Question about negative ground on fuzz with PNP trannies.

[Bluesman]

Hi.

I've read someware (don't remember were)  that it can be problem with osciallation and "motorboating" to have negative ground on a fuzz with PNP transistors. Please let me know what you think about this, should i build my fuzz with positive or negative ground?.

Bluesman.

[mac]

RG wrote that, I guess. Fuzzcentral.tripod.com advice not to use negative gnd with pnp trannies.
Others say (muzique.com ?? ) that the transistor can not tell the difference since the bias are set by the voltage drop, not by the polarity. If motorboating or oscillation occurs then  a big cap from + to - solves the problem.

Who is right , who is wrong? I do not know. I guess trying neg gnd before soldering is the way to go.

BTW, some say that neg and pos gnd sound different (pisotones.com  in spanish).

mac

[petemoore]

  I can't tell much difference between pnp and Npn FF's.
  There may be some, but I have a coule PNP FF's with lotsa variability built in, and I've swapped plethora's of transistors, caps...mainly it's the gain of the transistors that seems to matter.
  And I just got some NPN's...this type is available from this site, built FF from those and I can't really tell a difference between this new npn FF and the PNP one I've been using and 'perfecting' for years...I wasn't particularly looking for a difference, maybe there are some, a great sounding FF does me just fine.

[R.G.]

I guess I ought to weigh in on this.

The origin of my comments is that I have built a number of inverted-ground effects - that is, PNP fuzzface styles with negative ground and NPNs with positive grounds. There have been some cases where nothing I could do would make the inverted-ground setup stop oscillating. In all the cases I've seen with this problem, reversing the grounds to the "natural" ground stopped the problem dead.

I am NOT a beginner at this. There are probably stabilization tricks I don't know, but I do know a lot of them from doing this for a long time.

I am NOT a beginner at electronic theory either. The theory, sure enough, says that the transistors don't know which end is ground. The theory says that if oscillation or motorboating occurs, a big cap (or a regulator, or a smaller, high frequency cap, or different wiring, or... on and on and on) will fix it.

The theory on this is incomplete. There are some cases that I have personally encountered where no amount of bypassing, fiddling with ceramic caps, or rewiring works.

That being the case, it is downright silly for beginners who have trouble soldering or figuring out which end is the cathode to be trying to do inverted ground setups. And that is my advice - don't do this. If what you want to do with your life is enough research to find the prescriptive cure for why inverted ground fuzzes sometimes - but not always - oscillate, that's great, and I think it's good that someone wants to dedicate their life to it.

But if what you want is a working effect, step to the side of that partcular rock and walk around it instead of trying to batter your way through it.

This water will eternally be muddied by the fact that sometimes you can do inverted ground setups and have them work perfectly.

Until that one time when they don't.

[Bluesman]

I guess I ought to weigh in on this.

The origin of my comments is that I have built a number of inverted-ground effects - that is, PNP fuzzface styles with negative ground and NPNs with positive grounds. There have been some cases where nothing I could do would make the inverted-ground setup stop oscillating. In all the cases I've seen with this problem, reversing the grounds to the "natural" ground stopped the problem dead.

I am NOT a beginner at this. There are probably stabilization tricks I don't know, but I do know a lot of them from doing this for a long time.

I am NOT a beginner at electronic theory either. The theory, sure enough, says that the transistors don't know which end is ground. The theory says that if oscillation or motorboating occurs, a big cap (or a regulator, or a smaller, high frequency cap, or different wiring, or... on and on and on) will fix it.

The theory on this is incomplete. There are some cases that I have personally encountered where no amount of bypassing, fiddling with ceramic caps, or rewiring works.

That being the case, it is downright silly for beginners who have trouble soldering or figuring out which end is the cathode to be trying to do inverted ground setups. And that is my advice - don't do this. If what you want to do with your life is enough research to find the prescriptive cure for why inverted ground fuzzes sometimes - but not always - oscillate, that's great, and I think it's good that someone wants to dedicate their life to it.

But if what you want is a working effect, step to the side of that partcular rock and walk around it instead of trying to batter your way through it.

This water will eternally be muddied by the fact that sometimes you can do inverted ground setups and have them work perfectly.

Until that one time when they don't.

