System Restart Problems

Man of Honour
Joined
17 Feb 2003
Posts
29,640
Location
Chelmsford
I have a PC that has been brought to me that appears to be having a few problems but before I start, here is the starting Spec:

Q6600,
AC 7
BioStar t power 45 i45 (latest BIOS)
2x 2gb Corsair PC2-6400
PNY 8800GT 256mb
Hitachi 500 GB HD.
Jeantech 450 PSU
Vista x64 Prem
Antec 300 3x12cm case fans


The system has been apparently dropping to desktop when playing games, although when I have tested this I don’t get any problems. However, I do get an error when running Futuremark 3Dmark06:

Idirect3DVertexBuffer9::Lock failed: n/a (Unknown)

Interesting because windows says that the driver has “Stopped working but has restarted”

This causes the application to freeze and goes back to desktop. I’m assuming this is the same problems the chap is experiencing during gaming.

Anyway, he also gave me a spanking new Powercolor HD4870 to upgrade to.

So I changed the PSU to A Seasonic 600w, removed the NIVIDIA drivers, ran driver sweeper, installed the new card fired up and then after about 20-30 seconds, the system just restarts and the monitor looses connection and after which just a blank screen. This happens even when I’m in the BIOS, so it can’t be a driver issue.

I tested the PSU and the rails look fine.. I then reinstalled the 8800GT with the PSU and everything was fine.. except I still get that “IDirect..” error when running 3dmark.

I’m at a loss where to go from here?


I have a dilemma here because I’m beginning to feel it’s a faulty motherboard but at the same time, I’m wondering if I have two faulty graphics cards. Has the chap been unlucky and been sent a duff HD4870 which is clouding the issue?

What is causing the “Idirect3DVertexBuffer9::Lock failed: n/a (Unknown)” error when the 8800GT is installed? I’ve searched Google and from what I’ve read, I’ve done everything that has been suggested, i.e. standard card/driver removal.

Other things i have tried is:

Memtest86.
Drive test
CPU test in UBCD
Temperatures are all fine.

Any help would be great.

Thanks
 
Have you got a spare HD to temporarily whack in and install a fresh OS on to run some games/benchmarks? As that would rule out any software/driver issue hopefully.
 
IT could be the PCI-e slot (motherboard) - could you chuck in your own gfx card just to rule out the '2 dodgy gfx' card theory?

Have you tried stripping it down to a skeleton rig - 1 stick of memory, no hdd, cd etc and then tried to see if you can get into the BIOS using the Powercolor HD4870? (meybe even set it up outside the case)

The other thing deff worth trying is resetting the CMOS/taking out the battery.
 
Ouch. Probabilities point at the motherboard being on its way out rather than two dodgy graphics cards, but symptoms point at graphics cards.

That the motherboard will run one well enough to deal with windows and so forth, but can't even run the other for a minute looking at the bios points at both cards being faulty, the new one more so than the old.

Of course there is a risk that one of the components has developed an unfortunate fault which attacks other components. The motherboard may be killing graphics cards, making me reluctant to suggest putting another one into it. That would be properly unlucky, but perhaps no more so than having two faulty graphics cards would be.

You're basically going to have to stick one or both of his graphics cards into a working system I fear. The alternative of populating his motherboard with known working processor, ram, graphics etc is more effort.

Curious to see how it works out, good luck!
 
Of course there is a risk that one of the components has developed an unfortunate fault which attacks other components. The motherboard may be killing graphics cards, making me reluctant to suggest putting another one into it. That would be properly unlucky, but perhaps no more so than having two faulty graphics cards would be.

JonJ, makes a good point - i've witnessed threads where a gfx slot has killed gfx card after gfx card.

I bet that's put your mind at ease... ;)

Good luck!
 
The other thing deff worth trying is resetting the CMOS/taking out the battery.


Have tried this along with a BIOS flash.

Dam skim reading :o

As Plec says try your own GPU in the system, and in each of the PCI-e slots.

Will try this.



Yes this is my fear that I have two faulty cards. Unfortunately, the card in my rig has a water block and all my test cards require no aux power connections. They are just basic video cards. They appear to work ok but i don't feel this is an adequate test.

I have contacted a fellow PC technician to see if he can test the cards.
 
JonJ, makes a good point - i've witnessed threads where a gfx slot has killed gfx card after gfx card.

I bet that's put your mind at ease... ;)

Good luck!

:eek:

the only contrary to this, is that the 8800GT card doesn't have the same problem.


Edit - I have just found a PC I have floating around.. Will a 400w PSU provide enough juice for the HD4870? Other bits incl E6320, Asus P5N-E, one optical and one 250gb HDD..
 
Last edited:
:eek:

the only contrary to this, is that the 8800GT card doesn't have the same problem.

The mother board beguiled it in a totally different way. ;) (it is unlikely.)

Good idea getting the cards tested - if they come back ok you can then start giving the motherboard long hard stares and see if it blinks.

Have you tried a skeleton set up outside the case just to rule other components out?
 
Edit - I have just found a PC I have floating around.. Will a 400w PSU provide enough juice for the HD4870? Other bits incl E6320, Asus P5N-E, one optical and one 250gb HDD..

If it's a good one brand it should, if not it might be fine for 2D :p
 
Only one way to find out - but if it's a generic supply it will probably just groan in disdain when you switch it on.

As, Happy, said (gotta love your name Happy - makes me smile just typing it) if it's a decent brand it may be up for the challenge.
 
No problem, glad to hear you got it resolved.

You seemed to have had your suspicions in your first post - it was just proving it that was the problem. Motherboard problems are always annoyingly difficult to prove even when your 99% sure that it's the route of all your problems.

It's odd how many times we usually have a gut instinct, for the problems we get, yet we still have to trouble shoot to the n'th degree just to be certain. It's largely thanks to those random and inexplicable 1% worth of problems that always seems to involve: faulty cables or a wierd component arrangement which suddenly decides, on a whim, it can't get along anymore. God, i hate that 1%!

Still, at least your friend should stand you a few drinks for your efforts...
 
Back
Top Bottom