A semiconductor device having three electrodes: emitter, base, and collector. It is effectively a sandwich of two types of doped semiconductor, usually p-type and n-type silicon, and so contains two p-n junctions. When the region common to both junctions is p-type, an npn transistor is formed; when it is n-type a pnp transistor is formed. This central region forms the base electrode.

Bipolar transistors are so named because both charge carriers, i.e. electrons and holes, contribute to the flow of current. Current flow between collector and emitter is established by applying a forward bias between base and emitter. In linear (i.e. nonsaturated) operation, the magnitude of this current is proportional to the input current drawn at the base. The current flow is in opposite directions in npn and pnp transistors.

If the base current is increased but the collector current is restrained, so that the transistor effectively receives more base current than it would seem to require, the transistor is driven into a state of saturation, also known as bottoming. It then behaves as a very efficient switch since the base-collector junction becomes reverse biased and, in saturation, the collector-emitter voltage can fall as low as 20 millivolts. The device thus seems virtually a short circuit. Bipolar transistor switches, working into saturation, are the basis of TTL circuits. Saturated transistors do however have a fairly low switching speed. The much higher switching speeds of Schottky TTL and ECL circuits are achieved by using a nonsaturated mode of operation.