1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Did I fry my video card?

Discussion in 'AMD / ATI' started by EagleEye, Jan 24, 2009.

  1. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Okay, here's the whole story.

    About 2 years ago, I bought an ATI Radeon x1900 GT (v2) 256MB card from Newegg as part of an overall system upgrade that consisted of a mobo/cpu/video card. The whole upgrade centered around the video card, and it was my first PCIe, which is why I had to get a new mobo (I had AGP in the past).

    Here are links to all of the items I bought (from Newegg).
    Video card: http://www.newegg.com/Product/Product.aspx?Item=N82E16814102022
    CPU: http://www.newegg.com/Product/Product.aspx?Item=N82E16819103023
    Mobo: http://www.newegg.com/Product/Product.aspx?Item=N82E16813131035

    Problem 1: My PSU was a 350watt 2x12v rail, with 14amps on one rail and 15 on the other.

    I never really saw a problem with the card and it worked fine for that whole time. The CPU was an AMD 3400+ single core.

    Early this past December, I upgraded to a new AMD Phenom 4x2.3ghz, with a gigabyte mobo. Newegg links here:

    Mobo: http://www.newegg.com/Product/Product.aspx?Item=N82E16813128081
    CPU: http://www.newegg.com/Product/Product.aspx?Item=N82E16819103244
    RAM: http://www.newegg.com/Product/Product.aspx?Item=N82E16820231122

    Just prior to that, I got a 1TB drive for my server, so I moved the 120gig SATA drive from the server to my main gaming machine to replace my old 80gig IDE. This whole process of upgrading and moving drives required that I reinstall my OS (Which I suppose I should mention is Vista 32bit home basic).

    Anyway, prior to the upgrade, I had been getting artifacts, flickering, and odd stretched polys in WoW. I also played CS:S for a long time, but never had a problem there. The problems in WoW would go away if I alt-tabbed and went back in, which I guess forced a reset of the 3D view or something.

    So I do the upgrade and I start getting these problems a lot more... I do a little research and find out that this entire time I've been underpowering my video card. It requires 25 amps on the 12v rail, and recommends at least a 400watt PSU. It's not that simple, of course, because a 400watt PSU doesn't necessarily have all the amps you need.

    SO... I decide to get a new PSU. Now, being stupid, I didn't even check the rails on my current one until I got the new one and had the same problems.

    I got a Thermaltake 500watt with 2x12v rails, 14a and 15a respectively... then I look at the info on the 350watt that I just replaced and see it has EXACTLY THE SAME AMPERAGE on the 2 12v rails. *SIGH* and *DOH!*

    Still, I figure Thermaltake is a better product and 500 total watts is probably better anyway.

    Link to PSU: http://www.newegg.com/Product/Product.aspx?Item=N82E16817153052

    SOOOOO... I power it all up and it works fine for a bit... then BAM, same problem and worse than before. More flickering, and a new symptom: Complete video freeze! My other software is still running, I can talk over ventrilo and people can hear me, and I can hear them, etc... but my video just freezes.

    I have a dual monitor set up, and the game will freeze on the main one, and on my other one it just goes blank, like the video card stops feeding it any information (I get a "no signal" box on it).

    I have to manually reset the system to get anything working again.

    EVERYTHING works in windows just fine, probably because it's not under any sort of extreme load.

    I also play TF2 and since the new mobo/cpu/RAM, I have had worse artifacts and screen tearing than ever, and now it's starting to cause lockups as well, which never happened with the 350watt and single core CPU.

    This is where ATI Tool comes in.

    I wanted to see about underclocking the video card while waiting for the new PSU to arrive, hopefully to stave off the problem for a while so I could at least log in to WoW and do a few things. I also wanted to use it to test to see if it really was the video card being under powered, so I tried finding max core speed and it was insane... it shot up quickly, then get this... artifacts found over and over and it stepped back down to UNDER the default core speed (513mhz), all the way down to 400mhz before I just aborted the test.

