Maybe the most important thing there was is that everybody should have a surge protector (UPS is even better). Still not many do. And I've seen lots of (not my) stuff being broken by not having one.

In Vancouver BC, I doubt very much that AC line instabilities cause even 1% of PC failures -- assuming reasonable home wiring. We have very good AC power.

The single biggest cause of PC/component failures over the past 10 years that I can identify is me. Careless handling; playing inside live, powered PCs and components; connecting power plugs incorrectly; and other user errors. Yup, PCStats got it right: user error is PC enemy #1. In my case, it comes with the territory -- hardware reviewing is bound to result in some mishaps, that's just the nature of the work.

But I think it holds true for most PC enthusiasts, that the worse & most common problems are caused by user errors.

Author left out simple NEGLECT.

Opening up computers and finding thick carpets of dust covering everything, and fans that are choking or even dead.

As a coworker says (from experience):

"Wine, screwdrivers and motherboards don't mix."

-D

They also left out sabotage. I remember when I was a kid I had a really old 300mhz pentium 2 and my dad was the type of person who would use things till it's broke. I couldn't convince him to replace it so I introduced Mr. cpu with miss 9 volt battery. I think that was the precise moment in my life when I got interested in computing. Now I'm a computer science/accounting major.

I remember a nasty buzzing in the speakers of a relatively new TV way back when. The tech who received it at the warranty station had a smirk as he unscrewed the front facia. He'd seen/heard this problem before. When the grill over the speaker was removed, a handful of pennies fell out. My toddlers had stuffed them through the grill slots. Buzzing problem solved. The tech said there'd been a dozen such cases before me.

Very interesting article.. and more than a few good chuckles.

Im also one of the crowd who fried a PC with bad power.

A friend and myself were busily fragging each other one nite when the PC he was using just up and rebooted, and mine hung. Both of us rebooted again, but the new PC he was using was suddenly VERY slow.

Seems a power surge had come through the house, through my power strip and left my older pc untouched, but fried most of the math functions on the other PC, thus our attempts to play Doom were severly crippled.
Lesson learned: new Chips can cost more than college tuition for 1 quarter, so dont "test" new PCs.
-- and yes that was a while ago.

Maybe I've just been lucky, but I've seen / done many stupid things to computers over many years, and had relatively few problems. Yes, there was one time when I upgraded a CPU and discovered to my surprise that my video card no longer worked. Never could trace that one down.

But even when my friend one time idly switched the red voltage switch on his PSU while the computer was on, didn't cause any lasting damage.

I haven't killed anything for a long time, since I managed to snap the power connector off the motherboard of my BBC Micro.

I've had a couple of near misses recently, probably due to advancing years and complacency.

Plugging my brand new £500 PCI-E RAID card into a PCI slot - that made for some blue smoke, burnt PCB traces and heart palpitations. Fortunately it's OK, but I doubt I'll be selling it second-hand on e-Bay.

Mounting a CPU heatsink the wrong way round on a Socket A CPU in my mates PC, that was also entertaining (it doesn't make proper contact so after 30 seconds of power-up there's lots of beeping and a shutdown).

Let me count the ways I've killed my numerous PCs:

• Bad Power supply during an upgrade - Dead MB
• Water leak (CPU block corroded) - Dead Video
• Water leak (bad install and no leak test!) - Dead MB
• Water leak (let my son get too close and he yanked a tube just enough to have a *very* slow leak) - Dead MB
• Plug in a pan header while pc is live - nice shock wave out of that! Deam wireless router, Antec 350 Silent PS. The machine works but the CPU is damaged.

I've also electrocuted myself on a modem - that is how I learned there was power in the phone lines! I had left the phone cable connected and tried plugging the modem into the motherboard.

Not computer related but I have tried to replace a plug with the power on - it was dark and needed the lights. After the shock dissipated, I decided to wait till morning.

Notice a pattern of stupidity? Having bad power in my area I do have a UPS but may need a brain surge protector

The article did sort of miss out one category - people killing their machines with software.

e.g. get so much mallware/viruses/just general cruft built up that the machine won't run, or the operating system won't boot, or crashes, etc.

While that is technically different; from a user's perspective it looks the same (i.e. computer doesn't do what I want it to). Is also responsible for significant number of machines being replaced prematurely (easier/cheaper to get a new machine than to have somebody clean out the software mess).

(And yes, you can even get snap-crackle-pop that way.
e.g. setting the video resolution/refresh rate too high for the monitor you are using. (Or in the case when I did it, forgetting to start in safe mode when plugging in a different monitor.) - One of the few ways I know to fry something from the keyboard/mouse.).

I can't believe how terrible everything is according to this article. If this article is to believed, as soon as any single part of your PC fails or you so much as yell at it while it's on the entire thing will fry. I've had tons of hardware fail on me and in general once I figured out what exactly broke, I could still use the rest.

Only thing that really causes total and utter failure is a cheap PSU (have seen two of my friend's PCs make their way to their final destination that way)

I go to yard sales and occasionally buy used components through craigslist. In the last couple of years I've taken apart at least four cases with dead motherboards. Open up the generic psu and inside are capacitors with swollen tops and the beginning of leakage. Have several even older towers with CWT or Fortron psu's, all working.

