SUMMARY: Ancient Sun3 ESDI problems

Date: Thu May 25 1995 - 22:22:00 CDT

Original inquiry:

> I have a venerable 3/60, still running 3.5 on a 141MB Micropolis 1355
> attached via an Emulex MD21. The drive has developed a spinup problem,
> and its intended replacement (a Maxtor XT-8760EH) won't format. Details:
> Yesterday afternoon, I shut the system down to avoid overheating. When
> I tried to bring it back up later, the drive would not boot -- the ROM
> reports some kind of sd error repeated about 64 times (4 groups of 16
> retries each) and then asks me to spin up the disk. I do hear the disk
> spin up at power on, but after maybe a minute it spins down a little,
> then back up, and after repeating this cycle about 4 or 5 times it
> spins down and stays down until power is cycled. I've tried replacing
> the MD21 with no effect on the problem, and the other drive on the same
> MD21 appears to be working OK. Power supply voltages look good.
> The only place I've seen such behavior previously is with a 3.5" Maxtor
> IDE drive, which recovers once it warms up for an hour or so, but even
> overnight warmup has not been enough to resurrect this 1355. There's
> nothing irreplaceable on it, but I'd like to avoid the reinstallation
> hassle if possible. Anyone seen this before? Any cures, even
> temporary ones?
> As to the Maxtor XT-8760EH, has anyone gotten one of these to work with
> the Sun3-Emulex MD21 combination, and if so how did you have the drive
> jumpered? This particular drive is reported to work properly on a PC
> using an Ultrastor controller, and I'm using the proper dimensions
> (1632 cyls x 15 heads x 53/54 sectors), but formatting fails strangely.
> I'm using the same approach which has worked previously with other
> unsupported drives, so I don't think it's pilot error in running
> stand/diag.
> * stand/diag is unable to find any defect list, and after "Extracting
> defects from controller" (and finding none) it reports the drive as
> unformatted. This is not totally surprising since it was previously
> formatted on a different type of controller.
> * When I try to format it with either 53 or 54 sectors specified, the
> mode select fails. That problem goes away if I specify 52 sectors,
> but the format fails after about 30 cylinders -- I can hear the seek
> between cylinders -- reporting a media error on block 0. I've tried
> using fewer cylinders, even fewer than 30, but always get the same
> failure reported. I've also tried marking the beginning of track
> zero as defective.

Thanks to:!gecko!oconnor!bminer (Bob Miner OPS)

So far, no success with either problem. I ended up reinstalling on
a different drive.

Two responses suggested power-supply output problems (noise, ripple,
etc.) which might not show on a voltmeter. Good suggestion, but not
the culprit this time: the drive fails the same way in a different
shoebox, and other drives (including another Micropolis 1355) work
properly in the original shoebox. Another response suggested applying
the sticktion cure: whack it to loosen up the spindle lubricant. I
haven't tried that -- visions of inducing a head crash, and it does
spin up although I suppose it may not quite get all the way up to
normal speed and that may be why it doesn't assert ready. Finally,
one response suggested updating to newer technology...

Regarding the Maxtor XT-8760EH, I got one report of the combination
working but no details and I have not been able to accomplish a
similar feat. Maybe I can use that drive on something else.

Perry Hutchison
Not speaking for Intel, of course

This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:10:25 CDT