Positive ground reversal in roger Mayer Classic fuzz question

[pete_g]

I've made a roger Mayer Classic fuzz and it seems it can't use a daisy chain of an adapter, so what I'm wondering is can u reverse it so it is a negative ground and if so what needs to he done? Thanks

[dwmorrin]

Flip C1 as well (+ side towards Q1 base).
A good "ground flipping" reference: http://www.muzique.com/lab/fuzzface.htm
A lot of people report problems when doing this (oscillations) so it may or may not work for you.

[R.G.]

Note that you'll have much better luck if you put a big (100uF or bigger) electrolytic the right way round from power to ground, paralleled with a 0.1uF capacitor.

Even then, you may not be lucky. Or you might.

[chi_boy]

How about using a MadBean RoadRage as an inverter??

http://www.madbeanpedals.com/projects/RoadRage/docs/RoadRage.pdf

[pete_g]

Flip C1 as well (+ side towards Q1 base).
A good "ground flipping" reference: http://www.muzique.com/lab/fuzzface.htm
A lot of people report problems when doing this (oscillations) so it may or may not work for you.

Do I have to flip C4 around also?

Ive read the above article and Ill give it a shot

[dwmorrin]

Do I have to flip C4 around also?

No, C4 is directly connected on one side to +9V.  Keep the + side going towards the +9V, - side on the fuzz pot.

[R.G.]

It's worth noting my standard warning here: it may or may not work. There are poorly understood things going on with ground/power reversals. Sometimes they are intractable.

Using a charge pump inverter like the LT1054 to make a local -9V supply and let the circuit use a solid ground is a much more certain way, and sidesteps the issues of what gets reversed and so on entirely.

[dwmorrin]

There are poorly understood things going on with ground/power reversals.

The mystery seems to be as old as the hills.
What's stopping us from figuring it out?  Is it really completely random, like 2 people could breadboard the same reversal and get different results?  Or is there a trick like avoiding certain transistors, leakage, etc that we just haven't uncovered yet?
(Tricks other than the power supply caps.)

I have no desire to own a "neg. ground converted fuzz" for personal use, but I've been tempted to breadboard them just to try and study this peculiar problem.

R.G., did you ever mess around with trying to solve the riddle?  Not the "fixes" but the mysterious cause?

[pete_g]

Thanks guys, it worked. Just had to make sure lug 3 on the fuzz pot was connected to power.

[R.G.]

The mystery seems to be as old as the hills.
What's stopping us from figuring it out?  Is it really completely random, like 2 people could breadboard the same reversal and get different results?  Or is there a trick like avoiding certain transistors, leakage, etc that we just haven't uncovered yet?
(Tricks other than the power supply caps.)

I have no desire to own a "neg. ground converted fuzz" for personal use, but I've been tempted to breadboard them just to try and study this peculiar problem.

R.G., did you ever mess around with trying to solve the riddle?  Not the "fixes" but the mysterious cause?

Actually, I did. I've posted on this topic here many times. I ran into it in about 1990 when I built a custom distortion for a player, and with no trepidation at all, did the swap for power and ground. It oscillated. I'd done this before, several times, so over about two weeks of spare time tinkering ( I was working for TIC at the time) I tried everthing I could think of, and that was a lot. I tried everything I've ever seen on the topic, and many I've never seen disclosed. It oscillated when the power and ground were inverted, and promptly quit when they were put back.

To cut a long saga short, what's stopping us from figuring it out is that it's hard. I've tried all the simple and many of the abstract ways of both fixing it and of instrumenting such circuits to find out what's wrong. I'm fairly certain that it's tied up in the impedance between power and ground, where the internal circuit thinks that "ground" is its power supply and its "ground" is what the external circuits think is power. If the impedance between those is zero, then no oscillation is possible (from this cause, at least; there is a forest of other opportunities).

Oscillation requires  forward gain, a feedback path with a loop gain (i.e. the total gain from input, through the amplifier, and back through the feedback path to the input again) of unity or more, and phase inversions to make the feedback at the input be N*360 degrees at some frequency. In the real world, there is always a feedback path; the only question is how much it attenuates the output signal. If this is less than the forward gain, oscillation doesn't happen. Where you get into trouble is increasing the gain, and the battle cry of effects hackers is "More gain!!". As gain increases, you have to work on making the many feedback paths be higher attenuation, by moving things further apart, shielding, careful arrangement of wires and traces, using more or different wires (i.e. star ground/power) to eliminate current-induced feedback voltages from the grounding traces, enhancing power/ground decoupling, and in RF work, elimination of antennae, both sending and receiving. It's not widely appreciated, but cell phone RF would probably be impossible without SMD parts to keep the wires (antennas!) between parts short.

There is a limit on what complexity can be spent on a problem. When it's simpler to do it another way, you're wasting time, money and energy on solving it from first principles. This bites with a vengeance on this particular problem. If a couple of caps and some shielded wire fix it, GREAT. But the complexity of the solution can't be allowed to exceed that of a one-IC charge pump. In fact, the one-IC charge pump is so good, it's tough to justify anything else. It's the old issue of "Doc, it really hurts when I do ... this..." to which the doctor replies "Well, don't do that".

Worse yet, the problem is inconsistent. Enough people get away with it that in spite of my best warnings, the "technique" remains alive in the memes of effects culture. So people poo-poo the issue, read about success stories and wade into the water. The either get lucky or unlucky. The original poster in this thread got lucky, so he is now and forever a believer that if he just tinkers enough, it'll work fine. And that's where I was, until I ran into a couple of situations where I simply could not make it work, and neither could my professional-EE-circuit-designer-friends who were mostly ham radio operators for fun in their spare time could not help me solve it either.

So go dig in and mess with it if you want. But be aware that it's layout specific as well. Doing work on a breadboard may or may not translate into success inside a vero, perfboard, or PC inside a box with wires. It's tempting to arm oneself with the attitude that we'll just go beat on this til we solve it. But that can be a very costly exercise, and it's one where there has been a lot of work expended already.

So that's why I warn people - as well as why they persist in just wading in, then popping up here with the question.

[zombiwoof]

I have a Voodoo Labs Proctavia that requires an isolated power supply, so I either just use a battery (don't know if you built your pedal to use a battery or not), or just plug in a separate adapter for that pedal.  IMO Fuzz Faces sound better with a battery anyway, usually a cheap 9volt is suggested, so that's what I use with my Fuzz Face.

Al