Random Reboots. Previously Stable Win10 x64

[Formaleins]

Wonder if anyone has any ideas on this one. The PC is randomly restarting (it reboots on its own which seems strange as I have it set to be off if there is a power failure - so this suggests some sort of restart maybe?)

There are no Minidump files and no related errors in the Windows Logs, everything there appears fine. (I get the normal error 10016 Kernel Power, but I think that is just a warning from Windows that it was shutdown unexpectedly)

I haven't really made any changes lately so I cannot think of any software that would cause it.I have removed all overclocking and set everything back to standard and it still does this randomly.

 

The system is quite new, motherboard Asus Maximus IV Code,  Intel i7 7700K, Samsung NVMA 250GB M2 drive (boot), a couple of WD hard drives an Intel 520SSD. RAM is Corsair Vengeance 3400MHz. (downclocked to 2200MHz for now)

Windows is genuine x64 Pro.

I am beginning to suspect it may be something to do with the PSU, I have been running a log using HWInfo to watch the temperatures and the 12V, 3V and 5V but nothing is showing up.

 

Anyone have any ideas? Sometimes it will run all day then at other times 3 reboots in 20 minutes.

 

Cheers

 

(Forgot to mention, I have been running MEMTEST for over 8 hours and no problems picked up there) Graphics is Gigabyte Nvidia GE 1050Ti.

Edited September 28, 2017 by Formaleins

[mtls2005]

Continue to evaluate the memory, if you have more than one module remove all but one and test each one separately (a pain, I know).

 

Completely remove the graphics driver - there are some good total removal programs available, often from the GPU/card manufacturer. Reinstall the most current driver.

 

Check heat issues, esp. CPU cooling.

 

Check power (mains) supply.

 

Probably memory (faulty or clocking issue) and GPU driver issue? But these usually leave BSOD trace.

 

Assume you have all the latest Asus drivers/MB firmware?

 

 

[Formaleins]

1 minute ago, mtls2005 said:

Continue to evaluate the memory, if you have more than one module remove all but one and test each one separately (a pain, I know).

 

Completely remove the graphics driver - there are some good total removal programs available, often from the GPU/card manufacturer. Reinstall the most current driver.

 

Check heat issues, esp. CPU cooling.

 

Check power (mains) supply.

 

Probably memory (faulty or clocking issue) and GPU driver issue? But these usually leave BSOD trace.

 

Assume you have all the latest Asus drivers/MB firmware?

 

 

Thanks,

cooling isn't a problem, it has liquid cooling and never gets above about 75C on the CPU.

I am going to have a look at the removal of the graphics drivers as that sounds possible - What I have noticed is that it usually crashes when it is running Edge with YouTube open, whether running or paused.

 

Cheers for the suggestions, hopefully I will find a fix as it is a massive PITA!

 

[RichCor]

1 hour ago, Formaleins said:

Anyone have any ideas? Sometimes it will run all day then at other times 3 reboots in 20 minutes.

 

Cheers

 

(Forgot to mention, I have been running MEMTEST for over 8 hours and no problems picked up there) Graphics is Gigabyte Nvidia GE 1050Ti.

 

mtls2005 has given some great suggestions.

 

If the issue is somewhat consistent, I'd try unplugging the HD and boot some test OS off a USB stick and then mem test. If you do get issues, try also booting into BIOS / SETUP and see if it resets there. Hopefully these two tests would help say it IS or ISN'T a boot OS-based issue.

 

If it did continue to crash while booting off USB or BIOS/SETUP then that would more clearly indicate a hardware issue ...which could be anything from a failing/weak Power Supply, a failing subsystem module or component, or something causing a bus to reset. Hopefully *something* would stand out as the direct replaceable culprit. 

[Formaleins]

7 hours ago, RichCor said:

 

mtls2005 has given some great suggestions.

 

If the issue is somewhat consistent, I'd try unplugging the HD and boot some test OS off a USB stick and then mem test. If you do get issues, try also booting into BIOS / SETUP and see if it resets there. Hopefully these two tests would help say it IS or ISN'T a boot OS-based issue.

