Measuring hfe of a transistor by multimeter ?

[Nhan95]

Some cheap multimeters have a transistor socket to measure hfe. But some new multimeter, the transistor socket is not available. So Is there any way to measure the hfe by any multimeters that do not have transistor socket on them ?

[tautech]

Yes I've noticed this too, and is the reason why I still have a Vichy 99.
I have found faulty transistors with it, matched pairs, and selected subsitutes to fall within a HFE range as listed in datasheets.
IMO exact measurement of gain is not so important, but comparison and confirmation of a working or failed device is when it becomes very useful.

[rdl]

Get one of those cheap component testers from eBay. It's one of the most useful items I have. price starts around $20 or less, though I'd recommend spending a bit more and get one that comes in some kind of case. Just search on eBay for transistor esr tester. There's a many pages long thread about them here somewhere.

edit: here it is, was only a few threads away

https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/

[bson]

I think mainly because getting a quality (usable, if you depend on it) measurement of some property of a transistor is a bit more involved.  The most obvious problem is that hfe changes with temperature - as the DUT warms up hfe will increase.  Temperature changes with Ic... so there's no single hfe to measure. 

The diode tester will tell you if a BJT good...  An NPN BJT can be checked as two diodes with a common anode (KAK).  A PNP is the opposite.

[Electro Fan]

another device worth considering (not as cost-effective as a built in DMM feature, but handy if you are into transistors):



They also have another model that does more via a PC interface.

[tom66]

One of the main reasons meters have stopped including the transistor tester is it prevents them getting any kind of  safety certification. If a transistor is attached, it will be connected to the multimeter's internal common ground on one pin, in most cases. If you then probe mains voltage, the transistor is floating at mains voltages. Not pleasant if you touch it.

[XFDDesign]

Beta (typ HFEdc) is a measure of how inferior a BJT is. As others have mentioned, it varies on temperature. It also varies on bias. Your beta term will be one thing at 1uA of bias current, something else at 10uA, and so on. Not all BJTs have the same beta values. For a practical, simple test, the diode check is sufficient. It's a quantity that is tossed into a datasheet to give a general idea.

[tec5c]

I don't mean to hijack this thread or anything but this got me asking myself, how do those cheaper DMMs measure hfe? The majority of them don't even have the capability to measure temperature so I don't imagine they would be doing an accurate (to the theoretical formulae) algorithm to get the term...

Any ideas anyone?

[SeanB]

They measure the current flow by biasing the transistor under test with a resistor, and measuring the current flow. Not too accurate, and will give an error with a leaky transistor as well showing an apparent high gain.

[ModemHead]

I don't mean to hijack this thread or anything but this got me asking myself, how do those cheaper DMMs measure hfe?

To add a little detail to Sean's answer, the hFE test on a a typical cheap DMM uses one socket and three basic 5% resistors.

The ICL7106 clone chips in a basic 2000-count DMM have a built-in voltage reference that regulates the analog ground at 2.8V below the positive battery terminal. This regulated voltage, along with a 220K base resistor, yields a base current of about 10uA for a silicon device with a 0.6V Vbe junction voltage.  A 10 ohm shunt to ground is used in the emitter leg for the NPN socket, or the collector leg for the PNP socket. The voltage across this shunt is fed directly to the 200.0mV full-scale input of the ADC, but with the decimal point turned off. So the reading on the meter represents the approximate ratio of collector current to base current at a 10uA base current level.

Technically for the NPN configuration, it would be wrong to measure the emitter current, but with only 10uA of base current, and a 10 ohm shunt, the error is only one count on the display.  This allows the 10 ohm shunt to remain connected on the "low" side, requiring nothing extra in the way of switch contacts.

edit: 3 resistors, not 2.

[Neilm]

One of the main reasons meters have stopped including the transistor tester is it prevents them getting any kind of  safety certification.

Technically no - they can get safety certification, the tester just needs to be isolated from the input jacks.

[smjcuk]

My dinky little Uni-T UT33A has no conduction path to the probe ports and the transistor tester pins by the looks.

Transistor hfe tester is very useful even if its not accurate. If you've partially buggered, toasted, zapped, vaped or otherwise mutilated a transistor then the hfe can drop to 1/5 of its colleages' value (transistor pain = less gain). Its also good for spotting naff ones shipped by the hundred from aliexpress.

I got through 5 of them building a current source the other week. Sort of 'oh that one is a bit hot', hfe check, hmmm 10 instead of ~200 trip to bin, get a new one out.

LTspice is probably cheaper so I don't recommend my design methodology

[paulvo]

It is very interesting to put a transistor in a simple test configuration using 3 resistors and an adjustable power supply, and measure simultaneously the collector current and the base current.
For a BC547, you can use a Rc=2K7, Rb=390K and Re=56ohms.
When you vary the base current from 10µA -> 20µA -> 30µA -> 40µA -> 50µA by increasing the supply voltage, note down each time the corresponding collector current.
Then for these various values, divide de collector current value by the base current value, to obtain the hFE
You will find that hFE will vary with the collector current (as was already stated by XFDDesign).

In addition, I measured the hFE using a TC1 component tester, and I can say the value shown by this tester was way out of line !
I would encourage you to do the same tests as well.

When measuring a power transistor like a 2N3055 you need lower values for the resistors: I took Rc=270ohms/5W, Rb=27K and emitter directly to GND.
Now adjust the power supply voltage so that the base current ranges say from 0,2mA upwards to 0,7mA or so.
Then note the corresponding collector currents, and calculate Ic/Ib=hFE
You may be surprised of the variation...

And if you have various transistors of the same type, it is also interesting to repeat the above test with a number of them.

I think you will find that hFE values are to be interpreted carefully.

[Jeff eelcr]

Fluke service manuals had a circuit you can build for units that measure seimens (conductance).
Its not perfect but does work.
Jeff

[Kleinstein]

The DMMs usualy measure Hfe as the large signal quantety and not hfe als the small signal parameter. The gain is not strictly a constant, but it can vary.

With modern circuits the actual gain is usually not a big deal, they are usually made to tolerate quite some variations, even if this means using more transistors. In addition the modern silicon transistors usually have a pretty stable gain. Measuring the gain was mainly an issue with germanium transistors, where the gain was more likely to change (e.g. with time or heat) and parts were scattering more.
Today the Hfe measurement is more like a function test and verification to have the right pin-out.

The DMMs removed the transistor test, because it is very difficult (nearly impossible) to meet the modern safty standards with this. So today the transistor tester is more like a separate instrument or a simple DIY circuit: send a small (e.g. 1 µA)  essentially fixed current to the base and measure the collector current. Power transistors anyway need a test at a higher current than small signal ones.

[Smokey]

Discussion/criticism of this thread here: https://www.eevblog.com/forum/beginners/how-to-really-measure-hfe-of-npn-transistor/?topicseen

Disregard if you are allergic to old threads