PC shuts down

Hey all, need some help. I have a Compaq Presario that I upgraded with more memory and a Radeon Galaxy graphics card. GW looks great with the new card and runs just fine. But, the hard drive runs up to full speed as soon as i get ingame and never slows down until about 15 minutes later when it shuts my PC down. I thought maybe it was overtaxing my drive so I just changed it out with a new 500g WD Caviar Green HDD. Still sounds like a hair dryer running full blast and still shuts down. All the fans are working and I even removed the side panel and pointed a 24" box fan in there, still shuts down. I've downloaded several SMART monitoring programs and can't seem to get any of them to work right. At this point I'm not real sure whats going on. Oh, and I upgraded the power supply also which obviously did no good. Any help or suggestions? Does this with Star Trek Online too so its not just GW. Thanks in advance.

Well, first of all, the hard drive always runs at full speed. You probably can't get the SMART programs to "work right" because there is no problem with the hard drive.

More than likely it is either the CPU fan or graphics card fan you hear, and my guess would be the video card. The fan speeds up to cool down the GPU on the card when you make more demands on it by playing a game. A certain amount of fan speed increase is normal.
It's possible though, that your video card is faulty and overheating. It's also possible that you have the graphics in the games turned up higher than the card can handle and/or you are running at a high resolution that is over-taxing the video card.

So, first thing to do is to put the side back on the computer and get rid of the box fan. Opening it up like that and directing a fan at it can (sometimes) actually make the airflow in the case worse.
Secondly, set the graphics resolution on your desktop and in-game, to the "native resolution" of your monitor. The native rez would be listed somewhere as x pixels by x pixels - for example, 1680x1050
Thirdly, reduce the graphics settings in game. Start by turning down AA to x2 (if it's set higher). In GW you can also use the "Auto detect" button to set the graphics closer to what your system can handle.

If doing the above doesn't help, you may have a bad video card, or there could be some other problem in your system configuration, or a virus, or a power supply problem.

Please post more details of your setup, such make/model of the computer, the CPU, and the video card. (What model of Presario and what model of Radeon - Galaxy appears to be a brand.)

Thanks for the reply. All the noise I was hearing turned out to be my CPU cooling fan. I unplugged it for a quick sec to verify and thats what it was. My Compaq is a SR5152NX with an AMD Athlon 64X2 Dual core at 2.10GHz. Upgraded RAM to 2gb and a 450 watt power supply. Graphics card is actually an Nvidia 9500gt. Once I cut way back on the textures it runs OK with GW now but Star Trek online at lowest settings still runs the CPU fan at max. GW looks like crap but at least I can play it.

I'm assuming at this point my processor just won't handle it? Looking at PCs online 2.10GHz doesn't look that bad. Anyway, opinions or advice would be appreciated. Thanks.

Also the CPU heatsink is clean.

Eek! Don't do that!
I thought it was odd that Galaxy suddenly switched to making Radeons after being an nVidia partner
Processor looks fine to me...

9500Gt is suspect, though. But, if you're certain that it's the CPU fan making the racket (after your potential meltdown test...) then that's where your problem lies. Check your CPU core temperatures ( http://www.alcpu.com/CoreTemp/ ) then check fan speed settings in BIOS to see if CPU fan is set to 100% always for some odd reason.

Thanks for the reply Snograt. I downloaded the coretemp and ran guild wars with low graphic settings. After sitting in Kamadan for about 3 minutes, my CPU was showing 70% load and 108c for temp (core#0) and 85c (core#1). Fan isn't set for 100%. I only hear it run like that when i'm ingame or just for a few seconds when I download or switch websites, etc. Most times just regular surfing its quiet. Maybe the graphics card is just too much for it?

23c is a huge difference between the two core temperatures.

I would suspect that the heatsink/fan on the CPU is either not properly positioned/secured, or that there was insufficient TIM applied.

Note: I use italics when I'm not certain about stuff - let's wait for an egghead to turn up with a second opinion

Something is seriously wrong with your CPU - temperatures of 85c on that model are lethal and 108 is higher than I've ever seen for that chipset. AMD states the max temp. for that particular CPU as 72c - and I know from personal experience that they shouldn't even hit 50c under load with a stock cooler. Do you have any thermal paste on there and what heatsink are you using???

I agree with Elder, your CPU temps are quite high. Been a while since i've had a desktop, but even in notebooks, the CPU temps shouldn't reach the 80C range. I'd suspect insufficient thermal paste or a detached heatsink like snograt.

Thanks for the replies. I've just got the stock heatsink and fan. The heatsink is clean and secure and the fan is working. I'll pull the heatsink tomorrow after I pick up some paste and check that out. I'll keep you guys informed. Thanks again.

I'll offer what I can from personal experience. 1st off the CPU fan revving and the spontaneous shutdowns may not be related.

Check the airflow through the PC. Adding a new graphics card can dramatically increase case temps especially if your airflow isn't good. Investing in a $5 exhaust fan is a really effective way of lowering case temp and inproving airflow.

Your CPU temps are way to high and the core temps differ so greatly that it suggests that your heatsink isn't correctly seated. Stock fans are OK but getting a herking great "flower" heatsink will never, ever hurt. It maybe that the shutdowns are being triggered by thermal cutoff, but I think there's another suspect...

PSU. I've seen 450 Watt spec PSUs really struggle when new GPUs are added. People always quote math saying that lower spec PSUs should cope, but the reality is that they don't. Investing in a decent (600+ Watt) PSU will most likely see your shutdown problems go away.

My recommendation would be to go for all 3. You've spent hard earned cash on memory and GPU, so chucking a little change at the PSU, heatsink & and exhaust fan is well worth it!

good luck

Yeah that is a good point, a PSU can cause all sorts of problems and more often then not they just act up rather then dying completely. What brand and model of PSU do you have? 450 watt ought to be enough for a 9500 gt but not all PSUs are created equally.

Jeez, with already purcasing new RAM, GPU and now maybe a PSU you would have been better off buying a whole new rig.

As the new GPU is a 9500GT, I assume the OP is on a rather tight budget.

When you say the heatsink is clean, do you mean that there is no paste or pad affixed at all? If that's the case, that would definitely explain why you're experiencing such high temperatures, and why the temperatures of the two cores is so vastly different. As such, you'd definitely want to put some paste on there before you even turn the machine on again if there's nothing at all.

Also, considering the 9500gt barely pulls 150 watts under load, I doubt greatly that the problem is related to the PSU not being rated to supply enough power, though it's not unimaginable that compaq / HP would toss a less then stellar PSU in the machine that doesn't preform to spec, or perhaps even failing.

Actually, this no longer seems to be the type of problem that can be solved just by suggesting things in this forum. It's gotten to the point where some hands-on maintenance is required.

Don't always believe what is says on the "tin". I've seen a 4770 pull down a 500 watt PSU to brown out. Like I said, just doing the math doesn't always give the full picture. Most stock PSUs are pretty dreadful.

If it's not a CPU thermal cutoff then a PSU brown out is the most likely problem.