Blue Screen of Death too frequently

[zahpod42@tampab]

Hello. Recently I started experiencing system restarts for no known reason. I have had the same hardware in my system for a least six months, and all I have done software wise is to do a fresh install of Win XP and install the latest drivers of my hardware. The restarts happen seemingly randomly, and not just during any specific application. Sometimes I can use my computer for a whole day without a restart, and other times it will happen frequently. I have even had it happen within seconds after my computer booting.

I changed the restart automatically option the other day, so that I would get the BSOD and could write down the error messages listed there. So far I have only written down one of these messages. It basically said that the error was "caused by the following file: dxg.sys..." "page fault in non paged area..." "dxg.sys - Address BFF83AF1 base at BFF80000...Date Stamp 3d6fa2a..."

This sounds to me like possibly a hard drive that is going bad, or memory modules that have suddenly started to fail...of course, I don't know what I am talking about either.

Could this be a result of the last time that I updated my video card or motherboard drivers? The BSOD does not happen systematically enough for me to be able to narrow down the problem. And, since I do a fresh install of windows each time I update my drivers, I cannot exactly just roll back to a previously working version of Windows. Like some commercial said, I don't even know enough to know that I don't know. Can someone please tell me something specific that I can try to diagnose this problem. And, I don't know how to edit my registry, or do a lot of things that people might be inclined to suggest. So, if anyone is kind enough to offer me help, would you please give me very detailed instructions, or point me somewhere that will give me specific instructions? Thank you for your understanding.

Also, I have recently added PC Cillin 2004, although I did not necessarily notice that the problem started right after that or anything, as the crashes are normally so infrequent, that I didn't even really take note of them at first.

Thanks in advance,

zahpod

[Senor Panadero]

Those restarts can be from several things. That message

"caused by the following file: dxg.sys..." "page fault in non paged area..." "dxg.sys - Address BFF83AF1 base at BFF80000...Date Stamp 3d6fa2a..."

May hold a clue, but it might also just be what was in memory when the system went unstable and not a direct cause of it. All the same, you can try Start->Run->dxdiag and see if any errors come up in the tests. If so, you might re-install DirectX then re-install your video drivers.

Since you recently installed a new video card and drivers, you may have a problem there. You can try using a driver cleaning tool and search for and uninstall any perevious drivers that may be left over in your system. There are instructions and utilities for doing that in each of the Video Card Forums sticky posts. Afterwards, you may want to re-install your current video drivers and/or try a newer/older/different version.

Might be unrelated to Direct X, might be a hardware issue. Check your cpu temperature in the bios Hardware Manager screen. Let it sit there for a while, see how hot it gets.

Could be case heat. Try taking the side panel off of the case, get some air moving in there and try to duplicate the problem.

Edit: Scratch this, we have psu info.
One thing. Several have had problems with this particular power supply. No instability that we could really blame on it, but the 5 volt line does seem to dip alarmingly, and that is the line that the video card and cpu draw on heavily with this board. Just throwing the thought out there.

Could be memory problems. Yours looks good, but the best memory can have flaws. These flaws can add up over time and start to corrupt system files and give instability. Try testing with Memtest. Download it from www.memtest.org . Either get the zip file and follow instructions to make a floppy, or download and burn the iso. Set these disks to boot, and allow the tests to run for at least one, better two or three passes of the seven tests. Any errors that show up are a problem.

Let us know what turns up.

[zahpod42@tampab]

Hola Senor,

I probably didn't explain everything well enough, but to clarify, I have not installed a new video card, only new drivers. I have had no hardware upgrades for more than six months, but the BSOD's only started in the last two weeks or so.

I ran memtest, seven times through the default test, and one time with the "all tests" setting on. Out of the seven times that I let memtest run on the default setting in a continuous loop, it only found one error. And it found no errors on the "all tests" setting. I don't know how big of a deal that one error is, but it was on Test 4, Failing address 000376981a4-8865MB. When I was clicking around trying to get more specific info on the error, it said something about, "badram=0x376981a4, 0xfffffffc." I don't know what any of that means, but like I said, the error only occurred once, out of all seven times + once on "all tests" that I ran memtest.

