Do you find PNP negative grounded circuits noisier?

[darron]

Hi all.

Quick question.


Do you find that when you make PNP (especially germanium) circuits as negative wire grounded they add a bit more hiss/germanium-rumble? IE a flipped Fuzz Face or Rangemaster? I'm not talling about oscillations or hum.

I'm not the biggest fan on switching charge pumps so I don't mind converting the circuits. With thoughtful wiring and filtering I never have oscillation problems. I do feel like I'm usually loosing the noise battle though with germanium...

[mistahead]

Not in my experinces beating the Rangemaster into five forms of submission (first play thing)... but are we referring to on their own in a chain or on their own and how are we powering them?

[WaveshapeIllusions]

I don't think there should be a difference. Unless you're using a charge pump to generate a negative voltage; it could introduce some extra noise.

A cursory scan of data sheets does show a difference. If I'm reading it right a PNP 2n3906 has less noise than an NPN 2n3904. I have no idea why this would be though.

The noise figures shouldn't change based on negative or positive ground. Voltage is just a relative measure; the transistors only care about the difference between their three terminals. One sees PNPs and NPNs flipped all the time in amplifiers, with no difference in sound as far as I can tell.

[darron]

Not in my experinces beating the Rangemaster into five forms of submission (first play thing)... but are we referring to on their own in a chain or on their own and how are we powering them?

just on their own with no other pedals. either battery or power supply.





like i said. no charge pumps. i notice that usually NPN germanium circuits built correctly with a negative ground can be quieter, but i've always assumed that is from the later production and natural structure of NPN being quieter. I haven't done any real tests to compare positive VS negative ground PNP and thought other's might have some opinions.


I'm glad to have 2+ saying 'no' so far. that means I can keep going on the way I am!

[Thecomedian]

I've read that DIYers in the old days used PNPs because they were quieter.

Russian transistors are 25-40% more PNP in the old days.

[kingswayguitar]

no

[R.G.]

Do you find that when you make PNP (especially germanium) circuits as negative wire grounded they add a bit more hiss/germanium-rumble? IE a flipped Fuzz Face or Rangemaster? I'm not talling about oscillations or hum.

I'm not the biggest fan on switching charge pumps so I don't mind converting the circuits. With thoughtful wiring and filtering I never have oscillation problems. I do feel like I'm usually loosing the noise battle though with germanium...

A positive grounded PNP circuit of the normal type has the signal ground tied to the power supply ground by a wire. A negative grounded PNP circuit of the normal type has its signal ground tied to the power supply "ground" by a capacitor. The "grounding" is heir to all the ills of the differences between a capacitor and a wire for being a "short  circuit".

To the extent that the capacitor is not perfect, the "ground" is not perfect.

So yes, it's more vulnerable to issues of noise, as well as oscillation and hum. Hum and oscillation are the psychopathic versions of the problem, noise is the subtler issue.

Germanium is, as noted, generally more noisy than silicon because the processing techniques for germanium, especially old stock germanium was cruder than today's silicon processes. Germanium processing stopped evolving when silicon took over in industry. No one builds new germanium fabs. Even new-manufacture germanium uses "vintage" processing equipment.

By demanding germanium transistors, you're demanding a product that is not as well made in certain ways. By demanding PNP contorted to negative ground, you're demanding a less rigorous ground. Will there be issues? Sometimes. Will these issues be ones which can be made less severe. Sure. Will these issues ever be completely resolved? No.

[Thecomedian]

If I'm understanding you correctly, a capacitor is better for grounding the signal? Is that through the transistor's emitter or through the volume control after the DC decoupling cap?

http://www.muzique.com/schem/bird.gif

is this schematic "more ill" because it doesn't have a cap going to ground, and the signal shares ground with DC through the base and emitter resistors?

I modeled the Fuzz face in PNP for positive ground via the usual method of stuffing PNPs in the circuit, and also via negative ground via inverting the PNP, and both results in LTspice gave identical outputs. What should I look for in the circuit to determine changes?

[darron]

I'm glad you chipped in R.G. with a definitive answer. As I know that you LOVE commenting on the issues. *cough*. Usually not in your favour when people are given dramas.

Well, I never really have any problems with hum or oscillation that you'd normally associate with grounding issues, so I guess I'm flying alright. It's just the normal headaches of germanium noise. I'd normally filter closely with at least 470 / 1,000uF with a low ESR cap.

It seems a bit counter intuitive to keep proceeding with germanium! It died out for many reasons. What's wrong with us? Lovely mushy break ups and fun to collect. I wish there was a less noisy way to incorporate them in other than just using them as a diode junction.

[Thecomedian]

Well I doubt there's anything wrong with germanium in of itself. People are making GeSi composite transistors to get even faster speeds than could be done by either alone. I cant wait for tubes to get refined enough by NASA to be turned on by a single volt, as opposed to 10 or 150, as an example of an old technology that's seeing a resurgence in real world application beyond audiophile.

I'm not sure what the market on tube based hi-fi/pre-amps was before, but there's tons of this stuff coming out of china now, in very "modern" designed packages. eBay is loaded with it.

There's nothing wrong with the element per say, it's the fabrication technique and the doping elements, the design of the circuit, so many things that can contribute to either making it shine or or flop.