Note: This guide is assuming that the system was known to be working prior to this fault occurring, although these steps can be used to trouble shoot a new build there are some variations and other possible causes. For that reason if you’re experiencing the described fault on a new build and the steps below don’t resolve the issue, contact our technical support department on 01274 471204.
Let’s face it; computers can go wrong from time to time. Their engineering complexity yet operational simplicity can be a blessing as much as it can be a curse. Computers are simple machines, they generally do as they’re ‘told’ and faults are usually a result of the instructions it’s been given being incorrect.
But if you’re not sure what these instructions are or how they work, you’re going to struggle to diagnose the problem.
So in this handy guide I’m hopefully going to arm you with the skillset to troubleshoot problem components by using a methodical approach based on the theory of a computer’s basic operation, rather than teaching you how to diagnose a specific fault, in specific circumstances and in a specific way - that I have come to call ‘parrot’ fashion.
‘Parrot’ fashion is fantastic if you’re encountering the exact same problem as the one you were shown how to troubleshoot, which isn’t much help in a real world scenario.
For the majority of this guide, I’m going to assume that you either have access to an old computer to ‘borrow’ parts from or access to other components that you have left over from upgrades. At the very minimum you will need to have either a PC Speaker (a motherboard integrated diagnostic speaker – not a set of Creative 5.1 surround sounds) connected or a motherboard PCI diagnostic card which you can find very cheap online.
As the title of this guide suggests we’re going to tackle POST (Power on Self-Test) issues. These aren’t usually as simple as pressing the power switch and nothing happening, although that is one symptom. The usual manifestation of a POST error is that the PC appears to power on correctly but you don’t get a display. This causes a lot of end users to believe that either the monitor or the graphics card is faulty; but this is very rarely the case.
Before we can start discussing methods to troubleshoot the problem I first need to quickly explain in a simple way what a computer actually does when it performs its POST and what a POST actually involves.
A power on self-test is a check your system performs to ensure that all hardware is initialised correctly and appears to work. The first part of the power on self-test is instigated when your computers CPU is initialized, this prompts the BIOS to check that memory is installed correctly and then initialise the graphics card.
Once the system has detected that the memory is functioning and it is about to initialise the graphics adaptor it will give us a single beep to let us know that everything appears to be ok.
Knowing this process allows us to determine quickly and easily at which point the system is getting stuck during its powering on phase.
The Symptom: The computer is powering on but it isn’t beeping or displaying graphics
This is a quick and easy one to narrow down and the most common issue we come across with a system that isn’t powering on. People often believe that because the fans are spinning and the computer ‘appears’ to be functioning as normal that it must be an issue with the graphics card. However the lack of a single beep is telling us otherwise.
Follow the steps below to troubleshoot and determine the problem:
Remove the RAM from the system and power on without any memory installed. (For information on how to install or remove memory follow this guide here Computer Assembly Guide Part 2: Memory Installation & Fitting)
If the system beeps continuously when powered on without memory, most of the time this means it’s faulty RAM.
Solution: Try the RAM one stick at a time in the primary memory slot to identify which stick is causing the system to get stuck during POST.
Once you have identified the stick that is causing the problem, replace the stick or return under warranty for replacement accordingly.
If each stick of memory works and the system starts correctly with the memory in the primary slot, try the memory in each slot as you may have a defective memory slot on your motherboard in which case it is likely the motherboard – consult step three.
If you didn’t get any beeps continue onto the next step.
In instances where the system isn’t getting stuck due to faulty memory, the second most likely cause is insufficient power. There are many different tests you can perform on a PSU (Power Supply Unit), however most PSU testers and multi-meters are an inefficient test due to the fact that they don’t simulate the load that a PSU would be under when booting up.
Solution: Connect the power supply to the motherboard by just connecting the 24 pin ATX power connector and the 4 or 4 + 4 pin ATX 2 CPU power connector. (For information on how to install or remove a PSU follow this guide here Computer Assembly Guide Part 5: How to Fit and Connect a Power Supply Unit)
Although this isn’t going to allow to system to boot into the operating system or power up the other devices, it is going to allow the system to power on and beep at us if it can.
If with the replacement PSU the system gave us a single beep to let us know it had initialised the memory and everything was ok then connect your other devices to the ‘test’ power supply individually, if with all other devices connected the system still boots correctly – then we know that we require a replacement PSU. If when connecting a specific device such as a hard drive or optical drive the system shows the original fault again then we can identify that device as causing the problem and we know that device requires replacement.
If you still didn’t get any beeps continue onto the next step
If the memory or power supply weren’t to blame then the chances are that unfortunately it’s a fault with the motherboard. In these instances we would suggest contacting our technical support department who can arrange for the motherboard to come back to us for testing.
By knowing the sequence in which to tackle this one and thinking to ourselves “Is it getting past the memory or not, if it is what’s after, if it’s not what’s before” etc. we should all be equipped to break down this process and track the fault through each of the steps to narrow it down to the component at fault.
Any questions on this guide please feel free to leave a comment and I will endeavour to respond with any more information that I can.