    Then I did something stupid... I accidentally hit the max mem thing, and before I could cancel it, ATI Tool shot up the memory speed WAY high and my system just crashed.

    So I reboot, go back in to ATI Tool to make sure the settings weren't saved that way (fearing another crash), and I find it saying "Default (0.00 / 0.00)" in the selection box.

    I go in to settings and attempt to select the video card, and it says I have to restart ATI Tool to use that card. Fine, I do. Go back in to settings, and WTF? It's not selected... select the card again, and it says I have to restart ATI Tool, etc... rinse repeat over and over.

    This is all with ATI Tool 0.24.

    So... I see there's a new .26 out... Uninstall .24, install .26. Shoot past the stupid vista warning... and same thing.

    So I just avoid ATI Tool and try with RivaTuner... that shows things are right, but only allows me to go UP on my core speed and memory speed.

    Meanwhile I have a 650watt 1x12v52a rail PSU on order...

    This one: http://www.newegg.com/Product/Product.aspx?Item=N82E16817139005

    That PSU came yesterday.

    I install it, and again things work okay for a while (making me think there's an overheating problem)... And BAM, it all starts happening AGAIN!!!

    So... I've got ample amperage (if I may abuse an alliteration)... so that can no longer be the problem.

    The thing is... could it be that the ATI Tool crash when it did the max memory testing caused physical damage to my video card?

    Could it be that running the card under-powered for 2 years could have done this? The problem DID start way before ATI Tool came in to the picture... but it's gotten a LOT worse since my upgrade to quad core.

    A few other tidbits...

    I have taken the cooling mechanisms off my north bridge AND GPU, and replaced the stock "grease pads" (if you can even call it thermal grease at all...) with Arctic Silver. I don't know if it's an overheating problem, but I did it just to be sure.

    Also, I have rigged a regular 80mm case fan inside my case to blow on my north bridge.

    Speedfan reports my "temp3" as being between 78 and 80deg celcius at all times, which is waaaay too damn hot. That's 176 degrees for christ sake! BUT, Speedfan is notoriously inaccurate in my experience (it once reported my PSU was providing only 9 volts on my 12 volt rail, and an actual voltometer showed the rail was just fine). No other diagnostic tool (including the Gigabyte one) shows any temperature problem... but the north bridge IS fairly hot to the touch (I can put my finger on it for about 10 seconds before the temp becomes too hot for me to stand anymore).

    So... is my video card simply failing after 2 years of trusty service? Did I kill it with my stupidity or shenanigans? Is my quad core simply taking too much power from it? (It's a 95watt quad core, which i understand is fairly low power consumption). Could the motherboard (specifically that north bridge) be faulty and require an RMA? Hell, is it my system RAM?

    Here are some other raw stats:

    Gigabyte Easytune 6 reports:
    CPU Fan: 2402 RPM
    System Temp: 34deg c
    CPU Temp: 42deg c
    CPU2 temp: 0? (no sensor I presume? It's not a dual socket mobo).
    Voltages: VCoreA 1.050V on -12V rail
    +3.3V = 3.340V
    +12V = 11.920V
    CPU Vid: 1.250V

    RivaTuner reports:
    Radeon X1900GT -
    Core Clock 513
    Memory clock 657
    Core temp 48.5deg c
    Ambient: 42
    Fan duty cycle: 40.91%
    Fan Speed: 0RPM (It *IS* spinning though, so no idea why it's not reporting)
    Voltage Regulator Output: 1.16V
    Votage Regulator Current: 7.26A
    VRM phase 1 temp: 46deg c
    VRM phase 2 temp: 52deg c (132deg F? Wow...)

    AND FINALLY... after doing an uninstall of ATI Tool .26, and running CCleaner, then reinstalling .26... ATI Tool is recognizing my card again and showing the proper core and memory speeds... But I run an artifact scan at default speeds and after about 20 seconds I get 2 (and always in the same spot), and shortly after that, more and more and more... until the system crashes.

