The Great Debate: PNP and NPN in harmony?

[Govmnt_Lacky]

I have been racking my brain about this and wanted to throw this out there.

PNP positive ground based circuits (Ger Fuzz) utilize the ground as the V+ and also send the V+ through the Input sleeve/ring switch to turn the battery on/off. I have been told that the Input switch wiring is what causes the problem when using the same PS to power PNP and NPN effects.
So, this is my mitigation. If you ONLY power a PNP Fuzz circuit with a DC jack, why not run your V+ from the DC jack directly to the 3PDT and from there, run it to the ground rail of the board and the LED. Then you just wire everything else as an NPN circuit (Input/Output sleeves, ground for the board, LED ground, and DC jack ground connected)
This way, when the switch is engaged, your V+ from the DC jack is ONLY being sent to the board ground rail (PNP) and the LED. The Input jack remains connected to ground.

Would this work?

[Hides-His-Eyes]

The circuit board needs to see the same ground as the output jack!

[Govmnt_Lacky]

The circuit board needs to see the same ground as the output jack!

This was in my OP

Then you just wire everything else as an NPN circuit (Input/Output sleeves, ground for the board, LED ground, and DC jack ground connected)

[slacker]

It's nothing to do with the sleeve/ring switching. It's because in the positive ground circuit v+ is tied to ground, and in a negative ground circuit V- is tied to ground. If you try and use both negative ground and positive ground circuits off the same power supply then you end up with both V+ and V- being tied to ground, shorting out the power supply.
There's no way round it as far as I know.

In your example, where are you connecting the other power supply point on the board to?

[Govmnt_Lacky]

It's because in the positive ground circuit v+ is tied to ground, and in a negative ground circuit V- is tied to ground.

Ok. Maybe this is where I get confused. If by V+, you mean +9VDC...and... by V-, you mean Ground, then how does this differ from ANY OTHER CIRCUIT? In all other circuits, you have V+ and V- running through different components as well and there are no problems. Take any dual op amp for instance. 9V goes into pin 8, and ground goes into Pin 4. No problems there 

In your example, where are you connecting the other power supply point on the board to?

The "other power supply point" in this case would be ground. As stated before, ground would be attached to the circuit board where it would normally go. Use this layout for example:

http://www.generalguitargadgets.com/pdf/ggg_ff5_lo_pnp.pdf?phpMyAdmin=78482479fd7e7fc3768044a841b3e85a

According to this layout, the changes I would make are:

-Remove battery all together
-Connect Input sleeve, Output sleeve, DC jack ground, And board ground (where the DC jack ground connects to the board in the layout) together.
-Remove black jumper from 3PDT
-Wire V+ from DC jack directly to center lug of 3PDT.
-Run 2 wires from bottom middle lug of 3PDT. One to LED and the other to the "ground rail" which is now the V+ rail.
-All other wiring would remain the same.

[slacker]

Ok. Maybe this is where I get confused. If by V+, you mean +9VDC...and... by V-, you mean Ground,

I said V+ and V- because in your average power supply what you have is two connections with a potential difference of some voltage between them, 9 volts for example, so one of the connections is 9 volts higher than the other. They only become +9 volts and ground if you decide to call V- ground (negative ground), if you decide to call V+ ground then you have ground and -9 volts (positive ground). The side that you decide to call ground then gets connected to what you've decided is ground for the rest of the system, which is normally the sleeve of the guitar lead. For one power supply you can't decide that both V+ and V- are ground.
Hope that makes sense

Take any dual op amp for instance. 9V goes into pin 8, and ground goes into Pin 4. No problems there 

Sorry I'm not sure what you mean.

