6800GT strange hwmonitor readings and card instability

[Wolf2000me]

Hey,

With any driver my rivatuners monitoring program gives strange readings on the voltage and current monitoring lines.

Take a look at at this picture. It's an example of the type of readings I get and I've seen worse ones like this.
http://users.telenet.be/denthomas/AOA/monitoring.JPG

It is driver and clockspeed unrelated.

I also have stability problems which may be unrelated to these monitoring readings. Usually monitoring software is rather unreliable but I don't trust this for one bit.
My stability issues always happen in the same 3d games in exactly the same way with exactly the same error message. It also happens with any driver and at any clockspeed. A description:

In Trackmania nations I have a chance to get a hardware lockup during the server listing screen. It may happen within 2secs to 5 minutes. Usually if I'm fast enough at choosing a server I can play as long as I want with temperatures not exceeding 75°C.

In the Heroes 5 demo I can get the same kind of lockup during normal play. It never happens during combat however. The lockup happens in the same timeframe as with the above stated lockup.

I have had other games doing this and sometimes it stopped doing that. It almost appears as if the card can't cope with a specific graphic feature used in those games locking my system up. I'm totally stable otherwise and in most other games or benches.

The lockups always give me this error message:

The driver nv4_disp for the display device \Device\Video0 got stuck in an infinite loop. This usually indicates a problem with the device itself or with the device driver programming the hardware incorrectly. Please check with your hardware device vendor for any driver updates.

The errors source is "nv" and it is event nr. 108.
I have tried a dozen drivers like official, modded or WHQL ones. I'm pretty much out of options. I bought this card 2nd hand from some bloke a the nearest town. He has been promising me the invoice of his purchase of the card for over a year and I have not heard from him in several months. It may be time for a surprise visit.

[dsio]

I would suggest that your problem and the monitoring error are related in as much as when the driver crashes, the monitoring capability, which is also provided by the driver, also goes crazy.

Your temps look fine to me.

What power supply are you using, and also, is the +12V connection from the power supply to the video card dedicated?

By dedicated I mean, is there a single +12V wire going from the PSU to the GPU, with no forks connected to anything, no spilts connected to anything.

That odd behavior could be caused by something like a hard disk spinning up during play, so I would check carefully over that connection to ensure the GPU is the only thing on that lead from the PSU.

If your PSU is on the cheaper and nastier side, a 6800GT, and more importantly an overclocked 6800GT will have some issues.

[Wolf2000me]

Thanks for the answer. I am currently using an Antec 500W PSU.

I don't have the option for a dedicated molex due to the many disk devices I'm using. However I doubt that causes my problem. I don't have any problems using games which require much more power and have much more disk activity. Like fear, Oblivion, etc. I have never encountered this problem in heavy graphic card games like those.

[dsio]

Disk activity has nothing to do with it. If there is a hard disk on the same line as your 6800GT, its extremely likely that its the problem, and in any case its damaging to your card.

Spin up, spin down, and hard disk speed changes are what bug the card. Nothing to do with how much the disk is being used at all.

[Wolf2000me]

Right, I understand but that raises another question. Shouldn't I be in trouble in every game then? Like Trackmania nations, there it will only crash during the server listing screen. If I'm fast enough at choosing a server I can play the game for as long as I want.

I will however give it a shot when I have time. That will require a total restructuring

[dsio]

Different games mean different disk usage patterns, which means different current draw patterns. If a game, or an action within a game allows the disk to spin down, and then spin up, that could be a major problem. If it starts madly reading from the hard disk, that may be a problem as well.

So, two questions.

1: Does your hard disk have a spin down / power off setting in your APM settings?

2: Do you actually have a hard disk running off that rail?

3: Do you have much warranty left on your card? Because you may find yourself using it.

[Wolf2000me]

Makes sense.

1. No, not in the OS and not in the bios
2. Yes
3. That's a problem. I bought the card second hand from some bloke on a forum. He doesn't live very far, I can get there in 15 minutes but the dude has been promising me to mail me the invoice for 7 months now. He said it was somewhere else. I have not been able to contact him for a few months so it may be time for a surprise visit.
I have yet to try to change the power distribution in my system. I seem to be lacking spare time atm.

Thanks for the input dsio, it's greatly appreciated

[Wolf2000me]

I gave the 6800 a dedicated molex by disconnecting all but the OS drive. No help, the crash still happened exactly the same way.