Micropolis 1325 repairs
R. D. Davis
While trying to revive a Micropolis 1325 hard drive, that spins down
after being powered up, I've noticed something which may (or may not)
possibly be of any interest, and I'm wondering if this makes any
difference. First, I'll include bits of some previous replies from an
earlier thread, along with some comments.
psu0...@odin.cc.pdx.edu (Don Taylor) writes:
|In message <53505v$k...@news.intur.net> - Bruce Mitchell
|<[8]mi...@cvfn.org>5 Oct 1996 06:45:19 GMT writes:
|>The drive unlocks the spindle, starts to spin up, reaches what I am
|>guessing to be operating speed, then either holds that speed for a
|>few seconds or spins right back down.
|In my experience over the last 8 years that is THE failure mode for
|the 1302/1304/1325 series of drives. I talked to a drive repair shop
|about this particular model of drive and they said they were aware of
|the problem. They claimed that they had been able to recover these
|drives by opening the drive in a clean room and tightening a bolt that
|gradually worked it self into a different position. When this change
|took place the drive was unable to find the servo within the time limit
|at power up and would shut it self down. As it spun down the logic
|would then take over and try to start the drive back up.
Urgghhh... as most of us don't have a clean room (yet), this won't do
much good. When these drives are repaired in this manner, are steps
normally taken, such as an adhesive like Locktite being used, which
solves this problem for a long time into the future?
It seems to me that there should also be some electronic solution,
such as altering some servo signals, to get around the clean room
approach. I haven't given this much thought, but perhaps it merits a
bit more thought.
|I stumbled across a trick get some of these drives up. If you put the
|drive in the refrigerator and cool to an appropriate temperature and
|then pull the drive and immediately try to power the drive up. Getting
|it too cold seems to not work and the behaviour is the same. If the
|drive won't come up try just a bit more cooling. I have successfully
|managed to get failed drives up and spinning and been able to use them
|until the power would fail and I had to start all over. It does seem
|that after doing this repeatedly that it might get harder to make them
|work.
I'll have to experiment some more with this... I didn't want to get
the drive too cold, so I didn't leave it in the refridgerator too
long the first time.
ar...@eng.cam.ac.uk (A.R. Duell) writes:
|One good reason for repairing older hard disks is that it's getting very
|difficult to find drives with the ST506/ST412 interface. And some older
|machines _require_ that interface - you can't go and get a SCSI or IDE
|card for a classic PERQ or a DEC Pro350.
Well said.
|This is, alas, a stock fault on these drives, and I _think_ it's due to
|the head assembly sticking to the end stop inside the HDA. It's not
|normally an electronic failure. The following procedure sometimes gets
|them to spin up so you can recover data from them :
|1) Flip the drive over, undo the 2 screws holding the logic board down,
|swing it up
Check.
|2) Unplug the 2-wire flexiprint cable from the motor/servo board (on the
|HDA). This is on the right hand side, looking from the back of the drive
|with it upside down.
Check.
|3) Apply power. The motor should start, get up to speed and remain up to
|speed
It did start, but it spun back down again. :-(
|4) Plug in the flexiprint and hopefully the heads will find track 0, and
|the drive will go ready.
No such luck. For drives where this works, I suggest installing a relay
and timer circuit, or at least a switch. This would make the drives a
bit more convenient to use.
|5) Back up the data :-)
|BTW, if anyone has the service manual/schematics for these drives, I'm
|looking for a copy as well. AFAIK, it did exist at one time...
Has this information turned up yet? Manufacturers which don't make
such information available, at a reasonable cost, even for their older
producs, should be placed on a Usenet "buyer beware" list. The
manufacturers could be placed into one of five categories:
0. Never make schematics and other service information available.
1. Schematics and service info. available for new(ish) products at a
reasonable cost.
2. Schematics and service info. available for new(ish) products at a
rather high cost.
3. Schematics and service info. available for old and new products at a
reasonable cost.
4. Schematics and service info. available for old and new products at a
rather high cost.
Back to the 1325. I've noticed that:
A. When mine spins up with the data and control cables disconnected, I
can hear the heads load and then, shortly thereafter, the drive spins
down completely and does not restart.
