Testing germanium transistors....HFE keeps slowly rising

[armdnrdy]

Hey all,

I think that the title describes most of the issue so...I'll fill in the missing details.

I set up R.G.s HFE/leakage tester, and ran through seven different types of 2SA and 2SB germanium transistors.

Four of the types...no problems...very stable...expected results. The other three...no such luck.

I am inserting the transistors into the test circuit with tweezers, at a fairly stable room temperature, away from any heat source such as a light.

When first testing for leakage, the reading is stable. When testing for HFE...the reading never stabilizes...the reading keeps slowly rising.

This morning I inserted a transistor into the test circuit at 9:28 am with a leakage reading of .102 and an initial HFE reading of 1.40. (before the math)
I took various readings over the next hour. By 10:30, the HFE reading was up to 1.70!

One thing of note...I have never seen the HFE reading come back down....it just keeps rising.

Has anyone encountered this behavior with germanium transistors?

[blackieNYC]

Yes.
That's a really low gain but I have some of those.
Standardize the test and go with... 120 seconds? Or, until the rate of increases slows down. And don't breathe on it.  As far as trying to match the desired hfe of a circuit, I don't think anyone is gonna spec 1.7. Are you talking about 170?  140 to 170 - not surprised at that variation. A bit frustrating. Close enough.

Edit:right, RG's is the x100 figure. 140, 170 - time for a listen I would think. See what PJP has to say.

[armdnrdy]

Yes.
That's a really low gain but I have some of those.
Standardize the test and go with... 120 seconds? Or, until the rate of increases slows down. And don't breathe on it.  As far as trying to match the desired hfe of a circuit, I don't think anyone is gonna spec 1.7. Are you talking about 170?  140 to 170 - not surprised at that variation. A bit frustrating. Close enough.

Thanks for the reply.

Yes...the numbers I gave 1.4 and 1.7 were the HFE readings before the math.  140 - 170

I did record numbers after a short period...and the gain and leakage (very low) look very usable. Leakage as low as 31uA @ 132 HFE

I guess I would have to plop these into a fuzz circuit...keep it fired up for a while...on a warm day...to see what these things do.

[italianguy63]

Yeah-- I've seen this regularly too... I got to the point where I would just wait a couple minutes, take the reading and move on... 

MC

[thermionix]

Yeah-- I've seen this regularly too... I got to the point where I would just wait a couple minutes, take the reading and move on... 

MC

Same here.

[armdnrdy]

Thanks for the replies.

Good to know!
Everything else about these transistors seem to be winners.

When you've encountered this slow "HFE to the moon" anomaly...how did they sound?

[thermionix]

When you've encountered this slow "HFE to the moon" anomaly...how did they sound?

I'm not sure it is an anomaly.  Sound fine.  If you're seeing numbers you like, albeit unsteady, try 'em out.

[anotherjim]

Maybe you're just seeing the benign start of thermal runaway. Any volt drop between two points in a circuit in which some current is flowing has a heating effect. Heat increases turn on = increased current = increased heat. The heating effect is small, but happening.
Unless you're keeping them in a cooler and then testing in a warm room, heating up from ambient heat is not a likely cause.

If you have any of those clip-on finned heatsinks, it might dissipate internal heating better to keep it stable, though if the Q's are glass, probably not.

[armdnrdy]

Thanks for the reply Jim.

Two of the types in question have four legs...one attached to the metal case.

I tried connecting the case leg to "gnd" to see if it would help stabilize the slow "push forward"....but no change.

I'll try putting a heat sink on one of the transistors while under test to see if there's a difference.

[anotherjim]

4 legs are RF with a screened can connection. Still usable for AF.

Do test insulation between that screen and the semi-conductor. They are prone to grow metallic hairs inside due to the electrostatic influence of the grounded can, similar to the way lead-free solder can grow hairs. May be usable if you don't connect the screen. I have a hairy one in my FF, AF12x I forget exact number.