Another newbie needing help

[MattS2K]

Hello,

I've been trying to build my first PC and having put it all together, I pushed the power switch only to have my hopes dashed as the fans started for a fraction of a second then everything stopped. There was no beep and apparently no life after the initial burst.

I've packed it all away as I was getting very frustrated and didn't want to make anything worse so I found this forum and started browsing to see if others have had the same problem. As I've packed it all away and its getting late in the UK I haven't tried the 'build out of the case' procedure yet, but I will do so tomorrow.

However, I have noticed people commenting on the way the motherboard has been screwed into the case causing shorts and I wonder if that might be part of my problem. My case has metal standoffs and earlier I spotted that I missed one of the screw points - could this missing screw be causing a problem? Also, should I be using the washers or not? I'm sure I haven't got any extra standoffs under the board, but I have forgotten to screw one of them down. Any comments on this or any other suggestions would be gratefully received.

The stuff:
MSI K9N Platinum nForce 570 Sli PCI-Express DDR2 motherboard
Athlon 64 X2 4600+
Corsair 2GB (2x1GB) DDR2
2xBFG GeForce 7900 GT 256MB
Hiper TypeR 580W PSU
Western Digital 250GB SATA-II
LG DVD Rewriter
Jeantech Butterfly V3 case

Cheers,

Matt.

[Cricket]

A missing standoff won't cause what you're seeing.

Not screwing down the motherboard to one stand off won't cause what you're seeing either.

You really don't need the washer. I never use them.

But a standoff in the wrong location will.

Do the out of case troubleshooter.

:) Cricket

[glc]

http://forum.pcmech.com/showthread.php?t=132409

[MattS2K]

Firstly, thanks for the help so far. I have now managed to get her working, but I now have another problem.

It turned out I'm just more stupid than I thought. When I was taking it apart to rebuild out of the case, I realised that I had used a wrong power connector... My motherboard has three main power connections, the main 20/24 pin connector (PWR1), a straight 4 pin connector (PWR2) and a square 4 pin connector (PWR3). The motherboard guide has a typo meaning that it describe PWR3 twice, once as providing power to the CPU and once as providing power for the graphics card. Out of interest, does anyone know which ones which?

My power supply has a 20pin main power connector and a separate 4 pin section making up the 24-pin supply. However, I had used the 4 pin bit in the square PWR3 and it was only when I was taking the stuff apart to rebuild it out of the case that I noticed it was slightly the wrong shape. I moved that 4 pin connector into PWR1 with the other 20 pin bit and put a new power cable that was exactly the right connection into PWR3 and tried again. It now powered up and went to the bios screen.

I then installed Windows (XP professional) and that went fine.

Now, I have started having intermittent resets. It doesn't freeze and I don't get a blue screen crash or anything, its exactly as though I had pushed the reset button. The PC reboots and once I log into Windows again I get a 'Notify Microsoft' message as the system had a 'Serious Crash'. I've checked the CPU temp in the BIOS after this has happened and it seems to be running about 38-40 C which I think is fine.

Anyone got any ideas what might be causing these random resets or what I can do to fix them?

Thanks again,

Matt

[Cricket]

Random reboots can be caused by a faulty or undersized (overworked) power supply, a faulty video card, faulty RAM or a bad motherboard. They can also be caused by software conflicts. Normally when I get random reboots I swap out the power supply with a known good one and work from there.

Are you certain you have all the correct power connections from the power supply plugged into the motherboard now?

:) Cricket

[Chris_Rootes_Wizard]

My aunts Pc does that sometimes (had it for ages, runs along time, then reboots several times after some months of good running) Under no load. Any suggestions to that then?

[Alaron]

Chris, you're hijacking someone else's thread.

[Chris_Rootes_Wizard]

Yes, but because this Pc has been working for ages, the power cables must be in place, so what would be the solution then? Sorry, maybe the same senario with MattS2K?

REPLAY TO ONE BELOW:::Okay, sorry. I'll leave it anyways, it was just as you guys were on the topic. Plus im tired and want to go to bed, lol.

[Cricket]

Chris_Rootes_Wizard: Start a new thread for that problem. This thread belongs to MattS2K.

:) Cricket

[MattS2K]

Random reboots can be caused by a faulty or undersized (overworked) power supply, a faulty video card, faulty RAM or a bad motherboard. They can also be caused by software conflicts. Normally when I get random reboots I swap out the power supply with a known good one and work from there.

Are you certain you have all the correct power connections from the power supply plugged into the motherboard now?