B. With the data and control cables disconnected, and the two-wire
thin cable mentioned above disconnected, the drive spins up, the heads
_fail_ to load, and then, shortly thereafter, the drive spins down
completely and does not restart.
C. With the data and control cables connected, and the two-wire thin
cable connected or disconnected, I can't hear the heads load when the
drive spins up. Shortly after the drive spins up, it slows down
slightly, spins back up for a brief period of time, slows down
slightly, spins back up, etc. over and over again with the same
timing. The diagnostic LEDs on the system's main board never indicate
that the drive has become ready.
Does anyone have any more ideas?
Next, it will be time to get another 1325 working which doesn't spin
up at all.
--
R. D. Davis Eccentrics have more fun! :-) http://www.access.digex.net/~rdd
r...@digex.net, r...@mystica.uucp Computer preservationist. Many types of
Home telephone: 1-410-744-7964 unwanted older computer systems disassembled,
Work telephone: 1-410-744-4900 removed for free (locally) and preserved.
A.R. Duell
r...@access1.digex.net (R. D. Davis) writes:
>Urgghhh... as most of us don't have a clean room (yet), this won't do
>much good. When these drives are repaired in this manner, are steps
>normally taken, such as an adhesive like Locktite being used, which
>solves this problem for a long time into the future?
Sometimes you can get away with dismantling the drive in as clean
conditions as possible, not touching the surface of the heads or
platters, and then holding it over the clean air stream from a
demountable hard disk (something like a DEC RK07), putting the cover on
and screwing it down. Then back up the data!
Don't try this on modern drives (the flying height is much lower) but
it's worked on some of the old full-height 5.25" 40Mbyte units, etc.
[Spinning up a 1300-series with the positioner open-circuit)
>|3) Apply power. The motor should start, get up to speed and remain up to
>|speed
>It did start, but it spun back down again. :-(
Odd... That doesn't sound like the classic failure mode for these drives.
Have you checked the power transistors on the driver board (the little
one near the spindle). Are the power lines correct _and stable_ at the
drive.
On the older 8" micropolis 1200's you could inhibit the spindown by
pluging in a special test box. This box would then give you an error code
that you looked up in the service manual, and it would indicate the fault
it had detected. A few more tests would normally find the dead component.
[...]
>|BTW, if anyone has the service manual/schematics for these drives, I'm
>|looking for a copy as well. AFAIK, it did exist at one time...
>Has this information turned up yet? Manufacturers which don't make
No, I'm still looking for it. I have a number of these drives still in
use, so I need to be able to keep them running.
>such information available, at a reasonable cost, even for their older
>producs, should be placed on a Usenet "buyer beware" list. The
>manufacturers could be placed into one of five categories:
I agree 100% here. Computer manufacturers are the worst here - almost
never are schematics available, and in many cases their so-called
technical support people are downright insulting. They can't believe that
a customer could possibily understand how the device works, and could
repair it. No matter that said customer has rebuilt minicomputer
processors (1000+ TTL chips), replaced hard disk heads, designed more
interfaces that the tech-support ape has had hot dinners, etc. Said
customer couldn't _possibly_ replace the spindle motor in a CD-ROM :-)
>0. Never make schematics and other service information available.
>1. Schematics and service info. available for new(ish) products at a
> reasonable cost.
>2. Schematics and service info. available for new(ish) products at a
> rather high cost.
>3. Schematics and service info. available for old and new products at a
> reasonable cost.
>4. Schematics and service info. available for old and new products at a
> rather high cost.
I think you need to reorder those :
0 (The best) - Service info for all products available at reasonable cost
(or free!)
1 - Service info for newish products available at reasonable cost
2 - Service info for all products available at high cost
3 - Service info for newish products available at high cost
4 - No service info available.
>Back to the 1325. I've noticed that:
[...]
>B. With the data and control cables disconnected, and the two-wire
>thin cable mentioned above disconnected, the drive spins up, the heads
>_fail_ to load, and then, shortly thereafter, the drive spins down
>completely and does not restart.
That's not suprising. You've unplugged the positioner coil, so the heads
can't load.
