Sign in with
Sign up | Sign in
Your question

BSOD when idle, fine otherwise....

Last response: in Memory
Share
November 29, 2008 8:38:25 PM

ASUS P5Q Deluxe
2.13 C2D (no OC)
Crucial 1066 1GBx4 (Stock Specifications: 2.1 V 5-5-5-15)
ATI 4750
Antec True Power Trio 650
Vista 64 & XP 32 dual boot, separate drives
Running AVG & CCC

I've got a problem that's really stumping me. It started about a month ago and I don't think I had made any hardware or software changes prior. After leaving my computer idle for about 15 hours one day(only power save feature I have is monitor off after 30 minutes), I came back to it and it was sitting on Blue screen. I restarted and it said my OC had failed. The only "OC" I'm doing is setting the RAM to 1066 and 2.1 V because the board autodetects it at 800.

I took out one of my DIMM's and it rebooted fine. Using ram test, I confirmed that one stick was in fact damaged and the other 3 were ok. I ordered a replacement. Everything ran fine for a while. Got the new ram stick and put it in. Ran fine for a while, but now it's behaving the same way. I come back after leaving it idle for a period of time...enough time for the monitor to shut off...and either I'll have a BSOD, or the monitor won't come back on and I have to restart and change a BIOS setting to get it windows to load. If I don't, it'll get to the Windows loading screen, BSOD and restart automatically. This last time it happened, I had to lower the ram speed a notch to 887 to get windows to load. This leads me to believe it's certainly a RAM setting problem, but I don't know why this just started happening all of the sudden.

This doesn't happen every time I leave the computer idle, although I'm not sure if it happens after a certain amount of time or if it's random.

What makes this hard to understand, is that I play some intensive games and I have had no trouble with stability while I'm using the system. This has only happened while I was away from the computer and it was idle. This made me wonder if it has something to do with a power saving feature somewhere that I don't know about that's causing some sort of memory corruption...or is it possible that something is damaging my ram?

I re-tested the 4 sticks of RAM that I currently have with the VISTA tester and they check out ok even after several BSOD occurances.

My next step is to leave it idling in XP to see if it does the same thing. But I'm curious if anyone has any ideas? I'm going to leave it at 887mhz to see if the problem continues...but it's 1066, I don't know why I would have to run it at 887.

More about : bsod idle fine

a c 80 } Memory
November 29, 2008 9:17:52 PM

Use memtest and let it run overnight to make sure your RAM is fine. I doubt running it at 1066 instead of 800 makes a significant difference. THat Crucial memory doesn't seem to be on the QVL for 4 modules (1 or 2 modules only).
a b } Memory
November 29, 2008 9:45:11 PM

^+1. Also note many people have trouble with the Crucial DDR2 1066 RAM. That is why I recommend getting DDR2 800.
Related resources
November 29, 2008 10:26:37 PM

I had a similar problem in the same time frame. Random BSODs after periods of inactivity with no apparent pattern to the error number, Memory checked out fine with Memtest, no recent SW or HW changes. One of the BSODs occurred just as chkdsk /r was completing (thought maybe there was a disk problem), hosing my XP installation. Had to restore my system from an image (thanks Acronis!). Never did conclusively determine the source of the problem, but I did notice that Adobe tried to push through an Acrobat Reader update (8.1.3) onto my freshly restored system, and the update didn't go well. Wound up uninstalling Acrobat and installing version 9. At the same time, I disabled the power saving feature of my USB root hubs in Device Manager just in case XP was shutting down the hubs during periods of inactivity and causing problems with the drivers for my USB devices. Never did conclusively determine the source of the problem, but I haven't had a BSOD in 3 weeks now.
December 4, 2008 12:53:34 PM

Well, I found that two other sticks were dead. Unfortunately I didn't mark the new one I got recently so I don't know if it fried as well, or if it was all the original sticks. At the moment, I have two in channel A and it's been running just fine. Hopefully it was just a power spike or something that fried 3 sticks, and not something that is continually killing my ram in B channel or something. I'll find out I guess when I get replacement ram for B channel.
December 13, 2008 12:01:53 AM

*Sigh* I still have the two sticks in that I thought were good and everything was running fine since my last post above, until now. I was playing Fallout3 and I get a BSOD. I restart and start replying to a post, and get another BSOD. Hopefully I can get through this post before it dies again. Is it possible that my motherboard is frying my ram? But intermittently? This is really getting irritating.
a c 80 } Memory
December 13, 2008 12:59:55 AM

How much voltage are you applying to your ram? What's the part number of your modules?
December 13, 2008 9:36:23 AM

It's getting 2.1 which is what the specs on the RAM say. Or is that maximum voltage they should be getting?

Sorry, I'm at work right now so I don't have the part number, but it's Crucial Ballistix 1066 1GB sticks.
December 13, 2008 9:44:39 AM

Wtf, I just looked at newegg to link you to it and I noticed there are two 1066 Ballistix with different rated voltages.... 2.2 (BL2KIT12864AA1065) and 2.0 (BL12864AA106A). I know the sticks I have are 2.1. I specifically remember seeing 2.1 V somewhere, either on the packaging or on newegg's specs.

I just looked at my Newegg order history and sure enough, the last 2 sticks I bought to replace the first two are the **106A's which are the 2.0 V. Could 0.1 V be frying them?
a c 80 } Memory
December 13, 2008 2:53:59 PM

Having the voltage at 2.1V definitely shouldn't fry them. I know that the P5Q Deluxe is very stable using 800Mhz G.Skill modules, but I haven't tested 1066Mhz modules. To make absolutely sure that memory is causing the BSODs, lower the speed to 800Mhz and see if it resolves the issue or not.

