Voltage/Resistor/Q1/Fuzz Face

Frequently asked question on transistors: types, substitutions, how to test, use and misuse them.
User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

Messin' around with a FF circuit on my breadboard right now.
In Q1 I really dig a 2N3440 with a hfe at about 50-60, but the thing is that with a 33k resistor on the collector it sits at about 2v or higher.
With a 47k it lands on about 1.8v and with 100k it reads about 1.5v

I've got a trimpot on Q2's collector, so I can always bias this to 4,5v (or whatever I feel sounds "best") but there is a different sound/feel depending on the voltage of Q1.

Would like some input regarding the voltage/resistor in Q1 just to hear you ideas experience and technical understanding :)

User avatar
DrNomis
Old Solderhand
Information
Posts: 6804
Joined: 16 Jul 2009, 04:56
my favorite amplifier: Self-Built Valve Amp Head :)
Completed builds: Dallas Arbiter Fuzz Face,Tone Bender Professional Mk 3,Tone Bender 3-Knob,Baja BK Butler Tube Driver,Baja Real Tube Overdrive,Roger Mayer Octavia,EH Soul Preacher,Tech 21 XXL Distortion,MFOS Weird Sound Generator.
Location: Darwin,Northern Territory Australia
Has thanked: 98 times
Been thanked: 275 times

Post by DrNomis »

polifemo wrote:Messin' around with a FF circuit on my breadboard right now.
In Q1 I really dig a 2N3440 with a hfe at about 50-60, but the thing is that with a 33k resistor on the collector it sits at about 2v or higher.
With a 47k it lands on about 1.8v and with 100k it reads about 1.5v

I've got a trimpot on Q2's collector, so I can always bias this to 4,5v (or whatever I feel sounds "best") but there is a different sound/feel depending on the voltage of Q1.

Would like some input regarding the voltage/resistor in Q1 just to hear you ideas experience and technical understanding :)


As the 2N3440 Transistor is a Silicon High Voltage (300V DC Max Collector-Base Voltage) NPN type, I'd be expecting something like 1.2V on it's collector with a 33k collector resistor when biased correctly, what are you using for Q2?, it may be that there's something funny going on with it........ :hmmm:
Genius is not all about 99% perspiration, and 1% inspiration - sometimes the solution is staring you right in the face.-Frequencycentral.

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

DrNomis wrote:
polifemo wrote:Messin' around with a FF circuit on my breadboard right now.
In Q1 I really dig a 2N3440 with a hfe at about 50-60, but the thing is that with a 33k resistor on the collector it sits at about 2v or higher.
With a 47k it lands on about 1.8v and with 100k it reads about 1.5v

I've got a trimpot on Q2's collector, so I can always bias this to 4,5v (or whatever I feel sounds "best") but there is a different sound/feel depending on the voltage of Q1.

Would like some input regarding the voltage/resistor in Q1 just to hear you ideas experience and technical understanding :)


As the 2N3440 Transistor is a Silicon High Voltage (300V DC Max Collector-Base Voltage) NPN type, I'd be expecting something like 1.2V on it's collector with a 33k collector resistor when biased correctly, what are you using for Q2?, it may be that there's something funny going on with it........ :hmmm:
Q2 is a 2N929 with a hfe at about 160.

User avatar
DrNomis
Old Solderhand
Information
Posts: 6804
Joined: 16 Jul 2009, 04:56
my favorite amplifier: Self-Built Valve Amp Head :)
Completed builds: Dallas Arbiter Fuzz Face,Tone Bender Professional Mk 3,Tone Bender 3-Knob,Baja BK Butler Tube Driver,Baja Real Tube Overdrive,Roger Mayer Octavia,EH Soul Preacher,Tech 21 XXL Distortion,MFOS Weird Sound Generator.
Location: Darwin,Northern Territory Australia
Has thanked: 98 times
Been thanked: 275 times

Post by DrNomis »

polifemo wrote:
DrNomis wrote:
polifemo wrote:Messin' around with a FF circuit on my breadboard right now.
In Q1 I really dig a 2N3440 with a hfe at about 50-60, but the thing is that with a 33k resistor on the collector it sits at about 2v or higher.
With a 47k it lands on about 1.8v and with 100k it reads about 1.5v

I've got a trimpot on Q2's collector, so I can always bias this to 4,5v (or whatever I feel sounds "best") but there is a different sound/feel depending on the voltage of Q1.

