[Amps] voltage swing in BJT amp

[Manfred Mornhinweg]

Drax,

> Does it matter how much the voltage swings on the output of a RF power
> amplifier?

It does.

It's true that 1 millivolt at 1 milliampere is 1 microwatt, while 1 volt at 1 
microampere is also 1 microwatt, and so is 7 millivolt at 140 microampere, 
roughly. But while the power is the same, there's still a big difference between 
them: In the first case you have an impedance of 1 ohm, in the second case it's 
  1 megaohm, and in the third case it's 50 ohm!

If you try to handle RF signals at the 1 ohm impedance level, you need very 
wide, ultra low inductance circuit tracks, very low value coils, very high value 
(but high Q) capacitors, and so on. It's not practical in low power circuits. In 
practice, stray inductance will eat you alive and will render the circuit 
inoperative.

If instead you try to handle those signals at the megaohm impedance level, it's 
the stray capacitance that will eat you alive! You can handle audio signals at 
the megaohm level, of you are careful, but not RF.

So you need to design a circuit to work at an impedance level that gives you an 
acceptable trade-off between the problems caused by stray capacitance and stray 
inductance. For small signal RF circuits that's typically in the range of about 
50 to 300 ohm. The farther you stray out of this range, the harder things get, 
meaning in practice that your upper frequency limit moves down.

When you design a small signal RF amplifier, you have some choice regarding the 
impedance you will load the collector with, and that will define what voltage 
and current you get for a certain signal power. If you design for higher 
impedances, that is higher voltage, you tend to get more gain, up to a point. 
Also more efficiency, that means you need less supply current to produce a 
certain output power level. At the same time the distortion tends to increase too.

If you are making a power amplifier, then it's mandatory to use nearly the full 
voltage swing available, otherwise the efficiency will be far too low. Most 
"linear" power amplifiers are actually driven slightly into the saturation 
region, for the sake of efficiency! It's typical to see their output power rated 
at the 1dB compression point. Nonlinear power amplifiers are driven as much into 
saturation as possible, for improved efficiency. An exceptionally clean linear 
amplifier might be driven to a peak voltage of only about 70% the supply 
voltage, but this will result in really poor efficiency.

Manfred


========================
Visit my hobby homepage!
http://ludens.cl
========================