BSOD. This has me stumped

[Exfso]

In the last few weeks, my Windows 7pro desktop has been getting the dreaded BSOD on a pretty regular basis. Eventually the whole shebang shat itself, could not even get the damn thing to boot up. Anyway a computer tech mate of mine managed to get it up and running, he had to do a complete re install of the OS, which I thought might have been the case. He tested it, in all conceivable ways, ran the memtest for 7 iterations, and that confirmed it was not RAM. I got it back last Monday and ran like a top for 4 days, and you guessed it, BSOD again. I have nothing attached to the computer, it is basically a virgin! Only running Office, and my ESET internet security.
So I rang him and he was dumbfounded, looking like it might be the Motherboard which is around 5 yrs old, so no way is there any replacements available. So any of you gurus out there have any ideas on this one.:screwy:
Obviously any help much appreciated.:thanx:

[luka]

Did you check/swap the power supply? They are one of the main causes of instabilities and would be my first guess.

If that has been ruled out reset the BIOS (no overclocking or similar) and after that start swapping components until you find the culprit.

Memtest only checks RAM. Google PC stability testing to find some software which will test all the hardware.

But swap the power supply first.

[Exfso]

Thanks Luka, don't have a spare power supply, will mention this and the bios to my mate and get his take on it. He builds and repairs PC's for a living and has done so for 20 years. been hammering the PC today and it is ticking over nicely, bloody intermittent faults are a real pain..

[floyd_2]

I've had BSOD caused by the video card before as well. Sometimes the diagnostic report generated by Windows after a BSOD can give you an idea of the driver / component giving you the headache.

[multiweb]

Yes, 99% of BSOD are hardware or drivers related. Before changin BIOS or shuffling things around look for these.

[Exfso]

When the OS was rebuilt, all the drivers were updated, I have no cards on the system, everything is onboard as in Video, sound etc.

[multiweb]

There's got to be something wrong with the MB, RAM or any other hardware component if you did a fresh install and it still shuts down intermittently.

[Exfso]

Mark, my mate did a ramtest run with 7 iterations when he re-installed windows, and it checked out perfect. Yep, leaning toward the motherboard.

[redbeard]

Hi Peter,

As Dean mentioned, the bsod writes a log file which may help. I’ve also noticed that when the bsod is actually happening, if you look towards the bottom of the screen before it reboots, sometimes there is a reference to a hardware driver that might be causing the issues.

A windows update can also update a driver you already had loaded and cause issues.

Cheers
Damien.

[Redshift13]

Yes, BSOD generates (among other data) something called a STOP code.


The STOP Code is in the format 0x00000000, where (usually) it is the last two bits that identify the problem. For example, 0x00000050 or 0x0000007B.


If you can identify what the STOP code is for your situation, this should point you in the right direction for resolving the issue.


Also, try Googling "BSOD STOP codes" for more information.

Given that the problem is persisting even after the OS has recently been reinstalled from scratch, this does seem to indicate it is likely to be a hardware fault, rather than a corrupted software driver.


But you never know - Windows is a strange beast...


Good luck!

[multiweb]

Here's a comprehensive list (https://www.lifewire.com/blue-screen-error-codes-4065576). 50 shades of blue. :lol:

[Exfso]

Thanks chaps. the screens I photographed before the rebuild had a stop code ending in 0003B. the latest efforts did the dump before I could photograph the screen.

[multiweb]

Sounds like the MB then :sadeyes:

[redbeard]

Take the lid of your PC and remove and re insert all the cards and memory. Twice.

Check for obvious crud/dust in the process.

Do the same for all the connectors going to the devices and the motherboard.

After 5 years, something is bound to have gathered a little crud.

I have done this on numerous PC's and at least half have gone on for a few more years, the others I threw away.

The other way to look at it is 'Wow that's a nice shiny new fast computer in the shop'! :thumbsup: After 5 years I usually update to something way better. :D

Cheers,
Damien.

[JA]

It certainly would be a good idea to change out a power supply to test the fault. Also take a careful look at all the electrolytic type capacitors on the motherboard and look for any sign of bulging on top or sides, leaking or spewing out of internals.