Would like some input regarding the voltage/resistor in Q1 just to hear you ideas experience and technical understanding :)


As the 2N3440 Transistor is a Silicon High Voltage (300V DC Max Collector-Base Voltage) NPN type, I'd be expecting something like 1.2V on it's collector with a 33k collector resistor when biased correctly, what are you using for Q2?, it may be that there's something funny going on with it........ :hmmm:
Q2 is a 2N929 with a hfe at about 160.

Okay, I googled that and the transistors you are using are both NPN silicons, check with a multimeter to see if you're getting at least .6V across each transistor's Base-Emitter junction, that would indicate that both are biasing up correctly and are working, if one of them shows 0V across their B-E junction then that transistor has gone B-E short and needs replacing, hope that helps, also, check to make sure you have the transistor oriented correctly too.... :thumbsup
Genius is not all about 99% perspiration, and 1% inspiration - sometimes the solution is staring you right in the face.-Frequencycentral.

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

I've got an Atlas Peak so that I can check the transistors and they are ok and oriented correctly.

I've tried several low gain (below 100) transistors in Q1 and it seems like "the lower the hfe the higher the voltage".
With a BC109c/33k resistor the voltage is 1.2v.

User avatar
DrNomis
Old Solderhand
Information
Posts: 6804
Joined: 16 Jul 2009, 04:56
my favorite amplifier: Self-Built Valve Amp Head :)
Completed builds: Dallas Arbiter Fuzz Face,Tone Bender Professional Mk 3,Tone Bender 3-Knob,Baja BK Butler Tube Driver,Baja Real Tube Overdrive,Roger Mayer Octavia,EH Soul Preacher,Tech 21 XXL Distortion,MFOS Weird Sound Generator.
Location: Darwin,Northern Territory Australia
Has thanked: 98 times
Been thanked: 275 times

Post by DrNomis »

polifemo wrote:I've got an Atlas Peak so that I can check the transistors and they are ok and oriented correctly.

I've tried several low gain (below 100) transistors in Q1 and it seems like "the lower the hfe the higher the voltage".
With a BC109c/33k resistor the voltage is 1.2v.


That's interesting, the BC109c typically has an Hfe of about 450 or greater,it may be that the transistor you're using for Q1 might have too low a gain to work correctly in the Fuzz Face circuit with the stock standard resistor values, the stock resistor values were originally chosen to work with Germanium transistors, my theory anyway, I'd say try tweaking the resistor values till the circuit works with the transistors you're using..... :thumbsup
Genius is not all about 99% perspiration, and 1% inspiration - sometimes the solution is staring you right in the face.-Frequencycentral.

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

DrNomis wrote: I'd say try tweaking the resistor values till the circuit works with the transistors you're using..... :thumbsup


That's what I'm doing :D

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

Just for fun I decided to see which value transistor that was needed to get the voltage down to about 1.2v on Q1's collector and it turned out to be 470k!

With about (I've got a trimmer on Q2) 6.8k on Q2's collector I land on 4.5v.

Haven't been able to listen to this yet (the family doesn't like my fuzz building...) but I will.

Seems like one needs totally different values on the collector resistors - maybe also in the feed back loop!? - when using low gain si transistors.


What I'm trying to do here is to use SI trannies with the classic GE Fuzz Face values.

User avatar
dwmorrin
Breadboard Brother
Information
Posts: 76
Joined: 07 Apr 2013, 20:03
Has thanked: 3 times
Been thanked: 14 times

Post by dwmorrin »

I'm very curious what that will sound like. Please let us know.
At 470k, Q1's load impedance is dominated by the base input of Q2, which can be estimated as beta x 1k, which is much less than 470k. Usually the 33k is much less. So you're really changing the operation of the circuit.

When changing transistors, all of the resistor values may require changing. If you study the various versions of the fuzz face, and it's extremely close variants (check tonebender mk1.5, mkii, vox tonebender, roger mayer, etc) you may get some ideas.

As a gain stage, Q1 has an open loop gain of the collector load divided by the internal resistance of the transistor (25Ω is a good guesstimate value), so increasing the 33k might be an attempt at maxing out the first gain, but is such maxing desirable? The actual, closed loop gain is attenuated by the 100k feedback resistor. Play with this as well to set Q1's operating point. Lower value=more neg. feedback/attenuation, eventually reaching a non-functional point. Raise value=less neg. feedback, possible increase of gain, again only to a point. You may find adjusting the 100k and the 33k will keep you closer to the stock values. Any of the resistors (including the 1k fuzz pot) are fair game for experimentation.