    Crazy, eh?

    So please, tell me what the problem is... it's driving me batty! :banghead:
     
  2. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    I THINK I FIGURED IT OUT!

    ATI's CCC has a problem with dual monitors and AERO.

    Disabled AERO, and uninstalled all ATI stuff, then reinstalled Driver-only... and BAM, no more problem!
     
  3. tigger

    tigger I'm the only one

    Joined:
    Mar 20, 2006
    Messages:
    10,187 (2.97/day)
    Thanks Received:
    1,402
    glad you sorted it,but how many words a minute do you type :p
     
  4. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Hahaha, about 100wpm. I'm a programmer and a long time computer user and internet forum/chat type person. :)

    Oh, also another thing I noticed... my CPU has gone down from about 42-44 degrees, to 36-37... just by getting rid of the CCC thing? Maybe there was something going on in that app that was spiking my CPU? *shrug*.
     
  5. tigger

    tigger I'm the only one

    Joined:
    Mar 20, 2006
    Messages:
    10,187 (2.97/day)
    Thanks Received:
    1,402
    Aah well,its sorted and your cpu is running cooler,bonus.

    100wpm :O,i think i'm about 20 lol
     
    EagleEye says thanks.
  6. r9

    r9

    Joined:
    Jul 28, 2008
    Messages:
    2,144 (0.84/day)
    Thanks Received:
    284
    Just when I was going to say that it is not PSU problem and could be also artifacting if you have RAM memory problem and if it is not RAM and underclocking don`t resolve problem that you have faulty mem chip on the GPU so the GPU runes fine until that chip is loaded with data but I was wrong :D
     
    EagleEye says thanks.
  7. FilipM

    FilipM New Member

    Joined:
    Dec 30, 2007
    Messages:
    802 (0.29/day)
    Thanks Received:
    94
    Location:
    Bitola, Macedonia
    Now if everyone made such good full reports when they have a problem...glad you sorted it out!
     
    EagleEye says thanks.
  8. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Thanks for the compliments... it must help that I am a long time computer tech. For one thing, I know how to think my way through a problem... for another thing, I have a lot of experience with all sorts of things.

    The thing that really throws a wrench in to all of this is that there were SO many changes in such a short amount of time.

    1) New CPU
    2) New RAM
    3) New Mobo
    4) New HD (well, new to the system)
    5) New install of Vista
    6) Ultimately a new PSU

    All of these things pretty much prevented me from narrowing down the problem... and as it turned out, it was the new install of Vista that really threw me. A lot of people don't think about it this way, but... reinstalling your OS could mean installing something with default settings that you had turned off previously... or installing something with a newer version that is buggy that was stable in your previous install of it... or installing something in a different way than you did the previous time, which causes a problem...

    The problem with the whole situation was, I was focusing on the obvious (at least at first), and thinking it MUST be a problem with the hardware I got. (And I can't help but ask how many RMAs and reports of bad hardware are due to something like this, where the board is 100% fine, but the user doesn't realize what the real problem is...)

    Anyway... I'm excited about finding a solution. :)
     
  9. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Update: just played 3 hours of TF2 without a single bad poly, no flickering, no crashes.
    Also played about an hour of WOW without any problems.
     
  10. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Another update...

    3:15am after playing WoW for 3 or 4 hours, I started getting poly stretching, and ultimately a hard crash requiring a reset... :(
     
  11. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Okay, another update...

    I'm STILL crashing every so often... still often enough to be annoying.

    Now the crash has changed so it's not generally a hard lock up (though that still happens)... WoW crashes to the desktop, and HL2 (while playing TF2) crashes so I have to end task on it.

    At this point I'm thinking it may not be the video card at all, but my Mobo or RAM.

    I'm about to take out one of my 2 sticks of RAM to test... if it crashes, I'll swap them and test again.
     
