Random Access   chris, kp & rob
In The Lab
PC Troubleshooting Part 1: 5 Steps to Solve Common Problems
by chris

We have covered multiple topics in Random Access, including wired and wireless networking, Build-Your-Own-PC, and even Doom 3. While we try to include some precautionary notes and a bit of what-if scenarios, we really have not covered much in the way of troubleshooting. To be really effective at troubleshooting means not only being knowledgeable about products and technology, but understanding their interaction or relationship with the system as a whole. Things go much faster if you are able to apply common sense when approaching the task. I have found that some of the best Technical Support representatives are those who can solve problems by asking questions, even when they may have little or no personal experience with a product. While I can cover the general process of troubleshooting problems here, without some level of product knowledge and a bit of logic in your approach, arriving at a solution may take a while... or in some cases be more frustrating than the problem itself.

A big part of troubleshooting is asking the right questions; doing it well involves knowing which questions to ask and being able to interpret the answers you get. For example, suppose you are trying to determine why a system will not connect to the Internet at home. The first thing you need to do is define the real problem. The symptom is that you cannot connect to the Internet, but the problem could be that your network is hooked up wrong, that the system is still configured for connecting at work, the Internet connection might be off-line, or maybe that you have a loose or bad network cable. This is where common sense comes in. Did you consider obvious things like configuring the router on your wireless network?

Step 1: Define the problem
Some of the basic questions to ask when defining the problem include "Did it work before?" If it did, then what has changed since the last time it worked? If it never worked, then your task is more difficult, because you must not only troubleshoot the problem, but also try to determine if it can work in the first place. Installing a new high-end processor in an older motherboard may never work because the motherboard cannot operate at the required speed. Worst case, the processor could be damaged because it requires a much lower voltage than the board is capable of providing. Go back and check the all of the specifications or product requirements first. (This is a polite way of saying read the manual.)

Another basic question useful in troubleshooting is "Is this the only computer displaying the problem?" If the answer is no, then the problem source may be something that could affect multiple systems, not just yours. Try to identify what is common between the systems having the problem. If you have one system that works and another that doesn't, start to compare settings or configurations between the two, looking for obvious differences.

Can the problem be reproduced? If your answer is yes, this is actually a good thing since you now have a way to determine when the problem is corrected. If the answer is "no," or "not every time," then you have an intermittent problem. Intermittent problems can be nasty for several reasons, besides the obvious one that you can't be sure that the problem is truly fixed. On computers, intermittent problems can be caused by a software conflict, but are more often triggered by hardware problems such as bad memory. However if the cause is hardware related, it is also probable that other intermittent problems will occur. It is unlikely that bad memory is only going to cause just one program to fail.

Are there any error messages? Error messages can be a great help in identifying the problem. Even when the message does not tell you what went wrong or how to fix it, the error message itself may be a way to research the problem. When troubleshooting Windows or other Microsoft products, you can search their on-line support knowledgebase by specific products, and then enter the error message or the error code to narrow the results to a few specific articles. The same holds true for many manufacturer web sites, including Apple, Symantec, Corel, and others.

Step 2: Identify possible cause
To identify the cause of your problem, whether it is software not doing what it should, hardware not being found, or your system locking up, you need to be able to isolate the problem first. One method to isolate some hardware problems is by removing everything that isn't required to run. If you are building a new system from parts, and it won't start with everything assembled, have you tried a minimal configuration first? For example, you could start with a motherboard, memory, processor and video (no drives) and see if the system will power on. You should be able to enter the system Setup and see the start-up logo or device detection stage. One thing even veteran system builders sometimes overlook is the possibility of unseen problems like a motherboard shorting out against the case. You can set up a motherboard on an insulated pad to test this theory; it's not good for any sort of normal operation, since it is all too easy to short a cable or knock cards loose while exposed this way. If you have spare components, you can use these to help isolate the cause as well. The best way to use spare components is to exchange them one at a time, and then if this does not correct the problem, change it back to your original component. The idea is not to introduce multiple variables, but to test one item at a time, allowing you to isolate the possible cause. The reverse approach to handling this type of testing is to start with a working system, and then replace a working component with an unknown, one at a time. If all components work individually in another system, then the cause may be more fundamental (such as your new system board), something you are overlooking (such as the power supply), or a combination of components that result in a conflict.

