What do you do when you get noise when fuzz = 100% ?

[LucifersTrip]

I've built a bunch of difference fuzzes and one thing has burned me more than anything.

Especially in the Fuzz Face variants, I get some type of wailing/hiss/noise/feedback when the fuzz is turned to 100%.
If I dial back to 95% it goes away, but I lose a bit of fuzz, sustain & treble. I probably have that problem with 1 out of every 5 to 8 fuzzes
I've built.

I've tried adjusting the bias & tweaked various resistors with little or no luck. The best luck I've had was dropping a relatively low cap (~.001 uf) across random points on the circuit unwanted noise vanishes. With the Tonebender I lucked out and the wail disappeared without changing the tone, but more commonly, of course the tone will change.

I've had some luck with using a slightly lower gain Q2, but that will sometimes lose best tone so I'll pull it.

Are there any other tricks for getting 100% without general noise on those hairy circuits?

thanx

[nbabmf]

Some silicon fuzz builds will squeal like mad because they are so high gain.  You just have to remedy that with a small value capacitor "across random points" haha.  Harbor Freight sells a cheap ($3.99) multimeter with an hFe setting so you can choose lower gain trannies (if they're silicon), but sometimes you still have to use the cap.

[yeeshkul]

Also, the proper wiring can help - keep distance between the output and input wires - that applies for your PCB as well. Hi gain units don't like to have in and out parts too close.

[earthtonesaudio]

1. Make sure the fuzz can't get to 100% with a small value resistor in series with the gain pot
2. small value caps in strategic* (surprisingly actually not random) locations. 
3. small resistors in strategic locations, note R7 here



*Generally you want to place the cap to make an integrator.  So (for a Fuzz Face) either between collector and base of the first or second transistor, or between the base of the first transistor and the emitter of the second transistor.

[kupervaser]

Did you build a NPN version? They do these kind of things occasionly. You can try the search function and you will find lots of threads. I beleive it was a capasitor across collector and base of the transistor.

[petemoore]

wailing/hiss/noise/feedback when the fuzz is turned to 100%.
  That's probably it, self-oscillation, another ''invention'' the FF's can discover.
  Many ways to control gain, the B/C capacitor is a way to reduce the HF gain, often this is where...a bit of gain reduction allows otherwise high gain to not oscillate. Axis Fuzz has added capacitor across the fuzzfaces 'little' resistor [top of most schematics: across the Resistor between 9v and Q2C resistor].
  Could try either one, or both, trick for keeping 'fuzz' [which is HF distortion element] is to make it just large enough to quell the HFs from being out of hand but not change the circuit eq too radically. Otherwise makes fine 'distortion' tone.

[Ronsonic]

I don't know that making a Fuzz Face unconditionally quiet and stable with the fuzz turned all the way up is a worthy use of time.

That isn't how they are generally used and if it's an issue for you some other design is probably a better approach.

Just my two ...

[LucifersTrip]

Also, the proper wiring can help - keep distance between the output and input wires - that applies for your PCB as well. Hi gain units don't like to have in and out parts too close.

It's not boxed yet. The input/output are spread out on the floor a couple feet from each other....but I do separate when boxing. thanx

1. Make sure the fuzz can't get to 100% with a small value resistor in series with the gain pot
2. small value caps in strategic* (surprisingly actually not random) locations.  Wink
3. small resistors in strategic locations, note R7 here

1) yes, the easiest solution but don't want to lose the fuzz, sustain, treble
2) that was my best luck so far...i know not "random", but for me it is since I don't know circuit theory well
enough. On the one I'm currently testing [see below], all cap placements so far reduced fuzz quality
3) thanx...that's something I didn't try yet

Did you build a NPN version? They do these kind of things occasionly. You can try the search function and you will find lots of threads. I beleive it was a capasitor across collector and base of the transistor.

It's actually a mos version. I'll definitely continue to try other cap locations.

  That's probably it, self-oscillation, another ''invention'' the FF's can discover.
  Many ways to control gain, the B/C capacitor is a way to reduce the HF gain, often this is where...a bit of gain reduction allows otherwise high gain to not oscillate. Axis Fuzz has added capacitor across the fuzzfaces 'little' resistor [top of most schematics: across the Resistor between 9v and Q2C resistor].
  Could try either one, or both, trick for keeping 'fuzz' [which is HF distortion element] is to make it just large enough to quell the HFs from being out of hand but not change the circuit eq too radically. Otherwise makes fine 'distortion' tone.

yes, reducing hfe was the first thing I tried, but I don't think I have as many options with a mos. I will be reducing the
cap value in one of the "working" positions to see if I don't lose quality

I don't know that making a Fuzz Face unconditionally quiet and stable with the fuzz turned all the way up is a worthy use of time. That isn't how they are generally used and if it's an issue for you some other design is probably a better approach.
Just my two ...