  12. DRDNA

    DRDNA

    Joined:
    Feb 19, 2006
    Messages:
    4,805 (1.39/day)
    Thanks Received:
    573
    Location:
    New York
    make sure your video cards air cooler has not built up dust blockage ...hate to go to the obvious but well you know..
     
    EagleEye says thanks.
  13. Cold Storm

    Cold Storm Battosai

    Joined:
    Oct 7, 2007
    Messages:
    15,014 (5.25/day)
    Thanks Received:
    2,999
    Location:
    In a library somewhere on this earth
    Well, here is something that I came to hate.. Vista on that system is a wrong idea... i say it because I owned 1950, 3400+ and tried Vista... If you look in the Vista trouble thread that Polaris started, I have posts when I first started.. Well RM made them...

    You can't get the system stable enough for Vista to run. It's not really going to happen.. I tried my hardest.. Couldn't, and that's with my system oc'ed to 3.4 and oc on everything... Vista didn't like Aero at all with ATI on it with that card..... It was to the point I had to reinstall XP till I gave it to my sister..
     
    EagleEye says thanks.
  14. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Unfortunately, I don't have an XP license to fall back on.

    I have Aero disabled.
    I don't have the CCC app installed with my ATI drivers.
    Unfortunately, the Omega drivers aren't available for my x1900.
    I've stopped using my dual monitor setup...

    I'm down to thinking that maybe the mobo or RAM is bad.
     
  15. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    I've cleaned it completely. :)
     
  16. Cold Storm

    Cold Storm Battosai

    Joined:
    Oct 7, 2007
    Messages:
    15,014 (5.25/day)
    Thanks Received:
    2,999
    Location:
    In a library somewhere on this earth
    Do you have Solaris's boot dvd? Or Ultimate Boot CD? If not, I suggest grabbing it. restart the computer with one of them, and run memtest86+. Do one full pass. Should take a half an hour or so... Then you'll know if it's ram..
     
  17. Wile E

    Wile E Power User

    Joined:
    Oct 1, 2006
    Messages:
    24,324 (7.53/day)
    Thanks Received:
    3,778
    Did you happen to update to SP1 yet?
     
  18. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Yes to SP1. No to the memtest yet... I used to fix computers for a living... I'm familiar with Ultimate Boot CD... very nice, but I don't have it yet.

    I'm TRYING to boot to my USB thumb drive, but it's not working.
     
  19. Thrackan

    Thrackan

    Joined:
    Oct 10, 2008
    Messages:
    3,482 (1.40/day)
    Thanks Received:
    656
    So what's the temps you're hitting in say FurMark?
    You should be able to check the temps in the CCC (Overdrive section), so you don't need to guess if Speedfan's tem3 is the correct one :p
     
  20. EagleEye New Member

    Joined:
    Jan 24, 2009
    Messages:
    11 (0.00/day)
    Thanks Received:
    0
    Riva shows my VRM Phase 2 at 50deg C when in windows. When playing 3D games it shoots up to 77 or 80 deg celcius.

    That seems VERY high to me... That's 176 degrees F!!!
     
  21. Thrackan

    Thrackan

    Joined:
    Oct 10, 2008
    Messages:
    3,482 (1.40/day)
    Thanks Received:
    656
    It's high, but not totally unacceptable for the card itself. The solution still eludes me :(
     
  22. Cold Storm

    Cold Storm Battosai

    Joined:
    Oct 7, 2007
    Messages:
    15,014 (5.25/day)
    Thanks Received:
    2,999
    Location:
    In a library somewhere on this earth
    The reason that my set-up couldn't work was due to the fans on the card.. the card was getting to hot for Vista. The reason I say that is, Vista has a fell safe on their os, It tries to catch BSOD before it even happens, and fixes itself. The heat of the card isn't that bad. You'll see the newer cards that have same temps.. But, I believe Vista just knows with drivers that the cards are suppose to be that hot. As for the old ones... I don't think so..
    It can be heat that is caused, but not because it's to hot, just the aging of the card..
     

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)

Share This Page