Get inside your hard drive; how it works and how it is designed to fail

Introduction
A hard drive is a mechanical device inside of your personal computer designed to store and retrieve information; you sometimes hear it spin to life when you turn on your computer. It might be easier to think of a hard drive as long-term memory, storing all of the important applications and files that make your computer such a handy device. For instance, Windows is generally stored on a hard drive and so is Word, or any of the new games you like to play. When the drive malfunctions, you are unable to start Windows, or open any of your applications or files.

Imagine your long-term memory being wiped out, not remembering your family or friends. Imagine not knowing where you are from, or where you live. When a hard drive fails it can lose bits of files or whole patches data – such as entire applications, or an operating system. Worse yet, it can have a meltdown and you can lose everything! I hope you start to get a feel for the severity of the problem; a failing hard drive is a very bad thing, indeed.

It is a matter of fact that any mechanical item is bound to fail, regardless of how well it is built or who it is built by. Gears grind down, time takes its toll, and in the end, our drive is left inert and ineffective. This problem plagues modern industry; to the extent that companies spend billions each year trying to improve upon designs that give their products increased durability, and ultimately, lengthier longevity.

This brings us to our topic at hand; hard drives as a storage medium. A Hard Disk Drive is a mechanical device that consists of a series of high speed moving parts, with tolerances measured in microns that we, as end-users, expect to last forever. In this article, we dive right into what makes up a hard drive, how it works, and how you will ALWAYS lose your data if you are not careful; the risk factor of losing your data is always 100%. However, with the steps in this article, you can gain the knowledge needed to protect your data so that when your hard drive does fail, you are not left up a creek without a paddle.

The Mechanics
I was going to explain the history of a hard drive, but honestly… you can read about that elsewhere. However, the mechanics behind the drive and its fundamental parts are important to us for the purpose of this article. A hard drive consists of one or more hard, circular, disks that are driven by a high-speed rotary motor; hence the name Hard Disk Drive.

Each of the hard disks inside of the drive is referred to as a platter. Each platter is usually made from glass or aluminum, but in all cases, the material is non-magnetic. The glass or aluminum is then brought into spec using microscopic tolerances and the finished circular disk is coated with a thin layer of magnetic material (usually iron, but more recently a cobalt-based alloy). The end result is a mirror-like finish with a nearly perfect magnetic surface. The magnetic properties of the disk are pivotal to the storage of data. With this technology, a hard drive stores data in a similar way as our old cassette tapes with magnetic spools of tape; hard drives are merely rotary instead of linear to minimize physical size and maximize storage space. That last blurb was probably useless without going into the technology behind cassette tapes – you’ll have to trust me.

Inside of the hard drive platters are spun at high speeds and data is written magnetically by alternating the magnetic charge of the platter. The read-and-write head rests nanometers above the magnetic surface of the platter and has the ability to magnetically charge and detect magnetic areas of the platter as they spin past it. The individual magnetic areas are known as sectors and binary data is stored in 1’s and 0’s – which at its roots is as simple as on / off, or charged / not charged.

Now we’ll take a short aside: When you run your disk diagnostic utilities, you might encounter a “bad sector”. This simply means that the magnetic properties of that particular section of the platter are flawed. Being out of spec, that portion can no longer be accessed by the drive. Consider this problem non-repairable. Fortunately, newer drives remember which sectors are bad, and continue to function just fine ignoring the bad sectors. After all, a bad sector is only 1 unit of binary data. To put this in perspective, one bad sector is exactly 1/8th of an alpha letter (A, B, C … Z) that you can no longer store; not really a huge loss when you think about it. However, enough strategically placed bad sectors can tank your whole system.

What really gives the hard drive its power is the ability to read different tracks along the horizontal axis through the use of a motorized arm attached to the read-and-write head. This allows the drive to move around the platter in a grid-like fashion, entering or recording data the entire time. This is truly where the hard drive surpasses the limitations brought on by conventional linear cassette tape. However, due to the high speeds and constant movement, the hard drive is simply doomed to fail; to quote the largely popular movie, Armageddon, “It is not a matter of if, it is a matter of when”.

