• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

XFX 280GTX randomnly crashing (Fan revs,black screen, yet no system crash)

JGBR

JGBR

JGBR

Associate
Joined
18 Mar 2007
Posts
1,838
Had this problem with xfx 280gtx for awhile:

Randomly the XFX 280GTX will crash - the GPU fan will go to full power, the screens go dark yet the system still work - it happens at random times

IE: Can do a week without a failure, then may do it on bootup, or maybe after 5 hrs of FSX or 20 minutes of internet surfing. Very random.

(Its not drivers or power supply)
 
Lost proof of purchase

It cant be temp as it crashes randomly after days switched of, or even within 5 secs of bios
 
JGBR said:
Lost proof of purchase

It cant be temp as it crashes randomly after days switched of, or even within 5 secs of bios
Click to expand...

If a warranty return is not an option, then as mentioned above -
  • Remove heatsink and cleanup/replace thermal paste.
  • If that does not sort your problem, possibly look at flashing the software.
You say its not a PSU problem - what have you done to prove this ?
I take it from your earlier post you have obviously reloaded drivers etc ?

A quick search for similar problems seem to point towards a PSU fault (if you do not see an obvious overheat problem when the fans spins up before failure). What is your PSU, have your tried another or your friends for example ?
Also I take it you have the correct PCIE slot used ? Which motherboard are you using ?

http://www.nvnews.net/vbulletin/showthread.php?t=115245&page=2

http://forums.nvidia.com/lofiversion/index.php?t74802.html

Also dependant upon your Motherboard, it could be a simple voltage tweek.

Need more info really if you are asking for advice :confused:
 
Last edited:
My xfx 280 gtx is currently off on rma, i could only use it with the fan set manually to 100%, on auto setting the card would start artifacting at 70-75c, considering the fan on theese cards when set to auto, are only set to rise from 40% speed once temps reach 85c. Still waiting to hear from xfx.

Few screenshots.
Photo-0005a.jpg

Photo-0005.jpg
 
Tried a different PSU and mutliple different ways of powering it. Heat is not the problem.

Its hooked up into a 850W Seasonic PSU and Asus P6TD Deluxe

The crashes are random. With no set pattern that I can detect. Have changed OS, Pulled parts out the machine, changed Memory twice, Change CPU twice and so on.

It's soemthign to do with the card. Can not RMA it
 
JGBR said:
Tried a different PSU and mutliple different ways of powering it. Heat is not the problem.

Its hooked up into a 850W Seasonic PSU and Asus P6TD Deluxe

The crashes are random. With no set pattern that I can detect. Have changed OS, Pulled parts out the machine, changed Memory twice, Change CPU twice and so on.

It's soemthign to do with the card. Can not RMA it
Click to expand...

When you say "Heat is not the problem" do you mean you have monitored the graphics card while the fans begin to speedup and the system goes blank ? Or general heat with the system (obviously not related!).

Fair enough -
Looks like either -
  • Remove heatsink and cleanup/replace thermal paste. (wont fix anything if if NONE of the IC's are overheating)
  • If that does not sort your problem, possibly look at flashing the software. (can be risky - possible bricking of Graphics card - Read up on this fully before completing).
Are the only options ???
 
Yeah kept an eye on the temps

The card sometimes crashes when the machine is just warming up around room temp, overall the machine temp is quite low for a failure. Checked out the card, cooling(paste) seems fine.

So kind of stuck on this one..
 
Reading the forum posts, I have a funny feeling that it might be a rail/voltage problem.

Were not running a modular PSU and having to use some converters to get the 6+8 Pins in.

Think ill go buy a 850W Modular. If not, new graphics card time i guess..unless anyone else has heard of anything to resolve this.
 
Before you get a new psu, is it possible if you could get someone to try the card in their system, may help to rule out the psu. Also was the card a factory oc'ed XT version, seen higher reports of the oc versions failing, mine was a 640mhz core clocked XT variant.
 
Coffington said:
Looks like an issue with graphics card ram and/or GPU soldering. Anyway the problem is in the graphics card.
Click to expand...

If this is the case, a real good look when the HSF is removed may reveal a dry joint. If you are handy with a low wattage solding iron with a very small tip you may be able sort any possible dry joints quite easilly (should not be any need for extra solder, just touch each solder joint in turn with heated soldering iron), if you are not confident with this get someone that is -otherwise you could (again) easilly write the card off totally.

With how intermitent the snag is a dry joint would make sense, can you cause the fault by moving the gfx ports/vibrations etc ?

  • Remove HSF
  • Check for dry joints/signs of damage
  • Solder all suspect joints (if competent)
  • Remove and replace thermal paste
  • replace HSF
  • Test to see if sorted ;)

Last resort - reflash firmware .

Jobs a good un - as a they say ;)
 
Carlos said:
If this is the case, a real good look when the HSF is removed may reveal a dry joint. If you are handy with a low wattage solding iron with a very small tip you may be able sort any possible dry joints quite easilly (should not be any need for extra solder, just touch each solder joint in turn with heated soldering iron), if you are not confident with this get someone that is -otherwise you could (again) easilly write the card off totally.

With how intermitent the snag is a dry joint would make sense, can you cause the fault by moving the gfx ports/vibrations etc ?

  • Remove HSF
  • Check for dry joints/signs of damage
  • Solder all suspect joints (if competent)
  • Remove and replace thermal paste
  • replace HSF
  • Test to see if sorted ;)

Last resort - reflash firmware .

Jobs a good un - as a they say ;)
Click to expand...

That's not realistic approach in home environment.

Most likely some soldering point underneath GPU or RAM chip has failed.

Hot air blower or owen trick might help. Using hot air blower it would take 10-15 quick sweeps or 7-8 minutes for the PCB in 200C owen. This requires technical skills and there's no guarantee the card will work. If the card still has warranty left on it, I'd just RMA it.
 
Last edited:
JGBR said:
Lost proof of purchase

It cant be temp as it crashes randomly after days switched of, or even within 5 secs of bios
Click to expand...

Coffington said:
That's not realistic approach in home environment.

Most likely some soldering point underneath GPU or RAM chip has failed.

Hot air blower or owen trick might help. Using hot air blower it would take 10-15 quick sweeps or 7-8 minutes for the PCB in 200C owen. This requires technical skills and there's no guarantee the card will work. If the card still has warranty left on it, I'd just RMA it.
Click to expand...

OP has already stated no proof of purchase or chance to RMA.
As far as being realistic - IT ALL depends on the capability of the OP as far as what he can and can't solder. Yes the "OWEN" (also know aas Oven :) ) trick has been know to fix a number of PS3's XBOX's Etc and could work. I personally would try and use a decent soldering iron on the contact pads (where IC's are soldered) before baking the PCB in the oven. Obviously (to state the obvious) should you try the "OWEN" method - make sure you remove everything you can from the GFX card (hopefully just leaving the PCB and its components) to reduce the damage this could do (200 Deg C - OOOUUCCHH !)

Heres one of many guides - Obviously use at your own risk !!!!!!!
http://www.overclockers.com/the-oven-trick-repairing-your-broken-video-card-with-an-oven/
 
id check the power supply first, just for elimination, got my wifes 8800gtx doing the same thing under load, new psu with dedicated rails sorted that issue out.
 
You must log in or register to reply here.
Back
Top Bottom