Exactly...I expect some & even higher noise than other fuzzes, but in this case turning up to that last 5% (or only last 2-3%) is too noisy to be usable. I don't want to completely eleiminate it, just get it to a "normal" level without losing fuzz quality.

thanx for all suggestions. One of the  things I haven't tried yet is adding a strategic resistor.

There's little discussion on this forum about the Mos Face, especially about what I'm asking. This one had some info:
http://www.diystompboxes.com/smfforum/index.php?topic=2597.0

[LucifersTrip]

*Generally you want to place the cap to make an integrator.  So (for a Fuzz Face) either between collector and base of the first or second transistor, or between the base of the first transistor and the emitter of the second transistor.

Update:
A 220pf cap across C and B of Q1 did the trick on this one. It removed the wail with ZERO change to the tone. I tried lowering to 22pf and it barely removed anything.  Btw, I forgot to mention that I used a 2N7000 for Q2.

thanx alot

[us_jim]

hi, i tried to do this "mod" a 224pf cap between B e C of Q1, solve the hiss issue, but i lose volume and gain...
it's normal?

ah... i use 2n2222 and bc108/bc107 transistor

[duck_arse]

hello and welcome us_jim.

can you clarify the value of the cap? do you mean 220pF or 220nF (224)? if you search-up the "tantalus", or the "silliface I or II", or the "great cheddar", or the "axis face" and etc, all sillicon fuzz-face variants, you'll see mostly Q2 has the B-C cap. there are some other ideas for hiss reduction shown in a couple of those circuits too.

[mth5044]

Don't forget the importance of PCB layout. Even a PCB from shop can be laid out 'poorly'.

[R.G.]

What's going on is that there is enough high frequency gain - that is radio-frequency gain - to get into oscillation above audio. This needs attacked two ways - power supply decoupling and killing circuit response above audio. Actually three ways; layout, as noted, can undo all the good that any parts-connection design does.

Using a MOSFET implies including a device that has useful gains from 500MHz - 1GHz in most cases. These things are FAST. Bipolars have frequency limitations in most cases that are much lower. A MOSFET used as a follower can self-oscillate from capacitive loading on the source. It simply sits there counting its own toes at 800MHz. This can be too high a frequency for most 20MHz oscilloscopes to even see.

To fix the problem in the generic sense, you (1) make the power supply and grounding immaculate with both layout and decoupling, and (2) use bandwidth limiters on the devices to kill the really high frequency gain.

In the case of the MOSFET, what really helps is a gate-stopper resistor. This is simply a resistor in series with the gate as close to the physical gate lead as you can get it. The resistor works with the gate-source capacitance to limit high frequency gain. By having very short lead length to the gate, it also limits the ability of other signals to couple to the gate and cause oscillation. Try 1K or even more. If this simply moves the oscillation down, you also have to work on other things. Killing the high frequency response of the first transistor with a C-B capacitor helps too, but you're still open to funny RF/black-magic issues with the MOSFET.

Apparently killing the bipolar RF gain was enough. So keep the gate stopper in mind for the next time you use MOSFETs.

[us_jim]

thanks to all...

made a mistake, i used 224 cap means 220nf. i'm going to change it with a new one. 120pf () i think!

also try with 1k input res, after this will leave my feedback about this.

i made this fuzz face from a stripboard schematic found on the web.

i have another 2 questions but i think to open a new one post

thanks again

[deafbutpicky]

A small resistor from E to ground on Q1 might help too...

[us_jim]

will try also with the res in your position...
months and months with this f fuzz face and can't be able to makes it sounding good!
now my ff lose lot of fuzz... also if i remove the cap across transistor.
would like to take all and put out of the window 

[DDD]

The most frequent cause of such feedback is a parasitic capacitance of a stomp switch.
Putting a small cap from base of the first tranny to ground helps a lot.
Capacitor value is usually from 200 pF and up to 1nF (for low-impedance inputs such as as Ge Fuzz Face have).
This capacitor do not involve the tranny into the process so we can treat it as the most "clean" way to reduce the feedback.