30 Jan
2003
30 Jan
'03
5:53 a.m.
Folklore for PC's was that if you recorded data on a hard drive with the
platters horizontal, remounting it vertically could cause read errors. Any
truth to this?
-----Original Message-----
From: John Allain [mailto:allain@panix.com]
Sent: Thursday, January 30, 2003 10:32 AM
To: cctalk(a)classiccmp.org
Subject: Re: OT: the 1U system
<snip>
I wonder if disk drives fail at the same rates horizontally
as they do vertically (an open question) ?
John A.
30 Jan
30 Jan
6:06 a.m.
New subject: Mounting hard disks at strange angles (was RE: OT: the 1U system)
--- "Feldman, Robert" <Robert_Feldman(a)jdedwards.com> wrote:
Folklore for PC's was that if you recorded data on
a hard drive with the
platters horizontal, remounting it vertically could cause read errors.
Any truth to this?
I remember that was a commonly accepted concern back in the days when
we used to have to do our own low-level formatting (ST506/ST412 drives,
primarily). I think since modern drives have things packed so tightly
that the thermal expansion and contraction of the platters causes the
tracks to perceptibly migrate, causing drives to be *much* smarter
about positioning than they used to, might have made that concern
(if there ever really was one) obsolete.
I've heard cautions for more modern drives that they can be mounted
any way except one, typically either nose down or circuit board up,
presumably due to bearing loading issues. The Amiga 4000 had the
drives routinely mounted "upside down" with no concerns, but I
would think that one would *never* want to do that to, say, an
RK05.
Anyone have any personal experiences with oddly-mounted drives?
-ethan
6:41 a.m.
New subject: Mounting hard disks at strange angles (was RE: OT: the 1U system)
On Thu, 30 Jan 2003, Ethan Dicks wrote:
--- "Feldman, Robert"
<Robert_Feldman(a)jdedwards.com> wrote:
My venerable IBM XT spent it's long life standing on it's side next to
my desk. In the fulness of time, I upgraded, and it was purchased by a
friend. He came over, we put it through some confidence tests, and he
took it home. And called me a half-hour later, because it wouldn't boot
at all, giving a fatal HD fault.
Long story short: he put it on the desk, on it's "feet", and the old 10M
Seagate promptly failed. Turning it up on edge restored it to life and
DOS, in that order. So we took out the HD, wrapped cardboard around it to
insulate the PC board, and left it outside the PC case on it's edge.
Ran for another two years like that, doing statistics number crunching
(SPSS).
Cheers
John
Folklore for PC's was that if you recorded
data on a hard drive with the
platters horizontal, remounting it vertically could cause read errors.
Any truth to this?
Anyone have any personal experiences with oddly-mounted drives? 
8:18 a.m.
New subject: Mounting hard disks at strange angles (was RE: OT: the 1U system)
Hi,
John Lawson <jpl15(a)panix.com> said:
On Thu, 30 Jan 2003, Ethan Dicks wrote:
My almost equally venerable AT was used horizontally in work but when I
acquired it and took it home I ran it vertically under the desk with no
problems. Still works fine, but now horizontally again.
--
Cheers,
Stan Barr stanb(a)dial.pipex.com
The future was never like this!
--- "Feldman, Robert"
<Robert_Feldman(a)jdedwards.com> wrote:
My venerable IBM XT spent it's long life standing on it's side next to
my desk. In the fulness of time, I upgraded, and it was purchased by a
friend. He came over, we put it through some confidence tests, and he
took it home. And called me a half-hour later, because it wouldn't boot
at all, giving a fatal HD fault. Folklore for PC's was that if you recorded
data on a hard drive with the
platters horizontal, remounting it vertically could cause read errors.
Any truth to this?
Anyone have any personal experiences with oddly-mounted drives? 
31 Jan
31 Jan
3:16 a.m.
New subject: Mounting hard disks at strange angles (was RE: OT: the 1U system)
I haven't had any problems hanging them with CB up, down, or the drive
(whole computer) on its side. My biggest drive issues were power spikes,
jolts to a running system, and heat expansion/overheating. 15 years
experience has gotten me out of those scrapes...
Ed
San Antonio, Tx, USA
it vertically could cause read errors.
Any truth to
this?
I remember that was a commonly accepted concern back in the days when
we used to have to do our own low-level formatting (ST506/ST412 drives,
primarily). I think since modern drives have things packed so tightly
that the thermal expansion and contraction of the platters causes the
tracks to perceptibly migrate, causing drives to be *much* smarter
about positioning than they used to, might have made that concern
(if there ever really was one) obsolete.
I've heard cautions for more modern drives that they can be mounted
any way except one, typically either nose down or circuit board up,
presumably due to bearing loading issues. The Amiga 4000 had the
drives routinely mounted "upside down" with no concerns, but I
would think that one would *never* want to do that to, say, an
RK05.
Anyone have any personal experiences with oddly-mounted drives?
