Tester for Germanium devices, is there one?

[Specmaster]

I have 2 of these Chinese multi purpose transistor/L/C/R testers that work pretty well for Silicon devices but gives unpredictable results on Germanium devices. Does anyone know of a device that will test Germanium devices and give reasonable results? I'm trying to select some pretty well matched sets of transistors for use in push pull amps for radios.

[Paul Stenning]

I have a basic analogue transistor tester purchased from Maplin decades ago which tests GE and silicon transistors fine. Unfortunately it is in storage and I can't remember the brand name.

It has a knob for hfe which you increase to the point just before the LED goes off to get a reading. Probably not that precise for the actual figure, but fine for comparison and matching. Single switch for NPN or PNP, a button for power and three croc-clips on short leads and that's it. Uses a single 9V battery.

If something like this is available second hand it would do the job nicely.

[Radio Wrangler]

AVO transistor testers turn up fairly often and aren't expensive. They do Ge and Si quite properly as well as diodes, zeners etc. Can do 1000v for breakdown tests, too.

David

[David Simpson]

The AVO CT446 Transistor Analyser is the apogee of accurate testing of vintage transistors. Both PNP & NPN. Particularly Germanium ones from the 1950' & 60's. The original MK1 came out in the late 50's & used a battery compartment containing a number of 1.5V cells which now are hard to come by. The later MK2 was mains operated.
Best bet is to ask around the Forum for someone near you who has one & knows how to use it. They would need an AVO Tr. Data Manual &/or a TITS.

Regards, David

[John_BS]

Peak DCA75 is normally reliable; I supplied some Ge transistors to them when an older release of their tester gave false readings: they then refined the test algorithms and all came good. There are some Ge transistors which are so poor (in terms of xs leakage etc) that they will confuse almost all automated testers. You can build a simple analogue test meter yourself to deal with these devices.



John

[paulsherwin]

Originally Posted by Paul Stenning

I have a basic analogue transistor tester purchased from Maplin decades ago which tests GE and silicon transistors fine. Unfortunately it is in storage and I can't remember the brand name.

It has a knob for hfe which you increase to the point just before the LED goes off to get a reading. Probably not that precise for the actual figure, but fine for comparison and matching. Single switch for NPN or PNP, a button for power and three croc-clips on short leads and that's it. Uses a single 9V battery.

If something like this is available second hand it would do the job nicely.

You can build one using a couple of resistors, switches and a transistor socket. The transistor socket can be knocked up from an IC socket. An external multimeter can be used to measure the transistor current if you don't have a spare meter to hand.

[AC/HL]

Originally Posted by Paul Stenning

I have a basic analogue transistor tester purchased from Maplin decades ago which tests GE and silicon transistors fine. Unfortunately it is in storage and I can't remember the brand name.

That sounds like the one I have, Archer (Radio Shack).

[Refugee]

I still have a home made one.
It has 5 collector current settings and you have to calculate the hfe.
It dates from the late 1970s.

[thermionic]

Originally Posted by AC/HL

That sounds like the one I have, Archer (Radio Shack).

I had one like this, purchased from Maplin many moons ago. I seem to remember it being called a ‘ Seesure’ transistor tester.

Not sure who Seesure were, but I imagined they were a small outfit.


Cheers. Simon.

[Craig Sawyers]

I have a Peak transistor tester DCA55. When you first put the clips on the transistor, your fingers heat a Ge transistor up from body heat. So it takes a minute or so for it to get back to ambient and give stable readings of leakage and beta.

Craig

[Electronpusher0]

I use an old TTC Model C3022 transistor checker I bought from my local car boot recently, still in its box, looks like new. Photos show it testing an AC125 for gain.
I also attach the schematic from the instruction booklet

Peter

[paulsherwin]

I have a possibly identical tester branded Eagle.

[David Simpson]

Aye Specmaster, to get back to your original post regarding matching Ge transistors - - back in the 60's & 70's I remember several of the electronics/radio hobbyist magazines often featured "Curve Tracer" project circuits, for use with an oscilloscope. Maybe some Forum folk still have copies lying around. Also, in the true spirit of the Forum - why not build your own simple tester, like Refugee & Paul advise. Then perhaps someone near you with an AVO CT446 could do a quick calibration check with one or two Ge transistors of known characteristics.
Me - I'm a few hundred miles north of most of the Forum community, but if you become desperate, post me a wee handful of Ge's & I'll test them on my CT446 MK2, plus a comparison test on a chinko DCA55.
I'm sure if you get your arris to the nearest BVWS Swapmeet, you might meet someone who shares your semi-conductor interests.

Regards, David

[WME_bill]

Tester for Germanium devices.
I was put off by the GE in the heading,as I read it as implying General Electric (of US), who made a few germanium transistors, which can be tested in the same way as any other germanium transistor. (And they were the last firm to make tunnel diodes, for those who are searching around).

You chaps all have your commercial testers.

