I installed my new NVIDIA GTX 660 just like the directions said. After playing Skyrim for about 10 minutes, my computer shut off. It felt hot and smelled like something burnt. My power supply is 700 watts. I have 4 fans going, inside the computer. Any idea what happened? I'm afraid to turn it back on.

_________________
RIP Shiloh - I'll never forget ya

Have you upgraded your drivers recently? I know that one of the recent ones have caused some people's Nvidia cards to overheat or in some other capacity, die. I know this was a huge problem for the people who recently bought the new 700 series cards, as they were the ONLY drivers available for that series.

Seems to be 320.18 that's causing the hullabaloo.

_________________
-Formerly Movien, greatest Beastmaster evah

This is exactly the kind of thing I fear when tinkering with my computer. I should have taken it up to Geek Squad and had them install it for fifty bucks extra.

As for updated drivers, yes I just updated it when I installed it. Windows auto-updated it.

_________________
RIP Shiloh - I'll never forget ya

There's (usually) nothing wrong with just installing the driver yourself. Considering the performance fixes that drivers can fix sometimes, you'd be handing them 50$ more than you really need to.

Anyways, that might be your problem. I wouldn't trust windows to find shit for me and install it. Always get the drivers from either the manufacturers or Nvidia's site.

_________________
-Formerly Movien, greatest Beastmaster evah

Make sure those fans are actually working.

If the fans work, make sure you didn't dislodge the heat sink from the CPU or something.

Overheating CPU is more likely than overheating GPU. Most GPUs would rather run so hot they start to mess up the on-screen image rather than emergency cut off the power supply. That's very much an overheating CPU feature.

Power supply is fine, shouldn't cause any kind of overheating even if it wasn't.

In most modern PCs you can check the temperature of the CPU inside the BIOS, usually. This is handy because it's quick to open and you can quickly turn the PC off too.

If you suspect overheating, but the computer stays on long enough, you should get some monitor software running asap once you're in Windows. Something quick and simple like HWMonitor (64bit link, just run exe inside zip).

Bear in mind graphics cards run really damn hot, and that's normal. You can easily expect them to run at 150F while a game is going. CPUs and GPUs will only start to keel over once they reach ~190F. If your computer is doing nothing they'll probably idle around 100-120F (depends on overall cooling).

As for drivers, I really doubt it has anything to do with it, but go to http://www.nvidia.com and download them yourself. The drivers windows installs by itself are never the latest or even that great.

I turned it back on. Updating drivers via the installation disk. It's a good sign that the installation disk is taking about 10 minutes to update drivers while the windows auto-update only took like 1 minute. My computer is idling at 111*F, and 2 of the fans are going at 5000 RPM & 1100 RPM.

_________________
RIP Shiloh - I'll never forget ya

5000 RPM is max speed for a computer fan, and it's quite odd for them to be doing so unless the computer is struggling to cool itself. 1000-2000 is much more normal.

Other reason a fan would be at max speed is that it's connected directly via the power supply, instead of via a fan header on the motherboard. The power supply is unable to regulate a fan, and will just spin it at max speed.

Are one of your fans oddly loud like a vacuum cleaner? Which fan does it say is that fast? Or nothing out of the ordinary? To be honest it's also possible that the reading is wrong. Many of these temperature sensor programs can't get /everything/ correct due to how many different kinds of motherboards there are.

Also, in all honesty, when it comes to NVidia cards, just use the drivers NVidia supply from their website instead of any installation discs. The chance of that installation disc being up to date is ludicrously slim. There's also no need to ever use manufacturer branded (EVGA, Gigabyte, whatever) graphics card drivers either. The NVidia ones will always be better.

You can leave the temperature sensors running when you next try to run a game and keep an eye on it.

EDIT: And finally, you should probably clean the dust out of your computer. If you've had it more than a year or two it'll probably be full of it. Compressed air can will do wonders.

the fan (actually going 6000 RPM now) is the ASUSTeK Computer INC. M4A77TD fan

_________________
RIP Shiloh - I'll never forget ya

i ran a benchmark on my video card and here is the result:

_________________
RIP Shiloh - I'll never forget ya

81 celsius is about 180 * F, not too far from the boiling point of water

i wonder if i picked the wrong power supply cord to plug into? there were like 10 cords in there to pick from, and i just randomly picked one cause i figure theyre all the same.

_________________
RIP Shiloh - I'll never forget ya

it looks like the driver from my installation disk is about a year old, so i will try the actual website.

_________________
RIP Shiloh - I'll never forget ya

81 Celsius is ok if that's the temperature it maxes out at during a thrashing.

If it's idling at that temperature when you're just sitting at your desktop, then its a problem.

It should idle around 50 C, 110 F if it hasn't been doing anything for a while.

The power cord you've used will make no difference. If the PCIe power connector cable fit and you didn't use some kind of adapter to make it fit, then it's rated correctly for that graphics card. Temperature is entirely the onus of the fans, the heatsinks, and the paste used to join them (which you shouldn't ever have to touch on a graphics card).

