Most modern designers reach for a 555 timer or a microcontroller to generate timing pulses. But when you need to trigger a Silicon Controlled Rectifier (SCR) in a high-noise industrial environment, or build a rugged relaxation oscillator that survives massive dV/dt transients, the transistor unijonction (UJT) remains the undisputed king of the workbench. While English datasheets typically spell it "unijunction transistor," the device physics and applications are identical regardless of the regional spelling.

Here is the direct answer: a UJT is a three-terminal, negative-resistance semiconductor device used almost exclusively for timing and thyristor triggering. The safe default part number is the 2N2646, and it fires when the emitter voltage reaches roughly 63% of the base-to-base supply voltage. Below, we break down the pinout, biasing math, a complete SCR trigger circuit, and how to test one with a multimeter.

Pinout, Symbol, and the Intrinsic Standoff Ratio

Unlike a standard bipolar junction transistor (BJT), the UJT does not amplify current. Instead, it acts as a voltage-controlled switch. The physical structure consists of a lightly doped N-type silicon bar with two ohmic contacts at the ends: Base 1 (B1) and Base 2 (B2). A heavily doped P-type region is alloyed into one side of the bar, forming the Emitter (E).

The circuit symbol looks like a standard transistor, but the emitter arrow points inward toward the base bar at a sharp angle, distinguishing it from an NPN or PNP device. The key to understanding UJT operation is the intrinsic standoff ratio ($\eta$). Because the emitter is physically closer to B2 than to B1, the silicon bar acts as an internal voltage divider. The ratio of the internal resistance from the emitter junction to B1 ($R_{B1}$) versus the total interbase resistance ($R_{BB}$) is defined as $\eta$. For the ubiquitous 2N2646, $\eta$ typically falls between 0.56 and 0.75.

Bench Tip: Never confuse B1 and B2. While the device is somewhat symmetrical physically, the internal doping gradient means swapping B1 and B2 will drastically alter the standoff ratio, usually resulting in a circuit that fails to oscillate or produces severely distorted sawtooth waveforms.

Operation Regions and Biasing Requirements

To use a UJT, you apply a DC voltage ($V_{BB}$) across B2 and B1. The emitter is connected to an RC timing network. The device operates in three distinct regions depending on the emitter voltage ($V_E$) and emitter current ($I_E$).

Operation Region Emitter Voltage ($V_E$) Emitter Current ($I_E$) Device State
Cutoff $V_E < V_P$ (Peak Point) Very low (leakage only) High impedance; capacitor is charging.
Negative Resistance $V_P > V_E > V_V$ (Valley Point) Increasing rapidly Device fires; internal $R_{B1}$ drops to near zero.
Saturation $V_E < V_V$ High (limited by external circuit) Fully on; behaves like a closed switch.

The critical threshold is the Peak Point Voltage ($V_P$). The UJT will remain off until the capacitor charges to this voltage, calculated as:

$V_P = (\eta \times V_{BB}) + V_D$

Where $V_D$ is the forward voltage drop of the emitter P-N junction (typically 0.7V). Once $V_E$ hits $V_P$, the emitter injects holes into the N-type bar, flooding the B1 region with carriers. This causes $R_{B1}$ to collapse from thousands of ohms down to roughly 10–20 ohms in nanoseconds, dumping the capacitor's charge into the B1 load resistor.

Safe Default Part Numbers and Ratings

If you are sourcing UJTs for a new build or repairing legacy power electronics, stick to the proven industry standards. According to the onsemi 2N2646 Datasheet, these are the safe defaults:

  • 2N2646 (TO-92 package): The absolute workhorse. Max $V_{B2B1}$ is 35V. Total device dissipation ($P_D$) is 300mW at 25°C. Peak emitter current ($I_{EP}$) is 2A (surge). Ideal for low-to-medium power SCR triggering and audio-frequency oscillators.
  • 2N4891 (TO-92 package): A slightly higher-voltage variant. Max $V_{B2B1}$ is 40V. Often used in industrial motor controls where the DC bus rail is slightly higher.
  • 2N2160 (TO-5 metal can): Harder to find today, but rated for higher power dissipation (500mW) and used in older heavy-duty ignition systems.

Note: Always check the interbase resistance ($R_{BB}$) column on the datasheet. The 2N2646 typically ranges from 4.7k$\Omega$ to 9.1k$\Omega$. This resistance dictates your minimum and maximum timing resistor bounds.

Complete Application Circuit: UJT Relaxation Oscillator

The most common application for a transistor unijonction is the relaxation oscillator, used here to generate a pulse train that triggers an SCR gate. For a comprehensive breakdown of the waveform mathematics, Electronics Tutorials provides an excellent reference on the sawtooth generation.

