Hard Drives: Recovering Data From A Troubled Drive

When it comes to a computer's internal organs, few are as vital to the system as the hard drive, and none is as valuable to the user. Sure, you might not be able to use your computer temporarily if the processor goes kaput, but after a quick transplant, it's back to business as usual.

But the problem isn't so easily solved when your hard drive dies. Lose your hard drive, and you'll lose millions of bytes of irreplaceable data. You can get a new hard drive, of course, but you still have to find a way to get by without the lost information. Talk with anyone who's ever experienced a hard drive crash, and you'll quickly learn how frustrating it is to lose a hard drive. Most hard drive problems don't have such devastating results. With a little know-how, you can remedy the small problems before they become big ones. And with a little more know-how, you can prevent the small problems from ever occurring. But we'll discs that a little later. The truth is, you can't do anything for your drive unless you know a little about it first.

Hard Drive Basics. A hard drive is a desktop or portable computer's primary means of data storage. The hard drive consists of five main components: a number of thin, rigid, circular disks, called platters, on which data is stored; the read/write heads that read data from and write data to the platters; an access arm that holds the read/write heads; a head actuator, which controls the movement of the access arm; and a spindle motor, which spins the platters at a high speed. These internal mechanisms typically are encased in a metal box that is connected to the computer's frame. A personal computer sold today probably comes equipped with a hard drive that has a capacity of four gigabytes (GB) or more. A gigabyte is a unit of data measurement equal to a little more than I billion bytes. The average hard drive a few years ago was capable of holding only a few hundred megabytes (MB; a unit of data measurement equal to a little more than I million bytes). The first personal computers didn't have hard drives.

Today's hard drives can hold substantially more data than their predecessors, but this also means that today's hard drives can cause substantially more problems to today's computer users. To avoid frustrating situations, computer users must be diligent in performing drive maintenance and must be aware of situations that could prove disastrous to a hard drive and the data stored on it. We've divided our troubleshooting tips into two categories. The first category deals with problems that occur during or immediately following the installation of a hard drive. The second category includes problems that occur anytime after that. Most of these latter problems have to do with the way data is stored on the drive. The interesting thing about troubleshooting hard drives is that it's possible to avoid almost all the things that can go wrong with them. Except for manufacturer's flaws or cases of drive abuse (shaking the hard drive, exposing the platters, putting it under an electrical magnet, etc.), the hard drive is fairly stable. Nevertheless, problems happen and things go wrong. We've come up with answers to some of the more common hard drive dilemmas.

Installation problems. It's important to point out that the majority of all hard drive problems occur during installation. If you can get the drive up and running, you'll eliminate the source of most hard drive problems.

Every computer installation is fraught with peril. We've all heard the stories of systems that worked perfectly until new components were added. Of course, in some of these cases, the new component actually was the culprit. But most of the time, problems occur because the person performing the installation didn't read and follow all the directions, says Richard Haggan, technical support supervisor at Maxtor Corp.

"They tend to jump right in there and they forget some steps, whatever they may be," says Haggan. "Then we have to backtrack." As a preventative measure, read the directions before beginning an installation. That way, you'll be prepared for what comes next. And because you'll understand the process, you'll be less likely to skip a step you think is unimportant. Of course, the installation instructions aren't always the easiest thing to understand and problems will crop up. Some of these are caused by human error. For example, if the hard drive's power cable isn't connected correctly, then the drive won't work. One way to identify a problem with the power connection is to look at the lights on the front of your computer. One of those lights is linked to the activity of the hard drive. When the hard drive is working the light flashes. If that light doesn't come on when you turn on your computer, the hard drive probably isn't getting power and you need to make sure the power cord is connected.

Another example of a problem caused by human error is an improperly connected data cable. If the hard drive indicator light flashes when you turn the computer on, yet you can't access the hard drive, then you can suspect this is the problem. This isn't the only reason for an inaccessible hard drive, but it's an easy one for you to fix. Besides, you don't want to waste your time calling technical support only to be told you didn't correctly connect the hard drive to your system.

To double-check the cable connections, remove the cover from your computer and locate the hard drive. It looks like a metal rectangular box and will be secured to the frame of the computer. The computer's users manual should tell you where it is if you can't find it.

