4 x 18A 12v rails for a GTX 260 Black Edition?

Corasik · 22 Dec 2008 at 17:49

duffbeer666 said:
The PSU is a Seasonic S12 650w and I have just noticed that it has 4 12v rails at 18A each.
I assume the PCI-E connectors on the PSU would have a rail each which would give the GPU 36A, but I assume that as with most things Seasonic exaggerate their figures slightly so it would never quite achieve its maximum output.

Seasonic are top notch PSU's. Under most lab tests they have met, or even exceeded their labels. Those quad rails have been tested by several labs, and appear to actually be a single rail anyway. The Corsair TX650 is almost exactly the same as the S12 650 (with the possible exception of slightly uprated capacitors, Seasonic is the manufacturer of the TX650) Yet the Corsair claims a single rail. Most likely Seasonic only write it as 4 rails to comply with the exact specification for ATX12V Psu's which say a maxumim of 240VA on a single rail.

It's pretty unlikely that you will overload the seasonic, and if you did manage it, Seasonic's overload protection should kick in anyway. In the world of PC PSU's Seasonic are excellent, certainly one of the top brands.

marktime · 22 Dec 2008 at 18:11

LOL - well that was fairly conclusive!

Furmark stability test lasted 129 seconds! Ran it again, exactly the same - black screen, hard reboot. Within seconds of the card getting to 77c (which is well within specs) it's gone.

Set the clock/mem to standard 575/1000 and it's been running for an hour no problem.

So, get furring folks and post your results!

http://www.ozone3d.net/benchmarks/fur/

TenaciousTomato · 1 Jan 2009 at 17:32

Hey all, I am new to these forums but I've been having the same looping sound, crashing issue with my EVGA GTX 260 Superclocked. So first I got the 'nvlddmkm.sys' BSOD, when my computer would lock up after playing about 15 minutes of any game, and crash.

I read up online about uninstalling the old drivers, cleaning them, then reinstalling the new ones. So I did that, but to no avail. After that I tried underclocking my card to 575mhz core and 1000mhz mem, respectively. This did not do anything to help my situation, besides let me play a game for 10 minutes precrash with worse graphics.

For power, I am currently using a Corsair 650tx, a powerful and high quality PSU, I've got both of my 6-pin connectors connected as well. I am not sure if this is a low measure of power, but my 12V says it's running @ 12.096. According to what I've read my PSU has 52A on one 12V rail. So power shouldn't be an issue, but at this point it may be.

One thing I am really concerned about is the temperature my card runs at when playing any games. It idles at 60*C+ and when playing Crysis it's easily past 110*C after 15 minutes of playing. Should my card really get this hot? I've got the GTX 260 fan @ 100%, and the rest of my case fans at high power to move the heat out. Maybe the high temps are causing my card to crash, or perhaps my card is just faulty. Whatever the case I really wanna fix it!

GSDog · 1 Jan 2009 at 19:38

110C does seem extremely hot. I have yet to see my card exceed 80C when playing Crysis and other games.

GSDog · 1 Jan 2009 at 20:02

marktime said:
LOL - well that was fairly conclusive!

Furmark stability test lasted 129 seconds! Ran it again, exactly the same - black screen, hard reboot. Within seconds of the card getting to 77c (which is well within specs) it's gone.

Set the clock/mem to standard 575/1000 and it's been running for an hour no problem.

So, get furring folks and post your results!

http://www.ozone3d.net/benchmarks/fur/

Did you RMA your card in the end?

I just ran FurMark on mine and within seconds this happened;

I wasn't expecting that because the card has been stable when gaming. When I restarted the PC the screen stayed like that and my motherboard made an error beep. I had to switch the PC off and wait 60 seconds before turning it back on.

What settings in furmark did you use?
I just used my native res (1920x1080) and put MSAA to 16X and ticked the Xtreme burning mode.

EDIT: It just happened again, this time without any MSAA and Xtreme burning mode disabled. Temps reached 82C then about 120secs into the test it the screen went green and when I restarted I get error beeps.

VortexA1 · 2 Jan 2009 at 04:34

GSDog said:
Did you RMA your card in the end?

I just ran FurMark on mine and within seconds this happened;

I wasn't expecting that because the card has been stable when gaming. When I restarted the PC the screen stayed like that and my motherboard made an error beep. I had to switch the PC off and wait 60 seconds before turning it back on.

What settings in furmark did you use?
I just used my native res (1920x1080) and put MSAA to 16X and ticked the Xtreme burning mode.

