Its definitely a bad card, it has something to do with the heatsink shroud.
This is definitely the first time Ive gotten a Bad card on arrival. the last time i sent a Video card in was an ATI AIW 9600XT and that was because the fan locked up. every other card was fine, im not one of these "I got 1 bad card lets swear off XFX and ATI for Life people", i understand that some defects can slip through QA lines, and these things are so complex nowadays its not even funny (I remember my Trident 9680 1MB PCI Card, and it was small and the PCB was fairly empty, small Chip, 2 Memory Modules, 2 Sockets for expansion.)

IF the RMA goes through, hopefully the replacement comes back fine. in 2 months I'll be fully re-located and $800/less a month for rent so I'll be able to maybe upgrade his system for giggles.

Sometimes the system wont boot up unless i put upward pressure on the heatsink then it will boot, as soon as i take my finger off the bottom of the shroud the image will start to flicker etc etc and cut off eventually or lock up the system.

I put my 8800 in the slot and none of the image flickering occured even with wiggle, the only problem from that was the ATi/nVidia driver mess (since I couldnt boot to uninstall the drivers for the ATI card.

once i got past all the BSODs and wiped the drivers and cleaned them out i was fine.
(it took me forever to get win7 stable again after this)

so I put the 5830 back in for now until XFX gets back to me, then I'll have to ship this puppy out and wait for it to come back.


[EDIT 03-13-2011 12pm]
Its weird because for a WEEK, he played AvP, TotalWar ShoGun II Demo, a few other DX10 Games all max graphics settings no problem.

Then all of a sudden 2 minutes into the first rendered scene on FFXIV everything goes back to blah.

Like i said though, I've had an open Ticket with XFX since I assembled his system, but its definately something with the heat shroud. if I wanted too, I could Prolly put a tight rubber band around the heat shroud just after the slot bracket to maintain the pressure, but why go as far as modding something thats defective. lol. (im the type of guy that doesnt like to RMA and return crap unless i have too.)

the shroud itself is sturdy and nicely mounted on the rear end where the power plugs are, the front end by the PCI Brackets is very loose, and I mean VERY, I can move it about 1/4 to 1/2 inch without force (and moving the shroud moves the heatsink and pipes on the inside as well), it hangs when its mounted in the system. which is why its only stable when I use my finger to apply pressure.

so either its loose and something is shorting, or its loose and the GPU is over heating (which GPU-Z and CCC report upper 40s and mid 50s when the flickering starts.

now none of his games Run stable, they all crash out with unknown error after a few minutes of play.

Well, as a result of the countless BSODs and the ram timing Issue, the windows 7 SFC Utility is reporting corrupt files (and the store files as well, so they cant be repaired), he’s already getting application crashes as a result, so I guess I get to wipe and reformat/re-install again, yippie!!! AFAIK Ram is good that the set timings (tested again in Memtest and Windows Diag, also ran checkdsk with the 2 switches that took 3+ hours but reported no errors or bad sectors.

On the GFX Card Note:, I have a lil Cell phone book holding the heat shroud up so the system will boot stable without video flickering etc. I’m gonna get a nice size rubber band and wrap it around the card to hold the heatsink/shroud to the card.

When I take the card out of its slot and look at it from above, the PCB is Curved/warped a little.

Either xfx isn’t seeing the picture or they are beating around the bush with giving me an RMA,

These are my last 2 ticket posts and responses:
[ 3/8/2011 10:39:20 PM] The windows driver timeout was only happening in Adobe Flash. The instances where the image would flickering w/ corruption and/or cutting off was happening at random from the time the system is powered on and is directly related to the heatsink shroud or PCI-E Slot. as applying gentle upwards pressure on the heatsink shroud would resolve the image flickering temporarily. However the card has not exhibited this problem lately.

[DANIEL_E 3/9/2011 1:02:59 AM] Continue to keep an eye on it and let us know if anything else comes up.

[ 3/12/2011 8:56:16 PM] I tested the card playing some new games. when playing the image would start to flicker on and off or to a grey screen with lines, and the mouse cursor gets corrupted if in motion, and eventually the game crashes and windows says that there is no DirectX Device present and press ok to close the program. Even after exiting the game the mouse cursor remains corrupted til I reboot the system.

[DANIEL_E 3/15/2011 4:05:21 PM] Hi, go into the catalyst control center. Then go to over drive. Once in overdrive unlock it and then check enable overdrive. Then click defaults and ok. Then click apply. Then uncheck enable ati overdrive and then click ok. Daniel



Ok, I don’t get why they would see that as a possible solution, I didn’t overclock it at all, the only thing I did was turn up the Fan speed, I noted that in my opening ticket post, that it wasn’t overclocked and was problematic right out of the box, and is new not refurbish.

So what does “resetting the defaults” gonna do besides put the fan speed back, its not a software /setting problem if I can use my finger and stop the flickering/corruption by applying upwards pressure on the shroud over the heatsink.

