Hard Disk Tilt?

[Decker87]

Does it matter what position the hard drive is in? What precautions, if any, need to be taken if you are to mount a hard drive at an angle such as 35 or 45 degrees?

[defakto]

Do not, I repeat, DO NOT mount a harddrive anything other than horizontal or vertical orientation(upside down is ok also). They are not designed to run at an angle. This debate came up here before and was resolved out with a phone call to the harddrive manufacturer to verify anything other than above mentioned can significantly increase chance of drive failure.

[mikeblas]

How close to perfectly horizontal or vertical must the drive be mounted to be safe?

[Laforge]

My latin's a little rough.. does your sig say "Who's watching the watchers"?

[ChingChang]

How close to perfectly horizontal or vertical must the drive be mounted to be safe?
I can't remember exactly, but I thought it was like 10-15% is safe.

I think I read it on some hard drive manufacturer website... maybe WD.

[mikeblas]

And what did the drive vendor say about lap-top drives? Or other small-size portable drives, like in MP3 players, or in PCMCIA- or CF-size cards?

[ChingChang]

I think it was only for the desktop 3.5" HDD. Not sure about the other drives, maybe it's not as important because they are lighter and spin slower usually.

I'll try to find the link tomorrow. I think it was Seagate or Western Digital. Pretty sure it wasn't on Maxtor's site. I might have a bookmark, but I'll have to dig through my collection :)

[Decker87]

What about upside-down?

Also, what side of the hard drive produces the most heat?

[unhappy_mage]

If orientation is such an issue, then why don't drive (http://www.seagate.com/cda/products/discsales/enterprise/tech/1,1084,698,00.html) manufacturers (http://westerndigital.com/en/products/Products.asp?DriveID=133&Language=en#jump88) mention (http://maxtor.com/_files/maxtor/en_us/documentation/data_sheets/maxline_iii_data_sheet.pdf) this (http://samsung.com/Products/HardDiskDrive/SpinPointTSeries/HardDiskDrive_SpinPointTSeries_HD300LJ.asp?page=Specificatio ns) on (http://www.hitachigst.com/hdd/support/7k500/7k500.htm) their product pages? According to this page (http://www.pcguide.com/ref/hdd/op/packOrientation-c.html), it shouldn't matter what orientation the drives are used in, but it doesn't specifically address 45 degree angles etc.

As for most heat, the base of the drive is designed as a heat sink. Most of the heat ends up there.

http://www.hardfolding.com/ftag1.php/mem/150072 (http://www.hardfolding.com?go=38&id=150072&tm=33)

[mikeblas]

This old Seagate manual (http://www.seagate.com/support/disc/manuals/ata/3655pma.pdf) (for the ST3655 family, not bigger than 500 megs back in 1993 or so) says the drive can be mounted in any orientation. It's hard to tell if that means any of six orientations, or any arbitrary rotation.

To answer my own question, the Travelstar drives want to be nearly level (http://www.hitachigst.com/hdd/support/e7k100/e7k100mec.htm) which is surprising because my laptop is never, ever level.

[defakto]

What about upside-down?

Also, what side of the hard drive produces the most heat?


As long as the heads are moving mostly perpendicular or parrellel to the force of gravity you should be fine.

If orientation is such an issue, then why don't drive manufacturers mention this on their product pages? According to this page, it shouldn't matter what orientation the drives are used in, but it doesn't specifically address 45 degree angles etc.

As for most heat, the base of the drive is designed as a heat sink. Most of the heat ends up there.

Have you ever seen a production computer case that mounts a drive at an angle other than zero or ninety degrees, ever? Not only that if you just consider the physics behind a drive head moving, it makes sense that large angles are bad because as the head accelerates to read from seperate portions of the drives its force perpendicular to gravity could overcome the force that the cushion of air, that the heads rest on, causing the heads to touch the drive. After a bit of research most drives are designed to handle something like 50,000 head touches(they have a little slider that prevents damage but will wear out because head touches were/are a part of a drive starting up) before it becomes an issue. If the drive is mounted flat rightside up/down or on it's end/side the heads will move in such a way that the force of them moving will never exceed the force of the air cushion during normal operation.

This old Seagate manual (for the ST3655 family, not bigger than 500 megs back in 1993 or so) says the drive can be mounted in any orientation. It's hard to tell if that means any of six orientations, or any arbitrary rotation.
If you look at page 31 of that manual it gives the recommended orientations for mounting the drive. It's doesn't explicity say that angular mountings are out, but it is implied by the diagrams they provide.