I still have my home built one, based upon a design by Waddington in Wireless World June 1970.
For the sort of activity of many on this Forum in repairing faulty equipment, it has the huge advantage that it indicates how a transistor has failed. As well as the usual leakage, hfe, Si/Ge and npn or pnp. It will also indicate quickly the connections for e,b,c, as an alternative to using an ohmmeter and messing about with changing leads around.
I have used it to match pairs of transistors. Although only at two different currents, it has proved quite close enough.
I attach the article.
I also have an Avo CT537 as a superior version,testing at all currents and volts,as mentioned by Radio Wrangler, but rarely use it. Photo attached for those not familiar with it. Produced later than the early CT446 mentioned by David Simpson and other styles which Avo tried. I think I saw a CT537 on the usual site recently, though the price seemed more than I expected. My Waddington design could be built for much less than that.
wme_bill

[David Simpson]

Hey Bill, that's spot-on. Wireless World June 1970. Ideal for Specmaster. Right enough, The CT537 would be a fine improvement on the 446. Bit after my time.
June '70 - I'd just returned from 3 years in Cyprus, and was on disembarkation leave prior to heading off to Cosford as an instructor. Halcyon days.

Regards, David

[usradcoll1]

Originally Posted by Paul Stenning

I have a basic analogue transistor tester purchased from Maplin decades ago which tests GE and silicon transistors fine. Unfortunately it is in storage and I can't remember the brand name.

It has a knob for hfe which you increase to the point just before the LED goes off to get a reading. Probably not that precise for the actual figure, but fine for comparison and matching. Single switch for NPN or PNP, a button for power and three croc-clips on short leads and that's it. Uses a single 9V battery.

If something like this is available second hand it would do the job nicely.

It sounds similar to the one I have! Mine uses a neon lamp and some form of a power oscillator. It's a Japanese import, that I must have from the late 60's. Runs on a single "C" cell.
Best regards! Dave USradcoll1.

[ionburn]

I have a Russian multimeter that I bought in the 1970's. It has a transistor tester which I have used recently; the first time in ages as my modern meter was useless with Germanium. I think it is reletively accurate for normal purposes. The meter itself was one of my better buys. A bit agricultural with its steel case but I have found it excellent, even if the ranges are a bit strange. Probably a quarter the price of an Avo ((second hand prices perhaps) and (comparing my Avo which has been little used, of similar vintage which has rotted case stitching and a dicky reset button) better robustness and reliability. Cannot have everything though as it just doesn't have a 10A range.

[Argus25]

Originally Posted by Specmaster

I have 2 of these Chinese multi purpose transistor/L/C/R testers that work pretty well for Silicon devices but gives unpredictable results on Germanium devices. Does anyone know of a device that will test Germanium devices and give reasonable results? I'm trying to select some pretty well matched sets of transistors for use in push pull amps for radios.

For your application there is a much better approach. The simple transistor tester will not give you enough data to match the transistors well for the application. A curve tracer would be better, but the better way in practice to do it is to set up your push-pull output stage with sockets.

Drive your amplifier with a sine wave and have small low value emitter resistors say 1.5 ohms or similar for each transistor. Adjust the bias just out of cross over distortion.

When the transistor's are closely matched:

With no drive signal the voltage drop across each emitter resistor will be near identical indicating closely matched emitter currents. Connect a resistor dummy load on the transformer output, instead of a speaker and have the scope across that.

With drive signal the waveform will be symmetrical above and below zero and clipping will be symmetrical with overdrive on each half of the sine wave. Then you can be sure your transistors have closely matched static and dynamic characteristics, just an hfe tester won't get this result.

Don't forget if you are testing RF germanium types, their junctions are very fragile and can be damaged sometimes with just a few mA base currents, Ideally don't test them on vintage multimeters or some modern testers designed for silicon types.

[Radio Wrangler]

I bought a CT537 for a reasonable price at a radio rally. Got it home and found the meter coil open circuit.

Proves it was a real AVO!

Found another meter of the right sensitivity in the attic, but I still need to make a new scale for it.

I recall posting a macro photo of the actual bit of blitzed wire on the forum.

Anyway, beware of the AVO transistor testers. They have the same competency of priotection as their valve testing brethren.

David

[David Simpson]

Right enough, as David says, AVO's 30uA fsd meters, used in MK3 & MK4 VCM's & the CT160, & the CT446 Tr.Analyser, were/are a weak point. Does the CT537 have the same meter David ?
If Specmaster is considering going into matching Ge or Si transistors in a big way, then a curve tracer or Argus25's advice would be a great idea.
The problems with Ge transistors, mostly jolly old now, have been well discussed & documented on this Forum. However, if any member has one or two fine & reliable examples, along with their curves &/or decent test results - perhaps they might consider taking them along to major swapmeets etc. so that folk with super dooper modern chinko wee digital testers can get them checked out Ge-wise.
I haven't the time, but I've often thought that after the reasonable success of my free 6AQ5/6AU5 graphed St'd valve project a couple of years back - that it would be a boon for Forum members if some younger keen enterprising semi-conductor buff were to distribute some "Standardised Transistors" out to the membership.

Regards, David