HEY ALL YOU FUZZ BUILDERS...

[brian wenz]

Hello Hello--
 I'm interested in getting  everyones experience on this-----
 Using 2-tranny and 3-tranny fuzz circuits as a point of reference  [ Fuzz Face, Tonebender, etc.]  and assuming everything else in the pedal is equal [layout, wire, enclosure, metal-film res., good quality caps., etc.], what do you consider to be the main source of the "running water" type of hissing that you hear  when the volume and fuzz is set pretty high and the volume control on yer guitar is all the way OFF??    Transistor type?  Bad layout??  Some circuits are just noisier then others??
I know with germ circuits the leakage of the tranny can make for a pretty noisey "idle" [also when the guitar's volume is turned up.....] and I'm wondering if the same logic transfers over into silicon circuits in general.
[Actually, some si circuits are seem pretty "whooshy" , too.]
Thanks!
Brian.

[rx5]

Hi,

Ill refer to the 2 tranny fuzz... My quick guess is maybe you have interchanged Q1 for Q2 and vice versa.... :?  putting a higher gain in place of Q1 than Q2 would sound something similar like youve mentioned...

Q1 gain must be 70-80 and Q2 120-130......

hope this helps....

[Paul Marossy]

In my experience, the biggest source of noise is the transistors themselves. General purpose silicon transistors can be very noisy in an audio circuit with high voltage gain. That's why something like a 2N5088 or 2N5089 are good to use. As far as Germanium transistors go, I think it depends on the type you get. I built the Boutique Fuzz Face at GGG no too long ago, and the Ge transistors I used are actually very quiet. But the tone is sometimes inconsistent, due to the sensitivity of material to ambient temperature. Ge is very sensitive that way.

[petemoore]

At least not any two that I've tried.
 I Bl Factis the Ge's can be noisy...the TB and FF with the matched sets are reasonabley 'normal' [not very noisey] noise level, Alot depende on the transistors...IME
 Gonna hafta try a couple Ge builds with 1% Resistors...just to see...take the same tranny's and place them in an otherwise same ckt but with the 'quieter' resistors.
 Comments on 1% resistor builds noise levels welcome..

[Paul Marossy]

Closer tolerances on the resistors may help some, but I still think it's mostly the transistor itself that generates the noise. Unless they are carbon comp types, but I don't think that would make any difference in a 9V DC guitar effect circuit. Check at that carbon comp article at GEO. It's very interesting.

[petemoore]

I'm running the longest efkt chains in my history.
 The noise floor is as low as ever.
 I attribute much of this to using batteries [mostly] and not chaining PS's.
 If I can't get the noise out of an efkt it gets canned...extra allowances made for high gain Fuzzers, as it's a balance there [noise to gain/Fuzz ratio]
 Nice to have the crunch fade to sustain that fades to taper that fades to nearly nothing...

[nightingale]

hey brian~
i've learned to stay away from 2n2222's they seem to be super noisey in my expeirments... the only high gain circuit that seems to be an exception is the jordan bosstone, maybe this has to do with the circuit using npn, and pnp tranny's...

i know both us of really like our fuzz circuits... the best thing i've run across lately is a bosstone with a big muff tone control spliced onto the end of the circuit...(tone pot wired to volume pot)... i drew up an express PCB board, but havn't tested it yet... lemme know if you want to check it out... there's loads of output, and nasty fuzz in the stock bosstone.... so it seems  to tame the circuit just about rightvIMO...

[smoguzbenjamin]

Wierd, I've never had problems with 2n2222. Probably 'cos I haven't used 'em much But I built the LPB-1 with a 2n2222 transistor and have had no problems, even after trying some clipping diodes.

I took the diodes out 'cos I didn't really enjoy the sound. 'twas too harsh for my sensitive ears

[brian wenz]

Hello Hello-
  Yeah, I think some of the problem is the trannys themselves......also the tip about putting the lower gain tranny in Q1 is something I'm gonna experiment with [and take notes on!] 'cuz lately I've been putting the higher gain in Q1.  
Bear in mind,  what I'm talking about is the "running water" sound I hear alot when the volume and fuzz pot are up pretty high and the guitar volume control is on ZERO.  Other circuit fizz and noise that creeps in when the guitar volume is turned up is a different matter.
Anyway, thanks and I hope more people chime in with their experience!
Brian.

[Joe Davisson]

When the guitar volume is down, the guitar cable becomes a small coil (inductance) to ground. There is also some capacitance. Ever see a crystal radio schematic? There is a small coil and capacitor to tune the desired station. The guitar cable acts the same way.

