Are hard drives supposed to go 'boing'?

Posted by: PaulWay

Are hard drives supposed to go 'boing'? - 27/04/2005 00:56

Hi there!

I accidentally dropped my little 30GB portable drive the other day getting out of the car. I was slightly worried but remembered the reassurances of the manufacturers specs that these things could take 500Gs when unpowered and survive. Then I plugged it in and, instead of making the usual 'start up the spindle' whirring sound it made a 'where are my heads' click click sound. No drive appeared on the USB chain. I tried a few different combinations and another machine, but nothing worked - it looks like I've adgered my drive.

It's really only the 100KB or so of source code on the drive that I'd like back; the 20GB of mixes I already have elsewhere. A pity, then, that I hadn't backed it up at all recently.

I looked around for replacement drives and, to my surprise, found exactly the same model of drive still on sale. Hmmm, I thought; maybe I could put the old platter in the new drive and I could get the data off. Of course, the risk of this is that I somehow adger the new mechanism or new platter and therefore end up having paid for two dead drives rather than one. I don't like throwing good money after bad for no result.

So my question is: has anyone else tried this kind of operation? What kind of success have you had from it? Any good pages you'd recommend on how to do it? I've tried searching Google but I mostly get stories about replacing your hard disk in portable players and laptops.

Preparing to write off my work,

Paul
Posted by: lectric

Re: Are hard drives supposed to go 'boing'? - 27/04/2005 01:36

I have had success in the past simply removing the cover, plugging in the drive, and when powing the computer on, spinning the platters by hand to give them a jump start. Once they started to spin, they kept on doing so. I then copied everything off immediately to a new drive and all was well. Keep in mind, this was important data and there was NO other option. If nothing else, try spinning them by hand before you try swapping platters. Just take care not to touch the platters themselves. Only touch the spindle holding them in place. Also, be prepared for the fact that there may be nothing you can do to save your data. In my case, when I powered the drive on with it in my hand, I could feel the drive shudder, but not spin up, so I had a pretty good feeling that a jump start would do the trick.
Posted by: SE_Sport_Driver

Re: Are hard drives supposed to go 'boing'? - 27/04/2005 02:21

Quote:
I have had success in the past simply removing the cover, plugging in the drive, and when powing the computer on, spinning the platters by hand to give them a jump start. Once they started to spin, they kept on doing so. I then copied everything off immediately to a new drive and all was well. Keep in mind, this was important data and there was NO other option.


I actually did the same thing with my notebook drive and it worked. But I found that as the drive heated up, it would freeze again. Got to be pretty painful, but I just repeated the process to get the data.

Since I had to do this a few times, I learned that a good "wack" on a hard surface, just as the drive was trying to power up worked too. I've heard this from many people as well.

However, what I learned was that if I waited to plug in the drive just as the computer was looking for it (during bootup) rather than have it plugged in prior, the PC would detect it without me giving it a good wack. The initial spinning of the platters manually may have helped loosen the bearing (which was my problem I think).

When you open the cover, the "axel" of the thing will most lkely have a star shapped slot on it. Have a screw driver of that type handy and hold your breath during the whole process. It'll encourage you to hurry up!

Good luck.
Posted by: Shonky

Re: Are hard drives supposed to go 'boing'? - 27/04/2005 07:08

Sounds like it's already spinning so I don't think the manually spinning method will help.

You seem to be suggesting swapping control boards on the drive. I seriously doubt the control board has failed. It's more likely to be the stepper motor for the heads. If it's the heads, you might have a chance, but I seriously doubt you could pull it off.

I have seen the control board replacement work when the control board itself has crapped itself.

As the others say though, if you are absolutely desperate, you'll be willing to try anything.

Oh and please explain "adger" to a fellow Australian (albeit from Brisbane)...
Posted by: mlord

Re: Are hard drives supposed to go 'boing'? - 27/04/2005 19:07

Quote:
reassurances of the manufacturers specs that these things could take 500Gs when unpowered and survive.


They probably can handled that. But going from ANY speed to ZERO instantly is much more than 500Gs. Which is pretty much what hitting the concrete does.

Cheers
Posted by: gbeer

Re: Are hard drives supposed to go 'boing'? - 28/04/2005 03:54

Quote:
It's more likely to be the stepper motor for the heads.


Do current generation drives have stepper motors for the heads any more? All them I've torn apart, or seen pictures of, seem to have a single energized coil suspended between two wafer like magnets. Doesn't seem to be much inside of current drives that can be damaged any more.
Posted by: bonzi

Re: Are hard drives supposed to go 'boing'? - 28/04/2005 11:47

Yeah, in order to keep deceleration to 500g after free fall from 1m, one would need the travel (provided by a well cushioned bag or such) of some 4mm, if I calculated correctly. Naked drive hitting bare concrete will experience several thousand g.
Posted by: jmwking

Re: Are hard drives supposed to go 'boing'? - 28/04/2005 11:56

As part of my freshman chem lab, they taught us how to put a foot under expensive glassware to keep it from hitting the floor full speed. And also when not to, of course - some of that stuff can be messy on your shoes, or remove your shoes for you. Permanently.

