Accuracy of Far East transistor testers. - UK Vintage Radio Repair and Restoration Discussion Forum

WaveyDipole · 13th Nov 2016, 8:52 pm

I recently acquired a vintage transistor tester from a member (called a Transitester). It seems to work, although it requires a somewhat off 22.5v battery. Since someone has already put a couple of wires through the back to connect to a PSU, I set my PSU to 22.5v and gave it a try. It proved useful in confirming a suspicious leaky germanium transistor which the far east tester seemed to indicate was OK.

I decided to test a handful of other transistors and comparing with the far east gadget, and also to the hFE function on my Maplin Precision Gold (WG020) DMM to see whether I would get consistent readings, but the results rather puzzled me. For the purpose of the comparison, I tested 3 transistors, 2 silicon (one of each NPN and PNP) and one germanium, and got hFE readings as follows:

2N4403 (Si)
FE tester - 286
Transitester - 320 (approx)
Precision Gold - 265

2N5088 (Si)
FE tester - 470
Transitester - 410 (approx)
Precision Gold - 410

GT322 (Ge)
FE tester - 69
Transitester - 75 (approx)
Precision Gold - 73

Vf readings on the FE tester all seemed to be within the expected range.
On the Transiteter, the Iceo on the two Si transistors was about 1μA, whereas on the Ge device was about 14μA.

As can be seen, the readings for the 2N5088 andf GT322 seem to correlate between the Transitester and the Precision Gold, whereas for the 2N4403 the FE tester and the Precision Gold are closer, although still out by a small margin. The readings on the FE tester do seem to vary between repeated tests as well, whereas on the other two meters they seem to be consistently the same each time.

I do think each instrument has its own uses as the FE gadget shows the pinout, hFE and Vf, whereas the Transitester shows hFE and leakage current and also tests ordinary and Zener diodes (a 15v zener tested slightly out at approx 14.5v). The DMM just shows hFE so is only useful as a cross reference for hFE. It would seem that the measurements on the FE tester are at least questionable (although I'm not complaining for the price!).

I am wondering what accounts for the varying readings and which one can I/should I trust?

paulsherwin · 13th Nov 2016, 9:04 pm

Those hfe readings are all quite close. Hfe isn't constant and varies with voltage and current, so it's only an approximate reading.

MrBungle · 13th Nov 2016, 9:08 pm

I've had a couple of these. It is true to say that if a man has more than one watch he never knows the time. Both had completely different readings on the same devices and neither agreed with a calibrated HP LCR meter. So I stole the displays, one of which can be found in my frequency counter, the zif sockets and slung the rest in the bin.

hFe is hardly worth measuring accurately. I have only ever found go or no go tests with cheap DT830 style meters to be necessary, although I did build a test rig to check germanium transistor leakage automatically a few years ago (when I had a bag of 200 to go through)

paulsherwin · 13th Nov 2016, 9:10 pm

I've found mine to be reasonably accurate, though it's obviously not a precision instrument.

Radio Wrangler · 13th Nov 2016, 9:15 pm

As Paul says, Hfe varies, so all it takes is a difference in the bias conditions between testers and they all could be right.

Don't get too fussed about Hfe values. New devices have a 2:1 or wider spread of values within their spec limits. Some have no upper limit. In use they vary a lot more whn you take into account temperature, bias Vce and bias Ic.

The real difficulty is in having designs which will tolerate the specified spreads of the devices they use.

A good design should be able to function properly with all of its semiconductors anywhere within their specifications. There've been recent threads about magazine articles and how iffy some of the designs are. Unfortunately, the majority of punters count the components and think that lower numbers indicate simplicity. It isn't so simple! And sensitivity to device parameters and selection of devices to find one which works is a common result.

David

MrBungle · 13th Nov 2016, 9:19 pm

Indeed. This is one reason integrated circuits, particularly op amps are rather popular.