Best
JA

[sil]

sounds power related but that might be bad caps on MB, bad PSU, even where in your home you have it plugged in to the mains voltage. Plug it in somewhere else in the house AWAY from where you have it plugged into. Homes have several power loops, not one loop covering all wall sockets or one fuse per wall socket. If you have cheap powerboards and cheap "dirty" (electrically) devices plugged in to the same loop it could be causing spikes going to your pc (any everythin else on the same loop). If you turn off the fuse that covers the powerpoint your computer is connected to, take note of powerpoints still active and move your computer to one of those and run it there. You probably have a cheap fan heater thats a high power draw and often poorly made on the same loop, or $2 powerstrip. You can test and narrow it down, but first move the computer to an entirely different power loop and run from there. If it doesnt bsod the computer is fine, your home is not. If it continues to bsod then it is in the computer somwhere. its simple process of elimination rather than jumping to false and expensive conclusions.

[Exfso]

Running through a UPS, so should be getting fine 240vac

[AndyG]

Hi Peter,

Just to entertain the whimsical... what exact mainboard model is in this PC?

Cheers, Andy.

[Exfso]

Andy it is a Gigabyte Z77MX-D3H.

[acropolite]

I recall that some time ago there was a virus or Trojans that caused BSOD. If you have a spare HDD might be worth running a fresh install to see if the problem persists.

[Exfso]

Phil, this is a clean install, was getting them before the re-install.

[AndyG]

Thanks Peter. I suspected it may be something of that vintage. I had 20+ of an Intel brand equivalent to this M/B until recently. I'd sling you one if I had any left. Sorry about the delay, I just got back to the office to check the "stock (https://ark.intel.com/products/59044/Intel-Desktop-Board-DQ77MK)". All gone though :(

[Exfso]

Thanks Andy, appreciate the thought.:)
Don't suppose you would know of any other mobo that would fit the bill and take my ram, cpu etc..clutching at straws here.

[AndyG]

Hi Peter,

It's not hard to find something that fits the bill. Here's an initial idea:
https://www.ebay.com.au/sch/i.html?_from=R40&_nkw=Z77&_sacat=0&_sop=15

All that essentially matters is that the CPU fits. The memory controller is on the CPU (which you're keeping), making them a matched pair. You can find the broader list of options by seeing what's in common with that CPU socket:

https://en.wikipedia.org/wiki/LGA_1155
(Alternative Chipsets are listed in the different Intel Chipsets in the 1st and 2nd tables). One simply needs to find a product wrapped around said chipset. As they are all old/monetarily worthless, simply aim for Q/H/Z77.

The limits of the first table may be an issue, which is easily clarified by stating which CPU you actually have - apologies if you did, my eyes are a little blurry this late :P

I'll have another look at work. $130 is too much to pay for this old junk. No offence meant - my fave home PC is exactly the same age...

[dpastern]

Try this for me - uninstall the AV software, reboot and re-test for a few weeks. I've seen AV do some very funny things over the years.

If you can, unplug the PC from mains AC, pop the sides of the case off and check the caps (capacitors) to see if they're bulging into a dome like structure on their top.

Cheap PSUs can do all sorts of weird things too, I *never* recommend going cheap when it comes to PSUs.

Another thought - reboot into safe mode and keep it like that for several weeks and see if it BSODs. if it doesn't, then it's almost certainly a driver or software related issue.

Curious to see what the outcome of this problem is.

[Exfso]

Dave, have done the uninstall of Eset, also have a pretty good power supply Thermaltake. Also looked at all the caps on the mobo, they look to be in good condition, no bulging/leaking etc. (I am a wire jerker by trade so have a pretty good idea what to look for)
The really strange thing is it was doing the BSOD before the OS was re-installed, so I doubt it is software. All the drivers are up to date. I disconnected all the external drives just to make sure it was not one of these.

[dpastern]

Thermaltake PSUs are quality, and uninstall of ESET should eliminate it from the equation. If Caps look good, then I'd say that they're good.

How many RAM slots are being used? If possible, keep it simple and just use One RAM slot/stick and see what happens.

When you did the wipe, did you just do a basic wipe of the HDD? If possible, use a bootable tool (USB stick) to do a low level format of the HDD, including the MBR, running the tool 3 or 4 times on a detailed low level format. Then reinstall Windows, and Windows ONLY and monitor for 2-3 weeks (keep it off the network/Internet during testing is my advice). I tend to not only format the drive, but run a fdisk on the /mbr.

My mate had issues with an older SSD a few years back, with BSODs and it was down to a firmware issue with said drive. It had to be sent back to base/manufacturer. I can't remember the brand.

