a push-pull output stage

A push-pull output stage is how an amplifier drives a real load — a speaker, a motor, a cable — with strong current in BOTH directions. It uses two complementary transistors, an NPN to PUSH current out into the load (source) on the positive half of the signal, and a PNP to PULL current back from the load (sink) on the negative half. They take turns, like two rowers on opposite sides of a boat, so the output can swing high and low with plenty of muscle in each direction.

Why two transistors? A single emitter follower can push current out hard but can only pull weakly through its bias resistor, wasting power and limiting the negative swing. The push-pull pair fixes this: the NPN handles the top half, the PNP handles the bottom half, and at any instant only the one that is needed is doing the work. In its simplest form (class B) each transistor is off until the signal exceeds its 0.7 V turn-on, so near the zero-crossing there is a moment when NEITHER transistor conducts — producing the notorious CROSSOVER DISTORTION, a kink in the waveform as the baton is fumbled in the handover. The standard cure is class AB: bias both transistors slightly on (with diodes or a Vbe-multiplier providing about 1.2 V of bias) so that each conducts a little past the crossover and the handover is smooth.

This is the output stage of practically every audio power amplifier and op-amp. The class AB compromise — a small idle (quiescent) current to kill crossover distortion without the wasteful full-time current of a class A stage — is one of the most important practical circuits in analog electronics. The honest pitfalls: that idle current must be set carefully and STABLE, because the transistors heat up, their turn-on voltage falls, and the idle current can climb — the seed of thermal runaway. Good push-pull stages therefore add small emitter resistors and thermally couple the bias network to the output transistors, so the idle current is policed as temperature rises.