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

As I understand absolutely NO theory electronics regarding electronics I can't say that I understand your reply :oops:

In the end I will of course go with what I like the best sound wise, but I DO like to understand things!!!

A circuit will of course sound behave different if Q1 sits at 1.2v compared with 2.1v with even if the resistor on the collector is 33k.

I'd like to be able to listen to the output from Q1, so WHAT/HOW do I do this?

IF I get to possibility to listen to the circuit (with 470k on Q1's collector) I will report back!

I will also check which resistor value that is needed on the Q2 collector in order to get 4.5v using 33k/47k/100k in Q1.

User avatar
dwmorrin
Breadboard Brother
Information
Posts: 76
Joined: 07 Apr 2013, 20:03
Has thanked: 3 times
Been thanked: 14 times

Post by dwmorrin »

polifemo wrote:As I understand absolutely NO theory electronics regarding electronics I can't say that I understand your reply :oops:
:D No problem. Simpler answers can be had.
I'd like to be able to listen to the output from Q1, so WHAT/HOW do I do this?
Stick an audio probe at Q1's collector and see what you get. General audio probe = 100nF (0.1µF) cap alligator clipped to a guitar amp input somehow. Google "DIY audio probe" for pictures and ideas.

Rewording my last post: raising the 33k to 470k seems extreme. I'm truly curious if the circuit still works at that point. I think I'll have to try this for myself.
Try messing with the 100k feedback as well.

When subbing different transistors, all the resistors of the circuit are subject to change.

Because Q1 and Q2 have a feedback loop, resistor changes can affect both of their operation.
This is a great site that includes a textbook explanation of this circuit: http://www.talkingelectronics.com/te_in ... index.html
The fuzz face like circuit begins at Fig 21.
This is required reading for fuzz face exploration: http://www.geofex.com/article_folders/f ... fffram.htm

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

It will take some time before I'll be able to get into this a bit more, but i did manage to listen to the "470k/Q1" circuit a little bit yesterday.

With a 470k resistor on Q1's collector the circuit produces much less fuzz compared to 33k-47k, and doesn't sound very good at all.
I did not have time to mess with the feed back loop resistor, which I feel is a must if using such a large resistor in Q1!

I feel that anything between 33k - 100k works just fine (did not try anything lower than 33k) with the lowest values producing the most Fuzzy tones.

I haven't had the time to measure my trimmer on Q2, but with 33k-47k in Q1 I end up somewhere around 2k-3k on Q2's collector.

Seems like the actual voltage in Q1 doesn't matter that much (with this particular transistor: 2N3440, with a hfe at about 60, it will sit at about 2v with 33k and about 1.8v with 47k) and it seems to be of more importance to stay within the 33k - 47k range on Q1's collector and then use whatever is needed in Q2 in order to get the "4.5v" or whatever one goes for.

Still a bit curious regarding how much of the differences in sound between GE/SI is due to the fact that a GE transistor in Q1 will operate at about 0.5v whilst a SI transistor usually sits at 1.2v.


PS If anyone wonders WHY I'm messing with all this, it's because I like a really low hfe in Q1, and that I happen to like the sound and cleanup of these 2N3440 trannies a lot 8)
2n2369's also sound good in Q1. Ds

User avatar
dwmorrin
Breadboard Brother
Information
Posts: 76
Joined: 07 Apr 2013, 20:03
Has thanked: 3 times
Been thanked: 14 times

Post by dwmorrin »

polifemo wrote: With a 470k resistor on Q1's collector the circuit produces much less fuzz compared to 33k-47k, and doesn't sound very good at all.
I did not have time to mess with the feed back loop resistor, which I feel is a must if using such a large resistor in Q1!
I feel that anything between 33k - 100k works just fine (did not try anything lower than 33k) with the lowest values producing the most Fuzzy tones.
This makes sense. If we "guesstimate" the input impedance of the Q2 to be around 100k, then Q1's fixed collector load (normally the 33k resistor) really needs to be below this value for the circuit to function. I would expect some lower gain if you lower the value.
Your 33k-100k range is probably about all that is practical... even with messing with the other resistors. You'd have to raise the 1k fuzz pot to raise Q2's input impedance to try and get around the limitation.
I've seen 27k many times for this circuit.
Seems like the actual voltage in Q1 doesn't matter that much (with this particular transistor: 2N3440, with a hfe at about 60, it will sit at about 2v with 33k and about 1.8v with 47k) and it seems to be of more importance to stay within the 33k - 47k range on Q1's collector and then use whatever is needed in Q2 in order to get the "4.5v" or whatever one goes for.
Q1's voltages are stabilized by the feedback loop. Start at the base and trace the dc path "back" to the collector. When tracing dc, consider all capacitors OPEN, or disconnected. You'll go from base, to 100k resistor, to Q2's emitter. Follow through Q2's emitter to base, and you're back at Q1, now at the collector.
Because of this loop, Q1's voltages tend to stay put. Why? Imagine a small increase in voltage at the collector... this goes through Q2's emitter, through the 100k, and creates a small increase at Q1's base. The base is locked to the collector.
Still a bit curious regarding how much of the differences in sound between GE/SI is due to the fact that a GE transistor in Q1 will operate at about 0.5v whilst a SI transistor usually sits at 1.2v.
The straight voltage differences are due to the physics of Ge junctions vs. Si junctions. These transistors are called "bipolar junction transistors" or BJTs for short. They are made on the same idea that Si and Ge diodes are made. An elementary view of a transistor is two diodes connected back to back. You should always visualize the base to emitter junction as a diode. The base to collector diode is also a diode, but for normal operation it is not so helpful to visualize this as a diode.

