Powercolor x850 pro agp CCC problems

[Digideath]

Sorry to bother people. Ive got an annoying problem with my new gfx card that Ive been working on for days. Im using an powercolor radeon x850 pro agp 256mb and cant get the catalyst control center to work. To save me typing out everything again, Ive posted an extract from a report I sent to ati's technical support staff. Im still waiting to hear from them.

Quote:

Still didn't help. Spent 2 days studying and experimenting with the problem. I'm apparently not the only one using an x850 to have had this problem. I just cant find a solution.

In windows xp x32 and x64, I get no ccc. It doesn't initiate. I tested using 5.9 ccc and 6.2 ccc.
Tried using the 5.9 x64 basic cp and driver set. The cp does initiate but the config tabs are missing.
Tried using the Omega set for x32. It gave me the same problem as above.
Tried using the DNA set for x64. Same problem as above.
My x64 installation was installed 1 week ago. A new installation of windows doesn't help the issue.
BTW, in-case you ask, I always install the .net fix. I need it for more than the ccc. I also understand about fixing it to work with the x64 os.
I also use the ccc uninstall tool from your website. ( seems to do a more accurate job of uninstalling ).

I uninstalled the gfx card and reinstalled my old 9800 pro. The ccc took to it like a duck to water. I uninstalled the 9800 pro and reinstalled my x850 pro. The lack of ccc problem returned.

My driver set failed on me last night. Was playing Doom 3 and the game locked up then crashed to desktop. I got an error message about a service exception error with the main ati driver (ati2evxx I think). My system dumped its memory, rebooted, then gave me a serious system failure message. This has only ever happened once. It happened with the 6.2 x64 ccc set on the x64 os.

I understand a decent amount about ati driver sets. Ive used an ati card for 3 years and have never seen this problem before. I'm completely confused???

I am using ati tray tools app at the moment to config my card. This does appear to be working. The card appears to be working in all respects it just wont take the ccc. IE I don't think there is a physical problem with the card other than it is extremely tetchy. There is something glitching at a low level point ( a bug within the driver set perhaps? ) for this to happen.

Is there anybody out there who knows what is causing this problem and if there is a solution for it. Ive been hunting on line for days now without any joy. I can find people with the same problem but no solution.
It was suggested to me that my sounblaster live card could be conflicting with my gfx card if it was installed in the first pci slot. I checked this and sure enough it was sharing an irq with my audio card. I moved the audio card down one pci port but as was suggested to me but it made no difference.

I don't want to return the card unless I know for a fact that it is damaged. The shop I bought it from would charge me a fortune if they get it to work for themselves ( they usually try to avoid returns ).

Massive thanks for any help or advice given.

Sorry forgot my system specs.

Windows xp sp2 + Windows xp x64 pro corp.
Asus K8nE-Deluxe skt 754 mobo
AMD 64 3200+ Newcastle
1 gig corsair pc 3200
160 gig western digital sata raid + 80 samsung ide
Powercolor radeon x850 pro vivo 256 mb agp
soundblaster live 24 bit

[elmers]

Welcome to TECHIMO

We're here to be bothered, so don't worry about it.

Have you tried installing a fresh copy of xp? How about trying a driver pack that doesn't come with ccc? Or maybe you need .net v2 or the other ones.

[Digideath]

Ive had to reinstall both windows instalations to help me set up my sata raid.

Ive also tried using the basic control panel. It does initiate but I get no configuration tabs.

I didn't know there was another .net. I thought there was only v1.1. I will look into this and get back if it helps.
Im sure I saw an error message in the x64 event veiwer/aplications section about part of the .net not working...

[elmers]

Sorry I thought it was .net v2 i meant v1.1

[Digideath]

Found out that there is a v2.0 .net fix. Its in beta stages but its there alright. Unfortunately it didn't help any.

I'm still not convinced that there is a physical fault with the card. Ive read about a lot of people using powercolors x850 agp cards having the same problems plus the card absolutely rips apart 3d mark 05. I get a score of 8000+ using the default free version test. Honest. Surprised the heck out of me. I was only expecting about 5000. Ill remind people that this card is only a 12 pixel line pro version. Ive never saw a faulting card perform like this one before. The frame rate and quality levels exceed those described online for the card by a huge margin.
I know the drivers failed once playing Doom 3 but if the driver sets aren't installing properly then the driver would cause problems. In fact this happened using the 6.2 catalyst set which is known for giving people problems. As soon as I installed the 5.9 catalyst, the card stabilized and the performance improved massively. Still no ccc or control panel though.

Some people will be asking why I'm bothering so much. I do have a means to config the card ( ati tray tools ) and the performance is great. It's just not something I was expecting or something that settles well on me considering how much I payed for the card.

I've just noticed something strange I thought I'd add in. The card is supposed to run with a core speed of 507mhz and a memory speed of 520mhz x 2 however the bios for the card shows the settings as a core speed of 500mhz and a memory speed of 500mhz x 2. I dont know if that means anything but I thought id add it in.

[Digideath]

I have some extra information on this subject. I formated and reinstalled windows x64 again. Still no catalyst control centre however whenever I double click the shortcut the install places on the desktop for the ccc, I receive this error message in the ACEEventlog.

Quote:

The description for Event ID (0) in Source (ACEEventLogSource) cannot be found. The local computer may not have the necessary registry information or message DLL files to display messages from a remote computer. You may be able to use the /AUXSOURCE=flag to retrieve this description; see Help and Support for details. The following information is part of the event: 0000000003: 2006-03-02 00:48:04:437 Exception: From method (Main) caught exception (Object refrence not set to an instance of an object) Exception Called by: ATI.ACE.CLI.Implementation.CLI::Main processID:03000 threadID: (CLICcc ) domainName: (CLI.exe) assemblyName: (CLI.Implementation, Version=1.2.2217.17118,Culture=neutral, PublicKeyToken=90ba9c70f846762e)

This message occurs three times every time I try to run the control centre.

Just saw the same message in my 32 bit instalation aceeventlog.

[Digideath]

OK. My card is faulting. Just had this confirmed with ATI. They think its possibly trying to run 16 pixel lines instead of 12. Ati credit this to an incorrectly fitted resistor that is supposed to block a quadrant.

The card bios module is also suspect. The clock values for the card are deviating and I'm not talking about the overdrive either. The bios for the card reported clock speeds of 500/1000 two days ago. Yesterday, it reported as 412/1000. Today it reports as 557/1040.
Default values should be 507/1040.
The card has to be returned. Not easy though. This is not an easy thing to prove.

ATI STATED THAT THE POWERCOLOR X850 PRO IS ATI MADE AND THAT THERE IS A POSSIBLE BAD BATCH OF THESE CARDS IN CIRCULATION.
IF YOU HAVE A POWERCOLOR X850 CARD THAT SHOWS THE SAME SYMPTOMS AS MINE AND HAS A MADE BY ATI LOGO INSTEAD OF THE POWERCOLOR LOGO ON THE HEATSINK, RETURN IT IMMEDIATELY. IT MAY GIVE OF HUGE PERFORMANCE LEVELS AT POINTS BUT IT IS OPERATING OUTSIDE OF ITS PARAMETERS AND WITHOUT OC'ING COOLING MEASURES, USING IT MAY DAMAGE THE CORE.

EDIT:- I think the reason the driver locked up and ctd then caused serious system failure is because the card overheat.

EDIT2:-

Quote:

They think its possibly trying to run 16 pixel lines instead of 12. Ati credit this to an incorrectly fitted resistor that is supposed to block a quadrant.

Big hint for anyone looking to unlock the other four pixel lines