Sign in with
Sign up | Sign in
Your question

GTX480 on XClio Greatpower 850w

Last response: in Graphics & Displays
Share
October 14, 2010 7:01:02 PM

So my question is because this PSU is 12v1 18A; 12v2 18A; 12v3 30A; 12v4 30A is this a problem with my 480. I have it wired using the 2 hardwired 8pin / 6pin leads. Using 1 8pin off of one lead and using the 6pin off of the second lead (both leads have an 8 pin and 6pin. I have been having COMPLETELY random crashes (NO BSOD) just the whole rig locking up (Gaming and internet video usage).

I am 99% sure it is not an overheat issue. The reason I say 99% is because I have everything water cooled (CPU, Northbridge, GPU (full coverage EK Block). The loop has a koolance 3 x 120mm rad and a single 120mm rad. I am air cooling the CPU Mosfet and the southbridge. The problem with my Mosfet is that EVGA sent me a FTW edition back from RMA and the Mosfet sink didnt line up with the mounting holes so I am using 3m thermal tape for the Thermalright HR-09 sink.

This is mind numbingly frustrating.......

The rig is:
EVGA 780i SLI FTW (SZ17 Bios; 15.53 Driver NVidia driver not EVGA)
q9650 (NO OC yet)
8GB geil black dragon ddr2-800 4-4-4-12 (4 sticks) (Tried with only 2 and still crash randomly)
EVGA GTX480 Driver 260.63
XCLio Greatpower 850w http://www.xclio.com/products/products-psu-great-850w.h...
October 14, 2010 7:11:05 PM

have you ran memtest and/ or prime 95 to check for errors?

plus it's never a bad idea to look at the actual temps, you may have incorrectly applied the thermal paste when mounting the waterblock or something
m
0
l
October 14, 2010 7:15:01 PM

First thing I would do is install MSI AfterBurner or another GPU temperature monitoring software. I would do this to ensure that you are not getting heat issues due to a poorly installed heat sink. It is always best to cover every base. Next try plugging a different power lead into the GPU. Is your computer plugged straight into the wall? Is there a surge protected power bar? What about an extension cord? If there is an extension cord, try to find a way to remove it from the loop. Also, if you can't remove the extension cord, try upgrading to a better extension cord. If you don't have a surge protected power bar you may wish to invest in one, the could have been a power failure that damaged your PSU or another component in your computer. You could have a memory issue, this can be something that is disruptive in an unexpected way. Try completing a memory test. Try these out and I can suggest other ideas if necessary.
m
0
l
Related resources
Can't find your answer ? Ask !
October 14, 2010 7:26:23 PM

So i have been running EVGA Precision. As well as Everest ultimate which gives me great reading for everything, temp wise right now (idle) temps are GPU 35° (underload I have seen as high as 51° under 3dMark Vanatage) GPU PCB temp is 29°. The Northbridge is showing 44°, Southbridge 54° and CPU 36° ( the weird thing about Everest is it is showing Core1 33°, Core2 30°, Core 3 32°, core 4 38°. Is this normal or is my CPU Block potentially not seated properly?)

The computer is plugged into an extension cord coming off of a Minuteman 3000VA true online UPS via 25' 12Gauge extension.

Will run a memory test and see..... I just installed all 4 sticks of mem now to do that.

Thanks guys I appreciate the help here, it awesome! :) 
m
0
l
October 14, 2010 7:38:50 PM

Sounds like the temps and power in should be good, let us know about the memory.
m
0
l
October 14, 2010 9:34:53 PM

Soooo memtest86+ finished 2 passes with 0 errors........ Unless this PSU is having random PFC problems or a rail has an intermittent issue is kind of where I am left thinking... I dont have another 850w PSU... I do have an Antec TruePower 650 I could run the rig on..... It is currently my 2nd PSU in this rig. I have it running all the fans, bay devices, water cooling and lighting for the case. The only thing the XClio is running is the mobo and videocard.

As far as the memory goes I am manualy setting the timings of 4-4-4-12, the bios's "optimal" setting is 5-5-5-12. I bumped voltage up to 2.050v mem, 1.5v southbridge, 1.6v north
m
0
l
a b Î Nvidia
October 15, 2010 1:29:39 AM

You can certainly try your antec 650W PSU just to rule out a power supply problem. Wouldn't be a bad idea.

Though I'm curious, with an 850W PSU and a single GPU, why are you using two PSUs in the first place? Do you not have enough power connections on that 850W to power everything?

Also, for troubleshooting purposes, you could easily set your RAM to it's recommended stock timings, even if memtest is testing properly. Just to rule that out as well.

EDIT: Also, make sure those PCI-E 8pin power connectors are plugged into the Red modular jacks on the power supply. Those are most likely the intended GPU connections (i.e the 12V @ 30A lines).
m
0
l
October 15, 2010 2:24:13 AM

The reason I had 2 psu's was because I was running tri-sli before. 2 gtx280's and an 8800gts g92. That and I had the true power sitting arround so it was a fun project!

Thats a great point about the socketed lines instead of the hard wired sets. Now to find the cables........
m
0
l
October 19, 2010 6:32:49 AM

So now that I have exhausted the PSU problems.... I tried the red socketed lines and still will crash (No BSOD just lock up)... So I thought maybe it is a thermal issue. So I put a 20" box fan on the side of my case and I can game for hours on end no problems....... Now the problem probably is in the mosfets because on this board there is no after market mosfet sinks that match up to the mounting holes (they are different on this board than a standard 780i). I used 3m thermal tape to sort of secure the HR05 to the mosfets. This leaves me with the option of thermal epoxy. Unless someone could help me out and let me know if there is a better option?
m
0
l
!