[unhappy_mage]

Have you ever seen a production computer case that mounts a drive at an angle other than zero or ninety degrees, ever?
Have you ever seen a production computer case that mounts a CPU or a video card at an angle other than 0 or pi/2 radians, ever? Just because it isn't done doesn't mean it can't be.
Not only that if you just consider the physics behind a drive head moving, it makes sense that large angles are bad because as the head accelerates to read from seperate portions of the drives its force perpendicular to gravity could overcome the force that the cushion of air, that the heads rest on, causing the heads to touch the drive.
The head on a new drive moves full stroke in 17ms. That's about 3cm. So it has to speed up, hit max speed, and slow down in that time. Even if we assume it hits max velocity only instantaneously (it's always accelerating, IOW), we get:
.03=.5a*.017^2
a=.06/.017^2
a=207m/s^2 about 20 gravities for the acceleration of the disk head. Suppose (the worst case scenario) we've mounted our drive at pi/4 radians. Then we get this diagram for the acceleration of the head:\200m/s^2
\
\
\
\|10m/s^2So the magnitude of the acceleration in the direction perpendicular to the plane of the disk is about g/sqrt(2). The disk heads can handle about 60G's worth while reading or writing. It's certainly an interesting point, but I'm not sure how much influence gravity would have on the head. Objections? Did I get the math wrong?

http://www.hardfolding.com/ftag1.php/mem/150072 (http://www.hardfolding.com?go=38&id=150072&tm=33)

[defakto]

Considering even something as small as bumping your case while the system is running can cause problems, and the fact that the average human fingerprint, or particle of smoke, is tall enough to cause a head crash, I don't think gravity can be ignored in that situation because even a tiny fraction of an inch wiggle in the head could be potentially fatal.

However, when all other arguments fail to work, I recommend just calling any drive manufacturer and asking if it is ok to mount the drive at a 45 degree angle, instead of flat, or on end and see what their tech support says.

Your math looks good though, I've never been 'A' grade at physics, I know no force should be ignored in a situation if it can cause problems.

[Lazn_Work]

Have you ever seen a production computer case that mounts a CPU or a video card at an angle other than 0 or pi/2 radians, ever? Just because it isn't done doesn't mean it can't be.

The head on a new drive moves full stroke in 17ms. That's about 3cm. So it has to speed up, hit max speed, and slow down in that time. Even if we assume it hits max velocity only instantaneously (it's always accelerating, IOW), we get:
.03=.5a*.017^2
a=.06/.017^2
a=207m/s^2 about 20 gravities for the acceleration of the disk head. Suppose (the worst case scenario) we've mounted our drive at pi/4 radians. Then we get this diagram for the acceleration of the head:\200m/s^2
\
\
\
\|10m/s^2So the magnitude of the acceleration in the direction perpendicular to the plane of the disk is about g/sqrt(2). The disk heads can handle about 60G's worth while reading or writing. It's certainly an interesting point, but I'm not sure how much influence gravity would have on the head. Objections? Did I get the math wrong?

http://www.hardfolding.com/ftag1.php/mem/150072 (http://www.hardfolding.com?go=38&id=150072&tm=33)

One question: at that angle, if you bump the drive/computer, is it more likely to headcrash with a 20G disadvantage to start, in a direction that it is not designed to brace against?

==>Lazn

[ekard]

Been reading these forums forever, but finally figured I would register to clear up some issues in here.

First, orientation does not have a significant impact on seek performance. The actuator is designed to be almost perfectly balanced about the pivot, meaning gravity has little to no effect. Gravity alone will never cause a head crash. Gravity pulling the drive down from a couple feet after it's slipped from your hands is another thing.

At that point orientation matters when you're talking about impact. If the drive lands on its side or end, you might be okay. The shock will cause the actuator to rotate, but the drive should notice and recover. Its when you land or shock the drive on the bottom or top that the forces try to push the heads into the disk. Not good. As always, notebook harddrives are built to take *some* shock; desktop drives not so much.

During normal use, orientation can affect the spindle motor however. Most hard drives these days use fluid filled motors to keep them quiet. They've all been tested and proven at various orientations, but it's the only component I've heard concern about when the drive is not parallel to the ground. Not all companies test the same though, so follow your manufacturers suggestions.

[mikeblas]

Older drives tie their error rates to the orientation of the drive.

[davidlem]

Simply don't do it. The larger the platters, the more rotational mass, the higher spin rates - don't give your desktop drive the opportunity to self-destruct.

On a side note...on more than one occasion I've had to free a burr from the motor of a particular manufacturer's seized 3.5" desktop drive. After performing this operation the drive will only spin freely if it is face down. During the recovery process it slowed to a crawl and I had to monitor its every movement. I was able to help it along by placing a 1cm tall object under one side of the case - the speed nearly tripled. It was just by coincidence that I discovered this as it is not documented in any of our methods. To be sure it was effective, I removed the object and saw an instant reduction of throughput. So being tilted just slightly helped this failing motor assembly stay up to speed. YMMV, but the angle made a difference in that failing drive - who's to say it wouldn't effect a good drive in the opposite manner. They're like hand-held gyroscopes, particularly the Maxtor D740X/D540X series for whatever reason.

[ketamine]

I've built a few Car-puters (http://www.mp3car.com/vbulletin/showthread.php?t=32035&highlight=Jetta) in my day and have mounted hard disks in vehicles in every which orientation. I've never had any problems directly related to hard disk failure with any installations. Keep in mind these are daily driving vehicles on roads so they also have to stand up to a lot of vibration. If you visit the site I linked you'll find that many others have had success with multiple orientations.