EDIT: It just happened again, this time without any MSAA and Xtreme burning mode disabled. Temps reached 82C then about 120secs into the test it the screen went green and when I restarted I get error beeps.

Just did some googling .

There could be a driver problem with furmark causing your prob if games are working properly.

marktime · 2 Jan 2009 at 08:51

I'm RMAing my Black Edition and elected for a refund. Will probably get a 55nm 260 or even a 280 if the prices drop a bit as a result of the refresh.

Don't think it's a driver issue as I've tried 3 different sets now with the same result. Fact is if I drop the clocks back to the default 575/1000 Furmark will run for hours. It will start to crash more regularly in proportion to the amount I raise the clocks until it literally crashes every time within minutes of running at 666/1150.

TenaciousTomato 100c (although technically still within spec I think) is extremely hot. Mine got to 80c max after an hour or so of Furmark on the xtreme burning mode. Doubt if it would get to that in a normal game within 15 mins of starting tbh. Does the fan spin up when in games? If so it suggests the hs hasn't been fitted properly.

crunchyfr · 2 Jan 2009 at 09:49

to op you mentioned a grey screen crash and complete lockup.

i have had the grey screen and sound loop crash but it was caused by me copying a large file to my external usb hdd and trying to play a movie from the same drive at the same time

the crash also caused me to lose a couple of files in emule that i was downloading which is quite unusal as i have safe met writing set to always... add to that that im using an ati gpu so its got to be windows thats the issue ? only difference i can figure is that i have recently enabled indexing and readyboost to try out my new patriot usb memory stick

my psu has 4x12v rails also but at 60amps

GSDog · 3 Jan 2009 at 20:59

marktime said:
I'm RMAing my Black Edition and elected for a refund. Will probably get a 55nm 260 or even a 280 if the prices drop a bit as a result of the refresh.

Don't think it's a driver issue as I've tried 3 different sets now with the same result. Fact is if I drop the clocks back to the default 575/1000 Furmark will run for hours. It will start to crash more regularly in proportion to the amount I raise the clocks until it literally crashes every time within minutes of running at 666/1150.

TenaciousTomato 100c (although technically still within spec I think) is extremely hot. Mine got to 80c max after an hour or so of Furmark on the xtreme burning mode. Doubt if it would get to that in a normal game within 15 mins of starting tbh. Does the fan spin up when in games? If so it suggests the hs hasn't been fitted properly.

I just tried 'underclocking' my EVGA GTX260 SSC card to the default clocks of the non-OC version and lo and behold it hasn't crashed once in FurMark...
Clearly my card can't support it's overclock so I will also be RMA'ing mine.

These are the clocks I used, they are correct aren't they?

FirebarUK · 3 Jan 2009 at 21:22

I've been following this thread and can confirm my Black Edition definately exhibits the same problem(s).

Anyone had any luck getting in touch with XFX about this?

marktime · 3 Jan 2009 at 22:54

GSDog said:
These are the clocks I used, they are correct aren't they?

Yep those are as near as damn it to the default 260 clocks. Had mine running for ages @ 650/1150 and thought it was stable but it finally ended up crashing in a game, so I've dropped the core back to 630 and no problems since.

FirebarUK said:
I've been following this thread and can confirm my Black Edition definately exhibits the same problem(s).

Anyone had any luck getting in touch with XFX about this?

When did you buy it, mate? If it's in the last few weeks just contact the retailer and let them sort it out.

FirebarUK · 4 Jan 2009 at 10:03

When did you buy it, mate? If it's in the last few weeks just contact the retailer and let them sort it out.

It was 2 months ago, so no chance now. Going to have to go directly through XFX.

However, just run 12 hours of Furmark @ 666/1400/3300 with the fan at 100% and it's been fine. Might just try reseating the heatsink.

Alfie · 11 Jan 2009 at 15:47

Hi folks.

Anyone had any luck resolving this? I'm having the exact same issue with my GTX 260 black edition. Mainly the grey screen and looping audio forcing a hard reset. I've already RMA'd once and the problem persists. I can re-create it every time at the first checkpoint in crysis warhead. Furmark runs fine for hours though.

Other issues include those bloomin nvdklm.sys errors. I hate those with a passion, it's been a while and we still don't have stable vista drivers. Shame on you Nvidia.

I'm running P45 chipset with an enermax 720w PSU and the system is 100% prime stable and was fine with my previous 8800gtx, trusty old dog that it was.

