Help me make a fuzzface

[karbomusic]

A slight diversion but I already had a scope when I began bulding fuzzes so I've always just adjusted bias via scope which comes out something like this (not my pic)...



I may or may not adjust bias to get a more symmetric duty cycle but either way, tens of fuzzes or so into this I can pretty much nail the sound (I expect to hear via personal preference) every time by starting with the scope then confirming by ear. I can check voltages too but my point is the waveform is a fingerprint of the overall character of the clipping or maybe better said, I don't care if Q2C is 4.5V if my waveform is what I expect.

[a2music]

It is important to note, for the beginning fuzz face builder, that those voltages do not have to be exact. Q1 does pretty well between -.5 and -.9v in a germanium fuzz face. I prefer the lower voltage in most cases but it's not that critical. Q2 doesn't need to be at precisely -4.5v. I usually set mine to 5v with a 10K pot (centered) in series with the biasing trim pot. This way I have an ideal bias in the pot's center position but can add or subtract 5K to/from the resistance which will starve the transistor of voltage creating high compression and splattery note attack or it will overload the transistor with voltage which can sound good up to a point. It makes for a brighter and more open sound. The other great benefit is being able to adjust the bias on the fly when you're playing on a hot outdoor stage or on a cold studio floor. 

When I was learning to build fuzz circuits I saw a lot of posts that were extremely technical and because I didn't understand the purpose of testing HFE, leakage etc. I tended to ignore those things but then wondered why I couldn't make a decent sounding fuzz face. Since you've ordered a transistor tester you'll soon be able to take matters into your own hands. Once you know a transistor's HFE and leakage you have the data you need.

As for those AC 128 that you are trying, they may just be the wrong gain or too leaky and, since you're not doing anything to bias the circuit you're not getting good results but you've already taken important steps to correct that. Most of the AC 128 transistors that you'll find on eBay are made by Tungsram and they're not great. That's not to say you can't make a good sounding fuzz face with them. You just might do better with a pair that's already been tested for use in a fuzz circuit. Smallbear is a great resource and there are at least one or two eBay sellers who actually test transistor pairs for the fuzz face. Since the common practice is to use a transistor pair where the second transistor's gain is a 1.5 multiple of the first transistor's gain, be cautious when shopping for "matched pairs". Some sellers seem to think that matched pairs should have the same gain. If you're looking around the Internet for information on germanium transistors gain for the fuzz face you'll see a lot of people referencing 80/120 as the gain standard for Q1/Q2. It pays to experiment. Different transistor types will perform differently and you should try non-conventional transistor pairs. My favorite gain pair is around 60/90. That tends to clean up nicely with the volume control and to roar when maxed out.  For many of us, the love of fuzz has led to seeking the acquisition of transistors from the four corners of the earth. Most people don't hear the subtle and nuanced differences between Japanese and American transistors but we do.


Sent from my iPhone using Tapatalk

[snarblinge]

I started down this path about 15 years ago, beginning with the fuzz face, I must agree with the above comments on transistors, I quickly realised singles was no way to buy, so bought up 50 bags, and boxes from Eastern Europe, Japan, local surplus stores, and augmented with the occasional collection of anything from local electronics stores, and the pedal parts suppliers, I have more stock than I will ever need, but auditioning transistor pairs is so satisfying. I bought a transistor test or early on, and sorted for leakage, faults and hfe.

External bias has never been overly useful to me. But then I don't play live

[Electric Warrior]

It is important to note, for the beginning fuzz face builder, that those voltages do not have to be exact. Q1 does pretty well between -.5 and -.9v in a germanium fuzz face. I prefer the lower voltage in most cases but it's not that critical. Q2 doesn't need to be at precisely -4.5v. I usually set mine to 5v with a 10K pot (centered) in series with the biasing trim pot. This way I have an ideal bias in the pot's center position but can add or subtract 5K to/from the resistance which will starve the transistor of voltage creating high compression and splattery note attack or it will overload the transistor with voltage which can sound good up to a point. It makes for a brighter and more open sound. The other great benefit is being able to adjust the bias on the fly when you're playing on a hot outdoor stage or on a cold studio floor. 

