Connect3D X1950pro 512 headacke!!

[Reeverdark]

Mornin all.

and it aint a fine one ether!

i bourt a 512mb DDR3 Connect3D PCI-E X1950pro at around £126 off the net 5 mounth ago and up till a mounth after perchise all was going well (i wasen't running anything that needed alot of power at the time) however apon running a gold addition of Spell fource: order of dawn, which work fine on my old pc, it started to freez in game go blank then the recovery centure poping up a error messege say the card or driver had stop responding and needed refreshing. I though at first it my be an incompatability with the game, however when
C&C3:TW came out i bourt a copy thinking i would run considdering i had only built my new pc this year. ran the game... played it for 15 min then the same thing happend. i left it in the game menu for a bit to get lunch and apon return, yet again the same thing. so i started a long investigation of my system to find the problem. i tested all my other games and found that it dosent matter how new it is... i tested Ground control 2 and it turned out to be the worst affected to the point of it being unplayable as it froz once every 10 to 20 seconds. even world of warcraft frezzes after 30min of playing. i tryed to reinstall windows on my pc and that made no difference (current operator is win XP home sp2). here is a list of my pc specs:

AMD Athlon x2 2.0Ghz
MSI K9A platinum
2GB DDR2 800
WD raptor 150GB

As far as cooling is conserned my pc has got within an excess of over 11 fans (including CPU, GPU, PSU. ect)and a WD tep controler.
PSU is a modular and has and output of 530w and as far as i can tell its running fine.

I need help asap people so please i beg you all, and if iv left anything out please ask.

[mitsirfishi]

that will be vpu recoverer kicking in which means the core is excessively overclocked and sets back to default and or your core is overheating try taking the heatsink off cleaning the gpu and base up and sticking some thermal paste on there and give it ago again

[Reeverdark]

but its on its defult settings i havent overclocked it at all. if it just that the heatsink is out of tuck with the gpu coz the dam manufacturer dident do his job then ok but as far as what iv done there is nothing to sujest that it should happen in the first place.
thanks and ill give it a try...
would you think that it might be the design of the heatsink? after all it was the flates one i could find as i tryed my best to alow airflow in my case and as its the only gcard in there it should have all the space it needs.
are there better alternatives Heatsinks out there with space in mind? i can only afford the standard fan type and not water of liquid metel cooler units.

[SpookyWillow]

what are your psu rails at when your running 3d apps? check the 12v - 5v and 3.3v.

how hot is your card getting when running 3d?

lastly, try underclocking your core by about 20mhx and see how it plays games.

post your results back here so we know a bit more ;)

[Reeverdark]

Right! heres the low down..
i was unable to get all the data required but...
i tryed Ground control 2 again to see.
my GPU standing temp (in windows decktop) is at about 45-48C. (this is the standerd and it will not go any lower).
apon running the game.. i was in the menu fidling with the grph settings then it frozz on me and i dident even play the dam thing!!!!
the standing temp at that point was at the most 55-57C.
i cant tell you about the moyority of my PSU as im a newb at that however i can tell you that it is a Hiper Type-R Series 530w modular supply and at time of purchice it set me back about £76.

[mitsirfishi]

psu is fine then tried different drivers ?

[Reeverdark]

Yes, iv tryed first updating to the newest one then uninstalling the card alltogether and installing it using the new driver pack but the same thing keeps happening.

[Sasqui]

Sounds like a problem that could be anything. I thought my X800GTO was bad in my prev compy, but it turned out to be a bad memory module... the crashes only showed up in games, go figure.

[ghanz]

If your Connect3D card is using the standard reference design which uses digital VRMs, the VRMs could be overheating. Try using a fan to blow over the VRMs.

For more info on this, check out vr-zone review of the 1950pro.
http://sg.vr-zone.com/articles/PowerColor_X1950PRO_256MB_With_Arctic_Cooling/4190-6.html

[Reeverdark]

i cant thats why im here.
the only thing i cant work out (apart for why is this happening in the first place) is why when im playing a game itll play it for a bit and be fine then do this to me and why Ground control 2 (a game that came out a few years ago) is more affected that the new Command & Conquor 3. at first at all sounds like hardware dammage but if you throw this into the equation its like there some little peace of code in the driver or the games them selves that trigers it to do this.
::- new found Info-::
i looked at event viewer in windows and it gave me this ::

Event Type: Error
Event Source: ati2mtag
Event Category: None
Event ID: 108
Date: 06/10/2007
Time: 18:03:55
User: N/A
Computer: BLADE
Description:
The driver ati2dvag for the display device \Device\Video0 got stuck in an infinite loop. This usually indicates a problem with the device itself or with the device driver programming the hardware incorrectly. Please check with your hardware device vendor for any driver updates.