After you locate it, see if it has two cables plugged into it One cable consists of up to four separate cords and probably has a plastic plug on one end. This is the power cable. The other cable looks like a wide ribbon. This is the data cable. Make sure the cables are connected securely to the drive. When you're certain they're connected, put the cover back on the computer and start it. If the drive works, then it looks as if you've solved the problem. But what if you've double-checked the cable connections and still can't get the drive to work? Here are many of the most common installation problems and their solutions.

Drive doesn't spin. For data to be read from or written to the drive, the platters on which the data is stored need to be spinning. As the platters spin, they make a soft humming noise. If you don't hear the humming noise and can't access the data on the drive, the first thing to do is check the power connection. If that checks out OK, then you might have a problem with the spin-up jumper.

A jumper is a tiny plastic box that has two holes in it. It slides over the tiny pins that are on the back of the hard drive. These jumpers, which let you manually configure the settings for the drive, are on/off switches for the most basic drive functions.

A spin-up jumper is a jumper that lets you control the spin-up function. The spin-up jumper, like all the jumpers on a hard drive, needs to be configured correctly, prior to installation. Every drive is different, so you'll have to check the hard drive's instruction manual to find out how the spin-up jumpers should be set. If you get confused about how to set the spin-up jumper, contact the manufacturer of the drive or computer and ask for technical assistance.

LLNO ROM BASIC." This error message appears because the primary partition on the master (primary) hard drive isn't set to active. What's a partition? It's an area on your hard drive that is treated like a storage unit. Each partition is divided into uniformly sized smaller components called sectors. Data is stored in groups of sectors, called clusters, and each duster can store a maximum of one file. For more information about the storage areas of a hard drive, see the "Out of Disk Space" section later in this article. Hard drives can be set to have more than one partition. Although this rarely happen with smaller hard drives, it happens regularly with today's multi-gigabyte drives. Each partition is treated as a separate hard drive and has its Own drive letter.

If your drive is divided into multiple partitions, one of these partitions must be designated as the primary one; this is where the operating system is stored. If the primary partition isn't set as active, then you'll receive the NO ROM BASIC error message. Setting the partition as active and solving the problem can be accomplished using FDISK, which is a DOS utility, or using the proprietary software that came with the hard drive. Which one you use will depend on which application you used to originally set the partitions. Because every manufacturer's software is unique, we'll show you how to set a partition as active using the universal PDISK. (NOTE: You'll need a bootable diskette if you plan to use FDISK. You should have created one of these when you first installed your operating system. If you don't have one, you can contact your computer's manufacturer and ask for one.)

Insert the bootable diskette into the diskette drive, then turn on the computer. When the DOS prompt appears, type fdisk and press ENTER. Select Set Active Partition. Follow the on-screen directions for setting your main partition. You'll need to know the letter (probably C) assigned to the partition. The program will set the partition active and return you to the main menu. Press ESC to exit FDISK.

Master or slave? When you add a new drive to your system, there's one way to create drive identity problems: Don't correctly configure the Ds jumper. This jumper must be set properly so your system can determine which drive is the master and which is the slave (the secondary drive).

Basically, if the DS jumper is present, the drive is designated as the master. If the DS jumper is absent, the drive is designated as the slave. Before you insert the drive, make sure the DS jumper is properly set. When they're shipped from the factory, most drives are set to be master drives. So if you're installing the drive as a master drive, you probably won't have to change the DS jumper. If you're setting the drive as a slave drive, you'll probably have to make a change.

(NOTE: Every drive is different. The only way you'll be able to tell which jumper is the OS jumper is to read the users manual. If you're nervous about messing with the jumpers, call the drive manufacturer and a technician can talk you through it.)

Dual drives, single boot.

Adding a second hard drive to your system is one of the more dangerous upgrades you can perform. Advancements in storage technology have led to the introduction of other storage options, including high-capacity portable storage devices such as the Jaz drive, that are much easier to add to your PC. Before you add a second hard drive, give one of these other drives a look. It might save you a few headaches down the road.