>C. With the data and control cables connected, and the two-wire thin
>cable connected or disconnected, I can't hear the heads load when the
>drive spins up. Shortly after the drive spins up, it slows down
>slightly, spins back up for a brief period of time, slows down
>slightly, spins back up, etc. over and over again with the same
Odd... I can't think of any reason why plugging in the controller cables
would make a difference. Have you checked the PSU lines _at the drive_. I'm
wopndering if the extra grounds on the controller cables are improving
the power to the drive.
>timing. The diagnostic LEDs on the system's main board never indicate
>that the drive has become ready.
>R. D. Davis Eccentrics have more fun! :-) http://www.access.digex.net/~rdd
--
-tony
ar...@eng.cam.ac.uk
The gates in my computer are AND,OR and NOT, not Bill
R. D. Davis
In article <5814j2$q...@lyra.csx.cam.ac.uk>,
>Sometimes you can get away with dismantling the drive in as clean
>conditions as possible, not touching the surface of the heads or
>platters, and then holding it over the clean air stream from a
>demountable hard disk (something like a DEC RK07), putting the cover on
>and screwing it down. Then back up the data!
That's a good idea. I still think that I should get around to
building a miniature clean room, or clean box, which shouldn't cost
very much to build.
>[Spinning up a 1300-series with the positioner open-circuit)
>
>>|3) Apply power. The motor should start, get up to speed and remain up to
>>|speed
>
>>It did start, but it spun back down again. :-(
>
>Odd... That doesn't sound like the classic failure mode for these drives.
>Have you checked the power transistors on the driver board (the little
>one near the spindle). Are the power lines correct _and stable_ at the
>drive.
I'm sure that the power lines are ok, as this problem exhibiots no
ramdom patterns. I would think that if I measured the time from
start-up to spin down, etc. that it would be nearly the exact same
time period each time.
>On the older 8" micropolis 1200's you could inhibit the spindown by
>pluging in a special test box. This box would then give you an error code
>that you looked up in the service manual, and it would indicate the fault
>it had detected. A few more tests would normally find the dead component.
With all of the computers and hard drives in existance, I don't
understand why no generic test-box has ever been mass produced, and a
version made available in kit form as well. Too many people assume
that when a hard drive stops working, for any reason, it's not
repairable.
>>|BTW, if anyone has the service manual/schematics for these drives, I'm
>>|looking for a copy as well. AFAIK, it did exist at one time...
>
>>Has this information turned up yet? Manufacturers which don't make
>
>No, I'm still looking for it. I have a number of these drives still in
>use, so I need to be able to keep them running.
I don't understand why such information should be so difficult to
obtain. People should refuse to purchase products from manufacturers
which don't make such information available.
>I agree 100% here. Computer manufacturers are the worst here - almost
>never are schematics available, and in many cases their so-called
>technical support people are downright insulting. They can't believe that
>a customer could possibily understand how the device works, and could
>repair it.
Emulex comes to mind here. They absolutely refuse to let me have, or
buy, schematics for their 4016 terminal server, which has a direct
short across the 5V supply somewhere. They claim that there is some
sort of FCC regulation prohibiting them from giving out schematics to
a non-authorized repair center! Furthermore, only one authorized
repair center exists for products used in tne US, which is located in
Puerto-Rico or some place like that. Worse yet, I've been told by
someone at Emulex that this is a typical failure, hence, the same part
probably fails, which means that this could possibly be a simple thing
to fix. All that Emulex is willing to do is take the old unit, and,
for a fee of $400, replace it with a refurbished one. I suspect that
these are badly designed on purpose just so that Emulex can continue
to extort funds from gullible idiots. Beware of Emulex!!!
>No matter that said customer has rebuilt minicomputer
>processors (1000+ TTL chips), replaced hard disk heads, designed more
>interfaces that the tech-support ape has had hot dinners, etc. Said
>customer couldn't _possibly_ replace the spindle motor in a CD-ROM :-)
Indeed. They don't want us to do this, and their employees are so
brainwashed by company propaganda that they refuse to believe that
many of us are capable of doing such things. I've even seen fuses
inside of power supplies with "no user serviceable parts inside"
stickers on them. That's taking things to the limits of lunacy, I
think. Even an utter moron should be able to replace a fuse.
>>0. Never make schematics and other service information available.
>>1. Schematics and service info. available for new(ish) products at a
>> reasonable cost.
