NOS leaky germaniums

[McMurdo]

I've just received a couple of NOS AC176 npn germaniums as outputs on a 1960's transistor radio; this sort of high-tech is not my usual forte. (edit: o.ps are npn / pnp complimentary).

The original fault was tin whiskers in the NPN output transistor, temporarily cured with a little tap for an hour or so.

I have plenty of AC128's so I thought I'd match one with one of the new ones with my chinese component tester.

Both the new AC127's fail as 'unknown' in the tester. Not unusual; its not foolproof. The AC128's all come back with sensible readings.

On an avo 7 the new transistors measure leaky C-E both ways; about 500R one way and 25R the other. I know Ge's leak but this seems excessive.

As such I can't match a pair.

Putting one in the radio and it works but is very distorted and I've had to turn the bias pot down to prevent the stage taking excessive current.

Opinions, please!!

[paulsherwin]

Tin whiskers?

You could just convert to a BC327/337 pair or something similar. You just need to adjust the bias components.

The Chinese analysers do get confused by Ge components.

[Chris55000]

Hi!

I wouldn't advise using a 1000ohm/V AVO 7 to test a germanium device - there were many warnings about this in 1960s and 1970s Servicing articles, because AVO 7's need a larger ohm-meter current and it can be enough to overheat and damage the devices!

Chinese wonder-testers aren't a good bet as you've discovered as well!

My recommendation is to download R.E.F. Street's book "Practical Television Circuits" off the Internet Archive - there's a good simple design for a Ge-friendly Transistor Tester in it!

In the meantime, the best way of testing your AC176 transistors is to fit one, together with it's complementary AC128 in your set, connect a mA meter (AVO OK for this!) in one battery lead and monitor the current for 15 minutes carefully on zero signal - volume at minimum.

If the current starts to rise within a short time of applying power, the O/P transistors are potentially suspect. If the reading remains stable within 15 min of applying power, your transistors are sound!

Chris Williams

PS!

Just re-read your post - it seems conclusive!

You might be able to get a good pair of AC187/01 and AC188/01, which are newer devices, alternatively look to see what there is on eBay in the ex-Soviet transistor line - the Soviets made excellent transistors that don't "Tin-whisker", are long-lasting and don't leak if run within their voltage/current ratings!

[G4_Pete]

Hi,
The other comment I would make is that leakage indication at least on my peak atlas transistor tester is VERY temperature sensitive. If I put an OC71 in the fridge it tests perfect, with uA leakage indicated, but if i warm it with my fingers it comes up into the 1 -2 mA leakage range. This is mentioned in the user manual FAQ's.
Pete

[Philips210]

Originally Posted by Chris55000

My recommendation is to download R.E.F. Street's book "Practical Television Circuits" off the Internet Archive - there's a good simple design for a Ge-friendly Transistor Tester in it!

Hi Chris.

I had a look through that book which contains some very interesting projects but unless I've missed something, I couldn't see a transistor tester listed.

Regards
Symon.

[Philips210]

Hi

Attached is a list of test gear projects featured in R. E. F Street's book.
As I say I couldn't find a transistor tester listed. Perhaps Chris55000 can enlighten us.

Regards
Symon.

[Chris55000]

Hi!

My apologies Gents - I must have mixed up my Transistor Tester with a EHT Voltmeter!

The October 1964 Issue of Practical Wireless has a nice simple circuit, suitable for measuring I'co and hFE 0-100 & 100-250 on pages 54-56, etc., and Germanium-friendly - get it from the American Radio History site!

(You'll need a 0-1mA meter movement to make it tho!)

Sometimes if you're testing/reusing old components, you have to use test-gear circuits from the same era - as the OP found, and I have also found, Chinese Wonder-Testers fall down with old semiconductors!

Chris Williams

[Philips210]

Hi

Thanks for getting back.
I agree the Chinese testers have problems with germanium devices and you're better off with an old analogue type tester from the period. My simple gain tester was from a RC design which allows you to check both germanium and silicon transistors but that doesn't check leakage current. I enjoy building test gear circuits and thanks for the link to the Oct 1964 PW circuit. The most comprehensive transistor tester I built was from a PW design (Jul 1976) and I also included the follow up circuit for FETs as well in the same enclosure but I haven't tested this completed project thus far. Attached a few pics.

Regards
Symon

[McMurdo]

I do have an old eagle transistor tester, one of those that uses the transistor as an oscillator and then you read the gain off a calibrated dial when the neon starts to light.

I'd already measured the idle current of the radio with its original whiskery transistors at 16mA (when it was behaving).

AFAIK the equivalent russian output transistors have their case connected to one leg (base?) which might make the heatsink mounting a little awkward. The thermistor also has to live on the heatsink.

I've ordered another AC176 from Langrex this time who have always been ok. They can also match them for you if you ask.

[paulsherwin]

See this thread for a silicon conversion: https://www.vintage-radio.net/forum/...ad.php?t=54342

[McMurdo]

The alternative AC176 arrived, one of the slimmer packages but the heatsink clip tightened around it no problem.

This one tested on the chinese tester with a gain of 56, not far off the original 'green spot' matched pair at 44. So in it went and it's working fine