[Fixed] power surge? lost all my devices

[Cyan]

I got my PC back online, using an old video card.
the hard drive are fine and have all my data.

I used the PC for 6 hours, and it crashed on blue screen while doing nothing special. (I wasn't using it, but maybe a program did in the background)
Maybe the RAM is the issue, I don't know.
I should check minidump? I don't know how do read it.

I looked in minidump folder, and I have one generated from the day my PC crashed (february 16) and one from now.
someone know how to read it?

edit:
found this one :
http://www.nirsoft.net/utils/blue_screen_view.html
I'll try.

first issue :
021617-25568-01.dmp 16/02/2017 09:08:23 0x00000116 fffffa80`0d8534e0 fffff880`0625eb78 ffffffff`c000009a 00000000`00000004 dxgkrnl.sys dxgkrnl.sys+5d134 x64 ntoskrnl.exe+72a40 C:\Windows\Minidump\021617-25568-01.dmp 4 15 7601 792 280 16/02/2017 09:10:06

dxgkrnl.sys dxgkrnl.sys+5d134 fffff880`064f7000 fffff880`065ec000 0x000f5000 0x539e411c 16/06/2014 01:58:04
dxgmms1.sys dxgmms1.sys+37f3c fffff880`01c5b000 fffff880`01ca1000 0x00046000 0x5164dc13 10/04/2013 04:27:15
nvlddmkm.sys nvlddmkm.sys+83ab78 fffff880`05a24000 fffff880`064f7000 0x00ad3000 0x55b0323a 23/07/2015 01:15:54

seems related to Nvidia drivers.


today issue:
030617-20389-01.dmp 06/03/2017 00:31:33 0x00000124 00000000`00000000 fffffa80`09c5b028 00000000`bf800000 00000000`00000124 hal.dll hal.dll+12a3b x64 ntoskrnl.exe+72a40 C:\Windows\Minidump\030617-20389-01.dmp 4 15 7601 293 984 06/03/2017 00:32:51

ntoskrnl.exe ntoskrnl.exe+18c9a3 fffff800`03064000 fffff800`0364c000 0x005e8000 0x5507a73c 17/03/2015 05:02:04 Microsoft® Windows® Operating System NT Kernel & System 6.1.7601.18798 (win7sp1_gdr.150316-1654) Microsoft Corporation C:\Windows\system32\ntoskrnl.exe

hal.dll hal.dll+12a3b fffff800`0301b000 fffff800`03064000 0x00049000 0x4ce7c669 20/11/2010 14:00:25


I don't know what hal.dll is or does.

 

[Madridi]

That's some great news I suppose. Did you contact an electrician an your insurance? Surely you're not going to pay for all of these again?

About the minidumps. My guess would go to the bios reset, and now that you're using a different video card, things might need to be reconfigured again.

 

[Cyan]

Yes, I thought it might be a bios issue (the thing is I never noted the previous values before reseting it)

I also changed my graphic card, but it's still a NVidia, so the drivers should be good. Maybe I should reinstall the driver to get a proper card detection ?


I searched what Hal.dll is, and it seems responsible of IRQ and HDMA (direct memory access), hardware manager (HDD).
I found it could be either an issue with the RAM, or a bad sector on HDD. I suspect it could be the BIOS with bad settings too, or the mother board bus which is also defective?

I'll search how to test my RAM.
I'll try a nvidia driver reinstall.
I'll try a chkdsk for my drives (will take a long time, if it crash while scanning, I won't know what's wrong)

I could try a Ghost reinstall of my previous backup (from july 2016), in case it's system/OS related, and not hardware.

I found an option in the BIOS named "Above 4G decoding"
with description "for 64bit compatible PCI devices", but with no explanation what it's doing.
I enabled it, because I'm on 64bit, but maybe I should disabled it ? It could be my issue, right?

It's only for PCI, so it won't affect my usable RAM (8GB).
it's only for on board memory, for example graphic card's RAM? I don't have more than 4GB GPU RAM. I can disable it, right?