>>2. Schematics and service info. available for new(ish) products at a
>> rather high cost.
>>3. Schematics and service info. available for old and new products at a
>> reasonable cost.
>>4. Schematics and service info. available for old and new products at a
>> rather high cost.
>
>I think you need to reorder those :
>
>0 (The best) - Service info for all products available at reasonable cost
>(or free!)
>1 - Service info for newish products available at reasonable cost
>2 - Service info for all products available at high cost
>3 - Service info for newish products available at high cost
>4 - No service info available.
That makes sense. Such a list is indeed needed rather badly. I'm
hoping that if enough companies start receiving enough bad publicity
on the 'net, perhaps some will change their ways.
>>C. With the data and control cables connected, and the two-wire thin
>>cable connected or disconnected, I can't hear the heads load when the
>>drive spins up. Shortly after the drive spins up, it slows down
>>slightly, spins back up for a brief period of time, slows down
>>slightly, spins back up, etc. over and over again with the same
>
>Odd... I can't think of any reason why plugging in the controller cables
>would make a difference. Have you checked the PSU lines _at the drive_.
Well, this definitely appears to make a differnce when the control and
data cables are connected and the computer system is turned on. I'll
have to dig up some of my information about the control signals for
these drives and see if I can get this to make some more sense.
>I'm
>wopndering if the extra grounds on the controller cables are improving
>the power to the drive.
Perhaps! When I got this hard drive, it had been inside a system that
appeared to have been stored in a damp area, as it was coroded rather
back in a few places. I'll clean the power connectors and see what
happens.
>-tony
Jeff D McMahill
R. D. Davis wrote:
>
> While trying to revive a Micropolis 1325 hard drive, that spins down
> after being powered up, I've noticed something which may (or may not)
> possibly be of any interest, and I'm wondering if this makes any
> difference. First, I'll include bits of some previous replies from an
> earlier thread, along with some comments.
>
I've run into this problem with several drives (something like 8 out
of 9 that I have!). I have tried both refrigerating the drives and
disconnecting things to bring them up. If these worked at all, it
was only temporary - the drive would fail again after a few power
cycles. As noted before, the failure seems to be caused by the
drive not being able to seek the heads. This is apparently due to
a rubber bumper that gets sticky over the years and prevents the
heads from moving during the initialization.
I finally decided to try the "clean room" approach, only
without the clean room, with the following results...
(This should be done in as clean a place as you can find, and as
quickly as possible to avoid getting dust in the drive. Also,
keep in mind that while this worked for me, it is a somewhat risky
process and can result in ruining the drive. This is very much
a hack, and may not work long term!! Don't say I didn't warn
you, etc...)
1. Remove the screws on the top of the drive - some may be covered
by the label.
2. Carefully remove the top cover.
3. Locate the head assembly and the rubber bumper.
4. Carefully move the head assembly (ie, do what the drive does
to seek the heads). It will probably stick a little bit at
first (which is why the drive can't move it!)
(now for the really scary part)
5. Cut out a small piece of paper so that it will fit in between the
bumper and the head assembly leaving about an inch or so sticking
out. The idea here is to keep the head assembly from sticking
to the bumper and the paper does this quite well.
6. Bend the paper over and carefully tape it to something that doesn't
move. If you can find a better way of attaching the paper than
tape, go for it. Who knows what might happen if the tape
deteriorates...
7. Replace the top cover and all the screws.
8. Cross your fingers and power up the drive...
I've done this to two drives so far and haven't had a single problem
with them for almost 2 years now. Now maybe I'm just lucky, so
do this at your own risk and only on a drive that you don't care too
much about!!
Good luck,
-Jeff
Jason D. Pero
In article <581prl$c...@access1.digex.net>,
There is another common problem with this micropolis 1300 series spindown
caused by gummy rubber stop. This thing stuck so strongly to the rotory
coil that drive cannot unstick itself! That favors the rubber stop in
parked position. Take that stop out by unscrewing two small screw, be
careful of strong magnet nearby and mark the position the stop is in!
Then scrape out the old rubber and replace it with something soft like
plastic glued or sticked on. Reinstall the stop in right position marked
out and check to make sure the heads does not touch the spindle.
I saved a old 40 megger this way and got the data out.