The "other power supply point" in this case would be ground. As stated before, ground would be attached to the circuit board where it would normally go. Use this layout for example:

http://www.generalguitargadgets.com/pdf/ggg_ff5_lo_pnp.pdf?phpMyAdmin=78482479fd7e7fc3768044a841b3e85a

According to this layout, the changes I would make are:

-Remove battery all together
-Connect Input sleeve, Output sleeve, DC jack ground, And board ground (where the DC jack ground connects to the board in the layout) together.
-Remove black jumper from 3PDT
-Wire V+ from DC jack directly to center lug of 3PDT.
-Run 2 wires from bottom middle lug of 3PDT. One to LED and the other to the "ground rail" which is now the V+ rail.
-All other wiring would remain the same.

If you ignore the fact that you're using the 3PDT to switch the power, what you've done is make it into a negative ground circuit, because you're connecting V- to the sleeves of the jacks. This does work, but you have to flip the transistors and any polarised caps round. Have a look here for a better explanation http://www.muzique.com/lab/fuzzface.htm. This can lead to other problems if you search for "PNP negative ground" you should get loads of info on this.

[R.G.]

Ok. Maybe this is where I get confused. If by V+, you mean +9VDC...and... by V-, you mean Ground, then how does this differ from ANY OTHER CIRCUIT?

That is indeed a common point of confusion. V+, V-, and ground are completely unrelated except by how we connect them.

Imagine a battery. It has two terminals, a more-positive one we'll call V+ and a more-negative one we'll call V-. Which one is ground?

Neither of them. A battery is completely floating. Either one of them - or neither of them! - can be connected to ground. If it's a 9V battery and we hang two equal value resistors in series across it, we can even ground the center of the resistors and have the voltages be +/-4.5V.

The idea that the negative voltage is ground is caused by the preponderance of modern circuits that do it that way. There is no real need to call the minus side of a supply ground.

GROUND IS ONLY A PLACE WE DECIDE WE'LL MEASURE ALL OTHER VOLTAGES FROM.

[Taylor]

Understanding is good and should be encouraged. I continue to try to deepen my understanding of this concept, because it is confusing. Although I've never built a Fuzz Face.

However, just to add some more info, one way to achieve this harmonious unity of PNP and NPN you seek is to use a charge pump to derive -9v from +9v. RG's tired of linking it but he has info on this at geofex, so I'll link it for him:

http://geofex.com/circuits/+9_to_-9.htm

[Govmnt_Lacky]

If you ignore the fact that you're using the 3PDT to switch the power, what you've done is make it into a negative ground circuit, because you're connecting V- to the sleeves of the jacks.

Maybe I over-simplified what I am saying or this concept is just not sinking in! Exactly HOW have I changed the "circuit" to a negative ground by connecting V- to the sleeves 

V+ is STILL going to the same place on the circuit board as the original.

V- is STILL going to the same place on the circuit board as the original.

The input signal is still going to the same places.

The output signal is still going to the same places.


The only change that I see is that the enclosure is not tied to V+ anymore so there should be no problems with the Input jack at a different potential as the rest of the pedals in the chain.

I might need a deeper explanation than this because I am just not seeing the problem with the way I want to wire this. Not trying to be difficult. Just trying to understand the logic. 

[edvard]

I have wired Fuzz Faces exactly as you describe and haven't come out any worse for wear.
I HAVE heard that sometimes they are prone to feedbacks and such, but it never happened to me.

I also concur with RG; all our terminology is simply a methodology to label points of reference so we can easily describe a given signal path.
We might as well relabel positive and negative "Rizzo" and "Wednesday" as long as we understand there are 9 volts of potential between those two points, and which one is connected where in our rig setup so as to avoid shorts.

[slacker]

Maybe I over-simplified what I am saying or this concept is just not sinking in! Exactly HOW have I changed the "circuit" to a negative ground by connecting V- to the sleeves 

I had another look at the layout and I think it has an error which is causing the confusion. If you look at the  schematic you'll see that the battery positive or the DC plug positive are connected to the ring of the input jack, this then gets connected to the sleeve when you insert the guitar lead, and this is ground. The sleeve of the output jack is also ground and so are lug 1 of the volume and fuzz pots. The battery negative or DC plug negative then goes to R11, R2 and R4. Although the grounds are shown as separate points in the schematic, that's just to make it easy to read, in reality all the points marked as ground are connected together, so the DC plug positive is ground, making the circuit positive ground, with the supply voltage being -9 volts.