I found a Asus help, saying it "should" be disabled, as graphic card MUST not use address above 4G.
I might go disabled it. no risk trying to restrict to 4G I guess.
https://www.asus.com/support/faq/1004170/
no idea what GRID or other acronyms are.

 

[sarkwalvein]

I'll search how to test my RAM.

Install memtest86 in a usb flash drive. Let your PC boot from it, and leave it testing RAM for a couple of hours.

I'll try a nvidia driver reinstall.

I will strongly suggest you doing so. I bet THIS is the problem.

I'll try a chkdsk for my drives (will take a long time, if it crash while scanning, I won't know what's wrong)

Sure, but do that last because it will take a lot of time.

 

[Jack Daniels]

@Cyan hal.dll i think is direct 3d related (hardware acceleration libary)

 

[orangy57]

If they have those wires with the three holes or whatever, try to replace those. I had a power surge, but only those wires on my monitor and computer stopped working. I must have been lucky as hell, but nothing else was burnt up.

 

[Cyan]

When I thought everything was finally fixed, it backfired at me again


So, I got PSVR back from sony (working)
I got graphic card from manufacturer as replacement. (it's said never used on the box, bu it had dust on the card. Not sure to which extend it's never used).


I decided to install the new card today.
But just before I do that, my computer decided to crash on blue screen again ! hal.dll (hardware Abstraction Layer)
I was doing some HDD actions (file moving, backup, etc.)

Since then, it crashed a lot of times, in randomness, sometime it's 2min after boot, sometime it's 30mins.
Most of the time it seems to be related to HDD access, but I cant be sure.

I decided to make a ghost of my OS (acronis), and I used a bootable disc. So I'm not using windows at all, and the computer rebooted while acronis was verifying the backup integrity.
It looks like the PC crashes, not windows....

even more reason to think it's a hardware issue.
reinstalling windows wouldn't help.

I thought "I have this issue since I'm using my old video card, GS8800, so I'll try to put my new GTX9600.
I did, had drivers issues not detected etc. (nvidia driver not providing proper files, can't extract their installer, pffff, wrong day really)
I finally managed to get old drivers, which BSOD while installing. again, random. Next reboot it properly installed the drivers.
I thought "good, now it's installed, it should work", but no, I still got few BSOD.

I tried booting into my session and not touching anything, but I guess background softwares are active (antivirus, etc.) and it crashed again 30 mins later.
I'm now trying to wait on the "insert password" screen. but if it's hardware issue, that will happen too.


I Hoped strongly that the hal.dll was a hardware issue because I used an old graphic card, but it's not
I now suspect it's a HDD issue, but I tried chkdsk and it didn't find any errors. I didn't scan all the drives for sectors though, it was long so I stopped the check to do other things more important first.


so, if its hardware, I guess it could be :
Mother board : maybe the "power surge" affected it. It's not hot (25°c)
CPU : same, or maybe a cache problem? it's not hot (30°c)
RAM : I didn't test them yet
HDD : I will try to backup all the data, between multiple BSODs sessions, until they die ?

It's really annoying, and I'm starting to think I might just buy a new computer u_u



Strange thing : the HDD LED notification on the PC is blinking a lot, even on the "insert password" screen. which app is running in the brackground ?
Windows and network, probably, but why would my hdd be very active if nothing is using them nor accessing them from network?

I don't know what to do now.
Maybe check the RAM.
I should find a USB Fash drive I could use to test RAM.

 

[kumikochan]

And where do you live. Don't you have surge protection

 

[Cyan]

I do.
That's old news.
I'm now struggling with BSOD issue
it's probably related to the surge (if it's a surge), because it first happened since that day.

Or maybe a bad contact with one of the hardware? maybe I affected something when opening the PC and touching all the components?
What is very strange is that I didn't get hal.dll bluescreen for more than a month !
I thought the issue was really related to the "Unlock 4G access" option on the bios. I disabled it and no more crash, I thought it was related.
Now I know it wasn't.