If you apply voltage to a diode, starting a 0.001V and increasing slowly, the diode does not conduct much at all. Keep going, and eventually the diode "turns on" and conduct's heavily. It's "resistance" changes from very high to very low suddenly.
This "forward voltage drop" which every diode has is usually 0.100V-0.300V for Ge diodes, and usually around 0.600 for Si diodes.

If you've got 2 Ge's with 0.250V drops, and 2 Si's with 0.600V drops, add the two together and you get your voltages : 0.5V for Ge, and 1.2V for Si.
Follow the loop again. Q1 from base to emitter must be 1 diode drop. Q1's base will always be around 0.100-0.300V for Ge, and around 0.600 for Si. Keep going back up the loop, and you get the Q2's base emitter diode drop. This means Q2's base, which is also Q1's collector must be another diode drop higher. Make sense?

Why the Ge circuits and Si circuits sound different is a wholly greater can of worms, and I don't think we've settled on a scientific answer yet. I would guess it's part of the reason why so many people keep building the circuit all these years later.

User avatar
DrNomis
Old Solderhand
Information
Posts: 6804
Joined: 16 Jul 2009, 04:56
my favorite amplifier: Self-Built Valve Amp Head :)
Completed builds: Dallas Arbiter Fuzz Face,Tone Bender Professional Mk 3,Tone Bender 3-Knob,Baja BK Butler Tube Driver,Baja Real Tube Overdrive,Roger Mayer Octavia,EH Soul Preacher,Tech 21 XXL Distortion,MFOS Weird Sound Generator.
Location: Darwin,Northern Territory Australia
Has thanked: 98 times
Been thanked: 275 times

Post by DrNomis »

dwmorrin wrote:
polifemo wrote:

Why the Ge circuits and Si circuits sound different is a wholly greater can of worms, and I don't think we've settled on a scientific answer yet. I would guess it's part of the reason why so many people keep building the circuit all these years later.

It possibly has something to do with the turn-on characteristics of Ge diodes versus Si diodes, Ge diodes have a softer, more rounded knee compared to Si diodes, so therefore I'd expect a Ge Fuzz Face to sound smoother and less buzzy, although when you compare the output signals of a Ge Fuzz Face and a Si Fuzz Face on an oscilloscope, the clipped waveforms actually look very similar, maybe the feedback system behaves slightly different in Ge Fuzz Faces and Si Fuzz Faces, if you put a scope probe on Q1's collector and monitor it while feeding in a steady sinewave, you see this waveform in a Ge Fuzz Face as shown in the attached screenshot, notice how the bottom portion of the waveform is rounded?, if you do the same with a Si Fuzz Face you get a waveform with sharper corners:
Attachments
Ge Fuzz Face Q1 Collector Waveform.
Ge Fuzz Face Q1 Collector Waveform.
Genius is not all about 99% perspiration, and 1% inspiration - sometimes the solution is staring you right in the face.-Frequencycentral.

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

Thanks for sharing your knowledge guys!

Found some time to check the circuit a bit more this afternoon and with 47k on the collector of Q1, Q2's collector will sit on 4.67v when using a 2.7k resistor.

2.7k is far from 8.2k, which is the recommended value, but I can't hear any "problems" :D (What do you guys say?)