If compare that to the layout you'll see it doesn't match. The battery positive and DC plug positive and lugs 1 of the pots are all connected together on the board. The sleeves of the jacks are connected together, but they aren't connected to anything else, they should be connected to the board where the pots are.
If it's easier compare the layout with the late 60s fuzz layout which uses the same board as that is correct and matches the schematic.

If you try to connect this to a pedal that has the DC plug negative connected to the sleeves of the jacks, and run them off the same power supply then because the sleeves of the jacks are connected together via the guitar lead you are connecting V- and V+ together which will short out the power supply. This is the problem you are trying to prevent.

Like you said with your mod what you have done is connect the sleeves of the jacks to V- instead of V+ so V- is now ground making the circuit negative ground. So it will now work with other negative ground pedals on the same supply.

I might need a deeper explanation than this because I am just not seeing the problem with the way I want to wire this. Not trying to be difficult. Just trying to understand the logic.  Cool

No problem it took me ages to get my head round this.

[Govmnt_Lacky]

I just dont understand. I think that the solution that I have above is SO simple that others would have thought of it first. Is there ANY fault or problems with the way I want to wire this?

[caress]

Understanding is good and should be encouraged. I continue to try to deepen my understanding of this concept, because it is confusing. Although I've never built a Fuzz Face.

However, just to add some more info, one way to achieve this harmonious unity of PNP and NPN you seek is to use a charge pump to derive -9v from +9v. RG's tired of linking it but he has info on this at geofex, so I'll link it for him:

http://geofex.com/circuits/+9_to_-9.htm

+1.  this is so simple to do and will solve your problems... the cost of the parts are around $2-3 and will take around 15-30 minutes to perf up.

[R.G.]

PNP positive ground based circuits (Ger Fuzz) utilize the ground as the V+ and also send the V+ through the Input sleeve/ring switch to turn the battery on/off. I have been told that the Input switch wiring is what causes the problem when using the same PS to power PNP and NPN effects.

You have been told incorrectly. It is not the input switch wiring that causes the problem with using the same power supply for positive ground and negative ground effects. Instead, it is the fact that the positive ground effects ( PNP ) need to ground the + side of the power supply to function, and the negative ground effects ( which you call NPN effects) need to ground the negative side of the power supply. What's important is that whichever side of the power supply is used as ground in an effect, that power supply terminal is connected to the signal ground sleeve on the input and output jacks, and must be for the effect to work without massive hum.

When you have one of each kind of effect, one of them attaches the + side of its power supply to the signal ground (which is connected to the input and output jack sleeves), and the other attaches the - side of its power supply to the signal ground (input and output jack sleeves); it must be this way, or the effects won't work, or will have massive hum.

As long as there are two different power supplies, or if there is no connection between the input and output jacks, all this works well. But if you (a) power both effects from the same power supply and (b) connect the two effects with a cable which carries a signal ground (and you must do this to use the effects), then both the + and - terminals of the single power supply are connected to the same place - signal ground - and it shorts out the power supply.

Would this work?

No. It doesn't matter where the DC jack goes, or how the bypass switch is wired. For the effects to work individually, one needs the + side of the power supply grounded, the other needs the - side grounded, and both signal grounds are connected in common. The power supply gets shorted. This cannot be avoided by clever switching.

It can be avoided by forcing either the positive ground or negative ground pedals to not, in fact, use the ground they think they're using. This is the "negative ground for PNP effects" hookup. This theory is based on the idea that power supplies have a zero impedance for AC signals, and that the + and - side are "the same as ground" for AC signals. So the common practice for this dodge is to break the signal ground loose from the + side of the power supply for positive-ground (your PNP) effects and tie signal ground to the - side instead. This works to the extent that the gain of the pedal and other factors do not amplify the imperfectness of the + side of the power supply up enough to make the effect not work right. Sometimes it works, sometimes it does not.

