25 Jun
2003
25 Jun
'03
3:08 p.m.
Quothe Tony Duell, from writings of Fri, Jun 20, 2003
at 10:12:28PM +0100:
You think I _didn't_ keep the bits of the old fans? Of course I did :-).
And some of them have already proved useful in restoring other fans....
Not always... I mentioned those I2S machines in
another message. Some of
the fans in those had been powered for so long while stuck solid that the
windings had overheated and burnt out. And the bearings were in very poor
shape. Yes, I _could_ have rewound them, flash-tested them, and fitted
new bearings, but it made more sense to buy a new fan....
Yes... in a case like that, buying a new one makes much more sense,
and then save the old ones for spare parts, or, for later repair
in case the fans turn into unobtanium. the other day, I repaired a Westclox alarm clock...
the "real" type of
clock with hands and a dial, not one of those blasted digital display
Actually, IMNO the 'real' type of clock doesn't use electrical parts at
all. It's entirely mechanical. And yes, I do fiddle with those from time
to time too...
annoyances. A very simple repair, actually, which
involved replacing
the neon lamp which illuminates the dial. Granted, the clock cost
under US$10 IIRC, but I still think that this repair was more sensible
than replacing the clock.
Of course...
Interestingly, I learned something interesting about failing neon
lamps. The failing lamp would light in the daylight, but flicker
dimly in the dark. I never realized before that neon lamps were
light-sensitive. Can anyone explain this?
I first saw this on the PSU neons in my 11/45 system (see, it's on-topic
:-)). They would light steadily with the CPU pulled out on the slides,
but would fliker when it was pushed back in. I first thought it was a bad
connection and spent a long time checking.
Then I realised it was due to light falling on the neon bulb. Apparently,
light will cause some photoemision from the cathode in the neon, and
maybe will also cause a bit of photoionisation of the neon gas itself.
This causes the bulb to strike at a lower voltage.
This was a problem when ROMs consisted of an array of neon bulbs with a
photomultiplier pointed at them. It was common practice to illuminate all
the neons just before one was required to strike (and to gate out the
resulting pulse from the PM output) (See, for example 'Cold Cathode Tube
Circuit Design').
-tony