Positive vs. Negative ground in Fuzz Face

[MikeH]

Is there any difference in sound between the positive ground and negative ground fuzz face versions?  The tone pad pcb allows either design, and I'm trying to decide which one to go with.  Has anyone out there tried it both ways and heard a difference?

[petemoore]

Is there any difference in sound between the positive ground and negative ground fuzz face
  One is better than the other in blind tests?, I use a NPN ones , have built PNP FF's...these sound great.
  I can't say the difference between PNP Pos Gnd and NPN Neg Gnd, soundwise, is very great.
  I noticed at least the same amount of difference between a PNP FF and a PNP FF [pos. Gnd. of course].
  I remember reading something of the technical reasons, but I can't remember the theory or whether it could be test/proved.
  Mostly just 4 nodes need touched, swapping the +/- on polarized components to convert to other ground scheme. I socketted the 22uf on one board for just such occasions, also structuring V+/- as accessable above board for easy reverse PS re-works.
 
 
 

[captntasty]

IMO the difference between + & - ground doesn't have so much to do with the sound as it does the behavior of the effect.  I have built 2 FFs and a "Boutique" FF - all neg ground.  Lot of problems with oscillation (unwanted crazy spaceship noise, unless that's your thing).
I swapped one FF to pos ground and plan on getting around to swapping the ground on the others also.  Absolutely no oscillation problems.  Pretty sure there are workarounds for oscillation on neg ground FF's, a search on this site would reveal tons of great info that I simply am not qualified to speak about.  The newly pos FF is now one of my favorites.  I think the idea behind neg ground is that you can use it powered by common DC adaptor just like most other effects whereas pos ground requires a DC adaptor with different polarity.  Neg ground introduces something into the circuit that the original design doesn't like, thus the oscillation problem - not a very technical explanation, again a search will reveal TONS of info about neg ground, oscillation, etc...  Anyhow, I didn't notice a difference in the sound of the effect from neg to pos but the pos behaves better... it seems the neg ground FF is the wild child in the family and needs some discipline

[DuncanM]

Pretty sure there are workarounds for oscillation on neg ground FF's, a search on this site would reveal tons of great info that I simply am not qualified to speak about.

The standard workaround for a oscillating negative ground fuzz face is to stick a big capacitor across the battery terminals. I ended up using a 1000uF electrolytic. That should sort out most problems, but there are still some builds that misbehave even with that done - apparently...

Positive ground is, by all accounts, inherently more stable.

[R.G.]

I have a long history with positive and negative ground FF's.

There are some circumstances under which a PNP FF build set up with a negative ground cannot be made to quit oscillating. But people just keep on trying it anyway. Sometimes, even most times it works. But there are some situations where it can't be made to work.

My best advice is if you use PNPs in a FF, use positive ground. If you use NPNs, use negative ground.

[RedHouse]

This is one of those places in life where the rubber-meets-the-road so to speak, where Electronics walk's-the-walk and may not talk-the-talk. I had to find the answer to this question and ended up building a testbed board that has reversible ground.
(it's a PCB, there are traces on the other side)

In my experience with this board, if it oscillates with (-)Gnd it also oscillates with (+)Gnd.
(YMMV)



In the lower left of this pic you can see a 3-pin socket which is for a LM7809 and there is also a header/socket with black-n-white leads this is a reversible connection, next to it you can see the electrolytics are in sockets so they can be reversed.
(the 6-pin sockets and trimpots below them are for dialing-in Brett/Piggyback transistors)

My test unit has shown me that it just doesn't matter very much, ...if it squeels, it needs power supply bypassing (PNP or NPN) and usually also needs a 100pF cap between the base of Q1 and collector of Q2.

Generally the FF is a unique circuit in that except for the electrolytic caps it doesn't really know what/where ground is (power ground and system grounds are isolated by caps) so it works equally well as +/- ground (with proper layout and power supply bypassing, if you care see Jack Orman's website for more on this)

[R.G.]

In my experience with this board, if it oscillates with (-)Gnd it also oscillates with (+)Gnd.
(YMMV)

That's very true. My mileage has varied from that.

