Tony Duell wrote:
Yes, this
drive did have a sticky bumper. I resolved it by putting a very
thin sliver of tape over the sticky bit. I did not loosen any bolts, but
I wonder if that bumper is no the wrong size, and that is what's causing
the problems.Maybe the heads have to be able to find something o nthe
disk for hte drive to not find an error.
-tony
I've outright removed the bumper with no ill effect. Reason for that
is unsticking it once didn't work as it would stick again. Only solution
was remove the offending material. Result was a reliable working drive
that when powered off emitted a noticeable "clunk". That has had no
impact on reliability, least not for the last 10 years.
I'd suspect board level death or possible cable/connector issues
between boards. I've found in the dead boards, toasted resistors
in the motor drive and head control and random failures
around the local CPU (servo control). Drives that are received
non functional may have been fiddled by those unknowing and
have incurred further damage.
I've found that I have RD53s with serviceable HDAs (after fixing the stick)"
and bad boards and some where the HDAs had other issues Grafting
good HDAs to the working boards was the solution. Still have a stack
of RD53s I haven't looked at yet beyond verifying they have sticky
bumper as I haven't needed them. All told I run 5 of them and have
a more for spares other than they run hot and get the sticky bumper
they are decently reliable if well cooled. The bad board RD53s always
seemed to come from BA23/123 or microVAX2000 boxes with failed
fans. My assumption is the fan failed and the drive was the first to
signal that by failing electronically. This was also common with the
hot running RD32 (ST250) if cooling was poor. Oddly enough this
was seen most commonly inside DEC (ex-digit, the Mill) where small
systems like that would live unofficially under desks and such without
"service". Most migrated out of DEC as scrap or as employee purchase
and would die at home where I'd later acquire them as part of a system
rescue.
The best drive of the lot was the RD52 (Quantum D540) 31mb. I watched
as one was flung into a waste pail as "too small). I recovered it as a
possible set of boards and found it was still fully functional with no
new bad sectors. I have about 15 of them in various PDP11 and uVAX
systems and the oldest is going on 21years old and none have ever
failed.
Allison