Unijunction Transistor
A unijunction transistor, or UJT, is a three-terminal semiconductor device with only one pn junction, built on a bar of lightly doped silicon that has an ohmic contact at each end (called base 1 and base 2) and a small p-type emitter region near the middle of the bar. Unlike a bipolar transistor it is not an amplifier: it is a voltage-triggered switch. Held below a particular emitter voltage it passes almost no emitter current, and once that trigger point is exceeded it snaps into a low-resistance state, which makes it an unusually simple building block for timing pulses and for firing thyristors. The device was developed at General Electric in the 1950s, and its best-known member, the 2N2646, became a standard hobbyist and industrial part for decades.
The behaviour of the UJT follows from the silicon bar acting as a resistive voltage divider with the emitter junction tapped into it. With a supply voltage across base 2 and base 1, the bar sets up a potential gradient, and the point where the emitter touches it sits at a fraction of the supply called the intrinsic standoff ratio, written as the Greek letter η (eta), typically between about 0.5 and 0.8 for common parts. The emitter junction stays reverse-biased until the emitter voltage rises to roughly η times the interbase voltage plus one diode drop of about 0.6 volts. This threshold is the peak-point voltage, and because it scales with the supply it gives circuits a trigger level that tracks the supply voltage rather than drifting with it.
Once the emitter fires, the device shows negative resistance, and that is what makes relaxation oscillators possible. When the emitter junction conducts, holes injected into the lower part of the bar sharply reduce its resistance, so the emitter voltage falls even as emitter current grows. In a relaxation oscillator a resistor charges a capacitor toward the supply, the UJT fires at the peak point and discharges the capacitor through base 1 in a brief burst, and conduction ceases when the current drops below the valley point, after which the cycle restarts. The oscillation period is approximately R × C × ln(1 / (1 - η)), so with η near 0.6 it is close to R × C. Three components and one transistor therefore give a sawtooth on the emitter and a sharp pulse on base 1, a circuit that predates easy integrated timers by many years.
The pulse from base 1 is well suited to triggering silicon controlled rectifiers and triacs, which was the UJT's most important industrial job. Phase-control circuits for lamp dimmers, heater controllers and motor drives used a UJT oscillator synchronized to the AC line so that each cycle fired the thyristor at an adjustable delay, with the charging resistor acting as the control knob. Other uses included staircase generators, sweep circuits, long-delay timers where a large timing capacitor and high-impedance emitter input allowed seconds-long intervals, and tone generators. The 2N2646 was a typical choice, with datasheet values of η of roughly 0.56 to 0.75 and a very low emitter leakage current, which is the property that permits those long delays.
A common pitfall is to treat the name as a hint that the UJT is a close relative of the bipolar junction transistor, when electrically it is a different species. A bipolar transistor has two junctions and gives current gain, while the UJT has one junction and gives a threshold switch with a negative-resistance region. The programmable unijunction transistor (PUT), such as the 2N6027, adds to the confusion: it is actually a four-layer thyristor-like part whose trigger ratio is set by two external resistors, and it mimics UJT behaviour without the bar structure. Ordinary UJTs also need a resistor in series with base 2 and one in base 1, and the emitter supply must be able to deliver at least the peak current while not supplying so much that the valley point is never reached and the part latches on.
Modern relevance is mostly historical, but the ideas live on. Cheap integrated timers such as the 555, microcontrollers and dedicated phase-control chips replaced discrete UJT oscillators in new designs, and true UJTs have become scarce, with some types discontinued. Yet the circuit is still taught because it exposes negative resistance, hysteresis and relaxation oscillation in a form a student can build on a breadboard in minutes, and the same behaviour appears in neon lamp oscillators, diacs, tunnel diodes and Schmitt triggers. Readers meeting an unfamiliar threshold-and-reset circuit can often understand it as a UJT-style relaxation oscillator in modern clothing.
Read unijunction transistor as a resistive bar with a one-junction trigger: a voltage-divider threshold that, once crossed, drops into negative resistance and dumps a capacitor into a pulse, switching rather than amplifying.