3B1 Hard Disk Woes (Plea for HELP!)

Kevin W. Wall kww at cbnews.ATT.COM
Tue Jul 25 12:13:50 AEST 1989


I am having trouble with my hard disk on my 3B1 (a 67Mb Miniscribe).

A while back, the power supply on my 3B1 failed.  I since have replaced
the power supply, but its failure apparently caused the hard disk to fail
as well.

At the time, I thought that was no big deal; I figured I could just reformat
the hard drive and everything would be okay.  Well, when THAT didn't work,
I thought I might have had a head crash, so I borrowed a colleague's
"UNIX PC Reference Manual" and turned to the diagnostics section (Chapter 3)
so seem I could confirm whether or not it was indeed a head crash.  As it
turns out, even the so called "Expert Mode Diagnostics Program" is virtually
useless, but THAT'S another story.  However, I did learn that 1) all hard
disk diagnostics apparently first try to "recalibrate" the drive, and 2) this
"recalibration" always would fail.  For example, if I run the diagnostic
to park the heads on the hard disk, I will get the following output:

		WINCHESTER DISK TEST

		Hard disk restore failed
			**ERROR**
		Test:		Hard disk test (drive 0)
		Subtest:	Park Disk Heads
		Error:		WINCHESTER: Can't Recal; Response = 4
			Enter y[Y] to Abort, Return to continue

Anything that I tried run that concerned the hard disk resulted in the same:

		Error:		WINCHESTER: Can't Recal; Response = 4

however, all diagnostics on other devices (e.g., floppy, CPU, memory, etc.) 
passed with no errors.

Now my questions are these:

	1) What exactly does "Can't Recal; Response = 4" mean?  In particular,
	   what does this most likely indicate?  (E.g., a problem with the disk
	   controller, with the hard disk media, etc.)
	2) Can I safely assume that the problem is in the hard disk UNIT
	   (all the pieces inside the smaller cage containing the disk itself),
	   as opposed to being some kind of problem on the mother board?

	   [The reason that I want to no this is that I have a 72Mb Miniscribe
	   (Model # 6085) that I could (would like to) install to replace the
	   (bad?) drive, but don't want to do this it there is a problem some-
	   where on the mother board which will cause this one to fail too.]

	3) Is there any way, short of opening the sealed drive itself, to
	   tell if the problem is a head crash (vs. say, the failure of the
	   hard disk controller)?  [Note that the no portion of the 3B1 is
	   still under warranty, so I am not adverse to opening the disk
	   unit and peering inside.  However, I would like to send the disk
	   drive to be refurbished, so I don't want to break the seal if
	   this will keep companies from trying to fix it or if I have a
	   good chance of making things worse than they already are.]

	4) Assuming the disk is repairable (how do can I tell, especially
	   without opening it?) does anyone know of a reliable and reputable
	   company that you would trust fixing it?  [Recovery of the data is
	   only of secondary interest.  I have (about) everything either on
	   floppy, or on a mainframe at work, or am willing to do without.
	   I estimate to recover everything will take approximately 72 hrs.
	   of connect time at 2400 baud so if recovery is CHEAP enough, I
	   may be interested; it depends if I could have them selectively
	   recover certain directories (e.g., /usr/local and my $HOME
	   directory).]

Some final info that might be pertinent.  I was running version 3.51 of the
operating system when the system "crashed", and ran version 3.51 of the
diagnostics disk to analyze the problem.  The specifics on the disk itself
follow.  (Fortunately, I wrote this info down a long time ago, as the
diagnostics to get this information now fail!)

	Volume Name: mi67-4
	1024 Cylinders. 8 heads per Cylinder.
	Configured as:
		Total space:	65,536 blocks
		Partition 0 =	64 blocks
		Partition 1 =	5000 blocks
		User space =	60,472 blocks

One final note: please E-mail reply directly to me as opposed to congesting
this newsgroup with a bunch of follow-up discussions.  If enough interest is
expressed, I'll summarize to the net.

Please try to explain in layman's terms; I'm a UNIX hacker, not a hardware
jock. :-)

Thanks in advance for your help!
-- 
In person: Kevin W. Wall			AT&T Bell Laboratories
Usenet/UUCP: {att!}cblpf!kww			6200 E. Broad St.
Internet: kww at cblpf.att.com			Columbus, Oh. 43213
		"Death is life's way of firing you!" -- Hack rumor



More information about the Unix-pc.general mailing list