Bluescreen = dead RAM right?

5014 Views 19 Replies 3 Participants Last post by Movitz, May 26, 2008

Movitz

Discussion Starter · May 23, 2008

Hi guys, quick question which I think I already know the answer to.

So some files on my comp has been showing some CRC-errors when I download or copy between my drives. I recently also got a bluescreen with the inevitable auto-restart after some 2-3 seconds, this was while watching a of video clip. This only happened once, last night and has never occurred since, not even during heavy load (gaming) which should exclude the PSU. The power supply should be sufficent anyways, which basically means that one of the 4 Corsair sticks are damaged right? Oh, and memtest gives errors instantaneously.

I'd apprieciate your opinon before I start going through the memory modules one by one via memtest and doing the whole return to manifacturer ordeal.

Also, specs:
Asus p5k/C2D E6750/2xCorsair TWIN2X 6400C4 DDR2/Gainward GeForce 8800GT/Seagate Barracuda 7200.11/Antec TruePower Trio 650W

Status: Not open for further replies.

1 - 20 of 20 Posts

Underclocker

· #2 · May 23, 2008

The only way to verify that your memory is good is to make sure your BIOS has the correct timings and voltage set up to suit, then Memtest each of the modules individually.

Movitz

Discussion Starter · #3 · May 23, 2008

Yeah, I suspected as much. I just wanted to hear if there was some other explanation before I started. It's pretty time consuming since I probably want to run the sticks for an hour each and then start combining them.

Underclocker

· #4 · May 23, 2008

You could set up the correct timings and voltage, then do an ORTHOS test overnight with large FFTs.

Movitz

Discussion Starter · #5 · May 23, 2008

I have actually never heard of this program. I'll test it right away.

Right now the memories are running at 5-5-5-18 (instead of 4-12-12-12) because that's what the system defaults them to, I never saw any performance gain with CAS4 so I just left them at default. Is there any way of them beeing more unstable at CAS 5? Shouldn't CAS5 be just as stable since it's also taken from the SPD table? I never though of it.

Underclocker

· #6 · May 23, 2008 (Edited)

Sorry, forgot to give you a link to ORTHOS: http://sp2004.fre3.com/beta/beta2.htm

Choose the "Large, in-place FFTs" test.

Your system should be just as stable at 5-5-5-18, if not more. You could try CPU-Z to see the recommended timings and voltage on the SPD tab. http://www.cpuid.com/cpuz.php

Then just change the BIOS settings according to those values.

EDIT: It might take a couple of hours before you get an error, which would indicate either an unstably clocked CPU or bad RAM. If it's been running for more than 24 hours and still no errors, your system should be stable and problem-free.

Movitz

Discussion Starter · #7 · May 23, 2008 (Edited)

Thanks. I already found ORTHOS and I was using CPU-Z for finding the SPD :smile:

It boggles me though. Is there any reason to why memtest gives errors instantly but ORTHOS might be ok?

[Edit] To clarify my logic. Why would you recommend me to use ORTHOS when memtest already fails. Is there any scenario where memtest produces errors that ORTHOS does not

Movitz

Discussion Starter · #8 · May 23, 2008 (Edited)

So I've been running ORTHOS now for 8,5 hours without any problem. I'm turning it off because I'm going to bed and I'm not too comfortable leaving the comp on overnight since I've had two PSUs fail and start smoking on me. Fire while sleeping, not good.

I'll continue running in the morning but I'm still curious to why ORTHOS has no trouble while memtest insta-fails.

Also, why CRC-errors on my HDD if it's not the RAM/CPU (as ORTHOS seems to indicate)? The Seagate diagnostics tool says my disc is ok.

