FUZZ FACE TRIM POTS - Transistors with leak

[percyhornickel]

Hi, I have some germanium transistor (2SB492) and I´m planning to build a FF but they have some leakage as I can see,  could I still use them if I use some trim pots to set the bias in both transistors collectors or just in Q2?

Using geofex circuit test (with 2.472K and 2.2M resistors) I am getting this values:

Q1
0,82V (2.2M base resistor not conected)
0,33 mA leak
1,45V (2.2M base resistor conected)
hfe = 63,1

Q2
0,88V (2.2M base resistor not conected)
0,35 mA leak
1,59V (2.2M base resistor conected)
hfe = 71,3

Saludos and Thank you

[Big Monk]

Hi, I have some germanium transistor (2SB492) and I´m planning to build a FF but they have some leakage as I can see,  could I still use them if I use some trim pots to set the bias in both transistors collectors or just in Q2?

Using geofex circuit test (with 2.472K and 2.2M resistors) I am getting this values:

Q1
0,82V (2.2M base resistor not conected)
0,33 mA leak
1,45V (2.2M base resistor conected)
hfe = 63,1

Q2
0,88V (2.2M base resistor not conected)
0,35 mA leak
1,59V (2.2M base resistor conected)
hfe = 71,3

Saludos and Thank you

The best thing to do would be to breadboard the circuit and see how the voltages come out. I imagine with higher leakage transistors you will need to tweak collector resistor values to get in line with established Fuzz Face bias values.

This says nothing about how they will sound, of course, as the gain of the transistors and their inherent noise/hiss may disqualify them but at least you'll know prior to boxing it up.

Plug your values into here prior to breadboarding to give yourself an idea of where you want to start, component wise:

https://www.diystompboxes.com/biascalc/ffbias.html

[percyhornickel]

Thank you for the answer, and yes I am planning to breadboard the circuit but I still have not all components in here so I will have to wait a little until they gets to my house, I hope the next week.

My main question is if the important thing is to get VC1=-0.5v and VC2=-4.5V (electrosmash.com) setting the bias no matter the resistors values are used for this?

I´ve build a npn rangemaster before with an AC127 and a 100k pot to set the bias, sounds great..   ..never knew the hfe or leak but I set this almost the normal bias Vc=7V and the sound is very good.

I thought it probably could be the same with the ff.

[Big Monk]

Thank you for the answer, and yes I am planning to breadboard the circuit but I still have not all components in here so I will have to wait a little until they gets to my house, I hope the next week.

My main question is if the important thing is to get VC1=-0.5v and VC2=-4.5V (electrosmash.com) setting the bias no matter the resistors values are used for this?

I´ve build a npn rangemaster before with an AC127 and a 100k pot to set the bias, sounds great..   ..never knew the hfe or leak but I set this almost the normal bias Vc=7V and the sound is very good.

I thought it probably could be the same with the ff.

Let's use the calculator and your values. In the Fuzz Face, leakage is going alter the collector voltages for Q1 and Q2. Here are your values unaltered with nominal Fuzz Face part values:

Q1 = 61 hFE, 330 microamps Leakage
Q2 = 73 hFE, 350 microamps Leakage



R1 = 33k, R2 = 8.67k, R3 = 1k, R4 = 100k

Vc1 = 0.202 vDC, Vc2 = 8.51 vDC

The interesting thing here is you basically have a close approximation of the Q2 and Q3 stages from a Tonebender Mk II. Leakage is a most agreeable element in a Tonebender, but not so much a Fuzz Face. Let's illustrate the effect leakage would have on the Fuzz Face by dropping the leakage amounts in the calculator:

Q1 = 61 hFE, 100 microamps Leakage
Q2 = 73 hFE, 100 microamps Leakage

R1 = 33k, R2 = 8.67k, R3 = 1k, R4 = 100k

Vc1 = 0.549 vDC, Vc2 = 5.55 vDC

Here we see that dropping the leakage down to 100 microamps for each brings us nearly in line with generally accepted Fuzz Face voltages. Now let's see how some resistor tweaks can help you get the voltages in line using the leakage values you actually measured:

Q1 = 61 hFE, 330 microamps Leakage
Q2 = 73 hFE, 350 microamps Leakage

R1 = 19k, R2 = 12k, R3 = 1k, R4 = 100k

Vc1 = 0.493 vDC, Vc2 = 4.85 vDC

So, depending on how you set the trim pots, you should be able to tweak the circuit to hit the right voltages. With that said, it says nothing about how it will sound, which could be a function of the transistor hFE value, naturally occurring hiss, etc.

[Electric Warrior]

My main question is if the important thing is to get VC1=-0.5v and VC2=-4.5V (electrosmash.com) setting the bias no matter the resistors values are used for this?

no, those numbers are pretty random, anyway. They don't even match. With a Q2C around 4.5V, Q1C will probably be closer to 0.6 or 0.7 - but then again Q2C depends a lot on the fuzz pot's tolerance.

