Solved Rebuilt PC with random BSOD occurrences

[Useless User]

Dear forum,

after a Windows and Driver update in December (the PC had not been in use for six months, switched off, not moved or tampered with), continuous BSODs ensued, the update also never fully worked. In short: Last week, I updated the BIOS for the first time and re-installed Windows (re-formatting the disks) and drivers from the original DVDs from January 2012 (PC part purchases and build date). I have not updated anything since last week, apart from Windows' own updates. Now, I observe BSODs at random intervals, no matter if I run a software or just switch the PC on and let it sit idle. All I can see is that sometimes, but not always, there are only 12 instead of 16 GB memory installed. I swapped the sticks around and also cleaned the interior of the PC very carefully, to no avail. I have attached the recommended ZIP-file to my post.

Any help is much appreciated!

ps: While typing this post, Windows updated. While it installed updates, another BSOD 0x0000000a IRQL_NOT_LESS_OR_EQUAL raised its ugly head. I have not included the new dump file of that one, because this error was among the three contained in the ZIP file originally attached.

pps: Now nearly continuous BSODs 0x0000000a (3x) or 0x0000001e (8x) and 0x00000124 or 0x00000024. Windows updates eventually completed in between BSODs, also with only one RAM bar, default BIOS, CPU at 38°C, not OC'ed.

ppps: While in safe mode (CoreTemp won't run in safe mode), at stock speed, etc. as mentioned above, running prime95, one worker failed after one hour.

 

[Useless User]

In safe mode, I can't see, if the CPU or single RAM stick overheats, for example. Maybe the ZIP file I provided contains some clues?

I am quite surprised by all this mess, because I re-installed Windows 7 and all drivers from the original 2011 DVDs with no hardware change, so I (naively) expected everything would be fine. Did having the PC switched off for half a year after two years 24/7 operation without a single problem do harm to it in some way?

 

[Useless User]

Anyone has any idea what I could do?

ps: again, while PC is still in safe mode as before, worker #1 failed after two hours. Still only one RAM stick used. New test, new RAM stick in a different slot. Latest ZIP-file attached, too.

pps: again, worker #1 failed after one hour and 20 minutes, with one RAM stick in a different slot. At least, today while in safe mode since this morning, no BSOD. What gives?

 

[Useless User]

Is the motherboard damaged? A memory slot? All memory slots? A RAM-stick? All RAM-sticks? The CPU? In safe mode, it can't be a driver issue, right? How on Earth could I find out which hardware component needs to be bought?

 

[Golden]

*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

Use !analyze -v to get detailed debugging information.

BugCheck 1E, {ffffffffc0000095, fffff80002f3bea8, 1, fffffa800fd23010}

*** WARNING: Unable to verify timestamp for [COLOR="Red"]MpFilter.sys[/COLOR]
*** ERROR: Module load completed but symbols could not be loaded for MpFilter.sys
Probably caused by : [COLOR="Red"]luafv.sys[/COLOR] ( luafv!LuafvNormalizeNameComponentEx+162 )

Followup: MachineOwner
---------

http://carrona.org/drivers/search.php?id=luafv.sys
http://carrona.org/drivers/search.php?id=MpFilter.sys
Perform a Windows Update - install all available updates, including the optional ones.

*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

Use !analyze -v to get detailed debugging information.

BugCheck 1E, {0, 0, 0, 0}
Probably caused by : [COLOR="Red"]intelppm.sys[/COLOR] ( intelppm!MWaitIdle+1b )

Followup: MachineOwner
---------

Check for updates for the Intel Processor driver by following this link:
Driver Reference Table - intelppm.sys

Afterwards, perform a stress test on the CPU using the FULL FUNCTIONAL test as described here. Upload the testresults.txt file here.

http://www.sevenforums.com/tutorials/322483-intel-cpu-diagnose.html

Next perform a thorugh test of your RAM with all RAM DiMM's inserted into the motherboard. You must run MemTest86+ for a minimum of 8 passes. If you run less passes, then the RAM will not be adequately tested.

http://www.sevenforums.com/tutorials/105647-ram-test-memtest86.html

 

[Useless User]

Thanks; I'll do that right now (hope Windows does not BSOD on me while updating and driver searching). Maybe all problems are a result from me reinstalling Windows and drivers from the original (2011) Microsoft and ASUS DVDs - but Windows being updated after reinstall, so that the drivers supplied on the ASUS DVD are malfunctioning, because they are outdated/incompatible with respect to a current Windows 7 installation...

ps: the last safe mode prime95 run over night has yielded the same result as before; after one and a half hours, worker #1 failed.

pps: after much BSODing, I managed to update Windows 7 (bar IE10 - the update always fails with Code 9C48). I then followed your link to https://downloadcenter.intel.com/Default.aspx and ran the Intel Driver Update Utility, which showed me that the currently installed Intel Chipset Software Installation Utility (Chipset INF) is valid (version 9.2.3.1020), Wired Networking needs updating (version 11.13.51.0 to 18.5.54.0) and that my Graphics Driver, Audio Driver for Intel Desktop Board, Wireless Networking (Wifi) are unknown and unsupported (so I have to look for them manually).

ppps: the IPDT test fails after half a minute with "IPDT has not fully completed Testing. This is possibly due to either the user closing Test prior to all test Diagnostics completing or All IPDT diagnostics did not complete". Only the initial steps after starting IPDT described under http://www.sevenforums.com/tutorials/322483-intel-cpu-diagnose.html are performed up to "Executing CPU Load ..." and the TestResults.txt is zero bytes : ( I have copied what's being shown in IPDT's grey prompt window and attached it here.

