[Mr. Ksoft]

Well, you guys helped me a bit about a year ago choosing parts when I built this so I thought I'd come back now that I'm having trouble.

I built this machine and it's been running well for a year and a few months, but recently it has been having problems:
ASUS P6X58D-E motherboard
Intel Core i7-930
6GB DDR3-1600 RAM
1 TB, 1.5 TB, 250 GB hard drives
Radeon 5870 1GB
Linksys WMP600N wireless card
Windows 7 x64

Since around July I have experienced random freezes. Everything just stops on screen, audio cuts off, USB devices (ex. iPod) think they're disconnected, the keyboard caps lock light won't change, etc. There is no BSOD or anything in the Windows event log upon rebooting that would hint at a problem. Anyway, these freezes started out relatively rare; I think it only happened twice in July. In August they started happening more often, like once or twice a week. Recently they've gotten completely unbearable, and more random-- sometimes I'll go 15 minutes without it locking up, other times two days. It is not dependent on CPU/GPU load-- it often happens while the machine is idle, and has also happened while playing TF2 and Minecraft. I can't do anything with this machine any more if this keeps up.

So, I figured I messed something up when I built it that was coming back to bite me in the ass. Checked BIOS settings-- everything okay. Heard that bad hard drives can sometimes cause lockups so I ran Western Digital's drive diagnostics on all the drives (they're all WD drives) including full surface scans-- nothing is wrong with them. Ran memtest86 to check the RAM, no problems. However, I suspected it was the memory anyway because I had started out running the (DDR3-1600) RAM at DDR3-1066 speed and had only gone back and set it to the "proper" speed after a few months (during last winter). Sometimes after the computer froze, it would fail to boot up at all (the machine would whirr to life for a few seconds, turn off on its own)... but it would boot up okay if I pressed the ASUS specific "MemOK" button that would mess with the RAM timings until it found something that would boot. I suspected that I had damaged the RAM or something and that was making it unstable, so I RMA'd the RAM. I got new sticks back and installed them (and also set them back to DDR3-1066 again because that would stress the parts less and I couldn't tell the difference anyway), and the problems continue, so it isn't bad RAM. Hell, I can even run a stress test like Prime95 or LinX absolutely fine for hours on end... the freezes just happen completely randomly. It's not even Windows specific, so it isn't a driver problem-- happens on my third hard drive's Arch Linux install too.

I'm not sure what else can cause random freezes. Anyone have any ideas as to how I can make this thing stable again... or is it undiagnosable, and it'd be better if I just cobbled together a new computer somehow? (Although I have no idea how I would do that... I'm in college right now so my purchasing options are very limited)

[Sik]

Just to make sure, unplug all cards and put them back again. Maybe some card isn't connected properly and is causing the entire system to hang because something is mistimed.

[Mr. Ksoft]

I reseated everything... still trouble. In fact, more trouble :/

Now it won't boot most of the time at all. After I started it following reseating the cards, it ran for a few seconds—nothing coming up on screen-- and then shut off, sat for a few seconds, and rebooted itself. It started up, but froze during the Windows boot (which has never happened). So I shut it off and tried booting. It ran a few seconds again, shut itself off, and rebooted yet again, but this time the BIOS never came up at all. I shut it off again, let it sit for a few seconds, turned it back on, and it booted fine except froze during the BIOS. It seems to be running less and less each time.

Whoa... hold on. I'm actually working on it whilst typing this. It booted through the BIOS this time. At the very end it told me that "Overclocking failed!" which is weird because I don't partake in overclocking. So there is a thread about this a few below this thread, and I just looked at it. Unfortunately it isn't much help, as the RAM I'm using is on ASUS's compatible memory list for this board-- although at DDR3-1333 speed which is okay since I'm running at 1066, and I've already gone through a bunch of attempts to adjust the RAM (ex. the RAM list says I can run the RAM between 1.5-1.6v, so I've tried both; I've set the timings manually in case Auto was messing it up, etc) Anyway... I tried going into the BIOS settings to see if it had set some ridiculous settings this boot and it locked up immediately. So... shit.

This post has been edited by Mr. Ksoft: 06 September 2011 - 08:02 AM

[Oerg866]

.

It is definately no free-win-situatioon like "reseat cards" .

Now let me actually bring some technicalities in here..

I deduced a few things from your post (entirely helpful and no offense intended):

* The machine is not OEM (selfbuilt by somebody)
* The person who built it or picked out the parts does not have a lot of skill (yet)

You have have not listed the power supply and I think I probabyl know why. The power supply picked out by the person who built it is not a very fancy and probably cheap one. My guess is that this unit is at fault.

Please list Power supply brand and model. Do you still have warranty on it?

Cheers
Oerg866

This post has been edited by Oerg866: 06 September 2011 - 10:51 AM

[Mr. Ksoft]

This is indeed no OEM machine; I built it myself. It was my first build so I'm sure I made at least one mistake, but I was warned ahead of time to do a lot of research on the power supply. And so I did, and certainly didn't get anything cheap. I got this one, a Seasonic X750 Gold. Far as I can tell it is pretty highly regarded. Anyway, Seasonic says they give a 5 year warranty, but RMAing it is actually a risk-- not only am I paying shipping both ways but it says that if they find that it doesn't need service they're going to charge me another $25 for wasting their time. Thus, I only want to RMA it if I can definitely be sure there is a problem with it.

[HighFrictionZone]

I'm sure you already tried this, but I figure I'll offer this option:
Is there a build-up of dust and fine particulate matter on the fan/heatsink/graphics chip which would cause overheating.

Because seriously, I had a computer which had a buildup of such dust, and would shutdown due to overheat.

Then again, if it was that, you'd have more consistency when it comes to thermal shutdown. Is the motherboard defective?

[Mr. Ksoft]

When I can get it to run (ran for probably ~4 hours last night) I've been monitoring temperatures. I'm at around 40C idle, 60C while running at 100% CPU. Also checked GPU temps while running various 3D stuff and they were similar. I've looked over the case to make sure everything's spinning, and it is. I guess it could be the board... I can't really see what else. The fact that it is random makes this really hard to pin down.