Hi All,

I am having major problems with my new HIS ATI RADEON X1950 PRO ICEQ3 Turbo 256MB GDDR3 HDTV/Dual DVI (AGP). Before I seize the bull by the horns and spend anymore time on the trouble shooting process I need to know if the problems may be due to the card being faulty. I don't want to waste anymore time if all I need to do is request an RMA from Overclockers. To that end, I would appreciate your consideration of my problem and your experienced advice.

For background info my system:

Athlon 64 3200+ 2.2 Gig Processor
Asus K8VSE Delux Socket 754 MoBo
2 x 512 GB 400 DDR RAM (1024 MB) - 1 x Crucial, 1 x another brand (Tested using Windows Memory Diagnostic tool - No Faults found)
Jeantech (JN120F-450-AP) 450W PSU
Windows XP Home SP2 with all updates installed (inc .NET)
SB Audigy 2 ZS
Maxtor 160GB ATA IDE HD
Maxtor 80GB ATA IDE HD
NEC DVD Writer

My installation is about 2 years old and the system was built by Carrera SSC.

This whole story started when I got Ghost Recon Advanced Warfighter for Christmas. It wouldn't run for long on my old Nvidea Geforce 6800GT and used to crash/lock up after a few minutes. I eventually uninstalled the game and decided it was time to upgrade!

To cut a long story short, I ordered my swish new X1950 from Overclockers on 16 Jan 07. Delivered on 18 Jan 07, I removed my old GC (to pass on to my son) and uninstalled it using Add/Remove Programs (judging by the advice here that was my first big mistake). I fitted the new X1950 Pro the same evening and rather than use the 'out of date' drivers as shipped, I downloaded and installed Catalyst 7.1 (2nd mistake perhaps?).

Everything looked fine, so I installed Ghost Recon Advanced Warfighter and patched it up to the most recent version. Having auto-detected the card, GRAW started at 1280x1024 and impressively launched into the skydiving scene. After playing for about 10 minutes I started to get some artefacts appearing which basically took the form of little squares flickering all over the screen. I couldn't capture a screen shot but here is one from COD2:



The longer I played, the more abundant the artefacts became until I just couldn't play anymore - it was like a Virtual Migraine!! On escaping back to the desktop, the screen was covered in tiny blue pixels which stained any window I opened and moved.



I have since observed the same artefacts in:

Half Life 2
TOCA Race Driver 3 (plus textures slipping into other areas)
IL2 Forgotten Battles
Battlefield 2
Call of Duty 2
Call of Duty United Offensive
Medieval 2 Total War

The artefacts appear in both the 3D game and after that, also the 2D game menus.



I have also seen vertical slices in the display like somebody has got a knife and sliced the image causing a section to go missing or slip down lower. The game menus also smudged, ie; it looks like somebody smudged the menu options as if they were wet paint!



Rebooting does not clear any of these problems, in fact with each boot-up the artefacts seem to get progressively worse until eventually my system locks up during the transition to the Windows Login screen. The only way to clear the lockup is to boot in SAFE MODE and uninstall the ATI Software and drivers. I can then reboot and go from there.

I have since installed 6.12 without any improvement and reinstalled 7.1. My system passes all DirectX Diagnostics before I go into a game, but fails the DirectX 8 and 9 tests if I retest after seeing artefacts. I have not over clocked the X1950 and the card temperature appears to be between 40-45 degrees when idle, 55 to 60 degrees when I come out of game.



The fan is normally Very Quiet, especially when compared to my old 6800GT, but at the moment that is the only thing that impresses me about this card. Having wasted 7 days of my life trying to rectify this issue I finally stumbled across this forum and the wealth of information it contains.

In the last 3 days I have also now tried several clean ups with DriverCleaner (Nvidia and ATI), LiveUpdate and reinstall of the 7.1 CCC drivers, but the problem remains. I have also tried ATI Tray Tools without success though it did show up my AGP voltage at around 10.## ? Could this be the problem? Maybe my Jeantech isnt pushing out enough voltage?

So to pose my original question – should I carry on troubleshooting this, the ultimate act being to reinstall windows or am I just wasting my time. Is the card screwed and should I just request an RMA? Should I buy a new PSU before I proceed anymore. Anyone else seen this type of Artefact?

Really appreciate your advice/help.

I suspect power supply problems. A quick search of t'interweb reveals that your power supply has two 12V rails, one at 14A and the other at 15A. Which sounds enough.

However, I also found mention that the first one is dedicated to the motherboard, which may be a problem, as everything else in your system is then trying to run off 15A.

