We have confirmed what is perhaps the first rule of building your own PC: have a smart and PC-savvy friend. We’d like to thank Simon Bolduc, HUB: The Computer Paper’s resident PC doctor for coming by the office to help solve some of the PC’s issues and truly living up to his title. We’ll even forgive him for setting one of the test computers on fire.

Last month, we closed out the “Assembling the Beast” installment of Building the Ultimate Game PC by highlighting some issues we’d run in to and suggesting that the problems were well in hand and easily put to rest with a little problem solving. Turns out, that wasn’t quite the case.

When a machine is first powered on, whether it’s the first time or the five hundredth, it runs through POST, the power on self test. Or at least, it’s supposed to. The Ultimate Game PC turned out to be the Ultimate Pain in the Neck (and looked as though it might be the Ultimate Expensive Paperweight at some points in the process) as it simply wouldn’t go to POST. After trying the processor and other components in different motherboards and getting the same result — a complete refusal to POST, we were almost ready to give up.

Here are some of the potential problem areas:

Processor: The Intel Core 2 Duo Desktop Processor was replaced with the same model (E8500, $300, www.intel.ca). Both work flawlessly and, clearly with hindsight being what it is, the processor was not the problem. Indeed, unless there’s been some mishandling or overzealous overclocking, the CPU rarely turns out to be the problem. The LCD Poster (a small, monochrome screen that connects to the motherboard and gives system status updates) always got stuck on CPU INIT, initializing the processor, which led to the belief that it may have been a faulty processor. Fortunately, the problem was actually much easier (and cheaper) to solve. It was CPU related but it wasn’t the CPU itself. More on that in a moment.

RAM: Aside from the fact that the Corsair XMS3 DHX TWIN3X2048-1800C7DF 2GB DDR3 (2X1GB, $350, www.corsair.com) RAM had a larger than usual profile and thus required some creative component shuffling to get it to fit, it worked just fine. Or rather, it would have if the motherboard (ASUS Striker II NSE – not Extreme, as previously reported – $410, ca.asus.com) supported the 1800Mhz RAM, which it didn’t seem to. Instead, we replaced the Corsair RAM with 2GB of RAM from Kingston’s Value RAM series (KVR1066D3N7K2/2G, $170, www.valueram.com). It’s not very “Ultimate Game PC” worthy, but it works nicely. Ideally, we’d max out the RAM slots with at least 4GB of faster spec RAM, but we’ve yet to encounter any problems and the benchmark results are pretty impressive. That said, opting for the Value RAM allowed us to move the DuOrb CPU cooler into a configuration that doesn’t block any of the PCI slots and so we were able to install the Creative Sound Blaster X-Fi XtremeGamer Fatal1ty Professional Series sound card ($170, www.creative.com) giving us a sound experience worthy of the Ultimate Game PC moniker.

Video cards: The fact that we’ve never really experimented with nVidia’s System Link Interface (SLI), we thought that could be the problem, though it wouldn’t explain why the CPU wouldn’t initialize. We couldn’t get so much as a flicker on the screen; no error codes, no explanations. Nothing at all.

Motherboard: The ASUS Striker II mobo didn’t seem to be a likely culprit. Still, it warranted testing the components of the Ultimate Game PC out on another motherboard (Gigabyte GA-EP45-DS4P, $200, www.giga-byte.com). We got nothing on any of the other socket 775 boards we tested and so it was that much harder to narrow down where the problem lay.

Power supply: While the Thermaltake ToughPower 1000-Watt PSU ($365, www.thermaltake.com) didn’t seem a likely culprit either, given that all the fans spun up without issue and the mobo was clearly receiving power, it bore checking out. We grabbed an older PC that was in good working order, plugged our PSU in to the test PC’s power connector and Simon shorted the power switch pins to power the PC up. There was a whiff of ozone and the cloying smell of partially fried computer components in the air. Thinking it might have been an anomaly, we tried to short the power pins again. This time, looking at the front of the PC, we were greeted with a flickering orange-red light from the hard drive access light. I mentioned this fact to Simon, thinking the flickering light must mean that something good must have been happening. It turns out, what I was seeing was a small fire that had broken out on the motherboard, reflected through the HDD light lens. The smell of burning components became stronger and we were forced to take five to air the place out. In the end, the PSU turned out to be fine… though the same can’t be said for the PC that so selflessly gave its life in the process. RIP, little guy.

PEBKaC (problem exists between keyboard and chair): As is too often the case, the issue turned out to be sheer human error. Turns out that a lot has changed since I last built a PC; I neglected to plug in a connector from the PSU to the motherboard that is specifically designed to power the CPU. That would explain why, throughout all the building process, the boot seemed to get stuck before it even really started when trying to initialize the CPU. Lesson learned. Egg on face removed. Time to move on.

Next: Off the POST

Pages: 1 · 2 · 3