A bad transistor can sometimes be detected by its partly burned or distorted appearance, but more often there is no visible indication. One approach to troubleshooting is to substitute a known good component, but that is a costly way to go. Also, it is not reliable because an outside defective component can instantly destroy the replacement with no visible evidence. The sensible alternative is to test the transistor. A common multimeter can quickly do in-circuit tests, which are not totally definitive but generally provide acceptable go/no-go information, using either the meter’s diode-check or ohms mode.

The usual testing procedure is for use with a digital multimeter in the diode test-range with a minimum of 3.3 V over d.u.t. (diode-under-test). First consider the procedure for testing an enhancement-mode MOSFET (i.e. where the device is non-conductive with 0 V applied to the gate, operating like a switch). Connect the Source of the MOSFET to the meter’s negative lead. (Hold the MOSFET by the case or the tab but don’t touch the metal parts of the test probes with any of the other MOSFET terminals until needed.) Touch the meter positive lead onto the MOSFET Gate. Now move the positive probe to the ‘Drain’. You should get a low reading. The MOSFET’s internal capacitance on the gate has now been charged up by the meter and the device is ‘turned-on’.

With the meter positive lead still connected to the drain, short the source and gate. The gate will discharge and the meter reading should go high, indicating a non-conductive device.

MOSFETS that fail often go short-circuit drain-to-gate. This can put the drain voltage back onto the gate where it feeds (via the gate resistors) into the drive circuitry, possibly causing voltage and current levels exceeding the limits of components in that section. The overload will also affect any other paralleled MOSFET gates. Thus it’s best to check the drive circuits of dead MOSFETs. To avoid overloads, some designers add a zener diode between the source and gate–zeners fail short so they can limit damage in the event of a MOSFET failure. Another tactic is to add subminiature gate resistors. They tend to open (like a fuse) when overloaded, disconnecting the MOSFET gate.

Another frequent failure mode for a FET is a shorted drain-to-source. An ohmmeter test can verify the problem. Connect the gate of the device to the source terminal. If the drain-source path is healthy, applying the ohmmeter probes in one direction should read a short. The other direction should measure infinite resistance – or at least a few megaohms. The diode junction measured is the FET body diode. The body diode will show cathode at the drain for an N-channel device and at the source for P-channel device.

Unfortunately, modern multimeters use a low excitation to measure resistance (1–2 V) to ensure simple active probing of circuit elements doesn’t damage them. The problem is that testing a FET with a modern multimeter alone becomes problematic. The reason is most power FETs require a gate-source voltage bias of at least 4-5 V to switch on. Logic-Level FETs can be switched on at voltages of between 0.3 and 1.5 V.

A simple circuit shown here for an N-channel FET helps detect whether the device works correctly as a switch. The multimeter should read a voltage between points 2 and 4 that is quite low. Measurement of the device’s R_dsON begins by removing the link between points 1 and 2, then measuring between point 2 and point 4 to obtain an approximate reading of resistance on the multimeter.

With points 1 and 2 shorted together, measure the voltage between point 2 and point 4, then short point 3 to point 4. You should see the voltage vary from low in first test to the actual battery voltage applied (usually 9 V).

You can determine if any residual leakage between Drain and Source by shorting point 3 and point 4, then measuring the voltage across the 100KΩ resistance feeding point 1 from the battery. Then leakage current in milliamps is approximately = (multimeter reading in millivolts)/(10⁴). To measure the nominal V_gs threshold (starting-to-turn-on voltage) of the FET, short point 2 and point 3 and then measure the voltage between point 2 and point 4 as before.

When examining p-channel enhancement-mode MOSFETs, just invert the battery polarity and use the same circuit. All multimeter probing polarities will be reversed, but the same procedure still applies.

Now consider JFETs. Checking a JFET as a diode (gate-to-channel junction) with an ohmmeter should indicate low resistance between gate and source with one polarity and a high resistance between gate and source with meter polarity reversed. If the meter indicates high resistance with both the polarities, the gate junction is open. On the other hand, if the ohmmeter indicates low resistance with both polarities, the gate junction is shorted.

Now consider testing continuity through the drain-source channel. If you know which terminals on the device are the gate, source, and drain, it’s best to connect a jumper wire between gate and source to eliminate any stored charge across the capacitance of the gate-channel PN junction which could hold the JFET in a pinched-off state without any external voltage applied. Without this step, any meter reading of continuity through the channel will be unpredictable because a charge may or may not be stored by the gate-channel junction.

A good strategy to follow is to insert the JFET pins into anti-static foam prior to testing. The conductivity of the foam makes a resistive connection between all JFET terminals. This connection ensures all residual charge built up across the gate-channel PN junction dissipates, thus opening the channel for an accurate test of source-to-drain continuity.

