Tin whiskers and other packaging problems

[Pat Pending]

We trusted those devices, you could cook them till too hot to touch and they always worked.
They were batch loaded into a carousel tester and sorter, little bins collected the different types, the markings being printed on later.
Happy days, on £6 1/9p a week!

Thanks. This is the kind of information I was tilting at! You won't get a manufacturer to admit as much, but I'm sure getting the impression from the "factory floor" that different types were selected from single line. HOW do yoY test for maximum voltage and current ratings, without destruction? Regards. Andy.

[Refugee]

The actual voltage can be tested with a high value resistor and DMM. They work like a zena diode at there stated voltage plus a safety margin.
I would use a regulated HT power supply monitored with an accurate DMM and another DMM connected in series with the junction under test.
Subtract the readings and you have the actual reading ready to de-rate with your safety margin.

[Refugee]

Here are a few more variants on the E-line package.
Top left are BC171A in the classic grey package and 3793 (no data sheet).
Top right is a BC182L and 2N3704 in a very early package.
Bottom right is a 2SK19 also in a very early package.
Bottom right and center are three examples of Japanese E-line clones. 2SA677 in a greed package, 2SC634 in black and 2SC403C in a brown package.
Bottom right are 2SC458 and 2SC673 in a square package with a sloping top to denote the lead outs.

[pwdrive]

Those Japanese ones can suffer from some sort of growth/migration problem, a thin conductive film starts around the center lead and spreads to the others over time, I first discovered the problem back in the mid '70's whilst working for Sony (service department) the guys were having real problems tracking down some intermittant faults such as general noise, crackling, rustling noises etc or sometimes complete stage failure, after a fair bit of a determined attempt I traced the problem to that as described above.

Gently stroking or rocking the device or heating the leads usually made the symptoms worse or stopped the symptoms altogether, sometimes tapping the PC board would have the same effect which often led service engineers to think it was a traditional dry joint, which it wasn't, even though it manifested itself as such.

The problem was found in both Sony and Hitachi branded transistors of various package styles, and with the rectangular ones in all the usual colours eg: black, green and brown, the conductive film can be spotted with a sharp eye or a magnifying lens, it has a semi iridescent look to it.

Lawrence.

[Refugee]

Back in the late 1970s we used to find Sony transistors that behaved as if they had an internal dry joint.
Perhaps they might clean up.

[pwdrive]

The stuff present in the symptoms I described could be scraped off as far as I remember, more often than not though we just fitted a new transistor.

Lawrence.

[Pat Pending]

Refugee/Powerdrive. The to18 package (bc107-177) devices that were used in the pid application I mentioned in first post in this topic had a light grey bottom seal, and the "dendrites" (as I have been told its called) looked like rust stains on paint when viewed through with a magnifier, and it presented itself electrically as a very unstable form of leakage. I had originally put this down to corrosion in the bottom of the can or leads. I wonder if this phenomena is similar to what you observed. We were tight and would remove them, clean them up with a small fiberglass brush, dip them in urethane or black enamel, let them dry, and then refit them! If I come across any more examples then I will examine them more closely.

Regards Andy.

[Refugee]

We never had any idea that the problem could be fixed and anyway the workshop stores had them in stock.
Once they were out new ones went back in.

[Pat Pending]

Scottish heritage. And very often it was a mercy mission on my part when I was out and about, as we were never  allowed to carry spares. Officially it was because we didn't have room in the vehicles, but in reality it was so that the accounts could charge two call outs + traveling twice!

Regards. Andy