The technologist’s guide to troubleshooting hardware

Anyone who has ever done desktop support in any capacity, whether it was for an office with hundreds of employees or just for your family, knows that calling to get warranty support can be one of the most frustrating aspects of the job. You’ve got to run through a troubleshooting script with someone who assumes you don’t know anything, and there’s no secret code that techs can use to signify that they know what they’re talking about. You can get bumped from department to department before you finally talk to someone who can help you, and often have to re-state your name and all of the pertinent information each time. It’s a huge time sink, and it’s one I prefer to avoid if at all possible.

Learning how to expedite this process is a vital part of any technologist’s skill set. So, drawing upon years of sad practice, we’ve put together a master guide to troubleshooting for PCs, Macs, phones, and tablets in a way that should minimize the amount of time you spend on the phone with support people who are always going to start by asking you whether your broken hardware is plugged in.

Known-good parts: the troubleshooting silver bullet

Before we get into specific symptoms and fixes, there's one silver bullet that's guaranteed to get you past most of the support person’s troubleshooting script and right to what you want: the known-good part. That is, a power adapter, stick of memory, hard drive, or other component that has been plugged into another system and is known to be working properly.

Let’s say you’ve got a laptop that won’t power on. If you switch its power adapter for one that is known to work (or if you use its power adapter with a laptop that will power on and charge), you can say with a fair degree of certainty that the power adapter is not the problem. This method does require you to have working spare parts available for testing. But if you tell a phone tech that you’ve tested a particular problem with known good parts, you’ll automatically skip through a lot of the script—and quite possibly to the end of the conversation.

PC troubleshooting

PCs are always getting simpler and more streamlined, but there are still a lot of different parts to most of them, which means that there is a lot more that can go wrong with them. We’ll go through potential problems component by component, matching symptoms to issues and telling you the best way to inform your friend on the other end of the phone. Pay attention here, because many of these symptoms and procedures are also going to be useful when troubleshooting Macs, phones, and tablets.

Some computer manufacturers may ship (or make available for download) special diagnostic tools intended to detect problems with particular components. It's not always necessary to use these tools to diagnose problems, but getting support will often be easier if you have the error messages and codes generated by their tools. Having these error codes handy is the ultimate phone support shortcut, and if you open with them, you’ll almost always skip straight to the part where they set up the dispatch for you.

Power problems

Symptoms: Computer won't power on, battery won't charge.

If the computer simply isn't responding to any attempts to turn it on, you may be having power problems. Remember that there's a difference between not powering on and not booting—a computer with power problems won't light up or make any noises when the power button is pressed. If lights and fans are coming on but the operating system won't load, you may have a memory, hard drive, or even motherboard error instead.

As a first step, unplug the computer from power and remove any batteries, then press and hold the power button for 10 to 15 seconds. This will completely power cycle the computer, draining out any electricity that may be left lingering in its circuits (some desktop motherboards have a light on the motherboard that will stay on for a while after the computer has been unplugged—once this light goes out, you've discharged all of the power). If you plug the computer back in and still have no luck, it's time to start troubleshooting the different stages of the journey between the wall and the computer:

Start with the surge protector. Does the computer behave the same way if connected directly to the wall, or to another outlet that is known to be working normally?

Look at the power brick if you've got a laptop. Most power bricks have two cords: one that runs from the outlet to the brick, and one that runs from the brick to the computer. If either of these cords can be detached from the brick, try again with a known good cord if you have one. If you've got a desktop, you'll usually just have one cable to check, the one that goes from the outlet to the back of the computer. If your laptop’s cables and adapters are working normally, you’ve probably got a motherboard problem, and it’s time to call support.

If you’ve got a desktop, your problem could be either with the motherboard or with the system’s internal power supply. Again, a known-good power supply will tell you exactly which is the problem, but be sure to check for things like the aforementioned motherboard status light—if it lights up when the computer is plugged in, it may point to a motherboard issue rather than a power issue.

If your computer will turn on but your battery won’t charge, you’ve almost certainly got a bad battery. As always, try a known good battery in the computer (and, if you can, try the suspect battery in a laptop that is known to charge) and make sure it’s not an issue with the contacts in the computer.

If you do have a bad battery, it likely isn’t covered under warranty unless it failed prematurely. If the battery is less than a year old, you may be able to get a replacement. But if the battery is over a year old, any loss of capacity or breakage will generally be seen as “normal wear and tear” and you’ll have to buy a new one. Most laptop manufacturers will insist you buy a first-party battery to avoid voiding the warranty on the rest of the computer.

Memory

Symptoms: Blue screens or crashing applications, computer powers on but will not boot, other erratic behaviors.

