Thermionic Creep, A Peculiar Occurence

[SgtRock]

Greetings EEVBees:

--I have a Heathkit IM-1210 DMM. It is not anything special in terms of resolution, but the bright red LED display makes it a joy to read. I was cleaning it up and preparing to give it a fresh calibration so I can flog it on Fleabay, when I noticed it was functioning correctly, only occasionally, and was very sensitive to vibration. The display was lighting up, some of the time, and giving decent readings some of the time but mostly either no display or gibberish.

--At first I thought it must be the IC socket interface, so I removed all the ICs and cleaned them, and put a touch of DeOxit on them and reseated. No joy. I examined the soldering, no visible problem. I then proceeded to wiggle everything, only getting zapped a couple of times. Eureka! There are six test points where a red jumper wire can be pulled out of its socket. Please see the attached pictures. I deliberately left wire "C" out of its socket, so you could see it.

--Here is the weird part. I live in Florida, and I do not use air conditioning, so on some days there is a rather wide temperature swing. Wire C had managed to back itself out of the socket and came to rest just touching the socket, but not held by the socket, and the other five jumpers had done exactly the same thing, they had all backed out but remained just touching. So if you held your mouth right the durn thing would work just long enough to confuse you. Who wooda thunk it? Duh.

--I jammed them all home and Bobs your uncle. When I get er done, in the next hour or so. I will flog it, and soon I will have enough money in my PayPal account to pay for a slightly used Fluke 87V, which I will snipe for bottom dollar. I have been turning green watching DJ gloating with his paws all over his yellow beauty.

"I told you so, you damned fools."
Hugo Gernsback 1884 - 1967

Best Regards
Clear Ether

[Rerouter]

not actually that uncommon, a few long term bodged breadboards i have had to replace with hard wired systems had similar issues (not my breadboards) where the thermal cycles cause it to wiggle its way loose until its just touching and thus can expand and contract freely, part of that problem is from it having those self guiding wire holes, if they where tighter, the wire would flex towards the middle more and greatly slow down the process,

[free_electron]

Hearing /cooling cycles applied to a mechanical junction of disimilar metals will do that yes.. The expansion coefficients are different and you get a 'caterpillar' effect. The 'clamp' in the contact works one way. Wire goes in easy but is hard to pull out. Heat it up and it reverses... So the temp cycle actually slowly pushes the wire out in the way a caterpillar moves.

You can have this effect on power transistors as well. The lead wiggle themselves loose from the solder joints over time. To alleviate them the leads need a 'bend' in them. This bend translates the up/down motion to a left/right motion so there is no stress on the solderjoint anymore.

This is also the reason that transistor and ic pins are not simply stupid copper but a special alloy called
'alloy 42'' ( sometimes wrongly called kovar... Kovar is for glass sealed elements like vacuum tubes as kovars expansions contraction rate is equal to that of the glass of the tube. Invar is a particular type of kovar material in the sense that it has no expansion contraction . Nvar may be unsed on very expensive integrated circuits like side brazed packages or ceramic packages. For power ic's or transistors in to3 they use kovar as the pins go through a glass bead into the package.

These materials are iron-nicke or iron nickel cobalt alloys. The 42 defines the mix ratio. The ratio determines the expansion rate.