WHY do germanium transistors test so WEIRD ??

[rca2000]

So I have this rock-ola juke box amp to re-cap. I have done this....no real problems. BUT...when I went to check on the ckt. resistance on the negativbe B- supply rail (for the output stage-which is PNP GERMANIUM transistors) ..it checked LOW_-about 50 ohms or so across it. (n ONE direction ONLY though.) SO i checked the output transistors--delco PNP germaniums and they checked LEAKY...or so I THOUGHT. From E to C--they checked around 50 ohms or so ONE way...BUT the other way--they did NOT do this..the fluke meter just flashed "ol"... crazily. on diode test they do NOT show right at ALL.

B-to E and B to C checked similarly weird one way MUCH worse than the other. All FOUR transistors showed this. So I figured I had bad trannies. But this amp was supposed to WORK--it just needed re-capped to make it "whole".

THEN...I recalle that germanium transistors DO check different than silicons. I went and stole a couple of them out of a Zenith amp that is supposed to be working...and they checked about the SAME way.. Somewhere about 125 ohms or so one way E-to C and high the other way.

WHY do they show like this Normal si trannies do NOT . Is this normal for them ?

[Username1]

Yes, Ge transistors are not quite the clean crisp on-or-off like the Si transistors.
Ge's are better looked at in a circuit than taken out and tested.....

You can do something like this....

https://www.youtube.com/watch?v=dnvi2OFDLBk

.

[Electronic M]

Germaniums also have a different conduction threshold voltage, lower than the .7 of silicon devices....IIRC it is .5 or close.

[DavGoodlin]

Originally Posted by Electronic M

Germaniums also have a different conduction threshold voltage, lower than the .7 of silicon devices....IIRC it is .5 or close.

Using the diode test on most DMMs gives the value of this voltage drop. Most Si devices, when reading forward across a junction, show about .6 -.69 volts.
A 1959 car radio I worked on recently had a (good) Delco DS-501 transistor for the output. Anyway, it measured much less and inconsistently on my Fluke, like Tom says.