Memory errors can be hard to diagnose since they're often intermittent, but they present most often as general system instability: individual applications or the entire operating system may crash, the system may sometimes refuse to boot. And you may even experience graphics corruption, since the integrated graphics processors used by many computers today use the same memory as the rest of the system.

For more serious memory errors, the computer may power on and beep or make the power light flash a certain number of times without attempting to boot from the hard drive. Consult your computer's manual to see if these correspond with any known error codes.

Your first step in troubleshooting this problem is going to be a good memory diagnostic tool. Some computers will have a memory test tool built into the BIOS; the vendor’s support center will typically ask for error codes generated by those tools when replacing memory. But there are some good general-purpose alternatives if your computer shipped without one of these tools. Windows 7 comes with its own memory diagnostic, which can be run from within Windows 7 or from the Windows 7 install media.

My personal favorite memory test tool is Memtest86+. To run Memtest, you’ll need to download its disk image, burn it to a CD, and then boot your computer off the CD. The tool will automatically start testing your memory and will keep making additional passes until you shut the computer off. Generally, if the tool hasn't found an error after two or three passes, it's not going to.

Once you’ve verified that you’re dealing with memory errors (and assuming that your computer has multiple memory modules installed, as almost all of them do these days), take each module out and test it individually. This can help you isolate the issue to one of the RAM modules. Once you’ve got a module that you know is good, be sure to test it in all of the slots as well—this will reaffirm that the problem is with the memory stick and not with one or more of the RAM slots on the motherboard.

Hard drive

Symptoms: Slow or inconsistent performance, errors when attempting to access files, computer unable to boot, louder-than-usual drive clicking or activity noises (for mechanical HDDs only).

Losing the hard drive in a computer is one the most devastating failures you can experience, since the data is often the most valuable part of the computer. Even if all of the other components fail, the drive can still be pulled and the data transferred. But data recovery services for failed hard drives can cost thousands of dollars, and they aren’t foolproof. If you don’t have a good backup system in place (and you should: there are plenty of products that do it), you should be checking your drive for errors regularly—detecting a failure early is the best way to prevent data loss.

As with RAM, some manufacturers (particularly business-class systems from the likes of Dell or HP) include their own diagnostic tools with their computers, either in the BIOS or on a disc—if they do, they’ll prefer data gathered with those tools to data gathered by others. Even so, you can generally convince them that you’re having problems if you tell them you’re experiencing one or more of the symptoms listed above along with confirmed bad sectors found by a tool like Microsoft’s Chkdsk. Whatever tool you run, it’s vitally important that you back up any data from a suspect drive before you run any of these scans, as they are quite intensive and may actually exacerbate problems in the process of detecting them.

Chkdsk is normally run in one of two ways, depending on whether you can get your computer to boot or not. If your computer can boot, you can initiate the scan from within Windows. In a Windows Explorer window, go to Computer and right-click the drive you’d like to scan. Click Properties in the menu that pops up. In the properties box for the drive, under the Tools tab, click “Check now” under the Error-checking section, check “Scan for and attempt recovery of bad sectors,” and click Start. The computer will then offer to schedule a disk check for the next time you start the computer; accept the prompt and restart the system.

Finding your Chkdsk results in the Event Viewer takes a little digging.

When the disk check is done, it will display the results of the scan, but they’ll likely flash by so quickly that you'll miss them. To see this log file after the fact, open up the Windows Event Viewer (type “Event Viewer” into the Start menu’s search field and it should come up), expand the “Windows Logs” drop down, and select “Applications.” The Chkdsk log should be near the top of this list (the Source column should say “Wininit”), and if you scroll down under the “General” tab you should see the results of your test, as shown in the screenshot below. If you see anything more than “0 KB in bad sectors,” you should replace that drive—it’s not long for this world.

The other way to run Chkdsk is from the Windows install media—we showed you how to get this in a previous article. Boot to the media and before you do anything, press Shift+F10 on your keyboard to bring up a command prompt window. Type chkdsk c: /r (assuming the drive you want to check is drive C) and wait for the results. Again, anything more than 0 KB in bad sectors on the disk means that it’s time for a replacement.

Display problems

Symptoms: Computer won't display output on monitor, display is corrupted or garbled.

Shattered LCD screens like this one typically aren't covered under warranty, since screens are most often broken by being either dropped or stepped on.

Credit: Sam Kim

Your first task in this case is to determine whether the problem is caused by the monitor or the graphics processor; the tech on the phone is going to ask you if you've tried connecting the computer to another monitor to test this. If output displays normally on the monitor, then the problem is with the computer's screen; if output is still corrupted or garbled, you've got a problem with the GPU (or perhaps with the memory, as noted above—make sure you run thorough memory diagnostics before blaming the GPU). Telling the phone tech that you've tested with another monitor will usually get them to skip their script and go right to fixing your problem.

