View Full Version : Leadtek 5900Ultra Problem

08-17-03, 10:50 PM
Hi guys
my problem is simple, i cant get anything to run for long periods of time.
When in 3d mode the game ,benchmark or test will close to desktop instantly with no error to show for it, this happens randomly by the way
I havent successfully played anything for more than a half hour.
I was running it in an Asus A7n8x-x board with winxp and 2800xp barton
and i tried it in a gigabyte GA-7VRXP with same cpu and it still happens
my psu is 400w and i dont think that heat is a factor cause straight after it crashes i check the temp of the card and its never above 60c
ive got 2 case fans not inclueding the cpu fan 2 hard drives and one cd drive

any ideas's would be greatly appriciated

08-17-03, 11:24 PM
Is your Leadtek 5900 Ultra come with TwinTurbo II cooler or the reference cooler? I am surprised your card is more warm than mine with reference cooler run at 48C with 430W Antec power supply, my case do not have any case fans. May I suggest you to try disconnect the case fans 1 a time to see if it will stop the games and applications exit to desktop. I think I remember when I read Leadtek manual advised not to place a case fan directly next to 5900 Ultra cos it more likely to disrupted the reference/TwinTurbo II cooler fan(s).

08-18-03, 11:05 AM
It has twin turbo cooling fans the ambient temp is at 21c and the gpu hovers around 28 when idle.
ill try the fan idea though.
i hust thought id also mention that turning off AGP Fast Writes seems to delay the crashes significantly but still not satisfied state.

08-18-03, 11:29 AM
Heres something interesting, i unplugged both fans and every time a couple of seconds into windows it would memory dump on me the entire system would then shut down but when i plugged them back in using all of the power connectors it was ok.
Im starting to think the 400w psu is the culprit.
Any thoughts...

08-18-03, 09:36 PM
It look certainly that your motherboard will need the latest BIOS to fix problems and improve stablity, did you flashed the latest BIOS 1006 from ASUS Update? Also remember to backup your current BIOS before flash the new BIOS.

All Nvidia cards need AGP Fast Write enabled, did you tried set the AGP to either 4x or 8x in the BIOS? If nothing will help, try to load the BIOS to default settings, if it still failed. The last thing you have to try is to formatting your hard drive and reinstall Windows XP to prevent it memory dump every time XP start.

Also can you try to quick scanning the screen every time until you find out what caused the memory dump.

08-19-03, 06:41 AM
I think ive found the solution....
I saw on this site that the fx was tested with apic mode off so i turned off mine and so far nothing has gone wrong although so far ive only tried the dawn demo but i could never even open it before ontop of that i installed win2k.

08-19-03, 10:44 AM
nope that didnt do it
now the only error i get is a blue screen page fault in non paged area memory dump proclaiming that nv4disp.dll is the culprit
any ideas

08-19-03, 12:05 PM
Well you can try use a house fan and cool the entire comp just see if it might be heat :|
Longshot but why not try :)
Try downclocking the card? Maybe it isn't totally stable at that clockrate.
What brand of PSU?
Try different Mainboard BIOSes

You haven't been able to game for more than 30 min.
But are you able to just idle in windows browse write etc etc for longer periods?

08-19-03, 12:53 PM
my psu is generic 400watt, there are two fans in an opened case and living in canberra au the haedware and myself most of the time is damn cold.
ive been doing a lot of research and it might end up being one of two things.

1 the psu is crap and i need a more reliable one
2 my ram is crap and i need a more reliable one

my ram is 512 ddr pc2700 seitek
i know, i never heard of it either and its worked all the way up untill this upgrade.
Majority of the large number of people getting this problem with various other hardware specs have claimed to have fixed it by messing with ram timings or replacing then or changing various voltages or replacing the ram

i dont want to rplace my ram or psu but if a final solution presents itself ill take it

08-19-03, 02:11 PM
from what i've heard, not experienced, a bad PSU results in random reboots and not the things you've described.
I forgot though what bad RAM brought with itself

08-21-03, 01:22 AM
Originally posted by Valias
nope that didnt do it
now the only error i get is a blue screen page fault in non paged area memory dump proclaiming that nv4disp.dll is the culprit
any ideas

It the driver, what version of the driver you run now caused the memory dump? I think you probably used it from the supplied CD or from Leadtek site. The latest driver is Detonator 45.23 WHQL here (http://www.nvidia.com/object/winxp-2k_45.23).

08-21-03, 06:08 AM
i just solved the random reboot problem on my new system. it turned out to be the psu getting overheated, because i placed the system too closed to the wall. it just did sudden reboots, no BSoD, no error messages, and autoreboot option was disabled. once i avoid the psu getting overheated i can run ut2k3 for more than 10 hours straight without any problem. hope this helps.

08-21-03, 09:40 AM
nv4disp.dll errors are usually indicators that your RAM is having problems. This problem usually manifests itself through driver failures, because you are getting page errors internally in the RAM. I had the same problem with my old dual MP system, blue screens constantly and memory dumps with driver failures, until I tracked it down to a bad stick of memory. Luckily, it was crucial memory :). Try replacing your memory and see it that fixes it.

Hope I was some help.

08-22-03, 01:55 AM
thanx people im pretty sure that its the RAM to.
i did a load of research in page errors and blue screens in general and found out that there is actually very few causes for blue screens that not many people know and these few causes manifest themselfs in as you said driver failures and a load of other sources.
Though just going over forums here and there ive notice people evrywhere getting page in non paged area faults and this particular error is in the ram
im going to be getting a stick of 512 xms corsair ram when i can next afford it.

08-24-03, 02:24 AM
Nah must be something else cos my Leadtek ultra FX 5900 works fine on my machine... but I am running XP...

08-30-03, 12:36 PM
[the problem is fixed
i was stuffing up at the high default cas settings so i lowered them to run faster and it works now but to do this i had to turn thr ram frequency down to 133 (266mhz)as apose to 166 (333mhz)

i borrowed a stick of corsair ram from a house mate of same specs and it worked at full speeds no tweaking needed
so for future reference it is gennerally the ram cas speed settings that causes blue screens with new systems

08-31-03, 09:25 PM
Originally posted by Valias
It has twin turbo cooling fans the ambient temp is at 21c and the gpu hovers around 28 when idle.

Are you sure those numbers are correct ? are we talking about your video card GPU temperature here ? if that's the case please teach me how to get these temperature that is awfully LOW :)
and how does it go up to 60C from those numbers :rolleyes:

I idle at 35C GPU
and never goes up more then 50C and I game a Lot. :confused: :confused:

09-01-03, 02:53 AM
during the day with the side panel on it goes around 38 but at night it drops to around 28 keeping in mind though i live in a place with a cold climate

Canberra Australia