Wanted to add to the other comments: Artifacting can also be due to something being run out of spec - you're obviously not overclocking, but the fact that things get worse slowly suggests something might be slowly overheating making things progressively worse. Hard to say what exactly, could be GPU, could be video RAM, could be both. Have you got monitoring software for the card's temps, and what are they? Try (as a test) taking the side off the PC, and pointing a desk fan at it's internals to help cool everything down, and see what happens.

Your 10.xx voltage reading in ATI tray tools would back up the underpowered PSU theory.

I have just re-built my system and decided the PSU wouldn't be enough so got a Corsair 520W as recommended by a few people on here. I have had no problems with my HIS 1950 AGP so far...

I think you may have some borked ram! maybe rma time!

LOoks like bad memory mate, rma it!

Ok Guys,

Thanks for your time and advice. I think I'll go for the RMA option. I'll let you know what happens.

Cheers!

STILL ON THE HORSE! X1950 Pro & PSU Voltages

Hi All,

Well I RMA'd it and OCuk tested it for an hour or so and didn't find any problems. They reckon I have a compatibility issue. Anyway, £33 later I've got the same card back and I'm trouble shooting all over again.

I think the first thing I need to rule out is my PSU and I'd therefore appreciate some help in interpreting these charts. Here are 2 screen shots of my voltages; first running under a normal 2D load and the second taken after ALT-TAB'ing out of the the HL2 Lost Coast Video Stress Test during the church scene.


Normal Operating Conditions


Video Stress Test

As you can see the 12 Volt rails have not changed at all (10.90 and 2.61), in fact the only real change (other than my CPU temp) is that voltage readings are not so steady. I read at www.playtool.com that the 12v rails will normally never be exactly 12 volts, but should always be within 5% of that.

I quote ... According to the official specification all three of those voltages must be within 5 percent. But in real life it's better if they are closer than that. So 3.3 volts should be between 3.2 to 3.4 volts. 5 volts should be from about 4.8 to 5.2 and 12 volts should be from 11.6 to 12.4. If the voltages are outside of that range, it's not proof of a power supply problem but it's not good. Some video cards tolerate voltages which are off better than others. With most new video cards, the 12 volt value is the one which is most likely to cause problems. If the 12 volt value suddenly drops when you start up a 3D game, that's a bad sign. But ultimately, the only way to know for sure that a power supply is the cause of your problems is to have the problems disappear after swapping in a better supply.

If this is correct then both of my rails are well off! I'm not even sure which rail is feeding my X1950 Pro, I can say however its the dedicated 6 pin PCI-E.

So firstly, what do you PSU guru's reckon to my charts? Do I need to blow another £60 - £100 on a new PSU or are these readings normal?

Secondly, scouring the Forum for a solution hasn't provided much new info, but on the surface there do seem to be more ASUS Mobo users experiencing the brunt of the problems. Anyone got any thoughts on that?

As always, I really appreciate any advice you can give.

Cheers.

Sorry to tell you but the psu is pants. 10.90 is extremily low. I wouldnt consider a system powered enough if it wasnt at least 11.80 or more.

If you do get a new psu i recommend the corsairs 620w as i got it and i get 12.288 at the mo on idle. I think it usually is 12.33 or more tho i aint sure to be honest but 12.288 is what pcprobe says at the mo. The lesser corsair 520w should be easily good enough too. The 520w is 70 ish and the 620w is 100 ish.

Hmm strange that the vcore is 1.49v tho. Surely that cpu would be fine at a lower voltage?

Hi Mate,

No need to apologise for telling my PSU is crap ... I had a feeling it was after looking at the charts! Anyway, it served me well for 2 years and only cost me around £35. At the end of the day, you get what you pay for!

My only concern is that I shell out some more dosh for a new PSU only to find my card X1950 Pro still won't work ... still at this stage I think it's a risk I'm going to have to take.

I'm a bit miffed that OCuk billed me so much for P&P on my RMA - I can live with the testing fee £11.75 which is very reasonable, but the £10.62 P&P eachway seems to be daylight robbery! I did query this via email explaining that I not only intended to purchase a new PSU from them, but also future system upgrades but I was not offered any kind of refund or sweetner. Dissappointed by the response, I guess I just have to vote with my feet.

So the moral of the story so far is, if you are going to RMA something make sure it is faulty otherwise you get seriously stung for delivery costs!!!

Hi All,

Just to keep this thread up to date, I am pasting a reply I got from JEANTECH regarding my PSU ... Looks like I definately need a new one then.

