to be connected togheter.
The emiter of Q4 is at a diode drop, did you put the 2k2 or 2k7 emiter resistor, or just the diode?
Guys, please PM me, i'll do my best to help you.

mac

nbabmf · March 10, 2010, 08:32:46 PM

Quote from: Jarno on March 10, 2010, 06:34:27 AM
By the way those CV7355's are PNP and not NPN, you did remember to reverse the polarity of the electrolytics and powersupply?

Yeah... I did that right!

Quote from: mac on March 10, 2010, 08:57:55 AMC2 and B2 seems to be connected togheter.

That's what it seems to be, but when I use the diode/continuity function, it doesn't beep at me. There's a small resistance between the two... same as it would be with a diode there.

QuoteThe emiter of Q4 is at a diode drop, did you put the 2k2 or 2k7 emiter resistor, or just the diode?

Just the diode. I could put a resistor in series with the diode, since the diode is of a nominal resistance anyhow...

mac · March 11, 2010, 11:12:33 AM

QuoteC2 and B2 seems to be connected togheter.

That's what it seems to be, but when I use the diode/continuity function, it doesn't beep at me. There's a small resistance between the two... same as it would be with a diode there.

I mean that there is a short between C and B in your breadboard or pcb, check C5 and C6 or a bad connection from C to B.
Is R9 present?

About Q4, the diode is the problem.
If you use only a diode, and not the diode+emiter_resistor (see my schem at my gallery), not only it won't bias but Q4 gain will be too high, about 12k/0.1k*=120. Without mods gain is 12k/2.7k=4.4, often called recovery stage after the losses at the tone control.
There is no need to use a diode there, even with the original resistors you should get some volts at the collector.
But if you want a higher collector voltage replace R22 with a 560k or 680k. This should sit C at 4v or so.

* I estimated the resistance of the diode as 0.1k. Can be different but always much less than 2k7.

mac

nbabmf · March 12, 2010, 04:32:13 PM

R10 was 390, not 390k. @#$% ME.

Jarno · March 12, 2010, 04:48:33 PM

Don't you just love when that happens?

nbabmf · March 12, 2010, 05:58:24 PM

Q2 and Q4 are both chilling around 2.5-3v now, but it sounds killer. I'll play with it for a while... then maybe tweak it some more.

mac · March 12, 2010, 09:37:39 PM

QuoteQ2 and Q4 are both chilling around 2.5-3v now, but it sounds killer. I'll play with it for a while... then maybe tweak it some more.

Cool!

QuoteI built one with the same transistors. I first tried the resistors as calculated by the big muff bias calculator, but that didn't work. So I just stuck in trimpots, adjusted, desoldered them, measured them and put in appropriate metal film resistors.

Can you PM about this?

mac

News:

Germanium Muff bias issues