If you've already installed a second hard drive on your computer, you may as well make the best of the situation. This means trying to fix all the problems such a setup can create.

Here is one of those problems: You turn your computer on, and it only recognizes one of the hard drives. So you press ALT-CTRL-DELETE to restart. When it recovers, it recognizes both drives. Not a major problem, but it is an annoying inconvenience each time you use your computer. How can you avoid such frequent frustration? Reconfigure your BIOS.

The next time you turn your computer on, look for those words that grant entrance to the Basic Input/Output System (BIOS; software that controls the startup process of a computer). Usually, you'll see something such as "Press Fl to enter Setup" or "Hit F2 for Setup." Follow the command. You have to be quick or the operating system will load and you'll have to reboot and try again. When you've done it correctly, the BIOS Setup screen should appear. All you have to do now is follow our instructions, paying special attention to the Boot Options area of the BIOS. Disable both the Fast Boot Option and the Quick Power On Self Test options. Enable the Above 1MB option. If you see anything that mentions RAM Count, enable it. Set the Hard Disk Initialization Time Out option to 30. You may find all these settings in your BIOS, you may find only a few, or you may find none. Select the Exit And Save option when you're done.

The tricky part is that every system is different, so your BIOS might look different from any of the BIOSes we've seen. The BIOS is a fragile part of your system, so if our solution doesn't seem to match your BIOS, don't touch anything; you could foul up your computer. Instead, contact your computer's manufacturer and tell someone in technical support you have a BIOS timing problem. If that person can't help you, ask for the number of someone who can.

Usage problems. After repeatedly writing and erasing data from the hard drive, the platters start to get dirty. We're not talking about dust and grime; we're talking about pollution of bytes and bits.

When you write data to a hard drive, it is divided into sections (unless it's a small file, then it stays in one piece) and stored methodically on the drive. The location of the data is registered in the file allocation table (FAT; keeps a record of where data is stored on a drive) so the computer knows where it is when it wants it. Your hard drive becomes like a colossal warehouse, with thousands of crates of information stacked in neat piles. And if all you ever did was write data to a drive, these piles would always stay in the same place and you wouldn't have to worry about data integrity.

But you'll probably have to delete files from your PC, and that causes problems. The file actually isn't deleted from the drive. The reference to the file is deleted from the FAT. When new data is stored on the drive, it is recorded over the old data. And instead of storing the data in contiguous sections, it fills in the gaps left behind by deleted data.

Think of it in terms of a colossal warehouse. When you want to remove an object from the warehouse, the crate in which it's stored isn't removed. Rather, the object is removed from the case and the crate is marked as available in the warehouse foreman's register. When a new shipment of objects arrive, they don't get a new pile, or even new crate, to themselves. Instead, they're distributed among the many crates that are marked available in the warehouse foreman's register. As more objects leave, more arrive to take their places. Soon, similar items are stacked in various piles and the organization of the drive depends on the warehouse foreman's register.

When this phenomenon of scattered data happens to your hard drive, it's called fragmenting. As you add and remove data from your drive, it becomes more fragmented. As it becomes more fragmented, it takes longer for your hard drive to locate files and execute programs. Performance may be a bit sluggish. Some people may think this means the computer needs more RAM or a new hard drive. Actually, what it needs is defragmenting.

Defragmenting (or defragging as it's often called) is as easy as a few points and clicks in Windows 95 (Win95) and Windows 98 (Win98). Open the Start menu, select Programs, and choose Accessories. Select System Tools from the list of options, then click Disk Defragmenter. Up pops a Select Drive window, asking you to choose the drive you want to defrag. Select the drive. Win98 users can click the Settings button to tell Disk Defragmenter to check the drive for errors or to organize the program files for optimal performance while it defragments the drive. When you're ready to defragment the drive, click OK. If your drive is in bad shape, the Disk Defragmenter starts defragmenting the selected drive.

But if your drive isn't in bad shape, you'll see a message indicating you don't need to defragment the drive now. Because defrag-mentation takes a long time, especially if you have a high-capacity drive, we recommend you click the Exit button and wait to defrag-ment your drive until you need to do so.

