Announcement

**profJim** · 10-28-2009, 12:06 AM

Re: Wow, you guys are awesome

Glad we could help out.
If you post a screen-shot using cpuz or Memset SPD tab, you'll see the rated voltage specified for your memory at several memory speeds.
This voltage is what the manufacturer requires for each memory speed range.

Screwing it all together is easy but the technical aspects of tweaking the system to get it to work optimally (or just work) are all totally new to me.

Yes, the devil is in the details

Several threads here deal with Q9550 overclocks up to 3.80 - 3.90GHz and higher.

**Lsdmeasap** · 10-28-2009, 12:21 AM

Re: Wow, you guys are awesome

Great to see you found some help!!

Sometimes a small voltage increase is all that is needed, sadly sometimes users do not want to hear or try that.

Post back anytime!

**wallypiper** · 10-28-2009, 05:12 AM

Re: Wow, you guys are awesome

I have depended on the BIOS defaults mostly to handle all the settings. I recently uninstalled ET6 because it became a suspect in the BSOD storm so I don't currently have a quick way to check on temps but at the end of 90 minutes of Memtest86+ both the CPU and System temp were around 40 and both began to cool as soon as the test stopped.

Here's some cpuz shots. Tell me what you see. Given that I know hardly anything about most of the settings and specs, you can't hurt my feelings.

**profJim** · 10-28-2009, 05:33 AM

Re: Wow, you guys are awesome

Running your memory at the rated 1066MHz speed (533 MHz in cpu-z), your DRAM voltage should be 2.10 volts (possibly a bit less) for stability. Memtest86+ is good a finding memory hardware problems and "gross" bios configuration errors, but it is not a stress test.
You need to run Prime95 or OCCT for 10 - 12 hours or longer to verify that your system is truly stable.

**Chike** · 10-28-2009, 05:39 AM

Re: Wow, you guys are awesome

Originally posted by wallypiper View Post

So I got Memtest86+ (plain 86 couldn't test both sticks together, it kept crashing). First try I got 80 errors halfway through the first pass.

Turned the ram voltage up from 1.80 to 1.84.

Then I got 20 errors on the first pass. Stopped the test and turned it up again to 1.86.

3 passes, no errors.

This is interesting.
With DRAM voltage set to auto the memory would have never been set to 1.8v. With XMP enabled it would have been set to 2.1v and if it was disabled to 1.9v, so which were the actual values that failed?

memtest96+ would not be rnough to prove the settings stable, some more aggressive tests are needed, Prime95/Orthos blend or OCCT large data set.

**wallypiper** · 10-29-2009, 05:42 AM

Re: Wow, you guys are awesome

Immediately prior to running Memtest I flashed the bios to the latest version and, per the instructions on this forum, set the bios to optimized defaults. I didn't make any changes to the bios settings before running Memtest.

When I entered the BIOS setup after the first memtest run, I went direct to the DRAM voltage setting. It may have been set to Auto but when I typed ENTER, the setting which was selected in the list was 1.80. Perhaps when you change it from Auto it immediately defaults to a lower setting when it shows you the list. But the sequence of events following that is exactly as I reported - 80 errors on the first test, 20 errors at 1.84, 0 errors at 1.86.

As for running a more agressive test, I might do that. But I am not prepping this system for sale. It's my personal use computer at home. If it runs without errors as currently set up, I'm satisfied. Running a different test and generating errors that would never occur during normal use might be interesting, but it wouldn't really make my experience using the computer any better.

**Chike** · 10-29-2009, 06:19 AM

Re: Wow, you guys are awesome

What makes you think that memtest86+ represents your normal use better than Prime95?
I have core temps monitored and quite often temps get near the temps generated by Prime95 at normal use.

Perhaps you will zip/rar some files one day, play a new video game, it will fail/crash. or the computer will, maybe not, then subsequent random BSODs, then you'l understand why testing stability is important.

**wallypiper** · 10-29-2009, 07:40 AM

Re: Wow, you guys are awesome

Chike,

I appreciate your input.

My computer only does two things that I consider "important".

I use it to track my personal finances and business expenses with Quicken.

I use it to access the internet to download financial information from banks and lenders into Quicken.

Everything else it does is entertainment and is not "important".

Given the relatively light load placed on the system by those two tasks, and the fact that I have gone from BSODs every hour or less to zero BSOD's in two days regardless of what the computer is being used for, I feel that its current state of stability is adequate.

If, at some point in the future I have problems with it, I will again take on the task of tracking down the source of the problems and solving them. I may even explore tweaking the BIOS settings in my computer beyond the small DRAM voltage adjustment I just made, just for the fun and challenge of it. It is something I know almost nothing about and I might enjoy learning about it. But right now, its working and I have many other issues in my life that need attention that have, frankly, gotten shorted over the last week or so while I worked on this problem.

I look forward to exploring this forum some more and mining what is obviously a broad based and deep resource of information about the extremely esoteric and highly technical realm of modern computer operation at the board level.

