• Announcements

    • HEAVY265

      New job posting.   02/23/2019

      The Community Management team is looking for war correspondents to provide news stories or after action reports to be published on our website and our Steam forums.  Player correspondents imbedded with a squad is fine along with reporters overall who might want to produce stories about the campaign, a piece of equipment, a battle or a skirmish - all stories are encouraged and welcome.  If intertested in a volunteer war correspondent position apply at Tman@corneredrats.com
Sign in to follow this  
Followers 0
Randazzo

On the verge of 8800GTS card failure?

19 posts in this topic

Starting tonight, about 2 hours ago, my graphics started acting up in game. I had played for about 30-40 minutes as infantry with no trouble, then spawned a plane and the game froze and the screen went into a rainbow of colored pixels.

Tried a restart and got the same thing as soon as the game loaded, on the play axis or allied screen. Restarted again, had colored pixels on the post and bios, and blue screened (can't remember the last time this happened, never on this system) trying to load into windows.

At this point I pulled the card out and checked all the connections. Everything looked kosher but I took off the coolor and removed the unexpected dust bunny, cleaned off and reapplied thermal paste, put it back together and plugged it in. Fired up windows, same colored pixels everywhere. Shut it down again.

Came back into safe mode, everything loaded fine so I thought it might be a bad driver (although I have the latest one, do the updates automatically). Downloaded the driver install from NVidia, removed the old drivers and tried to install from the new file. It told me it couldn't detect any compatible hardware. Reboot the system, windows detects the card, installs it's own driver. I loaded up the game, no crash. I got into a plane (2 hours later) and flew for about 5 minutes, got bounced and with 1 second left on the despawn timer the game froze. As it was in the process of crashing out, I saw a big yellow caution indicator on my taskbar and moused over it, getting

"display driver nvlddmkm stopped responding and has successfully recovered"

This has never happened before and considering it just started in the middle of the game I don't really see how it can be a driver issue.

Pny 8800GTS 512MB

Core 2 Quad Q8200 2.33

Vista Home 64-bit

Share this post


Link to post
Share on other sites
How's the heat in the case?

Not sure about the ambient temp in the case but the GPU runs at about 76-80 degrees celsius under load, mid 60's at idle. This could improve if I downloaded a fan control program but I haven't had the need. Supposedly they can withstand sustained temps of 90ish.

CPU runs no hotter than 34 celsius, ever. I did have an issue with it a couple weeks ago where I had managed to break one of the cheap plastic pins holding the stock fan to the chip and didn't realize it until my system started setting off red alerts and alarms and buzzers (only thing good about the POS MSI board) warning me of overheats. I have since installed a heatsink that looks like a radiator and has an 80mm fan on it. Case has three fans, one on each side (80mm) and one on top (120mm), PSU obviously has it's fan too.

I've been trying to find stuff on the internet to explain it and I've pretty much come to the conclusion the card is toast, that seems to be the consensus when it's putting up lines and artifacts at the post and bios screens. It's apparently common with the 8800, so I got a good run out of it.

Share this post


Link to post
Share on other sites

I had a 6800 and a 9800GT card both go kaploo to heat issues in the last couple of years, only reason why I asked... Kinda figured as you'd even gone so far as replaced the thermal paste.

Perhaps a 450 or something is right around your corner! ;)

Share this post


Link to post
Share on other sites

I used to get that error all the time when I had Vista. I did some research into it and couldn't ever really fix it so I just upgraded to Win7 and never had another problem.

Share this post


Link to post
Share on other sites

It was definitely the card. I bought a 9800GTX to replace it and everything is fine.

If anyone else gets this with the 8800, artifacts on post and bios screens, the card is toasted.

Share this post


Link to post
Share on other sites

Read this... it does work! That is, IF you have no other options to purchase another GPU... this is your last resort. Follow the guidelines exactly as suggested!

http://hardforum.com/showthread.php?t=1421792

I would still purchase a new GPU, and keep the faulty one as a standalone physx card.

Edited by valis

Share this post


Link to post
Share on other sites
Read this... it does work! That is, IF you have no other options to purchase another GPU... this is your last resort. Follow the guidelines exactly as suggested!

http://hardforum.com/showthread.php?t=1421792

I would still purchase a new GPU, and keep the faulty one as a standalone physx card.

Aye, I read that.

(But I have no excuse to buy a better card if I did it! :D)

Share this post


Link to post
Share on other sites
Read this... it does work! That is, IF you have no other options to purchase another GPU... this is your last resort. Follow the guidelines exactly as suggested!

http://hardforum.com/showthread.php?t=1421792

I would still purchase a new GPU, and keep the faulty one as a standalone physx card.