Troubleshooting problems in the operating system can also be handled in a similar fashion by disabling drivers and programs that could be running in the background. On Windows systems, you can do this through the MSCONFIG utility, on older Macintosh OS versions, there is the Extensions Manager. Both tools allow you to selectively stop programs from loading at startup, and then restore them one at a time without having to uninstall and reinstall software.

Step 3: Research the problem
Whether you fully manage to isolate the problem or not, the next step is to research the problem. As I mentioned when identifying symptoms, error messages are often the easiest to research since they often identify a very specific problem. Lacking this level of detail, you would need to research based on the product and the symptoms you are experiencing. If nothing can be found on the manufacturer's web site, try doing a few searches using Google or some of the other search engine. You may find links to companies other than the one you were looking for; don't dismiss these out of hand. Take responses you find in blogs and user forums with a grain of salt. Solutions offered may or may not work, and some suggestions are so far off base that following them might make things worse and you might have to start over. One thing these search results may do is offer suggestions or courses of action.

Don't forget to check the help files that may be installed with your operating system or on the CD that came with the product. Most recent versions of Windows, including Windows 98, ME, 2000 and XP, all have troubleshooting wizards. These wizards walk you through a series of steps, most with suggestions or methods to correct a particular cause. To find a list of troubleshooters in Windows XP, open Windows Help and Support Center, search for "troubleshooter" and look in the full text results for a list of troubleshooting wizards.

Step 4: Select and Test Solutions
If you find a single solution, this part is easy. If you have multiple solutions, you will need to use some common sense and select which solution appears to be most likely to solve your problem. One thing you should consider is if your solution does not fix the problem or worse, what happens if it creates a different problem or breaks things so that other solutions won't work either? In the case of Operating System or application troubleshooting, one suggestion would be to make a backup of the hard drive first. Some updates can be un-installed from Add/Remove programs in the control panel. One recommendation with Windows is to create a copy of the registry or use the system recovery feature to create or restore a breakpoint. With Windows XP, you can launch the System restore utility from a button on the main MSCONFIG screen.

What if one of the solutions you find is to apply a BIOS flash to your drive or your system? If a flash update you use is for the wrong component, your computer can easily be transformed into a high-tech paperweight. In the case of a manufacturer's system (like our PowerSpec or WinBook systems), applying a flash from anywhere other than the specific manufacturer's site is a bad idea, since both Windows and your Recover CD depend on a model-specific BIOS image. The original motherboard BIOS lacks this information, and may actually create new issues that were addressed in the production BIOS. In this case I would consider trying all other solutions first. If nothing else works, and you found mention of a BIOS fix that addresses your specific issue, contact the manufacturer with the details; they should be able to duplicate your problem and test a BIOS fix for you, or point you in a direction to resolve the issue. BIOS flash programs often have compatibility checks involved to prevent flashing the wrong system or loading an incompatible image, but there are no guarantees. System flash utilities may have a way to make a backup copy of the old BIOS before overwriting with the new, if so, use it.

Step 5: Prevention
An often overlooked aspect of troubleshooting is the prevention aspect. If you regularly use Windows Update to look for security patches, you are taking preventative measures, not only from the various security threats, but updates that correct problems you didn't even know you had. The most common problem seen in our Service Departments is spyware and adware installations, either causing problems (like system or program crashes), or just slowing things down (AKA performance issues). Using a simple firewall and a good antivirus program won't stop these problems, but some will be blocked, and the firewall should alert you if anything suspicious is going on. Adding spyware blocking software is a highly recommended to detect and block most of these problem applications. The other approach to prevention is "If it ain't broke, don't fix it." Applying every update and software patch might never cause problems, but if you are not encountering an issue, why do it? A security patch is one thing, but updating versions may be a waste of time depending on the reason for the update.

Get Random Access

Understanding Tech

Print this article

Shop Online


Your Name:

Your Email:

Your Friend's Name:

Your Friend's Email:

 © Micro Center