HD 3850 causing reboots, game crashes, and windows xp pro freezing.

Status
Not open for further replies.

phaero

Distinguished
Jan 10, 2010
3
0
18,510
I just purchased a HD 3850 to go in an old AGP system I had so that my girlfriend's son could play Lord of the Rings Online. Web browsing and gaming are it's only uses. I don't want to purchase him a new system just for the 8-10 hours a week he will be using it. It continues to crash windows xp pro sp3, or crashes game to desktop, or completely freezes up the system to where only a reboot will fix it. I've tried everything I can think of, various bios settings (have the voltage raised to 1.7 right now), using CCC/not using CCC, using Revo Uninstaller Pro to uninstall everything ATI and then a reinstall, about 8 different driver releases (using the hotfix from Dec 09). I just can't figure it out, any suggestions?

p4 3.4ghz 478pin / 4gb ddr / ati radeon hd 3850 agp / abit vt7 pt880 chipset / windows xp pro sp3 / antec ea-650 psu / seagate 500gb sata hdd (x2)in raid 0 / maxtor 250gb boot/os drive
 
First off, give yourself a leval playing field by setting the BIOS back to defaults to eliminate any problems that may be caused by over/under clocking and over/under volting.
Check the temperatures next, use Speedfan or Coretemp for the CPU and GPU-Z or Rivatuner for the card. All of these monitors can be left running to log temperature changes.
Let us know how these first tests pan out.
EDIT:
Also check the motherboard manufacturer site to see if there is an updated BIOS, I have seen issues with VIA chipsets and later AGP ATI cards resolved by a BIOS update.
 

phaero

Distinguished
Jan 10, 2010
3
0
18,510



Yeah I should have noted that before. Have run multiple bios settings from base to upping just AGP voltage, to upping just AGP and DDR voltage, to just CPU, and all combos of the 3 that I can come up with. All leave temps about the same, +/- 5 deg with max load running at 76 at the highest point on all tests. Now the board only gives the ability to raise AGP to either 1.55 (base@63 deg idle), 1.60, or 1.65 (only 76 deg at near full load). Only allows +/-50, base, or +100 for CPU (highest I've seen it get is 1.46 with temp of 43 deg) and runs normally at 1/34v. I know it's not an overheating issue. But I have read that this card needs a lot of voltage to run properly.

Voltage:

CPU VCore 1.34 V
AUX 3.26 V
+3.3V 2.96 V
+5V 5.27 V
+12V 6.26 V

Core CPU Temp
TMPIN0 30deg C
TMPIN1 41deg C

Fans 1654-1687 RPM

GPU Core 62deg C 669MHz(stock)/829MHz(stock)
GPU Core 62deg C

I used CPUID HWMonitor to watch temps and voltage, CPU-Z to verify clocking (CPU and GPU settings), and RivaTuner v2.24 to set/monitor fan speeds and play with OC/UC of GPU. Seems like everything should have been worked out by now but for some reason it just keeps giving me these issues. I keep reading how it could be PSU issue but the PSU I got should be sufficient, unless I should be using 2 PCIe connectors with an adapter to allow them to merge into one out port. Hell idk.

I'm running the newest version of Bios and the 9.12 hotfix from 12/2009 for the vid card.

Could it be an issue with too many HDD on the same rail or something like that with nothing to do wih the graphics? Maybe just the graphics load pushes it over it's limit? I also have Page Filing turned off on main HDD but have it set to 1000MBmin/2000MBmax on a different HDD since everything I have read suggests that I should do that or turn it off completely with over 2GB RAM, I just don't want to get a "not enough virtual memory" issue so I left it on but reduced it to half of recommended size.

I've had the PC running since around 7pm CST Sunday without opening any graphical software (i.e. games, movies, websites with embedded graphics...etc.) without it freezing or crashing. But before that it couldn't go 5 minutes into a game without either freezing, crashing to desktop, or restarting.

Please help, I'm all ot of ideas.

EDIT:

Just rebooted on it's own when I tried to use ConvertXroDVD 4 to burn a DVD.
 
Status
Not open for further replies.