Thanks so much for giving me help with this question, i have built  a few stomboxes and one tubeamp, have good soldering skills from a lot of soldering in my work, my skills in electronic theory is just at the beginning but gets better and better for every projekt i finish but the main reason for building this fuzz is to get a good functioning and good sounding fuzzpedal, not a experimental project so i will go the safe way instead an build the fuzz with positive ground.
The only thing that is a drawback with that version is that i have to power it with a battery instead of the powersource in my pedalboard, i would really like to have a  led that indicates on off on the pedal but i guess i have to skip that to get a decent lifespan on the battery.

Bluesman.

[petemoore]

  I was kind of waiting for a fix post, it appears an intensively focused effort may hold some probability of finding it, that or a concerted effort via internet...
  But kids seeing the schematic and lure of Neg Gnd IMO is less likely to produce it.
  Perhaps putting that more toward the bottom of the list of FF choices with a warning or caveat would save some newbie-'ice cream' headaches.

[R.G.]

I was kind of waiting for a fix post,

I just keep posting this advice, hoping that someday it will take hold.

The idea is so seductive, in a sophomore EE kind of way - "Kewl! The power supply's just an AC short, so we can turn the actual power supply any way we want and it'll still work!"

The worst part is that many times it does work OK. That makes for newly-minted "experts" with a little experience immediately posting their success to the web, where it's picked up and taken as gospel because someone sees it in more than one place.

I guess I'll just keep posting the warning. This is not something to recommend to beginners, even moderates. Sometimes it won't work, and you need to be prepared to un-do the inverted ground to get it to work.

I tell you three times - don't use inverted grounds on PNP distortion circuits unless you're prepared to unwind the ground inversion.

[hairyandy]

Hey R.G.,

I'm curious, have you tried reverse-ground with lots of different germanium circuits?  I'm sure you have but I guess my question is do you think that it's possible that for some circuits it may work better than for others?  For instance, I've built many (probably around 20) Rangemasters with PNP/neg. ground and I haven't really had any problems, and I know that Keeley does it as well with his Java Boost.  I have not built any FFs like this though, mainly because I'm just not a big FF fan.  It seems that people have more problems with the FF circuit in this regard.  Could it be that it has to do with the added complexity of the circuit?  (I know a FF isn't really complex but it's about twice as complex as the RM and a Tonebender is moreso.)  I also understand your comment that it sometimes will work fine until that one time it doesn't, sometimes that's just how the world works.  I was just curious about what your opinion is on this as this thread got me thinking.  I don't even think twice about building PNP/neg. ground RMs anymore and I've AB'd mine with Butler's and Keeley's and it's right there with them, at least within the ballpark of transitor-to-transitor variance.

Thanks,

[R.G.]

There are a few things about this that make sense and match theory. Theoretically, the higher the circuit gain, the more grounding and power supply impedance matter. That follows what  you're saying. A FF has higher gain than a rangemaster, and FF variants with buffers and boosters in front of them have higher gain still. By building rangemasters, you're playing in the shallow end of the gain pool, and that's a lot less likely to get you into trouble. It's a lot like swimming - the shallower the water and the closer you stay to shore, the less likely you're going to drown. But since the reason why this happens is not clear, it's really hard to say that you're never going to have a problem.

I think my advice applies just as stated - if you're exerienced, and you're prepared to ditch inverted grounding, go ahead. But it's not something that a beginner should ever try, because we see a relatively steady stream of them here with the common complaint "I built my ? from the inverted ground plans at xxxx, and it oscillates. What's wrong?"

[Bluesman]

The idea is so seductive, in a sophomore EE kind of way - "Kewl! The power supply's just an AC short, so we can turn the actual power supply any way we want and it'll still work!" 

You really make me feel like a fool with that kind of comments to my thoughts powering a positive grounded fuzz in my pedalboard R.G, with the risk to be concidered a complete idiot i have to ask you to explaine what you mean by turning the power supply (I'm from sweden and my ability to read and write english is not the best) .
In my pedalboard i have a boss adapter with a cable that can supply 6 pedals, if i understand you right so can i by changeing polarity on the plug that i use with the fuzz get it to work that way, does that not require that i have isolated in and outputjack on the fuzzpedal or am i totally wrong here?.

Bluesman.  

[mac]

Who is right , who is wrong? I do not know. I guess trying neg gnd before soldering is the way to go.

I want to correct myself.