If your computer runs Windows 3.x or DOS, the process is basically the same. Exit Windows so you're looking at a DOS prompt, then type defrag and press ENTER. This initializes DEFRAG, the DOS defragmenting utility. Select the drive you want to defragment from the list that appears on-screen, then press ENTER. DEFRAG then analyzes the data stored on your hard drive and suggests a defragmentation option. We recommend heeding the suggestion. If your drive doesn't need to be fragmented, DEFRAG will tell you that. In that case, select

OK, press ENTER, then select Exit, and press ENTER again to exit DEFRAG and return to the DOS prompt. When do you need to defragment your drive? Carl Salter, a senior technician with Maxtor Corp., says that depends on how you use your computer. If you install and uninstall lots of programs and files, you should defragment often, perhaps monthly, Salter says. But if you tend to use the same programs and don't add and remove files often, you're probably safe defrag-ging your drive every six months.

(NOTE: The DOS and Windows defragmenting utilities are meant to be used on uncompressed drives. Disk compression is a means of reducing the amount of disk space required to store a file; data is squeezed so it fits into a smaller space on the drive. Most drive compression utilities have their own defragmenting applications; use them.)

File Not Available & other corruption woes. Another typical problem of active hard drives is data corruption. Yes, that means data is destroyed, and that's always a bad thing. But this isn't necessarily the end of the world for your hard drive. A little data corruption is to be expected when you give your computer a vigorous workout every day.

(NOTE: We're talking about minor forms of data corruption. If the entire hard drive becomes corrupted or inaccessible, this indicates a bigger problem probably a crash. The following solutions won't remedy a drive crash.)

The three most common types of data corruption are file fragments (also called lost allocation units), cross-linked files, and bad sectors. Maxtor's Salter describes file fragments as incomplete files or file entries. Essentially, that means the computer can't find all the data associated with the particular file. File fragments most often occur when an application is closed prematurely, before it had a chance to tell the system where it stored the data it was using.

For example, if you accidentally unplug your computer while you're in the middle of typing a document in your word processor, the portion of the document that has been typed since you last activated the Save com-mand will end up as a file fragment. The second type of data corruption is cross-linked files. This means there has been a mix-up on the hard drive and the FAT incorrectly recorded the location of some data. Think of it like this: Your friends live at ~OO Main Street, but in your address book you incorrectly entered their address as 1000 Main Street. When you try to visit them, you're greeted by strangers because you knocked on the wrong door.

Similarly, when the computer looks for data that has become cross-linked, it won't be able to find the data because it won't be where the computer thinks it is. When this happens, you'll lose the data that has the incorrect FAT address. You might have cross-linked files if you can't access a file you've previously accessed. Finally, there are bad sectors. As opposed to a good sector, which is a sector where data can be stored, a bad sector is a sector where data cannot be stored.

All of these problems have the same solution: ScanDisk (or Chkdsk as it's called in early versions of DOS). This utility diagnoses and repairs minor flaws, including file fragments, cross-linked files, and bad sectors, on a computer's drives. If a repair is impossible, ScanDisk or Chkdsk finds a solution.

To run ScanDisk in Win95 or Win98, click the Start button, select Programs, and choose Accessories. Highlight System Tools and click ScanDisk. Choose a drive to scan, then click Start. If your computer runs DOS or Windows 3.x, get to a DOS prompt (exit Windows if necessary) and type scandisk. (Type chkdsk if your computer runs DOS 6.1 or earlier.) As with Disk Defragmenter and DEFRAG, you select the drive you want to scan and let the program run itself.

If ScanDisk finds file fragments, it saves the fragment as a .CHK file. You then can view the file with a word processor. Just go to File Manager (in Windows 3.1) or My Computer (in Win95 and Win98) and open the C: drive. Double-click a .CHK file, which should have a name such as FileOOOl or FileOOO2. You'll be greeted by an Open With box, asking you to select a program with which to open the .CHK file. Highlight the Notepad option, then click OK. This brings the file up as a text document. Odds are you won't be able to make sense of what you find, but you might uncover a few usable fragments from a document you thought you lost.

