In comp.periphs.printers John Garza said:
All the responses seem logical though - time dries out electrolytic
caps, and some slight corrosion or dust in a switch could prevent
conduction, etc.
Sometimes it's really simple. This spring, my circa-1993 Laserjet 4
refused to power on, right as I wanted to print out my tax return. A
little digging revealed that a wad of dust had gotten into the line
power switch. Removing the dust made it go and it's been working ever
since.
Maybe I need to hermetically seal all my machines.
Or, get a zillionaire interested in them, and he will pay a staff to
maintain them.
http://www.livingcomputermuseum.org/
Maybe a big airtight antistatic bag filled with nitrogen?
Protip: You used to be able to get free nitrogen from the phone company,
but they are letting the copper infrastructure rot on purpose, so they
don't care about leaks anymore. Go to a shopping mall or similar large
parking lot and get free nitrogen from car tires that have green valve
caps on them. (If you're feeling really polite, replace it with a 78%
nitrogen/21% oxygen/1% otherstuff mix.)
I'm starting to feel like the repairmen in those old stories you hear
about the huge ancient computers. As soon as someone replaced a
vacuum tube, another would burn out some place else!
The story I heard was that they ran the tube filaments at nominal
voltage, or maybe a little under nominal. On Tube Day, they ran the
filament voltage up to 10% or so above nominal, and maybe flicked it on
and off a few times, and replaced all the ones that burned out.
Matt Roberds