-ethan 
30 Jan
30 Jan
6:33 a.m.
Yes, but that was with older MFM drives. The latest generation of drives to
not suffer from that disease. (Newer drives position the head with servo
tracks; the older MFM drives positioned the head mechically which could
change depending upon how the drive was mounted.)
Douglas Wood
Software Engineer
dbwood(a)kc.rr.com
ICQ#: 143841506
Home of the EPICIS Development System for the PIC
http://epicis.piclist.com
----- Original Message -----
From: "Feldman, Robert" <Robert_Feldman(a)jdedwards.com>
To: <cctalk(a)classiccmp.org>
Sent: Thursday, January 30, 2003 10:56 AM
Subject: RE: OT: the 1U system
Folklore for PC's was that if you recorded data on
a hard drive with the
platters horizontal, remounting it vertically could cause read errors. Any
truth to this?
-----Original Message-----
From: John Allain [mailto:allain@panix.com]
Sent: Thursday, January 30, 2003 10:32 AM
To: cctalk(a)classiccmp.org
Subject: Re: OT: the 1U system
<snip>
I wonder if disk drives fail at the same rates horizontally
as they do vertically (an open question) ?
John A.
8:37 a.m.
Folklore for PC's was that if you recorded data on
a hard drive with the
platters horizontal, remounting it vertically could cause read errors. Any
truth to this?
It might be true for the old hard drives that used stepper motor
positioners, or some other system where the positioning was 'absolute'
rather than taking a position reference (servo) signal off the platters.
With those drives, the force of gravity on the head assembly might move
it slightly off-track.
Servo-tracked drives position the head by reading signals off the
platters, and will correct automatically for gravity, etc.
-tony
9:02 a.m.
Tony gets a cigar!
Yes. It is true that hard disk makers didn't warrant drives that were
installed in funny positions such as on their face's or corners or
whatnot. These days though it seems you could mount one on any angle
you like. I've seen lots of machines with drives mounted with their
'faces' down.
Back in the day, drives relied on a passive stepper scheme for
alignment. These days, the stepper is replaced by a voice coil similar
to a speaker's cone. The coil throws the arm around and the arm's
pickup(s) 'sniff' at pre-coded positioning information on the disks to
fine-tune the coil's final location.
I think the reason for the switch to voice coils+wedge-servo encoded
positioning was a function of cost. It takes some local intelligence on
the part of the hard disk to do all that and in the early days it was
just too expensive.
One of the first drives I'm aware of that used this method was the
popular server drive by Seagate. The ST4096. It differs from today's
drive by having dedicated a whole platter to the wedge positioning
information. Now they use something called 'embedded servo' where the
positioning 'map' is factory encoded on the same platter(s) that user
data is stored.
Regards,
Jeff
-----Original Message-----
From: cctalk-admin(a)classiccmp.org [mailto:cctalk-admin@classiccmp.org]
On Behalf Of Tony Duell
Sent: Thursday, January 30, 2003 2:24 PM
To: cctalk(a)classiccmp.org
Subject: Re: OT: the 1U system
Folklore for PC's was that if you recorded data on
a hard drive with
the
platters horizontal, remounting it vertically could
cause read errors.
Any
truth to this?
It might be true for the old hard drives that used stepper motor
positioners, or some other system where the positioning was 'absolute'
rather than taking a position reference (servo) signal off the platters.
With those drives, the force of gravity on the head assembly might move
it slightly off-track.
Servo-tracked drives position the head by reading signals off the
platters, and will correct automatically for gravity, etc.
-tony
12:55 p.m.
Tony gets a cigar!
Sorry, don't smoke :-)
Yes. It is true that hard disk makers didn't warrant drives that were
installed in funny positions such as on their face's or corners or
Some manufacturers specifically stated that some mounting positions were
not allowed, even if you formatted the drive in that position (it was
often things like panel-down). I assume this was either a cooling issue,
or the fact that the servo amplifier (if servo tracked) would have to
work harder to keep the head on-track in that position.
One of the first drives I'm aware of that used
this method was the
popular server drive by Seagate. The ST4096. It differs from today's
There were older ones, I suspect, if you go to larger platters. Many of
the 14" 'washing machines', and things like the RL01 and RK07 were
servo-tracked. The Micropolis 1200 series (8" Winchesters) were also
servo tracked.
drive by having dedicated a whole platter to the wedge
positioning
information. Now they use something called 'embedded servo' where the
I beleive one of the first embedded servo drives was the RL01.
-tony
7843
days inactive
7844
days old
8 comments
8 participants
participants (8)
-
ard＠p850ug1.demon.co.uk -
dbwood＠kc.rr.com -
erd_6502＠yahoo.com -
ETILLMAN＠satx.rr.com -
jpl15＠panix.com -
Robert_Feldman＠jdedwards.com -
stanb＠dial.pipex.com -
Technoid＠30below.com