Why are old computers so vulnerable to temperature changes and moisture?

Question

Interesting discussion by someone thinking of procuring a PDP-11/34, though having difficulty finding suitable space for it: https://www.reddit.com/r/retrobattlestations/comments/dztvci/minimum_viable_system_for_pdp11/

A comment that caught my eye:

Keeping it in the garage will destroy it with the humidity and condensation as the seasons change. I've worked on machinery over the years that's been kept in garages and sheds, corrosion - and these are basic machines, it will be much less kind to electronics.

That seems surprisingly pessimistic! I don't have experience with mainframes or minicomputers, but I've been working with personal computers of various kinds for going on four decades, and I have never seen hardware rendered inoperable just by being in a location that is not climate controlled.

Is there something about the technology with which the older minicomputers were built, that makes them more susceptible to moisture damage?

Answer 1

If you watch enough YouTube videos of repairing old micros — and even ones that aren't that old (486s didn't come out that long ago, did they?) — you'll see a lot that have suffered some sort of damage due to storage over the years. Some damage, such as corrosion due to battery leakage or cracked circuit boards, is not due to climate. However some, such as oxidized contacts or corroded circuit board traces, is.

Old mainframes and minicomputers have a lot of points of failure compared to micros. Their construction often involves a lot of socketed components, plugs and connectors, and wire-wrapped connections. Each of these is an unprotected point where corrosion can occur. Repeated heating and cooling cycles can unseat socketed components and connectors, or weaken iffy solder joints. Humidity can corrode exposed wire, pins, and traces. Many old printed circuit boards do not have protective solder mask layers, which leaves them exposed as well. And there are a lot of boards, discrete components, and wired connections, which means a lot more pieces that can go bad.

Contrast this to the construction of most micros: more modern circuit boards, often (but not always, depending on age) with a protective solder mask layer. Many components are directly soldered to the board, and there are often only one or two boards. Almost all the connections are internal to the circuit boards. There are fewer discrete components, because there are more complex integrated circuits. There are far fewer exposed connections that can flake out, or components to be damaged.

Also, don't discount age as a factor. Depending on the computer, the quality of construction of some of the components may simply not have been as good because no one yet knew how to make them better, so ambient conditions may damage them too.

Saying "will destroy" may be something of an exaggeration depending on climate. A moderate or even hot, fairly dry climate will likely be reasonably kind to electronics if they're sheltered. A cool and damp one without too much temperature variation may not be a death sentence. In my opinion, a warning like that is still a good idea to someone who doesn't know the ins and outs of old electronics, since it will prevent disappointment and wasted money.

Answer 2

Is there something about the technology with which the older minicomputers were built, that makes them more susceptible to moisture damage?

No, it's about any human made - in fact even any pysical item at all.

I've been working with personal computers of various kinds for going on four decades, and I have never seen hardware rendered inoperable just by being in a location that is not climate controlled.

The issue isn't about being climate controlled or not, but what conditions it experiances over each day and years. Stuff in a shed in SoCal will stay much longer without damage than in a nice garage in the North-West.

That's why you find way more worthwhile project material when searching for an old car in Nevada vs. Oregon. And the same is true for old computers and anything else.

Bottom line: What people tell you about their experience when storing computers in garages or alike, depends mostly on their values for $GARAGE and $KLIMA which need to be mentioned to make the story worthwhile.

Answer 3

Someone can correct me if I am wrong, but I think larger components/contacts create larger shear forces due to the different rates of thermal expansion in different materials.

For instance, a tiny solder joint will have a much smaller difference between the expansion of the solder and of the plastic.

Answer 4

The specifications for most computers include (or used to) prohibition of 'condensing moisture' conditions. The 'condensing moisture' means dew. On a ship, below waterline, if there's no air conditioning, you'd violate that condition.

The reason is that the electronic parts of a computer have multiple different small metal parts. Adjacent dissimilar metals (copper traces, tin/lead/copper/silver/antomony solder, silver/gold/nickel/zinc plating, steel/brass screws, etc.) are subject to electrolytic corrosion if they get wet, and are SMALL parts, so it only takes a small bit of corrosion to ruin them.

Most working and living spaces don't get the moisture, or temperature swings, that produce that dew (unless it's on a windowpane) but in my garage, it certainly CAN happen. It wouldn't be good for the electronic parts of a computer, or even the mechanical ones.