Dying video card?

[mauser]

According to ATI's specs, the 9700 Pro supports up to 1920x1440 at 75Hz refresh and maxes at 2048x1536 at 60Hz in 2D.  It can go up to 2048x1536 in 3D (9700 Pro Features).  My monitor manual said the recommended resolution was 1920x1200 I think (the GDM-FW900 is a 16:10 ratio, 24 in. trinitron tube monitor), and maxes out at 2304x1440.  

*sorry SA2, didn't see your edit until after I was done *

It appears to me that box stock, both of the components should be able to handle the resolutions I was running at (never went above the recmd 1920x1200 and chose only the resolutions offered by the driver).  It would be disappointing to find out that the card would require non-standard voltages or cooling systems to run at the upper ends of it's supported resolutions.   My knowledge in this area is really limited, as I've never dreamed of having such a big screen or running more than 1024x768 for any game.  

mauser

[Siaf__csf]

The card may just have bad ram. Most users never use resolutions that high so who knows, maybe they get slipped through quality control.

Those lines look awfully familiar to me, after pushing the video ram clockspeed through the roof while overclocking.

[eagl]

"I bought a vid card on ebay and it doesn't work right"

You may be hosed, and the seller may have known about the problem beforehand.

I agree that it's most likely a heat problem.  I'd suggest 2 things.  First, remove and replace all the heatsinks on the card.  Use a decent thermal paste.

Second, get one of the vid card utilities like atitool and check the bios version and clock speeds against the card's manufacturer specs.  It may be running a hacked bios, in which case you need to re-flash with an ati approved bios.

Make sure it's not overclocked.

Make sure your case isn't too hot inside.

Try different driver versions.

[AmRaaM]

Load latest driver ... and/or

most people have their comp box set up so that the exhaust fans are restricted by being too close to the wall or dust, check for that first.

check the position of the nearest pci card and what it performs, if its a big heat generator change it to another slot further from the agp,

ck the card fan or sink  if installed for dust buildup dust is a great insulator , works just like cellulose blowable insulation in your attic also disrupts the blades ability to transmit air properly.

Check your fan settings if you are using a controller.

One way to bypass all the individual chk ups is to open the case and run it that way , even maybe place a room fan to cool it.


I've also seen that with atis that have been flashed "up" in performance.

I've also seen that with atis that have bad memory.

Test on another system

if after all above and still fails to display properly, use a water hose to cool it or just contact Ebay PRIOR to 30 DAYS passing from purchase date and get your money back...after 30 days they may not be able to provide all the help they can prior to 30 days.

[mauser]

Unfortunately, it's been more than 30 days since I got the card.  It was from a Power Seller that had over 98% positive and was selling a bunch of these 9700 Pros at the same time.  ATITool and Omega drivers reported the card running at it's design speed 324/310MHz, and the RAM chips are Samsung unlike the slower Infineon chips on my dead 9500 Pro.  

As far as heat is concerned, when I swapped it into my parents machine it had been sitting unused and disconnected for almost a day.  As soon as I installed it and switched it on, I could already see the lines on the screen as soon as Windows was loading.  That's how I figured it must be dying already.  Swapped back the old GeForce2 Pro that I gave to my parents and the machine was fine.  

My case at home is an Antec full tower with two exhaust fans plus the power supply's fan, and one intake fan at the front where the HD's are.  None of the intakes or exhausts (even the side intake which can accomodate another fan) are blocked and the unit is kept under my desk away from the side and rear panels of the desk.  After a few months with this tower, I never bothered to reinstall the temp monitoring software, nor bothered to check the temps in BIOS since they were in the acceptable range now.  

However, my concern right now is to prevent the loss of any more video cards.  From your experiences, should I not even bother running at resolutions higher than 1280x1024 for AH2 and the desktop?  Or would I need to up the AGP (if I go that route) or PCI-X voltages?  Go directly to aftermarket cooling?

mauser

[StarOfAfrica2]

Last thing I'd do is start upping voltages on your card slot.  As long as it was rated to work at those resolutions and refresh rates, it should work.  You may have just been unlucky and had them both go at the same time.  Unlikely yes, but not unheard of.  Do you have a way of checking the voltage at the AGP slot to make sure it is within specs?  I'd also check to make sure your BIOS settings ARE set to proper specs.

[68DevilM]

what should the 9800 pro's temp be around anyways

[mauser]

Originally posted by StarOfAfrica2
Last thing I'd do is start upping voltages on your card slot.  As long as it was rated to work at those resolutions and refresh rates, it should work.  You may have just been unlucky and had them both go at the same time.  Unlikely yes, but not unheard of.  Do you have a way of checking the voltage at the AGP slot to make sure it is within specs?  I'd also check to make sure your BIOS settings ARE set to proper specs.

Hi SA2

I had upped the AGP voltage to 1.6V through the BIOS to no avail.  I could go up to 1.8V I think in 0.1V increments.  Previously, all voltages were set to "AUTO."  I have a multimeter, but I'm not sure how to check voltages in the AGP slot.  I have another day off tommorrow so I may try and stick the 9700 Pro back in and give it another go.  

After doing some more research at Rage3D, I found out I'm not alone:

Attn: Radeon Users Displaying Screen Corruption during POST / Boot / In Windows.

ATI Display and Corruption Problems during POST / BOOT / & Windows Operation

The WinXP startup screen shot from the second link (Figure 5) looks like what I'm getting.  A lot of RMA'ed cards in the first thread, not to mention new power supplys.  The PS supply problems mainly occurred with units that put out less than 350W, and a couple had success after changing to the same PS I have.  I realized after doing additional reading elsewhere that I may not need to change my mobo/cpu (no matter how much I want to, we've spent enough money these past two months on our place).  I probably won't be buying video cards off Ebay anymore either...

mauser

[StarOfAfrica2]

Reading through the posts in your link, what I find funny is that I would have diagnosed 90% of those problems (initially anyway) as monitor problems, not video card related.  After reading through them, the most common thread I can find is power.  Either fluctuating power supplies or AGP slots with improper voltages, or both.  But then there are the ones who had cards that ATI actually found errors on.  As far as I can see THEY did not have bad power supplies or AGP slots.  

I think the one poster probably nailed the cause of it, from whichever source.  Voltage drop not allowing the card to properly initialise.  Whether this came from overheating and damaging part of the card, a bad power supply, faulty AGP slot (which could have been caused by a bad power supply if the power spiked), or just a plain faulty card.  

There's more than one way to catch a cold, but once you have it it's the same as anyone else's cold.  There's more than one way to fry a video card too it seems.  

There are programs you can  use to monitor voltages.  I use Motherboard Monitor

http://www.pcworld.com/downloads/file_description/0,fid,7309,00.asp

It monitors AGP voltage, CPU voltage, temps inside the case and at the CPU, and fan  RPMs.  You can also try SiSandra Lite, I think it shows AGP slot voltages.  Both are free programs.