I suspect a lot of cpu and motherboard failures could be traced to lousy psu's. The psu's will still work, but how much ripple or other garbage they're putting out will not be detectable by most users. Instead blame will fall on the motherboard or cpu. And many times the "dead" cpu will in fact be a dead motherboard.

Connectors and sockets increasingly being keyed to prevent the "where's pin 1?" problem is helping a lot. I killed a (speedy at the time) 486 CPU once because I had it rotated in the socket on power up.

There are still too many connectors on motherboards that are not keyed in any way and have tiny labels. The worst is the header where all your front panel connectors go, which should have been standardized long ago.

this probably has fried more components than reported. At times, this will not kill the whole board though so it might not be reported. THis is most retarded aspect of motherboard making that I can think of. I sit there for an hour at times trying to figure out what exactly this "high quality" (yet built like crap) enthusiast motherboard is indicating for which wires. The 30 dollar cases are more evident in their wiring than 200+ dollar motherboards.

had a thumbdrive burn a hole in my south bridge when I inserted it in a front slot.

Static electricity's a bitch.

Especially when combined with poorly designed cheap circuitry, like ich5r on the p4p series.

http://www.google.com/search?hl=en&clie ... f&oq=&aqi=

Moving a thumbdrive from one pc to another (all with ups) while remaining seated at your desk, should not cause your motherboard to fry. Isn't that how usb is supposed to work, "hot swappable"?
Now shuffling along the shag carpet in a lightning storm wearing wool socks and using an ungrounded plug would be another story.

Speaking of static, I feel I need to address this. It's commonly stated that touching your computer case frame will ground you and prevent you from damaging your electronics. This is only true when the power supply cable is plugged into the wall socket and installed in the computer. An unplugged power supply does not ground your case. I've seen too many youtube videos where someone says to the camera "better touch that case to ground yourself!" while everything is unplugged and the computer is on the carpet floor.

Will sitting on tiled floors in the lower groundfloor help?

And if you actually have an earth. Americans, I'm looking at you.

that has been my best bet as well.
and brand names. don't count on those either. I had to dismantle an earthwatts 380w, to strap down a critical piece of plastic...
You want a computer? read read and read some more. there are a few epiphanies once the feminine man in sales gets out of your face...or in contrast, the apeman with the 35GHZ computer 62 fans and 120000 watt psu...it is the extremists of the pc world.


12th system on the way in coming days...

I have got alot to say on this subject.... and ya know what...the computer and its manufacturing is its own #2 enemy..#1 is the sales jerk believing it to the point of an outrageous fairy tale....

I just read in the fbi news of counterfeit cisco stuff from china, complete with labels. I'd swear I didn't have an intel sometimes, with my intel chipsets....
and another note....
chipsets after the 2000 revisions (all related to security and firmware hubs) could possibly run into the 15 year mark, for the first time in history. From my own knowledge, and thorough reading, that means socket 370 and onwards could start up and browse the net safely....socket 370 is already ten.
Pci. agp, opr pcie vid cards? they are all going to benefit from the load taken off cpu shortly...(now THAT is an evolution!)

I have just purchased my 4th socket 478 chipset mobo...and truly look forward to another run.

and I'll leave off with just one tip of many...

start with the pc case.
SECC metal .8mm and thicker, and be prepared tp modify, even the greatest of brands...and when paint is necessary. think of the stinky stuff, it most often has two parts either for urethane or epoxy, the rest is absolutely a hopeless green push and fake product..including acrylics...

Bought a pair of tired looking but quality stereo speakers at a yard sale and before they made it to their destination in my bedroom I knocked one onto the floor from several feet up. When I picked it up something was rattling. Cursing myself for my carelessness, I removed the woofer to survey the damage. It was a Hot Wheels style toy car, undoubtedly inserted through the bass reflex port. Speaker is fine.

Stereo in my car would sometimes 'change' volume and l/r balance as I went around corners. When I finally took it out to install a new one, a few coins fell out of the cd slot. After taking it apart I found .37 in change inside. Still works fine to this day.
My ashtray is right below the cd slot, is full of coins, and I had little kids at the time.

Hmm.. in my experience the biggest problem of all failiures is a bad PSU.

For example:
I build PC's for people with the intention a stable system. My mate does that too. But his clients dont have a pc longer than 5 years becaus the pc is dead. My clients pc's (even the older ones) are still gong strong.

The difference:
Imagin e the same hardware setup.
He: Gets the cheapest case with space for a buttload of fans and a included psu (for 30 euro's total) and puts the money he has left for the Case + PSU in extra fans.
I: Get a case with a quality psu (or get a case with psu from a decent brand). Thats the difference.

Also. I dont trust CCFL's anymore. Client of mine wanted one and when i tried to switch it on or off the pc went off. DO the same with LED strips and the pc keeps running. The inverter is evil from those CCFL kits.

Also i always work with a ESD cord (offc connected to a grounded wall outlet OR case with PSU which is off but the wall cord still connected).