Beyond a joke now

Auraomega · 8 Jun 2011 at 15:37

I've had nothing but issues with my HD5850 since I bought it but now it's pushed me over the edge this time.

Having reinstalled Windows a number of times the only thing I can blame the slew of issues on is the graphics drivers, every single time my computer decides to stop working a graphics crash occurs right before, what follows is:
1) Boot into Windows, various programs start to crash, followed by a BSoD.
2) Following this happening at least once, the computer will boot but all I'll have is a black screen and a cursor.
3) Next, boot into safe mode, all works fine, try and system restore but I get nowhere.
4) Insert Windows 7 disc, use system restore from there, works for a couple of weeks.
5) Repeat up to step 4, where the system restore fails.
6) chkdsk /R
7) Repeat step 4, works again but only for a few weeks at most.
8) Point of no return, impossible to get up and running again (and believe me, I've spent days trying to), at this point I reinstall.

I've even gone as far as actually reinstalling the OS rather than using an image (incase anything was corrupt). As far as I can tell my system is perfectly stable, I've gone back to stock settings and the problems still occur, my system stays well within recommended temperatures overclocked so not an issue at all at stock. I get various BSoDs but they always seem random (I.E. not indicating any hardware issues). I'm at the end of my tether, I don't have enough money to replace the card but I'm at the stage now where I'm about ready to give up on the PC which seems a shame considering the vast amount of money I've spent on it!

Dominor · 8 Jun 2011 at 15:44

Sorry to hear that dude

If it is the card causing the trouble you could try selling the card on Ebay and switching to Nvidia to see if that fixes the problem, only thing I can think of really. I had plenty of driver problems in games with ATI but never stuff like you have so I feel your pain!

maxilive · 8 Jun 2011 at 15:44

Messed up hard drive?

I'm not sure tbh but sounds annoying! Can you try the card in another PC to see if its the card alone causing the issues?

wolvers · 8 Jun 2011 at 15:56

Sounds more like an HDD dying. Did you check the error logs?

Auraomega · 8 Jun 2011 at 16:06

Sorry should have added every time it's been on a different hard drive, all brand new. 3 hard drives failing a few months after purchase seems a bit far-fetched (I did have that thought though and checked all 3; all perfect).

I would swap it into another PC but this is the only PC powerful enough to run it, all my others are low power systems, the PSUs couldn't run the card let alone the rest of their system as well

I have an old 7800GTX which I picked up from my last job, only problem is it was a faulty card which I cooked to working goodness, I don't think swapping a problem component for a previously-faulty component is a good move; I might grab myself a really crappy PCI card to see if it causes any issues, it's just frustrating because it can take months for the fault to develop and the 5850 can barely keep up with my monitor so I'd be limited on games.

This is my first ATI card and will definitely be my last as in the past year I've suffered more crashes from this card than I have collectively in the countless years I've been using PCs. I know nVidia aren't exactly angels but in comparison they've got one hell of a halo.

wolvers · 8 Jun 2011 at 16:34

Checked the RAM? Did you actually check the error logs though to see what they say?

Even if it does turn out to the the GPU, don't let one piece of hardware muddy your opinion of AMD stuff. Generally it's very good and so are the drivers, as is NV stuff.

uv · 8 Jun 2011 at 16:43

Have you run memtest?

skanky · 8 Jun 2011 at 16:58

Dominor said:
Sorry to hear that dude If it is the card causing the trouble you could try selling the card on Ebay

sorry but thinking like that is exactly why i dont buy from ebay. if its a problem with the card, all is going to happen is that the OP will get found out and have to refund it the person who brought it?

if its still in warranty, RMA surely?

whipgeez · 8 Jun 2011 at 17:09

speaking of rma's (off topic) my rma has been confirmed as refunded on my account page today, how long does it take from that to see the money back in my account?

joxang · 8 Jun 2011 at 17:14

If it is the card causing the trouble you could try selling the card on Ebay and switching to Nvidia to see if that fixes the problem,

Lol yes...because OP is having an AMD issue

.

Marine-RX179 · 8 Jun 2011 at 17:40

What PSU are you actually using? It might be that your PSU is not feeding the graphic card stable enough power and causing it to keep crashing...

Also, have you tried the card on a different PCI-E X16 slot (if your board has it)?

Operation · 8 Jun 2011 at 18:49

skanky said:
sorry but thinking like that is exactly why i dont buy from ebay. if its a problem with the card, all is going to happen is that the OP will get found out and have to refund it the person who brought it?

if its still in warranty, RMA surely?

Indeed. 'I've got a problem so I'll transfer the problem to someone else'.

I think that attitude stinks personally.

james.miller · 8 Jun 2011 at 18:51

when did you buy it, and why havent you sent it back for a replacement sooner if its cuase you nothing but problems?

