How to repair a dying PC
If your problem doesn’t appear to be as a result of a driver, or if you can’t isolate the cause, you can always use System Restore to try rolling back your PC to an earlier state and see if this solves the problem. The most drastic measure is to completely reinstall your OS. However, before resorting to this, try booting into a different operating system such as Ubuntu Linux (perhaps from a live CD if you don’t want to set it up semi-permanently on your hard disk). If complications occur here too then you know the problem isn’t specific to your Windows installation, so it’s time to shift your focus to possible hardware issues.
Sometimes Windows will crash with a blue-screen error, yet there won’t be a driver identified – or, perhaps, different crashes will report different filenames. In this case your underlying problem could be faulty memory causing the operating system to try to execute impossible instructions or operate on invalid data, and to crash when it can’t.
Diagnosing this type of memory error can be difficult because, by nature, the error is intermittent. (On the rare occasions when a DIMM fails completely, the computer will normally refuse to start until it’s removed.) In fact, your best clue to a memory error is the seemingly random nature of the crashes you experience. If you suspect your RAM is playing up, you can test it by pressing F8 as Windows boots to access the Advanced Boot Options screen. Then press Escape to see the list of operating systems available, use the Tab key to move down to the “Windows Memory Diagnostics” item and press Return. The standard test should take less than ten minutes if all is well – you can change the testing options by pressing F1 at the main test screen. You can also launch the Memory Diagnostic tool from the System Recovery options on a Windows installation DVD.
If the problem is a memory error, you may assume you simply need to buy new RAM. This isn’t expensive, so if you’ve been intending to upgrade anyway, you may as well consider this an opportunity to take the plunge.
However, you may not need to splash out on new DIMMs to restore your system to stability. We’ve often found that when memory modules appear faulty, it’s because the BIOS is trying to run them above their rated speed, and they can’t reliably keep up. (This is a particular risk if you’re combining DIMMs of different types, a practice that enthusiasts often advise against for precisely this reason.)
To remedy this situation, check the frequency and timings of each of your DIMMs, either by examining the label or by looking them up on the manufacturer’s website. The former will be a speed, such as 1,333MHz, and the latter will be a series of numbers, such as 7-7-7-21. Then access your BIOS, disable automatic RAM settings and manually enter the correct settings. If you have lots of different DIMMs, choose the lowest quoted frequency and the slowest quoted timings – that is, the highest numbers. You won’t be extracting the maximum possible performance from every DIMM, but you may well find that, magically, you have a stable system once more.
Hard disk failures
A hard disk failure can be catastrophic, since it usually results in data loss. However, a disk can often alert you to looming problems before it fails. This is courtesy of a technology called SMART (Self-Monitoring, Analysis and Reporting Technology), which detects warning signs such as increased heat output and vibrations. As long as SMART is switched on in your system BIOS, you’ll see a warning when you turn on your computer if one of your drives is reporting trouble. This is your cue to back up your personal files immediately and make plans to replace the disk. If you want to check a drive’s SMART data for yourself, there are plenty of free tools that will do the job, including SpeedFan, which we’ll cover below.
Occasionally, despite what the SMART data may say, a hard disk will abruptly give up the ghost while you’re using it. This might cause a blue-screen error, but it’s more likely that Windows will freeze and/or shut down; the sudden loss of the entire filing system, plus the contents of the page file, is too drastic for Windows to crash “cleanly”.
After an event such as this, the disk may be permanently dead, in which case your PC will naturally refuse to reboot. Or it may restart smoothly, making it difficult to correctly diagnose the problem. For mechanical drives, two possible clues to hard disk problems are a regular clicking sound coming from the drive, or a whining noise. Both of these spell doom.