A Bad G-Card

[DaFees]

Hello everyone, (and please do excuse me if this in the wrong spot, heh)

My current computer which I had for a few years now, is really starting to act up on me. I'll explain that this previous summer my PSU died on me and I had to get a new one, and yes I learned my lesson: Never get a PSU just because the price is cheap!

Anyways all the time when I had that first crappy PSU, my computer would randomly restart and I'd get this error message:

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

Nvidia Power Indicator Error message:

Problem report:
The Nvidia system sentinel is reporting that Nvidia-powered graphics card is not receiving sufficient power.

To protect hardware from potential damage or causing a potential system lockup, the graphics processor has lowered its performance to a level that allows continued safe operations.

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

Normally, being the person that I was at the time, I just said whatever, and continued to operate my computer as normal.

Now a days it seems like whenever I play a game for a half or an hour (kind of depends on the game) my screen goes crazy, in that at first I get a screen full of multicolored diagonal lines, then my monitor flash on and off. Then it either restarts automatically or locks up on me. In addition my computer will sometimes just randomly goes crazy and lock up if I'm not doing anyhing except maybe downloading something or my screen saver is running.

I doubt this is relevant but one night my graphics card was over heating really bad, so I took it out removed the cover thingy and cleaned all the dust out that seemed to help the over heating issues, but I still get the problem mentioned above.

Now before everyone goes off saying that it is my PSU, as I mentioned earlier when my first one died I got a new one, but before getting a new one I used these very forums to get advice on a new PSU.

Now for those who find it relevant here's my system specs:

AMD 3200+ Proc
Some K8N-Ultra socket 939 gigabyte mobo ('m pretty certain it is K8N-Ultra, I'm posting this while at school so that's why I'm uncertain)
An antec 480w PSU (it has a blu led fan that I know)
An evga 6800gt graphics card (it's standard nothing overclocked)
2x512mb DDR400 OCZ platinum rev. 2 memory
SB Audigy 2 ZS sound card
1xLite-on DVD burner
1xSamsung DVD burner
1x36.7gb raptor hard drive
1x80 Wester Digitial IDE hard drive
1x320gb 7200.10 Seagate barracuda hard drive

Well there you have it. Is my graphics card bad? Or is there some other problem?

 

[kaotao]

Your PSU was/is still the problem. Why, after it telling you it wasn't getting enough power, did you continue on thinking everything was ok? Normally that PSU should have been enough, but with all the other stuff running in your set-up it wasn't. You may or may not have killed your card under powering it for so long, but you PSU is the problem.

 

[Gary_Busey]

Are you sure it's his PSU? He has an Antec 480w, which should be plenty.

 

[AdamBomb42]

Try the 6800 in another computer and see what happens.

Also, Antec PSUs are crap and I don't trust them.

 

[kaotao]

Now that I think of it, perhaps the OP forgot about the power connector.

 

[Gary_Busey]

That could be a possibility. That's a simple, easy mistake if you're new to hardware.

Like previously stated, check all your connections first, especially your power connector to your card. If those are fine, if possible, try the card in another rig. See what happens.

 

[DaFees]

Well thank you everyone for your repsonses, I don't have a readily available extra PC that I can test my card in, but when I get home I will certain check all my connectors. I can say my computer locks up and such but I don't get that system sentinel error anymore, I'ven't had that error since my first PSU like I mentioned before.

In addition to be fair I don't think my PSU could totally be the problem because as I said above I got advice on that PSU from users on this forum, and I don't think the people on here would steer me wrong.

 

[DaFees]

Well I can say this much, I'm home now and I checked all my power connectors, and everything is nice, snug and secure. My PSU has two PCI-E 6-pin cables on it and I've tried switching them, and that didn't help. I mean I was playing some Half Life 2: episode 1 (I totally cannot wait for episode 2, btw) and while my computer didn't lock up, it was not running smoothly like it should, and I could tell it was ready to lock up at any moment.

 

[sirrobin4ever]

Sure sounds like that think is overheating to me. Download ATI tool. I know, you have an Nvidia.....yes it works fine. Start it up and press the button on the left that says scan for artifacts. Watch the temps of the GPU as you do this. Does it come up with any artifacts? What are your temps when this happens? Report your findings here.

Best of Luck

 

[DaFees]

I ran the program / test like you said. I let it run for 5 minutes and it didn't find any artifacts, but my temperatures rose up to 85° (core) and 62°(ambient or idle)

normally my core temps are in the 60s and the idle is in the 40s or 50s,

Since I've been typing this my temps having going a bit crazy they've been jumping from the 60s to the 70s and back again. At one point it shot back up into the low 80s.

If I need to run the artifact test for longer I can.

 

[sirrobin4ever]

Temps don't actually sound too high....hmmm......

I would try running the ATI tool artifact test for a few hours. I'm curious as to whether the PC would lock up after a while. It sure seems like a GPU problem, but I could be wrong.

Best of Luck

 

[DaFees]

I strongly suspect it is a gpu problem as well, so when I go to bed tonight I'll let it run for a few hours like you said, and report my finds back to you when I can, but in the meantime before I start the test again. Anything you want to suggest I should change in the settings if necessary?

 

[DaFees]

Well if this helps, (I still had ATI tools open for some reason) a few moments after I posted my previous post, my computer locked up for a moment giving a screen full of multicolored diagonal lines, and then everything suddenly was reduced to extremely crappy settings (640x480 w/ with like 4 bit color) and then this error popped up (on my tv because it for some reason it disabled my dvi monitor from displaying):

"the nv4_disp driver has stopped working normally. Save your work and reboot to restore full display functionality. The next time you reboot the machine a dialog will be displayed giving you a chance to upload data about this failure to Microsoft."