Once I breadboarded a pnp neg gnd circuit and it worked fine, but when I soldered it... motorboating. Even the cap trick did not solve the problem. Bad soldering, most likely. 
I mean, no way to know until turning the final thing on.


mac

[R.G.]

You really make me feel like a fool with that kind of comments

In that case, I really do apologize. I did not mean to point a finger at anyone. In retrospect, that's the kind of thoughts *I* had when I literally was a sophomore in EE. I'd been shown some things and so I thought the whole electronic world worked the way the simple things I'd seen said. I meant that it is easy to think that if something works once, it always works; the reality may be complicated.

i have to ask you to explaine what you mean by turning the power supply (I'm from sweden and my ability to read and write english is not the best)

.
I promise you, your English is much better than my Swedish. At least you have a second language. 

In my pedalboard i have a boss adapter with a cable that can supply 6 pedals, if i understand you right so can i by changeing polarity on the plug that i use with the fuzz get it to work that way, does that not require that i have isolated in and outputjack on the fuzzpedal or am i totally wrong here?.

You are wrong here. I am saying that (a) there are some schematics that tell you to change the polarities on the plug and some internal connections and that (b) that does not work all the time. I'm telling you that the advice to do the inverted ground is incorrect at least some of the time. I'm telling you that you *can't* always do this.

Consider a pedal in isolation, no other pedals, just the guitar, the pedal, and the amplifier. For this case, assume the pedal is battery powered. The pedal only sees a 9V battery, and its ground reference is the signal ground from the guitar and to the amp. In most cases, we connect the negative terminal of the battery to that signal ground.

As long as all of the effects use a negative ground, we can connect many other pedals to this chain, and as far as the pedals and the power supply are concerned, they are all happy.

The problem is with a PNP fuzz face style distortion pedal. This pedal is designed to connect the positive battery terminal to ground, and power the effect from what is then -9V. As you are aware, this is a major pain, because you have to have an isolated power supply of some kind to run that one pedal.

The idea that the circuit inside the PNP FF style pedal can't tell the difference between "ground" being the + battery lead and the - battery lead because the signal is isolated by capacitors from the actual DC voltages is the simplistic view.

In fact, in theory if the power supply was a perfect DC source and the wiring were perfect, the theory says that this should work. That's the "sophomore EE" idea that I mentioned. In theory the pedal does not care about whether the external ground is on its positive or negative power supply lead. This **is** the "inverted ground" connection I spoke of.

If this worked first time, every time, no one would ever use a positive ground setup because it's just too convenient to use only a single pedalboard power supply.  But it doesn't always work. Sometimes, some combination of things that I admit I do not fully understand causes the effect to oscillate uncontrollably when it is run with an inverted ground.

So my advice is to NOT use the inverted ground arrangement. It cannot be trusted to always work. I wish I could tell you "do this, it will now work perfectly" but I would be leading you astray if I did that.

[hairyandy]

By building rangemasters, you're playing in the shallow end of the gain pool, and that's a lot less likely to get you into trouble. It's a lot like swimming - the shallower the water and the closer you stay to shore, the less likely you're going to drown.

But most shark attacks happen in less than 3 feet of water so watch out!    

Really though, that makes sense to me R.G. that the lower the gain, the lower the chance for oscillation.  That's kinda what I was thinking with my earlier post but I wanted to put the question out there to hear your view.  Put another way, the problem that might cause the oscillation is potentially still there, it just might not be as apparent under the circumstances of lower gain in an RM as in a higher gain circuit like a Tonebender, FF or Octavia.  I'd love to read a study into exactly what is taking place here...

[aron]

I've built a few negative ground PNP circuits, but in the end, I just use NPN Ge transistors. It's just easier in the end for me.

[Fp-www.Tonepad.com]

If you build the fuzz face with tonepad layout, changing polarity is as easy as changing a jumper (and a cap too IIRC ).

[R.G.]

I'd love to read a study into exactly what is taking place here...

So would I. I've read a lot of things that purport to tell the story, from "just build it" up through scholarly mathematical evaluations of modelling power supplies and circuits, and there isn't anything that I have found that has what I consider an adequate explanation.

There is a lot about "theoretically it should work", and also a lot about how to make your power supply better so it's fine either way.

