To test a Bipolar Junction Transistor (BJT) using a digital multimeter, set your meter to Diode Test mode and measure the forward voltage drop across the base-emitter and base-collector junctions. A healthy silicon transistor will read between 0.500V and 0.800V (500mV to 800mV) when forward-biased, and display OL (Over Limit) when reverse-biased. If you are testing a MOSFET, stop here—this specific diode-drop method applies strictly to BJTs (NPN and PNP), as MOSFETs require a gate-charge threshold test.

Testing a transistor is essentially testing two internal PN-junction diodes wired back-to-back. By systematically probing the three leads, you can determine if the junctions are intact, shorted, or leaky without needing a specialized semiconductor curve tracer. Below is the exact bench procedure, the numeric boundaries for pass/fail, and the decision path to get your circuit back online.

Meter Setup and Safety Category Requirements

Before you touch the probes to the component, you need to configure your multimeter correctly and verify the safety environment. Testing semiconductors requires the meter to output a small, regulated constant current (typically 1mA to 2mA) to forward-bias the silicon junction. The standard resistance (Ω) or continuity modes do not provide a reliable test current and will give you meaningless, fluctuating numbers.

⚠️ SAFETY & CAT RATING WARNING:
Component-level bench testing falls under CAT I measurements. However, if you are probing a transistor inside a power supply or an inverter circuit that interfaces with mains voltage, the environment is CAT II or CAT III. You must completely de-energize the circuit, unplug it from the wall, and bleed all filter capacitors using a high-wattage bleeder resistor before testing. A charged 400V DC bus capacitor will instantly destroy your multimeter's internal shunt and can deliver a lethal shock, regardless of the meter's CAT rating.

Meter Configuration Block

  • Dial Position: Diode Test mode (symbol: an arrow pointing into a line ->| ). On some meters, this shares a position with continuity and requires a press of the 'Mode' or 'Select' button to switch from the buzzer icon to the diode icon.
  • Red Lead Jack: VΩmA (or VΩ on higher-end Fluke/Brymen meters).
  • Black Lead Jack: COM (Common).
  • Display Verification: Touch the red and black probes together. The meter should read between 0.000V and 0.010V. Separate them; the display should immediately snap to OL (or 1 on older displays). If it doesn't, your leads are broken or your meter battery is dying.

Probe Placement and the Diode Test Method

A standard TO-92 packaged transistor (like the ubiquitous 2N3904 or BC547) has three leads. When you hold the transistor with the flat side facing you and the leads pointing down, the pinout from left to right is typically Emitter (E), Base (B), Collector (C). Always verify this against the specific manufacturer's datasheet, as Japanese 2S-series transistors use a different pinout (E, C, B).

The Base is the common terminal. You will test from Base-to-Emitter and Base-to-Collector. You do not need to test Collector-to-Emitter directly, as that path should be open in both directions when the base is floating.

Testing an NPN Transistor (e.g., 2N3904, 2N2222)

  1. Forward-Bias B-E: Place the Red probe on the Base. Place the Black probe on the Emitter. Record the voltage.
  2. Forward-Bias B-C: Keep the Red probe on the Base. Move the Black probe to the Collector. Record the voltage.
  3. Reverse-Bias B-E: Swap probes. Black on Base, Red on Emitter. Record the reading.
  4. Reverse-Bias B-C: Keep Black on Base, move Red to Collector. Record the reading.
  5. C-E Isolation: Place probes on Collector and Emitter in both polarities. Both should read OL.

Testing a PNP Transistor (e.g., 2N3906, 2N2907)

The procedure is identical, but the polarity is reversed. The Black probe goes on the Base to forward-bias the junctions, and the Red probe touches the Emitter and Collector. Reverse-biasing requires the Red probe on the Base.

Expected Readings: Good vs. Bad Junction Values

The numbers on your screen represent the forward voltage drop ($V_f$) required to push the meter's 1mA test current through the silicon PN junction. According to ON Semiconductor's 2N3904 datasheet, the base-emitter saturation voltage typically hovers around 0.65V to 0.85V under active load, but a multimeter's low test current will usually yield a reading right in the middle of the silicon threshold band.