 

If it did continue to crash while booting off USB or BIOS/SETUP then that would more clearly indicate a hardware issue ...which could be anything from a failing/weak Power Supply, a failing subsystem module or component, or something causing a bus to reset. Hopefully *something* would stand out as the direct replaceable culprit. 

It seems to be the previous graphics driver combined with Edge.

I cleaned out all the old graphic drivers as suggested and reverted to the one that came with the 1050Ti. It has run all day (a good 8 hours) without a reboot.

I just need to sort something out with Edge as it is misbehaving and locking up, Chrome is fine it has been left on all day running Youtube autoplay without an error.

 

Thanks for the help!

[Minnehaha]

I have owned so many computer and similar devices in Thailand that I begin to see patterns. 

 

The most common failure is power supply - either the connection male -female breaks or gets tweaked in some way, or the power brick fails or the cable fails.  

[Svadhistana]

Unfortunately, this is a kind of problem that, sooner or later, almost everybody experience. In few cases, it is a hardware problem, but that is usually not too difficult to detect. In all other cases, it is a software conflict between the OS and "something else" and you can become crazy trying to sort it out. My warm suggestion is: reinstall Windows and your problem will be surely solved (unless it is a hardware problem, as I said before).

[writeshack]

This may sound simplistic given the knowledgeable advice you  get here, but I had a similar problem with Win 10 and a quad computer running i7  with a dedicated AMD graphics card.  Nothing worked, not even online help. I ended up using reimage, and I now reject graphics driver updates (Radeon in my case) and image suite updates. I run reimage occasionally, it updates the basics, and have very few problems and my computer works  well.

[klauskunkel]

Random shut-down followed by immediate independent restart, I experienced this mostly when there is a short power fluctuation in the grid, i.e. power outage of a few seconds.

Edited September 28, 2017 by klauskunkel

[Formaleins]

Well, I thought the graphics driver had fixed things but it hasn't. The damn thing has rasn fine all day running YouTube with a couple of browser windows open and file explorer open.

Shut those down and went to run X-Plane, the program fired up fine but when I went to exit, there is that dreaded CLICK from the PC case and the power is turned off, then it reboots - Despite the BIOS being set to remain off after power loss.

 

Back to testing step by step. I may swap out the PSU tomorrow from my son's PC to see if it makes a difference.

 

Damn!

[Formaleins]

38 minutes ago, writeshack said:

This may sound simplistic given the knowledgeable advice you  get here, but I had a similar problem with Win 10 and a quad computer running i7  with a dedicated AMD graphics card.  Nothing worked, not even online help. I ended up using reimage, and I now reject graphics driver updates (Radeon in my case) and image suite updates. I run reimage occasionally, it updates the basics, and have very few problems and my computer works  well.

I have heard of this program but wasn't sure if it was just a gimmick - I will give it a trey and see what happens. Thanks.

[writeshack]

Hope it helps. Just remember to create a restore point first, although reimage will do that too. From memory, the first use was free. Not sure if that's still the case. Not everyone likes it, but  very many do.  

[Formaleins]

1 hour ago, Svadhistana said:

Unfortunately, this is a kind of problem that, sooner or later, almost everybody experience. In few cases, it is a hardware problem, but that is usually not too difficult to detect. In all other cases, it is a software conflict between the OS and "something else" and you can become crazy trying to sort it out. My warm suggestion is: reinstall Windows and your problem will be surely solved (unless it is a hardware problem, as I said before).

Been there done that and got the T-shirt, deep down inside I know that I am going to have to do this sooner or later, just hoping that it could be something else with any luck.

Thanks.

[Formaleins]

Regarding the program mentioned by Writeshack - I downloaded it and ran it and it did find that Edge was a major contributor to the crashes, unfortunately it did not detect any crashes from today ( I have had about 3 after midnight last night andthen this crash tonight - none of these were detected. Plus they want about 40USD to do the repair.

 

It is looking less like software the more that it happens - i.e. no mini dump files, no blue screen freezing, just a reboot and the only three files listed in the error report logs which are basically reports that Windows didn't shut down properly, not the actual cause.

 

As far as the power from the mains goes, it is not the best here where I live but my system is pretty resilient compared to some of the other stuff in the house. My PC will still run at 180V AC when the power drops and everything else is switching off.  