Hrmm….. I’ll do as the tech says but im 100% sure its not a solution, and the sooner I get an RMA# the sooner a replacement gets here the sooner I can wipe the driver and re-install windows 7

Not to mention download the 48 GB worth of games from steam AGAIN (btw, I don’t use steam much except America’s Army, but the account I made for my brother has FFXI Ultimate, AvP, and about 10 other games (yes he went bonkers downloading), if I completely format the drive re-install everything his steam account will have a record that he purchased these items and its just a matter of re-downloading them?

The HAL of the system should remain the same just storing CLEAN O/S, because playing with the ram caused corrupt files, then the BSODs with the GFX Card, oh and something I found that was corrupting files on my System and his.

WINDOWS 7 TURNS OFF THE HDD AT RANDOM TIMES EVEN IN “HIGH PERFORMANCE” POWER MODE.
I had to Override it to “TURN OFF : NEVER” to get the random power downs to stop, like it would power down in the middle of downloading and copying files.

Re: the Steam d/ls, yes. All you need to do is log in and d/l the games again. That's one of the things I like about Steam.

The rest of the stuff you're going through is so far over my head I see contrails. I do believe you're getting jerked around re: your GFX card; physical abnormalities should be an automatic RMA IMO.

XFX finally gave me an RMA Number, so now i get to pack this puppy up and ship it to Cali....

So XFX Gave me an RMA#, and I'm in the process of shipping that out,

Random BSODs have crept up lately citing, You guess it " MEMORY MANAGEMENT ",

As well as random Programs crashing, Especially Steam, (Says it wassnt installed correctly, wait a minute double click again and it works fine?), lol, Strike Fighter 2 Installed Fine, SF2 Vietnam's Installer gave an Installer Error while decompressing files, same with Europe and Israel, But all Installers work fine on my systerm. Windows Live Messenger Fails to load past teh login, and All browsers crash when going to youtube (the Flash Plugin crashes alot too.)

SO I think the one stick that test bad is, in fact bad, even though after manually setting the timings in the BIOS seemed to help it pass memtest for a while (but it passed before, then failed miserably, then passed, but only one stick, the other stick always passed regardless of slot it was in, so It shouldnt be the mainboard. I also stumbled apon a EuP jumper setting in the book, that would limit power to the mainboard, gonna take that off as saving power isnt important to me.

Ran CHKDSK /r and the other switch to pass the 5 phase test, no errors or bad sectors.
SFC/ Scannow reports files are corrupted as well as the stores (ntfs.sys is the main one that catches my eye, along with a few other dlls).

So The GFX Card will be shipped out, and replaced, which i have a temp workaround of using a Custom built "support" to hold the Heatsink shroud up to the card which stops the video card issues.

Ram is back to causing problems and Will have to be RMA'd,

But, Im gonna wait a few weeks order antoher matching pair of sticks when I order a better heatsink/fan, the same as the ones I have now, and use those and send the older ones in for RMA (through G.Skill). The New Sticks will prolly get here before the Video Card comes back from RMA (East coast to West coast = 4 days shipping UPS Ground, few days to process RMA and crap, and another 4 days back). So I'll test the new ram right off the bat, and if it passes, ship out the first set that has the bad stick, and when it comes back and tests clean he'll boot 8GB.

Im gonna run Prime95 and some other stuff again to see if the CPU passes, which it should.

I might end up getting a better mainboard in a month or so. if problems persist.

Either way Im gonna end up wiping his Drive and re-installing windows fresh, again.

A little update on the system.

The problems with the memory and windows come and go,

The one thing that seems to be causing everything though is the Graphics card shroud. The shroud is loose towards the front of the card and has way too much "wiggle" and when it is mounted in the system the shroud hangs about 1/4 to 1/2 inch from the card's PCB at the front while it remains flush with the PCB towards the back where the 6 Pin connectors are.

I temporarily resovled the issue for my brother while I am going in and out of town by placing a small paper back book in his system vertically under the card pushing up on the shroud, keeping it close to teh PCB (about 1/8 to 1/10 of an inch instead of 1/4 to 1/2.) I also kept the side off. The book works loose every few weeks or so.

Last night I gave him a visit and his system, ran some error checks and put heat probes in a few spots.

Took his GFX Card out, wrapped the rubber bands aroudn the frotn end to Put the shroud flush with the PCB of tehGFX card to eliminate that GFX issue. MOved his Ram from Dimms 00/01 to 02/03 (the ones away from the CPU, as his heatsink blows the hot air onto the Ram and stuff). re-mounted the fan on teh side panel to be overtop of the ram/cpu blowing fresh cool external air onto the CPU/Ram area.

I'll Eventually RMA the 5830, but I want to have a replacement in hand before i send it out, I'll prolly buy him a 5850 or 5870 in a month or so when they go on sale again, and keep the 5830 for myself to run that on my system in place of my 8800 GTS until I upgrade.