Fuzz face "too bright"

[goo]

hello, guys!
i`ve build fuzz with pre-gain (it works awesome) and input cap switching.
i put the sockets into the scheme and i`ve used old soviet gt402 (ГТ402, 180-200 hfe).
pretty gainy and treble.

and then i`ve tried some soviet transistors as well - mp25b (МП25Б, 120-140 hfe but datasheet shows about 50-60).
less gainy and less treble.

here`s some videos. see description for more information.

---

[pinkjimiphoton]

indeed!! nice work goo!

my very first fuzzface used nte 123's in it. hFE? bah!! humbug!! germanium is all you need to worry about!!- back then, anyways.

really low gain in a fuzzface starts to sound almost like a dirty compressor!

i love this circuit. i have probably 100 variants, quite a trick with so few components.

[mac]

I'll try putting a diode into my Fuzz Face clone today to see what happens..

When I tried this I noticed that a similar transistor wired as a diode worked better than the diodes I have. Maybe the same temperature coefficient, bigger surface, similar hfe and leakage, etc.

> old radio recorders schematics, and some have a FF kind of circuit right after the AUX or MIC input

"FuzzFace", with different values and lower signal level, was a VERY common low-level preamp.

Recently I bought a National Panasonic RQ-445 all Matsushita Ge with built in FF that sounds pretty good as a guitar practice amp
Very creamy and loud through a 12".

mac

[goo]

yes, treble-freqs also depend on the transistors i think. the lower gain - less treble.
but input cap is very important too.
btw, 500K volume pot makes a bit darker sound, yeah (and more volume, right). buuuutttt here`s the question: do putting 500K resistor in parallel with 500K volume pot can make "darker effect" (dispite the volume pot "is" 250K in this case)?

[pinkjimiphoton]

a 500k resistor in parallel with a 500k pot would be a 250k load ultimately. that's what i use usually, it won't make it darker, but brighter.

in series to the 500k resistor will cut the volume a little bit, i would imagine. i dunno if it would cut the highs as well, pretty much i think it would just limit the available output to half the pots rotation.

if you want it brighter as you turn down, use a treble bypass cap between the input and wiper. as you turn up, no effect, as you turn down it will get brighter. you can do the opposite as well and put the cap from wiper to ground (i think, memory dodgy at the moment) and it wil get darker as you turn it down.

do a google search for both high and low pass filters and you can learn to tailor the tone to whatever you want anywhere in a circuit. a cap into a variable resistance is a low pass, a variable resistance into a cap is a high pass . or something like that.

[Electric Warrior]

I'll try putting a diode into my Fuzz Face clone today to see what happens..

When I tried this I noticed that a similar transistor wired as a diode worked better than the diodes I have. Maybe the same temperature coefficient, bigger surface, similar hfe and leakage, etc.

Yeah, to make up for temperature changes they should behave as similar as possible. Using matching transistors is probably the easiest way to achieve this.

a 500k resistor in parallel with a 500k pot would be a 250k load ultimately. that's what i use usually, it won't make it darker, but brighter.

Do 250k really make an audible difference? With a 0.01µF cap, the cutoff frequency of the high pass would be around 60Hz. The lowest fundamental of the guitar is just over 80Hz, so it would only affect sub harmonics.

When using a 250k pot like AnalogMan, even if the pot reads 20% lower than its marked value it shouldn't harm the fundamental much.

do a google search for both high and low pass filters and you can learn to tailor the tone to whatever you want anywhere in a circuit. a cap into a variable resistance is a low pass, a variable resistance into a cap is a high pass . or something like that.

it doesn't need to be variable.

High pass:


Low pass:


These calculators can come in quite handy: http://www.learningaboutelectronics.com/Articles/Low-pass-filter-calculator.php
http://www.learningaboutelectronics.com/Articles/High-pass-filter-calculator.php#answer1

in series to the 500k resistor will cut the volume a little bit, i would imagine. i dunno if it would cut the highs as well, pretty much i think it would just limit the available output to half the pots rotation.

It limits the the range of the pot, yes.
But it doesn't cut the high end. By putting a resistor in series with the pot you're just shifting down the cutoff frequency of the low pass filter. With the stock 500k it's at 32Hz, with 1000k at 16Hz.

It won't have an audible effect on the output volume either.

A volume pot is wired up like an L-Pad. The ratio of its internal resistances determines the degree of attenuation. The value doesn't matter This calculatur shows pretty well what's going on: http://www.sengpielaudio.com/calculator-voltagedivider.htm

It substitutes 0Ω with 1Ω (hello real world!), so if you were to simulate a really, really small cranked volume pot (like 5Ω) you'd get quite a bit of attenuation - but of course volume pots are always significantly larger than that.

[pinkjimiphoton]

quite true don't need variable resistances. i just suggested that cuz its easier to see what the circuits do with a pot than swapping resistors out.

i found 500k in a FF can be really muddy. 100k is noticeably brighter, but ya lose a bit of output.
i use 250k as a compromise, to me it seems to get enough overdrive volume (to me fuzz always sounds best if pushing an amp or other dirt box into overdriving) to keep me happy, but seems to lose about an octave of sludge.

i'm still formulating my own ideas on that stuff, but certain thngs seem to ring true... like to drop the range an octave, double a cap. to raise it an octave, cut it in half.

i buy decades of everything and when i'm really monkeying will try stuff up and down the scale to see what it does. if an order of double or 10 fold doesn't do anything, i will go up 100 fold or so. and keep monkeying til its fine tuned.

i am just barely beginning to grok some of this stuff i think. maybe by the time i'm dead

[Electric Warrior]

quite true don't need variable resistances. i just suggested that cuz its easier to see what the circuits do with a pot than swapping resistors out.