Freddie1980 · 8 Jun 2011 at 19:07

Its hard to believe that these so called driver issues are exculsive to only the OP, if it were driver issues with the problems you have outlined I would have expected to have seen over a dozen threads with people with the same issues. It looks like you have a faulty PSU and/or motherboard or the video card could be faulty but that's hardware not saoftware.

I don't know why but drivers seem to be an easy scrapegoat for people who all manner of problems.

drunkenmaster · 8 Jun 2011 at 19:16

I'd say, list entire system, also I'd say why do you think the system is totally stable if the system isn't stable, just because temps were low overclocked means almost nothing in terms of stability.

First things to check when you upgrade to a (usually) more powerful gpu with an existing system is voltages. Add an extra load onto the psu and an overclock or even a stock speed can drop a little on voltage beyond stable levels. Plenty of mobo's undervolt underload, infact, most do to some degree, drop an extra load into the mix and that undervolt can become unstable.

Likewise a more powerful gpu can simply ultilise a cpu/mobo/memory/psu a little more than that system was stressed before.

Your description of crashing isn't detailed enough, what makes every crash a graphics crash before your system stops working? What programs crash before the system BSOD's, what is the BSOD you get. For the record, AMD and Nvidia BSOD's are insanely common when both the gpu is at fault and the system being unstable for really any reason. Mostly as its the single most complex driver thats the least stable so both AMD/Nvidia suffer a lot of "the gpu is at fault" BSOD's when the cause is usually memory or undervolting a system.

I would normally start with a bump on memory voltage, then cpu, northbridge, not miles just a notch or two. Check spd settings on memory aren't overly aggressive, more so if you're running a larger amount of memory, ie a system with 2x1gb sticks might run 1T with no issue, 4x1gb sticks might be unstable in anything but 2T.

For me the biggest reason to suspect gpu's(of any brand) is games with all kinds of texture bleeding, lots of white dots everywhere, banded lines of dodgey green/pinkness.

Crashing out of games without those things, or windows, or BSOD's would normally be system instability rather than gpu.

I'd also suggest maybe the mobo's on its way out, the issues do kind of sound like HDD issues, IE slow break down in stability of windows, cured by a reinstall, slowly getting worse then needing a reinstall. If its not the HDD's then I'd go with mobo causing that data corruption. Again in terms of gpu, if it was crashing, it would reinstall windows and crash and kill the OS straight away if it was going to, not last a while.

Dodgey cpu's the same, a dodgey cpu will rarely last weeks on an OS install, it will just be right out the window, gpu/cpu are essentially too intergral to work but only for a few weeks. Hdd/memory/mobo, or even virus type OS screw ups would be more likely with that kind of pattern of degredation.

drunkenmaster · 8 Jun 2011 at 19:17

Freddie1980 said:
Its hard to believe that these so called driver issues are exculsive to only the OP, if it were driver issues with the problems you have outlined I would have expected to have seen over a dozen threads with people with the same issues. It looks like you have a faulty PSU and/or motherboard or the video card could be faulty but that's hardware not saoftware.

I don't know why but drivers seem to be an easy scrapegoat for people who all manner of problems.

To be fair, almost any system instability that doesn't result on a no boot or a no load into windows, will normally cause a driver to cause a crash, so BSOD's often point to drivers as the fault, and Nvidia/AMD cop a lot of flak as gpu drivers tend to be the first to go..... coupled with gamers generally overclocking/having higher end systems/upgrading more often making them more liable to dodgey parts.

Auraomega · 8 Jun 2011 at 20:39

Whew big post here, I'll try and address all points. Finally got the system up and running by completely removing the ATI drivers and reinstalling again. Been playing a few games and all seems ok for now. Actually, just noticed an issue whenever I try and use the manage window:

"Application failed to start - side-by-side configuration incorrect", on EXPLORER.EXE, I can see everything else that I've tested so far, if it's not one thing is another :mad:

Marine-RX179 said:
What PSU are you actually using? It might be that your PSU is not feeding the graphic card stable enough power and causing it to keep crashing...

Also, have you tried the card on a different PCI-E X16 slot (if your board has it)?

OCZ ModXtream 700w. I only have 1 PCI-E x16 at x16, the other is x4 I believe, I could give it a try.

james.miller said:
when did you buy it, and why havent you sent it back for a replacement sooner if its cuase you nothing but problems?

Everything was bought in January 2010, the graphics card was more likely February due to how difficult these things were to get at the time. I'd check my OcUK account but it's disappeared (we've been in contact about the issue, hopefully it'll be sorted soon). I've always believed and still do believe that the drivers are the issue hence why I've never RMA'ed the system.

drunkenmaster said:
I'd say, list entire system

i5 750
Asus P7P55D LE
G.Skill Ripjaw 2x2GB (2000MHz)
Club3D 5850
OCZ ModXStream 700w PSU
500gb F1
1tb F3
Xonar DG

drunkenmaster said:
, also I'd say why do you think the system is totally stable if the system isn't stable, just because temps were low overclocked means almost nothing in terms of stability.