Well I rebooted and I didn't get a dialog like the error message said, but I booted up speed fan and now I'm curious as to if this could be part of it. Speed fan is telling me the following:

Vcore1: 1.39v
Vcore2: 2.56v
+3.3v: 3.31v
+5v: 3.09v (this number jumps as low as 2.xx to 4.xx, as far as I can tell)
+12v: 11.78v
-12v: -4.13v (this number jumps as low as -0.xx to -9.xx, as far as I can tell)
-5v: -4.91v (this number jumps from as low as -0.xx to 5.xx, as far as I can tell)
+5v: 3.41v (this number jumps from as low as 2.xx to 4.xx, as far as I can tell)
Vbat: 4.08v

also I am not 100% sure what temp3 is but it is at 70C it hasn't changed in the last five minutes and as far as speed fan is concerned this temperature is hot!

 

[sirrobin4ever]

Hmm. The voltages sound a little wonky. Does your BIOS display voltages? If so, I'm more likely to trust the readings from it. Repost your BIOS read voltages. If these voltages are true, I'm surprised that your PC is running at all!

 

[DaFees]

Well I checked my bios, it didn't give me numerical readings, it just said for the +12v and +5v: "ok"

all I can say is that I am baffled by this, and I wish I could figure it out.

UPDATE - I tried a different driver for my graphics card, I went with the latest official whql drivers from nvidia and so far I've been playing HL2:E1 for roughly an hour, and have had no problems beyond the norm. So I'll probably jinx myself with this update, haha, but so far I don't appear to be having any issues.

hmm, maybe it was a driver issue all along?!?

 

[DaFees]

Well it happened, I jinxed myself. Last night I played HL2:E1 for an hour straight, with no problems. Then today I played the same game for about 2 hours straight, no problems. Then just a few minutes ago, I tried playing some of sid meier's railroads! and no more then 5 minutes into the game it locked up on me.

So it looks like I haven't solved my problem

 

[sirrobin4ever]

Do you have access to a voltmeter? If so, just start up the PC and take one of the molex connectors and test the voltages. The red wire is 5v and the yellow wire is 12v. The black wires are the neutrals. Start up a game. At this point, the PC should be under high strain, so the voltages will drop to their lowest levels. Test with the voltmeter again. What are the read voltages?

 

[DaFees]

I do not readily have access to voltmeter, but I could maybe sometime next week hook up with my old physics prof at my old high school, I'm sure he has one I could borrow.

EDIT - I had a friend suggest the idea of a hard drive going bad, could this be the issue?

 

[sirrobin4ever]

Hmmmm......I don't really think that the hard drive could be it.....but perhaps....

It seems like the hard drive would be the last thing to suspect. Before I would suspect the hard drive, I would check the graphics card, the power supply, the motherboard, and the ram.

I don't think its the graphics card, so now were onto the power supply. Since you can't test it for a while, we can skip it for the time being and try the ram.

Google SP2004 (stress prime 2004) and download it. Run the program. Set the test to priority 9. Run it for like 4+ hours. Let me know if it locks up or not.

When you get that voltmeter, test those voltages.

Best of Luck

 

[DaFees]

Well I ran a stress test it was in blend mode at priority level 9 for 5 hours 42 minutes 38 seconds and I ended up with 0 errors and 0 warnings.

 

[sirrobin4ever]

Well, this proves that your cpu and ram are not the problem.

This really makes me suspect the PSU. Tell me when you get those voltages.

As for other solutions, I'm running out. The only other thing that I can think of is some possible Windows error. That or high temperatures in your PC.

Do a couple things for me.
1. Clean your PC. Open it up and blow the dust out of the heatsinks. How much dust was there? When was the last time you cleaned the PC?

2. Lets test your buddy's hypothesis, just in case. Run a Windows disk error check. (Open up my computer, right click on the drive, click properties. Select the tools tab. Click error checking. Select both options and click start. If it says you need to restart, do it. Let the check run its course. Report back the results.) Any errors?

 

[sailer]

Well I rebooted and I didn't get a dialog like the error message said, but I booted up speed fan and now I'm curious as to if this could be part of it. Speed fan is telling me the following:

Vcore1: 1.39v
Vcore2: 2.56v
+3.3v: 3.31v
+5v: 3.09v (this number jumps as low as 2.xx to 4.xx, as far as I can tell)
+12v: 11.78v
-12v: -4.13v (this number jumps as low as -0.xx to -9.xx, as far as I can tell)
-5v: -4.91v (this number jumps from as low as -0.xx to 5.xx, as far as I can tell)
+5v: 3.41v (this number jumps from as low as 2.xx to 4.xx, as far as I can tell)
Vbat: 4.08v

also I am not 100% sure what temp3 is but it is at 70C it hasn't changed in the last five minutes and as far as speed fan is concerned this temperature is hot!

I'll add my 2 cents worth. I don't have a lot of trust in Speedfan. The reasons are: it shows me as having 0 voltage to Vcore2, the -12v is listed as -1.58, the -5 as -8.78, and for the number one hilarious report, it tells me that temp3 is -128c. There is no way that temp3 can be -128c.

There is a possible reason to these strange readings, and that's because my motherboard is not among those listed as supported by Speedfan 4.31. It could be that your motherboard is not among those supported, so you're getting off readings as well.

 

[sirrobin4ever]

I don't have a lot of trust in Speedfan.

I don't either. It correctly outputs my core temperature, but it lists temp1 as 17C, which is below ambient.....

 

[krogoth1989]

How long did you run it with the crappy power supply (insufficient power). Usually when I get that message I immediately shut down and make sure its not loose. The way you make it sound you had the crappy one for about a year.

 

[DaFees]

Well I doubt it is my hard drives, the tests proved I had no errors.

As for how long did I have my first crappy PSU, I had it for about a year or so until it died on me.