Nuts,..just wondering diode bias

[petemoore]

could diodes be strung between Gnd and V+ to bias?
 run from ground enough diodes to bring about 1/2v, then run from V+ enough diodes to get about 1/2v, maybe a trim resistor to match more closely and connect them for 1/2v.
 ...or run them offbias a bit to mess with the signal swing as it crosses 1/2V..

[brett]

Hi Pete.
You're a cryptic bugger sometimes.  But I think I get what you mean.

The answer is yes, and it's used in a few circuits.  John Hollis was clever in the ring frobnicator - it uses 2 x red LEDs to get about 4 1/2 volts to run the LFO.  A similar system would also work in most op-amp circuits that need a Vbias.  A side benefit is that the LEDs show that everything is ok bias and battery-wise.

I have a small class A amp that uses a 3.3 volt zener to bias a transistor "on" (in a constant current source).  Instead of a 3.3V zener, the bias could have been set up by lots of combinations of other diodes.  I borrowed the idea from a circuit at the Elliott Sound Products page.  There's a LOT of good information at that site, especially for those of us somewhere between newbie and expert.

cheers

[petemoore]

That'd be most of us and me for sure.

[MartyMart]

I'm pretty sure that Torchy's Phaze45 and ZW44 use zeners for Vb too.
Sure, you could use that same trick anywhere !

Marty.

[petemoore]

The only reason I can think for doing so is if they'd provide a 'difference' to resistors biasing 1/2v...or anything else.
 If they're the same/same at this, that's probably why we invariably see mostly resistors doing this.
 I was thnkking maybe some type of Distortion that happens right close to 0v of the signal wave swing.

[RLBJR65]

Pete,
I have an old article about zener diodes when they were new. As I recall the mag. considered them to be the cure all for biasing issues. I can scan and email it to if you would like.
PM me.

Somewere I have a schem. of a dist. that (if I remember right) only used 3 diodes 2 caps and 1 transistor. I never tried it and can't seem to locate the schem. right now, but I'll keep looking.

Richard

[Sir H C]

With integrated circuits voltage dividers are often done using strings of "diode connected" FETs.  These take up less space than resistors in some processes.

As to the diode string, the problem is that you have a pretty small range over which the diodes conduct safely.  You can either bias them really cold (low current) or normal and if you do the latter you could smoke them if the supply goes too high.  Also with the temperature dependence of diodes you can get a really high current as the temperature goes up (and that will cause the temperature to go up more...)

Diodes can work well for biasing, but you have to be ready for the big variance over temperature (unless you are using zener diodes).