BJT and MOSFET Amplifiers

Pick by impedance, not by gain. Common source/emitter inverts and gives voltage gain; common drain/collector is a buffer; common gate/base is a current buffer with low input impedance.

A resistor in the source or emitter trades gain for linearity and raises input and output impedance. It is negative feedback in its simplest form.

The capacitance between input and output appears at the input multiplied by the gain — which is usually what sets the upper cut-off.

A mirror copies a reference current, scaled by the device ratio. A differential pair splits a fixed tail current between two branches according to the input difference.

Biasing sets the Q-point; coupling keeps it there while the signal passes through. The whole art is making the operating point depend on resistor ratios — which are stable — rather than on $\beta$ or $V_{BE}$, which vary between devices and with temperature.

Every amplifier question starts by killing the DC sources and replacing the device with its model. $g_m$ comes from the bias current, which is why biasing is solved first and only then the gain.

Three configurations, three jobs: common emitter/source for voltage gain, common collector/drain for buffering, common base/gate for a low input impedance and good high-frequency behaviour. Memorise the table — deriving it under exam pressure wastes the time the question was testing.

Negative feedback divides the gain by $1+A\beta$ and does the same to distortion and to sensitivity, while multiplying the bandwidth by it. What it does to the impedances depends on the topology, and that is what the questions test.

A cascode puts a common-base stage on top of a common-emitter one: the lower device sees almost no voltage swing, so the Miller effect nearly vanishes and the bandwidth rises without giving up gain.