Fuzz Face warble problems

Started by RicF, June 22, 2004, 06:40:23 PM

Previous topic - Next topic

RicF

For my first pedal project, I tried a germanium transistor Fuzzface, using a pcb from General Guitar Gadgets and transistors from Small Bear (PNP, negative ground). It works, and doesn't sound too bad, but I get some oscillation when the fuzz is all the way up, which seems very sensitive to the pickup combination and volume setting on my guitar. It's almost like a switch, when the guitar volume (strat) hits about 8, instant warble. I used shielded  wire on the hot lead coming from the input jack to the footswitch, and same thing from the switch to the output, but I'm guessing this isn't enough. I'm used to building tube amps, so I know the importance of lead dress, but I don't know much about high-gain transistor circuits. Can anyone offer some debugging advice?

Ric

Lonestarjohnny

It's not common for a Germ. Face to oscillilate, a Silicon fuzz face, yes, maybe one of the guys that has done this build will post some voltage reading's for you to compare.
JD

R.G.

I tell you three times:

The PNP/negative ground connection in a distortion cannot in general be relied on to not oscillate.

Sometimes, maybe even most times, you get away with it, but there are inherent difficulties in the connection that can rise up and make your circuit oscillate.

I would be surprised if you got oscillation if you rewired to positive ground. It might happen, but I bet not.

There are are two kinds of people: (1) those that have had problems with the PNP/negative ground setup and (2) those that haven't - yet. People who have not - yet - had problems will swear that it's flawless and works every time.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

RDV

A PNP FF with positive ground uses so little battery juice that I've never quite seen the point of making them negative ground. I've built a couple and they still have the same battery in them for about a year. They're just so much easier to deal with, and an FF of either type(Ge or Si) sounds better with a battery to me IMNSHO.

RDV

bwanasonic

Quote from: RDVA PNP FF with positive ground uses so little battery juice that I've never quite seen the point of making them negative ground. I've built a couple and they still have the same battery in them for about a year. They're just so much easier to deal with, and an FF of either type(Ge or Si) sounds better with a battery to me IMNSHO.

I second these sentiments.  I originally went the PNP/ Neg. ground route in my first FF attempt, only to get strange low pitched warblings/oscillations. Started from scratch with a pos. ground design and have never looked back. Use carbon batteries and they will last a ridiculous amount of time.

Kerry M

Lonestarjohnny

I've never had that problem, what I always hated was an afternoon gig where the temp changed a lot, the tone on my old Germ F/F always changed right with it for better or worse and 99 % of the time it was the latter,  :lol:
JD

RicF

Okay, I'm sold. This weekend I'll convert it to positive ground. That'll give me an excuse to clean up the leads too. Any other advice to avoid hum, etc? Thanks a ton for the help so far guys, I've learned a lot reading through the posts here. 'Course, now I want to build absolutely everything.

Ric

bwanasonic

Quote from: JDI've never had that problem, what I always hated was an afternoon gig where the temp changed a lot, the tone on my old Germ F/F always changed right with it for better or worse and 99 % of the time it was the latter,  :lol:

Soon after I had finished my DIY Ge FF, I was all exited to debut it live. But sure enough, our next gig was an outdoor *tent* affair, with the sun managing to find it's way to my pedalboard. Well, at least it was a good lesson on the thermal sensitivity of Ge trannies.

Kerry M

Paul Perry (Frostwave)

Before u rip it up, try a large electrolytic across the battery connections to the board. Say 100mfd or more. And a .1 ceramic bypass across there as well for luck.

Paul Perry (Frostwave)

Before u rip it up, try a large electrolytic across the battery connections to the board. Say 100mfd or more. And a .1 ceramic bypass across there as well for luck.
I think the problem is, any small resistance in the power switch gives a feedback effect. Equivalent to a dying battery.

phillip

QuoteSoon after I had finished my DIY Ge FF, I was all exited to debut it live. But sure enough, our next gig was an outdoor *tent* affair, with the sun managing to find it's way to my pedalboard. Well, at least it was a good lesson on the thermal sensitivity of Ge trannies.

Do the reverse biased Germanium diode trick (nearly at the bottom of the page):

http://www.geocities.com/SunsetStrip/Studio/2987/britface.html

That quick and easy little trick will allow the transistor to take quite a bit more heat before bad things start to happen.  It says to perform the mod to the input tranny, since Q2 has the pot on the emitter to adjust things.  The transistor and diode both have to be Germanium.

Phillip

Ge_Whiz

I've just completed a silicon FF with the 'Miss Piggy' mod. First transistor has a gain of 200, second one has a gain of 700. This gives me 'hot' mode. For 'vintage' mode I switch in another 200 gain transistor across the emitter-base of the second transistor. I haven't measured the gain of the combination, but I swear to you I cannot tell the difference between 'vintage' mode and the sound of my Ge FF, except the silicon doesn't give me 'Deutsche Welle' radio when I turn down the guitar volume.

RicF

Just a follow-up; yesterday I rewired the pedal for positive ground, and bingo, no more LFO problems. I'm not sure why the recommended way was negative ground, the positive ground setup isn't any more difficult.

Now the problem is, the pedal sounds way too muddy. It's only useable with the bridge pickup and treble settings on the guitar and amp maxed. Is this a common thing with this pedal? Do I need to twiddle input/ouput cap values? Also just for kicks I swapped the transistors (moved Q1 to Q2 and vice versa) and didn't notice any change in sound. But then maybe I wouldn't through all that mud.

R.G.

