7
\$\begingroup\$

Is there a simple method to do a sanity check on a bipolar junction transistor whether it is still functional or defective?

schematic

simulate this circuit – Schematic created using CircuitLab

Say for example I have a circuit that used to function, and I suspect the transistor (NPN, PNP). How would I go forward to make sure the transistor is defective or how can I reasonably be sure it isn't? Do I need the datasheet for the part?

\$\endgroup\$
  • \$\begingroup\$ I was surprised not to be able to refer to a question on this topic. For many it may be simple, but for new EE's it may not be trivial. I'll wait with answering my own question, for someone who likes the easy rep ;o) \$\endgroup\$ – jippie Aug 19 '13 at 19:24
3
\$\begingroup\$

Equipment

To test a bipolar junction transistor (NPN or PNP) you don't need any special features on your multimeter other than resistance or even better the diode test.

I am personally not a fan of h(FE) measurements on a DMM as it is not defined what is actually being measured. The actual h(FE) in your circuit may be an entire order of magnitude off due to the fact that h(FE) depends on the exact DC bias of the transistor.

Equivalent circuit

To test a BJT it is essential to understand what the equivalent ciruit looks like. Here is the equivalent for an NPN:

schematic

simulate this circuit – Schematic created using CircuitLab

Method

A total of 3 × 2 = 6 simple tests suffices:

  1. Base - Emitter diode. To test the diode, check with your multimeter (preferably in diode range, otherwise ohms) if the diode's measures is around 600mV in one direction and HIGH (O.L.) in the other direction. (a) Put one probe on base, the other probe on emitter, (b) then reverse. A short or an open indicates a defective transistor.
  2. Base - Collector diode. Same procedure as with BE-diode, but with different pin.
  3. Collector - Emitter not shorted. From the equivalent circuit it can easily be seen that it is impossible for current to flow between collector and emitter. With the diode tester check (a) if the two pins measure HIGH (O.L.) in (b) both directions. A short or lowish value indicates a defective transistor.

Do you have to desolder the transistor to measure it correctly? To be absolutely sure: Yes, but it doesn't hurt to check it in circuit. If you measure a short, take out. If it measures OK in circuit, you have a fair (though not 100%) chance that the transistor is OK.

\$\endgroup\$
3
\$\begingroup\$

You can do a simple check with any multimeter that has an ohmmeter function. Use the multimeter on a low ohms scale.

Check the resistance between the base and emitter with the red lead on the base and the black lead on the emitter, then repeat with the red and black leads reversed. If the transistor is good, you should get a low reading and a high reading. Repeat this process with the base and collector. You should also get a low and a high reading.

What you are doing is checking the diode junctions; in one reading, the junction is forward biased by the battery in the multimeter and will yield a low resistance, in the second reading, the junction is reverse biased and will yield a high reading.

Also measure the resistance between the collector and emitter and then reverse the multimeter leads as before. In this case, you should get a high reading in both measurements since you are measuring across back to back junctions. If all of the measurements are OK, then the transistor is probably good, If any of them are not, then the transistor its probably bad.

By the way, multimeters vary as to the polarity of the battery during resistance measurements. In most, the positive terminal is connected to the positive battery terminal but some multimeters have this reversed.

\$\endgroup\$
2
\$\begingroup\$

I check BJT's and diode's with the BJT and diode testing receptacles on my multimeter. It will report the forward drop for diodes, and the DC gain (HFE) for BJT's, both PNP and NPN. It is easy to insert TO-220 and TO-92 packaged transistors into the receptacles, which are spaced at 0.1". Each receptacle has four holes: EBCE, so that it can accomodate transistor pinouts that have the base in the middle, or the collector in the middle.

If you don't have this feature in your meter you should get one that has it! Or else you have to build a test jig, perhaps on a breadboard or more permanently.

There are simple tests you can do for obvious failures like opens and shorts. You should be able to make a simple circuit using a voltage source and a resistor, where the base diode (VBE) is used as a diode. Current should flow and the VBE drop should be a reasonable figure like 0.6-0.7V. Another test: the collector should not be shorted to the emitter.

If a transistor is in a circuit and cannot be removed, then one thing to do is to find another instance of exactly the same device and compare measurements. Failing that, if the transistor is part of a repeated circuit (e.g. channel in an audio device) or part of a complementary pair arrangement (where the voltages and currents in both halves should be similar), that can provide clues that the transistor isn't working. Failing that, what we can do is understand the circuit and use that to our advantage. We can directly measure the node voltages at the transistor's terminals when the circuit is powered up. Currents can be estimated by measuring voltage drops across the resistors (assuming the resistors are themselves not faulty and we know their values).

\$\endgroup\$
  • 2
    \$\begingroup\$ It gets harder when the transistor is soldered into a circuit and cannot easily be de-soldered. If you don't have a circuit schematic, then you won't know how much other components will get in the way of trying to measure the particular transistor you suspect. \$\endgroup\$ – Jon Watte Aug 19 '13 at 19:56
  • \$\begingroup\$ One meter that has the BJT receptacle and can measure hFE is the RadioShack 46-Range Digital Multimeter Catalog # 22-039. \$\endgroup\$ – tcrosley Aug 19 '13 at 19:59
0
\$\begingroup\$

I find the following procedure very straight forward and easy. On your multimeter, set the range to diode measurement. Put the red lead in the center pin and the black lead on any the side pins. If you get a reading and only one reading in just one of the sides you have an NPN transistor. Second case would be set the black lead from the multimeter at the center pin and try to get a reading from the other 2 pins touching them with the red lead. If you get a reading then we have a PNP transistor.

This method works for bipolar junction transistor since the principle is remembering that a BJT is simply a diode with an extra layer of either N or P material. So essentially you are testing a diode or a PN junction.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.