When the volume is up, the capacitance and inductance of the guitar cable is in series with the volume control, which has a large resistance. This blocks the noise sufficiently.

Inserting a 1k resistor before the input cap can usually lower the noise without botching up the tone too much.

[Paul Marossy]

That is a very good point Joe. I think that would explain the noise problem more than just noisy transistors, especially since the guitar volume is all the way down.

[R.G.]

Noise comes in two main classes: resistor noise and junction noise.

Resistor noise consists of thermal noise, which is an irreducable factor determined by the resistance and the absolute temperature of the resistance, and excess noise, which is determined by the "sloppiness" of the material and connections. Wirewound resistors come closest to having only thermal noise. Metal film is a near second. Carbon film is a bit worse, but usable for almost everything. Carbon composition resistors are notorious for excess noise, some manufacturers being worse than others.

For lowest noise, use low-value metal or metal film resistors - and keep your circuits cool.

Junctions make several kinds of noise. At low currents, there is noise from the fact that the electrons themselves are quantized, so there is the rain-on-the-roof effect at very low currents. At higher currents, the basic noise of the junction predominates. Transistors exhibit flicker or 1/f noise as well as white noise hiss, and burst or "popcorn" noise. The "running water" noise you hear is midband 1/f noise. It's a characteristic of the transistor.

To a very great extent, transistor noise is determined by two things: care in manufacturing and the history of the specific part.  Transistor noise used to be bad because the germanium or silicon chips were inadvertently let get contaminated before getting packaged, and the contaminants would form noisy leakage paths around the chip. Today's glass passivated chips, high-cleanliness fab rooms and all-automated heremetic packaging have largely removed this as a factor from modern transistors. All bets are off with NOS TO-1 Germanium devices, though. Some are OK, some are very bad.

Transistor history can make a quiet transistor get noisy. See "When good opamps go bad" at GEO.

There is a very long history of study of noise in amplifiers. For a short cut, I'll list some of the results:
1. Make the input stage as quiet as you possibly can. Its noise, whatever it is, will be amplified by all following stages. This is especially important for gain-of-a-million distortion pedals.
2. Is really a corollary to 1 - make the biasing on the input stage as quiet as you can. If you are designing, make this "noiseless" biasing; if you are paint-by-numbering, do not use carbon comp resistors.
3. Never, never, ever reverse break the base-emitter junction of bipolar transistors unless you want them to be noisy. Consider putting a reverse biased diode across the BE junction to prevent the junction from ever being broken over.

To sum it all up, if you're using carbon film or better resistors on someone else's circuit design, the problem you hear is in the transistors. The front end of the circuit is the most critical. Newer manufacture transistors are quieter. If at all possible, use noise-rated transistors like the 2N5088 and 2N4250. MPSA18 is also quite good.

[brian wenz]

Hello Hello--
   I've been using a lot of newer Fairchild Bc182, 183, and 184 in Fuzz Face and Tonebender  [2 and 3 tranny circuits] .  I guess I'm gonna have to put sockets in 'em and test out the trannys for noise.  It seems like keeping the gain down on Q1 would help, also.  As stated in my first post, I use metal film resistors and usually shield the input and output wires, too.  Some circuits are just NOISIER, though, and there is a certain percentage of "gonna-have-to-live-with-it" attitude I have to adopt!
So, the 2N5088 is low noise??  I sort of avoided it 'cuz of it's high gain properties [also, I don't remember if I liked the tone...], but maybe I should look into it some more for certain circuits.
R.G.---What is "reverse breaking" the base-emmiter junction?   I will be using any-and-all methods to cut down on noise, so I'm soliciting all the tutoring that anybody cares to offer!
Joe-- Yeah, thanks for reminding me about the 1K before the input cap. Some of the circuits actually have a resistor  [10K, 33K.....sometimes even a 100K!] already there....and some of 'em are real noisey! I'm gonna play around with that some more, though.
HEY RYAN!-- I  really like the Bosstone!!  The one I made is actually pretty quiet, too!   If you have any mods or  "design tips" I'd like to see 'em! Thanks.
Paul--Are the 5088 and 5089 quieter then the low-noise BC's??
Thanks  everybody--
Brian.

[Rick]

My first Tonebender build sounded like rain. It was a constant hiss. In my case it was the GE transistor, a very leaky AC128. Once swapped out for a better tranny, the hiss was gone. Socket your GEs at least for testing and you are saving mucho time.