12 Nov
2006
12 Nov
'06
4:16 p.m.
The RD53 drive in my 11/73 has finally gone to silicon heaven, after a
brief resurrection. Given the reputation these drives seem to have, I'm
not too surprised. So a few nights ago I dug through my stacks of
ancient hard drives and found three possible candidates for replacement:
- A Microscribe 6085, which seems to have the same drive geometry as an RD53
- Two Seagate ST-251s, which seem to be identical to an RD32
All of these are in "unknown" condition (I haven't used them in years),
but spin up without making any evil noises and are recognized by the
RQDX3 controller. To the best of my recollection, all three were
formerly used in PCs.
I've used zrqch0 (booted via vtserver) to format all three of these
disks. In all three cases the formatting appears to be successful --
the drives format for a few minutes, then go through three verification
stages for another 20-30 minutes. During this time I can hear the heads
being stepped, and the drive access light blinks. At the end of the
format no errors are reported.
I'll note that in order to do the format on these drives, I had to go
through the process manually (as opposed to the "AUTOFORMAT" mode of
zrqch0) -- for the Microscribe this was because zrqch0 didn't recognize
the drive (showed up as "unknown"), for the ST-251s this was because the
fault tables could not be read off the drives. I suspect this may have
something to do with the issue I'm seeing, but of course this is pure
speculation. I've tried various permutations of the formatting options,
but I've had no joy there.
After formatting the drives, no other utilities seem to be able to make
use any of the drives -- running the disk exerciser (zrqah0) fails after
the first test step (and after a very small amount of disk activity)
with an error similar to the below (this varies slightly based on the
test being run):
FATAL I/O ERROR
* DISK 0 WENT OFFLINE
* SUB_CODE: NO VOLUME MOUNTED OR DRIVE DISABLED BY SWITCH
* COMMAND: READ_COMPARE
Interestingly, 2.11BSD's 'disklabel' is able to partition and write the
label to the disk. Mkfs, however, fails with:
ra(0,0,0) error op=A2 sts=23
write error 3
Exit called
I've tried this with two different RQDX3 controllers with the same
results (and these controllers worked fine with my RD53 until it
failed). Is there a magic step I'm missing? Is what I'm attempting to
do (use non-DEC drives) not possible? Any ideas or suggestions?
Thanks for any help here...
Josh
14 Nov
14 Nov
7:56 a.m.
On Sun, 12 Nov 2006 16:16:56 -0800
Josh Dersch <derschjo at msu.edu> wrote:
Is what I'm attempting to do (use non-DEC drives)
not possible? Any
ideas or suggestions?
Do you have a VAXstation 2000 / MicroVAX 2000? You can use it
to format
(non DEC MFM) disks for use on a RQDX3.
--
tsch??,
Jochen
Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/
8:12 p.m.
Josh Dersch wrote:
The RD53 drive in my 11/73 has finally gone to silicon
heaven, after a
brief resurrection. Given the reputation these drives seem to have,
I'm not too surprised. So a few nights ago I dug through my stacks of
ancient hard drives and found three possible candidates for replacement:
RD53... Micropolis 1325 I believe. I ran one of these (I believe it was
an RD53
actually... not just a generic 1325) in a PC years ago.
It one day started not wanting to initialize... startup... think think,
chunk chunk,
spin down....
With nothing to lose, I began to experiment...
What I found was that it couldn't apparently locate track 0. If the HDA
os opened,
it has two rubber? bumpers for the extreme ends of head motion. One for
track 0
and one for last track ?
They look something like:
----
| \
|____ /
with an adjustable slide 'hole' in the middle.
What I'm guessing went wrong is that the material hardens and/or shrinks
with age.
By carefully tweaking the adjustment on this item to allow a mere few
thousands of
an inch inward... the drive started initializing fine every time.
I did not use the drive much after that, so how drive longevity would be
effected by
the introduction of contaminants into the HDA assembly I don't know (I
of course
don't have a clean room :-) ).
However, if you have a fairly clean space, and can work carefully as
such not to
stir up contaminants, you may be able to rescue your RD53.
Another problem I read of someone having a problem with an RD53 had to do
with the brake solenoid assembly needing adjusting as the brake pad was
dragging
(it is located under the circuit board on the bottom as I recall... no
inside the HDA).
-- Curt
- A Microscribe 6085, which seems to have the same drive geometry as
an RD53
- Two Seagate ST-251s, which seem to be identical to an RD32
All of these are in "unknown" condition (I haven't used them in
years), but spin up without making any evil noises and are recognized
by the RQDX3 controller. To the best of my recollection, all three
were formerly used in PCs.
