Base
The base is the thin, lightly doped middle region of a bipolar junction transistor and the terminal connected to it. A tiny current flowing into (or out of) the base controls a much larger current flowing between the other two terminals, the emitter and the collector, which is the entire reason the region exists: it is the control input of the device. The word has nothing to do with the base of a number system or the base of a logarithm, and it also differs from the foundation or substrate of a chip; in a transistor it simply names the central layer of the three-layer sandwich.
The base must be very thin and lightly doped, because the device works only if most carriers cross it without being lost. In a typical discrete silicon transistor the base is a few microns thick, and in modern high-speed bipolar and silicon-germanium devices it shrinks to tens of nanometres. Carriers injected from the emitter into the base are minority carriers there (electrons in the p-type base of an n-p-n device), and they move across mostly by diffusion. Because the base is light on dopant atoms and narrow compared with the distance carriers normally travel before recombining, only a few percent or less of them recombine on the way, and the rest reach the far edge. The few that do recombine are replaced through the base terminal, and that replacement flow is the base current. If the base were thick or heavily doped, most carriers would be lost inside it and the device would lose its gain.
The name comes from the early hardware, and Shockley's theory later turned it into a layer. In the point-contact transistor demonstrated at Bell Labs on 23 December 1947, two metal contacts touched the top of a small block of germanium while a third electrode connected to the block's underside, and that block electrode was called the base. When William Shockley worked out the theory of the junction transistor in 1948, with the first working junction device following in 1950, the same term was kept for the middle layer of a sandwich of three doped regions. The label survived because it still described a common reference electrode, even though in the new device the base was no longer a bulk slab but a sliver sandwiched between two junctions.
A small base current controls a large collector current because the base-emitter junction sets how many carriers are injected. Forward-biasing that junction by about 0.6 to 0.7 volts in silicon (roughly 0.2 to 0.3 volts in germanium) lets the emitter pour carriers into the base, and the number is exponential in the voltage. The collector collects nearly all of them, so the collector current is a large multiple of the base current, a ratio treated under current gain. In a circuit the base is therefore usually the input: a sensor, a microcontroller pin, or the output of a previous stage drives it through a series resistor, which limits the current so the junction is not destroyed. A forgotten base resistor is a classic beginner failure, because the base-emitter junction behaves like a diode and will draw whatever current the source can supply.
Base width is not constant, and the effect of that is named after James Early. The reverse bias on the collector side widens its depletion region, which eats into the neutral base and narrows it as the collector voltage rises. A narrower base means less recombination and a steeper carrier gradient, so the collector current creeps upward with collector voltage even when the base drive is fixed. This base-width modulation was analysed by James Early at Bell Labs in 1952 and is called the Early effect. It is why real transistor output curves slope upward instead of being flat, why a single transistor amplifier has finite output resistance, and why designers quote an Early voltage on datasheets and in simulation models. Pushing the base thinner for speed makes the effect worse, a trade-off that circuit designers have to manage with current mirrors and cascode arrangements.
Common mistakes concern the base as a voltage node and as a current node at once. The base is not a clean high-impedance input like a field-effect transistor gate; it draws steady current, and the base-emitter voltage falls by about two millivolts for each degree of temperature rise, which can lead to thermal runaway if biasing is careless. Base resistance, the ohmic resistance of that thin sliver, adds noise and limits switching speed, and in radio-frequency devices it is a primary design parameter. The same word appears in other terms (base station, number base, base metal), none of which relate to the transistor terminal, so context matters. In modern integrated circuits, bipolar bases still sit in analogue amplifiers, voltage references and fast silicon-germanium radio chips found in phones and radar, while the gate of a MOS transistor has largely taken over the role of the control terminal in digital logic.
Read the base as the transistor's control layer: a sliver thin enough that nearly every carrier crosses it, yet wired out so that the few that do not can be counted, steered and amplified into a much larger current.