Common Emitter with Split Supply?

[Scruffie]

Trying to wrap my head around this one, sure it's simpler than i'm making it.

Say I had a common emitter stage in my circuit but was using a split supply in the rest of it, such as the the schematic below (ignore the clipping diodes, no I don't want an 18V Electra, just using it as an example );



So if I gave it +/-9V... i'd still want my collector biased to 5V or whatever... but then... that doesn't seem right, if I wanted the extra headroom, how would I go about it?

[thelonious]

I just asked a similar thing in a different thread. If what you're after is headroom (and/or simplicity of keeping it all split supply instead of having to create a Vref for the transistor), could you use a resistor to bias the base and connect that bias resistor to ground as the "Vref"? (After connecting emitter resistor to V-, of course.) That would keep your base at half supply. (Btw, the thread I posted in is http://www.diystompboxes.com/smfforum/index.php?topic=104536.0)
______
(Edit: I was mistakenly thinking common collector instead of common emitter; see next post)

[Scruffie]

I just asked a similar thing in a different thread. If what you're after is headroom (and/or simplicity of keeping it all split supply instead of having to create a Vref for the transistor), could you use a resistor instead of diode to bias the gate and connect that bias resistor to ground as the "Vref"? (After connecting emitter resistor to V-, of course.) That would keep your gate at half supply...

Ah! Knew it was simpler than I was thinking, that got the ball rolling and I just need to bias the base negatively... I think.

It would depend on the application, if it was a common collector (i.e a buffer) you'd want to bias it to ground, assuming this is the right way to go about it yes as ground is your reference point.

[thelonious]

Ah, sorry - yes, I was thinking about my input buffer thing from the other thread, which is common collector as you said. It does seem like in your case you'd just need to make the base negative by a certain amount.

If you're not planning to use diodes, the schem you used as an example is kind of funky because it looks to me like a diode is involved in biasing the base like in the Bazz Fuss. So maybe the LPB-1 is a better discussion piece -



I am really interested to hear your results on this and/or what other folks say. I've searched quite a bit, and for some reason I'm coming up with very little on running a transistor from a split rail.

[thelonious]

This is similar to the LPB-1 but with values tweaked for ~17V like would be coming from a charge pump voltage doubler. Sim says 2n5088 collector would be at 8.5V, base at 1.47V, emitter at .87V.



I'm going to try breadboarding this with a single supply and then switching it to a split supply with V- connected to the bottom resistor of the voltage divider & the emitter. The transistor is isolated by its input and output caps; I think it'll just see 17V either way and work the same.

[Scruffie]

I wasn't paying attention to the diodes working as bias, I was thinking of just Collector, base and emitter voltage and signal swing.

Anyway, tried it, worked, at least for my purpose... but I preferred the lower headroom at the end of it, useful for later though.

I searched too and there isn't much on it... not that usual a thing though at least in stomp boxes.

[thelonious]

...tried it, worked

Very cool. Definitely going to try it with the input buffer for this Sparkle Drive relative I'm working on. Since it's a buffer, headroom is a good thing. Plus it'll save a few parts if I don't have to make a Vref for it and can run it right from the split supply.

[Digital Larry]

Remember that the circuit "ground" is the voltage reference for things going into and coming out of this stage.  So what you have done seems appropriate, instead of using a "split" supply, just double the voltage of a single ended supply.  You'll probably be able to swing 10 volts on that, so I hope whatever is downstream can handle it!

[Scruffie]

Remember that the circuit "ground" is the voltage reference for things going into and coming out of this stage.  So what you have done seems appropriate, instead of using a "split" supply, just double the voltage of a single ended supply.  You'll probably be able to swing 10 volts on that, so I hope whatever is downstream can handle it!

Not how I had it, it wasn't and for some reason had a stupid moment where I didn't just think of biasing the base negative to get a proper swing.

If you already have a split supply else where, why add more parts to double it.

[PRR]

Just because you have two seats in your car doesn't mean you have to use both of them.

Just because you have dual power supplies doesn't mean you have to use both of them on every stage.

You typically "want" dual supplies for your POWER stage. This will usually be more than 9V each side, and far more current available than any single small transistor could ever eat.

So given a one-9V stage, wire it ground to +9V and be done with it.

Just like if you need beer, and have a 2-seat car, you don't stretch-out to "fill both seats".

[PRR]

Another reason to go dual-supply is so the LARGE signals in the Power stage don't crap-up the ground reference. This 1-transistor stage uses the Emitter (parts) as signal reference, and isn't likely to throw much crap. So extending the Emitter leg to a probably crappy negative power supply rail seems like a poor idea.

[thelonious]

This 1-transistor stage uses the Emitter (parts) as signal reference, and isn't likely to throw much crap. So extending the Emitter leg to a probably crappy negative power supply rail seems like a poor idea.

Would you say it's also a poor idea in a common collector stage? This transistor is the only thing in my circuit that would require Vref (the op amps are running on split rails), and I was hoping to avoid adding the 2 resistors and cap to my already tight board. But maybe I'm just being lazy.
Here's what I was hoping to do:


When I breadboarded it and compared single supply to split, I couldn't hear any difference in the amount of noise, but then breadboards aren't exactly known for their low noise floor... it might have been covered up.