New system, cannot run stable stock

Sunkist

Limp Gawd
Joined
Jun 10, 2005
Messages
232
I just built a new system:
3200+ Venice
MSI K8N Neo2 Platinum
1GB Mushkin DDR 3200
6800GT
Audigy2
Antec SmartBlue 350 & SmartPower 400 (switch to see if that was the problem)

I just built it this week and I get BSOD randomly it seems with master reset error (or something like that) and playing games was basically like looking at a strobe light everything moved in slow spurts. I thought it was the power supply so I got the 400w and I reseated the processor which did seem to help with games but now they just crash after a while. Corecenter says the processor is at 49C when I was playing Half Life 2 when it crashed. Anybody have a clue on the problem cause I am stumped.
 
not enough power on the 12v rail.. that 6800gt is a power whore :p
you'll need a psu with at least 20a on the 12v rail
 
So do I need an even higher powersupply? the Smartpower 400 does 23A on the 12v. Ugh these things really add up.
 
oh, the smartpower should be enough then.. hmmm

in the bios, how much voltage is the ram getting, according to the sensor? also, which slots have you placed the ram in? (i'm assuming it's 2x512mb)

and i'm going for hardforum junkie, thus i'm a post whore right now :D
 
Memory is 2.7v and they are in the farthest two slots, they were next to the proc but I moved them just to see what could make a difference.
 
Damn thing just crashed on me again! Says MACHINE_CHECK_EXCEPTION on the BSOD. XP says either the proc, mainboard, memory, or powersupply caused the error due to overheating but I do not see any crazy temperatures happening although my proc has been at a consistent 45C right now which seems higher than I would expect a venice to be.
 
ok, get memtest, and run it. also, you should have the ram in the two closests slots to the cpu for dual channel on the neo2
 
If memtest fails, run it on each DIMM individually installed to see if you can determine which one is faulty.
 
Just had another thought....

Check your CPU, northbridge, and HTL voltage settings in BIOS (if applicable). Make sure they are all set to stock. If they are set to automatic in BIOS, try changing them manually to the stock settings.
 
Which memtest do you speak of? I just disabled all of the aggresive timing, and overclock features that the motherboard enables and all the voltage settigns appear to be stock so lets see how it goes.
 
What bios are u running? Also try to give the memory 2.8v. When I first got my Winnie, I had to swap the processor out, put a .130m processor in, flash the bios, and all was fixed. Although if that was your problem, you probably wouldn't get it to post.

I think the older bios have issue's with the Rev.E chips.
 
It was v1.5 when I got it started but I flashed to 1.8 now. Memtest ran fine, although I didnt fun it long cause the system still crashes. It appears that something is over heating cause it can run for quite a while when I first start it up but once it crashes it happens pretty consistently after that but I dont know what is getting hot. My motherboard says my venice is only at 22C which is pretty damn cool especially on the stock heatsink so maybe my motherboard has errors.
 
Stay grounded and touch things on the MB to see what's hot. MOSFETS, Shipset heatsink, etc. Also unplug every component not needed to get the computer to boot.
 
Sunkist said:
Memtest ran fine, although I didnt fun it long cause the system still crashes
did you do at least 3-4 full passes?
and 22c is abnormal :p

clown has a good idea. also, every time i look at this, power supply alarms scream in my mind :(
 
Yea the 22c is real strange, when I first installed everytthing it showed 38c then later after the 1.8 re-flash it showed 44c and now all of the sudden it shows 22c so I think the motherboard maybe the issue. I am probably just gonna RMA the parts back to New Egg and get replacements, I ordered the MB, CPU and memory from them but I dont know if I wanna send them all back. I will run memtest a few more times and longer, if I can, the computer seems to be crashing much more frequently.
 
Does memtest crash on you? You said you didn't run it long, you probably want to wait until it gets through one whole a pass or two.

IF memtest passes, I would format you HD and reinstall windows from scratch. When I'm having problems with new builds, I want a fresh clean install so I can rule out OS problems. Install NVidia Unified Drivers(I used the 5.10's) without the Nvidia IDE SW driver, install your video card drivers, and any driver that you need.

