My alienware randomly shuts off, any help appreciated!

[Netflixandchill]

Hi all,

So I recently attempted to upgrade my 6 year old alienware aurora r3 with a gtx 970 and an EVGA 750 G2 psu. The previous psu was the stock psu that came with the computer and was not modular. I'm not sure if I incorrectly installed the psu, but there is obviously a problem. Roughly 90 seconds after startup the computer will shutdown, no warning, no error beeps. It seems like somebody just pulled the plug out. I figured the cpu was overheating so I removed the stock cooler, cleaned off the thermal paste, applied new thermal paste and reinstalled the heatsink, and yet it still powers off after 90 seconds. I can't tell if the radiator fan is running or not. Would it possibly have something to do with the strip of fan connectors etc. on the top of the case not being powered? How would I do that without having the original non-modular power connector? Thanks for any help or advice, I appreciate it!

 

[cheapo gamer]

after typing this i actually googled your pc , it is watercooled and runs from that strip of fan connectors so plug them in


Hello , could you provide pictures of the computer just to see if you could be missing anything , also turn off your pc at the wall do not unplug it and hold the power button in for 10 seconds and try rebooting , make sure all the connections are plugged in , there should be a 4+4 pin socket(not a 6+2) for the motherboard and a 20+4 socket plugged into the motherboard make sure the wires are pushed in properly not just the adapters at both ends , with a g2 psu the plugs should be labeled , take out the gpu and blow down the pci socket and re-fit it , at worst take out the odd stick of ram or reset bios and check ram voltages , is your cpu air or watercooled it could be cutting off because your pump or fan might not be running if its the stock heatsink there should be a cable coming from the cooler you should plug that into a normally white or black or brown straight plastic stem with 3 or 4 pins stick out from the motherboard near the cooler area , i dont mean to give you many options you might have tried already but these are diagnostic options i would use

 

[SkyMembrane]

What were the original specifications of your R3? These came in a variety of different configurations and it would help to know what you originally had in it.

It could be a heat problem, but it looks like a PSU problem to me. If you want to check your CPU's temperature, use a program like HWmonitor64. It's not always 100% accurate, but it should give you an idea of what it looks like. Alienwares are always tricky when it comes to upgrades, and if I'm correct, the R3 originally used ATI GFX cards. It could be lack compatibility with NVidia's GPUs and be the cause of the crash.

 

[Netflixandchill]

after typing this i actually googled your pc , it is watercooled and runs from that strip of fan connectors so plug them in


Hello , could you provide pictures of the computer just to see if you could be missing anything , also turn off your pc at the wall do not unplug it and hold the power button in for 10 seconds and try rebooting , make sure all the connections are plugged in , there should be a 4+4 pin socket(not a 6+2) for the motherboard and a 20+4 socket plugged into the motherboard make sure the wires are pushed in properly not just the adapters at both ends , with a g2 psu the plugs should be labeled , take out the gpu and blow down the pci socket and re-fit it , at worst take out the odd stick of ram or reset bios and check ram voltages , is your cpu air or watercooled it could be cutting off because your pump or fan might not be running if its the stock heatsink there should be a cable coming from the cooler you should plug that into a normally white or black or brown straight plastic stem with 3 or 4 pins stick out from the motherboard near the cooler area , i dont mean to give you many options you might have tried already but these are diagnostic options i would use

Thanks for the response! I already tried reseating most of the components so I'm pretty sure that's not the issue. Hopefully these pictures help:

http://
Also, how do I connect the fan headers, when the strip was originally connected to a power cable that was intertwined with all the other ones in the old non -modular psu?

 

[Netflixandchill]

What were the original specifications of your R3? These came in a variety of different configurations and it would help to know what you originally had in it.

It could be a heat problem, but it looks like a PSU problem to me. If you want to check your CPU's temperature, use a program like HWmonitor64. It's not always 100% accurate, but it should give you an idea of what it looks like. Alienwares are always tricky when it comes to upgrades, and if I'm correct, the R3 originally used ATI GFX cards. It could be lack compatibility with NVidia's GPUs and be the cause of the crash.

Thanks for the response! ! My original specs were as follows:

i7 920 bloomfield @ 2.6 ghz
gtx 260
6 gb ddr2 (I think) ram
Stock water-cooling on the cpu
stock psu - not sure how many watts, but sufficient for the setup
a 1TB HDD at 7200 RPm

I'm not quite sure how to check the temp on my cpu/gpu when the computer only runs for about 90 seconds
Also, how do I connect the fan headers, when the strip was originally connected to a power cable that was intertwined with all the other ones in the old non -modular psu?

 

[cheapo gamer]

it looks like a sata socket the wide cream male plug looking thing on the fan power board , if it doesn't fit grab the cable that goes from the cooler to the board from the fan board and plug it into the white socket below your ram but that will only power your waterpump , you will need a fan controller

 

[SkyMembrane]

What were the original specifications of your R3? These came in a variety of different configurations and it would help to know what you originally had in it.

It could be a heat problem, but it looks like a PSU problem to me. If you want to check your CPU's temperature, use a program like HWmonitor64. It's not always 100% accurate, but it should give you an idea of what it looks like. Alienwares are always tricky when it comes to upgrades, and if I'm correct, the R3 originally used ATI GFX cards. It could be lack compatibility with NVidia's GPUs and be the cause of the crash.

Thanks for the response! ! My original specs were as follows:

i7 920 bloomfield @ 2.6 ghz
gtx 260
6 gb ddr2 (I think) ram
Stock water-cooling on the cpu
stock psu - not sure how many watts, but sufficient for the setup
a 1TB HDD at 7200 RPm

I'm not quite sure how to check the temp on my cpu/gpu when the computer only runs for about 90 seconds

Right, you should be able to see the temperature for the CPU in the system bios, along with the fans current RPMs. Did you ever update your Bios version? If you didn't, it is possible that it can't recognize the new GTX 970 and that leads to the motherboard shutting down the computer for safety reasons. If you haven't updated the Bios recently, you should put the GTX260 back in the computer and follow Dell's instructions for bios updates. Every bios are different when it comes to updates, and I haven't owned an Alienware, so I can't tell you exactly what steps to follow. If that doesn't solve the problem, it may be because Alienware's motherboard isn't compatible with that upgrade. It shouldn't be that, but with those pre-built machines, you never know what they did to them.

 

[Netflixandchill]

it looks like a sata socket the wide cream male plug looking thing on the fan power board , if it doesn't fit grab the cable that goes from the cooler to the board from the fan board and plug it into the white socket below your ram

I found the cream header, and it has a black header to the fan board, I plugged them in as well as another small header (not sure what it does) and hoped for the best. Unfortunately the fans still aren't spinning

 

[cheapo gamer]

no it will only power your pump so your temps will climb slower this time but hope it will boot to confirm its only a power issue then you can get a fan controller and be happy or find a way to use the old wiring loom on your new psu

 

[Netflixandchill]

What were the original specifications of your R3? These came in a variety of different configurations and it would help to know what you originally had in it.

It could be a heat problem, but it looks like a PSU problem to me. If you want to check your CPU's temperature, use a program like HWmonitor64. It's not always 100% accurate, but it should give you an idea of what it looks like. Alienwares are always tricky when it comes to upgrades, and if I'm correct, the R3 originally used ATI GFX cards. It could be lack compatibility with NVidia's GPUs and be the cause of the crash.

Thanks for the response! ! My original specs were as follows:

i7 920 bloomfield @ 2.6 ghz
gtx 260
6 gb ddr2 (I think) ram
Stock water-cooling on the cpu
stock psu - not sure how many watts, but sufficient for the setup
a 1TB HDD at 7200 RPm

I'm not quite sure how to check the temp on my cpu/gpu when the computer only runs for about 90 seconds

Right, you should be able to see the temperature for the CPU in the system bios, along with the fans current RPMs. Did you ever update your Bios version? If you didn't, it is possible that it can't recognize the new GTX 970 and that leads to the motherboard shutting down the computer for safety reasons. If you haven't updated the Bios recently, you should put the GTX260 back in the computer and follow Dell's instructions for bios updates. Every bios are different when it comes to updates, and I haven't owned an Alienware, so I can't tell you exactly what steps to follow. If that doesn't solve the problem, it may be because Alienware's motherboard isn't compatible with that upgrade. It shouldn't be that, but with those pre-built machines, you never know what they did to them.

I just tried to re-install the gtx 260 but the computer wouldnt boot. It gave me a 6 beep error, which apparently is the video card error message... You're right, the BIOS is ancient - it hasnt updated since 2009.

 

[SkyMembrane]

it looks like a sata socket the wide cream male plug looking thing on the fan power board , if it doesn't fit grab the cable that goes from the cooler to the board from the fan board and plug it into the white socket below your ram

If the problem is a cable plugged at the wrong place, the motherboard should shutdown automatically before leaving the Bios, not allow a full boot and crash after 90 seconds.

If the CPU wasn't overheating before you changed those parts, there are no reasons for it to be overheating now, unless the pump and the fans aren't turning on. If that's the case, you should see it in the bios, since you'll see the fans at 0 RPM. Make sure the pin connector from the fans is connected to ''CPU_fan'' and not to ''CPU_opt'' because that wouldn't cause a self-check failure and would let the computer boot on even without the fans. The fans often don't need an extra power source when they are with a water cooler, but they still need to be connected to the motherboard controller. If your fans don't have a header on them to plug directly onto the motherboard's CPU_fan plug, you won't be able to use them.

Here are a couple of things that could cause issues :

A GTX 260 runs OpenGL 2.1 and a GTX 970 runs OpenGL 4.4.

The computer might be bound to its PSU, since you couldn't put the GPU back in it with the EVGA 750G2. It's even more plausible since it seems your fan headers were attached to the PSU cables.

The Alienware Bios might not be up-to-date.

The 260 and the 970 both require a PSU with a minimum of 500W to run and both need 2 x 6-pins connectors to run. This means that your old power supply should be able to power your new 970, so I would try putting it back in. The fact that it didn't boot up when you placed the 260 back means that the only piece that wasn't originally there was the PSU. Try rebuilding your computer as it originally was and see if it boots on, because right now there's not enough information coming from the computer for me to tell you exactly what's going on.

 

[Netflixandchill]

it looks like a sata socket the wide cream male plug looking thing on the fan power board , if it doesn't fit grab the cable that goes from the cooler to the board from the fan board and plug it into the white socket below your ram

If the problem is a cable plugged at the wrong place, the motherboard should shutdown automatically before leaving the Bios, not allow a full boot and crash after 90 seconds.

If the CPU wasn't overheating before you changed those parts, there are no reasons for it to be overheating now, unless the pump and the fans aren't turning on. If that's the case, you should see it in the bios, since you'll see the fans at 0 RPM. Make sure the pin connector from the fans is connected to ''CPU_fan'' and not to ''CPU_opt'' because that wouldn't cause a self-check failure and would let the computer boot on even without the fans. The fans often don't need an extra power source when they are with a water cooler, but they still need to be connected to the motherboard controller. If your fans don't have a header on them to plug directly onto the motherboard's CPU_fan plug, you won't be able to use them.

Here are a couple of things that could cause issues :

A GTX 260 runs OpenGL 2.1 and a GTX 970 runs OpenGL 4.4.

The computer might be bound to its PSU, since you couldn't put the GPU back in it with the EVGA 750G2. It's even more plausible since it seems your fan headers were attached to the PSU cables.

The Alienware Bios might not be up-to-date.

The 260 and the 970 both require a PSU with a minimum of 500W to run and both need 2 x 6-pins connectors to run. This means that your old power supply should be able to power your new 970, so I would try putting it back in. The fact that it didn't boot up when you placed the 260 back means that the only piece that wasn't originally there was the PSU. Try rebuilding your computer as it originally was and see if it boots on, because right now there's not enough information coming from the computer for me to tell you exactly what's going on.

Thank you for all the help! So I replaced the gtx 970 with the 260, and I realized the problem wasn't the video card, but the fans - the radiator fan wasn't spinning at all. So I attempted to go into the BIOS and check out CPU temps and fan speeds. Unfortunately the BIOS is 6 years old and doesn't show any information except the type of cpu, Ram, and speeds at which they are running. I attempted to update the BIOS but the computer did not last long enough for the process to complete. So I checked to see if there were any connectors I was missing, and sure enough, the fan board was missing it's power connector. It threw me off originally because it was a molex male connection on the board, and I did not realize that it was the power connection. After supplying power to the fan board by using a molex to six pin power connector that was included with the 750 g2, the radiator fan turned on, and the computer has been running no issues.

I downloaded the software you recommended, HWmonitor64, and it shows that idle CPU temps are 95-100 C. This seems way too high, so could it be a cooler issue? Do you know how long water cooling systems are supposed to last?

 

[SkyMembrane]

it looks like a sata socket the wide cream male plug looking thing on the fan power board , if it doesn't fit grab the cable that goes from the cooler to the board from the fan board and plug it into the white socket below your ram

If the problem is a cable plugged at the wrong place, the motherboard should shutdown automatically before leaving the Bios, not allow a full boot and crash after 90 seconds.