The "name" of the fan usually just indicates what is providing the sensor reading. In this case, its the motherboard, so it could be any of the fans connected to the motherboard, including the case fans. If your computer doesn't sound like a vacuum cleaner though it's probably ok; just ignore it and pay attention to the temperatures instead.

alright. gonna reboot and then try skyrim again, this time with those diagnostics applications open. fingers crossed.

_________________
RIP Shiloh - I'll never forget ya

I tried and after about 5 minutes, my computer died. Drivers are updated, fans are working, it's not overheated. What am I going to do? I might as well go back to my 620 card and hope Tiger Direct has a refund policy

May not be able to play FF14 with you guys since that old graphics card was soooo slow

_________________
RIP Shiloh - I'll never forget ya

I had this problem the other week actually Whisp in a PC. I had a computer that, after I swapped the graphics card, would turn off after about 10-15 minutes.

Like you, I assumed the card was at fault in one way or another since it definitely wasn't overheating. Or I thought that perhaps my PSU (it was relatively small for the card I was putting in) couldn't handle the load so it was cutting out.

In the end, I put the old graphics card back in. Only to have the computer do exactly the same thing again, and turn off after 15-20 minutes or so.

So, evidently, it had nothing to do with the card, and I must have knocked something else inside the case. I swapped the cards back over, but not before making sure all the fan headers were re-secured, and the graphics card was properly seated and not loose at all. I cleaned out the dust, I tidied the cables, I even re-seated the RAM.

Computer then worked fine with the new card. No problems since.

Though I can't say for sure exactly what the problem was, I can say that there are so many things it could be that it's never right to just claim that a particular piece of hardware simply doesn't work. The only way to be sure of that is to have a second PC to hand you can put it in and try it there.

Reasons for a computer to *switch off*, not crash or reboot:

- Overheating, but we've covered this already.

- Overloaded power supply. But as I recall we already checked and your new card drains no more power than your old (good) card did.

- Motherboard/CPU safety measure. If the motherboard cannot sense the CPU fan, either the cable isn't connected or the CPU fan isn't spinning for some reason, it will immediately power down the computer.

- A power "short". Not a shortage, I mean an electrical short, which is where two ends of a power supply are connected with nothing between them. This can be caused by having a screw rolling about in your case, or by having a power cable in the wrong spot and slightly touching something with it's contacts, or by having an expansion card not quite in the slot properly. Simply reseating things inside your computer (taking them out and putting them back in) can fix this.

- Dust. Linked with the above. Having dust in some of your slots can cause a short. If you've recently taking something out and put something else back in, it's possible for some dust to have blown or been moved into the slots in the meantime. Good idea to have a compressed air can when you're fiddling with a computer that hasn't been cleaned in a while.

- Stuck power button. I know this one sounds stupid but it happens. If the power button is held down, the computer will usually force a power off. I've come across computers with a stuck power button that stay on once turned on because the button is stuck, but the moment the button unsticks and releases, the computer goes off. We're talking micro-switches here so it only has to be very very slight movement.

There are no other reasons I can think of off the top of my head for the computer to simply turn off. Every other circumstance would cause crashes, reboots, blue screens, etc. But to have the power supply simply cut out and your machine to turn off like you pulled the plug, it has to be one of the above.

Another thing to try: if the computer turns off, immediately turn it back on without waiting. If it cut off due to some kind of safety measure, it will usually not turn back on straight away, or will come on for a second then go straight back off again.

If you're absolutely certain the card is defective (like I say, having a second PC to try it in helps a lot here) then you'd be well within rights to simply have the card replaced.

That's never a good sign. Do you mind taking a few pictures of your PC case and posting them in here so I can get an idea of what we're dealing with?

Take a few pictures:

1. Lay the case on its side flat on the floor and take the side panel off, take a couple of pictures so you can see a good majority of whats inside.

2. Take a picture of the back of the case while it's standing up, with the graphics card plugged in.

Edit: Does your PSU (Power Supply) sit in the top of your case or at the bottom, a while back I had a similar problem with an old PSU. It was bottom mounted (fan facing down) and I could smell burning rubber. It turned out that the PSU fan wasn't working and was melting away the rubber insulation inside the PSU. Double check to make sure your fan is spinning.

Edit2: Also you did plug a 6 pin into the back of your graphics card right? all pins in the back card are filled? putting a 4 pin into a 6 pin adapter could get undesired results.

_________________

You don't happen to have chinchillas do you?

Here is my huge new video card, the Nvidia GEForce GTX 660.


The back of my computer, with the new video card installed. I noticed that it feels upside down, but there was no other way to install it.


My computer from a distance, lying on its side with video card still installed (light gray thingy).


Closer up. You can see EVGA written on it.


Here you can see it attaches to a blue strip of plastic.


The power supply connects to the video card via a 6-pin to 4-pin adapter. It's kinda hard to make it out but the white part (adapter) attaches to the black part (4-pin, from the power supply). Then the other end of the adapter is a 6-pin. Hold on computer acting funny, gonna post and edit more. In case it crashes.