Common Myths:

Enterprise drives Last longer.
FALSE – they are just as likely to die as normal, consumer, drives. In fact, some higher RPM enterprise drives are more apt to fail than slower consumer drives.

SMART Says my drive is ok.
FALSE – SMART drive error reporting does not always report errors, but when it does – it usually means that its time for a new drive.

Extra fans will save me.
FALSE – Heat does not overly affect the drive unless you are beyond 104F, at which point it still has little impact. Granted, if you are running a 10k Raptor at 200F, you’re going to burn up the oil and fry the drive. Generally speaking – heat has little to do with the issue.

The less I use the drive, the longer it will last.
FALSE – You can use your drive 1 day a year, or all day long for an entire year. The chance it will die is still random and no studies show that drives are less likely to die if they go unused.

Common Killers:

Dropping a Drive:
Dropping your drive, or even the computer is a bad idea. The tolerances between the head and platter are within nanometers, so if you cause the head to make contact with the platter by dropping the drive or shaking it, you’re bound to get some permanent damage. Reducing drive handling can improve reliability by 20% or more.

Voltage Spikes:
Voltage can ruin the controller board on the hard drive, or damage the rotary motor inside the drive. A spike can be from your wall outlet, or from static discharge as you handle the drive – either way, both are bad.

Cooking a Drive:
Temperatures that are so hot, you can cook an egg on top of the drive are bad. If you can cook an egg, chances are – you are cooking the drive and the oil that lubricates all of the moving parts.

Time:
The one thing sure to kill every hard drive out there is time. No matter how you baby a drive, time will always take its toll and the drive will fail. In fact, drives 3 years or older are five times more likely to fail. This does not mean your 3-year-old drive will fail, it just means that the chances are higher.

Environmental Variables:
Condensation, desert sand, flooding, etc. While these may or may not cause the drive to fail directly, they can often complicate the issue by putting more stress on other components, which in turn cause the hard drive to die.

What to do when the drive dies:

In the event that your hard drive renders itself inert and useless, I hope you have your information suitably backed up. However, if you are like me, and have “almost” just in time inventory, your drive is dead and the backup drive is in the mail – In case you did not notice, this is NOT an optimal solution… so we’ll cover a few ways to try and get the information from the drive.

Freeze the drive:
If the drive was wet, or stuck, sometimes a night in the freezer can shrink up and loosen the stuck parts. The drive will quickly start forming condensation, so get your information off it as fast as possible. Your drive might be stuck if you hear a grinding sound when you turn it on.

Replace the PCB:
If the primary controller board is dead, the drive will usually make no sound at all. You can likely replace it with a controller board from a donor drive of the exact same make and model. Make sure the firmware on the drive matches. You can usually verify the firmware using the sticker on the top of the drive.

Replace the Motor:
This is a sure fire way to kill your drive, but it might give you a chance to get your important information off of the drive. Replacing the servomotor or rotary motor requires a donor drive. It also involves opening the top of the drive and exposing it to dust. To really complicate the issue, it even involves taking the platters off of the drive and risking scratching them. It is a really complex procedure, and we can walk you through it, so you’re best off starting a thread in the forums and we’ll help you out.

Replace the Head:
Another sure fire way to kill your drive. The heads are the most difficult part to hardware level repair on a hard drive. The reason being is that the tolerances are so tight, that you often touch the platter so saving your data, might also mean corrupting some of it. Hey, it’s for the greater good, right?

Software tools for recovery:

LifeBoat by Tugboat Enterprises. Look for a full review soon!! I was able to browse my root level of a hardware damaged drive (bad motor). This was more than anything else granted. Like I said, look for the review soon!

BartPE / Windows Preinstall environments. These are useful CD’s and utilities for getting into a computer with a drive that has damaged sectors, as you are booting from a bootable CD.