Thyristor
A thyristor is any member of a family of four-layer (p-n-p-n) semiconductor switches that latch into conduction once triggered and stay on until the current through them falls to near zero. Unlike a transistor, which follows its control signal continuously, a thyristor is a regenerative device with two stable states, blocking and conducting, and a brief pulse at its gate is enough to flip it from one to the other. The family includes the silicon controlled rectifier (SCR), the TRIAC for alternating current, the gate turn-off thyristor (GTO) and the DIAC trigger diode, and together they handle some of the largest electrical power levels that any semiconductor device is asked to switch.
The four alternating layers behave like two transistors wired to feed each other, and that feedback is the whole secret of latching. Between the anode and the cathode sit four layers: p, n, p, n. Read as two transistors, the upper three layers form a pnp transistor and the lower three an npn transistor, with the middle two layers shared. The collector current of each one is the base current of the other. At low current the loop gain is below one and the device blocks, but a small gate current into the lower p layer raises the loop gain toward one, after which each transistor drives the other harder and the whole structure snaps on within microseconds. Once latched, the gate has no further say, which is why a thyristor is described as a latch rather than an amplifier. This two-transistor picture is a model, not a literal assembly of two parts, but it predicts the triggering, latching and holding behaviour correctly.
The family members differ mainly in how they are triggered and which directions of current they control. The SCR conducts in one direction only and is the workhorse of controlled rectifiers. The TRIAC is effectively two thyristors in inverse parallel sharing one gate, so it can be triggered in either half-cycle of an alternating supply, which is why it dominates lamp dimmers, heater controls and small motor speed controllers. The DIAC is a two-terminal device that conducts abruptly once the voltage across it passes a breakover value, and it is commonly used to produce the sharp trigger pulse for a TRIAC. The GTO adds a structure that lets a large negative gate current extract enough carriers to break the latch, so it can be switched off on command, at the cost of a considerable gate-drive effort and a modest turn-off gain. The SCR entry in this glossary covers the single device in detail; the thyristor view here is the broader family and what it is used for.
Thyristors control power by choosing when in each cycle of an alternating waveform to turn on. In phase control, the gate is triggered at an adjustable delay after the supply voltage crosses zero. Triggering early lets nearly the whole half-cycle through, triggering late lets only a sliver through, and the device turns itself off at the next current zero, so no separate turn-off circuit is needed. This is cheap, efficient control, because a switch that is either fully on or fully off dissipates little heat. The price is the chopped waveform, which injects harmonics and a lagging power factor into the supply, and the practical need to limit the rate of rise of voltage and of current, since a fast dV/dt can trigger a thyristor falsely and a fast di/dt can concentrate current in a small region of the die.
At the top of the power scale, thyristors become the valves of high-voltage direct-current transmission. The four-layer switching structure was studied at Bell Labs in the mid-1950s, and General Electric commercialized the silicon controlled rectifier in 1957. In the following decades, stacks of series-connected thyristors, often triggered by light through optical fibres to keep the control side isolated, replaced mercury-arc valves in converter stations that change alternating current to direct current and back. A modern thyristor valve may contain many devices in series, each on a wafer-sized silicon disc pressed between cooled metal contacts. Line-commutated converters of this kind move very large amounts of power over long distances and between grids that are not synchronized, and the thyristor remains a mainstay of that business even as newer transistor-based converters take on other roles.
A common misconception is that a thyristor can be turned off by removing its gate signal, or that it behaves like a power transistor that is simply bigger. For an ordinary SCR, taking the gate away changes nothing once the device has latched; the anode current must be forced below the holding current, usually by the alternating supply reversing or by a dedicated commutation circuit. Where a design needs fast switching off at will, as in motor drives and inverters, the insulated-gate bipolar transistor and the MOSFET largely took over from the GTO, because they turn off with a voltage on the gate and need no heavy negative current. The thyristor survives where its strengths matter: extreme voltage and current ratings, robustness against surges, a low conduction loss, and a long record of reliability.
Read thyristor as a semiconductor latch built from two transistors that feed each other: a tiny trigger turns it on, the supply decides when it turns off, and that blunt, rugged behaviour is exactly what makes it the right switch for dimmers, rectifiers and the valves of the largest power lines.