Which might work if the problem is bad soldering. If it's not then this won't make any difference at all.

Share this post


Link to post
Share on other sites
Which might work if the problem is bad soldering. If it's not then this won't make any difference at all.

not so much bad soldering (unless its something internal as a transistor, as you note)... the main reason GPU's fail is because of heavy use and heat, which cause the solder to deteriorate and cause bad / loose connections. the reason this works, as MOST manufacturers do with refurbs or QA (I personally used to work at IBM testing Tiger PCB/GPU boards, along side engineers - this is a common fix - but we had more sophisticated ovens), is by baking the card in a controlled environment which solves most of these common failures in GPU's... though, if you did have a bad transistor and blew during the oven process it would break/blow/ooze then it's obvious it was a hardware issue... then again outside a warranty you really have nothing to complain about.

If you had a bad solder from the start, you would have noticed issues earlier on... artifacting... overheating... etc...

otherwise, just follow the posters suggestion on [H] , and look at the photos of how he done it... if you follow it you won't do anything wrong, unless something is faulty with the hardware.

edit: just wanted to add that I had an IBM T30 with the known memory module problem. IBM put out a notice to have the mb replaced for those iwthin their warranty free of charge, unfortunately i bought the mb on ebay and hadn't known this and when I called IBM they wanted $600 to repair it, which of course was double the price I paid for the laptop. so after a year that dimm slot went bad because of how the laptop door was designed which caused excessive heat to the dimm slots and causing the soldering to break apart and a bad connection causing the module in that dimm slot not be recognized... I baked it and now it works fine. Today, I don't use the cover over the memory so it has free flowing air. :)

Edited by valis

Share this post


Link to post
Share on other sites
not so much bad soldering (unless its something internal as a transistor, as you note)... the main reason GPU's fail is because of heavy use and heat, which cause the solder to deteriorate and cause bad / loose connections. the reason this works, as MOST manufacturers do with refurbs or QA (I personally used to work at IBM testing Tiger PCB/GPU boards, along side engineers - this is a common fix - but we had more sophisticated ovens), is by baking the card in a controlled environment which solves most of these common failures in GPU's... though, if you did have a bad transistor and blew during the oven process it would break/blow/ooze then it's obvious it was a hardware issue... then again outside a warranty you really have nothing to complain about.

If you had a bad solder from the start, you would have noticed issues earlier on... artifacting... overheating... etc...

And this is about what I said, just with fewer words! ;)

Share this post


Link to post
Share on other sites

well every piece of PCB that gets hot suffers from "chip creep"

Chip creep is when a component on the PCB itself gets hot and expands which pushes the soft(er) solder away, and when you turn your computer off (often right after playing a game, leaving the system hot inside) the chip will contract back, leaving the solder just a weeeeeeeeeee' different then it was before, and over repeated cycles you can partially lose contact, or fully pull the chip from the PCB, sometimes even cross pins over.

There is also another thing that happens called "Tin whiskers" thats also caused from heat. i guess from what i understand of it, all Tin/Lead solder used today when exposed to a EMF (Electro-magnetic-filed) and heated will physically rearrange its self bit by bit in line with the EMF over time and eventually cross S@#t over and short-circuit things.

http://en.wikipedia.org/wiki/Chip_creep

http://en.wikipedia.org/wiki/Whisker_(metallurgy)

Both of those can be prevented by keeping your components as cool as possible all of the time, and 90C (194F) is insanely hot for any graphics card, sure its within its operating threshold, but thats the tip top of the threshold and while it'll operate, its certainly not good for it at all. I'd consider leaving the side panel of your computer off all the time, or upgrading to some very high CFm (cubic feet of air per minute) fans. stock 80's usually push 30cfm, while 120mm usually push 50cfm, i'd look at some 80's that push 50~60cfm and some 120's capable of over 100cfm.

http://www.frozencpu.com/products/8125/fan-499/Delta_120mm_x_38mm_Extreme_Speed_Fan_-_Bare_Wire_-_120_CFM_AFB1212HHE-F00.html?tl=g36c15s562

(I've got three of those suckers in my system, my GPU temp peaks at 134F, and idles at 89F)

http://www.frozencpu.com/products/6453/fan-397/Cooljag_Everflow_80mm_x_38mm_Fan_-_5666_CFM_F128038BU.html?tl=g36c15s699

never used these, i've got Thermaltake Volcano12, but i've heard good things about those particular ones.

