grounded emitter transistor biasing

[duck_arse]

I have a question about common emitter amp biasing and voltage divider bias. the base is always ~0V7 higher than the emitter, the base voltage/ emitter resistor sets emitter/collector current, and Vc from Rc. working backwards and forwards with emitter voltages and resistors, I can muddle the bias on paper close enuff to real life measured.

so now we ground the emitter, the base has to be 0V7. who wins the argument, the divider resistors or the transistor base? do I just calculate near 0V7 and let the device worry about it, use a series base resistor, or do I chuck 1 resistor and use collector / self bias? how do I set the collector current and Vc, or calculate the gain?

that makes 2 proper technical questions in one session!

[gjcamann]

I'm betting the transistor will win. It will control the current to make sure that the voltage drop across the lower resistor in your voltage divider is 0.7V. And once you do this, there isn't much point in bothering with a voltage divider for bias. You should probably just use a follower. Also, you generally want an emitter resistor if possible to keep the input impedance higher.

[R.G.]

I have a question about common emitter amp biasing and voltage divider bias. the base is always ~0V7 higher than the emitter,

Well, actually, it's not always. That just happens to be a very good approximation for situations where the exact voltage doesn't mess things up.

The base-emitter voltage is a logarithmic function of the base-emitter current. Or, put another way, the base-emitter current is an exponential function of the base-emitter voltage. The current into the base doesn't amount to much up to the first whispers of current flow at around 0.45V, it's really going at 0.5 to 0.6V, and is really, really  big (for a low-current diode junction) at about 0.7V ... for silicon! Each semiconductor has its own exponential function, of course.

If you look up a graph of base current versus base-emitter voltage ...

you'll see that it starts  conducting a lot between 0.6 and 0.7. It can go higher, but for most low signal biasing purposes, it's about 0.6 to 0.7.

the base voltage/ emitter resistor sets emitter/collector current, and Vc from Rc. working backwards and forwards with emitter voltages and resistors, I can muddle the bias on paper close enuff to real life measured.

Yes - that's the fast, easy way to calculate it.

so now we ground the emitter, the base has to be 0V7. who wins the argument, the divider resistors or the transistor base? do I just calculate near 0V7 and let the device worry about it, use a series base resistor, or do I chuck 1 resistor and use collector / self bias? how do I set the collector current and Vc, or calculate the gain?

By grounding the emitter, you have removed the ability to hand-wave away the variation in what Vbe actually is. The difference between a whisper of conduction at 0.5V and a lot at 0.7 now matters compared to the emitter voltage.

In fact, as one of my professors explained to me, one of the BIG points of using an emitter resistor is to make a static voltage that is large compared to the 0.2V or so variation of Vbe, precisely so the variations in Vbe will not matter to biasing. By invalidating that, you can no longer ignore the variation. And that makes the question about whether the voltage divider or Vbe win be irrelevant. Vbe will "win", but that tells you little about what is happening.

In fact, with a zero emitter resistance, you are better off looking at the current into the base, not the base voltage.

Approximations are useful as long as the assumptions that make the approximations true are valid. You've just walked off the edge of one of these.

[EATyourGuitar]

I just got done reading 3 books on BJT but none of them explained the important stuff as clearly as you just did

[Thecomedian]

I just got done reading 3 books on BJT but none of them explained the important stuff as clearly as you just did

that's usually how it goes with those books, pamphlets, and sheets.

[gjcamann]

I love these discussions! Really gets the brain going in the morning. Thanks R.G.

[duck_arse]

I thort it would end up more complexications than I could handle.

I ended up playing with one of these the other, and just stuck in a series base resistor. so I did end up worrying about current into the base. it worked, too.

[gritz]

I think that one additional variable that's not been mentioned is that the forward voltage of the p-n junction is very temperature dependant and without correction it may form a positive feedback loop:

Heat -> lower forward voltage -> more current -> more heat, etc.

This thermal runaway isn't a problem in yer average low current preamp application ('cos power dissipation is minimal at the kind of currents that we usually use), but the absolute temperature does toss another variable into that "forward voltage = 0.6V. Sort of" approximation. Hfe varies with collector current (and voltage) too. Our meters only measure these characteristics at one fixed point, so these variations aren't apparent to us.

It's all part of the fun.