Computer constantly crashes while gaming.

[E3Art]

I have a vista 64 x2 dual core 6400+
4g memory
ATI 2600xt video card
ECS Z780gm-a motherboard

All my problems started when i upgraded to the new hardware listed above and vista 64

I have all the most recent drivers for everything.


I have continual crashes when i play games. I play world of warcraft and my daughter plays spore. We cant even play either game for 15 minutes without it crashing. Most of the crashes just freeze up the game we are playing and we can get back to desktop to close the game and restart. Occasionally the crash will produce the Blue screen of death and restart the system.

I have worked with Blizzard tech support for a while and they seemed to think it was some sort of resource conflict with the sound or video card, but even disabling the onboard sound and buying a new sound card didnt fix the problem.

I saw on this forum that someone suggested getting Orthos stress test and see what happens. When i run it on Any of the tests it will fail after about 6-30 seconds of testing. Orthos just says to consult the stress txt file, but i cant seem to find one.

Can someone please tell me whats going on here....

Here is the latest round of Orthos test results.

Type: Large, in-place FFTs - stress some RAM Min: 128 Max: 1024 InPlace: Yes Mem: 8 Time: 15
CPU: 3200MHz FSB: 200MHz [200MHz x 16.0 est.]
8/9/2009 8:29 AM
Launching 2 threads...
Using CPU #0
Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 0 minutes 51 seconds - 1 errors, 0 warnings.
Execution halted.

Type: Blend - stress CPU and RAM Min: 8 Max: 4096 InPlace: No Mem: 3838 Time: 15
CPU: 3200MHz FSB: 200MHz [200MHz x 16.0 est.]
8/9/2009 8:31 AM
Launching 2 threads...
Using CPU #0
Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 0 minutes 15 seconds - 1 errors, 0 warnings.
Execution halted.


please help, I'm at wits end here

 

[NumericalMethods]

Hi,

My initial reaction would be to check a very common error. The bios of your motherboard has not selected the correct voltage to apply to your RAM. Random lock ups and bsod's is exactly what I would expect.

Check what voltage your RAM should run at (and the speed it should run) and enter the bios. Then manually set the voltages rather than leaving the board settings at Auto. I'm not saying that this is definately your problem but on a newly built system with those symptoms that is the first thing I would check. HTH

 

[E3Art]

i used to be up on all that stuff, but i confess im not anymore. How would i determine those things. I dont want to damage my system.

 

[E3Art]

Here is the memory i have

PQI POWER Series 4GB (2 x 2GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit Desktop Memory Model MAD44GUOE-X2 - Retail

 

[NumericalMethods]

Ok,

Easiest way is to look at the label on the actual sticks of RAM inside your pc. It will normally tell you what speed and voltage the sticks run at. If you aren't confident in doing that you could google your make and model of RAM and see what comes up from the manufacturers website. Of course, if you don't know what RAM you have you will have to open the case and take a look anyway.

As far as the bios is concerned, hit the delete key (normally) on start up of your computer (there is normally a prompt on the screen to do this) to enter the bios settings screen. I'm not familiar with your motherboard at all so I can't guide you specifically on what settings to go to. You should look for a section that covers the voltages your pc runs at. There should be a setting that looks like DRAM Voltage or RAM voltage of vDRAM.

See how far you get and post back when your unsure of what to do next. Good luck. Cheers

EDIT: Ok I see you know your brand of RAM. Cool

 

[ectx]

Did you upgrade one of the parts or the os? Which one? Or is it that you have new mb/cpu?

The errors you have for Large / blend tests suggest you have hardware problems.

I'd try small fft test to see to see if it is cpu-related. if you still error out quickly for small fft, most likely you need to nudge up the voltage a bit....

 

[E3Art]

Here is the test for small fft, as you can see, error after only 5 seconds.

Type: Small FFTs - stress CPU Min: 8 Max: 64 InPlace: Yes Mem: 8 Time: 15
CPU: 3200MHz FSB: 200MHz [200MHz x 16.0 est.]
8/9/2009 10:25 AM
Launching 2 threads...
Using CPU #0
Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 800000 Lucas-Lehmer iterations of M172031 using 8K FFT length.
FATAL ERROR: Rounding was 0.4921875, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 0 minutes 5 seconds - 1 errors, 0 warnings.
Execution halted.