Also, another suggestion, is to wipe (as per above paragraph to be thorough) and install GNU/Linux (my suggestion is Debian, but Ubuntu will suffice I guess). GNU/Linux is both very stable, and works well with hardware at a hardware layer, much better than MS Windows imho (sorry, but Windows is a terrible O/S) . GNU/Linux will also have many tools to help diagnose issues, including excellent logging. Yes, MS Windows logs stuff, but it's logs are truly HORRID. Having managed Linux and Windows mail/web/SQL servers for near 8 years in my last job, I know which operating system was more robust and nicer to use. Horses for courses though i guess - Linux is not everyones cuppa tea!

[Exfso]

I have 4 Ram sticks and my mate, a computer tech who has been building computers for years as his business, did the re install, so I would suggest he knows his stuff based on the amount of customers he has. My SSD is a Samsung, must admit, I am leaning toward this possibly being the culprit. I have another SSD with the operating system on it which was done via this computer a few months ago, so I will connect that one to the computer and see what happens. The BSOD reboots within 10 secs now where-as before it took about 20 mins to do the crash dump.

[dpastern]

I have seen memtest fail on detecting faulty memory in the past (admittedly, it's rare for memtest to be wrong from my experiences). Any hardware troubleshooting process with PCs is to keep it simple - i.e. the process of elimination.

I've seen people fail to spot things in the industry (20 years building my own PCs, 15 years working in IT) - for example, my last job, we had an older server that wouldn't boot. 2 other guys looked at it, and the boss, and gave up. I had a look at it and noticed popping caps on a daughterboard. Did some research, it was a known issue with this Dell server board, found a 2nd hand replacement for said daughterboard, told the boss, he went ahead and bought the 2nd hand part on my advice, we got it, swapped it out, and voila - Dell server started working again.

3 other people missed this. All of them pretty experienced and knowledgeable.

Always keep it simple. Linux is far better with hardware than any form of Windows, definitely worth testing the rig with that imho.

[luka]

If you want to use linux to test stability just download a bootable DVD or USB stick. It won't be very fast as it is running from a DVD or USB stick but there is no need to wipe your drive and install it.

[dpastern]

and if it's a SSD issue? Running Linux from a bootable DVD/USB stick is a hack solution. If you're gonna do it, do it properly. Since the machine is rather problematic, I doubt the OP has any data on it to worry about, so installing shouldn't be an issue other than time taken to do so. Most modern Linux installations are easier and quicker than Windows too...and detect all of your hardware OOTB, unlike Microsoft Windows.

[luka]

Why is it a hack solution? It will be slightly slower at worst and as you said it will detect hardware automatically. If it is an SSD issue then there will be no crashes as it won't be used. If it is issue with other hardware then there will be crashes.
Keep in mind that Windows was running nicely until recently and that a clean reinstall did not help.

And in this case it is not just the extra time of reinstalling Linux and then Windows again, Peter said that he had to ask his mate to do a install Windows for him.

Peter, if you do the test with your other SSD make sure you disable Windows updates. That may be another possible culprit.

[dpastern]

Installing on said SSD vs running a live disk is why it's a hack. Actually installing on the SSD and running the O/S from said SSD will stress the SSD from an I/O perspective.

Another thing with a live disk setup is that no log files will be written to disk, which will make checking problems more problematic.

I have a good number of years in the industry, and also troubleshooting problems for friends/family.

[Exfso]

I removed the old SSD and replaced with the new Samsung with Windows 7 Pro on it, so far so good, all working, I suppose now I have posted this bloody Murphy will intervene.:D

[DJDD]

two things:

SSD: we had a similar problem at work.
uninstalling and re-installing the Samsung Magician software resolved the problem.
we did not even need the latest version


BSOD reboots:
you can disable the automatic reboot of windows after a BSOD, which helps immensely

here's a link to instructions
https://www.lifewire.com/how-do-i-disable-automatic-restart-on-system-failure-in-windows-2626067

[Exfso]

Thanks for that, appreciated.:thumbsup:

[dpastern]

Ah, that's pretty cool, I didn't know that. That said, I don't use Windows @ home and in the work environment, reboots are a bad thing lol. another win for GNU/Linux - I can update without a reboot, including drivers. Windows, nope...



well that sounds encouraging, fingers crossed!

[Exfso]

Dave, I have everything crossed:eyepop:

[dpastern]

as do I! There's nothing more frustrating than an intermittent issue that's non-sensical in its nature! When you get a fix to one of those types of situations, it's a case for a huge party and dance imho!