And I still have the 1355 ESDI (dated 1987) still spinning.
Jason D. Pero
R. D. Davis
In article <32A47E...@cs.cmu.edu>,
Jeff D McMahill <jm...@cs.cmu.edu> wrote:
>I've run into this problem with several drives (something like 8 out
>of 9 that I have!). I have tried both refrigerating the drives and
>I finally decided to try the "clean room" approach, only
>without the clean room, with the following results...
[warnings deleted]
~
|
| platters ~
| |
|\ <---stop? |
|- |
| +-----+
| # | |
| # | O |
| # | |
+-------------------------+
>3. Locate the head assembly and the rubber bumper.
Is the rubber bumper part of that small area labeled "stop?" in the
above diagram? I didn't see any rubber there, just the hard plastic.
However, I didn't spend much time looking -- perhaps next time, using
a flashlight -- I didn't want to keep the drive open for very long.
Something else that I noticed:
In the above diagram of the inside of the hard drive, there's a
semi-rectanguar, flat, piece of metal -- it's in the area wher the 'O'
in the rectangle above is. Underneath of this piece of metal is a
round, flat-appearing, piece of rubber. The screw which holds the
flat piece of metal in place is not tight, and the piece of metal
spins somewhat freely around the screw. When rotated clockwise, this
piece of metal snd screw tighten. I suspect that this is not supposed
to be loose, and suspect that it's positioning affects something.
The heads were stuck, but now move more freely. Now, when the drive
is powered up, it doesn't seem to be spinning down like it was. I can
now hear a grunting noise (head positioning mechanism, I guess), which
repeats every so often. I suspect that the heads are not able to find
track zero, and they keep trying. Am I correct to suspect that the
loose piece of metal, as mentioned in the previous paragraph, not
being aligned properly, is causing the present problem?
What is the correct way to position this rectangle, and what is its
purpose?
>4. Carefully move the head assembly (ie, do what the drive does
> to seek the heads). It will probably stick a little bit at
> first (which is why the drive can't move it!)
Yep, they were stuck. There... that's part of the problem solved. :-)
>5. Cut out a small piece of paper so that it will fit in between the
> bumper and the head assembly leaving about an inch or so sticking
> out. The idea here is to keep the head assembly from sticking
> to the bumper and the paper does this quite well.
>
>6. Bend the paper over and carefully tape it to something that doesn't
> move. If you can find a better way of attaching the paper than
> tape, go for it. Who knows what might happen if the tape
> deteriorates...
I may try this once I get the data backed up that's on this hard drive.
>8. Cross your fingers and power up the drive...
^^^^^^^^^^^^^^^^^^
>I've done this to two drives so far and haven't had a single problem
>with them for almost 2 years now. Now maybe I'm just lucky, so
Excellent!
>do this at your own risk and only on a drive that you don't care too
>much about!!
Of course, if the drive isn't working, I don't guess that there's much
of any other choice, except for sending it out to someone else for
repairs.
>Good luck,
Thanks -- good luck with keeping your drive working for a long time!
A.R. Duell
r...@access1.digex.net (R. D. Davis) writes:
>That's a good idea. I still think that I should get around to
>building a miniature clean room, or clean box, which shouldn't cost
>very much to build.
Hmmm... The 'shouldn't cost very much' is somewhat unlikely when you
realise that a replacement absolute filter for a demountable hard drive
(at least in the UK) is somewhere over $150. The rest of it - the fan and
the box should be quite cheap, I think. I wonder if this would be enough
for minor hard drive repairs ?
>I'm sure that the power lines are ok, as this problem exhibiots no
Check them. Now. Seriously, if I had a pound for every time power lines
were at the root of an obscure problem, I'd not be reading this. I'd be
at home with my own Cray :-). Stick a voltmeter between the power pins of
some of the chips on the drive PCB. Don't trust the power connector make
a good connection.
[...]
>With all of the computers and hard drives in existance, I don't
>understand why no generic test-box has ever been mass produced, and a
Err.... That would be _very_ difficult. Out of the drives I have that
have _documented_ test boxes :
The Micropolis 1200 takes a custom thing that monitors an 8-bit port on
the CPU (also used for the servo DAC I think), and allows you to ground
certain status inputs on the drive.
