It’s the new year, which must mean it’s time for all our PCs to go belly up.

Or so it seemed to me in the past couple of weeks. I thought the tale of these three systems is worth telling, as a lesson in recognizing and solving different types of PC problems. Perhaps you’ll see something of your own PC troubles in this, but even if you don’t, there are still lessons to be learned.

We’ll begin this with the story of my daughter’s ever-slowing Vista system.

The Tale of the Clogged PC

This past weekend, I’d decided it was time to nuke my daughter’s system from orbit. It was the only way to be sure.

Emily’s been running a fairly decent, though not bleeding edge system. Core components include an Intel QX6850, Asus P5Q3 Deluxe motherboard and an older 512MB Radeon HD 4870. Over the past couple of months, the system had begun running slower than molasses.

Emily is a fairly typical teenage girl when it comes to PC use. She uses the web heavily, and happily downloads anything she thinks she might like. She is, however, more tech savvy than most teenage girls, so she doesn’t do really stupid stuff, like open phishing emails. However, she’s a happy user of WildTangent games, likes to have the Weather Channel bug running (ugh, I say), and related sorts of gimmickry that can act as brakes on a fast system.

Recently, though, her system had been really dragging – so much so, that she’d given up on using it, and was using the communal living room laptop to do her homework and even run some light games. (I confess: I got her hooked on Torchlight.)

It all began several months back, when Emily began complaining that her system was glitchy. At the time, it was running Windows XP. I’d built the system about eighteen months ago, and it had been running reasonably well. I’d never been entirely happy with the QX6850, though. Even with a beefy Scythe Ninja cooler, the CPU typically idled at 58 degrees C. The QX6850 ran at 3GHz, but was built with the same 65nm process technology used in the original Conroe CPUs.

So I did something that, in retrospect, planted the seeds of bigger problems to come: I thought it would be a good idea to perform an in-place upgrade from Windows XP to Windows Vista.

I should have known better. I’d read enough online posts to know that the in-place upgrade from XP to Vista can be problematic, particularly if your XP system registry has grown to huge proportions. My own personal experience with a Vista upgrade had been moderately successful, but even that system was eventually nuked and a clean Vista install performed.

So here was a PC, with lots of fairly useless (from my perspective), but fun (from her perspective) background tasks running under XP. In goes the Vista upgrade.

Did I mention that this system had 2GB of system RAM?

Within two weeks, her system had pretty much become unusable. I scanned it with several malware detectors, and other than complaints about a couple of adware items, there were no Trojans or viruses found. So the problem really came down to an already overloaded Windows XP system that had developed massive registry and hard drive bloat after the Vista in-place upgrade. So I decided to nuke and start over.

One of the cardinal rules of troubleshooting is: change only one variable at a time.

Unfortunately in my hubris, I ignored that rule. I put her system on the bench, swapped out 2GB of Kingston DDR3 RAM for a pair of Corsair 2GB DDR3 modules. I then swapped out the QX6850 for an E8500 3.16GHz dual core CPU.

When I powered the system up, I got no POST, but the fans spun up. Also, the PC beeper didn’t beep. This is, as anyone who has built systems will tell you, one of the most ambiguous and frustrating types of errors.

So I popped her old CPU back in and rebooted.

Same thing.

So I put the old memory back in, and the system POSTed properly. I checked her BIOS on the P5Q3 motherboard, and found out it was still running the 0704 BIOS. That BIOS predated Intel’s 45nm CPUs. On top of that, if you read the summaries of the various BIOS updates, you also see a number of them with text that reads “enchances compatibility with certain memory.”

Lessons (re)learned:

• Before upgrading CPUs or memory, check the BIOS version

• Never change more than one item of hardware at a time without testing it.

The Constantly Crashing P7P55D

Now let’s consider a different problem, also with an Asus motherboard at its core.

I’d built a system running a Core i7 860 and the Asus P7P55D. The GPU installed was an AMD Radeon HD 4890. The system booted normally, and Windows 7 installed with no hiccups.

Periodically, the system would go into sleep mode – this was normal – but would then lock up. This was decidedly not normal. Occasionally, the system would lock up during multiplayer gaming sessions, too. Updating the BIOS didn’t fix the problem.

Then I came across several posts on several forums suggesting an incompatibility that was specific to the Radeon HD 4890. So I swapped in an eVGA 260 GTX Core 216 SSC.

