17 Sep
2004
17 Sep
'04
9:36 a.m.
On Thu, 2004-09-16 at 12:43, John Foust wrote:
With platters, there are clean-room issues
Not this thing! Or should I say, a clean, room, OK, but not a CLEANROOM.
The memory is not sealed. The heads are on little lucite blocks held to
the top of the memory with two screws. No gasket, even. The heads
'crash' on the nickel-plated-steel platter, and lift when up to speed. I
assume dust just gets flung to one side centrifugally.
The San Diego 'Computer Museum of America' has an LGP-21, it was
apparently running when donated, and the disk surface on that thing
looks AWFUL, all scratched-looking, but we're talking 60 - 80 bits per
inch here, so as long as you don't score head windings in half it
probably just makes a wee noise and moves bits.
as well as very
difficult alignnment issues - imagine how difficult it would
be to re-align a read head with the concentric bands on a
platter that may have been slightly misaligned from its
original hub, especially at contemporary densities.
Contemporary densities being readable with the naked eye... :-) I assume
you are otherwise correct, the track is likely very narrow, and hard to
hit when a head is replaced. I think the manual actually tells the
recorded track width though.
I realize I haven't put up detailed photos of the innards of this thing
(like the memory details), I will do so in the next week.