For right now, Firefox, Quicken, MSmail and a handful of low end games that my wife likes to play (card games and mental puzzles) are all working great. Thanks again for your assistance.

**wallypiper** · 01-11-2010, 06:50 AM

Re: Wow, you guys are awesome

Well, I'm back and prepared to possibly eat some crow.

About a week ago, I started to experience some problems with Vista booting. I was getting an error that hal.dll was missing or corrupt. It wasn't missing and I tried replacing it with a copy from a machine that was working to no avail.

I decided to go ahead and do a clean install of Windows 7, which I have done. The install was uneventful but almost immediately I got some more boot failures with the same error.

I had not, however, installed new drivers from Gigabyte so I downloaded and installed drivers for the intel chipset, realtek audio and realtek lan. I also installed the latest driver for my EVGA graphics card.

Everything seemed to work fine for a day or so as I gradually moved data files etc. back onto the drive.

Then there were a couple of BSOD events but the system was rebooting so quickly after them that I did not see what the errors were.

BUT I have had several instances of the system reporting a boot disk failure. I have always been able to get a successful boot simply by either doing a hard reset or powering the system down and starting over.

I have run Seagate's test utility (self booting, not in Windows) and it found no errors.

I will gladly post whatever information is useful but I am not certain what that is.

When the boot disk failures occur, there is a sequence that seems to be pretty consistent. First, the Gigabyte splash screen is displayed for an unusually long time. Also, the HD activity light is on and stays on steadily, no blinking at all. So far, I have always been able to get the system to boot eventually, either by pressing the system reset button or, more often, by powering the system all the way down and restarting.

**wallypiper** · 02-12-2010, 05:08 AM

This looks seriously fatal

Returning from a few days away, I found my system in a BSOD, waiting for input. I tried the reset button but experienced the same sequence described above - long pause on the splash screen, HD activity light on steady and then a boot disk failure message. I powered it down and cold booted and it came up OK. Then overnight, it BSODed again and now it won't reboot no matter what.

I inserted a Seagate test CD and forced a boot from the CD. The Seagate utility could not find a hard drive.

I inserted a Windows 7 recovery CD and again tried to force a boot from the CD but now it won't boot from that CD either.

I think I have found the problem. I knew from reading here and elsewhere that the first instruction was going to be "check all the power cables and SATA cables". So I opened up the case and checked and checked and did not find any loose connections. So I hit the power button again and saw the following:

The CPU cooling fan was not starting. It would move a little bit, but not spin up. I gave it a little nudge and it took off. Then I noticed that one of the three cooling fans on the Corsair memory cooling array was turning slowly. I tired "helping" it also but to no avail. It kept turning somewhat slowly and erratically.

So I removed the memory cooling fans and tried again. This time, the CPU fan still needed some help to get going but I was again able to give it a little nudge and get it running. then the system booted normally.

So my theory is that the sticky fans are pulling the system voltage down somehow and messing with other subsystems on the board. I will get a new CPU cooler and contact Corsair regarding the memory cooling fans. I'm not sure they are necessary as there are many other fans in the system - two inlet fans to the case, one large exhaust fan from the case and the fans on the power supply and graphics card.

Comments? Suggestions?

**wallypiper** · 02-12-2010, 05:34 AM

Ooops

After a minutes just idling, not doing anything with it:

"STOP: c000021a {Fatal System Error}
The windows subsystem system process terminated unexpectedly with a status of 0xc0000006 (0xfdc031b0 0x00a1f150).
The system has been shut down"

**profJim** · 02-12-2010, 05:55 AM

Re: Wow, you guys are awesome

So how would you like your crow served?
raw? medium-rare? well done?

Don't feel too bad, read the last half of the first line in my forum signature.

The problem with not doing extended stress testing is that a system that isn't fully stable can result in data or file corruption.

Follow steps 1 - 4 in this post to provide the information we need.
If you are running stock memory and cpu speeds, your system will probably be stable using the recommended settings in the post.

Update the information in your signature to include the bios and motherboard versions and include a link to your memory's specifications.
Use the generic bios template in this post to provide your current bios settings.

Let us know if you want some help with running Intel BurnTest, Linx, Prime95 or other stress testing programs.

**Chike** · 02-12-2010, 06:12 AM

Re: Wow, you guys are awesome

Maybe your PSU is dying?
Check PC Health status voltages.

**wallypiper** · 02-13-2010, 04:22 AM

Re: Wow, you guys are awesome

The system is running at the moment after being completely shut down for about 24 hours. I'll get to work on the stuff you asked for.

I found this pretty interesting. When I google the exact error message I got, "The windows subsystem system process terminated unexpectedly with a status of 0xc0000006 (0xfdc031b0 0x00a1f150)", I only get one hit which is my post here. In all of googleland, nobody else has ever posted that message. I don't know whether to be terrified or proud.