[CCTALK] HD repair techniques?

From: Tothwolf <tothwolf_at_concentric.net>
Date: Wed May 15 16:21:40 2002

On Wed, 15 May 2002, Tony Duell wrote:

> > repair it with. The second drive has a bad head, but it affects one of the
> > lower platters. I've never had to replace a head before, so I'm looking
> > for advice, or even a step by step guide to replacing it.

I'm starting to think the head assembly of that second drive might not be
bad. The platters seem to spin at too high of speed, which makes me wonder
if the spindle motor is bad. That drive's logic board checked out ok on
yet another drive, so I know that much of it is good. It also (according
to the date written on it) failed right at 2 years after it was
manufactured. I think those particular drives actually had a 3 year
warranty on them, so I guess someone didn't know to send it in for a
replacement (it was marked 'bad' when I got it.)

> > The drives are both a Conner CFS1275A, 1.2GB, 3 platters, 6 surfaces. I'm
> > starting to suspect the top surface of the upper platter is the servo
> > surface, due to the behavior of the drive when it failed. Can anyone
> > confirm if this is the case, or if these type of drive even use a servo
>
> My guess is that these drives use 'embedded servos'. Basically, the
> servo bursts are in the sector headers on each surface -- there is no
> dedicated servo surface. The DEC RL01 is an early and understandable
> example. So if your top platter is unreadable, the servo will not lock
> when you try to read that surface, and this may be enough to cause the
> drive to spin down.

That makes sense. Do you think these drives stripe data across the
platters, or would data be more "randomly" placed? Would a drive normally
start from the top side of the top platter and go towards the bottom of
the stack, or do some drives use a more "random" scheme? I'm wondering if
there is anything left to recover with one side of one platter badly
damaged.

> I have to tell you that what you are doing is a 'challenge'. The
> flying height on such drives is much lower than on the older
> winchesters that some of us have repaired at home (== you need a much
> better 'clean box' to avoid further headcrashes). Also, there's
> generally no procedure for aligning the head. The heads are assembled
> at the factory, then the head arm is connected to a formatter
> (basically a very accurate stepper system :-)), then the tracks and
> servo bursts are written. There is no easy way to replace a head and
> get it aligned with the existing tracks.

I'm not one to bow out to such a 'challenge', though if I don't feel
comfortable in doing the work, I'll sometimes set things aside until I
have more information or better tools to work with.

How clean will the air need to be? Since I don't plan to use the drive
long-term, as long as it will hold up for about an hour or so, I'll be
happy. What about dust and debris caused by the initial head crash? Is it
likely the scratches on the platter will damage a new head, or is there
something I'll need to do to prevent that?

Would it be possible to replace the entire head assembly with that from
another (possibly working) drive? Would that have a better shot at keeping
the heads aligned? Anyone have a dead CFS1275A that just has a damaged
logic board? (Preferably one of the earlier versions with the green Conner
sticker, and close to a 9532 date code.)

I started going through the classiccmp archives for posts on this topic,
but it is very hard to sort out all the bits of information. Is anyone
aware of a faq/guide online somewhere that covers this sort of thing?

-Toth
Received on Wed May 15 2002 - 16:21:40 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:35:16 BST