 

This reboot is not like that at all - it is instantaneous and you can hear the click (like a hard reset) as the motherboard reboots.

 

Thanks for all the suggestions, appreciated.

[writeshack]

Sorry it didn't help. Did you ask reimage to update your system, not just search? That's how it works. It  has a cache of current Windows files it uses to replace your problem files.  Hope you get the problem worked out. Bit of a bummer, eh? 

Edited September 28, 2017 by writeshack

[writeshack]

Sorry it didn't help. Did you ask reimage  to update your system, not just search? That's how it works. It  has a cache of current Windows files it uses to replace your problem files.  Hope you get the problem worked out. Bit of a bummer, eh? 

[cme]

14 hours ago, Formaleins said:

Thanks,

cooling isn't a problem, it has liquid cooling and never gets above about 75C on the CPU.

I am going to have a look at the removal of the graphics drivers as that sounds possible - What I have noticed is that it usually crashes when it is running Edge with YouTube open, whether running or paused.

 

Cheers for the suggestions, hopefully I will find a fix as it is a massive PITA!

 

I'm no expert but 75C seems too hot to me. I would think that over 70C the CPU would shut down to prevent itself burning? Have you read the spec for the processor that you have?

[howto]

It is the PSU ,,, I don't care if is a 'brand name'.

They fail also.

 

Everything else would be throwing exceptions, blue screens w/ codes.

That is 'mobo, ram, hdd, gpu'.

Only a PSU can do this w/o throwing blue screens w/ codes.

 

This includes APPS, such as Edge with YouTube open .

You would get an exception, blue screens w/ codes.

 

Replace the PSU with a 'brand name' at least 'bronze' cert, say 650+ watts.

and get a decent UPS.

 

[BaaNahmLai]

Had this issue myself. Upgraded to a more powerful PSU and problem solved.

 

[tweedledee2]

8 hours ago, cme said:

I'm no expert but 75C seems too hot to me. I would think that over 70C the CPU would shut down to prevent itself burning? Have you read the spec for the processor that you have?

I agree 75C is too hot for a liquid cooling system. My fan cooled AMD 4 core runs cooler than that. High temperatures in the processor will cause the system to shut down.  Intel specs are between 64C and 68C. Are you sure the cooling system was installed and is working correctly? On water cooling systems overheating can be caused by lack of or inadequate thermal paste between contact plates or coolant in the system not circulating or flow being restricted. Also, inadequate airflow through the external heat extractor due to being dirty or the fan not working properly is the likely cause overheating.   

[mtls2005]

On 9/28/2017 at 8:18 AM, mtls2005 said:

Check heat issues, esp. CPU cooling.

 

On 9/28/2017 at 8:22 AM, Formaleins said:

cooling isn't a problem, it has liquid cooling and never gets above about 75C on the CPU.

 

Yikes. 75C isn't "a problem"?

 

Are you trying to power your PC by steam?

 

I think there is something wrong with the circulating pump on your cooler?

 

I have a similar build, but with a CoolerMaster Hyper 212+, I'm not sure I've ever seen it north of 38C.

 

This would seem to be your issue, or at least one that needs addressing sooner rather than later.

[Jai Dee]

You may find that it is a conflict between Edge and your graphics system that is the problem.

 

I found that Edge gave me problems not only with YouTube, but also accessing the ThaiVisa Forum.

 

No such problems with Chrome, Firefox, or Brave.

 

I don't use Edge at all now and have no problems.

[ubonjoe]

A off topic rant post about the windows operating system and the resultant replies have been removed.

[Formaleins]

75C on the  Kabylake is not over temperature when all cores are running at 100%, it doesn't begin to throttle the cores until well over 90C.

Its normal run of the mill operating temperature is around 48-60C. depending what programs you run, even the most mundane program will send individual core temps on the Kabylake to almost 70C, in random spikes - It is bad design by Intel, there is quite a few topics about it.

 

If you run Prime95 on the worst settings it will get near to 80C and there is still no throttling.

 

Besides, current ambient temperatures in my room (fan only) are already in excess of 40C after mid day so another 30C on top of that is hardly even warm.

 

Anyway, I think after all of this I have found out what is causing it. I just happened to look in the Windows Custom Views section of Event Viewer late last night and guess what I found?