I've used zrqch0 (booted via vtserver) to format all three of these
disks. In all three cases the formatting appears to be successful --
the drives format for a few minutes, then go through three
verification stages for another 20-30 minutes. During this time I can
hear the heads being stepped, and the drive access light blinks. At
the end of the format no errors are reported.
I'll note that in order to do the format on these drives, I had to go
through the process manually (as opposed to the "AUTOFORMAT" mode of
zrqch0) -- for the Microscribe this was because zrqch0 didn't
recognize the drive (showed up as "unknown"), for the ST-251s this was
because the fault tables could not be read off the drives. I suspect
this may have something to do with the issue I'm seeing, but of course
this is pure speculation. I've tried various permutations of the
formatting options, but I've had no joy there.
After formatting the drives, no other utilities seem to be able to
make use any of the drives -- running the disk exerciser (zrqah0)
fails after the first test step (and after a very small amount of disk
activity) with an error similar to the below (this varies slightly
based on the test being run):
FATAL I/O ERROR
* DISK 0 WENT OFFLINE
* SUB_CODE: NO VOLUME MOUNTED OR DRIVE DISABLED BY SWITCH
* COMMAND: READ_COMPARE
Interestingly, 2.11BSD's 'disklabel' is able to partition and write
the label to the disk. Mkfs, however, fails with:
ra(0,0,0) error op=A2 sts=23
write error 3
Exit called
I've tried this with two different RQDX3 controllers with the same
results (and these controllers worked fine with my RD53 until it
failed). Is there a magic step I'm missing? Is what I'm attempting
to do (use non-DEC drives) not possible? Any ideas or suggestions?
Thanks for any help here...
Josh
8:27 p.m.
On Tue, 14 Nov 2006, Curtis H. Wilbar Jr. wrote:
However, if you have a fairly clean space, and can
work carefully as
such not to stir up contaminants, you may be able to rescue your RD53.
I've heard that a bathroom is the best place to go about approximating a
cleanroom. Start off with a freshly-cleaned bathroom to get rid of skin
flakes and such. Get rid of stuff on the counters. Then run the shower
hot for a while to knock down floating dust. Then run the shower cold a
bit to drop the humidity.
--
David Griffith
dgriffi at cs.csubak.edu
A: Because it fouls the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
15 Nov
15 Nov
12:30 a.m.
Josh Dersch wrote:
I actually did something similar with my RD53. When I got the machine,
the drive was dead -- it'd spin up, do nothing for ~30 seconds and then
spin down.
I wagered that the heads were not moving, probably a stiction problem --
since I'd nothing to lose (same as your situation) I took the cover off,
and while the drive was powered up, I gave the heads a small nudge.
This got the drive working again. I was pretty pleased with myself :).
Ran long enough to get 2.11BSD installed on it, and everything was
sunshine and lollipops until last weekend when it died again.
It still spins up, but it never seems to get to a "ready" condition. Or
at least, that's how it _sounds_ -- it spins up, then sounds like it
starts to spin down a tiny bit, then spins up again in rapid
succession. It will do this for as long as I leave it running, with no
head activity at all.
So I'd like to get one of the spare PC drives I have running in this
thing, but nothing's ever that simple, it seems... :)
(On an almost completely unrelated note, I've noticed that my mails take
1-2 days to actually make it to the list. I sent out my original mail
on Sunday afternoon (11/12) and I didn't see it show up on the list
(i.e. it didn't show up in my inbox) until Tuesday morning! Anyone else
have this issue? I know I don't get delays like this sending email
elsewhere...)
Thanks,
Josh
The RD53 drive in my 11/73 has finally gone to
silicon heaven, after
a brief resurrection. Given the reputation these drives seem to
have, I'm not too surprised. So a few nights ago I dug through my
stacks of ancient hard drives and found three possible candidates for
replacement:
RD53... Micropolis 1325 I believe. I ran one of these (I believe it
was an RD53
actually... not just a generic 1325) in a PC years ago.
It one day started not wanting to initialize... startup... think
think, chunk chunk,
spin down....
With nothing to lose, I began to experiment...
What I found was that it couldn't apparently locate track 0. If the
HDA os opened,
it has two rubber? bumpers for the extreme ends of head motion. One
for track 0
and one for last track ?
They look something like:
----
| \
|____ /
with an adjustable slide 'hole' in the middle.
What I'm guessing went wrong is that the material hardens and/or
shrinks with age.
By carefully tweaking the adjustment on this item to allow a mere few
thousands of
an inch inward... the drive started initializing fine every time.
I did not use the drive much after that, so how drive longevity would
be effected by
the introduction of contaminants into the HDA assembly I don't know (I
of course
don't have a clean room :-) ).
However, if you have a fairly clean space, and can work carefully as
such not to
stir up contaminants, you may be able to rescue your RD53.