_________________
RIP Shiloh - I'll never forget ya

Alright so I removed my card. By the way I dusted earlier, so that wasn't what caused the malfunction. I see only the one strip for placing a video card.


Here is my power supply 700W


I found this little guy tucked away behind some cables. Maybe I can just plug it directly into my video card?


It took some tugging. That's what she said. But I got it to fit.


So I rebooted, I also uninstalled the driver software and reinstalled it. Yet I've been getting this update symbol every time I reboot.


Windows Update is panicking over my video card. Error Code 800F020B and 80070103


Do you think it's okay to just ignore this error?

_________________
RIP Shiloh - I'll never forget ya

I was able to play Skyrim for 20 minutes at medium graphics settings. So i cranked it up to high (the suggested settings according to Steam) and the computer crashed within 10 seconds. Now it wont turn back on. I think I killed it.

_________________
RIP Shiloh - I'll never forget ya

Expansion cards are always upside down (since they moved away from ISA/AGP 10-15 years ago, anyway). Don't worry about that.

The cable on the fan on the back of the case bothers me. It won't hurt it at all, but I think I can see it has one of the small fan headers on the cable too, in which case it'd be better to connect it to one of the motherboard's fan headers. That'll allow the motherboard to regulate the fan. Don't worry about this for now though, it has nothing to do with your problem.

Is the CPU's fan cable stuck under/on the motherboard chipset heat sink? If so, move it if you can. Those heat sinks can get pretty hot.

The red 6-pin PCI-e connector was fine to use. Those plugs have shaped pins, and will not fit at all if they're the wrong connector. If it fit, it was right. That said, check what else is connected to the same rail. Like, what other things are connected to the same cable as that plug, if anything. It's recommended that your graphics cards have a dedicated rail (don't share it with a CD drive or a HDD or something) because of the massive power draw. Also, try not to share the rail that fan is using either, they cause unstable rails due to the spinning of the motor.

The reason for the Windows Update thing is that it downloaded the Windows Update before you installed your own driver, and now the windows update version of the driver will fail to install because it's not as new as the one installed. To fix that, go into Windows Update, find the update it's trying to install and tell it you don't want that update anymore (I believe they call this "Hide Update" these days, from right clicking the update in the list.)

Is that long yellow cable your auxiliary P4 cable? You really ought to route that cable somewhere other than under your graphics card. It doesn't look long enough to go over the top, mind. It probably doesn't hurt it, but it's a pretty important cable (it powers your CPU).

As for the computer being dead:

Firstly, you said it "crashed". Did you get an error? A blue screen? Did the graphics garble? Did it freeze? Or did the computer simply immediately turn off with no prior warning or odd thing happening? This is pretty important when figuring out what happened.

You said it wouldn't turn on. Do you mean pressing the button does nothing? Or pressing the button makes it come on and then immediately turn off again? Or perhaps it simply won't POST (the nice little "everything's working" beep when you turn computers on) now?

Will it turn on now?

If it does. It probably cut off due to overheating. And wouldn't come back on until the temperature dropped.

If it won't come on, cut the mains supply to it (switch on back of case, or pull the plug), wait a few seconds, put it back in/on. Will it turn on now? If it does, it was definitely a safety cut out, meaning it was most likely a short or unstable power supply.

Another possibility: The PSU is 700W but that doesn't mean it's a "good" 700W. Power supplies come in all sorts of varieties and it's possible to have one with a high wattage, but with crappy amps on each rail. Although there are enough watts, when something tries to draw too much power it'll overload the rail, causing the voltages to drop which is enough to make everything mess up. Usually those power supplies are cheap, so you can tell just from how much the power supply cost. I find it odd because your old (good) graphics card worked fine and likely had a similar power drain though.

As a final measure. It is of course possible that the graphics card is defective. If you had another PC around I would recommend you try it in that. But if not, you can try putting your old card back in and see if everything is fine again.

As I explained before, I did this and discovered that everything wasn't fine all of a sudden. These are very different cards though, and if it's a power draw or temperature problem, the old card probably wouldn't trigger it.

If the card is defective, can you get it replaced? Here in the UK we have a lot of consumer protection laws preventing companies screwing you over for defective gear. In general, stores here have to replace your stuff on the spot, they can't just make you go chase it up with the manufacturer yourself (though some will try, its your job to know the law and tell them no, replace my item).

I feel bad that you're having some bad luck here though.

That was my guess a defective card I had the same problem with my GPU I tried everything and finally figured out why my games went black when I played them it was the damn GPU!

_________________

Oh god the computer still won't turn on the next morning. I'm fucked.

_________________
RIP Shiloh - I'll never forget ya

Give me a moment to try some of those steps

_________________
RIP Shiloh - I'll never forget ya

1. No it wasn't a blue screen type software crash. It just shut off. All four times this happened yesterday I also noticed my ceiling lights flickered, as if they were all affected by my computer. Also it made a quick bzhoooo noise as if a galka stepped on a taru.

_________________
RIP Shiloh - I'll never forget ya