Hard drive - crashes?

slash000

A hard drive "crash" generally refers to when the read/write head physically bumps, or crashes, into the platter above which it rotates.


Question: How susceptible are laptop hard drives to crashing in this manner?


I ask because a friend of mine told me that if you carry a laptop around, even if carefully, you are slowly causing the read/write head to make slight bumps onto the platter and causing regular damage to the drive.

I told him that's a load of bull, and that laptop drives are designed with a factor of safety in mind, and constructed in such a manner that every-day use and casual "carrying around" the laptop , even while turned on, this is not likely to happen - and even if we assume he's right, then stuff like oldschool MP3 players (that used to use laptop HDDs in them) which were carried around and thrown about and used in vehicles all the time, surely these would have been likely to crash, but cases of such were rare.



So what's the verdict on this?

Does a hard drive "crash" if you "carry it around" in a laptop too much, whether it causes a crash then and there, or does carrying it around cause it to have a slow, progressive "crashing" of the r/w head into the platter?

Or is it that laptop HDDs are durable enough to withstand moderate (but not excessive) forces of being "carried around?"



I don't buy his argument. Anyone know the answer and have a good factual basis for their reasoning?

Moe Fwacky

Okay, I'm no expert on hard drive innards, but I do work with some of the most abused laptops on the planet. These are basically the rental cars of laptops. I work on campus at a place that loans out laptops to students. We have about 200 laptops, and do anywhere from 300-500 loans per day. These things are dropped, banged into walls, opened and closed hundreds upon hundreds of times. Granted, these machines are retired after three years of service, but the hard drives hold up pretty well. Every so often we get a bad sector, but running a hard drive recovery utility found on Hiren's Boot CD usually does the trick.

In other words, he's full of shit. The most abused laptops in the world only suffer minor hard drive issues. Laptop hard drives are made to withstand greater G-force than 3.5 inch drives. Obviously, you'll incur less potential damage if the machine is off, though.

Ego

Ok, first, hard drives spin fast enough that if you get a head crash, you're probably going to know it. Tell your friend there's no such thing as a slight bump onto the platter or 'regular' damage to the drive from a head crash --at a minimum velocity of 5400 RPM, a head crash will instantly cause enough friction to strip an entire ring of the platter (so scandisk will find a LOT of errors) and usually generate enough heat that the drive won't even work until it cools off. Head crashes are rare, if all the hard drives I've taken apart (to salvage the fun and remarkably powerful magnets) are any indication.

Now then...

When turned off, a hard drive parks the heads, so you have basically no worry at all about moving a laptop around in this state at all.

Laptop drives are built to higher g-force tolerances than desktop drives as you surmised, and if you move them around carefully and set them down gently when they're powered on you'll be fine. Walking a powered-on laptop from one room to another isn't going to make the head slam into the platter. Some laptops (apples for example) have sensors to park the head in case of sudden motion. Letting your laptop thump down a couple inch onto a table? That's the sort of thing that might make the head smack the platter, despite improved g-tolerances. But I somehow doubt that's how you treat your laptop.

What you don't want to do is change the orientation of your laptop outside of the plane in which the hard drive is currently spinning. In other words spinning around a laptop that's sitting on a table is probably okay, but closing the lid and turning it over and over again (as if you were flipping a sheet of paper over to see what's on both sides) isn't. Even the relatively small drives in a laptop spin fast enough and have enough mass to have a very noticeable gyroscopic effect (a fancy way to say: inertia.) If you move a spinning hard drive along the plane in which the platter is already moving, the inertia is irrelevant and there's no gyroscopic effect. If you start turning the drive over, the gyroscopic effect puts a great deal of strain on the drive. At best your fluid bearings keep the platter away from everything and your drive heats up a bit. With greater movement the physical bearings come into play to stop the platter from slamming into it's enclosure or the drive head and your drive heats up a lot and gets a bunch of instant wear and tear. More movement still could certainly cause a head crash, although (not that I recommend trying this) I once tried to make it happen with a desktop drive and gave up on the idea when instead it just became uncomfortably hot.

Anyhow. If you treat a laptop gently when you move it, you're fine. Don't pick up your laptop while the drive's spinning, turn it vertical really fast and stick it into a case. Don't run with the damned thing. Don't thump it down. Sounds like you're already careful, and your friend is paranoid.

Just a quick note though, while I'm unfamiliar with most hard-drive based mp3 players, the ipod definitely would load songs from the hard drive into memory, then spin down the drive. I think this was mostly done for battery life (because really song tracks would suck ipods dry from the continual drive spin, and it was a reason they didn't have gapless playback) but no doubt it also improves survivability when people take their ipods jogging.

Sorry the post was so long, hope I got all your questions.

1ddqd

Think of it this way. If you spin a top really fast, then put your finger up next to it, when the top hits your finger it shoots off in a random direction. That's because it wasn't anchored to the ground. As you can see on this page, the platters are bolted to a spindle. There is about 2-3 mm between each one.

Now, if you drop a laptop from 5 feet up, it WILL generate some HDD errors. The platters *may* touch each other. The main problem is the arm - it is more likely to damage the platters than platters are. Laptop drives (most of them) now have inertia sensors - sort of like an accelerometer. If it detects that it is rapidly moving (up, down, left right) the arm is immediately parked. So yes, you're correct, and this is why.

slash000

Thanks for the info, guys. I actually appreciate the long and detailed responses, because I wanted to know the underlying reasons.

I was pretty sure that this was the case, but I just wanted to get some factual support, and so that I could understand it myself.