Oh. And I want you to know that I'm really grateful for the answers so far, thanks.

dai

· #9 · May 23, 2008

with memtest you need to run it on one stick at a time or you will receive errors where there are not any
the p5k board defaults at 1.8v my corsair ram i had to lift the voltage to 2.2v

Movitz

Discussion Starter · #10 · May 25, 2008 (Edited)

Firstly I want to thank Dai, upping the voltage helped me to get my memory stable at lower latencies. I was going to try this some day but never got around to do it until now.

I ORTHOSed my system for 8 h at 5-5-5-18 and 12 h at 4-12-12-12 without any problems. Since a faulty module would provide errors regardless of timing this effectively means that my system was stable for 20h (right?). Maybe I'll let it run for 24 h on one timing some day but for now I'm happy.

But it does not explain why I get the CRC errors. The bluescreen migh just have been fluctuations in the eldrich energies that governs the spacetime continuum and doesn't bother me as long as it doesn't repeat it self.

Should I consider the RAM non-faulty now since ORTHOS seems fine? Is is possible that the problem is with the motherboard or the HDD instead?

dai

· #11 · May 25, 2008

run
chkdsk /r /f

Movitz

Discussion Starter · #12 · May 25, 2008

That will only recover the files that are corrupted right? I'm looking for a way to understand why they get corrupted in the first place.

dai

· #13 · May 25, 2008

the main cause is improper shutdowns be it by the computer rebooting itself unexpectedly or the user not waiting for windows to close it down when turning off

Movitz

Discussion Starter · #14 · May 25, 2008

No not in this case. I can take a large file, copy it from my D:\ to C:\ and it will,almost certainly, be corrupted afterwards.

That's why i suspected the RAM, but since the system is so stable in ORTHOS I don't know what it migth be any more.

dai

· #15 · May 25, 2008

there is a explanation in the second post here
http://forum.osnn.net/showthread.php?t=18873

Movitz

Discussion Starter · #16 · May 26, 2008

Hmm. I seem to have the exact same problems as that guy. Too bad they don't present any solution in that forum. I tried changing the SATA cable and I've made sure that there's no overheating in that area.

Seriously, what the heck is the problem here? Is there anyone who ever experienced something like this?

Maybe this is an HDD issue and should be moved to the correct sub-forum.

Movitz

Discussion Starter · #17 · May 26, 2008

Can't move files, they're in use. Also, chkdsk says active processes

Hi guys.

I have two problems. I don't think they are connected though so I posted the first one here because I believed it to be due to faulty RAM. Basically, I get CRC-errors if I copy files around on my HDD but after 20 h of stable ORTHOS I don't think I can blame the RAM anymore.

The second problem is that I cannot move files between my D:\ and C:\. It copies the file then locks both the new and the old one down and the only way to remove the old one is to reboot and delete. I can however copy and then delete manually. This problem does not exist while in safe mode, there I can move stuff to my heart's content.

The C and D are both partitions on the same HDD. I also have another one (E

and I recently noticed that when I unplug it....I can move stuff between C and D again without auto-locking of the files!

The topmost problem (CRC errors) is most important to solve, it's driving me crazy and I might lose files to it. The second one is more of a minor problem but annoying non the less.

If anyone knows anything about this I'd appreciate any input whatsoever. I'm going crazy over here.

Specs:Asus p5k/C2D E6750/2xCorsair TWIN2X 6400C4 DDR2/Gainward GeForce 8800GT/Seagate Barracuda 7200.11 500Gb, SB 7200.8 200Gb/Antec TruePower Trio 650W

See less See more

dai

· #18 · May 26, 2008

with the p5k and the 8800gt the trio should be sufficient
personally i would be running a corsair 750

dai

· #19 · May 26, 2008

threads merged

Movitz

Discussion Starter · #20 · May 26, 2008

Is it possible to send the thread over to the HDD subforum? I don't think it's the RAM that causes the problem anymore. I also added another problem that doesn't have to do with the RAM at all. I suspect they might know more on this over att HDD.

Thanks mate

1 - 20 of 20 Posts

Status: Not open for further replies.