Because the JFET channel is a single, uninterrupted piece of semiconductor material, there is usually no difference between the source and drain terminals. A resistance check from source-to-drain should yield the same value as a check from drain to source. This resistance should be relatively low (below a few hundred ohms) when the gate-source PN junction voltage is zero. Applying a reverse-bias voltage between gate and source should pinch-off the channel and produce a higher resistance reading on the meter.

This brings us to bipolar transistors. It’s helpful to remember a bipolar transistor can be modeled as two diodes connected in series. The floating leads provide two test points, and the connected leads are a third center-tap test point. These two diodes will not function as a real transistor because the center-tap connection is not a semiconducting junction, and the two-diode model does not have three separate silicon layers as in a transistor. However, the hookup demonstrates the basic concept involved in transistor testing and identifying the terminals.

To test a transistor using a multimeter in the diode-test mode, insert the black probe into Common and the red probe into Diode Test or Ohms. Most manufacturers connect red to the positive terminal of the internal battery but this can vary, so it is best to check the polarity using a second multimeter in dc Volts mode. The usual test voltage is 3 V.

It is natural to assume the center lead on the transistor package connects to the base, but this convention is not universal. Connect the black probe to the base. Briefly touch the red probe to the emitter and note the voltage. Then switch the red probe to the emitter. If the readings are the same, so far so good. Removing the black probe from the base and replacing it with the red probe, briefly touch the black probe to the emitter and collector.

If the previous readings were high and these readings are low, the transistor passes the static test. If the previous readings were low and these readings are high, the transistor also passes the static test. If the two red probe readings are not the same or the two black probe readings are not the same when reversing the probes, the transistor is bad.

If base, emitter and collector identities are unknown, connect the black probe to one the transistor leads. In turn, briefly touch the red probe to each of the remaining leads. If both leads read high, the black probe is connected to the base, the transistor is NPN and it is good. If there are different readings on the other two leads, move the black probe to another lead and touch the red probe to the remaining leads. Repeating the test with the black probe touching each of the three leads in turn, you should have high resistance, and the transistor is either bad or PNP.
Remove the black prove and connect the red probe to one of the leads. Then touch the black probe in turn to each of the remaining leads. When each of the leads is touched and resistance reads high, the red lead is connected to the base and the transistor is a good PNP device.

If you get two different readings for the two leads, move the red probe to another lead and repeat the test. Connect the red probe in turn to each of the three leads. If the other two leads don’t give the same reading when touched by the black probe, the transistor is PNP and it is bad.

Multimeter tests determine if a transistor is blown (open or shorted) and provide a rough estimate of the transistor’s ability to amplify. But they do not report actual operating parameters. To get more information, the next step up is a service-type transistor tester. This instrument performs three measurements for bipolar transistors: forward current (beta), base-to-collector leakage current with open-emitter and short-circuit from collector to emitter and base. H_fe is measured and the transistor is reported good if this metric exceeds a specific level. However, the test will reject some functional but low-level H_fe transistors.

Some service-type transistor testers can check components either in or out of circuit, and they are capable of identifying unknown transistor terminals. Because H_fe varies with the device, service-type transistor testers can give erroneous readings and are not infallible.

A highly reliable, intuitive and convenient component test can employ an oscilloscope in conjunction with the scope’s built-in waveform generator or with an external stand-alone AFG. Capacitors, inductors, bipolar transistors and cables can be easily checked and their values determined. A signal from the AFG is applied to the component under investigation and the response is displayed in the oscilloscope. Typically, a 50-Ω impedance output from the AFG is fed through a T-connection to the device under test and to an oscilloscope analog input. Additionally, AFG OUT is connected to the oscilloscope Trigger IN.

The best transistor testers are laboratory-grade instruments. A related instrument is the semiconductor curve tracer. It contains a simplified oscilloscope in addition to voltage and current sources, which the user applies to the DUT. A swept voltage is connected to the input of the transistor under test, and its current output is measured and displayed as a graph on the instrument screen. The user can adjust the applied voltage, its polarity and series impedance. When a diode is subjected to the varying voltage, various parameters such as forward voltage, reverse leakage current and reverse breakdown voltage display.

A stepped voltage can be applied to the input circuit of an FET, or stepped current applied to a bipolar transistor. The result makes it possible to ascertain the gain of a transistor or a thyristor trigger voltage. To evaluate the performance of a transistor, the impedance presented to it (“load pull”) can be systematically varied. Load pull is applicable when variation of the load impedance causes the oscillation of the center frequency to shift from its nominal value.