ScanDisk also cleans up cross-linked files and marks bad sectors so the computer doesn't try to write data to these sections. It's a good idea to regularly run ScanDisk. For active computers (those that are used on a daily basis), we recommend running ScanDisk every week. If you use the computer less often, you can run ScanDisk less often, but it's a good idea to run it at least once each month.

(NOTE: A few warnings: Don't run ScanDisk on a network or from within a DOS shell application or Windows 3.x. Don't run ScanDisk on a compressed drive. Close all applications before running ScanDisk And it's a good idea to back up your drive before using ScanDisk.)

Can't boot from drive. A boot, when you're talking about computers, refers to a computer's process of executing the basic startup routine as dictated by the BIOS. During the process of booting up, the computer looks to a particular drive or drives (designated in the BIOS) where it can find the information it needs to get itself going.

If the computer looks to this drive and can't find this information, it will halt the setup and an error message will appear on-screen. If this happens to you, there are a number of possible reasons.

One reason is that you might have a diskette in the diskette drive. On most systems, the BIOS tells the computer to look first to the diskette drive and then to the hard drive. Usually, the diskette drive is empty and the computer, sensing that, proceeds to the hard drive.

But if a diskette is in the diskette drive, the computer will attempt to boot from it. The computer won't find the information it needs, and that causes the error message. If this happens, remove the diskette and reboot the computer.

Another reason for the error message is not as innocent; you may have acquired a virus that has affected the master boot record (MBR; the instructions that tell the computer which drive to access during a bootup). This virus takes the place of the MBR when the computer is turned on, making it impossible for the computer to access the information it needs to correctly boot.

Such a virus creates a Catch-22 for the computer age: You need to run an anti-virus program to remove the virus, but you can't access the anit-virus program because the MBR is corrupt. Fortunately, there is a solution, as long as you have a bootable diskette. Stick the bootable diskette in the diskette drive, turn the computer off, and then turn it on again.

Remember when we said a diskette in the diskette drive could cause problems when the computer is turned on? Well, in this case, the diskette will solve your problems. The bootable diskette contains the operating system information for which the computer is looking. You won't receive an error message; you'll receive a DOS prompt instead. From there, run your anit-virus program, making sure it checks for boot viruses.

A third reason for this error message is that the MBR may have become corrupt by the installation of more than one operating system on the hard drive. If you have reason to believe your system doesn't have a virus on it or the virus scan comes up clean and if you've installed more than one operating system on your hard drive, type fdisk /MBR at the DOS prompt. This will extricate the old corrupt MBR and replace it with a new clean one.

Out Of Disk Space. This error message pertains to the way your computer 5 hard drive is formatted to store data. As we mentioned earlier, a drive is divided into partitions. Each partition is divided into sectors, and a group of sectors is called a cluster. Data is broken down and stored in these clusters.

Every cluster in a partition is the same size. Files that contain more data than a single cluster can hold are broken into pieces and divided among a number of clusters. But-and this is important to remember-a cluster can't hold more than one file. This system of partitions, clusters, and sectors is great for keeping your hard drive organized, but on most computers it isn't efficient. To illustrate this point, let's consider a hard drive that has 32KB clusters. If you save an 89KB graphics file to this hard drive, the computer would divide the data into three parts and send the data to three separate clusters. Among these three clusters, there would be 7KB of unused storage space. That's not too bad.

But let's say you want to save an 8KB word processing file to this same hard drive. The computer stores the file in a single 32KB cluster. But that cluster alone will have 24KB of unused storage space. If you have lots of similarly sized files on your hard drive, you'll "till" your drive when it reaches a quarter of its maximum capacity.

To solve this problem, you need to reduce the size of the cluster. But you can't do that because it's the FAT that determines the size of the clusters. What you can do is control the size of the partitions, which is what the PAT uses to determine the size of the dusters. II the partition is large, the FAT sets large clusters; if the partition is small, the FAT sets small dusters. So if you create small partitions, FAT will create small clusters.

Before partitioning a drive, we advise contacting the manufacturer of the hard drive. The manufacturer can tell you the easiest way to do it and tell you about the limitations of your drive. Use the software the manufacturer recommends and follow the directions. Also, back up your data beforehand. You'll lose everything on the hard drive when you repartition it.

