Blue screen of death on boot

[CrimsonJustice]

Hi guys. I've been having a problem that just came up in the last 4 days on my homebuilt system.

I built my machine about 3 months ago and it has been perfect... until now. 4 days ago when I booted it from a cold boot (I turn it off at night since I can't sleep when it's running), it blue-screened on me after booting. I waited for the dump and it restarted and ran fine the whole day. The next day, it did the same thing. Likewise for the third. However, today when I started it it got 3 blue-screens in a row before now running for hours. I was a little worried before, but now I am quite concerned, and thinking it may be some failing component.

Everything is still under warranty, so if I can figure out what's the problem I have no concern sending it in. I know I should run a memory test, hard drive test, and video card test at the least. But I'm not sure which ones to run or in what order. Could I get some suggestions from you guys as to where to go on this?

Here are my specs:

AMD Phenom II X4 965 C3 (not yet overclocked due to busyness/not needed atm)
4 x 2GB G.SKILL F3-12800 Ripjaws (mobo set for 1600MHz clock speed)
ASUS M4A79XTD Evo
Corsair Professional HX850W
2 x XFX Radeon HD 4770 in Crossfire
Western Digital Caviar Black 640GB
Windows 7 Professional Edition x64

 

[capt_taco]

Memory is always the first thing to check when you're having random freezing or BSODs. Two possibilities come to mind with your RAM:

1) Voltage or timing is not the standard 1.5V and is not set up right in the BIOS, or
2) Since you have 4 sticks, you need a little over manufacturer's spec voltage and either you or the motherboard did not compensate for that.

It sounds like you are somewhat familiar with setting the memory timings/voltage, but it won't hurt to double-check what the manufacturer's recommended settings are and set them manually. And also, with 4 sticks, I'd give it an additional 0.05V-0.1V above that. One weird thing about voltage and timing issues is that sometimes a system will seem to be running fine for a few weeks or months and then start acting up all of a sudden.

If that doesn't fix it, run Memtest. If that doesn't find any errors, it's time to be concerned about something more complicated.

 

[JackNaylorPE]

When you did get in, did you examine Event Viewer messages or look at the dmp file ?

 

[Griffolion]

If you can get us what the blue screen says that would be brilliant. If its anything to do with a page fault then that will tell us a lot.

 

[CrimsonJustice]

@capt_taco: The voltages are indeed not the standard 1.5v. I had to set the timings manually and put them at rated 7-8-7-24 that the modules are at. They are rated at 1.6v, but I put it at 1.62v for a tiny bit of overhead. My BIOS goes in increments of 0.02v by the way. F3-12800CL7D Should I go up to 1.66/1.7 then? Or should I possibly drop my timings for a few days and see how that works out?

@JackNaylorPE: Windows 7 bring up the crash dialog to inspect the error right away, but went I went to view it it was absent from the Temp directory it seemed to be stored in. However, there is a report for all of them in the event viewer. Not very detailed though.

@Griffolion: Tomorrow when it happens again I will grab the info.

Thanks for the replies so far guys, I will update tomorrow to see if it happens again and get more info.

 

[CrimsonJustice]

Same thing this morning. Here's what I have for the info.

Problem signature:
Problem Event Name: BlueScreen
OS Version: 6.1.7600.2.0.0.256.48
Locale ID: 4105

Additional information about the problem:
BCCode: d1
BCP1: 0000000000000009
BCP2: 0000000000000002
BCP3: 0000000000000000
BCP4: FFFFF88006423F00
OS Version: 6_1_7600
Service Pack: 0_0
Product: 256_1

BSOD Message:

A problem has been detected and windows has been shut down to prevent damage to your computer.

IRQL_NO_LESS_OR_EQUAL

If this is the first time you've seen... -snip, your standard BSOD info-

Technical information:

*** STOP: 0x0000000A (0xFFFFFA7F0A77B1A0, 0x0000000000000002, 0x0000000000000001, 0xFFFFF80002C932A3)

Collecting data for crash dump ...
Initializing disk for crash dump ...
Beginning dump of physical memory.

Any other suggestions, guys?

 

[CrimsonJustice]

Well I'm working on solving it myself, since it did the same thing once again today. I took out the second channel of dimms (still have 4GB after so I think I can manage) and left the pc for a couple of hours so it would have to do a cold start again. After booting with just the two, I had no blue screen within 20 mins, but I decided to ru memtest on these modules to verify their integrity anyhow. They checked out 100%, so it'll be the other two that will get checked tomorrow. I'm going to update this thread further with results in case it helps people in the future.

 

[capt_taco]

@capt_taco: The voltages are indeed not the standard 1.5v. I had to set the timings manually and put them at rated 7-8-7-24 that the modules are at. They are rated at 1.6v, but I put it at 1.62v for a tiny bit of overhead. My BIOS goes in increments of 0.02v by the way. F3-12800CL7D Should I go up to 1.66/1.7 then? Or should I possibly drop my timings for a few days and see how that works out?

I would definitely try increasing it further when all four sticks are in. I wouldn't give up on that fix unless you get to 1.7V and it's still crashing. Don't worry, you have to give RAM a lot more juice than an extra 0.1V in order to damage it.

The fact that it seems to be doing better with just two sticks at a time is an additional hint that the four-sticks voltage issue might be the problem. I am not entirely clear on the technical reasons behind that particular quirk, but it seems to happen all the time.

 

[CrimsonJustice]

Actually, I did try what you suggested, capt_taco. I pushed the voltage up to 1.7V before I moved down to the two sticks, but it was still having issues. In fact it seemed to be having even more issues with programs randomly crashing during use at that time. Perhaps I need to modify some other board voltages at the concurrently? I was considering more, and certainly tried 1.76V, but couldn't get it to post at that point, so I loaded the boot-safe settings and re-loaded my profile.

Today, the RAM I had removed remained in the plastic container and it booted fine again. I will try booting with those tomorrow and test them, and then play with voltages some more if they check out fine.

 

[CrimsonJustice]

Well, seems no matter what I do they all have some problem, or it's not the RAM. I've tried undervolting, overvolting, and setting it to the recommended specs. My attempts to narrow down a specific stick were unsuccessful, for just as I thought the two were the culprits, I put in the ones that seemed fine and the problems still occurred. So I'm at a loss. Hopefully G-Skill Tech support will have an answer. After spending ~$200 on these things I'd hope they'd last longer than 3 months.

This has been quite a pain. Still, if anyone has anything to suggest besides getting a hammer...and yes, I've thought of that, I'm all ears.