Yesterday I moved/copied a lot of files and the HDD didn't cause any crashes. why today?

Edit :
Oh, I have 3 HDDs :
- 2 HDD in raid, with my OS.
- 1 HDD sata
I could try disconnecting the non RAID one, see if it's an issue with that drive.
But seing it's random, it might not be conclusive.

Edit again :
Ahhhh, it just crashed while waiting on the "enter password" screen.

I'll now try waiting in the BIOS.
edit: No crash in bios for 1H

I might try unplugging the lone HDD, at least that's still something to test.
But I'll see tomorrow, I'm tired of today's issues.

 

[Madridi]

Well I'm glad to see you got everything back to working order besides the bsod issues

As for the bsod, it could be multiple things, but from experience, it's 90% goes back to one of these 2:
- CPU
- RAM

I assume you don't have any overclock atm and everything is at stock levels. So I would consider testing ram first with memtest. Make sure it's a thorough test. (I like to do my tests for 24h) .. If that goes well, try linx for both CPU and ram testing.

I'm also noticing that since some drivers and stuff are not detected, as annoying as it can be, if I were you I would backup my files, and start with a fresh setup (after ruling out the CPU and ram issues)

If it's none of the above, then that's beyond my expertise. So hopefully it works out

 

[V0idst4r]

this is spookier than a creepy pasta

 

[Cyan]

All the crashes appeared after 5 to 30 mins.

I tested my RAM for 5 hours (did 5 passes) and found 0 error.
I have 2 HDD in RAID1, and 1 HDD non raid.

I don't know Linx, it's a way to test the CPU ?


Now I have to test my HDDs, but it's a little more complicated as I have some RAID HDD.

When I boot the PC, I get up to windows' session login page. I don't type any password and just wait on that page.
The HDD is heavily used (maybe windows is still doing some defrag work ?)

I tried plugging the non RAID HDD alone in all sata port, and the PC didn't crash.
I tried plugging the RAID alone, and it crashes after 30 min.

My non RAID HDD also have a windows installed and had as much HDD activity on that session login page. but didn't crash.
I kept that page for 1 hour on each Sata connector.

when I boot the OS located on the RAID, it crashes.

To me it looks like either a HDD problem, or a RAID problem from the mother board's RAID controller.


I'm eager to do more test by connecting only one HDD of the RAID1, as seeing windows is doing a lot of activity on the HDD, if the files are moved around, the two HDD will not be mirror anymore and will want to rebuild when I connect both again.


I also tried booting into a UEFI autoboot disc (Acronis 2016), to make a backup of my OS on the RAID.
I could make the backup, but when it tested the integrity, it crashed again at 3/4 of the test, ruling out Windows error. it's also crashing when not in Windows.

I tried booting into Acronis without doing anything, but it's not crashing (ON for the last 2 hours). I guess it requires activity on the HDD to crash. Maybe a defective sector?
I Hoped it would crash without doing any backup work, so I could test one HDD at a time without affecting the HDD's data and not disrupting the RAID mirror.


Is there a way I could boot an OS which is NOT editing the drive when the HDD is connected?
Maybe a disc autoboot Linux? So I could backup my data ? it's working with RAID, right?
I never used linux, not sure how to use it. I'll move that option last.

On next test, I will connect the RAID on 2 other SATA on the mother board. I will also use different SATA cable/Power cable to rule out the hardware.

Next, if still crashing, I *think* I will buy 2 new HDD and make them RAID too. I'll try to restore my Acronis backup to it, to see if the problem is the RAID controller/BIOS itself or the HDD.

If it's the RAID controller, I might ask to exchange my mother board. I hope I could find the same model so I don't have to reinstall any drivers to windows.


I have a question : If it's the RAID controller, and I don't get a new mother board, if I break the 2HDD RAID1 in 2 Non-RAID HDD, will windows still be working ? I installed it with RAID drivers and I dont know if it will work with non raid system without a full reinstall..... I really want to avoid any reinstall.