[Exfso]

The Samsung 850EVO which is the dodgy one has a 5 yr warranty and I purchased it 3 years ago. So if it is in fact faulty I will get a replacement. I rang Samsung Australia today and they will obviously want it back to be tested, all at their cost.

[dpastern]

Well, let's just test further first to make sure it is the SSD. If it is, that's awesome news. I have an older Samsung 840 SSD in my currently non-working PC (long story short, had removed PSU to lend to my best mate for him to test his VR PC with to isolate an issue, got it back, had it in its box on the floor of my room, got flooded and box got soaked, PSU prolly damaged I suspect). Lucky the PC wasn't on the floor. Said flooding did ruin a bunch of model kit boxes though, including a 1/18 Merit Dauntless SBD-3 :(

[Exfso]

Sounds like Murphy was definitely at your place Dave:D

[dpastern]

;)

Did you ever hear about O'Toole's law?

Murphy was an optimist. About sums up my life in general. Still, I keep smiling.

[Exfso]

I have had the new SSD and OS going for nearly a week without any BSOD, so unless "Murphy strikes", it is was almost certainly the other SSD.

[dpastern]

That is great news! May I suggest testing for another 2-3 weeks. The problem with random problems is, that, they are random!

[dpastern]

It seems that others are/were experiencing BSOD issues with the Samsung 850 EVO SSD and Windows 10 especially...a quick google search:

https://duckduckgo.com/?q=samsung+850+evo+ssd+bsod&atb=v105-1_g&ia=web

some pages of interest:

http://www.tomshardware.com/answers/id-3025797/problem-ssd-samsung-850-evo-bsod-problem.html

https://community.spiceworks.com/topic/1194579-blue-screen-in-windows-10-with-samsung-ssd-pro-or-evo

https://www.sevenforums.com/bsod-help-support/346536-bsod-crash-new-samsung-850-pro-ssd-256-gb-install.html

some are suggesting to install the Samsung magician software with the view of updating the firmware for the SSD and/or disabling "rapid mode".

I'm curious to see if the firmware on the old SSD was ever updated, and to try disabling rapid mode to see if it helps with the BSODs. It's prolly nothing, but I always wonder if things like this will fix an issue without having to resort to spending more money!

PS I had all sorts of trouble updating the firmware on my 840 SSD and gave up on it (Samsung software continuously crashed during the update process). I did contact Samsung several times, but their support was absolutely atrocious.

[Exfso]

Thanks Dave, it certainly looks like that was the culprit, but as you said, keep monitoring. I am expecting a satchel from Samsung to return the drive for analysis.

[dpastern]

You're welcome. i'll be interested to see what Samsung says.

[Exfso]

Well bugger me after 10 days no BSOD, I got one today, so back to square one..:mad2:

[dpastern]

Damn! I had hoped that you'd nailed the problem :/

[Exfso]

Yeah, me too. I have 16gb of ram in 4 sticks, I removed the 2nd pair leaving 8gb, within 20 mins the BSOD went off. Ok removed those 2 sticks and replaced with the previous removed sticks, so far so good. Seriously this really does one's head in..:shrug:

[dpastern]

Yeah, these sort of intermittent issues are the hardest to trace. It's exceptionally frustrating!

[Exfso]

Changing the ram modules over made no difference, so it is not a ram issue. It is not a video issue as I installed a video card I had that I know is fine, and that made no difference either, only thing left is either a dodgy power supply or the mobo is at fault. If it was power supply, I reckon I would be getting these crashes a lot more regularly and probably other issues as well.

[dpastern]

Agreed on PSU, but still, anything can happen! I presume the BIOs on the motherboard is up to date?

[Exfso]

More information to muddy the waters. Looks like malwarebytes has been causing BSOD regularly getting all knotted up with Ethernet controllers. Seriously does anything work as it should...:sadeyes:

[dpastern]

yeah, as i said, I'd do a fresh install and have nothing on it but the o/s - stock o/s, no updates, disconnected from network.

edit: nothing else at all installed, no 3rd party drivers, just whatever Windows finds OOTB.

