possibly overheated

anon

anon

New Member
My PC just shut down automatically without a bluescreen or any kind of warning. I smelled a faint burning smell, so I waited a couple hours with the side panel removed to let it cool. Now when I turn it back on, the fans blast at full speed, but no video signal is sent to the screen. I'm assuming it overheated, but I'm not sure which part to fix/replace.

Some background:
Relatively new (~1/2 year old) computer. Was in the middle of Half-Life 2 at the time, with the resolution set at the maximum to see if it looked better than normal. I couldn't tell a difference, but I left it at full resolution anyways just because... I don't know the actual temperatures, and I can't check now because it won't even boot properly. I heard during normal operation it's supposed to detect high temperatures and reboot, or at least do a controlled shutdown to protect itself, but it seems like it shutdown because it broke.

Anyways, the actual specs are:
Case: Antec Three Hundred with two front fans and one on top (though the front panel was covered in dust...)
CPU: AMD Phenom II X4 920 on a Gigabyte motherboard
GPU: HIS Radeon HD 4890
PSU: 700W OCZ ModXStream Pro (very unlikely to be the problem)

I put it together from components purchased at newegg, so I'm not sure if I have warranties on the parts, but it would at least help if knew what to fix.
 
Try unplugging the power and removing the CMOS battery. Then tap the power button a few times to be sure all the power is drained from the board. Give it a minute or two, reinsert the battery, plug it in, and see if you get any video.
 
I actually already tried switching the power button and unplugging the cord, but did not get a different response. I didn't try messing with the battery yet, though I doubt it will change anything. Just in case though, you do mean that little button cell battery on the motherboard, right? So I don't screw up something else on this bad luck streak.
 
anon said:
I actually already tried switching the power button and unplugging the cord, but did not get a different response. I didn't try messing with the battery yet, though I doubt it will change anything. Just in case though, you do mean that little button cell battery on the motherboard, right? So I don't screw up something else on this bad luck streak.
Click to expand...

It's a small battery about the size of a quarter. Just pop it off. It'll reset the CMOS and reset any errors it detected. Hopefully it'll POST after that.
 
Removed the battery, toggled the power buttons/plugs some more, etc. and still got the same result: lights turn on and the fans go crazy, but nothing on the screen. And it's been several hours now since it failed.

I just cleaned the little filter in front of the front-mounted fans, and it was covered in dust. I doubt those fans (2 of 3) were doing much at all. And again, I was playing a game at high resolution settings, and I smelled something burning, so I'm pretty sure this is in fact an overheating problem. And it's not likely to be the PSU. But again, there are supposed to be fail-safes for this exact sort of problem. I'm at a loss for solutions right now...
 
No dice. Removing the 24-pin cable didn't change anything either. BTW, the video signal is VGA and is coming from the videocard.

Is there some way to troubleshoot to determine which part broke? Like whether there is a different set of symptoms when the CPU overheats as opposed to the videocard? This would be a million times easier if I could just replace one part and not the entire computer.
 
Last edited:
I would take your gpu out and hook up your monitor to the onboard graphics if your motherboad has onboard. It sounds to me like it's either a power supply or graphics card thats going bad or has gone bad.
 
G25r8cer said:
I would take your gpu out and hook up your monitor to the onboard graphics if your motherboad has onboard. It sounds to me like it's either a power supply or graphics card thats going bad or has gone bad.
Click to expand...

yeah i would do that. if it turns out to not be your gpu, then do the same with the other components, eventually getting down to the bare minimum mobo, cpu, ram, hdd, and psu.
 
The 4890s run real hot have a couple of friends with them and they had to watch their temps one even had to buy a fan controller for his as it was overheating.
 
G25r8cer said:
I would take your gpu out and hook up your monitor to the onboard graphics if your motherboad has onboard. It sounds to me like it's either a power supply or graphics card thats going bad or has gone bad.
Click to expand...

I'm thinking it's the videocard, because it failed during a moderately graphics intensive game. But as far as I can tell, I don't have onboard video on the motherboard, so I can't just remove the videocard and see if it works again. (Pic of motherboard.) It would be very uncool to drop $200 on a new videocard when that wasn't even the problem. I doubt it's the power supply because it's rated to 700W. What has me second guessing myself about he videocard is that the fans go crazy too, and I don't think the videocard has anything to do with controlling those.

I could not find any obviously burned pieces after a visual check, but I think the problem would be hidden under the heatsinks for the CPU or GPU anyways. Especially the GPU, because modern videocards are typically covered by plastic casings, which hide most of the electronics.

Hsv_Man said:
The 4890s run real hot have a couple of friends with them and they had to watch their temps one even had to buy a fan controller for his as it was overheating.
Click to expand...

Well frick... I may end up taking a chance and just replacing the videocard, but I still don't understand the fans going crazy. Assuming for the moment that the videocard is the problem, would you recommend I just get a new 4890, or is there a similar model that runs cooler?
 
Last edited:
anon said:
I'm thinking it's the videocard, because it failed during a moderately graphics intensive game. But as far as I can tell, I don't have onboard video on the motherboard, so I can't just remove the videocard and see if it works again. (Pic of motherboard.) It would be very uncool to drop $200 on a new videocard when that wasn't even the problem. I doubt it's the power supply because it's rated to 700W. What has me second guessing myself about he videocard is that the fans go crazy too, and I don't think the videocard has anything to do with controlling those.

I could not find any obviously burned pieces after a visual check, but I think the problem would be hidden under the heatsinks for the CPU or GPU anyways. Especially the GPU, because modern videocards are typically covered by plastic casings, which hide most of the electronics.



Well frick... I may end up taking a chance and just replacing the videocard, but I still don't understand the fans going crazy. Assuming for the moment that the videocard is the problem, would you recommend I just get a new 4890, or is there a similar model that runs cooler?
Click to expand...

don't you have another pc or a friend who's pc you could use? or perhaps just any old videocard lying around?
 
The PC was my main computer, but I have a laptop that I'm writing this with, so I'm not dead in the water. Though I don't have another PC lying around, and I'm afraid I'd burn out someone else's card if it turns out to be a power supply problem. Do you think there would be any risk if I borrowed another card?
 
Last edited:
You must log in or register to reply here.
Back
Top