BSOD - Need Help!

[xxJohnxx]

Hi,

 

I started to get Bluescreens some weeks ago. It started with one in the first week of November, but now I am getting at least one per day, and it is getting very anoying!

I tried to update all the drivers, tried to remove hardware installed in November and checked the internet for the cause, but I couldn't solve my problem. So I need your help guys!

I attached the last two Minidumps.

System specs are in the signature, and if you need anymore information, please let me know!

 

Thank you and best regards,

 

Jonathan

Dump.zip

Edited January 15, 2014 by xxJohnxx
Adjustet title

[ALDEGA]

WinDbg says:

 

"Probably caused by : hardware"

 

Check Windows Event Viewer -> Custom Views -> Administrative Events

Any disk related errors?

 

tried to remove hardware installed in November

What kind of hardware did you add in November?

[xxJohnxx]

Hi,

 

"Probably caused by : hardware"

 

Check Windows Event Viewer -> Custom Views -> Administrative Events

Any disk related errors?

 

I get lot's of Warnings and Failures in there.

Some of those:

- Kernel-EventTracing

- WHEA-Logger

- e1qexpress

 

 

 

What kind of hardware did you add in November?

 

A USB Headset.

 

I got the next Bluescreen while I was typing this...

 

Thanks for your help!

 

EDIT:

 

Minutes before the first BlueScreen I got a crash of TeamSpeak and a minute later also Dwm.exe. I am not sure if that is related or not.

The expection code for Dwm.exe was 0xc0000005

[SkateZilla]

one was hal.dll the other was tcpip.sys

[ALDEGA]

Since the crashes are occurring in various exe/dll, I suspect a corrupt disk (most likely) or RAM.

To check the harddisk: run chkdsk on c:

To check the RAM: WindowsKey+R, type mdsched and press OK

[blackbelter]

Besides the above suggestions, you can also try lowering the cpu oc to 4.4 or 4.3 and see. It's quite normal that extreme oc-ing leads to system instability.

Edited December 31, 2013 by blackbelter

[Abburo]

Check your power supply first.

[xxJohnxx]

Hi,

 

Thank you all for your suggestions. I will try to work them out one after one! :)

 

Since the crashes are occurring in various exe/dll, I suspect a corrupt disk (most likely) or RAM.

To check the harddisk: run chkdsk on c:

 

chkdsk did not bring up any faults on any of my drives.

 

To check the RAM: WindowsKey+R, type mdsched and press OK

 

I will do that and report after doing so!

 

Besides the above suggestions, you can also try lowering the cpu oc to 4.4 or 4.3 and see. It's quite normal that extreme oc-ing leads to system instability.

 

I will definatelly consider that, altough it has been runing on 4.5 GHz for about 8 month now.

 

Check your power supply first.

 

I got 750 Watts. What do you think, is that not enough?

I got the bluescreens while typing in the browser, where the system doesn't have a very high power consumpiton, so I am not sure!

 

Best regards and thank you very much!

 

Jonathan

[blackbelter]

Other than its power output rating, another important specification of a power supply unit is its ability to keep its output stable. If the output is not stable, it may harm the electronic circuits on the components in the long run, ultimately leading to something like what is happening to your system. Specifically, is your power supply Bronze-, Silver- or Gold-rated? Obviously, the more expansive the Metal, the more stable the output.

 

I am not sure, but 750W for your computer specs seems a bit on the edge... Even if it's enough on the paper, the power supply unit might be pushed close to its limit, in which the output is likely to be unstable.

 

Not 100% sure. So please take my words with a grain of salt..

Edited December 31, 2013 by blackbelter

[xxJohnxx]

Other than its power output rating, another important specification of a power supply unit is its ability to keep its output stable. If the output is not stable, it may harm the electronic circuits on the components in the long run, ultimately leading to something like what is happening to your system. Specifically, is your power supply Bronze-, Silver- or Gold-rated? Obviously, the more expansive the Metal, the more stable the output.

 

I am not sure, but 750W for your computer specs seems a bit on the edge... Even if it's enough on the paper, the power supply unit might be pushed close to its limit, in which the output is likely to be unstable.

 

Not 100% sure. So please take my words with a grain of salt..

 

I got this one: link

It seems to be silver rated.

 

Thanks for the help!

 

Best regards,

 

Jonathan

[cichlidfan]

That's a good PSU but I might bump the output a tad. Nvidia rates the 780 as drawing 250w and recomends a 600w PSU (for a single card). If you are buying a new supply anyway, spend a bit more for at least another 100w. If it were me, I would use a 1000w PSU.

 

My two cents...

[pacotito]

It could be a weak power supply. According to this article 2 gtx 780s need 800w and that's without cpu/gpu overclock.

http://www.guru3d.com/articles_pages/gigabyte_geforce_gtx_780_windforce_oc_review,7.html

 

The 80 plus rating (silver,gold,ect.) has nothing to do with cleanliness of power delivered. It only deals with psu efficiency. Or how many watts is supplies compared to how many watts it actually uses. The tip I can give is to read lots of detailed reviews. They usually test everything including power cleanliness. Like cichlid fan recommended I would say 1000w or maybe even a little higher.

[blackbelter]

@Pacotito, thanks for correcting me. Good to learn something new... :thumbup:

[xxJohnxx]

Hi,

 

To check the RAM: WindowsKey+R, type mdsched and press OK

 

That check didn't find any errors.