---------------

Hi,

Our R&D department said that the RADEON X1950 card on the PC system will consume 250W-300W on its own. However, you need to take other parts into consideration, the 250-300W is not included anything like HD or any other parts you install, therefore, we don't think our JN120F-450-AP would be enough to support your whole system. You might need to consider a bigger PSU. The best thing is to find out how much power each part you use in your system then select the most appropriate PSU for it. Hope this help.

Regards,
Gina

ChemHazMat said:

Hi All,

Just to keep this thread up to date, I am pasting a reply I got from JEANTECH regarding my PSU ... Looks like I definately need a new one then.

---------------

Hi,

Our R&D department said that the RADEON X1950 card on the PC system will consume 250W-300W on its own.

Click to expand...

How much?? I don't think so! Try 66W!

Hope you get it resolved mate.

I would have thought 450w would have been more than enough tbh buddy!

ChemHazMat said:

Hi,

Our R&D department said that the RADEON X1950 card on the PC system will consume 250W-300W on its own.

Click to expand...

I don't know what they've been smoking, but your PSU is more than upto the job unless its faulty. What are the voltage readings in the BIOS hardware monitoring page?

Id try it in another machine if you can, or pop into your local if you can't and see if they'll quickly test it, what i done when i thought i had a faulty card before, as that looks like borked Ram on the card to me, and theres loads of faulty HIS AGP x1950pro's so id say it iss the card thats faulty, its a HIS.

Well lack of power can cause problems. As when i tryed my card with my akasa 400w with the 4pin to 6pin adaptor, it did boot and stuff but crashed in games after 10 mins. Also had monitor not coming on and also computer wouldnt boot sometimes. When i did game i did notice some fullscreen pink textures when exiting gta sa. When i got better psu all them probs went away. Only prob i have now is this card isnt compatible fully with gta sa mod samp which is a known issue in samp 0.1b but they said fixed in next release.

Based on everything I've read about these cards, I'd go out on a limb and say its more than likely to be insufficient cooling on either the card's RAM and/or the VRMs. If you fancy it, I's suggest completely removing the heat sink, clean everything up and reapply with a good thermal medium making sure contact with components is good.

Goosee:

Thanks for your advice but this is a brand new card - The last thing I want to do is void the warranty and try a DIY heatsink/thermal re-paste. If the VRAM is overheating then it must be a fault in design or workmanship and if that's the case, I'll be asking for my money back. In any case over the last few nights I have taken the side off my case and this has dropped my CPU temp by about 5 degrees across the board. Despite this the artefacts still appear.

Loadsa:

This card has already been RMA'd and tested by OCUK. They sent it back to me stating that the card was not faulty and in addition to the testing fee, stung me £22 for postage! I have to trust OCuK and assume that the card is working correctly.

The rest:

Thanks for all your comments guys. If you look above at Post 7 you will see my voltage charts which were recorded in a 2D desktop windows environment, and later during the HL2 Lost Coast video stress test. There was no marked difference in the charts which surprised me. My 12 Volt Rails came in at +12v Rail = 10.90 and -12V Rail = 2.61 respectively.

My Hardware Power readings in BIOS are showing as follows:

VCore: 1.488
3V: 3.216
5V: 4.972
12V: 11.520

No other values are shown and I can't see anyway of changing them - the only option I have available is to ignore them.

Definatly way too low, even my q-tec gave me 11.4 volts (under stress, when idle 11.8 v). 10.9 is definatly a way underpowered rail, under any conditions this should never fall below 11.7/11.6 volt and not too high either not above 12.5 or higher, as the line itself says, 12 volts is the aim voltage...

The thing you said about your 6800 card (about the lock-ups and crashes) only confirms a tottally rubbish psu, a slow gfx card causes low framerates but not crashes and since the 6800 series have shader 3.0 they should run any game out atm on at least lowest settings...

My 2p worth.

Firstly, it sounds like you need to try a new PSU mate. Those figures seem too low. Even if it doesn't solve your problem, it won't be a bad investment - just don't go mad thinking you need a 1k model for a not overly demanding card

The only thing I'm not comfortable with is that although a crap PSU can cause all kinds of locking and stability problems (as you mention you've previously had) it shouldn't cause artifacts to get worse for the longer you play - something else sounds a bit screwed. Try running with the case lids off (heat) or if you can remove 1 stick of RAM (you mention 754pin MB which I think will work with 1 stick?). If not, can you swap RAM with your son?

If you change both and the problems still exist, I'd be having another word with OC about the card... Until you've tried both going to be difficult to push that esp. as they've already said it's OK - Good luck fixing it, let us know

get yourself an Multimeter and check the rails voltage readings from software and BIOS mean nothing the PSU "should" handle your rig no problems

I would say borked RAM by the look of it