20... No, 25 years later (man, I'm gettin' old), I still either throw my foot under or jump way back when something is falling from the kitchen counter or the workbench.

-jk
Posted by: JBjorgen

Re: Are hard drives supposed to go 'boing'? - 28/04/2005 15:15

When I was in high school I worked in a grocery store stocking the dairy section. You'd be amazed at how fast and skilled you can get at catching yogurt containers with your feet. I still attribute working there to increasing my soccer/football skills.
Posted by: tanstaafl.

Re: Are hard drives supposed to go 'boing'? - 29/04/2005 00:28

some 4mm, if I calculated correctly

Oh, Dragi, you knew I couldn't resist the challenge, didn't you.

Let's see... V^2 = 2 x A x D where A = 980 cm/sec/sec and D = 100 cm

That gives... 442.72 cm/sec velocity after a drop of 1 meter.

Now, let's solve it the other way:

D = V^2 / 2A, where V = 442.72 cm/sec, and A = (980 cm/sec/sec) x 500. (that's your 500 Gs deceleration)

I get (442.72 x 442.72) / (2 x 980 x 500), or 196,000 / 980,000 and once all the units cancel out that comes to .2 CM, or 2 mm.

I'm guessing you dropped the "2" in the divide by 2A step?

Anyway, 2mm of deceleration would be enough to cushion it for a 500G stop. Calculating the actual G-load of a drop onto concrete is problematic, since the only "cushion" as such is provided by the deformation of the case and it's internal parts, and of course it is the deformation of those internal parts that does the damage to the drive.

Yeah, I know -- I should get a life!

tanstaafl.
Posted by: canuckInOR

Re: Are hard drives supposed to go 'boing'? - 29/04/2005 04:13

Quote:
Calculating the actual G-load of a drop onto concrete is problematic, since the only "cushion" as such is provided by the deformation of the case and it's internal parts,

<first year physics prof>
You may assume it lands on a frictionless surface, and that the deformation of the case is zero.
</first year physics prof>
Posted by: Roger

Re: Are hard drives supposed to go 'boing'? - 29/04/2005 07:02

Quote:
<first year physics prof>
You may assume it lands on a frictionless surface, and that the deformation of the case is zero.
</first year physics prof>


Imagine a perfectly spherical hard disk under simple harmonic motion...
Posted by: mlord

Re: Are hard drives supposed to go 'boing'? - 29/04/2005 12:07

Mm... 2mm sounds more like the likely ballpark. After I killed two (*TWO*) IBM (IoMega) 1GB Microdrives by dropping them 3" (three inches) on to a wooden tabletop on separate occasions, they finally modified the design to incorporate inset rubbery blocks in the corners of the case.

I imagine the rubber might deform as much as 1mm or so on impact, which would absorb a lot of G force.

But I've gone solid-state for embedded drives now. No looking back.

Cheers
Posted by: bonzi

Re: Are hard drives supposed to go 'boing'? - 01/05/2005 16:06

Quote:
I'm guessing you dropped the "2" in the divide by 2A step?

Exactly. Things like that kept happening in my student days - usually all problems solved, but a little "2" or "-" or "+/-1" got lost somewhere along the way

Quote:
Anyway, 2mm of deceleration would be enough to cushion it for a 500G stop. Calculating the actual G-load of a drop onto concrete is problematic, since the only "cushion" as such is provided by the deformation of the case and it's internal parts, and of course it is the deformation of those internal parts that does the damage to the drive.

I know some sidewalks here where crumbling of the concrete alone would provide those 2mm of travel...

Quote:
Yeah, I know -- I should get a life!

Me too
Posted by: PaulWay

Re: Are hard drives supposed to go 'boing'? - 03/05/2005 00:56

That explains the nice rubbery surround on most new Seagate hard disks...

Disk has now been sent to a firm in Sydney who specialises in this kind of repair. $64 and one week will buy me a prognosis on what it will cost to get some of the data off the drive. We shall see.

If I could stress this more, I would: MAKE BACKUPS! These days, with a host of different remote copy software available - use rsync, even! - secure remote access and relatively big pipes, there's no excuse to not find a friend and have a backup-swapping arrangement. Put it in your crontab as a regular job and you'll never have to worry again. I hope I'm preaching to the converted here...

Have fun,

Paul