03-25-2015, 10:34 AM
03-25-2015, 10:55 AM
When my PC wouldn't boot, I called in a technician because I have kids 
. He opened the case, desloted EVERYTHING and dusted with a brush all connectors, sockets and cards, slotted everything in and tried to boot. Only after he repeated the process three times(!) the computer booted successfully. Something to bear in mind.
. He opened the case, desloted EVERYTHING and dusted with a brush all connectors, sockets and cards, slotted everything in and tried to boot. Only after he repeated the process three times(!) the computer booted successfully. Something to bear in mind.03-25-2015, 11:12 AM
Well, I'm back at square one, black screen, no audio, fucking event viewer gives nothing but "The previous system shutdown at 9:51:46 on ?25.?3.?2015 was unexpected.", all other errors are from after it starts up Win again.
03-25-2015, 11:32 AM
Damn, at this point I'd call a technician. Worth every penny when considering the headache and the stress. Is that a possibility?
03-25-2015, 01:12 PM
I had a quick search over Tom's Hardware, and found this http://www.tomshardware.co.uk/forum/3078...d-computer
Basically, the guy had his motherboard replaced.
If the Mobo has a hairline fracture, it might work until the heat build-up is sufficient to deform the boar, resulting in a severed trace on the board. Or, the seating of the chips on the board might not be good anymore. Usually, you can remedy this by manually pushing them in.
A couple years ago (well actually around 25 or so) a friend of mine had an Amiga 500 that would not run stable. He brought it to the dealer, and, I kid you not, they lifted the machine up 20 cm over the table and dropped it. It worked afterwards. (DON'T TRY THIS AT HOME). The drop had reseated the socketed chips on the board.
Basically, the guy had his motherboard replaced.
If the Mobo has a hairline fracture, it might work until the heat build-up is sufficient to deform the boar, resulting in a severed trace on the board. Or, the seating of the chips on the board might not be good anymore. Usually, you can remedy this by manually pushing them in.
A couple years ago (well actually around 25 or so) a friend of mine had an Amiga 500 that would not run stable. He brought it to the dealer, and, I kid you not, they lifted the machine up 20 cm over the table and dropped it. It worked afterwards. (DON'T TRY THIS AT HOME). The drop had reseated the socketed chips on the board.
03-25-2015, 01:54 PM
But if it's mobo, why doesn't it crash when I'm not playing A3/Ins?
Today I've tried connecting the monitor to the mobo when it crashed in hopes something would happen and running A3 and Insurgency windowed. Neither crashed, then again I only have 4 hours of free time during workdays to spend at this, I wish the damn crash wouldn't take so long.
Today I've tried connecting the monitor to the mobo when it crashed in hopes something would happen and running A3 and Insurgency windowed. Neither crashed, then again I only have 4 hours of free time during workdays to spend at this, I wish the damn crash wouldn't take so long.
03-25-2015, 03:58 PM
Well yeah it points at the graphics card. At this point I would say only a professional can help, or trying to procure another graphics card and testing with that.
Could still be the PCIe slot that is damaged, but that is rather unlikely.
Could still be the PCIe slot that is damaged, but that is rather unlikely.
03-25-2015, 07:00 PM
It might be worth seeking help here http://www.msfn.org/board/forum/5-hardware-hangout/
I've found help there in the past and if there's any logs or memory dumps that Windows creates, they should be able to help you find and analyse them for clues.
I know you've got a new PSU but as you're only crashing when gaming, I wonder if the GPU isn't getting enough power when loaded. If it uses a PCI-E connector for extra power and you've got access to a voltmeter you could check that there's 12v on it (with it disconnected from the GPU).
It might be that Alwarren is right about there being a fault with the motherboard that only shows up under load/heat and maybe that's affecting the power supplied to the GPU through the slot and/or the data lines. You can only really rule that out by trying another GPU with similar power draw, or putting the GPU on another motherboard.
I know how annoying this must be for you. I went through 3 motherboards before I realised it was the PSU that was faulty, as it worked fine with another board I had, which obviously didn't load the PSU enough to trigger the fault.
I've found help there in the past and if there's any logs or memory dumps that Windows creates, they should be able to help you find and analyse them for clues.
I know you've got a new PSU but as you're only crashing when gaming, I wonder if the GPU isn't getting enough power when loaded. If it uses a PCI-E connector for extra power and you've got access to a voltmeter you could check that there's 12v on it (with it disconnected from the GPU).