I am unsure of how to change my voltage. There is an option in my bios to change voltage for DIMM. I am assuming that, that is the correct option? it is set at 12 right now and the instructions say that 1= +10 mV, 2= +20mV....ect
I dont know what to change it to. I changed it to 13 and then 15, but cpuz still says my voltage for memory is 1.8v under the timing table.

 

[E3Art]

Could my cpu be the issue here?

cpuz reports my voltage at 1.31v occasionally dipping down to 1.29v i think the specs on my processor say the voltage is 1.3/1.4

 

[ectx]

I have not used ECS borards before, so I am not sure where you can adjust the cpu voltage. But it looks like you are running anything at spces. So I'd worry if the cooling is adequate. What is your temperature? What coolling do you have.

Which parts / software did you upgrade when the errors start to occur?

 

[E3Art]

I dont have anything to moniter the temp, but every time i check it in the bios, its around 115F. It is set up to give me an alarm at 130F or so i think.

I upgraded every major component in the system. everything is pretty new

 

[ectx]

OK.. It is safe to say you have a system stability issue. (I was wondering why this would suddenly occur if you only made a minor upgrade ...)


115f does not seem to bad (I'd say it is on the high side, but I have not used Athlon x2 for a while now so I really can't say for sure - the reasing in bios is not the best way to tell, anyway).

I beleive core temp works for you cpu ... download 64bit version here Let's see what your temp is when you are in Windows. See how high it shot up after you strarted the small fft test.



You just have to debug this step by step. But small fft error suggests you must make the cpu stable first.

 

[terentenet]

I'd say check your temperatures and if you're in check, push the voltages a bit higher. First for the CPU, just a notch and see what happens. If it errors out very quick, push one more notch. Make sure to check the temperatures. If it takes longer to error out, push as far as temperatures allow you to maintain stability.
Do the same for memory, but do not exceed the factory recommended values by much.
If you can, run a simple memory test. Memtest 86 will do.

 

[E3Art]

ok...this is actually quite encouraging..

I pushed up the cpu voltage a bit. it is up to about 1.376-1.392 now. When i run orthos it dips back down to about 1.32, but so far it has stayed at 1.32 or above. And so far Orthos has NOT crashed.

I am running into a bit of an issue though. The processor is heating up rather quickly (at least it seems to me). When running Orthos the temp is approaching 90C. i shut the test down before it gets there though. What is a safe temp? should i back off of the voltage a bit?

 

[ectx]

the temp should not be in 90c for this processor. I'd reseat the hsf....

this is you cpu, right? I would not mind going a little above the specs but not 90c

Are you using the retail hsf? what kind of TIM did you use?

 

[E3Art]

I dont remember if im using the retail HSF or not. What is TIM?

I lowered the voltage a little bit and the temp is staying around 75-77c with Orthos running. Now i can get through about 5 minutes of testing with Orthos before it fails.

 

[BTRY B 529th FA BN]

I didn't read the entire list of questions asked but are you overclocking anything?

EDIT: bring your memory ratio to 1:1 and run memtest86 (whichever version) but make sure its running 1:1 If you fail memtest, you have faulty ram and need to replace them. Failing Large FFT's indicates mobo subsystem like NorthBridge voltages being too low.

Those are the two places I'd look first. RAM and mobo subsystem.

 

[ectx]

Did you put the system together recently? How can you not remember what HSF (cpu cooler, Heat Sink Fan) you have?

TIM = Thermal Interface Materail, or thermal spread, thermal grease....

using cpuid, you can see if your memory frequency is out of specs or not. (as the above post suggested)

I personally think you should take out you hsf, reapply TIM and start all over.

 

[E3Art]

Thanks for the replys everyone.

Ive been playing with the voltages and i have found a happy medium that gives me system stablilty and acceptable temperatures. And the best part is, my games arnt crashing anymore. So thanks to the people that steered me in the direction of cpu voltages, that seemed to be the key.

you guys rock.

 

[ectx]

glad to hear it. I'd keep an eye on the CPU temperature if I were you (for at least a few days).

 

[terentenet]

E3, you should really look for a better HSF. A cooler processor is more stable as well. Not to mention that using that, you may go up to 1.4v and perhaps reach perfect stability @ stock frequency. I don't know what's good for AMD these days, but I think a Thermaltake True will keep your temps in check.
What's odd is that this CPU/RAM/MOBO setup cannot maintain stability at stock speeds and Auto settings. I have never checked one of my computers for stability at stock speeds because I always thought they are stable.
E3, check ECS site for a BIOS update. A newer BIOS might improve stability for you.