I personally put a lot of time into finding out how to force this to work in all situations, and could never find a way. Sometimes it works, sometimes not. People who have only done pedals where they get away with it think it works all the time, and post their experiences on the internet as "The Way". This just means they have not yet run into the counterexamples.

In any case, no, you can't get around this with clever switching.

[PRR]

> I just dont understand.

Nor do I.

You can't paint word-pictures of a circuit network with more than 3 parts. Even if you understand what you mean, your readers WON'T. Our eyes glaze, we make unwarranted assumptions, we jump to conclusions.

DRAW US A PICTURE! (Not a reference to a picture plus a list of changes.)

"Switching" jacks using signal (ring) contacts are extra confusing. I think they are bad for several reasons; but the appeal is obvious. (Kudos for dropping that idea.)

> If you look at the schematic...

BTW, that shows the LED polarity wrong. The arrow should flow the same plus-minus direction as the emitter arrows.

Here is the best wiring for PNP amplifier on a negative-ground positive-hot power supply:



We re-define the negative power rail as "common ground". Bias and emitter bypass caps should go to common ground, not emitter power return, so they do not inject power crap. Observe polarities! If the positive rail is VERY clean and solid, bypass caps may return to it. But wall-warts hum and hiss; also you propose an additional NPN amplifier which could be a lot of total gain. For low-power stages an R-C power filter may clean and brace the positive power rail to be clean enough to be used as bypass return, simplifying re-layout. That cap might have to be hundreds of times large than any emitter bypass (though I bet 100uFd is ample for 99% of cases).

[Govmnt_Lacky]

Thanks RG and Paul for your inputs. I only have these questions in response:

If I wire the circuit in the manner that I described above, then how exactly am I grounding the V+ rail? Can you elaborate? (RG)

In my version of wiring, I do connect the Input and Output jacks to the V- board input, so why would there be hum? (RG)

How would I include a hand drawn picture with my post? (PRR)

Thanks all. I still do not understand why it would not work but I think I am getting some good information about it. Still trying to comprehend the logic behind it. 

[Taylor]

How would I include a hand drawn picture with my post? (PRR)

I'm not PRR, but top choices would be:

-scan your drawing with a flat bed scanner. export as jpeg.

-Draw in Photoshop, Illustrator, MSPaint....

-if you have none of the above, http://www.onlinesketchpad.com/

[slacker]

Thanks all. I still do not understand why it would not work but I think I am getting some good information about it. Still trying to comprehend the logic behind it. 

Like I said before I think what you're describing is the same as here http://www.muzique.com/lab/fuzzface.htm, if it is then it does work. The problem is that like mentioned at the bottom of that page and explained by Paul sometimes you get problems with oscillations, sometimes this can be cured, sometimes it can't.   

[BarnyardBill]

well, after researching this PNP vs. NPN stuff for an hours or so, I am still confused.      Maybe I will try this circuit for my next project so I can try to understand it.  At first intuition, I think it PNP sounds like a circuit with the negative and positive switched around.  As I read on, I see that a PNP or (postive grounded) circuit requires a -9V DC current.  I guess you could also use a wall wart that is reversed to achieve this?

[R.G.]

well, after researching this PNP vs. NPN stuff for an hours or so, I am still confused.      Maybe I will try this circuit for my next project so I can try to understand it.  At first intuition, I think it PNP sounds like a circuit with the negative and positive switched around.  As I read on, I see that a PNP or (postive grounded) circuit requires a -9V DC current.  I guess you could also use a wall wart that is reversed to achieve this?

Let's clear the air a bit. This is becoming a mishmash of several related concepts. The mishmash is properly confusing. Here are some relevant concepts that first need to be understood separately, and only then an attempt made to interrelate them.

