Tester for Germanium devices, is there one?

[Radio Wrangler]

It's a 30uA movement, though in a different housing.

I dismantled the movement and took the coil out. There was a fused strand quite visible on the top layer of the winding, right on one of the corners. I considered re-winding it, but hadn't any wire of the right gauge with thin enough varnish. It would have just been a matter of filing up a mandrel to sit over a rod to be held in my panavise and 30 minutes of careful winding. Anyway, I had an Ernest Turner meter in a newer style that does OK. Just need to make a scale.

David

[WME_bill]

Tester for germanium transistor: the Avo CT537.
Interesting David - RadioWrangler's comment on a 30ua meter in his CT537.
Avo must have changed the meter. I have a CT537 with a 50ua, 2.4k 1% meter. It is in a metal housing, with a small button knob reset on the lower right side of the panel.
My manual (1965) comments:
The military version CT537 uses an "Avo fuly sealed instrument with cut-out which breaks the It circuit. This cut out replaces the It fuse on the civilian model". The It fuse is 1 amp for the current test circuit.
The civilian version TT537 uses "a Taylor 50ua meter, with no cutout."
Both have a Int Rect 6F5 silicon diode (CV7384, 50v,6A) across the meter terminals as a limiter, + to +.

One reason I don't use mine often is that I have burnt out several transistors under test even at low currents, low voltage and with no heating. I rather suspect the switches marked break before make do not always operate as they are meant to. I would be interested to hear what experience RadioWrangler and others have with theirs.
wme_bill

[ionburn]

Originally Posted by David Simpson

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

Not sure how practical that is. As you will know, with Germanium transistors there is a far greater issue with temperature. My recent experience with my tester was with recently bought NOS Russian transistors. They came well within their spec so I would take them and the meter as Ok but my modern multimeter failed totally. If I were to go any further it would be with a homebuilt test jig, similar to that I used for valves (in fact I have used for point contact diodes and published results). Another point is that the old testers can cope (or at least imply the issue) with very leaky Germanium transistors. A 'standard', low leakage Germanium transistor would maybe be a poor example of the norm (but working) transistor.

My suggestion to anyone who cannot or does not wish to obtain an old tester would be to lookup the old circuits available and build one as they are fairly simple.

[avocollector]

PE had a really good design for a tester in the late 60's - basically a flip flop (2 x OC44) fed an ac signal into the base of the testing transistor and to another transistor whose output was used to 'null out' the test transistor's audio output via a pot which was calibrated for alpha(? going off memory = dynamic gain). A 5ma meter showed the input current/leakage via various switching arrangements. I built it an used it for years (still have it somewhere) but cannot recall the article or author's name. Awkward thing nowadays would be getting the original transistors to build it and the 95 ohm min speaker I used (presumably modern transistors could be used and a miniature radio output transformer). Mine never seemed to 'null out' completely but took it down to the lowest level.

[Radio Wrangler]

It might well be a 50uA movement, I was just running on memory from a few years back. It's definitely in a Taylor case.

No devices destroyed yet, but I usually have the bias backed off before moving any switches.

David

[David Simpson]

Just a quick thought - if anyone was considering a standardised spec for Ge transistors, perhaps something like a DTG1200 or 2400 as a power one, and a 2GT182(more modern OC71 eq.) for a small amp one.
After I'd started the st'd valve project, I was amazed at the software produced graphs that some participants produced. They put my hand-drawn efforts to shame. I'm sure these same computer experts could come up with something decent for Ge, or any type of transistors.

Regards, David

[David Simpson]

I'm sure Specmaster has by now quite a handful of advice. My final two-pennoth is some info from a 1965 Practical Electronics magazine. It takes up several pages. So if anyone wants them scanned, please let me know.

Regards, David

[David G4EBT]

The August 1965 issue (along with other issues of P.E. Mag from 1964, and most other UK radio mags too) is archived on the marvellous American Radio Hisory website. Here's the link to the Aug 1965 issue, with full constructional details of the transistor tester. Fittingly perhaps, it uses two long-since obsolete OC171 transistors:

https://www.americanradiohistory.com...cs-1965-08.pdf

[Argus25]

The notion of testing transistors in a simple tester is about or equally limited to to that of testing valves in valve testers. It probably explains why I do not own and have never owned a transistor tester of any kind.

There might be some benefit for screening out duds and rough comparisons. They are fraught with issues in "matching devices" unless there is a very complex test protocol at the frequencies of the intended circuit or at least over a range that covers that and curve tracing.

As is the case with valves, where a transconductance test alone is of limited but some utility value, an hfe on its own is too.

The best test you can perform is in the actual circuit the device will be used in, which includes the transistor's DC bias conditions in that circuit and its dynamic performance over the range of operating frequencies it will be used with. Signal generators & a good scope & meter are required.

Certainly for matching pairs of devices that are intended in a circuit where some sort of balance, either push pull amplification or signal nulling is required, dynamic testing is required for a close match.

Generally, testers can set the working conditions up which may well be far removed from the circuit you plan to use the device in.

Also, for RF germanium types the C-B feedback capacitance is very important and you would also have to measure that and the transition frequency. These values are a factor of 10 different for say an OC45 vs an AF125 and have a profound effect on the operation of a radio frequency circuit using them, but an ordinary transistor tester won't shed any light on this.

So for example, if you plan to use a germanium transistor in a 455kHz IF stage, set it up in one. Measure the signal gain, does it require neutralization, this can indicate a high range C-B feedback capacitance.

For an audio device set it up in a stage with the DC conditions it will be used in. Again apply a test signal output/scope measurements. Compare devices that way and it will not only tell you if the device is ok for the intended task, but if it is a reasonable specimen compared to its peers for its intended application and you can match them for some application if required.

['LIVEWIRE?']

I wouldn't use OC170/171 transistors in anything. They are essentially AF11x devices with different (pre Pro-Electron) type numbers, and are as prone to the 'tin-whisker' fault as others of their ilk. In a cupboard I still have a radio I built over 40 years ago, which uses 3 x OC170. If I wanted to make use of it again, I'd replace the OC170s with some of those Russian Ge transistors.

[Argus25]

Originally Posted by 'LIVEWIRE?'

I wouldn't use OC170/171 transistors in anything. They are essentially AF11x devices with different (pre Pro-Electron) type numbers, and are as prone to the 'tin-whisker' fault as others of their ilk. In a cupboard I still have a radio I built over 40 years ago, which uses 3 x OC170. If I wanted to make use of it again, I'd replace the OC170s with some of those Russian Ge transistors.

I agree with that, the AF11X and OC170/1 are completely untrustworthy now, but I would replace them with the AF178.