I will continue my efforts in trying to build SI Fuzz Faces with the traditional "GE resistor values", and I've got some more 2n3440's coming in on wednesday. Hopefully I will find a transistor in the 70-90 range for Q1 and in Q2 I'm very fond of my mysterious 2N929's (which measures within traditional Q2 120-140 values).

Using small caps (10pF) between collector/base there's a little bit of a "Germanium" character added.
Not really simulating "leakage", but not only a high cut!
Maybe something in between :hmmm:

Looking forward to what this experiment with "leaky diodes" will lead to: https://www.diystompboxes.com/smfforum/ ... c=101610.0

User avatar
dwmorrin
Breadboard Brother
Information
Posts: 76
Joined: 07 Apr 2013, 20:03
Has thanked: 3 times
Been thanked: 14 times

Post by dwmorrin »

polifemo wrote:Thanks for sharing your knowledge guys!

Found some time to check the circuit a bit more this afternoon and with 47k on the collector of Q1, Q2's collector will sit on 4.67v when using a 2.7k resistor.

2.7k is far from 8.2k, which is the recommended value, but I can't hear any "problems" :D (What do you guys say?)
If it sounds good, use it. No one is "recommending" you use 8k2 as the collector resistor - that is simply what was in the original '60s Fuzz Face... and the stories go that some sounded good and some bad. If there's any wide spread "recommendation" out there, it's that you do change the 8k2!

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

dwmorrin wrote:
polifemo wrote:Thanks for sharing your knowledge guys!

Found some time to check the circuit a bit more this afternoon and with 47k on the collector of Q1, Q2's collector will sit on 4.67v when using a 2.7k resistor.

2.7k is far from 8.2k, which is the recommended value, but I can't hear any "problems" :D (What do you guys say?)
If it sounds good, use it. No one is "recommending" you use 8k2 as the collector resistor - that is simply what was in the original '60s Fuzz Face... and the stories go that some sounded good and some bad. If there's any wide spread "recommendation" out there, it's that you do change the 8k2!
Hi,

I know that no one is "recommending" me to use 8.2k, but as my little "470k as collector resistor in Q1-test" proved, things can go very wrong if one drifts away to far from the "standard values"! (Of course 470k vs 33k is a much bigger difference than 2.7k vs 8.2k)

I ALWAYS trust my ears but I like to understand - as much as possible - WHAT happens and WHY!!!
I really appreciate your explanations :D

Another question: Is there any difference between setting the frequency in the output section by using a 500k volume pot and a 0.01uF cap compared to a 100k pot and a 0.047 cap? (They both create a filter at about 32-34Hz)

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

To continue on my question above:

When swapping the 500k volume pot, but keeping the 0.01uF cap, for a 100k pot one will raise the filter from 32Hz to 159Hz, right?

One could get more or less the same effect by reducing the cap to 0.0022uF and keeping the 500k pot, creating a filter at 145Hz, or?

User avatar
DrNomis
Old Solderhand
Information
Posts: 6804
Joined: 16 Jul 2009, 04:56
my favorite amplifier: Self-Built Valve Amp Head :)
Completed builds: Dallas Arbiter Fuzz Face,Tone Bender Professional Mk 3,Tone Bender 3-Knob,Baja BK Butler Tube Driver,Baja Real Tube Overdrive,Roger Mayer Octavia,EH Soul Preacher,Tech 21 XXL Distortion,MFOS Weird Sound Generator.
Location: Darwin,Northern Territory Australia
Has thanked: 98 times
Been thanked: 275 times

Post by DrNomis »

polifemo wrote:To continue on my question above:

When swapping the 500k volume pot, but keeping the 0.01uF cap, for a 100k pot one will raise the filter from 32Hz to 159Hz, right?

One could get more or less the same effect by reducing the cap to 0.0022uF and keeping the 500k pot, creating a filter at 145Hz, or?


Probably, give it a try and if you don't like the results you can always try something else..... :thumbsup
Genius is not all about 99% perspiration, and 1% inspiration - sometimes the solution is staring you right in the face.-Frequencycentral.

User avatar
polifemo
Breadboard Brother
Information
Posts: 140
Joined: 09 Nov 2007, 21:42
Location: Stockholm, Sweden.
Has thanked: 3 times
Been thanked: 7 times

Post by polifemo »

Yes I can!

I'm just asking because there's so much talk of the type: "Eric Johnson prefers a 100k volume pot" * when one as well could say that "E.J likes a slight bass cut at the end of the circuit", or is it something about the range/travel of the pot that makes 500k different from 100k, regardless of the cap/cut off frequency?

*which in fact turned out to be 470k reverse log :wink:

Post Reply