Does this sound like a defective motherboard?

Neon01

[H]ard|Gawd
Joined
Jan 22, 2008
Messages
1,048
(build in sig) I've been having problems with my PC for about 6 months now (ever since I built it, really), but they were never severe until recently. Here are the symptoms:

1. When I turn the PC on - especially in cold temperatures, like winter - it gives me all kinds of strange problems on boot up. Occasionally it won't even POST at first, and I usually just hit reset and it POSTs. Then when I get to the windows loading screen, it usually blue screens, giving me one of a number of different errors. "IRQL_NOT_LESS_OR_EQUAL" is the most common, but I've seen at least 2-3 others that I can't recall. I reset again, and sometimes I'll get a "Windows is loading files" white progress bar at the bottom of the screen, which inevitably puts me back in a blue screen. Eventually after several pass-throughs like this, it'll actually get to the login screen, and from there it's usually fine. It seems that it takes longest when it's really cold - for some reason if I have the PC running for quite a while, then shut it down and do a "cold boot", it boots up just fine (usually).

2. Just recently when I was switching between my monitor and my alternate display (a 47" LCD I use for "armchair gaming"), it lost all of my displays except 1, and refused to display anything but 800x600 or show anything but the default adapter. When I tried to load Nvidia control panel, it told me that it didn't recognize any Nvidia hardware as being present. I've asked my friend to test my video card for me and I'm using an older 8800 GT in the meantime (which it did recognize), but the other problems seem to be the same.

3. It will spontaneously blue screen me and restart on its own with varying frequency. Lately it's been happening a LOT (about every night - I leave my PC on all the time, usually). This is the one that really prompted me to start troubleshooting.

Background info: since I've had these problems I've replaced the hard drive, the power supply, the video card (see above), and even the motherboard (more on this in a minute). I've also run memtest extensively on the RAM and had no issues, and run extensive stress tests (with Prime 95 and others) on my CPU with no issues.

I had similar problems with the first mobo I bought, which was an Asus, but it was the P6T, not the P6T Deluxe V2 I replaced it with. The problems were lessened with the new mobo, so I sort of just ignored it since it really only made for a headache when booting.

I'm starting to suspect the mobo is really the culprit here, but I'd appreciate your advice, as I don't have the experience of much of this community.

Thanks
 
#1. Do you have the latest BIOS for your motherboard?

ftp://ftp.asus.com.tw/pub/ASUS/mb/LGA1366/P6T_Deluxe_V2/

#2. Are you using memtest86+(very unreliable) or HCI memtest?

#3. Is the outlet you are plugging the computer in to properly grounded and is the polarity correct? You can get a simple tester from walmart, etc. They have three lights on them and can tell about the ground and polarity.

http://hcidesign.com/memtest/

The IRQL_NOT_LESS_OR_EQUAL error is usually related to RAM (voltage or timings not being correct or faulty RAM) or the CPU (usually due to not enough voltage or overheating).
 
#1. Do you have the latest BIOS for your motherboard?

ftp://ftp.asus.com.tw/pub/ASUS/mb/LGA1366/P6T_Deluxe_V2/

Looks like I don't have the very latest (Nov '10) update. Under the description, this says the update is primarily for adding additional CPU support, so I'm not sure if that would help, but I will update this when I get home from work.

#2. Are you using memtest86+(very unreliable) or HCI memtest?

The former. I was unaware that it was unreliable. I'll try the other too.

#3. Is the outlet you are plugging the computer in to properly grounded and is the polarity correct? You can get a simple tester from walmart, etc. They have three lights on them and can tell about the ground and polarity.

http://hcidesign.com/memtest/

Again, not something I've tested, but I've been running computers on this very outlet for at least 7-8 years and never had a problem. Not saying that doesn't mean that there is a problem, but it seems less likely.

The IRQL_NOT_LESS_OR_EQUAL error is usually related to RAM (voltage or timings not being correct or faulty RAM) or the CPU (usually due to not enough voltage or overheating).

The CPU runs at very modest temps most of the time (about 55 C during normal operation, and up to about 70-80 C during Prime95). The problems have occurred with or without any modification to stock values, and taken with the fact that Prime95 never revealed any issues whatsoever even after 14+ hours of each type of stress testing (small, large, blend), it leads me to believe that the CPU isn't an issue. Also, this system isn't exhibiting any of the weird behavior of my last system - that I ultimately found to have a memory defect - like dropping data in torrents, faulty CD or DVD program installs, etc. I'm certainly no expert though, so I appreciate your input. I'll try those couple things you mentioned and see where that gets me.

But you say this doesn't sound like a mobo problem?
 
The former. I was unaware that it was unreliable. I'll try the other too.

memtest86+ can take days to find a single error while the HCI memtest program will usually find errors within minutes.

But you say this doesn't sound like a mobo problem?

It could be a motherboard problem, but I would eliminate other possibilities before swapping out the motherboard again.
 
memtest86+ can take days to find a single error while the HCI memtest program will usually find errors within minutes.

I second that. Technically memtest86+ is better but since it does not stress the ram the same way as running an os while testing ram at the same time in multiple threads memtest86+ can take many passes to find a single memory error.
 
I installed the latest firmware last night - no change yet, but I haven't gotten much chance to really test it. I've also set all BIOS settings back to defaults (except boot order and other minor items).

I downloaded HCI memtest and tried to run it, but it told me that since Win7 keeps programs resident in RAM even if they aren't running, I should run multiple instances of it, each running on 2047 MB, but that if the additional copies wouldn't load, to keep running smaller and smaller applications of it. So what if I don't cover all the RAM? It seems possible that it might miss an error or something. Any tips on how, exactly, to run HCI memtest?

I haven't checked the power polarity and ground yet.
 
Run multiple copies of it so that it covers almost all of your ram. Again this method is still better at detecting problems quickly than memtest86+. And my opinion comes from using both in a work environment not just my 3 home computers.
 
Ran three open copies HCI memtest of it all day with 0 errors.
 
If it is still giving you problems... try one last thing before deciding it is the motherboard.

Take the CPU out of the socket and clean the contacts on the CPU with a pencil eraser. Then clean the eraser residue off with some rubbing alchohol and a cottom swap or the like.

Reinstall the CPU and test again.

If it is still giving you the same issues I would say it is most likely the motherboard.
 
After a lot of searching on the net and more troubleshooting, I've come to realize that the problem was a defective stick of RAM. Tried booting up with one stick at a time. Two of the sticks gave me no trouble, the other made the problems 100x worse (wouldn't boot all the way to windows no matter how many times I tried). I then tried the other two again to verify and the problem disappeared. Put the bad stick bad in (by itself) and the problem materialized again. So I'm working through the RMA process with OCZ to get replacement RAM.

Weird thing is that the other RAM is now giving me different issues. Luckily (I guess) these issues seem to be shared by the community. After a lot of searching on the OCZ and Asus support forums (and even the 700 user reviews for the Asus mobo on newegg), it seems the RAM I was using doesn't like to report good settings to the mobo, resulting in unstable default settings. I've tried using some of the settings recommended by OCZ tech support for others with the same mobo, and I get lockups after awhile. They happen within 5 mins when I run Prime95 in Blend mode. The good part is that the lockups appear like they are attributable to overclocking (I'm actually only running the RAM at 1600), since they don't give me a blue screen, so I'm pretty sure it's not the same problem I was getting before.

What I might do is just swap the RAM for some Corsair dominator stuff and sell the RMAed RAM I receive back from OCZ.

Thanks for all the help guys, I really appreciate it!
 
Back
Top