Remember you'll want to regulate those fans, so a controller is always a good idea to keep the noise down when you don't need the extra power. they make rheobus style (knobs) or Digital ones, they've also got really sweet LCD thermister/tachometer controllers that control the fan RPm's by the temperature they read.

http://www.frozencpu.com/cat/l2/g34/c17/list/p1/Bay_Devices-Fan_Controllers.html

dscf5328.jpg

You can see my LCD tachometer up front in a CD-rom bay :)

Share this post


Link to post
Share on other sites

Nice rig there!

The 8800 was apparently an especially heat-heavy card.

The 9800GTX I replaced it with is only running at 51c under load, and the cooler installed on it is much smaller.

Edited by Randazzo

Share this post


Link to post
Share on other sites
Nice rig there!

The 8800 was apparently an especially heat-heavy card.

The 9800GTX I replaced it with is only running at 51c under load.

lol thanks, won rig of the month on thebestcasescenario a few months back.

yea i had a eVGA 8800GTS Super Super Clock edition....holy crap that card would overheat if it even looked at left4dead.

I had a adventure trying to keep that card cool.

x4q4n6.jpg

8800's under and behind all them fans.

One thing that REALLY made a difference is adding a fan directly on top of where the GPU itself sits, you can see i've got a 40mm on top of my ATi card, and i had two 72mm fans on my 8800GTS~SSC, 72mm fans fit right on top of Nvidia cards right on those four screws (look on top of your 9800GTX) it makes a good 6 to 8 degree difference and probably protects against creep and damage a lot as well.

Just use a couple pieces of double sided sticky tape, or a product called VHB made by 3M (strongest adhesive tape in existance) to hold the fan in place.

Edited by indo420

Share this post


Link to post
Share on other sites
The 8800 was apparently an especially heat-heavy card.

Indeed it was. My 8800 GTS failed a couple of days before yours, with the same symptoms. I replaced it with a 5850 with Vapor-X and the PC now runs so cool I have removed two of the four case fans, and it's still cool under full load.

Share this post


Link to post
Share on other sites

I have just had the exact same issue described....could all of our 8800 vid cards be giving out at once? Weird.

Guess I have to buy a new vid card....any recommendations of vid cards specifically for WWII online?

Share this post


Link to post
Share on other sites

Uhm, i have 8800 GTS 320MB (@ 620/1450/1900) and i run the game @ 1680x1050 @ max settings + 16x AF and 2x AA without any problem of temperature.....O_o

Try download MSI Afterburner and set a different fan speed. Download: http://event.msi.com/vga/afterburner/

I set my card in this way:

0 to 50 degrees ---> fan @ 50%

from 50 degrees, every time that the temperature step up of 1 degree the fan increase his speed of 2%.

In this way the max temperature reached by the 8800GTS is 66-67 degrees.

I hope this can help you.

Share this post


Link to post
Share on other sites
Uhm, i have 8800 GTS 320MB (@ 620/1450/1900) and i run the game @ 1680x1050 @ max settings + 16x AF and 2x AA without any problem of temperature.....O_o

Try download MSI Afterburner and set a different fan speed. Download: http://event.msi.com/vga/afterburner/

I set my card in this way:

0 to 50 degrees ---> fan @ 50%

from 50 degrees, every time that the temperature step up of 1 degree the fan increase his speed of 2%.

In this way the max temperature reached by the 8800GTS is 66-67 degrees.

I hope this can help you.

Out of the box, mine worked perfectly for three years and three months before failing suddenly and completely, which is an acceptable lifespan for a device having a hard life, in my opinion. It maxed out at 70C under full load (at 2560x1600), which is within limits. I'm not convinced that manipulating the fan speed would have made that much difference.

The temperature inside the case was far higher with the 8800 GTS than with the overclocked 5850 I replaced it with - so the new card should help the lifespan of all the other components. CPU, disk and motherboard temps are all down significantly since I replaced the 8800, so much so that I was able to remove two of the four 120mm case fans and it's still cool. No doubt I will have to put them back when I refresh the motherboard/CPU/RAM in the near future, but for now, it's made a big difference in noise level.

Share this post


Link to post
Share on other sites

I bought eVGA GF8800 GTS 512MB, but I hated the stock heatsink and the fan sucks

I mounted Arctic Cooling GPU heatsink plus 2 x 92mm fan with max RPM all the time.

It is about 45c idle and 55c MAX in summer day with all graphic setting to lowest in this game

Share this post


Link to post
Share on other sites
Sign in to follow this  
Followers 0

  • Recently Browsing   0 members

    No registered users viewing this page.