When a prototype board acts dead or a driver stage misbehaves, the bipolar junction transistor (BJT) is often the prime suspect. If you are searching for transistor how to check procedures, the direct answer is simple: use your multimeter’s diode test mode to check the base-emitter and base-collector PN junctions. A healthy NPN transistor will read between 0.500V and 0.700V when forward-biased, and show an open loop (OL) when reverse-biased. But knowing if the silicon is dead is only half the battle; understanding why it died and how to bias it correctly keeps the magic smoke inside the package.
Identifying Pins and the Safe Default Part Numbers
Before you can test a transistor, you must know its pinout. Visually, the NPN schematic symbol features an arrow on the emitter leg pointing outward (away from the base), while the PNP arrow points inward. For the physical TO-92 plastic package (the most common through-hole BJT), hold the flat side facing you with the legs pointing down. For the ubiquitous 2N3904 and PN2222, the pins from left to right are Emitter (E), Base (B), and Collector (C). Always verify this against the specific manufacturer's datasheet, as European BC-series transistors often use a Collector-Base-Emitter layout.
When designing a new circuit or replacing a fried component without a schematic, stick to these safe default part numbers:
| Part Number | Type | Max I_C | Max V_CEO | Package / Best Use |
|---|---|---|---|---|
| 2N3904 | NPN BJT | 200 mA | 40 V | TO-92 / Low-power logic switching |
| 2N3906 | PNP BJT | 200 mA | 40 V | TO-92 / High-side switching |
| PN2222A | NPN BJT | 600 mA | 40 V | TO-92 / Medium current drivers |
| TIP120 | NPN Darlington | 5 A | 60 V | TO-220 / High-current motors and solenoids |
| IRLZ44N | N-Channel MOSFET | 47 A | 55 V | TO-220 / Modern logic-level high-power switching |
How to Check a Transistor with a Multimeter
A BJT is essentially two diodes sharing a common anode (NPN) or cathode (PNP). We exploit this physical reality to test it. Set your multimeter to the Diode Test mode (usually indicated by a diode symbol). Do not use the resistance (Ohms) mode, as the low test voltage won't forward-bias the silicon junctions properly.
Testing an NPN Transistor (e.g., 2N3904):
- Find the Base: Place the Red (+) probe on the middle pin (Base) and the Black (-) probe on the left pin (Emitter). You should read between 0.500V and 0.700V.
- Check Base-Collector: Keep the Red probe on the Base and move the Black probe to the right pin (Collector). You should again read 0.500V to 0.700V.
- Verify Reverse Blocking: Swap the probes. Put the Black probe on the Base and the Red probe on the Emitter, then the Collector. Both readings must be OL (Open Loop). If you read a voltage drop here, the junction is shorted and the transistor is dead.
- Check Collector-Emitter: Place probes across the Collector and Emitter in both directions. Both must read OL. A reading of 0.000V or a low resistance means the transistor has suffered a catastrophic thermal short.
For a PNP transistor, simply reverse the probe colors: Black on Base, Red on Emitter/Collector for the forward bias readings.
Operation Regions and Biasing for the Job
Testing tells you if the transistor is alive; biasing determines if it will do the job you assigned it. A BJT operates in three distinct regions, dictated by the voltages applied to its junctions. For switching applications (like driving a relay from a microcontroller), you only care about Cutoff and Saturation. For audio amplifiers, you operate in the Active region.
| Region | V_BE (Base-Emitter) | V_CE (Collector-Emitter) | Collector Current | Primary Use |
|---|---|---|---|---|
| Cutoff | < 0.5V | Equal to Supply Voltage | 0 A (Leakage only) | Switch OFF |
| Active (Linear) | ~0.65V | > 0.3V (Variable) | I_C = β × I_B | Amplification |
| Saturation | ~0.7V to 0.8V | < 0.2V (V_CE(sat)) | Limited by external load | Switch ON |
The most common beginner mistake is relying on the datasheet's $h_{FE}$ (DC current gain) to calculate the base resistor for a switching circuit. Datasheets like the ON Semiconductor 2N3904 specification list an $h_{FE}$ of 100 to 300. However, that gain is measured in the active region. To force the transistor into deep saturation (acting like a closed mechanical switch with minimal voltage drop), you must use a forced beta of 10. The rule of thumb for switch biasing is: $I_B = I_C / 10$.