:) Cricket

Thanks again for your help. I *think* all the power connections are correct, but I'll double check them this evening...

Unless the PSU is faulty, it should be good I think (580w).

A couple of other observations:

This only started happening after I installed Windows. This is probably chance but could it be Windows causing the issue? The installation was SP1 and although I'm applying updates via automatic updates, SP2 has not yet been installed.

The hard drive is installed in SATA1, meaning that there is only one IDE device - the DVD-ROM. I have installed that as the master - is this right? I thought so, but there is a screenshot in my motherboard guide showing a CD-ROM installed as a slave when there is no master, so have I got that wrong as well?! Could it be linked?

Assuming the above are not causing my problems, I guess the only way to try to find the problem would be to start removing bits until it stops resetting. With this in mind, one final question:

I have 2 sticks of DDR ram installed (they are correctly in slots 1 and 3 as per the motherboard guide). Can I use just one of these at a time? and should I always install the test stick in slot 1?

Cheers,

Matt

EDITED TO ADD:
Just found the nVidia monitoring utility and its warning of some problems...

The problems its reporting are:
CPU voltage = 1.35v
Memory voltage = 1.8v

Its tooltip says that values shown with a red block (these are) indicates that the 'component is being overstressed and could suffer damage'

What values or ranges should these be?
How likely is damage - these have probably been running on and off for 5-10 hours, presumably at these dodgy settings?
What possible causes are there for this...
Dodgy PSU?
Me connecting stuff up wrong - I'm sure I've done it right, but what errors might I have made (assuming I'm an idiot)?
Any others?

My friend has exactly the same PSU and has been running it for months with no problems. If I can borrow it, would this count as a known good psu?

Sorry for all the questions and the long posts. I'm hoping that if I give as much detail as possible it will be easier for the experts to help diagnose the problem. I really appreciate everyones assistance so far.

[glc]

Even though that PSU is SLI-certified, pull one of your video cards and see if that stabilizes it.

That's AM2, right? It's VERY fussy about ram, is that ram on MSI's tested list?

Control panel, system, advanced, startup and recovery - uncheck restart on error and see what the bluescreen says next time it happens.

[MattS2K]

OK. I've done more investigations into the cpu and memory voltage because that really worried me. The bios seemed to confirm the values I am getting:
CPU voltage = 1.35v
Memory voltage = 1.8v
are ok. I also tried to understand the AMD datasheet and it looks as though 1.35v is correct. Can anyone confirm what these voltages should be?

An update so far...

I changed the PSU for one that has been working for my mate for months. No improvement.

I removed one of the video cards and it seemed better (previously it was refusing to run Ghost recon, crashing to desktop before completing the opening sequence). Now it ran the game fine for an hour or so...

Then, I installed 19 more Windows Updates (SP2 has now been installed) and it blue screened on rebooting to Windows (I had unchecked 'restart on error' as suggested above. The technical info given was a bunch of numbers (I can supply these if they'll help!)

I rebooted (successfully this time) and left it running for an hour to see what happened while I had my tea... I came back and it was still running. I then tried to run a game update patch (from the HD) and it immediately blue screened again (I think before it actually ran the update at all)

This was the same as the first blue screen except it also said
IRQL_NOT_LESS_OR_EQUAL
and the numbers in the tech info were different (again I have these noted if they are any use...)

Any ideas?

[MattS2K]

And there's more...

following the other advice above to make sure that the memory is on the MSI tested list, I checked and found it there - good.

The memory was listed in a table and stated the memory voltage as 2.1. As stated above mine is set to 1.8 in the bios....

I have adjusted it to 2.1 and can now boot and have managed to apply the game update that failed before.

I'll keep it running for a while and see how it goes.

Does this sound like a step in the right direction? And can anyone confirm the correct CPU voltage?

Cheers again,

Matt

[glc]

The CPU voltage autodetects, no worries there.

[MattS2K]

Hello again.

quick update to let you know that since I changed the memory voltage to 2.1, I've had no further crashes at all. Thanks to everyone who has chipped in to help me solve this.

I have now found one other problem (!)...

I downloaded and tried out 3DMark06 and find that if I run it with SLI enabled, the graphics tests are all screwy. It kind of looks like a badly tuned TV - you see the image is there, but its all out of sync. It runs fine with SLI disabled and runs fine if I remove the cards one at a time and run them on thier own (FWIW, card 1 scored 4702 and card 2 4699). Obviously, I would like to run it with SLI enabled and see what I get there...

Any ideas why SLI might not be working properly?