Drifting bias Fuzz Face

[rangermaster]

Hi

I've build a 1:1 copy of a fuzz face. Including the original design pcb, caps etc

Also 2x AC128 transistors

Q1: 75 Hfe
Q2: 120 Hfe
Both low leakage.

I'm already trying for 5 days now to get the bias right. -0.5v on Q1 and -4.5v on Q2

I tried different resistors and now i have placed 2 trimmers in place of the 33k and 8.2k spot. The bias keeps drifting.

When i set the bias correctly and wait an hour or so(cooling down) it stays stable but when i check the next day the bias is totally off.

Any one knows what this could be? Temperature?(could only be a couple of celcius in difference here). All solder joints are good and all wires are seated well. Everything looks and measures fine.

There is no led and it is battery fed. Exact build as the original arbiter but in a different enclosure.

Thanks.

[GibsonGM]

The bias is known to drift with them - generally temperature-based.  This might be what you're seeing.   Are you also measuring the battery at the same time you test bias?  Just in case there is something else going on (like a short).

How much is it drifting?  Is it only in one direction, or does the voltage go up and down?

I'd do more of my judgement based on listening now that it's working!

[mozz]

Yes, how much is it drifting? What is your temperature? Where are the transistors from? Did you measure the leakage or did the seller just state a number?

[rangermaster]

Transistors are all <290mA leakage

Battery is 9.01v

Yeaterday 10pm is set the bias on -0.53 on Q1 and -4.69 on Q2

This morning 5am i checked and Q1 had -0.74 and Q2 -5.87

I checked everything and there are no shorts.

Temperature here is between 20 and 23 celcius.

[mozz]

Readjust it then let it sit again.  How did you measure the leakage? Are you touching the transistors? Put your meter on the collector and watch, if the voltage is changing it's not settled.  If it keeps drifting your leakage is too high

Edit
Something doesn't sound right, q1 c goes up q2c should go down.  Get rid of the pot on q2, use 8.2k resistor.  Adjust q1c trim until q2c is 4.5v, don't worry about q1c volt. 

[amptramp]

A hairdryer and a refrigerator are all you need to diagnose thermal drift.  The Fuzz Face, particularly the germanium one, is subject to variations due to temperature.  Germanium is worse because it cannot be passivated against surface leakage currents (because germanium oxide is conductive).  Silicon is passivated by a layer of silicon dioxide (quartz that is not conductive).  Germanium leakage currents are normally about a thousand times higher than silicon and the leakage doubles for every 10°C rise in temperature.  The leakage current in germanium is often on a par with the circuit current.

[EATyourGuitar]

You can buy freeze spray for electronics testing purposes. I wouldn't use a refrigerator when I can use freeze spray.

[WoundUp]

You can buy freeze spray for electronics testing purposes. I wouldn't use a refrigerator when I can use freeze spray.

Canned duster upside down works well, also

[rangermaster]

Readjust it then let it sit again.  How did you measure the leakage? Are you touching the transistors? Put your meter on the collector and watch, if the voltage is changing it's not settled.  If it keeps drifting your leakage is too high

Edit
Something doesn't sound right, q1 c goes up q2c should go down.  Get rid of the pot on q2, use 8.2k resistor.  Adjust q1c trim until q2c is 4.5v, don't worry about q1c volt.

Ok so when i have -4.5v on Q2 and -0.33 on Q1 that isn't a problem ? I thought it should be between -0.5 / -0.7