Lockups still occurred, with the same frequency and symptoms. So the problem had nothing to do with graphics cards. Finally, I removed the memory – some early Corsair DDR3-1600 (though it was only running at 1333MHz) and swapped in some OCZ DDR3-1600 (also running at 1333MHz.)

The system has now been running smooth as silk over multiple gaming sessions.

These particular Corsair memory modules had actually given me fits in a number of different Intel-based motherboards, including P45, X48 and now a P55 motherboard. The Corsair memory is a 4GB kit, labeled CMS3X2048-1600C7DHX. What’s interesting about these modules is that when you drop them in an AMD AM3 based motherboard, the system seems to run perfectly fine. It’s quite a mystery, but illustrates how complex the issue of compatibility between components can be.

Lesson learned: nothing in particular, except to maybe to not trust forum posts, even multiple posts that may agree with each other. The graphics card wasn’t the problem here – once again, it was DRAM.

One other thought: despite rapidly falling DDR3 memory prices (or maybe because of rapidly falling prices), the state of the art in DDR3 modules seems to be evolving rapidly. This seems to create incompatibilities between specific motherboards and specific modules. I’ve had Kingston modules run great in Gigabyte motherboards and behave in a crashtastic manner in Asus boards. I’ve had OCZ memory behave badly in Intel boards, while running like a champ in eVGA X58 boards. Whether I adjusted timings, or ran in default modes, didn’t seem to matter.

How anyone can keep track of which modules run well on which motherboards at this point in time is a mystery. So the lesson here, I suppose is: if you don’t have a large supply of different modules and motherboards on hand – and most people don’t – make sure you buy your components from a reseller with a robust return and exchange policy.

The Slowly Dying Motherboard

My production system runs a Core i7 965 and 12GB of OCZ DDR3-1600 memory. The 965 clocks at 3.33GHz, and the memory runs at 1333MHz. In either case, it’s not really much of an overclock. The X58 chipset and Core i7 socket 1366 CPUs are rated at DDR3-1066 speeds officially. So technically, the memory is being overclocked.

The system was based on an eVGA X58 SLI motherboard – one of the early versions that shipped. I’d updated the BIOS as new versions came out.

Recently, I’d been having a weird series of glitches. My USB ports would stop working. Once every couple of days, I’d lose network connectivity. My audio would mysteriously stop working, even though I’d see audio activity in any media player, and the speakers seemed to be behaving correctly.

If finally occurred to me that maybe my I/O controller hub was slowly failing, or had developed an intermittent hiccup that created these seemingly unrelated issues.

So I decided to upgrade my motherboard.

My general philosophy when it comes to building the systems I actually use for productive work and personal gaming is to be somewhat conservative – if I can. Go with solutions that seem reliable and robust. I’d waited to upgrade my own system to Core i7 until I’d tested a slew of X58 boards, and the eVGA board had seemed stable. In fact, it had been running problem-free for months.

So naturally I ignored my own inclinations and bought a brand spanking new Asus P6X58D, which sported a discrete NEC USB 3.0 controller and a Marvell SATA 6gbps disc controller.

The board arrived, and seemed to install normally. Windows enumerated all the devices properly, and rebooted as expected. I installed the new drivers from the Asus DVD and rebooted.

Then a new symptom developed: after a warm boot, the fan on the Radeon HD 5870 would spin up to maximum speed and stay there. The sound was almost deafening. When this would occur, the system wouldn’t POST. Only shutting down the power supply and waiting a couple of minutes would allow for a normal boot.

Whiskey. Tango. Foxtrot.

After trying a variety of things – different CPUs, swapping in different memory, upgrading the BIOS (from 0201 to 0402) and even swapping in a different AMD graphics card, nothing changed. So I gave up. I headed out to my local friendly white box shop (Central Computer in San Jose), I picked up another Asus board: a P6TD Deluxe. The P6TD is a second (third?) generation X58 board from Asus, with more copper and 16-phase power. It’s really a souped up P6T board, but refined over the original design.

Swapped back in my original components, fired up the system and held my breath.

The system POSTed normally. Since then, I’ve run a number of applications, including some fairly system intensive games, and it’s all behaving as it should.

Lesson learned: when you need a working system for, well, work, then don’t go with the bleeding edge. I knew that already, but every now and then, I have to re-learn it. I suppose it’s my eternally optimistic nature, but I tend to believe that newer is often better. Alas, it’s not always the case.