1. What we call "ground" serves several different purposes.
1.a. Ground can be a measurement reference point that we just decide will be zero volts because we say it is. Think of this "reference ground" as where you hook the black lead of your voltmeter. Ground can be a signal reference point; this is an important subclass. I call this "reference ground".
1.b. Ground can serve as a shield to prevent capacitive coupling of noise and interference into signal wires. I call this "shield ground".
1.c. The conductor that carries the used electricity back to the power supply can be called ground. I refer to this as "sewer ground".

2. The actual polarity of the power supply that connects to the reference ground from 1.a. makes no real difference **=>>as long as only the one isolated circuit you're powering <<==** is considered. When you start hooking many circuits to a single power supply, which polarity is hooked to ground does start to matter, and this is especially true if the circuits are separate little boxes which internally make assumptions about which power supply lead is reference and sewer. This idea is what the people who promote reversed-polarity grounds count on working first time, every time. Sometimes it even does.

3. The polarity of the transistors has not much to do with which polarity of power supply is called ground in theory.

4. Any high gain, high impedance circuit is excruciatingly sensitive to whatever crap is on its reference ground. If a circuit thinks the + side is ground and instead the - side is ground, it's prone to getting confused and misbehaving.

First, NPN or PNP has nothing to do with what the power supply has to be. It has nothing to do with where the ground has to be. All that mattes with NPN is that the collector has to be more positive than the base, which has to be more positive than the emitter by about 0.6V for it to amplify. On the PNP side, the collector has to be more negative than the base, which has to be more negative than the emitter by about 0.6V for it to amplify.

In general, you can call any one point in a single circuit "ground", and as long as you measure all voltages from it, no difficulties arise.

HOWEVER: for the narrow niche of guitar effects pedals, what you call ground matters when you start hooking them up to other pedals in a pedalboard environment. This is because every single pedal thinks it's "ground" is the only ground, and will happily misbehave in a group of others if they don't all agree on what "ground" is. This is what causes the impossibility of powering a positive ground pedal and a negative ground pedal from the same battery or power supply and also running a signal cable carrying ground between them. Both sides of the same power supply can't be grounded and still have the power supply work right.

And the whole issue of which side of the power supply is grounded has been polluted by the few circuits from the Bad Old Days which use odd power circuits. Back when the Fuzz Face  and many of the PNP-germanium effects people are still imitating were first thought up, you pretty much had to use PNP transistors, because germanium works best in PNPs, and germanium was all anybody had or could afford. The practice at the time, carried over from tubes, was to ground the more-quiet side of the power supply, which was the negative side for tubes, and in a simple, single-transistor-polarity circuit made of PNPs was the positive side.

When silicon became available, people found that it worked better as NPNs, and as such for simple amplifier circuits, the practice of making ground the negative side seemed more natural. As silicon came to dominate the industry, people came to assume that the negative side was always ground, which was convenient but incorrect.

About 90% of the difficulty with grounding to eliminate hum and RF, grounding for power, and grounding for signal reference comes from the fact that what we call ground serves several purposes (item 1).

The issue with PNP and NPN transistor circuits using positive-side-grounded power supplies and negative-side-grounded power supplies respectively is partly historical. The power issue here is that you cannot have a positive-side-grounded circuit and a negative-side-grounded circuit working from exactly the same power supply, because that means both sides are grounded, and the power supply is shorted. You *can* fake the issue by grounding only one side of the power supply and trying to make one of the circuits think the polarity of supply it used to use as "power" is now ground. This works sometimes, but because of the use of ground for a signal reference inside the circuits, the one with the faked ground will sometimes misbehave. Sometimes this can be fixed, but there are cases where no practical effort will fix it.

It is far, far, far simpler to use a separate power supply for positive-ground and negative-ground circuits than it is to try to make one of the units accept the wrong-side ground. Good, reliable, quiet power adapters can be bought for about US$20. That's a remarkably cheap solution compared to spending several days flogging the internet looking for whywhathowwhenwho to get around it.