Stress tested as a whole then tested individual components after the first failure, nothing was picked up.

drunkenmaster said:
Your description of crashing isn't detailed enough, what makes every crash a graphics crash before your system stops working? What programs crash before the system BSOD's, what is the BSOD you get. For the record, AMD and Nvidia BSOD's are insanely common when both the gpu is at fault and the system being unstable for really any reason. Mostly as its the single most complex driver thats the least stable so both AMD/Nvidia suffer a lot of "the gpu is at fault" BSOD's when the cause is usually memory or undervolting a system.

The first thing is constant failing of the ATI drivers followed by the screen going messed up, sort of like jagged lines of interference, then the PC reboots (which when I think of it is odd). When I come back into Windows the first thing I see is various errors to do with ATI drivers and software, as soon as I start to load any applications they usually crash, eventually (could be 30 seconds, could be 30 minutes) the system will bluescreen with some message or another, bad pool, memory management, I saw something to do with data access once.

drunkenmaster said:
I would normally start with a bump on memory voltage, then cpu, northbridge

Memory is rated for 2000MHz at 1.6v, the CPU is at 3GHz using the same voltage as makes it stable at 3.5. No northbridge.

drunkenmaster said:
Check spd settings on memory aren't overly aggressive

Just running at what they're rated at.

drunkenmaster said:
I'd also suggest maybe the mobo's on its way out, the issues do kind of sound like HDD issues, IE slow break down in stability of windows, cured by a reinstall, slowly getting worse then needing a reinstall. If its not the HDD's then I'd go with mobo causing that data corruption. Again in terms of gpu, if it was crashing, it would reinstall windows and crash and kill the OS straight away if it was going to, not last a while.

I have suspected the mobo might be faulty, however I haven't one to try it against and as mentioned the system can be working fine for months then all of a sudden the slew of issues arise, as soon as Windows is reinstalled it's fine again (hence why I think software rather than hardware). It's hard to make it repeat and the format of crashing always makes me assume hardware.

Umm, one other thing that might be worth mentioning... very rarely but every so often when I turn the PC on all the fans turn on, but nothing else seems to, system doesn't turn on, idiot lights don't flash, hard drives don't spin and the monitors don't react.

Freddie1980 · 8 Jun 2011 at 22:32

Auraomega said:
Umm, one other thing that might be worth mentioning... very rarely but every so often when I turn the PC on all the fans turn on, but nothing else seems to, system doesn't turn on, idiot lights don't flash, hard drives don't spin and the monitors don't react.

You kept that part a bit quite, your whole PC sometimes fails to boot correctly and your having BSOD's and crashes all over the place yet you seem sure that that AMD's drivers have code built into them that exclusively gives your PC instability.

As I said before if it were driver related I would expect to see a dozen or so threads around the place or one giant thread but that's not the case so it seems very unlikely IMO that drivers are the problem (especailly since you have reinstalled Windows from scratch several times pretty much rules out software issues unless your running some dodge 3rd party/beta application). So from what you have told us previously and what you have just decided to tell us now all points to either a faulty PSU (more specifically a damaged or unstable 12v rail) or less likely a faulty motherboard. From you system specs I see your using an OCZ Mod Stream which on paper is more the capable of powering a 5850 but if the 12v rail on your PSU is unstable it means components such as your motherboard and video card i.e. the components that need power from a 12v rail are going to fall over and crash.

kingy666 · 8 Jun 2011 at 23:07

Just a thought, but do you have the latest Bios for your motherboard (which is version 2003)
Looking at the Asus website and there have been around 10 bios versions for that motherboard which most improve system stability.

Notes for version 2003
1. Improve memory compatibility
2. Improve system stability
3. Update CPU Level up function
4. Update Marvell Option ROM
5. Support new CPUs.

Auraomega · 8 Jun 2011 at 23:52

@Freddie1980: I assumed some incompatibility nothing more. I don't like the thought of having faulty hardware especially the mobo, and as mentioned the fact that I get one crash and the system then falls to pieces, then reinstalling magically fixes makes me think drivers, not hardware. What would be nice is some method of testing the mobo to see if it is indeed faulty, I know of none myself but maybe you do? I've checked the PSU before as best I can although it's hard to get an accurate reading with a DMM if the voltage fluctuates just enough to crash but not long enough to show. Could the constant corruption be linked to the mobo? I've had problems with certain sata drives before as well as data regularly getting corrupted after the initial crash.

@kingy666: I think it's more just an extra thing to throw in to the list, I'll try it now my system is mostly stable though and it can't hurt to try.