Circuit Parameters:

  • Supply ($V_{BB}$): 20V DC (regulated via a 20V Zener diode from a rectified AC source).
  • Timing Resistor ($R$): 10 k$\Omega$
  • Timing Capacitor ($C$): 1 $\mu$F (film or C0G ceramic for stability)
  • Base 1 Resistor ($R_1$): 100 $\Omega$ (pulse output)
  • Base 2 Resistor ($R_2$): 470 $\Omega$ (temperature compensation)

Step-by-Step Operation:

  1. Power is applied. The 1 $\mu$F capacitor begins charging through the 10 k$\Omega$ resistor toward 20V.
  2. Assuming a 2N2646 with $\eta = 0.63$, the peak voltage is $V_P = (0.63 \times 20V) + 0.7V = 13.3V$.
  3. When the capacitor hits 13.3V, the UJT fires. The internal $R_{B1}$ drops to ~15$\Omega$.
  4. The capacitor rapidly discharges through the emitter, the internal $R_{B1}$, and the external 100$\Omega$ $R_1$ resistor.
  5. This discharge creates a sharp positive voltage spike across $R_1$, which is routed directly to the gate of an SCR to latch it on.
  6. Once the capacitor voltage drops to the Valley Voltage ($V_V \approx 2V$), the UJT turns off, and the cycle repeats.

Frequency Calculation:
The time period ($T$) is roughly $T = R \times C \times \ln(1 / (1 - \eta))$.
$T = 10,000 \times 0.000001 \times \ln(1 / 0.37) = 0.01 \times 0.994 = 0.00994$ seconds.
Frequency $f = 1 / T \approx$ 100 Hz.

Bench Scenario: When the Oscillator Refuses to Turn Off

Theory is clean; the workbench is not. Here is a real-world failure mode that catches many hobbyists and junior engineers off guard when designing phase-angle controllers.

The Setup: I was building a pulse trigger for a 120VAC / 15A kiln heating element. The control circuit used a 2N2646 UJT relaxation oscillator to pulse the gate of a 2N5060 SCR. The DC supply ($V_{BB}$) was 20V. Wanting a very sharp, high-current trigger pulse to ensure the SCR latched reliably under heavy inductive load, I chose a Base 1 resistor ($R_1$) of 22$\Omega$.

The Outcome: Upon powering the circuit, the kiln turned on instantly to 100% heat and the control dial did nothing. Worse, the 22$\Omega$ gate resistor began smoking and the UJT case grew too hot to touch within seconds.

What Went Wrong (Load Line Analysis): A UJT relies on the external circuit to force it back into the Cutoff region. It will only turn off if the current flowing through it drops below the Valley Current ($I_V$), which is typically 2mA to 5mA for a 2N2646. By using a 22$\Omega$ $R_1$, the total resistance in the B1 path (including the internal ~15$\Omega$ $R_{B1}$) was only 37$\Omega$. Even when the capacitor was fully discharged, the 20V supply was pushing $I = 20V / 37\Omega \approx 540mA$ through the device. This is massively higher than the 5mA valley current. The UJT latched in the saturation region, acting as a dead short, and delivered continuous DC current to the SCR gate instead of a pulse.

The Fix: I replaced $R_1$ with a 1 k$\Omega$ resistor. This limited the maximum B1 current to roughly 20mA, allowing the UJT to snap off cleanly once the capacitor depleted, restoring the 100Hz pulse train and giving the kiln dial full control.

How It Fails and Multimeter Testing Steps

UJTs rarely fail from age; they fail from overcurrent (exceeding the valley current as shown above) or exceeding the 35V $V_{B2B1}$ maximum rating during AC line transients. When a UJT fails, it usually shorts internally between the Emitter and Base 1, or the silicon bar cracks, opening the B1-B2 path.

You can verify a suspect UJT on the bench using a standard digital multimeter (DMM). Follow these exact steps:

  1. Test the Interbase Resistance: Set your DMM to the Ohms ($\Omega$) range. Measure across B1 and B2. You should read a resistance between 4 k$\Omega$ and 10 k$\Omega$ (depending on the specific $R_{BB}$ grade of your part). If it reads open (OL) or near zero, the device is dead.
  2. Test Emitter to Base 1 (Diode Check): Switch the DMM to Diode Test mode. Place the red probe on the Emitter and the black probe on B1. You should read a standard silicon forward voltage drop of 0.500V to 0.750V. Reverse the probes; it should read OL.
  3. Test Emitter to Base 2 (Diode Check): Keep the DMM in Diode Test mode. Place the red probe on the Emitter and the black probe on B2. You should again read 0.500V to 0.750V. Reverse the probes; it should read OL.
  4. Check for Shorts: Measure resistance between the Emitter and B1, then Emitter and B2, with the meter in Ohms mode. Both should read OL (open circuit) in both directions. If you read a low resistance, the P-N junction has blown and the UJT is a paperweight.
Safety Caveat: If you are pulling a UJT from a legacy mains-powered SCR circuit (like an old dimmer or motor drive), ensure the main filter capacitors are fully discharged before desoldering. UJTs are often connected directly to gate-drive transformers that can retain lethal voltages if the bleeder resistors have failed.

While microcontrollers have taken over complex timing tasks, the transistor unijonction remains a brilliant, rugged solution for high-noise thyristor triggering. By respecting the load-line boundaries, choosing the right $R_1$ resistor, and understanding the intrinsic standoff ratio, you can build power control circuits that will outlast the equipment they are installed in.