Let me know how you guys get on resolving this issue.

coupe69 · 11 Jan 2009 at 15:55

marktime said:
At the moment my suspicion is that the card just can't run 100% stable at the default overclock. I just wish I could get conclusive proof enough to get a replacement without a lot of hassle

The easiest thing to do would be download rivatuner, knock the clocks down to stock and do some testing. If it runs ok @ stock then you have your answer. A guy on here the other week had exactly the same issues with a factory overclocked card, I suggested to take the clocks down to a standard GTX280 and lo and behold he had a stable card. A friend also had similar issues with an old MSI 8800 ultra which came with a stupidly high factory overclock, but the only problem was it just wasn't stable ie bluescreens and the old display driver has stopped responding etc etc. He took the memory down a bit and problems gone. He managed to get a refund eventually after a lot of hassle with the retailer which i hasten to add was not ocuk.

marktime · 11 Jan 2009 at 16:01

Well I returned the card and expecting a refund (hope so as I've already picked up a cheap GTX280 last week).

Interesting you say you can get Furmark to run ok - what sort of temps is it getting up to whilst it's running? Mine would crash almost straight away after it reached 77C.

Mine also got one or two nvdklm.sys errors but mostly it was black screen/hard lock ups. Really don't think it's driver related as the new GTX280 runs completely stable at much higher temps.

Good luck with finding a solution.

GSDog · 11 Jan 2009 at 16:39

Alfie said:
Hi folks.

Anyone had any luck resolving this? I'm having the exact same issue with my GTX 260 black edition. Mainly the grey screen and looping audio forcing a hard reset. I've already RMA'd once and the problem persists. I can re-create it every time at the first checkpoint in crysis warhead. Furmark runs fine for hours though.

Other issues include those bloomin nvdklm.sys errors. I hate those with a passion, it's been a while and we still don't have stable vista drivers. Shame on you Nvidia.

I'm running P45 chipset with an enermax 720w PSU and the system is 100% prime stable and was fine with my previous 8800gtx, trusty old dog that it was.

Let me know how you guys get on resolving this issue.

Hi.

You say you have already RMA'd it once - what happened? Did you get a replacement or did the shop say it was not faulty?

All going well my EVGA GTX260 SSC is going back on Tuesday and my main fear is that the shop will say it's not faulty and charge me £30.
My screen goes black or grey in FurMark after just 80 seconds and causes a hard reset so it's pretty clear it is unstable at SSC clocks. At default GTX260 clocks it's OK - runs a bit hot with temps hitting 86C but it doesn't crash at default GTX260 clocks.

Alfie · 11 Jan 2009 at 16:54

GSDog said:
Hi.

You say you have already RMA'd it once - what happened? Did you get a replacement or did the shop say it was not faulty?

All going well my EVGA GTX260 SSC is going back on Tuesday and my main fear is that the shop will say it's not faulty and charge me £30.
My screen goes black or grey in FurMark after just 80 seconds and causes a hard reset so it's pretty clear it is unstable at SSC clocks. At default GTX260 clocks it's OK - runs a bit hot with temps hitting 86C but it doesn't crash at default GTX260 clocks.

It was with ****.com who wont have tested it to be honest. big stores like **** and ****** might say they test em, but they never do, especially if you sound like an advanced user in the RMA note. Like i say, the replacement has the identical fault. When i get back to uni i'll try and underclock and let you know how it goes.

Wish i'd held on a month and got the ultra cheap 280 OCUK had in recently.

Good luck with you RMA and keep in touch.

FirebarUK · 11 Jan 2009 at 19:16

Have RMA'ed mine to ****.com, will see what happens...

GSDog · 12 Jan 2009 at 15:34

FirebarUK said:
Have RMA'ed mine to ****.com, will see what happens...

Me too, it's being collected Tuesday. I'm going to throw in a small note explaining the symptoms and what I have ruled out, etc.
I might ask for a refund and get the standard non-OC EVGA GTX260 55nm and just overclock it myself. I don't think I will buy a pre-overclocked card again to be honest.

marktime · 12 Jan 2009 at 17:51

GSDog said:
Me too, it's being collected Tuesday. I'm going to throw in a small note explaining the symptoms and what I have ruled out, etc.
I might ask for a refund and get the standard non-OC EVGA GTX260 55nm and just overclock it myself. I don't think I will buy a pre-overclocked card again to be honest.

As long as you register the RMA within 28 days of purchase you can request a refund, after that it's replacement only as I understand it.