Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!rochester!cornell!uw-beaver!tektronix!teklds!dadla!jamesa
From: jamesa@dadla.TEK.COM (James Akiyama)
Newsgroups: comp.sys.ibm.pc
Subject: Re: 30MB Hard Disk Problem
Message-ID: <2126@dadla.TEK.COM>
Date: Thu, 25-Jun-87 22:26:11 EDT
Article-I.D.: dadla.2126
Posted: Thu Jun 25 22:26:11 1987
Date-Received: Sat, 27-Jun-87 05:24:06 EDT
References: <4490004@hpcvc0.HP>
Organization: Tektronix Inc., Beaverton, Or.
Lines: 60


As most of you probably guessed, Seagate did have a thermal reliability
problem on many of their disk drives.  They also had some problems with their
stepper motors, I believe.  Since that time, Seagate has told us that they
corrected both problems (I haven't actually verified this since the disk drives
are no longer an engineering concern).

For the most part we (Tek. LA) have had good luck with the ST225 drives (from
which the ST238 was derived).

A couple of things that you might try:

    1.	Try backing up the fixed disk, then reformatting it with a low-level
	format routine while the drive is operating at nominal temperature
	(about 25 degrees C after running for approximately one-half hour).  Be
	sure to map out the listed bad blocks.  Usually its best to map out the
	entire track, although it is possible (through calculations) to derive
	and map out only the affected sectors.  Note that it is usually best
	to format the drive with the actual controller that the drive is to be
	used with (allows greatest margin on the data separation).

    2.	Be sure you are using the correct Write Precompensation.  For the ST225,
	the correct value is 12 nS. for both early and late bits from track
	300 to track 614.  This may be different on the ST238 (I don't have a
	ST238 manual).  Adaptec may have already set this value correctly.

    3.	Make sure your drive enclosure has adequate ventilation.  The Seagate
	Drives disipate approximately 15 watts and must run under 45 degrees
	C.  Generally this means have adequate heat conduction from the drive's
	frame to the enclosure and/or providing force air circulation.  I
	believe that IBM's derated temperature specification for the PC's is
	a result of this requirement.

    4.	Verify that the grounding strap is attached.  This strap serves to
	"bleed off" electrical charge which could otherwise cause errors during
	reads.

    5.	If possible, verify the buffered seek pulse widths meet the Seagate
	timing requirements.  Buffered seek pulses MUST be between 5 uSecs and
	200 uSecs for the drive to function properly in buffered mode.  Note
	that several disk controller chips (e.g. Western Digital WD2010)
	provide a FAST seek rate (3.2 uSec for the WD2010).  This fast rate
	may cause the Seagate drive to "miss" pulses, resulting in a seek
	error.

    6.	Finally, be sure the drive is reasonably shielded from stray magnetic
	and electrical fields.  This is especially a problem on drives which
	are not mounted in an enclosed cabinet.

Also worth noting is the fact that there are reports of problems with the
Seagate ST238 and the Adaptec 2070A (as well as the WD1003-RAH).  Western
Digital has told us that the problem appears to be in the Seagate Drive.  I
haven't followed up on this at all.  Hope this helps.

James E. Akiyama
Logic Analyzer Division
Tektronix, Inc.
P.O. Box 4600  Mail Sta. 92-725
Beaverton, OR 97075
(503) 629-1907