I have already swapped the RAM, loosened timings, underclocked the RAM, reinstalled XP from scratch, and swapped the MOBO.

Xp sp2, e6750 2.66, 2 gbs crucial ballistix pc-8500, antec sonata III w/500Watt, galaxy geforce 8800GT 512, Gigabyte GA-P35-DS3L.

I can run memtest no prob, run orthos burn in for 5 hours.
have used 1 stick of ram in all different ram slots,

here are my 2 most recent minidumps since swapping the MOBO (the last step i have taken) and a few from before my MOBO swap.

Any help appreciated

Attached Files

	Mini020608-01.dmp (88.0 KB, 4 views)
	Mini020608-02.dmp (96.0 KB, 2 views)
	Mini020508-02.dmp (96.0 KB, 2 views)
	Mini020508-01.dmp (96.0 KB, 2 views)
	Mini020108-01.dmp (88.0 KB, 2 views)

From what I have been reading your minidumps are pointing to a software/driver issue, not a hardware issue. However, 4 of your 5 dumps kept repeating that the correct symbols could not load. I've never experienced this before in trying to read dumps, thus it was hard to to find a definitive culprit or culprits causing your crashes.

Anyway, your first dump pointed to Rtenicxps.sys which a driver for RealTeck. Have updated your audio chipset drivers recently? If not, give that a shot.

One of your newer ones points to eamon.sys which is of NOD32 anti-virus. (I run and love it) Question: What version are you running, 2.7 or 3.0?

Just a few suggetions:

1. Run chkdsk and a full harddrive diiagnostics.

2. Scan for viruses, malware, rootkits, etc.

3. Update your audio chipset drivers.

I run nod32 3.0 and have the latest realtek drivers from realtek direct not the gigabyte ones I was running before reinstalling windows. Like I said besides the 2 newest dumps the others were from a previous install of windows. The 2 newest are from a complete full reformat and install.

Since my last post I have had another BSOD this time a page fault error. I know that the majority of my errors have been related to memory. page fault, irql not equal or less than, and memory_management. But have replaced both my board and RAM. What are the odds of bad memory on a faulty CPU.

Anyways here is the latest dump.

update:
chkdsk run no errors, AV run nothing, Adaware run nothing, and double checked I have newest driver for audio.

Attached Files

Mini020608-03.dmp (96.0 KB, 2 views)

Once again your minidump says that it could not load the proper symbols and once again it points to ntoskrnl.exe which is a critical part of your Windows OS. I am wondering if it is corrupted.

I also wonder if there are bad/corrupted sectors on the harddrive itself. You've already changed out quite a few items but are getting BSODs regardless. Have you run an HD diagnostics since the new mobo install?

I thought it might be your Crucial Ballistix because their DDR2-800 versions are being RMA'd big time; just like 2 years ago when Crucial was forced to yank their Ballistix PC3200 and especially their PC4000 off the market. I was one of the victims. But I haven't read on the net that the Ballistix pc-8500 is a problem.

By the way, for your own information, Adaware no longer updates their free version; haven't in a month.

Yeah I know I have Adaware Pro. I will give a HD scan a shot and let you know what I come up with.

No problems on HD ran through multiple types of tests error free.

What are the odds of the BSODs I have been getting being related to
Vid Card, CPU, or PSU? I have no other peripheral cards.

If I boot into safe mode would that take the Video Card out of the equation?

Have you attempted to do a Windows repair?

What is your PSU make and capacity?

If you haven't tried already, a) download the latest driver executible to your desktop, b) do a complete clean uninstall your current drivers, c) reboot and when Windows Install attempts to load its own drivers stop it, let it load back to your desktop screen, and then load the new drivers by clicking on the executible.

It would be interesting, as you suggest, to see if you get crashes in safe mode.

I had a similar problem and was getting ready to pull my hair out, when I figured out that my psu was not putting out enough power, then after I replaced the psu all of my problems went away

I loosened my timing on my RAM to 5,5,5,18 even though it should be able to do 4,5,5,12. I am running the RAM at 333 x 2.4 for 800 even though it can do 1066.
My CPU is now set to 8 x 333 for 2.66 and have since not had a BSOD in 24hours, knock on wood.

I did a repair originally and when that did not work did a full reformat and reinstall to no avail.

I have a 500 Watt Earthwatts PSU that came with my Antec Sonata III case.

I'll see what happens here and if I get another BSOD will try running in safe mode for a few days to see what happens.

I have read some good things on the Antec Earthwatts psu. It seems it just might be an overclocking issue after all. If I knew you had overclocked I would have suggested that you scale back which you had done.

By all means, let us know if you gain stability in normal mode since lower timings on RAM and edging back on your cpu.

It wasn't overclocked 1066 was the default clock speed. I am now underclocking from default. And have the timings lower than the SPD data I get for them.

Anyways was good for about a day and a half and then another BSOD here is my minidump.

The error was driver_irql_not_less_or_equal relating to tcpip.sys.

I have read that this is definitely a memory hardware error, what do you think, the only thing left to swap are the cpu, psu and gpu. what are the odds its the L2 cache on the cpu?

Attached Files

Mini020808-01.dmp (96.0 KB, 4 views)

Something is trying to access an IRQ level that it shouldn't but what that is I cannot tell. Your symbols will not load in your mindumps -- again.

Have you done a Windows repair?

Is there any way for you to test that RAM in another PC?

Yes I have done a windows repair but I will do another one.

I can probably test my RAM in another computer, yeah I don't know what the deal with the unreadable symbols is. I tried using the debugging tool before posting here the first time but just thought I was using the program wrong.

did a repair on windows, still no help.

checked RAM in another PC no BSOD, ran memtest no errors.