After the fresh install, try running prime95 to see if it is stable.

If it's not, pull the MB out and put it on an Anti-static surface and try running only necessary components and see if it is stable.

If it is stable, it probably is either a PSU problem or heat problem.

You said reseating you processor helped a little bit. Hmm. After you pull you MB out, you can pull your processor out, and make sure there is no bent pins. Also make sure there is no thermal goop on the socket. Reseat the CPU, place some AS's 5 or similar on the IHS, and remount the HS. (what HS are you using btw)

It's not an exact science.
 
Memtest does not crash and it ran fully but I only did about 1 1/2 passes. I plan to reinstall windows one more time just to check before I start sending parts back. No pins are bent and I had applied a large amount of compound before but I cleaned it up and used less when I pulled it out. I have a retail box CPU so stock heatsink fan with AS1 (classic ;) ) on there. How long does Prime run? my computer seems to only last about 20 mins now before it dies, or atleast it did this morning.
 
Prime should run as long as you want it to run, a good measure of stability is to allow it to run for 24 hours.

Also as you run prime(use In-place large FFT's since MEMtest passed) keep an eye on core center. If you are reaching 49c while you are playing a game, prime will probably give you considerable higher temps.

One other thing, go into the bios and load setup defaults.
 
Re-installed windows, videocard drivers, Nforce 3 drivers and only have one hd plugged in and its been running fine for about an hour, just downloaded Prime 95 and getting 3dmark to see if I can run graphics.
 
okay little update here, ran 3dmark 03 ran fine, ran Prime for 38 mins (I know to short) ran fine, temp got to 35c (still oddly low). I still have not installed Audigy drivers, plugged in two other HDs, installed Bluetooth logitech keyboard/mouse drivers, so one of these things may have caused the error. Oh yea I wanna thank everyone for all the help cause I have been pretty stumped, this has been the most chaotic computer build I have done and I have been doing this since the 486DX (mine was the AMD knock off :D )days. Gonna run more stuff tomorrow after work.
 
The K8N Neo 2 is known to report buggy temperatures, so you're not alone on that one.

As for the crashing, make sure your RAM timings aren't too aggressive. I was experiencing blue screens during gaming and 3d intensive stuff, and I found out the RCD timing was causing it. Memtest would pass every test I threw at it, but as soon as I got into Windows, programs would randomly crash and 3d programs would produce blue screens.
 
adurn03 said:
The K8N Neo 2 is known to report buggy temperatures, so you're not alone on that one.

As for the crashing, make sure your RAM timings aren't too aggressive. I was experiencing blue screens during gaming and 3d intensive stuff, and I found out the RCD timing was causing it. Memtest would pass every test I threw at it, but as soon as I got into Windows, programs would randomly crash and 3d programs would produce blue screens.

Yea I thought the v1.8 Bios were supposed to cure the faulty temp readings :rolleyes: I havent messed with they timing cause they did not appear to be aggressive when I looked so I just messed with the voltage before. Running Prime95 all day while I am at work to see if any errors arise. One new thing I noticed is last night I installed Corecenter and it seemed to cause my NB fan to slow down. In BIOS it runs at 5000+ rpm but after I installed Core center I noticed a fan got quieter and I looked to see my NB at only 2800+ rpm. Everytime windows starts up I hear a fan slow down and I noticed this noise before when I first had problems as well. It still ran fine last night but that was weird.
 
Ran Prime for 9 hrs with zero errors, gonna install the soundcard drivers soon and see it that causes any problems.
 
Sounds like you mostly ruled out your CPU/MB. I would probably say it was the windows install. Does you Windows XP come with service pack2? If not, I would slip stream it in, instead of running the service pack installer.
 
Yea something must have FUBARed on the previous install but it appears to run fine now. I do need to slipstream on SP2 which I was gonna do before I built this sucker but I didn't get around to it. Ran Prime another time and played some Battlefront 2 so its running smooth for now. Time to overclock!!!! Well maybe not just yet.....
 
Back
Top