But you made a general statement about filters. If you've built a Fuzz Face with a pot it's easy to add another resistor as a mod. For understanding what's going on it's easier to look at the filtering and the volume attenuation seperately.

i found 500k in a FF can be really muddy. 100k is noticeably brighter, but ya lose a bit of output.
i use 250k as a compromise, to me it seems to get enough overdrive volume (to me fuzz always sounds best if pushing an amp or other dirt box into overdriving) to keep me happy, but seems to lose about an octave of sludge.

Yeah, with a 100k the low end sure is noticeably tighter. As you're losing part of the signal, of course it can't be as loud as with the full frequency range intact.

I now can see why 250k makes a difference. I was underestimating the filter's shallowness.

This is a graph of the frequency response of a stock Fuzz Face high pass between 20Hz and 300Hz:



I set my amps clean with fuzz, but they're low watters providing rather warm cleans.. Pushing them hard sure can make things really mushy..

[diydave]

hello, guys!
i`ve build fuzz with pre-gain (it works awesome) and input cap switching.

What's the "pre-gain"?

[Gus]

A thread from 2012
http://www.diystompboxes.com/smfforum/index.php?topic=96381.msg838899#msg838899
The sims used a LPB like circuit
You also need to think about the cable after the effect, the input loading of what is after the effect and the output resistance of the effect

[Electric Warrior]

So the volume pot is also part of a low pass? Interesting..

[goo]

What's the "pre-gain"?

that`s 250K input pot for smooth fuzziness instead of 1K pot for fuzz.
joe gagan`s variant of FF ("easy face").

[shredgd]

buffers work fine after fuzzfaces.

Not really, actually, in the way we are used to conceive a pedalboard.
Guitar -> 3 to 5 meters cable -> fuzz face -> 3 to 5 meters cable -> amp: sounds great!
Guitar -> 3 to 5 meters cable -> fuzz face -> short patch cable -> buffered pedal (e.g. delay) -> 3 to 5 meters cable -> amp: sounds dreadfully harsh!

Why? That's not actually the buffer's fault itself, but it's the patch cable's. A 3 to 5 meters cable has a capacitance to ground of about 300-600pF (depending on the cable quality/characteristics), a patch cable has only about 30pF of capacitance. This means the high impedance output of the fuzz face doesn't create the low pass filter it usually creates together with the following long cable to the amp.
There are two ways to solve his problem:
1. Using a long cable between the fuzz face and the following buffered pedal
2. Adding a toggle switch to our fuzz face to add a 470pF cap from output (central pin of volume pot) to ground, so we can "mimic" that cable capacitance whenever we are using a buffer after our fuzz face

To answer Artr's original question, the most common way to tame treble in a fuzz face is to add a small cap between base and collector of Q2. Look for the Eric Johnson FF model (27pF in it) or the Fulltone 70 schematic.

Giulio

*if you like some numbers: when the 500k volume pot of a FF is set to unity volume into a clean amp, there is about 250k of series resistance in the pot (I measured mine), which creates a low pass filter with a cut-off of about 1200Hz with a 470pF long cable, and a cut-off of about 18900Hz with a 30pF short patch cable... 18900Hz is about the limit of a young healthy human, so using a patch cable after a FF and before a buffer is like totally removing that important low pass filter!

[pinkjimiphoton]

i can most definitely see that as an issue with most silicon fuzzfaces.  and i guess it depends on your ears too.
not disagreeing. but a ge ff into a buffered delay can be freakin' awesome
i tend to usually run a buffered reverb at the end of my chain literally after the long run from my board just before the amp. works great. for ME, anyways

the q2 bc cap trick works great. but... why not make it variable? you can control the treble content pretty well from there by simply faking a variable cap.

use a 100k-250kish pot. use a very small value cap on one side, say 10p or so, and 2.2n cap on the other. tie the free end of the caps together,  and connect that to your c (or b, can't remember which is which, if it works backwards reverse the two connections).. connect the wiper/center terminal to b (or c ) and now you can go from virtually no treble cut to a quite noticeable treble cut and find the sweet spot. you don't lose the volume you get with a standard tone pot, cuz its active.
been doing this for a while, if all ya need is a bit of treble cut give it a try. great trick. maybe my only real contribution to society, even.

[shredgd]

i can most definitely see that as an issue with most silicon fuzzfaces.  and i guess it depends on your ears too.
not disagreeing. but a ge ff into a buffered delay can be freakin' awesome
i tend to usually run a buffered reverb at the end of my chain literally after the long run from my board just before the amp. works great. for ME anyways 

You're confirming what I say! The long run from the FF to the buffered reverb gives you the right amount of cable capacitance to keep that low-pass filter in action!
Yes, the issue is most evident with silicon fuzz faces, but my germanium fuzz also got a brand new "magic cap" toggle switch to use it before buffered pedals

PS: Nice idea for the tone control!

[pinkjimiphoton]

yeah, having a buffered delay back at the board going back to the buffered reverb at the amp really makes a huge difference in terms of noise floor etc, basically line driver i guess.

thanks for the kind words, i think if ya try that treble cut you'll really diggit. came up with it for my monkey balls pedal a ways back, and it really seems to work well and is fairly adaptable to a lot of circuits.

peace