Table 1: BJT Junction Test Expected Values (Silicon)
Test Condition Probe Polarity (Base to E/C) Expected Good Reading Indication if Outside Range
Forward-Bias (NPN) Red on Base, Black on E/C 0.500V to 0.800V <0.400V = Leaky/Shorted; >0.900V = High resistance/Bad contact
Forward-Bias (PNP) Black on Base, Red on E/C 0.500V to 0.800V <0.400V = Leaky/Shorted; >0.900V = High resistance/Bad contact
Reverse-Bias (Both) Opposite of Forward OL (Over Limit) Any numeric value = Leaky junction (Failed)
Collector to Emitter Both polarities OL (Over Limit) Any numeric value = C-E Short (Catastrophic failure)
💡 BENCH TIP: Germanium vs. Silicon
If you are repairing vintage audio gear or guitar fuzz pedals (like a Dallas Arbiter Fuzz Face), you might encounter Germanium transistors (e.g., AC128, 2N139). Germanium has a much lower bandgap. A good forward-bias reading for a germanium BJT is 0.150V to 0.300V. If you see 0.200V on a vintage board, do not throw it away thinking it is shorted!

Common Mistakes That Give Misleading Readings

Even with the meter set correctly, environmental and physical factors can skew your $V_f$ readings, leading you to throw away good components or solder in bad ones. All About Circuits highlights several parasitic paths that ruin bench measurements.

1. Testing In-Circuit (The Parallel Path Trap)

If you try to test a transistor while it is still soldered to the PCB, the surrounding components (resistors, transformer windings, diodes) create parallel electrical paths. A 10kΩ biasing resistor in parallel with the base-emitter junction will pull your multimeter reading down to 0.100V or lower, making a perfectly good transistor look dead-shorted. Rule: You must desolder at least two of the three leads (lifting the Base and Emitter is usually sufficient) to isolate the transistor from the circuit.

2. The 'Finger Resistance' Error

The human body has a DC resistance ranging from 10kΩ (sweaty skin) to 100kΩ (dry skin). If you hold the transistor body in one hand and pinch the metal probe tips and the transistor leads together with your bare fingers, your body resistance parallels the PN junction. Because the multimeter's diode test is highly sensitive to parallel impedance, your finger resistance can drag a normal 0.650V reading down to 0.550V or cause a reverse-bias test to show a 'leaky' 1.500V instead of OL. Always use alligator clips, a PCB holder, or hold only the insulated probe shafts and the plastic body of the transistor.

3. Dirty or Oxidized Leads

Old TO-92 transistors pulled from a junk bin often have oxidized tinning on the legs. If the probe tip doesn't bite through the oxidation, the contact resistance adds to the junction drop, pushing your reading up to 1.100V or higher. Scrape the leads lightly with a hobby knife before testing.

Decision Tree: Keep, Trash, or Replace?

Use this decision matrix to interpret your measurements and take immediate action. Do not waste time trying to 'dry out' or 'recover' a failed silicon junction—once the internal crystal lattice is compromised by thermal runaway or overvoltage, the gain ($h_{FE}$) is permanently degraded.

Table 2: Transistor Diagnostic Decision Path
Symptom / Reading Diagnosis Action & Concrete Replacement Pick
0.550V - 0.750V (Fwd) and OL (Rev) on both junctions. Junctions are healthy and intact. KEEP. Reinstall or use in your prototype.
0.000V - 0.150V in both directions on B-E or B-C. Junction is dead-shorted. Likely caused by exceeding max collector current ($I_c$) or thermal runaway. TRASH & REPLACE. Buy ON Semi 2N3904G (NPN) or 2N3906G (PNP) (~$0.15/ea at Mouser/DigiKey).
OL in both directions (Forward and Reverse). Junction is blown open. Internal bond wire melted due to overcurrent spike. TRASH & REPLACE. For higher power needs, upgrade to ON Semi 2N2222AG (NPN, 600mA continuous).
0.600V (Fwd) but reads 0.800V - 1.500V (Rev) instead of OL. Junction is leaky. The depletion region is compromised; it will cause massive thermal drift in analog circuits. TRASH. Do not use for audio amplifiers. Replace with exact matched pair if in a differential stage.
B-E reads 0.650V, but B-C reads 0.450V (Fwd). Asymmetrical junction degradation. Common in transistors subjected to reverse-bias base-emitter breakdown. TRASH. The $h_{FE}$ will be highly non-linear. Replace with Fairchild/ON BC547B for low-noise general purpose.

By strictly following the diode-test method and referencing the numeric boundaries above, you eliminate the guesswork from semiconductor troubleshooting. If your readings fall outside the 0.500V–0.800V forward and OL reverse parameters on an isolated silicon BJT, the component is definitively failed. Bin it, grab a fresh 2N3904 or 2N3906 from your parts drawer, and get your circuit back on the bench.