Why won't my Fuzz Face bias anywhere near -4.5V with an 8k2 on Q2?

[BluffChill]

Hi all!

I'm building a Fuzz Face. I have built many. it ain't my first rodeo. The problem with this one is I cannot get the collector of Q2 to bias at -4.5 with anything remotely in the range of an 8k2 resistor. I'm making the classic circuit with no tweaks - the 33K on Q1 collector, the 470 after -9V before the Q2 resistor, and for some reason, it massively overbiases every time. I've had to use a 47K resistor to get it near -4.5V. What could I have done wrong?

I've checked all my other resistor values and they're all correct. I had to wire in a trimmer and keep turning it while I measured it with the multimeter, and it didn't hit -4.5V until I was at about 45K on the trimmer. I'm using old resistors - a hige bag of old carbon comps and 1/2 watts I got at a house clearance. They meter fine - my 8k2 read exactly 8k2. Anyone else had this problem?

[BetterOffShred]

What are the specs of your Q2 transistor?   

[BluffChill]

Actually reasonably low - it's a 90hFE OC75.

[BetterOffShred]

you should still be able to get that to work, does it have crazy high leakage ?

[willienillie]

The gains and leakages of both transistors are relevant.  It's a pretty rare pair that biases up correctly with stock circuit values.  Even when they get close to the desired Q2c voltage, you would still be better off with a trimmer to dial in the sweet spot by ear, and then to occasionally compensate for temperature variations, etc.

[BetterOffShred]

Yeah I have some MP416's that are like 80-100 and I usually use an ~80 in Q1 and 90-100 in Q2..  Virtually no leakage on these guys (tested with Atlas) and they typically will produce sound with stock values.   I also tend to tweak a bit.   I haven't really tried to make a Fuzz Face with very leaky transistors.

[PRR]

You know the drill. Voltages! V(Q1c), V(Q2e), and the V(Q2c) you get.

[mozz]

Yes, post all your voltages. Q1 may be way too leaky.

[BluffChill]

I tried it with a few different cans - a 2SB54 (high hFE) and a lower gain OC75 (56hFE).

In both cases, 0.3V at Q1c, 0.1V at Q2e. Then Q2c only is way too high at around 8V with an 8k2 resistor.

[mozz]

Q1c should be closer to.5- .6. I don't know how much that will bring down q2c but it will. Bad transistors, wiring error, bad pinout, check fuzz pot wiring,  old carbon comps are known to go high.

[Electric Warrior]

I tried it with a few different cans - a 2SB54 (high hFE) and a lower gain OC75 (56hFE).

In both cases, 0.3V at Q1c, 0.1V at Q2e. Then Q2c only is way too high at around 8V with an 8k2 resistor.

What makes you think Q2c measures way too high? Your circuit sure is biased on the hot side, but not unreasonably so. 0.3–0.4V at Q1C is not uncommon in Fuzz Faces. The voltages look ok. Does it sound any good?

Q2's hfe and leakage actually hardly make a difference on its collector voltage. That mainly depends on Q1's specs. With Q1C at 0.3V I'd definitely expect way over 7V on Q2C.

If you do need to bias a little colder, you need a lower leakage transistor on Q1 or a smaller Q1C resistor.