As for the temperature, I think I am ok there. After running memtest for about six hours (one pass through at the "all tests" setting) I restarted the computer to the BIOS screen an observed the temperatures listed there for about 15 minutes. Bear in mind that the room temperature near the computer was 82 degrees Farenheit. The MB temperature was consistant at 31 degrees Celcius, and the CPU temperature fluxuated between 33 and 34 degrees Celcius. So, I'm no expert, but I don't think that it is a cooling problem unless it is some freak hot-spot on a specific section of the MB, which I don't know how to go about testing for.

So, Senor, could you please tell me what you think of my memtest results? And is there a better way to check my hard drive than just running scandisk? How would I go about checking my PSU? I don't know what to do next. A complication is the fact that the BSOD can be so infrequent at times that it would be very difficult to narrow down the problem by trial and error. I think the only way that I can truly find the problem will be the result of some type of diagnostic test.

Thank You in advance for you help and expertise,

zahpod

[Senor Panadero]

I probably didn't explain everything well enough, but to clarify, I have not installed a new video card, only new drivers. I have had no hardware upgrades for more than six months, but the BSOD's only started in the last two weeks or so.

OK, well if the new drivers even roughly coincided with the problem, might be something to check into.

I ran memtest, seven times through the default test, and one time with the "all tests" setting on. Out of the seven times that I let memtest run on the default setting in a continuous loop, it only found one error. And it found no errors on the "all tests" setting. I don't know how big of a deal that one error is, but it was on Test 4, Failing address 000376981a4-8865MB. When I was clicking around trying to get more specific info on the error, it said something about, "badram=0x376981a4, 0xfffffffc." I don't know what any of that means, but like I said, the error only occurred once, out of all seven times + once on "all tests" that I ran memtest.

Opinions will differ, but that one is too much for me and I will explain. Good thing you wrote that down. IF you loop the default tests again a couple of times, AND you continue to get an error at that same address, then it might be evidence that you have a bad register in the memory. What I suppose would be a physical flaw. Every time that data is loaded into that area and therefore into that register, it will come out changed, corrupted. It is only a tiny amount of data, but over time that can add up. If you contacted Corsair, and told them that you were able to test and get an error in the same memory address consistently, they would swap the ram out for you.

The error can also be due to the FSB (CPU External Frequency in your bios) being too high.
I do not know if you are overclocking. But I know from testing with my Revision 1.0X A7N8X, that if I have the FSB up high enough to get one error every other pass of the tests, that is enough for me to get an intermittent reboot (blue screen) when playing games or running Folding at Home. Therefore I cannot use that FSB permanently and turn it down until no errors ever occur. If you are not overclocking, then this shouldn't be the issue. The error can even come from other sources than the memory, and you did not get it consistently, so maybe leave that alone for now.

And is there a better way to check my hard drive than just running scandisk?

Yes, you can use the drive diagnostics from the manufacturer. I don't see the drive listed in your sig, or must have missed it in your post. In any case, it will be available from the support section of the manufacturers home page. Google will get you there in one or two clicks.

Somehow I don't think you are overclocking. One thing out of the blue I might suggest, it's been lucky lately, set your CPU Vcore setting to "Menu" and raise your CPU vCore to 1.675.

Let us know how it goes.

[zahpod42@tampab]

Senor Panadero,

Ok, I have run some more tests, taken note of a few things and here is the info:

First off, I have run memtest, both the standard and "all tests", numerous times in the last couple of days, looping the tests over and over again, and I have never gotten a single error detected there except for that one that I already mentioned in my earlier post, and it has never shown up again.

Also, I have run the extended Powermax drive diagnostic from Maxtor, and the program said that my drive was certified error free. Oh, before I forget to mention it, my HD is a Maxtor 40GB 6L040L2, with 2MB Cache.