TonyDuell · 13th Nov 2016, 9:53 pm

I find that if you really want to know the characteristics of a transistor, you need a curve tracer. But that is a large, complex, instrument that takes a lot of time to use properly. Not useful for finding a dead transistor in a radio...

The Hfe range on my (Maplin) DMM has one use. Identifying the pinout of a transistor. After using the diode test range to identify the base lead and the polarity (NPN or PNP) of a transistor, I use the transistor test range. Base I know, so I try the other leads both ways round in the collector and emitter sockets. Normally the gain is much higher with one connection than the other. And that is the right one.

MrBungle · 14th Nov 2016, 12:39 am

Similar here. I use one of these. The price is ridiculously low!

http://cpc.farnell.com/duratool/d030...tal/dp/IN07220

David Simpson · 14th Nov 2016, 12:09 pm

Hello Wavey Dipole,
My advice would be to acquire a couple of "Standardized" transistors from someone near you who has an AVO CT446 Transistor Analyser. The CT446 uses the tabulated data seen in the 400+page AVO Transistor Data Manual. Similar data can be seen in a "TIT's" Data Book.
The "Far Eastern" testers I featured in my thread "Special Project.Quiz" back on the 20th Sept have a great advantage over a CT446, in that they immediately identify the type & lead configuration of the semi-conductor under test. Especially if they have lost their identification lettering/numbering. However, just using wee PP3 9V batteries, they are seriously limited as to Vc & Ic conditions for testing. A CT446 can test using up to 150V & 1A.
If you have a heap of NOS or 2nd hand transistors to test, then a cheapo one is ideal. But they don't like you testing 'in situ', and testing in live circuitry is a definite no-no. But, in fact their readouts for hFE or beta often compare reasonably with a CT446's for common user transistors.

Regards, David

WaveyDipole · 14th Nov 2016, 3:33 pm

Thank you for all the replies. I was just interested in understanding the differences in readings. The fact that different test voltages or methodologies within various instruments might lead to different readings is probably a sufficient answer to my question.

The Transitestor is interesting since it clearly shows the leakage. I had two germanium transistors which I removed from a Roberts R200. I knew something was wrong since they were getting very hot, but the FE tester showed that both were OK. With the Transitestor the leakage was all too clear. Initially showing higher than expected leakage, the meter needle gradually moved further and further over to the right as the leakage increased, presumably as the transistor junction heated up.

Therefore, I'm not sure that the FE tester can always be relied upon to provide the correct result, although a has been said, it is generally very good at identifying the pinout of devices and providing and indication of the device parameters. Of course, this may be an odd result on account the item of interest in this case being a germanium transistor.

DaveVoorhis · 14th Nov 2016, 4:00 pm

I think most if not all of the far eastern testers are based on this project: https://www.mikrocontroller.net/arti..._.28English.29

I don't know if it's designed to recognise germanium transistors, and as an experimenter project, it may have many limitations that the manufacturers don't mention (and probably don't know about). I've been meaning to get one of them to try it out.

paulsherwin · 14th Nov 2016, 5:25 pm

Most cheap modern transistor testers don't do a very good job with Ge devices. I do use the Chinese analyzer with them (and the DMM hfe tester too), but only as a quick go / no go test. I find a mid 60s Eagle tester (basically a meter, a battery and some switches and resistors) gives a more meaningful result with Ge.

The Chinese tester has trouble recognising Ge diodes at all.