 

[Useless User]

After a restart, IPDT failed again without TestRexults.txt - next try, it did work and completed all tests; see file attached. Odd.

ps: now, with all four RAM DIMMs in the board, I managed to get Memtest86+ 5.01 running off a USB key after many BSODs and let it run for half a day now.

pps: ~5 hours, 2 passes, no error so far.

 

[Useless User]

Still no errors with the memory after ~10 hours...

In the meantime, I read at Anandtech that processors degrade over time when overclocked, even with built-in OC settings as in the case of ASUS. It was suggested that, when always the same worker fails on multiple processors, one could try to raise VCORE by 0,05 or low multiples of that. The gist was that the more a processor is abused over years, the higher the voltage necessary to run it stable again before it finally fails. Maybe that could be the case here, if the RAM throws no error...

 

[Golden]

After MemTest86+ has completed its 8 passes, run this:

http://www.sevenforums.com/tutorials/220743-cpu-stress-test-using-intelburntest.html

 

[Useless User]

Cheers, I do that then... Intel website reveals Tcase to be 66,8°C max. which I then monitor over time. Can I use Core Temp 1.0 RC6, too, which has a logging function? I see from the nearly two years constant operation in ASUS Optimal mode that the temperature was always around 63°C-73°C (I used Core Temp to log this).

 

[Golden]

That might work.....I use HWInfo. Its up to you what you think is the best to use.

 

[Useless User]

Ok, I ran the Intel Burn Test in standard, high and very high mode. No error. See the temperature results in the attached screenshot.

 

[Useless User]

Works well now for over one day. A forum member gave me a simple but probably correct tip: all I had to do was to raise VCORE in the BIOS by 0,050 V - problem solved. No worker fails and Windows 7 works.

 

[Useless User]

Premature enjoyment, the PC is dead (for now). While it was on, everything worked fine: restart, hibernate, sleep, wake on LAN. This is what I did, in that order:

shut down
detached monitor, ethernet, mouse and keyboard
switched off PSU
detached power cable
screwed on side of the case
carefully carried the PC back to its place
attached power cable
switched on PSU
attached monitor, ethernet, mouse and keyboard
pressed "on" button on PC

Result: black screen, no POST, VGA_LED constantly lit, Q-Code 9A. When switched off by pressing the power button for four seconds, both USB keyboard LEDs remain lit. Now mostly Q-Code 9A (USB initialisation started), sometimes 0E (microcode not found). No change when monitor, ethernet, mouse and keyboard are connected or not. Same result. Occasionally, when using the reset button, I get Q-Code B2 (legacy option ROM initialisation) as an "alternative" with a blinking cursor top left.

What is going on? I am very puzzled. Obviously, the switching off and unplugging of the power cord did something. What can I do now?

 

[Useless User]

Ok. Tried another new disk with OS: 9A. Tried the graphics card in another PC: 9A. Tried again swapping RAM and slots: 9A. I am right back to where the whole mess started in December after reactivating the PCs - but even worse.

To me, this looks like a new fatherboard for each and maybe even new silicon. Still - how can these things go bad in the first place, if not used and disconnected from the mains for half a year? Technology is bad to me : (

 

[Useless User]

Before I order new cousinboards - anyone has a last ditch idea what the constant 9A Q-Code could mean? I seriously hope it is the daughterboards, otherwise, the last option then would be to buy new silicon - ouch!

 

[Useless User]

Back to square one, once again: I finally found a RAM DIMM from another PC which I fitted like so: | void | void | void | void | CPU | void | 4GB | void | void | With that, I get to the ASUS EZ-Mode BIOS setup screen again, select the SSD, and boot.

But now: what to do? Should I go to the Kingston website and select new 4x4GB or 2x8GB memory from their list pertaining to the ASUS P9X79, order and fit that? ASUS' own RAM QVL is still from 2011. Or - is there something else apart from getting new RAM I should try first? Test the CPU once again with the Intel Burn Test? Run prime95 for a day?

What I also don't understand is that the VGA_LED lights up and 9A shows some USB status - rather, with after two years now faulty RAM, I would expect the DRAM_LED to light up? Very confusing for non-technical users, these motherboards.

 

[Golden]

I thought you said you had run MemTest86+ and it passed......why are you swapping RAM modules?

 

[Useless User]

Well, yes, memtest86+ found no errors after many passes. After everything worked but then (after removing power to re-assemble the PC and put it back) got stuck at Q-Code 9A with the VGA_LED constantly on, no matter what I tried (swapping graphics cards, PSUs, HDDs with OS, Linux USB-sticks), I did not want to order new motherboards and processors for the PCs just yet, that would be very costly, compared to buying new RAM, for example.

So, I took another PC out of service, removed one of its obviously functioning DIMMs, and tried, if the bricked PCs can go at least as far as the BIOS and maybe even boot. And that worked. So, now I am back to square one - but I now know that the graphics card, PSU or HDD are likely not the culprit (because they worked in other PCs).

To me, it looks like I either buy new RAM as recommended by Kingston for the P9X79, or I find out what the "real" problem is, or I buy new motherboards, or I buy new CPUs ... unless you have other suggestions from your experience; I am happy to try other things. I am a regular PC user, not a pro, after all.

 

[Useless User]

I bought new Jamaican RAM. Cheapest option of the three. Let's see how that goes.