It's not that I'm uninformed, short on formal theory, unskilled at debugging, or unable to find and fix problems - but I have had multiple instances over the years where a PNP circuit with negative ground has oscillated, and no amount of bypassing, lowering power supply impedance, dinking with RF bypassing, ferrite beads, isolating ground leads, star grounding, killing RF response at transistors, base stopper resistors, or salt thrown over left shoulders by Icelandic virgins at a crossroads at midnight would stop it.

What did stop it in each case was reversing the grounds to positive ground. I have never run into a PNP positive ground FF that could not be stopped from oscillating by at least one of the above fixes. I had one case that just oscillated no matter what I did, but I think that one was just cursed from the start. I buried it quietly where no one would find the body. 

Of course, I've seen many, many fuzz faces and other PNP germanium devices over the last three decades of building effects.

In my experience, there exist instances of reversed-ground FF's that will oscillate. Some fraction of these can be cured by bypassing, or other circuit twiddles. Some number of them can not.

You had a particularly apt sentence:

This is one of those places in life where the rubber-meets-the-road so to speak, where Electronics walk's-the-walk  and may not talk-the-talk.

It is, however, a long road.

[zjokka]

This all sums up my impressions after building a PNP negative ground fuzz face.
But my somewhat belated question is:

Is this specific to the Fuzz Face? Or do other Germanium effect designs using PNP and negative ground are apt to have identical problems? Or: would e.g. a 1 PNP transistor gain stage tolerate the negative ground better?

hope I'm making myself clear here?

I lost my interest in PNP Germaniums long time ago, with the pos ground introducing some many practical problems in making a good reliable box. But perhaps this was a premature decision?

zj

[enigmur]

I built a NPN Si Axis Face (fuzz face), it sounded great.

I then got a hold of some PNP germaniums, simply swapped the polarity and it sounds even better.

I can't hear any additional problems due to it being positive ground now. The circuit is so simple I dont think it would make a different.

[R.G.]

As I have noted many times, swapping the ground from positive to negative works fine many if not most times. This is the bait which entices people to just do it because shoot, it'll probably work.

But there is a significant fraction of the time where flipping an originally positive ground circuit to negative ground setup will NOT work. And none of the fixes I know of will make it work.

"I know of..." is not a trivial statement, since I've been hacking electronics my entire adult life, and regularly teach experienced amp techs and other experienced electronics folks things they didn't know. I have had cases where I tried all of the recommended fixes for oscillation and noise in these circuits. I made a living designing power supplies for a while, so I know about power supply bypassing. I know about grounding. I know about RF interference. I know about Nyquist oscillation. I know about parasitics. I know about magnetic field loops and capacitance. There have been times where I have expended all of the tricks I could think of, to no avail. Reversing the ground back to positive fixed the problem.

Go ahead and play with reversing a positive ground if you like. It'll even probably work. If it doesn't, don't wast a lot of time with mystical fixes. Reverse the ground back and it'll probably work.

I DO NOT recommend trying to reverse a positive ground for beginners. It's hard enough for them without introducing a really hard problem.

This issue applies to all circuits normally designed for PNP and positive ground. The lower the gain, the less likely you are to have trouble. The higher the gain, the more likely to have trouble, because with enough gain, everything will oscillate.

[mac]

I guess that the circuit is unstable. Any perturbation at Q1 is amplified by Q2, and there is also a feedback resistor. It is negative feedback but, is it possible that under some circumstances it behaves like positive feedback? If this happens, any perturbation is returned to Q1 to be amplified again and again. What happens if the 100k is increased/decreased, does it stop oscillations?
Is there any FF perturbation analysis out there? I mean, if Y=F(t) is a solution of a system, and we introduce a small perturbation P does Yp=Y+P tends to Y after some time, or at least the difference |Yp-Y| remains constant or decrease?

I'm to lazy to fill pages and pages with eqs.   I use positive gnd and forget about oscillations.

... or salt thrown over left shoulders by Icelandic virgins at a crossroads at midnight would stop it.

No no no, no virgins, that's the problem!!!

mac