For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.
Data:
0000: 00 00 00 00 03 00 52 00 ......R.
0008: 00 00 00 00 6c 00 04 c0 ....l..À
0010: 00 00 00 00 00 00 00 00 ........
0018: 00 00 00 00 00 00 00 00 ........
0020: 00 00 00 00 00 00 00 00 ........

[Kursah]

Yep I get that message if I push my OC's on my x1950 too far sometimes...it's been an issue with x1950's for a while, and for many it has been fixed. ati2dvag can sill cause problems, but for most it shouldn't be as bad as it used to be.

What are the clocks of your card? I know reference x1950's were less than 600 on GPU, I think 570 or 580Mhz. You should be able to see your clocks in CCC Overdrive, or ATI Tool .27 b2.

Also, you could get ATI Tool .27 b2, run it in Artifact Scan mode to see if errors occur during stress. ATI Tool's fuzzy cube is well designed in stressing GPU's, and I use it for stressing every graphics card I've ran for the last few years.

[Reeverdark]

GPU is standing at 575mhz
mem at
688mhz
ran the cube test and i froz after 10 seconds with a temp peek at 60-61c
what the hell is going on???

Edit:
64c exact before crash.

[Kursah]

hmm, well those clocks shouldn't be harming you, and that temp isn't horrible. On the RH side of the card, or the end with the power plug, is there a vertical line of capacitors?

The other people who've posted help have also inquired to find out whether you have VRM's or Capacitors. I have an older VRM based x1950pro, and my VRM's get up to 100C under load, at 120C the card will crash. The capacitor based cards don't have this issue.

Maybe go to the x1950 Pro Modded Bios thread, I thought I remembered seeing a modded Bios for a Connect3D card, (well Ketxxx made a few primarily for the 512mb cards on PCI-e) that helps with stability and such. May be worth a look for ya!
:toast:

[Reeverdark]

http://geizhals.at/img/pix/218553.jpg

well from what i can tell from looking at my card in my case it looks like iv got one single phase and one dubble phase VRM.

[Reeverdark]

OK! peoples i need all the teckys and phisasists out there to give me a hand as iv come to the conclution that the only way to solve this headacke is get the VRMs a heat sink but there are peramiter problems with this as follow-

I can not return sed card or get it repaired.
I can not get a theird party cooler as my case design is required to do suff (cant go into detail for now) and as moste coolers are dubble PCI slot in thickness it would be inpractical for the required appliction.
the only option i have is to A. bye two small heatsinks approimatly 1cm square (for single phase) and one 1cm by 2cm (for dubble phase). tubble there is there aint much out there and i dont know where to look.
or B. make my own (im an enginreeing student so making it isent hard) the only thing is i need to work out the dimetion and serfice area required to disipate at lest 75% of the max temp of 120C (as tested by a site posted by ghanz & and a suggested by Kursah. thanks guys.) wile having the above base sizes. so far i know the best matirale to be copper but as for the lenth of the grills and as to there quantity is were i need help. if i know the required area then i can come up with the appropeat design to soot.
This is going to be a tuff one to crack by if anyone can help then please do.
I have prefrence to make my own as it would be a good experiment but if there is any online for purchese then please do tell.
thank you all for your support.

[Widjaja]

I your'e will to possibly put your card at risk of getting worse go for it.
Sounds like something is definitley overheating.

My X1950pro idles pretty high, even after letting rest after a few games.

Reaches a max of 82deg at times.
Have you tried installing your old card again to see if those games work fine?

Something else in your system might have gone funny after the installation of the pro.

[cdawall]

bus speeds on the bus out of whack? that made my 7800GS crash in games

[Reeverdark]

Sadle this is the only grph card its had and i dont own anyother.

and as iv said iv only run the card at manufacturers standed settings. that gose for everything inculding my main bourd, there is no resone for this to happen nomaly, iv had it looked at by a pc lab eng and he charged me £10 for looking at it and fixing some other ises but he said there was nothing wrong with it and yet this is happening.

[ghanz]

Hi Reeverdark,

I think a normal memory chip heatsink(such as those attached to graphic card memory chips) will do. You can saw one into half & use the 2 pieces for each VRM.
You can use thermal epoxy to glue on the heatsinks. I personally think that the VRM is too small to use the super glue with thermal paste method.

http://techreport.com/r.x/ocztitan3/cardangle.jpg

[Reeverdark]

well iv replaced the thermal payst on it which seemed to have an affect but its still realy bad....

[Widjaja]

Well,
One sure proof option is to see f you can install this card in someone elses PC.

This will definitley tell you if it's the card or not.
Just doesn't seem right, having such low temps for a stock X1950pro, then having it freeze on you.

[Ketxxx]

try my custom bios posted somewhere near the end of "get your x1950pro bios here"! thread. id go into more detail but im unwell atm an me brain feels like itsd been stuck in a blender.