I run a local computer repair and IT support business in Brisbane, Australia. When my phone rings with a number I don’t recognise it’s invariably a potential opportunity with a new client. I take on all sorts of jobs, both hardware and softwarerelated,andI’m usually keenly anticipating what any new client is going to say.
My phone rang the other day and my heart sank when the client told me about the issue she was having. She said that they had a desktop computer which was randomly shutting down. The reason my heart sank was because she told me that the shutdowns would happen maybe a couple of times a day. In cases like this it can be very difficult to confirm that the issue has been resolved due to the intermittent nature of the issue.
Causes of Spontaneous Computer Shutdowns
The cause of a spontaneous computer shutdown can be any one of a number of things. However, it’s much easier to diagnose the issue if the behaviour is regular and repeatable. For example, if the computer always shuts down a few seconds after startup, or even after being on for a few minutes, then I can begin to take various remedial actions and I can quickly determine whether any of these actions has been effective in resolving the issue.
It’s a much more challenging scenario if the shutdowns are only happening occasionally. Because I charge for my time in fifteen minute blocks it’s not practical or cost-effective for me to take some sort of remedial action and then wait for hours to determine whether this has eliminated the cause of the shutdown. For this reason, it can be a case of taking one or two actions and then leaving the computer with the client for a day or two so that they can determine if the issue has been eliminated. This can become a multi-stage process as I take one action after another until the issue is resolved.
In my experience, spontaneous shutdowns are usually caused by hardware issues rather than software issues. I’m referring to the sorts of shutdowns which happen totally spontaneously whilst working on the computer. A shutdown which is preceded by some sort of warning on the screen, or even a blue screen of death (BSOD), is a different matter. These can be caused by driver issues, or have some other cause which doesn’t point to faulty hardware.
Thermal Maintenance
My starting point will usually be to perform some thermal maintenance on the computer, especially if it’s an older machine. It is of course possible to install software which will monitor the temperature of the computer’s CPU. Use of such software means that you can see the temperature prior to any shutdown and this can give you a very good idea about whether a temperature increase is causing the issue.
The first step in thermal maintenance is to open the computer and remove any dust and debris. A vacuum cleaner can be used as long as you take care not to touch any of the circuitry or components. Following this I would remove the CPU cooler, wipe away the thermal paste from both the CPU and the cooler, and apply a new layer of paste to the CPU. I would tend to be generous with the quantity of paste used as it’s better to err on the side of too much rather than too little. After applying the paste I would then reassemble the cooler, switch the computer on, and see what happens.
Additional Troubleshooting Steps
This is exactly what I did for the client mentioned at the beginning of the article. But before switching the computer back on I also took several other steps. I swapped the two RAM modules to different slots (there were four slots in total). I moved the hard drive SATA cable to a different SATA connector on the motherboard, and I removed the graphics card. I then moved the monitor cable to the HDMI output from the motherboard and switched the computer on.
The computer powered on successfully and continued to operate. However the client preferred to have the best possible graphics and so we powered the computer down again and I reinstalled the graphics card, this time to a different slot. We now found that we couldn’t power the computer on so I removed the graphics card and tried again. Still the computer wouldn’t power on. I concluded that the issue wasn’t with the graphics card. I also removed the RAM modules one at a time and tried again, but still the computer refused to power on and remain on.
A Successful Diagnosis and Repair
I then reinstalled the graphics card and tried a different power supply which I use for test purposes. This time the computer powered on successfully and remained on. I then powered it off again and tried another power supply unit. Again the computer powered on successfully. At this point I left it with the client for a couple of days and they were happy to report that the computer continued to function well.
In this particular case, the spontaneous shutdown issue had been caused by a faulty power supply unit. It could, however, have been caused by almost any element of the computer’s hardware. My troubleshooting process eliminated thermal issues, faulty RAM, and the graphics card, before homing in on the power supply. I finished by installing a brand new power supply unit and the customer was left fully satisfied with the service provided by Norm’s Computer Services. Had the issue been caused by a fault with the motherboard it would have been a more complex and costly repair.
This diagnostic guide was written by Norm McLaughlin, founder of Norm’s Computer Services, a computer repair and IT support business in Brisbane, Australia.