Anoter option would be to go with a NAS.
I can connect my RAID1 setup in a synology NAS and it will be detected and mounted? or would the NAS ask me to format the drives to use them?

 

[Astral_]

Honestly if you had a surge which fried PS4+computer, I doubt the motherboard went through this unharmed. Crashing during BIOS is a bad omen. You should try and load the failsafe settings in BIOS. I'd replace the PSU as well... heck if it happened to me I'd replace everything before I lose my data !

 

[Cyan]

the BIOS does not crash.
When I had my issue, I couldn't see anything nor hear any BIP (it seems my mother board doesn't BIP when doing the POST test !)
I thought the issue was the BIOS and I reset it, so maybe I didn't put the correct values back.

I never took any screenshot of the old values, so maybe I did something wrong ?
It's almost everything in default value, except the RAID which I enabled.

I'm also thinking the PSU could be faulty, I tried a different one, which fried ! (BAD Luck !) I really really think it was just bad luck and not an issue with my Electricity company.

so, it still could be a PSU issue or CPU issue, but it's really strange that only the RAID HDD are crashing, it's not crashing when I'm using the non raid HDD. does it rules out the PSU/CPU out of the problem ?
I also don't think it was a power surge, To me it seems the graphic card had a glitch and fried other connected devices due to discharged electricity. but I still don't know where the issue originated (TV ? graphic card?)

 

[Astral_]

Just reread the whole thing... you replaced a 8800GTS with a 960GTX ? This is another level of power consumption, might be that the 2 extra HDDs puts the nail in the coffin.

What's the wattage of your PSU ?

 

[Cyan]

I originally had a 960 for 2 years without issue.
I used my old 8800gts as temp replacement until I got the new replaced 960 back

it's a 450W, not much but it worked for 2 years until today. I never play games so It's kind of unnecessary high end card for just watching videos and dual screen. but I don't think 1 video card + 2 HDD is too much.

Ahh, and also, it blue screen'd with the 8800 too.
It crashed once right after I reset the BIOS and plugged the card. I enabled an option in the BIOS which I thought was the issue (enable 4G access....something) and I disabled it right after the first blue screen and didn't have more for 1 month. I thought that bios setting was the issue, until last week.

It started to crash again, right before I change the video card. So it's not related to the 960


The BIOS default setting seems to be set on "high end memory usage" and "auto over clock"
But I don't want to overclock at all. is there a possibility the issue is the default over clocking ?
I can take screens of all my BIOS settings if needed. there are few options which I don't know their use.


Anyway, thanks to everyone who is trying to help me
I appreciate the fact users are helping me while I'm used to help them instead.

Spoiler: First edit

I found few tools, not sure if it's okay to test/run them :

this one works with RAID drives:
http://vm1.duckdns.org/Public/HardDiskValidator/HardDiskValidator.htm
it has a BootCD with WinPE and Mono.
I don't know these tools at all, but it might be a way to do it?

I didn't check the S.M.A.R.T. data of the HDDs
there's "HDD Health" which I found on a website.

I might try to restore the Acronis backup on a non-raid HDD and see how windows behave with Raid drivers on a non raid OS.

Spoiler: Edit sunday apr 9

I did few hdd tests:
- Another (non raid) HDD on all port : good
- Another (non raid) HDD + Another RAID : good, should rules out mother board and RAID controller.
- problematic RAID on another port, with different cables : fails, should confirm the issue is the drives and not the mother board or ports.
- Another (non raid) hdd + problematic RAID : the OS (windows) runs from the non raid drive. Seems not crashing.

On this last test, I stayed 1 hour and it didn't crash while it usually crash after 15min when the OS is running on the RAID one.
seeing it is still working, I launched HDD Validator (link just above) and it's now performing a simple "read" of all sectors.
I thought it analyzed each HDD of the RAID independently, but it found ony the raid volume not raid HDDs.
It's performing a cluster check of the raid volume. (I don't think it can do a per HDD sector check, as it's a multi HDD volume)

I don't know which information it will provide. If it can tell which HDD of the raid is bad or not.

