Original silicon Fuzz Face - bias drifting?

[gmr1]

Got a late 60's fuzz face in to work on. BC108 transistors. Had one of the electrolytic caps replaced at one point (with an ugly blue xicon...), so the owner isn't worried about it having parts replaced.

He told me it just quit one day. Kinda sputtered out. Turns out it still works, but needs so much signal to get the fuzz to open up enough to be audible, and then it's spitty and un-usable.

So I checked everything and noticed the bias was way off. So I pulled the 8K2, threw in a pot, got it to bias up nicely. 3K9 to get 4.5v on the q2 collector. It sounds ok. Pretty wooly with not as much top end as I'd expect from a si FF.

Is this just a case of the transistor going bad? I have the right replacements, but I'm afraid it's going to sound a little different than he originally remembered. It was a go to fuzz face for him, so I want to choose the correct solution.

Thoughts? Should I leave the trim (it's in the hole of the resistor, so I can easily reverse the work), and see if he can get some use out of it as is? I'm guessing the transistor is on the way out, yes? It's a bit on the noisy side, but i've heard worse

[zachomega]

Are you sure none of the other parts have drifted grossly?  I'd check the other resistors and see if they measure correctly.  Changing Q2's Collector resistor to a value of less than half seems like a big change. 

I doubt the transistor is bad.  They tend to either work or not work and when they are on their way out it doesn't take long for them to go. 

The easiest way to check the resistors is by disconnecting one lead from the board and measuring the resistance.  Measure the fuzz pot as well as that will have a big effect on the bias.  Also, if the cap bypassing the fuzz is shot, it will throw the bias off majorly. 

-Zach

[gmr1]

This is good info Zach, thanks. I was surprised to see that it would bias up at all, as I figured that Q2 had to be bad. The collector resistor needing to drop that much was also odd to me. I pulled the 100K, and that was within range (it just looked bad, so I figured while I was in there...). The odd thing is, how could a component drift so fast that it was working well one use, and the next time, nothing? Seems like it would be gradual enough that it would slowly go bad. I know the 2 electrolytics are good, so not much to check after that.

I'll give the rest of the circuit a check in the morning. Thanks again for the info. - Greg

Are you sure none of the other parts have drifted grossly?  I'd check the other resistors and see if they measure correctly.  Changing Q2's Collector resistor to a value of less than half seems like a big change. 

I doubt the transistor is bad.  They tend to either work or not work and when they are on their way out it doesn't take long for them to go. 

The easiest way to check the resistors is by disconnecting one lead from the board and measuring the resistance.  Measure the fuzz pot as well as that will have a big effect on the bias.  Also, if the cap bypassing the fuzz is shot, it will throw the bias off majorly. 

-Zach

[PRR]

Clean the board. Maybe someone spilled beer in it (and forgot). Or maybe just old-age and CT humidity seeped into the PCB.

Most of these fuzz things work best when they barely work, so it only takes a little leakage (transistor, PCB, beer-residue) to go out of order.

[zachomega]

2 quick anecdotes...funny enough both featuring Ampeg amps. 

One time I was repairing a 1950's Jet and the tremolo had stopped functioning.  Well actually it'd function for about 2 pulses and disappear.  Upon going over the circuit, one of the resistors had drifted 500% of its value upward.

Another time, I was playing through an early 1970's VT40.  Put the amp into standby, came back 15 minutes later, no sound.  Bad filter caps.  Sometimes things just stop working. 

-Zach 

[gmr1]

"Most of these fuzz things work best when they barely work"

This is a great quote, and so true! This is the main reason I don't want to go overboard "fixing it"... I have a feeling that it was at least slightly mis-bias'd and that was part of it's charm. No reason to make it so right, that it's wrong.

The board is clean. The enclosure looks like a bomb hit it, but minus cleaning up some solder splatter from the last guy that worked on it, the insides are quite tidy for a 40+ year old pedal.

Do I need to check the pots? They seem to function fine, taper seems good, no noise. I'm going to run through the board component by component today. If everything is within tolerance, the odd behavior has to be a transistor, yes?

[zachomega]

Being that the fuzz pot is what biases up both transistors, I'd say yes, you need to check the fuzz pot.  The volume pot should be fine assuming it is doing its job.

-Zach

[gmr1]

Being that the fuzz pot is what biases up both transistors, I'd say yes, you need to check the fuzz pot.  The volume pot should be fine assuming it is doing its job.

-Zach

yeah, that's what I figured... it just seems like it reacts correctly, but really I suppose I should start with it. Thanks again for the help.

[gmr1]

Checked every component. Nothing has drifted more than 5% or so (pretty much within orig. tolerance of the parts). Fuzz pot - almost dead on 1K. the .01uf, good to go. Both electrolytics are newer philips caps. I don't know why it now hits bias at around 4K on the Q2 collector. All that's left is the transistors... odd.