[Exfso]

Dave, I was getting BSOD with a fresh install, however I think I had installed ESET and Malwarebytes. When I googled the BSOD issue with Malwarebytes, there was a lot of people getting similar issues. As far as the bios goes, I have not re flashed it, unless my tech mate did when he did the fresh install, but I am a tad scared about flashing Bios as it can be a real pain if it stuffs up. As I said I have uninstalled MWB and touch wood no BSOD, time will tell. Thanks for all your positive input, very much appreciated.:)

[dpastern]

happy to offer help.

PS I've seen antivirus do similar things to what you're seeing. My last PC had similar issues to you, and it had me stumped. it would work, and then reboot on its own and refuse to boot the windows o/s. Wouldn't repair it either (said it couldn't, no matter what I tried). No safe mode. nada. No rhyme or reason to it - sometimes after a day or 2, sometimes after 2 weeks. I was using grub as the main bootloader, which handed off to the windows bootloader. i'd get the windows loading screen and it'd just hang there. Bloody weird.

3 months of trouble shooting and I gave up and gave the PC to my best mate for him to play with. Know what it was? Bloody well AV...I had a clean install of Windows 7 (8 had only just been released at that time), no 3rd party apps, no 3rd party drivers, full windows updates (also tested it without any windows updates, same problem). The only thing I did install was my backup data (onto an actual separate drive, so not o/s drive) and AV. I re-tested it with no backup data (after doing a low level format of the drive and fdisk'ing the mbr to be safe and paranoid) and still issues. I didn't even suspect the AV. I'd been using KAV for a good number of years without a single issue and there weren't any google posts on it (I did google it, so when nothing came back, I though nothing further of it).

What did my mate do when he got it? Uninstalled KAV, rebooted it and then tested it for 3 months. Not a single issue. He didn't tell me what he died for 3 months, just that he was still "testing" it. Then he told me and I was like 'WTH!, seriously!!!". I let him use the PC for the next 18 months lol as it was faster than his pretty decent PC and I didn't urgently need to use it as I had my MBP and an older Q6600 based PC if I really needed a PC. At that time, i had basically moved all my gaming to the PS3/XBOX, so didn't need a PC for gaming, and my MBP could run PS elements 12 which did most of what I wanted, so the PC wasn't a high priority. Of course, i got the PC back eventually. Mate needed a PSU to test his VR PC, I lent him the PSU, got it back a week later. Hadn't even reinstalled it back into the PC in question, had the PSU in its box on the floor in my room and woke up one morning at 8am to an inch of water throughout my room (heavy storm, didn't wake me up, water came down the drive, into the garage and then into my room) and has probably ruined the PSU (box was water damaged). Too heartbroken to open it and test . Lucky, I had the PC in my late dad's old vintage car up and away from water. Unluckily, mice had taken to living in the car and probably have made a nice home inside the PC and screwed the mainboard with mice pee and poo and chewed cables with my luck. Again, with my luck, I'm not game to open it lol. I'll eventually do it one day, and probably cry. Brand new PC, 2k parts, never really used, probably stuffed. *sigh*.

[Exfso]

That is a full on story, mate mine is very insignificant by comparison.:D

[dpastern]

Thought it'd give you a giggle and make you smile. :)

I'd like to say that there's a ray of sunshine at the end of every dark day, but imho, it's really a death ray from a Martian invader's spaceship ;-) (that is meant to be a jest people).

[Exfso]

Definitely confirmed that Malwarebytes is the culprit, they have put out a beta fix that appears to be working. I have not had a crash for a week. Initially I uninstalled MWB and this did the trick, so when they advised of the fix, I re-installed with the fix applied and no problems as yet. It took them 2 months to find the fix.

[dpastern]

Well, that is EXCELLENT news. I'm also glad it's not a hardware related issue. Well done Peter!

[Exfso]

Thanks Dave, yeah I am a bit stubborn, don't like things beating me, so I am like a dog with a bone. Still cost me for my tech mate to do a rebuild which in hindsight was not necessary as I can do that myself, but I thought it may have been something a bit more sinister initially, which would have been a bit above my pay grade to deal with. So if anybody here is getting BSOD and they have MWB installed, it is almost certainly the cause.
Thanks for all your help and suggestions.:thumbsup:

[dpastern]

Same here. I even have a few holes in the backyard that I've dug in my time :rofl:

[Shano592]

Good to see the end of this saga, Peter.

When I was building these boxes, I used to run this software overnight:

https://www.passmark.com.au/products/bit.htm

It stress tests all of the system components.

Something for the next drama!

[Exfso]

Thanks Shane, looks like a handy piece of software.:thumbsup: