Overclock.net › Forums › Cooling › Water Cooling › 45C/90C Watercooled...
New Posts  All Forums:Forum Nav:

45C/90C Watercooled...

post #1 of 53
Thread Starter 
Just when I finally think I can use my computer, something else goes wrong. I am so utterly sick and tired of this crap. If anyone wants to buy this rig, I'll sell it and you can deal with it. Here's issue #9646218:

Finally plugged in my computer the other day. I just finished installing Windows 7 and formatting my second SSD earlier today. Came home tonight with Black Ops, TF2, and BC2 installed. So excited.

I turned all of my video card settings all the way up high, since I never really had the chance to do that on a 7900 GTX Go. Antialiasing at 64x SLI, Transparency at 16x supersampling, Anisotropic at 16x. I was ready to go.

Started playing Black Ops with 80-120 FPS in 1920x1200 and about a minute into it my computer shut off. I said to myself...okay well it can't be the temperature cause it's watercooled. Maybe it's the BIOS or drivers or something. I updated my BIOS and reinstalled drivers.

So I might as well try out BC2. Heck, it coulda just been Black Ops right? Nope. BC2 shut down a couple minutes in as well. I got mad so I lowered my NVIDIA graphics through their tool. Anisotropic and EVERYTHING off or as low as possible. Then I loaded up TF2. Couple minutes in...crash.

I got real mad so I installed EVGA precision. Before I reveal that, let me explain my specs:

2xEK-FC580 waterblocks with an EK Dual Parallel attached by 2 EK-FC GeForce Links. Expensive but I figured it'd be worth it. I also have PLENTY of airflow throughout the case, so that's not the issue.

So here's my EVGA specs as of right now while posting in this thread:

GPU1 Temperature: 45C
GPU2 Temperature: 39C

The lowest I've seen them is 43C and 37C respectively. I loaded up TF2 and saw the temperatures SKYROCKET. During constant playing, I monitored closely and it reached 93C in a matter of 120 seconds. It would have gone higher I assume but I left the game at that point. No reason this should be at 93C at low graphics.

Did I mention nothing is overclocked?

At this point I was mad and wanted to listen to some music so I loaded up YouTube for some Bravery. You would think that would mean GPU temperatures would have no effect...oh and by the way, did I mention MY CPU TEMPERATURE RUNS FROM 20-27C (for all the cores) during this process? I don't understand. Here are the results:

GPU1 Usage: 0%
GPU2 Usage: 0%

Anyways, I monitored a nice gradual increase from the previous posted temperatures all the way to:

GPU1 Temperature: 80C
GPU2 Temperature: 70C

I am sitting here now with Core Temp opened with a YouTube screen, this thread, and EVGA precision...my temperature keeps rising. I'm about to shut down. GPU1 is at 86C and rising.

Ambient temperature is about 25C and CPU is still running at 27C max, 21C lowest.

We're at 88C GPU now.

17705130.png

I exit my YouTube and CoreTemp. GPU is dropping quite rapidly.

Sorry for the long post guys but I just don't know if I can handle this anymore. I have had so many issues along the way that were solved with a lot of help from you guys, some intelligence on my part, and it wasn't too big of a deal.

I feel like this is just going to push me over the edge. I honestly can't handle it anymore. I seriously feel like I should sell this to someone here who might know what they're doing and just buy an Alienware...it's that annoying.

Does anyone have any suggestions or any way at all that can help me get through this without having to do anything drastic? I just don't have the time or knowledge to take this apart again and troubleshoot. There has to be an explanation. Someone smarter than me please help.
post #2 of 53
Something is definitely wrong. Does your board have one of those switches to turn off the PCI slot so you don't have to take the card out? Is there a way you could test one card at a time without disconnecting them from the loop? Maybe just pop it out and rest it on something? I'm assuming you know what your doing with your loop as in proper rad size and blocks etc.
The Butthurt Box
(21 items)
 
   
CPUMotherboardGraphicsGraphics
Intel Core i5 2500K P67 Professional AMD Radeon HD 6950 unlock NVIDIA GeForce GTX 460 2gb 
RAMRAMHard DriveHard Drive
G.Skill Ripjaws G.Skill Ripjaws Intel X25-V Seagate Barracuda 
Hard DriveOptical DriveOSMonitor
WD Goflex USB 3.0 24x DVD R/W DL Win7 Ultermate 64 Acer 20in 1600x900 
MonitorMonitorMonitorKeyboard
Compaq 20in 1600x900 17in Dell 1280x1024 19in Gateway 1280x1024 Razer Blackwidow mech w/ cherry MX blues 
PowerCaseMouseAudio
Kingwin LZ-750 HAF 932 + sound activated lights Logitech G500 Rosewill USB headset 
CPUMotherboardGraphicsRAM
Pentium 4 3.2ghz w/ HT HP Intel IGP 4gb Mushkin DDR2 
Hard DriveOSPowerCase
80gb sata WinXP Pro HP OEM HP Small form factor 
  hide details  
Reply
The Butthurt Box
(21 items)
 
   
CPUMotherboardGraphicsGraphics
Intel Core i5 2500K P67 Professional AMD Radeon HD 6950 unlock NVIDIA GeForce GTX 460 2gb 
RAMRAMHard DriveHard Drive
G.Skill Ripjaws G.Skill Ripjaws Intel X25-V Seagate Barracuda 
Hard DriveOptical DriveOSMonitor
WD Goflex USB 3.0 24x DVD R/W DL Win7 Ultermate 64 Acer 20in 1600x900 
MonitorMonitorMonitorKeyboard
Compaq 20in 1600x900 17in Dell 1280x1024 19in Gateway 1280x1024 Razer Blackwidow mech w/ cherry MX blues 
PowerCaseMouseAudio
Kingwin LZ-750 HAF 932 + sound activated lights Logitech G500 Rosewill USB headset 
CPUMotherboardGraphicsRAM
Pentium 4 3.2ghz w/ HT HP Intel IGP 4gb Mushkin DDR2 
Hard DriveOSPowerCase
80gb sata WinXP Pro HP OEM HP Small form factor 
  hide details  
Reply
post #3 of 53
Thread Starter 
Quote:
Originally Posted by railfan844;12430259 
Something is definitely wrong. Does your board have one of those switches to turn off the PCI slot so you don't have to take the card out? Is there a way you could test one card at a time without disconnecting them from the loop? Maybe just pop it out and rest it on something? I'm assuming you know what your doing with your loop as in proper rad size and blocks etc.

I'm not too sure about the PCI-e slot but unfortunately I can't take the card out without disconnecting the entire loop.

The cards are connected in unison with a dual parallel. If one comes out, they both come out. This is just so upsetting.
post #4 of 53
i would check if everything is seated correctly.

you just got those so there is no possible way for this to be happening. even when running with stock coolers.

check it again and let us know
The Cleaner
(14 items)
 
  
CPUMotherboardGraphicsRAM
I7 980X UD-9 GTX 480 SLI EK BLOCKS Corsair Dom GT 2000MHZ TR3X 
Hard DriveOSMonitorKeyboard
Lots of UBER PR0N GIGS W7 H223BHMID Main Screen U2410x3 (storage for now) Filco Majestice Linear R 
PowerCaseMouseAudio
Corsair 1000HX HA a JOKE RATz9 Titanium HD 
  hide details  
Reply
The Cleaner
(14 items)
 
  
CPUMotherboardGraphicsRAM
I7 980X UD-9 GTX 480 SLI EK BLOCKS Corsair Dom GT 2000MHZ TR3X 
Hard DriveOSMonitorKeyboard
Lots of UBER PR0N GIGS W7 H223BHMID Main Screen U2410x3 (storage for now) Filco Majestice Linear R 
PowerCaseMouseAudio
Corsair 1000HX HA a JOKE RATz9 Titanium HD 
  hide details  
Reply
post #5 of 53
With ambient at around 25C, your CPU temp looks normal while GPU idle temps are not.
I suspect the waterblocks+paste to the graphics cards are not installed properly.

First, in order to eliminate the possibility of something being wrong with the watercooling system, you can stress test your CPU and see if the temp is OK. If it is, this means the radiators, pump....etc should be good.

I suppose this is a one-loop setup? Can you show a diagram of how the flow goes?
post #6 of 53
Thread Starter 
Thanks for trying to help guys. Honestly if you can help me I'll send you a check, I'm not even kidding.

How do I run a stress test on the CPU? Is there a link to a download for something somewhere?

This is an awful diagram and it's really hard to draw because the case is 2 sided. I tried. On the left side is the CPU and GPU. Obviously the top is the radiator.

On the right side is the reservoir, pump on the bottom, and radiator.

37873002.png

As you can see, the reservoir goes down into the pump. It is a dual pump setup but I am only running one for now. The pump pushes that long tube up into the radiator. The radiator goes from one side of the case (marked by the circle) into the other side of the case into the other radiator. That radiator falls down into the CPU, which goes down to the GPU. The GPU goes down through a hole and up a few inches into the side of the reservoir.

There are some better pics in my buildlog thread: http://www.overclock.net/water-cooling/939431-build-log-x1.html

I am just sick of all of the problems I've had. I really don't want to have to take it apart again because of who knows what.

I'm tired so I'm gonna hit the sack but please help me if you can guys. Please.
post #7 of 53
I idle around the same temps with my three 470's, but they rarely get more than 10C hotter than idle when I'm gaming. If the CPU temps are fine but the GPU temps are skyrocketing..., this is strange to me because if you had messed up with the TIM the idle temps would be a lot higher than they are, wouldn't you think?

Don't give up though. I've had leaks, dead parts (after my loop is bled), and all sorts of other issues. My rig is finally 100% now though and it was totally worth all the troubles I've gone through.

Have you ever tried those 580's with their stock air cooling? I'm curious as to how the temps are with those on there. If they are normal temps then it's obviously something going on with the waterblocks on the cards. Just a thought.
post #8 of 53
Don't give up though. I've had leaks, dead parts (after my loop is bled), and all sorts of other issues. My rig is finally 100% now though and it was totally worth all the troubles I've gone through.

Have you ever tried those 580's with their stock air cooling? I'm curious as to how the temps are with those on there. If they are normal temps then it's obviously something going on with the waterblocks on the cards. Just a thought.

^this^ Really hope you get this sorted out because thats a killer rig you have there!cool.gif
Pilgrim
(19 items)
 
  
CPUMotherboardGraphicsRAM
i7 6700k (4.8ghz@1.27v) Asus Maximus VIII Hero MSI GTX 980 Ti Twin Frozr 6G Corsair 16gb DDR4 Vengeance LPX 
Hard DriveHard DriveHard DriveHard Drive
Samsung 950 Pro M2  Samsung 850 Pro  Samsung 840 Evo  Samsung 840 Pro 
Hard DriveHard DriveOptical DriveCooling
WD Caviar Black Seagate Barracuda 1 x Samsung SH-222A Corsair H100I 
OSMonitorKeyboardPower
Windows 10 Pro x64 Rog Swift PG278Q Corsair K95 RGB Corsair AX860i 
CaseMouseMouse Pad
Fractal Design XL R2 Logitech G502 Proteus Core Logitech G440 
  hide details  
Reply
Pilgrim
(19 items)
 
  
CPUMotherboardGraphicsRAM
i7 6700k (4.8ghz@1.27v) Asus Maximus VIII Hero MSI GTX 980 Ti Twin Frozr 6G Corsair 16gb DDR4 Vengeance LPX 
Hard DriveHard DriveHard DriveHard Drive
Samsung 950 Pro M2  Samsung 850 Pro  Samsung 840 Evo  Samsung 840 Pro 
Hard DriveHard DriveOptical DriveCooling
WD Caviar Black Seagate Barracuda 1 x Samsung SH-222A Corsair H100I 
OSMonitorKeyboardPower
Windows 10 Pro x64 Rog Swift PG278Q Corsair K95 RGB Corsair AX860i 
CaseMouseMouse Pad
Fractal Design XL R2 Logitech G502 Proteus Core Logitech G440 
  hide details  
Reply
post #9 of 53
Thread Starter 
Quote:
Originally Posted by Methos07;12430702 
I idle around the same temps with my three 470's, but they rarely get more than 10C hotter than idle when I'm gaming. If the CPU temps are fine but the GPU temps are skyrocketing..., this is strange to me because if you had messed up with the TIM the idle temps would be a lot higher than they are, wouldn't you think?

Don't give up though. I've had leaks, dead parts (after my loop is bled), and all sorts of other issues. My rig is finally 100% now though and it was totally worth all the troubles I've gone through.

Have you ever tried those 580's with their stock air cooling? I'm curious as to how the temps are with those on there. If they are normal temps then it's obviously something going on with the waterblocks on the cards. Just a thought.

Thanks for the post. It's unfortunate because generally, 470's run a bit hotter than 580's. AND you're running 3 of them. Ideally I believe I should be idling at < 30 and loading out no more than 50.

I just ran a CPU stress test. I'm not quite sure if I did it right but I ran the IntelBurn Test at Maximum for 1 run. I left threads at "Auto" and my temperature never exceeded 44C. Minimum was 37C. It took like 3 minutes to complete though. Is that indicative of something? Is that bad?

My CPU idle temperature is 17C-22C...in a room with ambient temperature at 25C. Now that I'd like to figure out as well.

Could it be the fitting inside of the CPU block restricting flow to the GPU? I don't see that as possible because then it would also restrict flow to the CPU. I hate this.

I honestly don't think that makes sense. CPU on load is < 40C. GPU on idle is more than that. Why?

Like I don't get it? Could it be drivers? I HATE THIS. I just want a functional computer.
Edited by xyeLz - 2/18/11 at 1:15am
post #10 of 53
Quote:
Originally Posted by xyeLz;12430767 
Thanks for the post. It's unfortunate because generally, 470's run a bit hotter than 580's. AND you're running 3 of them. Ideally I believe I should be idling at < 30 and loading out no more than 50.

I just ran a CPU stress test. I'm not quite sure if I did it right but I ran the IntelBurn Test at Maximum for 1 run. I left threads at "Auto" and my temperature never exceeded 44C. Minimum was 37C. It took like 3 minutes to complete though. Is that indicative of something? Is that bad?

Could it be the fitting inside of the CPU block restricting flow to the GPU? I don't see that as possible because then it would also restrict flow to the CPU. I hate this.

I honestly don't think that makes sense. CPU on load is < 40C. GPU on idle is more than that. Why?

OK, I see that your CPU load temp looks normal.

And, your loop goes like: res-->pump-->radiator-->CPU-->(GPU1+GPU2 in parallel)-->res
And this also looks ok.

So, I am suspecting even more now that the 2 waterblocks+paste are not installed properly.
I have not gone through your build log yet. I do not now if you have used QDCs before and after important components (eg the graphic cards) so that a component can be taken out easily. I hope you have.

Wait a day to see if anyone has any bright idea. If not, I am afraid the next logical step is to take out the cards for examination.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Water Cooling
Overclock.net › Forums › Cooling › Water Cooling › 45C/90C Watercooled...