What was the BSOD error code?
December 13, 2008 10:39:45 PM

I'm not sure what the code was, I didn't write it down.

I lowered the voltage to 2.0 and played all day today and I had no problems. One would think that if I had a bad stick of ram that it would crash more consistently. What else do you think might be causing it? What steps can I take to determine whether or not for sure it's the ram?

I still have the old 2 sticks that I had determined were bad, but now I'm wondering if they really are. Vista memory check got errors on both of them, but that doesn't necassarily ensure that the stick is bad, does it?

If I get another BSOD this week, I'll lower the speed to 800 like you suggested. I'll leave the system on overnight, which is when it seems to have the most problems staying "alive".

In any case, I ordered 4 2GB sticks of OCZ 1066 today.
a c 80 } Memory
December 14, 2008 12:27:28 AM

But what does 1066 memory provide, other than slightly better benchmarks?
December 14, 2008 9:58:48 AM

I don't know, I just figured it would be faster than 800. Is 800 faster than 400?
December 14, 2008 12:05:29 PM

So I did some reading since my last post and I think I understand now. Basically, I wouldn't benefit from 1066 unless I was OC'ing the CPU higher than 400mhz fsb, correct? And that's assuming I don't want to OC my ram? This year I upgraded my mobo from a previous ASUS that only supported 667mhz RAM, just so I could get faster RAM. Basically this upgrade wasn't really an upgrade at all unless I overclock my processor past 333, right? I'm just trying to understand for future builds/upgrades.

If I now understand correctly, it begs the question, what is the purpose of 1200mhz RAM? Is it possible to OC a CPU past 533 Mhz FSB? That would turn my 2.13 into a 4.16+....and then into a pile of goo.

Well, it's already bought anyway and I don't mind paying $70 for 8 gigs.
a c 80 } Memory
December 14, 2008 12:56:59 PM

I'm not saying that 800Mhz is as fast as 1066, but give it a try to determine if it makes your system stable. I prefer a slightly slower and stable system over a faster one that sometimes crashes.

To be honest I don't really see the purpose of 1200Mhz RAM for the average user or even most overclockers. Faster RAM often is 800Mhz that's overclocked and requires more voltage and lower timings to run. It's useful if one must have the fastest system and achieve the highest benchmark scores.
December 14, 2008 6:32:37 PM

I appreciate your help through this.

A few minutes ago I got a back to back BSOD so I lowered the speed to 800. My new RAM will be here Wednesday and I'd like to know for certain that it's going to be safe to install. I don't want to burn up another four sticks, if that is in fact what is happening.

My motherboard auto detects 5-5-5-18 whereas the RAM specifies 5-5-5-15. I've been running hte RAM specification but do you think I should just let the board autodetect everything for a while? I have a friend who has almost the exact same system that I built for him and he's having no problems. The only difference is that I don't think I took the time to check the CMOS AI settings so his is probably running 800mhz and 5-5-5-18.

I'm going to do some Memtest on my sticks today to see what my damage is....
a c 80 } Memory
December 14, 2008 7:53:46 PM

Running at system defaults certainly can't hurt.
December 14, 2008 7:58:05 PM

*Big Sigh* After the most recent BSOD, I ran memtest86 with both sticks in channel A, which is where they've been. I got two errors within a minute, but no more afterwards. I then tested each stick separately and got no errors on either. Then I tested the old two sticks separately and guess what, no errors now, even though they had errors before.

Now I've got all 4 sticks in and it's running fine at the moment. I'm certain it will eventually crash and most likely while I am AFK, which has been the case 80% of the time.

Conclusion, it's got to be either the motherboard or the PSU. Because it seems random, this makes me think it's the PSU. Unfortunately I dont' have another to test with. I haven't had this one long, probably about 8 months. What are the chances that it's the mobo? I would think if it's the mobo, that it would be a more consistent problem.
a c 80 } Memory
December 14, 2008 10:11:47 PM

You should never get memory errors. I doubt it's the motherboard or even the PSU. Did you get memory errors running it at 1066 or at 800?
December 14, 2008 10:14:27 PM

800. And I got another BSOD since my last post. I changed everything to auto settings (266 FSB, 800mhz RAM, Auto voltage, auto timings) for the memory tests and left them there.
a c 80 } Memory
December 14, 2008 10:22:06 PM

Are those modules on the QVL? Can you get other memory modules on the QVL to test with?
December 15, 2008 9:12:28 AM

I haven't checked. The only other RAM I'll have access to is the OCZ that's coming in Wednesday.

I ran Memtest all night (almost 12 hours) with all 4 sticks in and got NO errors.
a c 80 } Memory
December 15, 2008 2:54:40 PM

Also run Prime95. Memtest is good, but it doesn't make the system work hard.
December 18, 2008 10:26:43 AM

Well, my IT director at work said it's probably not a RAM problem and he suggested to check drivers and PSU. I borrowed his PSU tester and it checked out ok. I updated my video driver and reinstalled my sound driver. It's been two days and so far I've had no problems! YAY!
December 23, 2008 9:06:13 PM

Sure enough, no problems! It must have been either the sound driver or the video driver. Just wanted to follow up.
December 23, 2008 9:48:00 PM

excellent, thanks for posting back, will give the next person with this problem some help.
!