Application Circuit: Driving a 12V Relay
Let us apply this biasing rule to a real circuit. You need to switch a 12V automotive-style relay using a 5V Arduino GPIO pin. The relay coil has a measured resistance of 160Ω.
Calculating the Load Current ($I_C$):
Using Ohm's Law, $I_C = 12V / 160Ω = 75mA$.
Calculating the Base Current ($I_B$):
Using our forced beta rule: $I_B = 75mA / 10 = 7.5mA$.
Calculating the Base Resistor ($R_B$):
The Arduino outputs 5V. The base-emitter junction drops roughly 0.7V. The voltage across the resistor is $5V - 0.7V = 4.3V$.
$R_B = 4.3V / 7.5mA = 573Ω$.
We select the next standard lower resistor value to guarantee saturation: 470Ω. This provides roughly 9.1mA of base current, safely within the Arduino's 20mA recommended GPIO limit, while ensuring the 2N3904 is fully saturated.
The Complete Component List:
- Q1: 2N3904 NPN Transistor
- R1: 470Ω 1/4W Resistor (Base current limiter)
- D1: 1N4148 or 1N4007 Flyback Diode (Cathode to 12V, Anode to Collector)
- K1: 12V SPDT Relay (160Ω coil)
The flyback diode is non-negotiable. When the transistor turns off, the relay coil's collapsing magnetic field generates a high-voltage reverse spike that will instantly punch through the collector-emitter junction, destroying the BJT. For a deeper look at BJT switching topologies, Electronics Tutorials provides excellent schematic breakdowns.
War Story: The Melted PN2222 and What Went Wrong
Theory is clean; the workbench is not. A few years ago, I was tasked with debugging a custom irrigation controller board that kept failing in the field. The board used a PN2222A (TO-92 package) to switch a 12V water solenoid valve. The schematic looked correct: a 1kΩ base resistor driven by a 3.3V ESP32 GPIO, with a flyback diode across the coil.
The Setup & Numbers:
The solenoid valve was rated for 1A at 12V. However, solenoids are inductive loads with a massive stall current (the current drawn the millisecond power is applied, before the plunger moves and inductance builds). I measured the stall current on the bench at 2.4A. The PN2222A datasheet lists an absolute maximum continuous collector current of 600mA.
The Outcome:
When the ESP32 pulled the base high, the transistor tried to pass 2.4A. Because the 1kΩ base resistor only supplied about 2.6mA of base current ($3.3V - 0.7V / 1kΩ$), the transistor was pushed deep into the active (linear) region, not saturation. In the linear region, the $V_{CE}$ voltage drop was roughly 6V. Power dissipation is calculated as $P = V_{CE} \times I_C$.
$6V \times 2.4A = 14.4W$.
The TO-92 package has a 625mW maximum power dissipation limit at room temperature. The transistor was asked to dissipate 23 times its physical limit. It violently popped, cracking the epoxy case and lifting the copper pad off the PCB.
What Went Wrong & The Fix:
The designer made two fatal errors. First, they confused the PN2222A (plastic TO-92, 600mA limit) with the original 2N2222 (metal TO-18 can, better thermal mass, though still not rated for 2A). Second, they used a BJT for a high-current inductive load without accounting for stall current or power dissipation.
The fix was to rip off the BJT and replace it with an IRLZ44N logic-level N-Channel MOSFET. MOSFETs are voltage-controlled, draw virtually zero steady-state gate current from the ESP32, and have an $R_{DS(on)}$ of roughly 0.022Ω. At 2.4A, the MOSFET dissipates only $I^2R = (2.4)^2 \times 0.022 = 0.12W$, running completely cool to the touch without a heatsink. When dealing with currents above 200mA, abandon the BJT and reach for a logic-level MOSFET.
