Emitter Follower Biasing

[petemoore]

Emitter follower biasing articles to read around here?
 I see some recent posts asking about bias's after posting what they have...I'd like to try being more help...
  I always wired 'em right and the always worked for me. I just test B/E 'sound' with an audio probe or signal injector and look for 1 to 1...same/same sound at B or C.
 I'd like to know the operational voltages necessary for E's [base and collector too] to follow.
 I think the 'top' lead of the transistor generally [always?] goes right to V+ [in neg gnd cct.].
 Source followers too...any good reads?

[R.G.]

Here's the entire secret to biasing bipolar transistors: in the linear region, the base emitter **can't** be anything other than one diode drop apart.

So - you want an emitter follower. That means that the signal comes out the emitter.

That means that for biggest signal swing up and down, the emitter needs to sit at half the power supply voltage (assuming you aren't trying to make the emitter follower clip for some reason; that's not usually why people use EF's).

That means that the voltage across the emitter resistor is half the power supply.

That means that the base HAS to be one diode drop above (NPN) the emitter. Well, actually, it means that to get the emitter to half the power supply, you have to pull the base up to half-supply plus a diode with some biasing voltage. The emitter... follows 8-)

Ohm's law now tells you what your resistor is. You decide what current you want flowing through the transistor (make this 10x any load current) and then Re = (1/2Vcc)/current. If you pick currents too big, you will need mondo biasing currents to make that come true, and the transistor will dissipate lots of heat. If your biasing network can't supply the base with current/hfe, then the base voltage will sag, and so will the emitter voltage.

What about collector resistors? The collector resistor eats voltage too. It eats current*Rc volts. For biggest signal voltage swing (if that's what you're after) Rc = 0.

And that's how to bias bipolars. Period. Even for gain stages, you just reapportion how much voltage is dropped over the collector and emitter resistors and set the base to make the emitter be where you want it.

As to source followers:

FETs are more complicated than bipolars because although the source does follow the base, the offset voltage is not as tidy as always one voltage drop. MOSFETs have a positive offset voltage of gate to source to get them to conduct. This is in the range of 0.1 to 5.0V. You have to know your MOSFET. Read "Designing MOSFET Boosters" at GEO for how to do this.

For JFETs, the offset of gate to source is negative and big. The offset must be in the range of voltages to reverse bias the gate-source diode. A reverse voltage of 0V is fully ON for JFETs and a reverse bias of Vgsoff is fully off. In the middle is the linear region. The range of Vgsoff is from as little as 0.1V to as much as 12V. This is what makes JFETs hard to work with.

In all cases, you decide what voltage you want your source to sit at and how much current you want to flow in the source resistor. Then you put the gate at the voltage that makes that come true. It's just harder to figure out what that voltage is.