When I was learning to build fuzz circuits I saw a lot of posts that were extremely technical and because I didn't understand the purpose of testing HFE, leakage etc. I tended to ignore those things but then wondered why I couldn't make a decent sounding fuzz face. Since you've ordered a transistor tester you'll soon be able to take matters into your own hands. Once you know a transistor's HFE and leakage you have the data you need.

As for those AC 128 that you are trying, they may just be the wrong gain or too leaky and, since you're not doing anything to bias the circuit you're not getting good results but you've already taken important steps to correct that. Most of the AC 128 transistors that you'll find on eBay are made by Tungsram and they're not great. That's not to say you can't make a good sounding fuzz face with them. You just might do better with a pair that's already been tested for use in a fuzz circuit. Smallbear is a great resource and there are at least one or two eBay sellers who actually test transistor pairs for the fuzz face. Since the common practice is to use a transistor pair where the second transistor's gain is a 1.5 multiple of the first transistor's gain, be cautious when shopping for "matched pairs". Some sellers seem to think that matched pairs should have the same gain. If you're looking around the Internet for information on germanium transistors gain for the fuzz face you'll see a lot of people referencing 80/120 as the gain standard for Q1/Q2. It pays to experiment. Different transistor types will perform differently and you should try non-conventional transistor pairs. My favorite gain pair is around 60/90. That tends to clean up nicely with the volume control and to roar when maxed out.  For many of us, the love of fuzz has led to seeking the acquisition of transistors from the four corners of the earth. Most people don't hear the subtle and nuanced differences between Japanese and American transistors but we do.


Sent from my iPhone using Tapatalk

It's not just that the bias guidelines -0.5 and -4.5 don't need to be exact. By no means does a Fuzz Face need to be biased anywhere close to that. The MK1.5 Tone Bender, on which the Fuzz Face is based on, was biased to around -0.15V/-8V and it still sounds great, but temperature is a real problem for these. No surprise the only tweak Arbiter made was lowering the bias a tad by swapping the 47k on Q1's collector for a 33k.

Some prefer matched pairs to mismatched ones. Some like higher or lower gain transistors than what is usually recommended. The "guidelines" are pretty random and not necessarily close to what Arbiter was doing back in the day. But there are many ways to set up a good sounding Fuzz Face.

I started down this path about 15 years ago, beginning with the fuzz face, I must agree with the above comments on transistors, I quickly realised singles was no way to buy, so bought up 50 bags, and boxes from Eastern Europe, Japan, local surplus stores, and augmented with the occasional collection of anything from local electronics stores, and the pedal parts suppliers, I have more stock than I will ever need, but auditioning transistor pairs is so satisfying. I bought a transistor test or early on, and sorted for leakage, faults and hfe.

External bias has never been overly useful to me. But then I don't play live

This. Buying just two transistors for a Fuzz Face is optimistic. You need more transistors in order to have something to work with.

Firstly thanks for the replies here guys this is all super helpful and sorry for a late reply.
I wasn't as familiar with biasing and the theory side of pedal building and electronics so all of this is new to me.

I'll breadboard, both the npn/pnp fuzz faces again, and try and get Q1 at 0.5v and Q2 at 4.5v. If I'm still having problems ill report back. Also, If i was to add a bias knob on the exterior of a fuzz face controlling Q2's collector resistance would that help this issue? and if its a good idea what value pot would you recommend?

I'm finding it hard to make a recommendation on how to fix your problem without knowing your voltages with stock values.

A bias knob on the exterior of the enclosure may be useful to make up for temperature changes with a germanium Fuzz Face. Q1's collector resistor would probably be a better choice than Q2's, as it sets the bias for both transistors.

Most of the high gain silicon units seem measure over 5V on Q2's collector, so my guess is the math is off a little.

In circuits, I can't say I've ever got voltages over 4.5V with the standard parts.   Not sure what is going on here.

Chucking the correct Vbe's in the calculator I get:
hfe     Vc2
200     2.94V
400     3.43V
800     3.68V

Those are in the ball-park of what I remember.
Maybe I need to dig-up some BC108s and check again.

I experimented with 2sb175s today. The calculated voltages were quite a bit off (about 1.5V on Q2C), even considering the wide tolerance of the fuzz pot. Maybe the behaviour of these things in the real world is not 100% predictable from the numbers we have to work with?