It might be that Alwarren is right about there being a fault with the motherboard that only shows up under load/heat and maybe that's affecting the power supplied to the GPU through the slot and/or the data lines. You can only really rule that out by trying another GPU with similar power draw, or putting the GPU on another motherboard.
I know how annoying this must be for you. I went through 3 motherboards before I realised it was the PSU that was faulty, as it worked fine with another board I had, which obviously didn't load the PSU enough to trigger the fault.
03-25-2015, 11:13 PM
(03-24-2015, 11:04 AM)Outlawz7 link Wrote:Try this tweak [size=1em]KBOOST[/size]Played ArmA3 for an hour with Chrome running. Last thing I did was uninstall Intel GPU drivers and this:
https://forums.geforce.com/default/topic...oblem-fix/
. :o ;D
edit: spoke too soon, still crashed. Removed the GPU now
Quote:Use "regedit", search after Card ID e.g "4D36E968-E325-11CE-BFC1-08002BE10318"http://forums.guru3d.com/showpost.php?p=...stcount=93
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Contro l\Video\{*gfx card ID*}\0000 or with IGPU 0001
Add:
- "PowerMizerEnable"=dword:00000001 [1 for on, 0 for off]
- "PerfLevelSrc"=dword:00002222 [or 3322 for different battery and AC settings] ... 22 is for fixed setting and 33 for adaptive
- "PowerMizerLevel"=dword:00000001 [1 = high, 2 = med, 3 =low] ... this is the setting for when using battery
- "PowerMizerLevelAC"=dword:00000001 [1 = high, 2 = med, 3 =low] ... this is the setting for when using AC
https://forums.robertsspaceindustries.co...x-nvlddmkm
03-26-2015, 12:31 AM
I've got around to running OCCT, it ran for 30 mins, caused a black screen, produced no logs, probably because of reset.
03-26-2015, 09:04 PM
Doesn't OCCT just load the CPU, not the GPU? If so, that points to a CPU / motherboard / RAM problem.
You can try HWinfo as it will let you log your voltages, etc to a CSV file so you'll be able to check what the last values before crashing were.
http://www.hwinfo.com/download.php
You can try HWinfo as it will let you log your voltages, etc to a CSV file so you'll be able to check what the last values before crashing were.
http://www.hwinfo.com/download.php
03-27-2015, 10:58 AM
You can do both CPU and GPU and I have http://openhardwaremonitor.org/ for logging.
Also I played Insurgency in windowed and it froze with a loading icon cursor after a while, not sure if that's the black screen or sth else, no damn errors as always.
edit: looks like Ins crashes at the end of training, not black screen
edit2: Ins also black screened in windowed, although if the following is supposed to work, I had power management mode set to Adaptive since I reinstalled nV drivers.
https://forums.geforce.com/default/topic...oblem-fix/
Also I played Insurgency in windowed and it froze with a loading icon cursor after a while, not sure if that's the black screen or sth else, no damn errors as always.
edit: looks like Ins crashes at the end of training, not black screen
edit2: Ins also black screened in windowed, although if the following is supposed to work, I had power management mode set to Adaptive since I reinstalled nV drivers.
https://forums.geforce.com/default/topic...oblem-fix/
03-28-2015, 03:49 AM
I still think you should test RAM because of crashing and Windows as seen in event viewer not catching the error, that looks like memory issue although it can be something else. Start RAM testing program and leave it while you are at work for example and if it goes without errors you can at least write off RAM as an issue.
03-28-2015, 06:43 AM
I agree. Boot into memtest, which will take Windows and any other software out of the equation and run it overnight or something http://www.memtest.org/ If it fails, it might be something's set wrong in the BIOS rather than faulty RAM but at least it will narrow it down or eliminate that.
03-28-2015, 06:54 AM
(03-27-2015, 10:58 AM)Outlawz7 link Wrote:You can do both CPU and GPU and I have http://openhardwaremonitor.org/ for logging.
HwInfo reads a lot more sensors, so might be more useful when trying to identify the problem. Whatever you use though, have you checked the logs to see what the voltages, VID, CPU load, speed, etc were at the time of the crashes? Once you've ruled out RAM, just test the CPU with OCCT and if that crashes you know it's not your GPU.