I reran memtest on my rig and got thousands or errors on test 2, then I can run memtest later and get no errors on multiple passes.

Can memtest detect errors in the memory of the CPU?

Attached Files

Mini020908-01.dmp (96.0 KB, 2 views)

Hope you don't mind me pasting this in.

Inside the new minidump file:

BugCheck 10000050, {e085c818, 0, 805c1053, 0}

***** Kernel symbols are WRONG. Please fix symbols to do analysis.

*************************************************************************
*** ***
*** ***
*** Your debugger is not using the correct symbols ***
*** ***
*** In order for this command to work properly, your symbol path ***
*** must point to .pdb files that have full type information. ***
*** ***
*** Certain .pdb files (such as the public OS symbols) do not ***
*** contain the required information. Contact the group that ***
*** provided you with these symbols if you need this command to ***
*** work. ***
*** ***
*** Type referenced: nt!_KPRCB ***
*** ***
*************************************************************************
*************************************************************************
*** ***
*** ***
*** Your debugger is not using the correct symbols ***
*** ***
*** In order for this command to work properly, your symbol path ***
*** must point to .pdb files that have full type information. ***
*** ***
*** Certain .pdb files (such as the public OS symbols) do not ***
*** contain the required information. Contact the group that ***
*** provided you with these symbols if you need this command to ***
*** work. ***
*** ***
*** Type referenced: nt!KPRCB ***
*** ***
*************************************************************************
*************************************************************************
*** ***
*** ***
*** Your debugger is not using the correct symbols ***
*** ***
*** In order for this command to work properly, your symbol path ***
*** must point to .pdb files that have full type information. ***
*** ***
*** Certain .pdb files (such as the public OS symbols) do not ***
*** contain the required information. Contact the group that ***
*** provided you with these symbols if you need this command to ***
*** work. ***
*** ***
*** Type referenced: nt!_KPRCB ***
*** ***
*************************************************************************
*************************************************************************
*** ***
*** ***
*** Your debugger is not using the correct symbols ***
*** ***
*** In order for this command to work properly, your symbol path ***
*** must point to .pdb files that have full type information. ***
*** ***
*** Certain .pdb files (such as the public OS symbols) do not ***
*** contain the required information. Contact the group that ***
*** provided you with these symbols if you need this command to ***
*** work. ***
*** ***
*** Type referenced: nt!KPRCB ***
*** ***
*************************************************************************
*************************************************************************
*** ***
*** ***
*** Your debugger is not using the correct symbols ***
*** ***
*** In order for this command to work properly, your symbol path ***
*** must point to .pdb files that have full type information. ***
*** ***
*** Certain .pdb files (such as the public OS symbols) do not ***
*** contain the required information. Contact the group that ***
*** provided you with these symbols if you need this command to ***
*** work. ***
*** ***
*** Type referenced: nt!_KPRCB ***
*** ***
*************************************************************************
*************************************************************************
*** ***
*** ***
*** Your debugger is not using the correct symbols ***
*** ***
*** In order for this command to work properly, your symbol path ***
*** must point to .pdb files that have full type information. ***
*** ***
*** Certain .pdb files (such as the public OS symbols) do not ***
*** contain the required information. Contact the group that ***
*** provided you with these symbols if you need this command to ***
*** work. ***
*** ***
*** Type referenced: nt!_KPRCB ***
*** ***
*************************************************************************
*************************************************************************
*** ***
*** ***
*** Your debugger is not using the correct symbols ***
*** ***
*** In order for this command to work properly, your symbol path ***
*** must point to .pdb files that have full type information. ***
*** ***
*** Certain .pdb files (such as the public OS symbols) do not ***
*** contain the required information. Contact the group that ***
*** provided you with these symbols if you need this command to ***
*** work. ***
*** ***
*** Type referenced: nt!_KPRCB ***
*** ***
*************************************************************************


Probably caused by : ntoskrnl.exe ( nt+ea053 )

Followup: MachineOwner

kimsland, he has the same readouts in his other minidumps as well. He has said he has done a repair as well as a full reinstall? What do you suggest? This sure is a stumper...

Mod1Fy, have you attempted another repair?

Your debugger is not using the correct symbols
Can be safely ignored

According to WinDebug ntoskrnl.exe at fault

Can be related to faulty hardware: Harddrive or Keyboard

First option is to Repair Windows including boot.ini (done)

Run Check Disk (done)

Optimize: This is a little vague, even if you run CCleaner and Defrag so forth, I still feel this may be more related to Hard Drive issues.

Virus (done)

Partition: You could try downloading Gparted and check the partition for errors

Windows repair ran yet again, and partitions checked for errors, chkdsk ran again as well for good measure.

No crashes yet but I well let you know as soon as I have one which I am sure will happen (as you can see I am real optomistic) I am going to swap something else tomorrow what do you think, psu, gpu or cpu first?

if it's out of those 3 only PSU

Got another BSOD can't upload minidump as I as from another system and have taken my into the place I bought all the parts so they can further diagnose it.

But the last BSOD was related to rtenicxp.sys which has to do with my realtek network adapter onboard, installed newest driver for it and then got another BSOD just a general page_fault.

I don't know if this is of importance or not but I only get errors on memtest after a BSOD. If I reboot properly and run it, no errors are detected on multiple attempts, at 10 hours of run time and most of the errors are on test #2 and sometimes test #1. Is there any way to only test the CPU or GPU memory and not the RAM?

Oh and I also ran a program called S&M which also tests, memory, PSU, and CPU and it passed with no errors.