Edit:
Result : no error found at all, no crash at all.
The cluster are fine (no cluster errors, and probably no sector errors either?)

So, I don't know what to test next.
I'll probably try to reinstall an image of my windows to a new HDD and see if Windows is the cause or not.

Spoiler: Edit April 12

I could test the HDD !

I found old previous spare HDD and decided to make them RAID1 too, just to see if the mother board had issues with any RAID or just my drives.
It turns out it's not ALL the raid, just my HDD which are bad.

I unplugged all my HDDs.
I created a new RAID1 from 2 old HDDs.
I tried to restore an image backup of my OS (using Acronis) but there weren't any partition or table so it couldn't do it.
I used Gparted booting CD to format the new RAID volume, which .... FAILED COMPLETELY.

Really, GPARTED IS BAD. I don't understand why people are recommending this tool.

It couldn't create a partition (froze when creating the partition table)
After reboot, I could add partitions, but failed in creating them physically on the drive.
It said "there are missing space on the HDD, do you want to fix it?" and doing so just destroyed the RAID volume and made 2 non raid HDD unformatted.
Gparted has issues with RAID?

So, I finally booted a different Windows OS I had on another HDD and plugged the re-created RAID volume, and Windows could format a new partition in 2 seconds !
Really, not hard to format in windows.... Linux seems to try too hard to make something easy


I then restored an OLD acronis image of my OS (current OS failed to backup, hopefully I had an old one from July 2016)

It boots, and could stay on it for hours without any crashes.
So it rules out the mother board or RAID controller. the issue is really one of my RAID HDD, now I don't know which one !


next test :
- Try and find which HDD is bad.
- Either : stop using RAID
OR
buy a new HDD to rebuild randomly, hoping I kept the correct HDD to rebuild with.
I could always reformat the new one and rebuild using the second HDD as source if the first source is the problematic drive.
OR
I keep the new test RAID as my main OS, and keep my failed raid as secondary partitions, and not access the first partition (which seems to be the one which have issues). Other partitions should work, except if the bad sector is shared among multiple plates.

- Try to restore my OS on a non RAID hdd. see if windows will still work as it was installed with RAID drivers.
But it should work, I could use windows on a non raid partition but still connect RAID hdd as secondary sata.

- I will try to break the RAID volume of the test drive to see if my data are still available and the drive can be used as non raid (it's a mirror with no parity, all the data is on each drive).

Spoiler: April 12, next

As next test, I tried to "break the RAID".
I did it, and the BIOS warned me all data would be lost.
I did it anyway, and surprisingly, the data were still there.

I booted with both drives connected, and windows started ! (from one drive)
Once in windows, I noticed I didn't have all the hdd recognized, only the one I booted from.

1min later, windows showed me a message "driver installed for SCSI...." etc.
I thought, it's ok, now the other drive should be detected.
I rebooted, and was greeted with the recovery menu because windows supposedly didn't shutdown correctly, while it did.
I chose "boot normaly" and got a blue screen right away.
So, having a broken RAID HDDs system is not a good idea.


Next step:
I kept only one drive, and restored my OS backup again, this time without any RAID.
windows worked ! (and still does)
I had a message for "installing new drivers" and "you need to reboot to use this !" (whatever "this" could be? I don't know).
I thought "it will break again, it detected I'm not on RAID system anymore and will crash to BScreen again?"
But no, it seems to work so far.

I'm copying my data from my non-RAID HDD to it, so I'll use it as main HDD and will plug my previous OS RAID HDD as slave.

I will leave the PC alone for few hours to be sure everything is stable again before reinstalling/recovering my programs, setting internet back, etc.

