Transistor testers and cheap DMM

[large_ghostman]

Whats the latest thinking on the HFE test on a cheap DMM? I ask because my 'decent meter' dosnt have a transistor test, and aprt from having true RMS and 3 digits after the decimal, it reads pretty much the same as my cheapo meter that does have a transistor test.

So how accurate is the hfe reading from a DMM or is it worth building a fixed current test circuit? Yes I could use the transistor tester I made but erm its bulky for what I want

 

[Nigel Goodwin]

Build yourself one of these:

**broken link removed**

or these:

**broken link removed**

I've built both, fun to make and absolutely essential tools for very little money.

But why do you want to measure Hfe anyway?.

 

[large_ghostman]

Build yourself one of these:

**broken link removed**

or these:

**broken link removed**

I've built both, fun to make and absolutely essential tools for very little money.

But why do you want to measure Hfe anyway?.

Cheers Nigel, yeah sorry for the sloppy missing capital H. I will have a look at building one. I want one because I have inherited a HUGE amount of old stock components many of which are obsolete but apparently still wanted. I have a huge range of transistors some of which I cant find info for, so I thought I would take a closer look at some and see what they were like. I have so many of some types I might sell some off to make room and I wont use most of them, so I thought I would test some and put them on sale.

Just looked at the links, i have a good LCR/ESR meter, I am looking for a way to characterize transistors mainly looking at the audio slant. I also want to see if I have any that can handle high frequency

 

[KeepItSimpleStupid]

Part of the problem is "Lead identification". The "component tester" guesses. I haven't finished building mine. i got side tracked.

I do have a $500 1980's USD transistor tester In any event it has:
1. signal trans
2) power trans
3) Signal FET
4) enhancement FET
5) Gain (lead ID)
6) Leakage (Icbo, and Idss)

The property of a transistor that it inverts is the basis of the LEAD ID. You manually move the selector until it starts beeping in 2 of the 3 positions. In 2 positions it makes noise. Then gain tells you what the lead ID really is. This is a good/bad and lead ID test.

You have Vbe which some DMM's can do.

Ft might be a bit harder. An amplifier and sweep the frequency or possibly measure rise time. Rise time is related to frequency response.

In any event, you could make a jig that does, lead ID using the component tester and whatever else it can find. Look at Vbe to see if you have a darlington, silicon or germanium transistor and do a rise time test using a scope with math.

e.g. Some sort of selector switch.
1. Component test
2. Set the leads based on component test, set NPN or PNP
3. Read Vbe
4. Select another position to measure leakage.
5. Put transistor in a circuit and have scope measure rise time - 10% to 90%

Now they are binned. The other stuff requires a curve tracer.

You might have to set the level of base drive. Hfe has a lot of dependencies especially for saturation, so measuring Hfe is going to be hard anyway.


Generally, the diode breakdown are non-destructive if you keep the currents low enough. So, you need a curve tracer. How high in voltage do you want to go.

 

[Nigel Goodwin]

Part of the problem is "Lead identification". The "component tester" guesses.

It doesn't 'guess' - it works it out

If nothing else it's a great way of identifying which lead is which.

 

[KeepItSimpleStupid]

Poor choice of words.

If nothing else it's a great way of identifying which lead is which.

and that was the point I was trying to get across.

 

[large_ghostman]

If I do all that I might as well rebuild the tracer I did, its clunky at the moment. I need to think about it. I just thought the test on the DMM was a nice simple way to do it

 

[tomizett]

Even if it's not got good absolute accuracy, it could still be useful for comparative measurements - that it, seeing if your batch of transistors have a wide spread of Hfe, finding matched pairs or seeing if a particular device is damaged. So the cheap DMM is always handy to have around. I keep one about because it's a quick way to find the pinning of a device if I can't remember it, or see if I've broken it by doing something daft.

The other matter is that, as you'll know, Hfe changes with collector current, so unless you know the test current the meter is using it will be hard to make comparison with published data or with the results from other instruments.

Those multi-component analysers are pretty cool though. Just my quick 2p worth!

 

[JimB]

My experience with transistor testers is as follows...

Most problems are found using the "two diode test", ie forward and reverse resistance tests on the Base-Emitter and Base-Collector junctions.

About 40 years ago I built a simple transistor tester which measure the E-C leakage current and the Hfe.
In the intervening time I must have used it all of 10 times.

I have a cheap (£5) DMM which has a transistor test feature, I have never used it.

Curve tracer? I have never used one, or had the urge to build or acquire one.
If I had one I would probably play with it for an afternoon and think "very nice, now what do I do with it?".

Just my random thoughts.

JimB

 

[cowboybob]

... Most problems are found using the "two diode test", ie forward and reverse resistance tests on the Base-Emitter and Base-Collector junctions. ...

Agreed. In-circuit, though, not so much.

But, for out-of-circuit testing for pin out and NPN/PNP typing, it's good enough for quick and dirty sorting,

Once that is done, test for Hfe as needed.