 

That's a good PSU but I might bump the output a tad. Nvidia rates the 780 as drawing 250w and recomends a 600w PSU (for a single card). If you are buying a new supply anyway, spend a bit more for at least another 100w. If it were me, I would use a 1000w PSU.

 

It could be a weak power supply. According to this article 2 gtx 780s need 800w and that's without cpu/gpu overclock.

http://www.guru3d.com/articles_pages/gigabyte_geforce_gtx_780_windforce_oc_review,7.html

[...]

Like cichlid fan recommended I would say 1000w or maybe even a little higher.

 

Okay. I probably will order myself a stronger PSU in the next few days if no more suggestions come up!

 

Thank you very much and best regards,

 

Jonathan

[trooph]

Just a small tip...

BSODS are not always caused by hardware failure. Did u try to do fresh windows install? It also may be caused by some drivers that "dont like each other" or simply by corrupt system files. I know this happens from time to time if u dont keep ur OS clean. I personally had issue like that 6 months ago where my PC suffered from BSODS during gameplay at first and later during system boot up. I literaly had to wipe my OS clean (no recovery disc helped)

[Abburo]

Your graphics have the followings specs:

 

up to 300Wadditional 6+8 pin PCIe power required

 

So, is safe to say that your system is firstly underpowered. Go for >1000W PSU in order to be on safe side. I have 900W for 2x770GTX and is working properly, but the PSU is getting quite warm while the system is in full load.

Edited December 31, 2013 by Abburo

[danilop]

The easy way to check for PSU overloading is to pull out one of GTX780s and do the stability test for several hours on single GPU. It looks, like your PSU was upto SLI power requirement while new and with aging lost some of it's power (quite common - that's the reason why you want some overhead in power).

 

Quality 850W PSU would be sweet spot.

 

Good luck!

Edited December 31, 2013 by danilop

[SkateZilla]

750w is not enough for 2x 780s in SLI, and a i7-3770K overclocked to 4.5GHz.

You Are Barely Cutting it or overshooting by about 50 Watts,

(300x2, and about 200w for i7-3770K @ 4.5Ghz),

 

You could get away with it for a while, by starving your CPU or 2nd GPU depending on what you're running.

But as soon as PSU begins to age and power output degrades, you're gonna run into low voltages on the NB, PCI-e Slots, and RAM Slots, then CPU will start to undervolt and Fail Prime 95, and eventually PSU will go Poof on it's own.

 

Basically, something isnt getting enough power now, While I would say Ram, it looks like the Mainboard itself isnt getting enough power to power the onboard LAN and Devices.

 

Microsoft's RAM Check is worthless, run Memtest from USB Stick, and sit in Awe of how many errors pop up because the RAM is not getting enough Voltage.

Edited December 31, 2013 by SkateZilla

[xxJohnxx]

Okay guys. Thank you for all your replies. I will go ahead and order myself a new PSU then. I have been reading good stuff about the Corsair AX1200i (1200W as the name suggests) and probably will order that one later tonight.

In the meantime I started Memtest86. It is running now:

 

 

I hope I set it up the right way. No errors reportet yet, but I will let it run for some time.

 

Thank you and best regards,

 

Jonathan

[SkateZilla]

usually takes 2 or 3 passes for stuff to finally start to show up.

 

I Usually run 8-10 Passes to verify RAM integrity.

[xxJohnxx]

3.5 hours done. 3 passes where made and no error yet. I will let it run for a few more passes.

 

EDIT: Does Memtest also put load on the CPU and GPU? The CPU is at 45°C now, and the GPUs seem to be very cold too, as the fans didn't spool up yet, nor is a lot of heat comming out of the case...

[SkateZilla]

shouldnt put any load on GPU, CPU Load would be minimal.

 

So you'd prolly stay below your 750w Ceiling as both GPUs would pretty much be Idling, and the CPU is prolly using Max 70w.

[xxJohnxx]

Okay, so would that RAM undervoltage even occur if I run Memtest? I mean it can't detect a problem that does not occur, because the system is not on load? Or am I getting something wrong?

[SkateZilla]

Most likely not.

 

Until the system is put under load, the PSU will be able to supply power to the components in use with no problem,

 

once you overload the PSU and it's not able to supply power, components will start to throw up errors, which can be exhibited on the Ram, the CPU, the mainboard, etc.

 

Easiest way is to see what your GPUs+CPU Power Consumption is, and what the Voltages are on various mainboard components, 12v rail, 3.3v, 5v etc.

 

Usually CPU voltages are the first to suffer,

when CPU Voltage drops too low, it will start to cause problems, computation errors, lock ups, BSODs when a Thread or Process requires a big enough workload etc.

 

if the main-board is starved of power, then PCIe Slots wont put out enough, Ram Slots wont have enough, etc etc,.. Each GPU still depends on 75w from the PCIe Slot. So you're board is sucking at least 150w alone for the GPU's, plus the Power from the PCI-6/8 Pin Connectors. If GPUs are sucking all the power to the PCI-e Slots, then Voltages on the Memory bus may drop, same with NB Voltages.

Edited December 31, 2013 by SkateZilla

[xxJohnxx]

Okay, thank you guys, and especially SkateZilla, for helping me out and explaining the problem to me! :thumbup:

I will place my order on the new GPU today, and I will report back if the stronger PSU could solve the problem once I get it.

 

Thank you very much and a happy new year!

 

Jonathan