In cases where the computer appears to be powering on normally but there's no output on the display, it's not uncommon for the monitor's backlight to have gone out. If you put the display under a bright light, you may be able to make out images and text. If you've got a desktop computer or an external monitor that won't power on, double-check the power cable and power adapter the same way you would for a laptop.

Motherboard

Symptoms: Most of the above.

Since the motherboard connects to all of the major components of the computer, its problems can manifest in all of the ways we've already talked about. A bad RAM slot can make it look like you have bad memory; a non-booting computer can look like a power issue; a bad integrated graphics processor might look like a monitor issue. One of the best ways to diagnose motherboard problems, if your computer doesn't come with built-in motherboard diagnostics, is to test all of the other components first. When you have eliminated the impossible, whatever remains, however improbable, must be the truth.

Mac troubleshooting

Diagnostics, OS 9 style: it's the Apple Hardware Test!

Troubleshooting a Mac is similar to troubleshooting a PC in most important ways, since the two have shared basically identical internals for the last six years now (and even before the Intel transition, the fundamentals were mostly the same). As such, the symptoms they present will be the same, and some troubleshooting steps (particularly for display and power issues) will be identical. Macs do use their own troubleshooting tools, though, and there are a few Mac-specific troubleshooting steps that the AppleCare representatives are going to ask you to perform when you call in.

Apple’s main diagnostic tool for Macs is the Apple Hardware Test, which ships with all new Macs and has remained nearly unchanged (right down to the OS 9-style windows and buttons) despite the Mac platform’s numerous software and hardware changes over the last decade. The tool can be accessed by holding down the D key at boot, but if you've wiped or replaced your Mac's hard drive since buying it, you'll need to go through a couple of extra steps: on older, pre-Lion Macs, you can find the Apple Hardware Test on one of the restore DVDs that came with the computer. Newer Macs can access the tool via the Internet thanks to the Lion Internet Recovery feature, again by holding down the D key as the computer boots.

Performing the Extended version of the test is recommended, since it’s the one that is more likely to find errors. It will do fairly robust testing on your memory, motherboard ("logic board" in Apple parlance), and other components—take note of any error codes or messages you see while the test is running. Other diagnostic tools (like TechTool Pro) have been developed for the Mac, and Memtest86+ will also work as a memory tester, but for the purposes of getting support, Apple tends to want error codes and messages generated by its own tools.

You can use Disk Utility's verify and repair disk options to check for some hard drive failures—it relies on the SMART self-reporting information from the drive to detect problems. My personal experience with SMART has been a bit mixed—it has detected drive failures in time to save the data, but it has also failed to catch errors before a drive or two became unreadable garbage. As is the case on PCs, a good backup strategy is your best protection against drive failure.

There are also a couple of Mac-specific things you can do to get rid of minor, intermittent problems, especially those related to trouble booting or powering on. Most Apple techs will ask you if you’ve already performed these steps, so even if they don’t fix anything, you might as well try them.

The first such trick to reset your PRAM, which contains settings for virtual memory, the computer’s startup disk, and a few other settings (the entire list is located on Apple’s support page. To reset the PRAM, turn your computer off, press and hold the Command, Option, P and R keys, and power on the computer. Once you hear the startup chime the second time, let go of the keys, and see if your problems persist.

The second Mac-specific fix is to reset the System Management Controller (SMC), which can fix fan speed and power issues (including both the system’s power lights and the ability to power the system on)—a complete list is available on Apple’s support page. On desktops and older MacBooks with removable batteries, resetting the SMC is just a fancy name for power cycling a PC—unplug the system, remove the battery, and hold down the power button for a few seconds to drain any residual power from the system. On newer MacBooks with built-in batteries, you must turn the computer off, then press (but not hold) the Shift, Control, Option, and power buttons at the same time.

Smartphone and tablet troubleshooting

Credit: Clifford Joseph Kozak

There's much less to say about tablets and smartphones than about some of the other hardware we've looked at—the more appliance-like a device gets, the fewer things you can try to get it working before you have to cave and call support or take it into your carrier’s store. If your device isn’t charging, the steps for troubleshooting a faulty power adapter are the same as they would be for a PC or a Mac—try a known good adapter with your device, or the suspect adapter with a known-good device—but otherwise there simply aren’t many diagnostics to run.

If your power adapter seems OK but you can’t turn the device on, there are a couple of things to try: if your phone or tablet has a removable battery, remove it and hold the power button down for a few seconds and then replace the battery to cycle the power. On iOS devices that don’t feature removable batteries, you can try to perform a hard reset by pressing both the power and the home button simultaneously—if there’s nothing wrong with the phone, it should turn on within a few seconds, at which point you can release the buttons. These troubleshooting steps can also help you out if your device has frozen or become unresponsive.