4.5V is pretty much as low as it gets for a vintage Fuzz Face. Don't be afraid to bias it hotter than that. Keeping Q1C over 0.3V is probably a good strategy, though.
The Tone Bender MK1.5 was often biased hotter, but then again it's infamous for crapping out at higher temperatures. When Arbiter cloned the circuit, they biased it a little colder with a 33k instead of a 47k on Q1's collector.

The interesting thing here is you basically have a close approximation of the Q2 and Q3 stages from a Tonebender Mk II. Leakage is a most agreeable element in a Tonebender, but not so much a Fuzz Face. Let's illustrate the effect leakage would have on the Fuzz Face by dropping the leakage amounts in the calculator:

I guess the MKII was the fix for the temperature issues that the MK1.5 had. The first stage keeps the signal over the threshold. Instead of getting farty, the pedals actually sounds really cool when gating a little..

The best thing to do would be to breadboard the circuit and see how the voltages come out. I imagine with higher leakage transistors you will need to tweak collector resistor values to get in line with established Fuzz Face bias values.

This. The calculator is a great tool to check out the mechanics of the circuit, but it may not be all that precise.
If the bias is off because your transistors are too leaky, you can simply make up for it by tweaking Q1's collector resistor. As Q1 sets the bias on Q2, both collector voltages will be adjusted at once.

Q2's leakage and hfe hardly have an influence on the voltages, btw..

[Big Monk]

This. The calculator is a great tool to check out the mechanics of the circuit, but it may not be all that precise.
If the bias is off because your transistors are too leaky, you can simply make up for it by tweaking Q1's collector resistor. As Q1 sets the bias on Q2, both collector voltages will be adjusted at once.

Q2's leakage and hfe hardly have an influence on the voltages, btw..

I've noted some inconsistencies with the calculator that causes it to bomb out for values i know are not out of line. This could have something to do with limits the author originally put on the calculations, i.e. min/max acceptable inputs.

I'm not the smartest guy in the room for sure and i don't have a grasp on the equations required to make something better or else i'd have had an Excel sheet written weeks ago!

[Rob Strand]

I'm not the smartest guy in the room for sure and i don't have a grasp on the equations required to make something better or else i'd have had an Excel sheet written weeks ago!

The calculations aren't that helpful.   If you adjust the 8.2k resistor on the collector of Q2 to get 4.5V you will be pretty much there.  Once you get that far you can experiment with changing that resistor to raise or lower the Q2  to suit your taste.

Most of the voltages in the fuzz-face are not easily changed.  The only thing to worry about is way off voltages which usually means a bug in the build.

[Big Monk]

I'm not the smartest guy in the room for sure and i don't have a grasp on the equations required to make something better or else i'd have had an Excel sheet written weeks ago!

The calculations aren't that helpful.   If you adjust the 8.2k resistor on the collector of Q2 to get 4.5V you will be pretty much there.  Once you get that far you can experiment with changing that resistor to raise or lower the Q2  to suit your taste.

Most of the voltages in the fuzz-face are not easily changed.  The only thing to worry about is way off voltages which usually means a bug in the build.

My starting place is always mathematical analysis. That's because day job is protection engineering at a utility. We don't have the option to do trial and error in the field so we do fairly rigorous analysis prior to doing work in a practical sense.

That carries over for me in pedal building.

I agree with you 100%. I was just trying to illustrate the points where mods to standard values would help here with the transistors he has in hand.

[Rob Strand]

My starting place is always mathematical analysis. That's because day job is protection engineering at a utility. We don't have the option to do trial and error in the field so we do fairly rigorous analysis prior to doing work in a practical sense.

That carries over for me in pedal building.

Normally I'm like that.    In order to get good numbers you need to get a lot of input parameters right.   You could measure the transistor gains with your DMM which might be at a 1mA collector current but the circuit is operating the transistor at 250uA so the gain in-circuit might be lower; hFE varies with current.  Using the DMM gain means the calculations will be off despite good intentions of measuring parameters up front.   Another variable is leakage, you might measure the leakage of Q1 but in-circuit the leakage ends-up being lower because the VCE voltage of Q1 is quite low.    For the FF we have one main goal of getting the Q2 collector voltage right and one parameter to control it, the Q2 collector resistor.  The experimental method gets you the right answer without a lot of up-front measurements - essentially it bundles-up all the variables.

[percyhornickel]

Thank you for the answer, and yes I am planning to breadboard the circuit but I still have not all components in here so I will have to wait a little until they gets to my house, I hope the next week.

My main question is if the important thing is to get VC1=-0.5v and VC2=-4.5V (electrosmash.com) setting the bias no matter the resistors values are used for this?

I´ve build a npn rangemaster before with an AC127 and a 100k pot to set the bias, sounds great..   ..never knew the hfe or leak but I set this almost the normal bias Vc=7V and the sound is very good.

I thought it probably could be the same with the ff.