If the CPU wasn't overheating before you changed those parts, there are no reasons for it to be overheating now, unless the pump and the fans aren't turning on. If that's the case, you should see it in the bios, since you'll see the fans at 0 RPM. Make sure the pin connector from the fans is connected to ''CPU_fan'' and not to ''CPU_opt'' because that wouldn't cause a self-check failure and would let the computer boot on even without the fans. The fans often don't need an extra power source when they are with a water cooler, but they still need to be connected to the motherboard controller. If your fans don't have a header on them to plug directly onto the motherboard's CPU_fan plug, you won't be able to use them.

Here are a couple of things that could cause issues :

A GTX 260 runs OpenGL 2.1 and a GTX 970 runs OpenGL 4.4.

The computer might be bound to its PSU, since you couldn't put the GPU back in it with the EVGA 750G2. It's even more plausible since it seems your fan headers were attached to the PSU cables.

The Alienware Bios might not be up-to-date.

The 260 and the 970 both require a PSU with a minimum of 500W to run and both need 2 x 6-pins connectors to run. This means that your old power supply should be able to power your new 970, so I would try putting it back in. The fact that it didn't boot up when you placed the 260 back means that the only piece that wasn't originally there was the PSU. Try rebuilding your computer as it originally was and see if it boots on, because right now there's not enough information coming from the computer for me to tell you exactly what's going on.

Thank you for all the help! So I replaced the gtx 970 with the 260, and I realized the problem wasn't the video card, but the fans - the radiator fan wasn't spinning at all. So I attempted to go into the BIOS and check out CPU temps and fan speeds. Unfortunately the BIOS is 6 years old and doesn't show any information except the type of cpu, Ram, and speeds at which they are running. I attempted to update the BIOS but the computer did not last long enough for the process to complete. So I checked to see if there were any connectors I was missing, and sure enough, the fan board was missing it's power connector. It threw me off originally because it was a molex male connection on the board, and I did not realize that it was the power connection. After supplying power to the fan board by using a molex to six pin power connector that was included with the 750 g2, the radiator fan turned on, and the computer has been running no issues.

I downloaded the software you recommended, HWmonitor64, and it shows that idle CPU temps are 95-100 C. This seems way too high, so could it be a cooler issue? Do you know how long water cooling systems are supposed to last?

95-100 degrees celcius is plain impossible while idle, and even under load your CPU would probably be busted already. The temperature sensor might be showing a temperature in Fahrenheit, which would make sense, but then the temperature would be 35-38 Celcius, which is still pretty high for idle. I would recommend that you download Prime95 and run a Small FFTs stress test to see how hot it actually gets. If you start running the test and see no difference in temperature, then the sensor might be defective, so stop the test immediately to avoid damaging the CPU. If the temperature sensor really is in Fahrenheit, it will go up, but the computer won't crash. At that point, do not keep the test running if the temperature shows anything higher than 162 (72 degrees celcius).

If the temperature instantly jumps higher than the limit and doesn't immediately start to go back down, then you might have a cooler issue.

If everything is alright there, update your bios and try reinstalling your graphic's card. If Zotac included any installation CD with the 970, try installing the drivers that came with the card. If the computer is running smoothly now, you shouldn't have any problem once the GTX970 is installed.

 

[SkyMembrane]

Since I know the USA doesn't usually use Celcius, I'll give you a point of comparison. 100 degrees Celcius is Water's boiling point, which is much too high for any processors out there.

 

[Netflixandchill]

Since I know the USA doesn't usually use Celcius, I'll give you a point of comparison. 100 degrees Celcius is Water's boiling point, which is much too high for any processors out there.

That's what I figured. It was definitely in C though, it showed 100 C (212 F) on the sensor display. It was also rapidly fluctuating between 95 and 100 C for all four cores. It seemed way too high. I will definitely try Prime 95, and hopefully something makes sense. I wonder if the liquid inside the cooler could have evaporated.... Anyways thanks for all the help - I truly appreciate it!

 

[SkyMembrane]

Since I know the USA doesn't usually use Celcius, I'll give you a point of comparison. 100 degrees Celcius is Water's boiling point, which is much too high for any processors out there.

That's what I figured. It was definitely in C though, it showed 100 C (212 F) on the sensor display. It was also rapidly fluctuating between 95 and 100 C for all four cores. It seemed way too high. I will definitely try Prime 95, and hopefully something makes sense. I wonder if the liquid inside the cooler could have evaporated.... Anyways thanks for all the help - I truly appreciate it!

No problem! Fluctuations in core temperatures by 3-8 degrees at idle is fairly common, so don't worry. HWMonitor64 takes the information the sensor sends and shows it as Celcius automatically, since it's the norm. If the motherboard is in Fahrenheit, it will display it as Celcius, because it doesn't know the difference.