PNP fuzzface with negative ground?

[amz-fx]

There has been a lot of discussion about negative ground with PNP fuzzface circuits and the theory is proposed that there are a select few transistors that are somehow defective and have slightly different characteristics from normal PNP devices. Go to my pedal blog to see a fun question about the probability that a "bad transistor" will be the cause of oscillation.

Note:  This is just for fun and I made up these numbers. It is not meant to support either side of the discussion.

Also, please don't post the answer here so that others who read this after you will have the chance to work it out on their own. Thanks.

regards, Jack

[QSQCaito]

couldn't you do a pnp neg ground using -9volt?

[zachomega]

What characteristics are we talking about when we say that one transistor is "bad"? 

This is really interesting...although personally I prefer positive grounded pedals.  They make more sense to me. 

-Zach

[brett]

Not to people with neg ground power supplies.
"Battery eaters" never make much sense to us.

[brett]

Hmmm...the link just takes me to... "Sorry, but you are looking for something that isn't here."
Or is there a further link that I've been too lazy to find?

[amz-fx]

I was editing the post to fix a typo and Wordpress just locked up. The database got corrupted and I had to load the backup which was 3 or 4 days old, so I lost the post.

It wasn't really about the fuzzface, but about statistics and probability. I made up the numbers as I said, but it's too much trouble to work it out again and retype the whole thing.

There's a post here in the Members section about Fuzzface bias that talks about negative grounds if you want to read more.

Sorry for the confusion 

-Jack

[R.G.]

Jack, since we reached a different conclusion about this issue over in the members-only section, could you reword this part of your page to be less pejoritive? Here's your original:

No amount of misleading mumbo-jumbo about dreaded oscillations should sway you from trying this technique since it is easily reversed.

If for some reason oscillation does occur, a quality low esr capacitor from the positive supply to ground will solve the problem everytime since it effectively places the power rails at the same ac potential. Put the capacitor on the pc board if possible. Also keep the positive and negative power wires as short as practical. I've used this method with two different fuzzface derivatives for years with no problems and one of them has a quite long positive power supply wire.

How about something like:

Some experimenters have reported oscillation when this technique is used. If you want to experiment with reversing your grounds, you should try it, bearing in mind that sometimes it will oscillate.

If it does oscillate, a quality low-ESR capacitor from the positive supply to ground will often solve the problem, since it lowers the impedance between the power rails.

I believe that's much more like what you said in the members-only area:

I have no doubt that you have experience with RF circuits and/or high current circuits that will oscillate on this type of ground trick, and I agree completely with you that problems can be encountered.

On the other hand (I hate it when my lawyer says that, which he does frequently), when we are talking about the limited subset of fuzzface type circuits, the method should work almost every time. The key is good circuit layout and solid ground connections, as you well know.

For anyone who wants to try it, keep the input and output well separated.  The emitter of Q1 should be tied to the same point as the power supply end of the 1k Drive control, with a fat wire as short as possible. We're looking to get a good low impedance (resistance) connection for both the emitter and Drive.

Since most people who are interested in this technique will be trying it with low gain PNP germanium transistors, it will work in this application.  Yes, if you have some super-hot silicon transistors and try this, you may be able to get it to squeal, but that's the exception rather than the rule.

[R.G.]

So what do you say, Jack? How about we go for consistency?

[R.G.]

So what do you say, Jack? How about we go for consistency?

[brett]

Hegemony and self-advocacy might win the battle, but mutual respect and the acceptance of others' versions of reality wins the war. 

Of course, that's just my view of reality !

[R.G.]

I'm with you Brett. The discussion of this issue reached a mutual acceptance of a version of reality, but in another, much less public venue. I very much respect consistency.

[m_charles]

digging up an old topic but wondered. The "quality low esr capacitor from the positive supply to ground (to prevent oscillation)", what exactly is a low ESR cap?
Are we just talking a quality electrolytic (vishay, CDE, etc.), maybe 100uf?
thinking of doing a pnp-neg gnd FF.
chuck (thanks in advance)!

[earthtonesaudio]

ESR is "effective series resistance," and has to do with the properties of the dielectric.  Avoid aluminum electrolytic caps, they're pretty bad.  Tantalum, multilayer ceramic, and film caps are also good.  Tantalum has the highest capacitance for its size, but personally I'd go for the multilayer ceramic for a good compromise between size, reliability, and environmental reasons (tantalum mining is a yucky business).


If you don't need a large value for "bulk" board capacitance, you might try starting with the typical 0.1uF multilayer ceramic cap for bypass.

[m_charles]

Thanks! So are you saying a regular old .22uf or .1uf METAL FILM would do the job?
I have a couple 10uf tants around as well. Didn't know about the mining thing...