QuoteI'm not sure why the recommended way was negative ground, the positive ground setup isn't any more difficult.
The negative ground recommendation was based on the growing trend to use an external power supply for a number of pedals by simply paralleling the 9V supply and daisy chaining the power jacks. This is very convenient - and introduces other problems as well.

A positive ground pedal in a sea of negative ground pedals offers a power problem. You have to provide a separate power supply, either AC or battery, for the positive ground pedal to keep from shorting out the positive ground and negative ground power supplies. It's a pain. I think that's why people want the convenience of the negative ground connection.

Like most conveniences, there is a cost. In this case, the cost is signal current running on a power line that isn't intended to carry current. That's where the feedback that causes the oscillation comes from.

The simple single ended circuits we use in effects have almost no power supply noise rejection, so anything on the power line gets reflected into the signal path.  While power and ground *should* be the same potential for AC signals, in fact, they're not. Batteries have some internal impedance, that is bigger as the battery gets older. We can't count on batteries to always "short out" any signal differences between power and ground.

AC power supplies that include a regulator do better, because the regulator drops the impedance at the regulator to a few milliohms. But the wires and jacks in the daisy chain have some impedance too. It doesn't take too much impedance to make a few millivolts of signal fed back to the input of a distortion pedal that has a gain of thousands.

Sometimes, as Paul suggests, you can locally decouple the power supply of a negative ground wired positive ground circuit by adding a local decoupling capacitor. Putting a BFC (Big Freaking Capacitor) across the 9V supply in a negative-ground-wired-positive-ground-designed pedal will often drop the local impedance of the power supply enough that power-line feedback will drop below the point that maintains oscillation. Often.

Many times. Usually. Most of the time.

There are hard cases where it won't. I used to be a believer in "I can always make it work" power supply design, as I used to design these things for a living. A modified Jordan Bosstone made a believer out of me. There will always be some combination of device gain, layout, and wiring that can make the reversed power supply connection oscillate.

So I dropped back to the practical view: if the reversed power wiring doesn't work for your situation, put it back as it was designed.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

jrc4558

Change the output capacitor to 0.01µF and the volume pot to 100kOhms.

RedHouse

Not wishing to fly in the face of those here who are much wiser than myself, but I'd like to contribute my 2¢ worth and possibly save you some re-build time.

I'm not saying I haven't had "issues" but IMHO changing the ground scheme should be like way at the bottom of the list (*I'll tell you why later) but I certainly have run across my share of oscillation issues but find that usually a part value change or addition of one or two parts will fix the issue.

I find that adding a 100pF monolithic-ceramic capacitor (56pF minimum) across the collectors of Q1 and Q2 usually stops 98% of oscillation issues. This is the one, single, best, anti-oscillation adjustment one can make to a Fuzzface circuit be it positive or negative ground.

Another place that seems to contribute to oscillations (which I would describe as "motor boating") is Q1's collector bias resistor (specified as a 33k resistor) often needs to be increased to 47k or greater, specially true with Si based circuits.

Now-a-days I always build negative ground Fuzzface boards unless one specifically requests a "vintage" positive ground board, and I get 100% of my Negative ground circuits working correctly through the use of my "test jig".

It's been mentioned here before, ad-nausium, but one really should build a perf-board "test jig" fuzzface circuit, there are very few parts and it certainly wont break-the-bank to build but the return on the investment is invaluable and enables you to build 1 or 100 fuzzface circuits... that work, every time.
(Perf-board NOT Proto-board)

Build it with sockets for the transistors, use 50k pots in place of Q1 and Q2's collector bias resistors and throw a 20k in front of Q1, and if you want to you can use a 50k resistor and 100k pot in series replacing the 100k feedback resistor. Using this "test jig" you can absolutely "dial-in" your circuit.

You WILL find that more often than not, you can dial-in a fuzzface to work perfectly BUT it quite likely WILL NOT use the part values originally specified on the schematics.

In fact after building nearly 2 dozen fuzzface's, NONE were able to use 100% of the originally specified part values, at the very least Q2's bias resistor must be modified to provide the correct bias.

Many times they can burst into oscillation prior to "tuning" or "dialing-in" the parts values, but I would not be quick to attribute it to the ground schema and here's why I say this...

* I made my perf-board "test jig" so it has both Positive or Negative ground options and when oscillation issues have arisen I have found that merely changing ground schema's DID NOT fix the issue, there were part values that needed tweaking to fix the oscillation issues.

RicF

Fantastic information for a beginner like me. Thanks again guys. Like I said, I did rewire it for positive ground (no other changes) and that got rid of all my oscillation problems. But the sound quality is still not at all what I hoped for so I'm definately not done messing around with it.  The "test jig" sounds like exactly what I need to do.  As usual, I need to spend a whole lot of time reading and experimenting.

Paul Marossy

I built a neg gnd Boutique Fuzz Face and I have never had any problems with it. I guess I got lucky on this, my first Fuzz Face circuit - it worked exactly as expected and had the sound I wanted. I did test all of my Ge transistors for leakiness first, though. I have tried it with both Si & Ge transistors and I like both of them equally as well. Si transistors are much more stable with changes in temperature.

Anyhow, what I meant to say is that Joe Gagan suggests using a Si for Q1 and a Ge for Q2. This gives you a lot more stablility with changes in temperature while retaining some of the Ge flavor. It also makes for a "more comlex tone", which I think sounds really good. There used to be a blurb about all of this at the old Nine Volt Nirvana website, but the website has changed since I was there last.