Triode
A triode is a three-electrode vacuum tube: a heated cathode that boils electrons into the vacuum, a plate (anode) that collects them, and a wire-mesh control grid placed between the two. Because the grid sits close to the cathode, a small voltage on it can raise or choke off a much larger current flowing to the plate, which makes the triode the first practical electronic amplifier. It is the device that the transistor was invented to replace, and the grid-controls-current idea survives, almost unchanged in spirit, in every modern transistor.
The triode began as a diode with one extra wire. John Ambrose Fleming's 1904 vacuum diode passed current in only one direction, which made it a detector of radio signals. In 1906 the American inventor Lee De Forest inserted a third electrode, a zigzag of wire and later a mesh, between filament and plate and patented the result, which he called the Audion. His first understanding of why it worked was imperfect, and early Audions contained residual gas that made them erratic. The decisive improvements came around 1912 to 1913, when high-vacuum versions were developed at the General Electric laboratory under Irving Langmuir and at AT&T and Western Electric, where Harold Arnold and colleagues turned the device into a dependable telephone repeater amplifier. Edwin Armstrong's regenerative circuit of 1912 showed that the same tube could also be made to oscillate, and with that the triode became both the amplifier and the oscillator of radio.
The grid works by modulating how many electrons escape the cathode's cloud of charge, not by touching them. A cloud of negative charge hovers around the hot cathode, and the positive plate pulls electrons out of it across the gap. A grid held at a small negative voltage relative to the cathode repels electrons and reduces the plate current; making it less negative lets more through. Because the grid is negative it collects almost no current of its own, so signal power drawn from the source is tiny, while the plate circuit can deliver far more. A load resistor in the plate circuit turns the changing current into a changing, larger voltage: that is amplification.
Three published numbers describe a triode, and they are linked. The amplification factor, written μ (mu), is the ratio of how much the plate voltage would have to change to match the effect of a one-volt change on the grid; typical small-signal triodes have μ from about 17 for the 12AU7 up to about 100 for the 12AX7. The transconductance, written gm, is the change in plate current per volt on the grid, in milliamperes per volt, usually a few milliamperes per volt for receiving tubes. The plate resistance rp is how much the plate voltage must change per unit change in plate current. They obey the relation μ = gm × rp, so a high-gain tube gives up current-handling ability and a low-impedance one gives up gain. Real stages deliver less than μ because the load resistor forms a voltage divider with rp, a first illustration of why gain and output capability always trade against each other.
The triode is the conceptual ancestor of the transistor, and the analogy is specific. The grid corresponds to the gate of a field-effect transistor: a voltage on an insulated or reverse-biased control electrode steers a current between two other terminals, the cathode and plate corresponding to source and drain. The bipolar junction transistor is a looser match, since its base takes a small current rather than just a voltage, but the cathode, grid and plate map onto emitter, base and collector in role. Bell Labs, under Mervin Kelly, wanted a solid replacement for tubes in telephone amplifiers and switching, and the first field-effect schemes of William Shockley's group were literally attempts to build a solid-state triode, in the same line as Julius Lilienfeld's 1925 to 1926 patents. The point-contact transistor demonstrated in December 1947 reached that goal by a different route and is often described as a solid-state triode.
Tubes lost on size, heat and lifetime, not on principle. A triode needs a glowing filament, typically consuming a watt or more to heat, and the filament burns out, and the glass can crack or lose vacuum. The ENIAC computer of 1945 used about 17,468 tubes and was notorious for the labor of keeping them working. A transistor needs no warm-up, runs on a small fraction of the power, and can be made in billions on a single wafer. The simple comparison that a tube is large and a transistor small is true but incomplete; the deeper change is that transistors can be manufactured by photolithography, whereas tubes were assembled one by one.
Triodes are not extinct, and the common view that they are purely obsolete is a misconception. Guitar amplifiers and high-end audio gear still use dual-triode tubes such as the 12AX7 for their distinctive distortion and behavior, and high-power vacuum tubes remain in broadcast transmitters, radar, particle accelerators and industrial heating, where they handle power levels that semiconductors reach only with great difficulty. Another pitfall is thinking amplification creates energy: the extra signal power is taken from the plate power supply, and the grid merely meters it. Today's chips, including those running AI workloads, are collections of tiny voltage-controlled switches obeying exactly this logic, with the grid replaced by an insulated gate a few nanometers from the channel.
Read triode as the first device in which a whisper of control voltage steered a flood of current: the three-electrode template, a controlled path with a gate, was proved in glass and vacuum in 1906 and then rebuilt in crystal, which is why the transistor inherited its job and much of its vocabulary.