My best guess is that it's related to the internal impedance of the power supply and the power supply noise rejection of the circuit. Opamps typically have GREAT power supply rejection, and I've never built an inverted power opamp circuit that didn't work. Of course, with opamps, there's no real need to unless you're doing something that's a bit bizarre. Generally the simplistic FF style fuzzes have zero power supply rejection - meaning that whatever noise is on the power supply comes right through to the output - and are also sensitive to the power supply impedance, as some commercial pedals use for effect. 

But if that's all that's going on, you should be able to cure a case of inverted-ground-oscillation by a combination of better wiring and bypassing the power supply. And that does work - sometimes. Other times, it simply doesn't. That tells me that there's at least one more cause hiding in there that I have not yet found. And I can't find that anyone else has either, based on what I've read and tried.

Which leads me back to - this is a poor place for beginners to be playing.

[hairyandy]

My best guess is that it's related to the internal impedance of the power supply and the power supply noise rejection of the circuit.

That's funny, the only problem I've had so far with a RM was because of weird power.  I built one for Peter Stroud and the next gig that we had was in Puerto Rico on some small little island.  The power for the whole stage was run off of a generator and he plugged the thing in and it made the most horendous hissing and squealing you've ever heard.  Of course all of my fellow techs were laughing their ass off!    I told him to take it home and I thought it would be better there.  Sure enough, at home it's really quiet and sounds great.  I wonder if it had been wired positive if it would have been any different.  I'm thinking it still would have been noisy as hell most likely but maybe not as  much oscillating...

[Bluesman]

You really make me feel like a fool with that kind of comments

In that case, I really do apologize. I did not mean to point a finger at anyone. In retrospect, that's the kind of thoughts *I* had when I literally was a sophomore in EE. I'd been shown some things and so I thought the whole electronic world worked the way the simple things I'd seen said. I meant that it is easy to think that if something works once, it always works; the reality may be complicated.

i have to ask you to explaine what you mean by turning the power supply (I'm from sweden and my ability to read and write english is not the best)

.
I promise you, your English is much better than my Swedish. At least you have a second language. 

In my pedalboard i have a boss adapter with a cable that can supply 6 pedals, if i understand you right so can i by changeing polarity on the plug that i use with the fuzz get it to work that way, does that not require that i have isolated in and outputjack on the fuzzpedal or am i totally wrong here?.

You are wrong here. I am saying that (a) there are some schematics that tell you to change the polarities on the plug and some internal connections and that (b) that does not work all the time. I'm telling you that the advice to do the inverted ground is incorrect at least some of the time. I'm telling you that you *can't* always do this.

Consider a pedal in isolation, no other pedals, just the guitar, the pedal, and the amplifier. For this case, assume the pedal is battery powered. The pedal only sees a 9V battery, and its ground reference is the signal ground from the guitar and to the amp. In most cases, we connect the negative terminal of the battery to that signal ground.

As long as all of the effects use a negative ground, we can connect many other pedals to this chain, and as far as the pedals and the power supply are concerned, they are all happy.

The problem is with a PNP fuzz face style distortion pedal. This pedal is designed to connect the positive battery terminal to ground, and power the effect from what is then -9V. As you are aware, this is a major pain, because you have to have an isolated power supply of some kind to run that one pedal.

The idea that the circuit inside the PNP FF style pedal can't tell the difference between "ground" being the + battery lead and the - battery lead because the signal is isolated by capacitors from the actual DC voltages is the simplistic view.

In fact, in theory if the power supply was a perfect DC source and the wiring were perfect, the theory says that this should work. That's the "sophomore EE" idea that I mentioned. In theory the pedal does not care about whether the external ground is on its positive or negative power supply lead. This **is** the "inverted ground" connection I spoke of.

If this worked first time, every time, no one would ever use a positive ground setup because it's just too convenient to use only a single pedalboard power supply.  But it doesn't always work. Sometimes, some combination of things that I admit I do not fully understand causes the effect to oscillate uncontrollably when it is run with an inverted ground.

So my advice is to NOT use the inverted ground arrangement. It cannot be trusted to always work. I wish I could tell you "do this, it will now work perfectly" but I would be leading you astray if I did that.

Thanks so much for your advices R.G, after help from you, petemoore and some other people on this forum I have decided to build the fuzz with positive ground and without indicator led and DC jack and power it with battery instead, just like the original Dallas Arbiter fuzz.
I plan to put two extra external pots on it, one at the input of the circuit that acts like the volume control on the guitar and one dial knob that controls the voltage to collector of Q2.

Bluesman.