13th Nov 2016, 8:52 pm	#1
WaveyDipole Nonode Join Date: Mar 2013 Location: Leicester, UK Posts: 2,535	Accuracy of Far East transistor testers. I recently acquired a vintage transistor tester from a member (called a Transitester). It seems to work, although it requires a somewhat off 22.5v battery. Since someone has already put a couple of wires through the back to connect to a PSU, I set my PSU to 22.5v and gave it a try. It proved useful in confirming a suspicious leaky germanium transistor which the far east tester seemed to indicate was OK. I decided to test a handful of other transistors and comparing with the far east gadget, and also to the hFE function on my Maplin Precision Gold (WG020) DMM to see whether I would get consistent readings, but the results rather puzzled me. For the purpose of the comparison, I tested 3 transistors, 2 silicon (one of each NPN and PNP) and one germanium, and got hFE readings as follows: 2N4403 (Si) FE tester - 286 Transitester - 320 (approx) Precision Gold - 265 2N5088 (Si) FE tester - 470 Transitester - 410 (approx) Precision Gold - 410 GT322 (Ge) FE tester - 69 Transitester - 75 (approx) Precision Gold - 73 Vf readings on the FE tester all seemed to be within the expected range. On the Transiteter, the Iceo on the two Si transistors was about 1μA, whereas on the Ge device was about 14μA. As can be seen, the readings for the 2N5088 andf GT322 seem to correlate between the Transitester and the Precision Gold, whereas for the 2N4403 the FE tester and the Precision Gold are closer, although still out by a small margin. The readings on the FE tester do seem to vary between repeated tests as well, whereas on the other two meters they seem to be consistently the same each time. I do think each instrument has its own uses as the FE gadget shows the pinout, hFE and Vf, whereas the Transitester shows hFE and leakage current and also tests ordinary and Zener diodes (a 15v zener tested slightly out at approx 14.5v). The DMM just shows hFE so is only useful as a cross reference for hFE. It would seem that the measurements on the FE tester are at least questionable (although I'm not complaining for the price!). I am wondering what accounts for the varying readings and which one can I/should I trust? Last edited by WaveyDipole; 13th Nov 2016 at 8:58 pm.

13th Nov 2016, 9:04 pm	#2
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,966	Re: Accuracy of far east transistor testers Those hfe readings are all quite close. Hfe isn't constant and varies with voltage and current, so it's only an approximate reading.

13th Nov 2016, 9:08 pm	#3
MrBungle Dekatron Join Date: Jun 2016 Location: London, UK. Posts: 3,687	Re: Accuracy of far east transistor testers I've had a couple of these. It is true to say that if a man has more than one watch he never knows the time. Both had completely different readings on the same devices and neither agreed with a calibrated HP LCR meter. So I stole the displays, one of which can be found in my frequency counter, the zif sockets and slung the rest in the bin. hFe is hardly worth measuring accurately. I have only ever found go or no go tests with cheap DT830 style meters to be necessary, although I did build a test rig to check germanium transistor leakage automatically a few years ago (when I had a bag of 200 to go through)

13th Nov 2016, 9:10 pm	#4
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,966	Re: Accuracy of far east transistor testers I've found mine to be reasonably accurate, though it's obviously not a precision instrument.

13th Nov 2016, 9:15 pm	#5
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,901	Re: Accuracy of far east transistor testers As Paul says, Hfe varies, so all it takes is a difference in the bias conditions between testers and they all could be right. Don't get too fussed about Hfe values. New devices have a 2:1 or wider spread of values within their spec limits. Some have no upper limit. In use they vary a lot more whn you take into account temperature, bias Vce and bias Ic. The real difficulty is in having designs which will tolerate the specified spreads of the devices they use. A good design should be able to function properly with all of its semiconductors anywhere within their specifications. There've been recent threads about magazine articles and how iffy some of the designs are. Unfortunately, the majority of punters count the components and think that lower numbers indicate simplicity. It isn't so simple! And sensitivity to device parameters and selection of devices to find one which works is a common result. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

13th Nov 2016, 9:19 pm	#6
MrBungle Dekatron Join Date: Jun 2016 Location: London, UK. Posts: 3,687	Re: Accuracy of far east transistor testers Indeed. This is one reason integrated circuits, particularly op amps are rather popular.