Let's use the calculator and your values. In the Fuzz Face, leakage is going alter the collector voltages for Q1 and Q2. Here are your values unaltered with nominal Fuzz Face part values:

Q1 = 61 hFE, 330 microamps Leakage
Q2 = 73 hFE, 350 microamps Leakage



R1 = 33k, R2 = 8.67k, R3 = 1k, R4 = 100k

Vc1 = 0.202 vDC, Vc2 = 8.51 vDC

The interesting thing here is you basically have a close approximation of the Q2 and Q3 stages from a Tonebender Mk II. Leakage is a most agreeable element in a Tonebender, but not so much a Fuzz Face. Let's illustrate the effect leakage would have on the Fuzz Face by dropping the leakage amounts in the calculator:

Q1 = 61 hFE, 100 microamps Leakage
Q2 = 73 hFE, 100 microamps Leakage

R1 = 33k, R2 = 8.67k, R3 = 1k, R4 = 100k

Vc1 = 0.549 vDC, Vc2 = 5.55 vDC

Here we see that dropping the leakage down to 100 microamps for each brings us nearly in line with generally accepted Fuzz Face voltages. Now let's see how some resistor tweaks can help you get the voltages in line using the leakage values you actually measured:

Q1 = 61 hFE, 330 microamps Leakage
Q2 = 73 hFE, 350 microamps Leakage

R1 = 19k, R2 = 12k, R3 = 1k, R4 = 100k

Vc1 = 0.493 vDC, Vc2 = 4.85 vDC

So, depending on how you set the trim pots, you should be able to tweak the circuit to hit the right voltages. With that said, it says nothing about how it will sound, which could be a function of the transistor hFE value, naturally occurring hiss, etc.

What a great explanation, I´m so greatful for that. Well I guess I will have to wait till parts come over and breadboard with I´ve got and evaluate the sound and hiss/noise (I am not a fuzz expert). After doing that I think I probably could think about building a Tonebender which I like too (with min attack). I´m from Venezuela so parts in here are so difficult/imposible to find so I need to work with parts I have right now. I found a couple of NTE103A which are like NPN AC127 that I will try too. I have no transistors with big hfe to put on the circuit for Q2 so let´s see how it works later.

Thank you!!!

[percyhornickel]

My main question is if the important thing is to get VC1=-0.5v and VC2=-4.5V (electrosmash.com) setting the bias no matter the resistors values are used for this?

no, those numbers are pretty random, anyway. They don't even match. With a Q2C around 4.5V, Q1C will probably be closer to 0.6 or 0.7 - but then again Q2C depends a lot on the fuzz pot's tolerance.

4.5V is pretty much as low as it gets for a vintage Fuzz Face. Don't be afraid to bias it hotter than that. Keeping Q1C over 0.3V is probably a good strategy, though.
The Tone Bender MK1.5 was often biased hotter, but then again it's infamous for crapping out at higher temperatures. When Arbiter cloned the circuit, they biased it a little colder with a 33k instead of a 47k on Q1's collector.

Ok thank you, I will have that in mind (Q1C over 0.3V), and yes..  ..temperature is a important factor and you can note this when measuring hfe, I have to wait a few mins while the transistors gets ambient room.

[Electric Warrior]

My main question is if the important thing is to get VC1=-0.5v and VC2=-4.5V (electrosmash.com) setting the bias no matter the resistors values are used for this?

no, those numbers are pretty random, anyway. They don't even match. With a Q2C around 4.5V, Q1C will probably be closer to 0.6 or 0.7 - but then again Q2C depends a lot on the fuzz pot's tolerance.

4.5V is pretty much as low as it gets for a vintage Fuzz Face. Don't be afraid to bias it hotter than that. Keeping Q1C over 0.3V is probably a good strategy, though.
The Tone Bender MK1.5 was often biased hotter, but then again it's infamous for crapping out at higher temperatures. When Arbiter cloned the circuit, they biased it a little colder with a 33k instead of a 47k on Q1's collector.

Ok thank you, I will have that in mind (Q1C over 0.3V), and yes..  ..temperature is a important factor and you can note this when measuring hfe, I have to wait a few mins while the transistors gets ambient room.

Right. As temperature rises, leakage increases, which brings down Q1's collector voltage and brings up Q2's.

[percyhornickel]

The best thing to do would be to breadboard the circuit and see how the voltages come out. I imagine with higher leakage transistors you will need to tweak collector resistor values to get in line with established Fuzz Face bias values.

This says nothing about how they will sound, of course, as the gain of the transistors and their inherent noise/hiss may disqualify them but at least you'll know prior to boxing it up.

Plug your values into here prior to breadboarding to give yourself an idea of where you want to start, component wise:

https://www.diystompboxes.com/biascalc/ffbias.html

Just wanted to let you know I just breadboarded the circuit with Q1 2SB405 / Q2 2SB492, not measuring leak or hfe.

First I did Q1 2SB492 / Q2 2SB405, set bias with two trim pots (22k-23k each) and sounded the best that case, then I swap transistors and set bias Q1C=0.45V / Q2C=doesn´t matter, I adjusted this looking for the "sweet point"..    ..this thing is amazing and my guitar just has a couple of hambuckers (Ibanez RX20) and a 15W amp. Even just a little hiss noise..

I think it is better letting a 100k pot outside to set bias according with what I want and temperature issues..

I can´t wait to weld all this in a pcb and test with a decent amp and a strat or tele with singles....


Saludos!!!