The RA60 takes a normal ascii terminal
The RX01 takes a KM11 (a custom DEC card set that also fits some models
of PDP11 and some other peripherals)
The R80 doesn't need one - it has LEDs and switches on the logic board
Some other winchesters have connectors on the logic board that _may_ be
for test boxes, but without the service manual I can't be sure.
Making a generic box to cover all of the above would be non-trivial.
[Service manuals]
>I don't understand why such information should be so difficult to
>obtain. People should refuse to purchase products from manufacturers
>which don't make such information available.
I already do. Before I purchase a product, I find out about the
availability of service manuals/schematics. I'll only buy something
without a service manuals if :
a) It's sufficiently close to a product I have the service manual for
that I can use said manual to repair it
b) It's sufficiently trivial that I can repair it without the service manual
c) There's no other alternative (at _any_ price).
[...]
>Emulex comes to mind here. They absolutely refuse to let me have, or
>buy, schematics for their 4016 terminal server, which has a direct
>short across the 5V supply somewhere. They claim that there is some
>sort of FCC regulation prohibiting them from giving out schematics to
>a non-authorized repair center! Furthermore, only one authorized
Err... Can somebody tell me what this regualtion states? I was told
something similar when I wanted a modem schematic - that it was illegal
to supply schematics other than to official service centres. The only
laws I could find relating to this where that (I think) modifying or
repairing (maybe the latter only if approved parts weren't used) would
invalidate the approval, and make it illegal to connect the modem to the
public telephone system. Nothing about it being illegal for me to have
the schematics in the first place.
>many of us are capable of doing such things. I've even seen fuses
>inside of power supplies with "no user serviceable parts inside"
>stickers on them. That's taking things to the limits of lunacy, I
>think. Even an utter moron should be able to replace a fuse.
Actually, I can believe that - particularly on switch-mode PSUs. When the
fuse fails in one of those, it means (normally) that other components
have failed as well, and replacing the fuse is a waste of time unless you
find the real problem. Of course some users are capable of repairing SMPSs...
>R. D. Davis Eccentrics have more fun! :-) http://www.access.digex.net/~rdd
Jeff D McMahill
> ~
> |
> | platters ~
> | |
> |\ <---stop? |
> |- |
> | +-----+
> | # | |
> | # | O |
> | # | |
> +-------------------------+
>
>
>>3. Locate the head assembly and the rubber bumper.
>
>Is the rubber bumper part of that small area labeled "stop?" in the
>above diagram? I didn't see any rubber there, just the hard plastic.
>However, I didn't spend much time looking -- perhaps next time, using
>a flashlight -- I didn't want to keep the drive open for very long.
No, the rubber bumper is in the 'O' section. If you watch as
you move the heads, you'll see something in there moving back and
forth between two pieces of rubber. The one you need to worry about
is the one that it rests against when the drive is powered off.
Also, in my previous post, I neglected to mention that there is
a locking mechanism on the heads that will keep them from moving
when there is no power to the drive. If you look in the area
with the '#'s on your diagram you will see a white plastic thing.
Just move that to unlock the heads.
>The heads were stuck, but now move more freely. Now, when the drive
>is powered up, it doesn't seem to be spinning down like it was. I can
>now hear a grunting noise (head positioning mechanism, I guess), which
>repeats every so often. I suspect that the heads are not able to find
>track zero, and they keep trying. Am I correct to suspect that the
>loose piece of metal, as mentioned in the previous paragraph, not
>being aligned properly, is causing the present problem?
>
>What is the correct way to position this rectangle, and what is its
>purpose?
Is this hooked up to a computer or powered up alone?
I'm not sure what the grunting noise would be exactly. As far as I
know, when the drive starts up, it seeks to the outer track and then
back to the inner track and then is ready. The only failure I've
seen is where it gets stuck and can't seek at all.
If you hook it up to a computer, can you access the drive?
I don't know about the rectangle thing. I'm not sure exactly what
part you're talking about. There is a flat piece of metal connected
to the seek motor which will move as the heads move. Is that
what you are talking about? If so, I don't know for sure what
it's for or what would happen if it's loose. My guess is that
it's probably some kind of balance thing or something.
If you do tighten it, just make sure that the heads can still
move freely!
-Jeff