Now, I did have two BSOD's yesterday. One while I was looking at a plain jane HTML website, with no animations or anything like that in in. The BSOD file or error listed then was "Bad_pool_caller." Then, later that evening, I was playing Planetside when I got another BSOD. The file or error listed then was, "Memmory_Management."

So, I am fairly certain that the cause of the BSOD is not my memory modules, my Hard Drive, or a temperature related issue. Which, I think leaves just the possibility of an issue with the PSU or a driver related issue. Well, I went in and looked at the voltages in the BIOS (by the way, I upped the CPU voltage to 1.675 like you suggested, but I did not do that until after the second BSOD, so whether this has helped or not I don't yet know). Anyhow, the 5 Volt line was reading steady at 4.91, so I do not know if this is within tolerance or not.

As for a driver being the issue, I installed the same exact drivers on my two other computers and they have not had this BSOD problem. One is an A7n8X Deluxe with an Athlon XP 1500 and a Geforce 4400; and the other machine is an A7n266-C with an Athlon XP 1800 and a Geforce 4200. So even though the drivers themselves are not "broken," perhaps they just work better with the other two computers' hardware configuration than with mine?

Thanks for you ongoing suggestions and advice. What should I check for next?

P.S. You are correct, nothing in my system is overclocked.

Thanks,

zahpod

[zahpod42@tampab]

Senor Panadero,

I upped the voltage like you suggested, and I did not have a BSOD for several days. Then all of a sudden I booted my computer a few minutes ago, attempted to open Outlook Express as soon as it had booted, and BAM I got a BSOD that said "Memory Management."

Please Help. Should I replease my PSU?

Thanks,

zahpod

[Senor Panadero]

Senor Panadero,

I upped the voltage like you suggested, and I did not have a BSOD for several days. Then all of a sudden I booted my computer a few minutes ago, attempted to open Outlook Express as soon as it had booted, and BAM I got a BSOD that said "Memory Management."

Please Help. Should I replease my PSU?

Thanks,

zahpod

Over the months I have soured somewhat on that particular PSU and the brand in general, but I am not going to tell you to buy a new one. Borrow and try one, maybe, buy no. It should be just fine for what you are doing. Is it (the psu) running hot? Try the case open thing again.

Your new BSOD, that is a very infrequent error, and you can't tie it to anything. would be much easier if it were repeatable. I'm oriented toward hardware, and keep looking to that as a cause, but that may be misguided. Could just as easily be software, an explorer/Office add-on, spyware, virus, outdated driver. Don't know.

But the only thing left I can suggest might be to look at the memory timings. If that memory was new six months ago, then the timings should be showing up in this screen as no lower than 6-3-2-2 going top to bottom, even 6-3-3-2 should be OK. If you get something else there, please list the numbers here as it might be worth a try changing them. Something ending with 2.5 might be a problem with that memory. So might something like 6-2-2-2. If it is older than that, say a year or so, the situation changes.

[zahpod42@tampab]

My memory timings from top to bottom are:
5
2
2
2T

This is by SPD though, and they are the same timings listed for this memory on Corsair's website. I don't have anything in my system overclocked, and the memory timings haven't been changed since I installed the memory about six months ago. The BSOD problem is a recent problem, less than a month I would say.

I wonder if the problem could be related to my integrated ethernet, because come to think of it, I can't remember every having a BSOD that wasn't somehow related to an online application. Outlook Exress opening, Online games, web browsing, etc. I don't remember ever having the BSOD when I was playing a single player/offline game. I wonder if I should try using my 3Com port instead of the nVidia port?

Thanks,

zahpod

[Senor Panadero]

As a test I'd suggest changing those timings. Set Memory Timings to User Defined and change to 5-3-2-2. While you are in there, might as well change Memory Frequency to 100%, a safe adjustment.
About the net connection, sure maybe it is. I don't think it would be caused by hardware though, so don't really know where to go with it beyond that.