13th Nov 2016, 9:53 pm	#7
TonyDuell Dekatron Join Date: Jun 2015 Location: Biggin Hill, London, UK. Posts: 5,215	Re: Accuracy of Far East transistor testers. I find that if you really want to know the characteristics of a transistor, you need a curve tracer. But that is a large, complex, instrument that takes a lot of time to use properly. Not useful for finding a dead transistor in a radio... The Hfe range on my (Maplin) DMM has one use. Identifying the pinout of a transistor. After using the diode test range to identify the base lead and the polarity (NPN or PNP) of a transistor, I use the transistor test range. Base I know, so I try the other leads both ways round in the collector and emitter sockets. Normally the gain is much higher with one connection than the other. And that is the right one.

14th Nov 2016, 12:39 am	#8
MrBungle Dekatron Join Date: Jun 2016 Location: London, UK. Posts: 3,687	Re: Accuracy of Far East transistor testers. Similar here. I use one of these. The price is ridiculously low! http://cpc.farnell.com/duratool/d030...tal/dp/IN07220

14th Nov 2016, 12:09 pm	#9
David Simpson Nonode Join Date: Mar 2008 Location: Aberdeen, UK. Posts: 2,857	Re: Accuracy of Far East transistor testers. Hello Wavey Dipole, My advice would be to acquire a couple of "Standardized" transistors from someone near you who has an AVO CT446 Transistor Analyser. The CT446 uses the tabulated data seen in the 400+page AVO Transistor Data Manual. Similar data can be seen in a "TIT's" Data Book. The "Far Eastern" testers I featured in my thread "Special Project.Quiz" back on the 20th Sept have a great advantage over a CT446, in that they immediately identify the type & lead configuration of the semi-conductor under test. Especially if they have lost their identification lettering/numbering. However, just using wee PP3 9V batteries, they are seriously limited as to Vc & Ic conditions for testing. A CT446 can test using up to 150V & 1A. If you have a heap of NOS or 2nd hand transistors to test, then a cheapo one is ideal. But they don't like you testing 'in situ', and testing in live circuitry is a definite no-no. But, in fact their readouts for hFE or beta often compare reasonably with a CT446's for common user transistors. Regards, David

14th Nov 2016, 3:33 pm	#10
WaveyDipole Nonode Join Date: Mar 2013 Location: Leicester, UK Posts: 2,535	Re: Accuracy of Far East transistor testers. Thank you for all the replies. I was just interested in understanding the differences in readings. The fact that different test voltages or methodologies within various instruments might lead to different readings is probably a sufficient answer to my question. The Transitestor is interesting since it clearly shows the leakage. I had two germanium transistors which I removed from a Roberts R200. I knew something was wrong since they were getting very hot, but the FE tester showed that both were OK. With the Transitestor the leakage was all too clear. Initially showing higher than expected leakage, the meter needle gradually moved further and further over to the right as the leakage increased, presumably as the transistor junction heated up. Therefore, I'm not sure that the FE tester can always be relied upon to provide the correct result, although a has been said, it is generally very good at identifying the pinout of devices and providing and indication of the device parameters. Of course, this may be an odd result on account the item of interest in this case being a germanium transistor.

14th Nov 2016, 4:00 pm	#11
DaveVoorhis Tetrode Join Date: Mar 2016 Location: Derby, Derbyshire, UK. Posts: 50	Re: Accuracy of Far East transistor testers. I think most if not all of the far eastern testers are based on this project: https://www.mikrocontroller.net/arti..._.28English.29 I don't know if it's designed to recognise germanium transistors, and as an experimenter project, it may have many limitations that the manufacturers don't mention (and probably don't know about). I've been meaning to get one of them to try it out.

14th Nov 2016, 5:25 pm	#12
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,966	Re: Accuracy of Far East transistor testers. Most cheap modern transistor testers don't do a very good job with Ge devices. I do use the Chinese analyzer with them (and the DMM hfe tester too), but only as a quick go / no go test. I find a mid 60s Eagle tester (basically a meter, a battery and some switches and resistors) gives a more meaningful result with Ge. The Chinese tester has trouble recognising Ge diodes at all.