Another problem I read of someone having a problem with an RD53 had to do
with the brake solenoid assembly needing adjusting as the brake pad
was dragging
(it is located under the circuit board on the bottom as I recall... no
inside the HDA).
7:55 p.m.
Josh Dersch wrote:
As I recall, that was the behavior mine had too... slightly tweaking
that bumper
solved it....
Bet if I dig it out of storage, it still works... other than the issue
(most likely
due to the materials aging), that drive was pretty 'bullet proof'...
-- Curt
Josh Dersch
wrote:
I actually did something similar with my RD53. When I got the
machine, the drive was dead -- it'd spin up, do nothing for ~30
seconds and then spin down.
I wagered that the heads were not moving, probably a stiction problem
-- since I'd nothing to lose (same as your situation) I took the cover
off, and while the drive was powered up, I gave the heads a small
nudge. This got the drive working again. I was pretty pleased with
myself :). Ran long enough to get 2.11BSD installed on it, and
everything was sunshine and lollipops until last weekend when it died
again. The RD53 drive in my 11/73 has finally gone to
silicon heaven, after
a brief resurrection. Given the reputation these drives seem to
have, I'm not too surprised. So a few nights ago I dug through my
stacks of ancient hard drives and found three possible candidates
for replacement:
RD53... Micropolis 1325 I believe. I ran one of these (I believe it
was an RD53
actually... not just a generic 1325) in a PC years ago.
It one day started not wanting to initialize... startup... think
think, chunk chunk,
spin down....
With nothing to lose, I began to experiment...
What I found was that it couldn't apparently locate track 0. If the
HDA os opened,
it has two rubber? bumpers for the extreme ends of head motion. One
for track 0
and one for last track ?
They look something like:
----
| \
|____ /
with an adjustable slide 'hole' in the middle.
What I'm guessing went wrong is that the material hardens and/or
shrinks with age.
By carefully tweaking the adjustment on this item to allow a mere few
thousands of
an inch inward... the drive started initializing fine every time.
I did not use the drive much after that, so how drive longevity would
be effected by
the introduction of contaminants into the HDA assembly I don't know
(I of course
don't have a clean room :-) ).
However, if you have a fairly clean space, and can work carefully as
such not to
stir up contaminants, you may be able to rescue your RD53.
Another problem I read of someone having a problem with an RD53 had
to do
with the brake solenoid assembly needing adjusting as the brake pad
was dragging
(it is located under the circuit board on the bottom as I recall...
no inside the HDA).
It still spins up, but it never seems to get to a "ready" condition.
Or at least, that's how it _sounds_ -- it spins up, then sounds like
it starts to spin down a tiny bit, then spins up again in rapid
succession. It will do this for as long as I leave it running, with
no head activity at all.
So I'd like to get one of the spare PC drives I have running in this
thing, but nothing's ever that simple, it seems... :)
(On an almost completely unrelated note, I've noticed that my mails
take 1-2 days to actually make it to the list. I sent out my original
mail on Sunday afternoon (11/12) and I didn't see it show up on the
list (i.e. it didn't show up in my inbox) until Tuesday morning!
Anyone else have this issue? I know I don't get delays like this
sending email elsewhere...)
Thanks,
Josh
16 Nov
16 Nov
2:52 a.m.
On Wed, 15 Nov 2006, Josh Dersch wrote:
It still spins up, but it never seems to get to a
"ready" condition. Or at
least, that's how it _sounds_ -- it spins up, then sounds like it starts to
spin down a tiny bit, then spins up again in rapid succession. It will do
this for as long as I leave it running, with no head activity at all.
I have two Micropolis 1335 with the same problem. I've found out that the
drive gets ready when you do the following:
- With power off, remove the small two conductor film cable from the
control board to the drive (this is the head actuator power cable)
- Power up the drive and wait until it has reached its speed
- Plug in the cable et voil?, the drive makes "pling" and is ready
This made me build a little dirty hack to the drive. I've put a small
relais (or what you call solenoid) between the logic board and the
actuator cable that gets energized after a delay of several seconds
(enough time to let the drive spin up). This is the automated version of
plugging the cable out and in... The drive works fine since then.
BTW when my drives started with this problem, they ran fine as long as
they were powered up. But as soon as I powererd them down they wouldn't
start normally, I always had to wait at least two weaks, then they would
start again. If I tried after one week, it didn't work.
Christian
PS:
I'm happy for each MFM drive I can get, even defective ones. They're
getting very rare now.
6612
days inactive
6615
days old
6 comments
5 participants
participants (5)
-
cc＠corti-net.de -
derschjo＠msu.edu -
dgriffi＠cs.csubak.edu -
jkunz＠unixag-kl.fh-kl.de -
rescue＠hawkmountain.net