We'll use FDISK to set up a two-partition hard drive. Boot from your bootable diskette, type fdisk at the DOS prompt, then press ENTER. From the resulting list of options, select the Create DOS Partition option, then press ENTER.

You need to create a primary partition first; this is where you'll store the operating system. Choose the Create Primary DOS Partition option, then press ENTER. Specify how much of the drive you want to allocate to the primary partition. We'll set the partition as 50% of the drive's capacity.

After you set the capacity of the primary partition you'll return to the main menu, where you need to choose the Set Active Partition option. The on-screen directions will guide you through this and return you to the main menu. Once again, choose the Create DOS Partition option and press ENTER. From the resulting menu, choose Create Extended DOS Partition, then press ENTER. We chose to set this partition as the remaining portion (50%) of the drive. When you return to the main menu, choose the Display Partition Information option. The resulting screen will tell you how the drives are configured. The primary partition probably will be listed as the C: drive. The extended partition probably will be listed as the D: drive. Take note of the drive letters.

Finally, you need to format the partitions. Insert the bootable diskette, type format c/s, and press ENTER. After the format is complete, remove the diskette and reboot the computer. This time it should start without the help of the bootable diskette. Format the D: drive by typing format d: at the DOS prompt. Now you're ready to reinstall the operating system and all your data.

It's best to create multiple partitions and set partition sizes when you first get a drive. Otherwise you'll have to reformat the drive, which entails erasing all the data from the drive. It's also important to think carefully about how many partitions you want and how' large each partition should be. Too many partitions can be confusing; too few can be a waste of drive space.

A useful partition size is something close to 500MB. That's big enough to hold large programs and small enough to make a noticeable improvement in storage efficiency. But if you're afraid to make that decision by yourself, contact the drive's manufacturer for some advice.

Of course, all of the above information is irrelevant if your computer is running Win98, your hard drive is larger than 1GB, you've backed up all the important data on the drive, you know you won't use another operating system on the computer, and the drive isn't compressed. If you meet these criteria, you can use ~Vin98's Drive Converter utility.

Drive Converter changes the FAT that's used by the hard drive. The FAT used by DOS and previous versions of Windows is called 16-bit FAT (FAT16) and was designed for hard drives that have capacities of less than 512MB. But as hard drive size has increased, FATI6 has become an impediment to drive performance so Microsoft added Drive Converter to Win98. This utility converts the hard drive to 32-bit FAT (FAT32), which provides smaller cluster sizes on bigger partitions.

To use Drive Converter, click Start, choose Programs, select Accessories, highlight System Tools, and click Drive Converter (FAT32). Follow the instructions as they appear on-screen to complete the drive conversion. Some of the software installed on your computer may not support FAT16 so Drive Converter may not let you complete the conversion. If the conversion completes, however, you can run Disk Defragmenter and continue computing as usual.

Drive crash. Probably the worst problem that could happen to your drive is a crash. This means your drive has died. A crash can be caused by hardware (such as the read/write heads hitting the platters) or by software (such as a virus scrambling the FAT).

The symptoms of a drive that has crashed range from the drive making strange grinding noises to an Invalid Drive Specification error message. If you can't access your drive regardless of what troubleshooting solutions you try, odds are your drive has crashed. At this point, if you've regularly backed up your drive, you're grateful that you have and you re annoyed that you're going to have to buy a new drive. If you haven't backed up your drive, you're probably enduring one of the most horrifying experiences you've ever had.

Fortunately, there are some companies such as DriveSavers (800/440-1904, 415/382-2000, http://www.drivesavers.com) and Ontrack (800/872-2599, 612/937-5161, http://www .ontrack.com) that can recover data from a crashed drive.

Before you call one of these places, however, you should do two things. First, call the drive manufacturer to find out if your drive is under warranty or if there's a way to save any of the data on the drive. Second, go buy a backup drive.

Safe & Secure. When you stop to think about all the data stored on your hard drive, it's mind-boggling that the drive can be efficiently organized. Thankfully, technology does most of the work. All you need to do is carry out a few drive maintenance tasks and periodically take note of how the drive is performing. Hopefully that's all it will take to keep your data safe and secure.