[comp.sys.dec.micro] Problem with Rainbow hard drive/controller ?

martin@gvax.cs.cornell.edu (Bill Martin) (01/03/90)

I have a Rainbow+ with the following configuration:

   896K memory
   DOS 2.11
   20 Meg Seagate ST-225 hard drive (2-1/2 years old)
   Clikclok

Lately, it has has developed a perplexing problem.  Perhaps
one of you might know what's going on.  Four months ago, the
machine would not boot up from the hard disk, after many months of
reliable service.

The hard drive made a long series of "head seek" sounds before
the terminal displayed a could-not-load-operating-system sort of message.
The Winchester diagostics said that either the controller card
or the drive itself was bad (failing in subtest 2 on the head seek
test).  Powering the system on and off didn't seem to correct
the problem.  I then ran full system diagnostics (main memory, screen,
floppies, etc.) with no errors reported.  Strangely, the system
would then boot off of the hard disk with no problems after I ran
the diagnostics.

This erratic behavior happened again a couple of days ago. This time I was
running WordPerfect 4.2, and the drive failed while saving a wp workfile. 
The Winchester diagnostics pointed to the same problem as before.  I called
Suitable Solutions to price out a new controller board. 
The representative I spoke with said that the cards usually only fail 
within the first month of service, and that my problem was probably 
in either the hard drive cable connections or the hard drive itself.  
I reconnected the hard drive cables and the system would still not boot
up.  Whle I considered my next move, I ran system diagnostics, but this
time only the individual memory and system interaction tests. 
Again, no errors reported. Incredibly, my Rainbow would then boot 
off of the hard drive.

Does anyone know what might be going on here ? 

Thanks in advance,

Bill Martin
Cornell University

bcw@rti.UUCP (Bruce Wright) (01/04/90)

In article <35667@cornell.UUCP>, martin@gvax.cs.cornell.edu (Bill Martin) writes:
> The hard drive made a long series of "head seek" sounds before
> the terminal displayed a could-not-load-operating-system sort of message.
> The Winchester diagostics said that either the controller card
> or the drive itself was bad (failing in subtest 2 on the head seek
> test).  Powering the system on and off didn't seem to correct
> the problem.  I then ran full system diagnostics (main memory, screen,
> floppies, etc.) with no errors reported.  Strangely, the system
> would then boot off of the hard disk with no problems after I ran
> the diagnostics.

I would tend to agree with the people at Suitable Solutions.  This is
probably either a drive problem or a cable problem.  Controllers rarely
go bad (though that's not impossible either...).  We have had lots of
hard drive problems with our Rainbow ... lots of ST225's gone bad, plus
a bad hard drive cable at one point.  All of these can have similar
symptoms (we verified that the ST225's had gone bad by trying them out
on a PC, so we know that *both* our drives *and* our cable have gone
bad independently on different occasions).

I'd try replacing the cable to the hard drive, if you can find a
replacement.  Around here DEC will replace it for a suitably outrageous
price (maybe around $100 as I recall...).  Might be worth it as insurance.
You might also try running the drive on a PC, if you can stand backing
it up and reformatting it;  you might find that the PC has problems with
it too.

Our experience with Seagate 225's is VERY bad.  For a while our yearly
mortality rate for several machines was running at nearly 200% -- having
to replace each 225 in each machine about every 6 months.  At the moment
we are running the Rainbow with 251's without any problems (over a year
now), although I think I will probably never really trust a Seagate drive
again.

							Bruce C. Wright