Conclusion :
The bluescreen seems to be caused by a bad HDD in my RAID.
The RAID system is supposed to help and save you from losing your data when 1 HDD is dying by hot-switching the bad one, and on my previous system it worked as intended.
This time (this mother board? windows 7?) it's not doing its job and telling me a HDD is dead and needs to be replaced, but instead just crash on blue screen ! (probably causing even more harms to the drives each time it hard resets and stop write access savagely)

I think I will stop using RAID for my OS.
I'll probably continue using RAID for my data, it "SHOULD" not crash on blue screen if the problematic partition is not the current running OS partition.
I'll buy myself a Synology NAS running a real hardware RAID 5 controller now for all my storage needs.
Maybe the bluescreen is not related to Raid but to bad sector and it would still happen on a non raid system, but I still need more tests.

 

[Cyan]

Hello everyone,

I'm posting here to tell you I finally fixed my computer's issue.
all these blue screen crash every 15min was due to a bad BIOS setting !

I spent more than a month wondering how to fix it, I even swapped a lot of HDD, RAID, hardware, and tried different OS to narrow the issue to the bios
I could have done it earlier....

When I reset the BIOS 2 months ago, it asked me which type of fan I have on my CPU, and based on my answer it decided to enable Overclocking with some random values.
It set the based clock to 102 instead of 100. Just a small difference made all the difference. There weren't any choice for "no overclock" at all, it was "XMP extra memory [default]" or "manual". I thought the "default" was really that and it managed overclock itself. it appears it wasn't the real default, and another real "default" exists when resetting the bios and not selecting the auto-setup based on hardware.


Now it's fixed, and I have a backup of all my data (lesson learned, for now).
I have a month of threads, gaming and scene news to catch up.
It's nice to be back.

 

[Dionicio3]

Hello everyone,

I'm posting here to tell you I finally fixed my computer's issue.
all these blue screen crash every 15min was due to a bad BIOS setting !

I spent more than a month wondering how to fix it, I even swapped a lot of HDD, RAID, hardware, and tried different OS to narrow the issue to the bios
I could have done it earlier....

When I reset the BIOS 2 months ago, it asked me which type of fan I have on my CPU, and based on my answer it decided to enable Overclocking with some random values.
It set the based clock to 102 instead of 100. Just a small difference made all the difference.


Now it's fixed, and I have a backup of all my data (lesson learned, for now).
I have a month of threads, gaming and scene news to catch up.
It's nice to be back.

Nice man, welcome back

 

[orangy57]

Hello everyone,

I'm posting here to tell you I finally fixed my computer's issue.
all these blue screen crash every 15min was due to a bad BIOS setting !

I spent more than a month wondering how to fix it, I even swapped a lot of HDD, RAID, hardware, and tried different OS to narrow the issue to the bios
I could have done it earlier....

When I reset the BIOS 2 months ago, it asked me which type of fan I have on my CPU, and based on my answer it decided to enable Overclocking with some random values.
It set the based clock to 102 instead of 100. Just a small difference made all the difference. There weren't any choice for "no overclock" at all, it was "XMP extra memory [default]" or "manual". I thought the "default" was really that and it managed overclock itself. it appears it wasn't the real default, and another real "default" exists when resetting the bios and not selecting the auto-setup based on hardware.


Now it's fixed, and I have a backup of all my data (lesson learned, for now).
I have a month of threads, gaming and scene news to catch up.
It's nice to be back.

nice job fixing the issue. My only confusion is that it took you a month to think to reset the BIOS? It's usually like two physical pins yo connect and the BIOS resets itself for you, and you didn't think of ever trying that?

well the time's passed already now

 

[Cyan]

nice job fixing the issue. My only confusion is that it took you a month to think to reset the BIOS? It's usually like two physical pins yo connect and the BIOS resets itself for you, and you didn't think of ever trying that?

well the time's passed already now

That's the first thing I did, and what caused the issue in the first place.
I didn't thought to re-reset it and this time don't let it auto-setup the options.