 

There were two processes, both from Hyper Visor running and one trying to access the graphics card and the other trying to get a hold of the network card.

 

Why these were even running is beyond me as Hyper V has not been used for weeks.

 

I removed Hyper V from Win 10 menu and disabled it in the BIOS and the damn thing has ran all last night and all of today without a glitch!

 

Really cannot understand how Hyper V was even running at all! It does not show up in task manager or any of the other processes, nor in the event viewer system logs.

 

Thanks for all the help though

Edited September 29, 2017 by Formaleins

[Formaleins]

6 hours ago, mtls2005 said:

 

 

Yikes. 75C isn't "a problem"?

 

Are you trying to power your PC by steam?

 

I think there is something wrong with the circulating pump on your cooler?

 

I have a similar build, but with a CoolerMaster Hyper 212+, I'm not sure I've ever seen it north of 38C.

 

This would seem to be your issue, or at least one that needs addressing sooner rather than later.

You must have aircon then as my ambient air temperature is more than 35C most days. My sone has the Hyper 212 on his i7 3770K and it is 48C before it does any work.

 

My temperatures at the moment are 37 /38/39/37 as it is rather cool here today. But it doesn't matter what program you run, this processor will jump in temperature for no apparent reason, it has been complained about widely.

Edited September 29, 2017 by Formaleins

[William T]

While on the subject of start up for PC i have a problem which comes and goes every few weeks.

 

I had Windows 10 installed about 2/3 months ago.I normally put my computer to Sleep a few times during the day and only shut down when finished for the day. Normally when i put to Sleep i have to be very careful when i close the drawer with my keyboard and mouse otherwise it opens up again. When i want to open from sleep i just wiggle my mouse and it opens.

However after a week or two for no obvious reason it wont open and i have to press the Power button and it opens but not the same as a proper start from shutdown. Then after a week or two it corrects itself and back to wiggling the mouse.

Then after a few more weeks and it is back to pressing the Power button. Now back again to wiggling the mouse.

 

Any ideas?

[mikecat]

Yes, it could be the graphics driver or heating issues. I had this problem and found that the computer was automatically set to shut down at a certain temperature. I have a eight year old Samsung laptop  running windows seven. I opened it up and cleaned the fans and heat sink etc, raised the laptop of the desk and use it in an air conditioned room I also updated some drivers. Haven't had a problem since. I would look at the drivers first,

[RichCor]

3 hours ago, William T said:

I had Windows 10 installed about 2/3 months ago. I normally put my computer to Sleep a few times during the day and only shut down when finished for the day. [...W]hen i put to Sleep i have to be very careful when i close the drawer with my keyboard and mouse otherwise it opens up again. When i want to open from sleep i just wiggle my mouse and it opens.

 

However after a week or two for no obvious reason it wont open and i have to press the Power button and it opens but not the same as a proper start from shutdown. Then after a week or two it corrects itself and back to wiggling the mouse. Then after a few more weeks and it is back to pressing the Power button. Now back again to wiggling the mouse.

 

Any ideas?

This question really should in a new topic thread.

 

It could be that after a bit your computer is shutting down power to USB/Mouse ports.

Check 'Device Manager' / each device "Properties" / "Power Management" tab and see if "Allow this computer to turn off this device to save power" is checked on for any Generic USB HUB or Root USB HUB.

 

If my mouse want's to play dead, I usually can tap the Space Bar to wake the computer from sleep.

[ignis]

On 29/09/2017 at 12:02 AM, howto said:

It is the PSU ,,, I don't care if is a 'brand name'.

They fail also.

 

Everything else would be throwing exceptions, blue screens w/ codes.

That is 'mobo, ram, hdd, gpu'.

Only a PSU can do this w/o throwing blue screens w/ codes.

 

This includes APPS, such as Edge with YouTube open .

You would get an exception, blue screens w/ codes.

 

Replace the PSU with a 'brand name' at least 'bronze' cert, say 650+ watts.

and get a decent UPS.

 

Yes also had a problem for months, 7 days ago bought a new PSU and fitted it last Saturday, no problem since,  Thermaltake Smart M 650 - 650W Power Supply 80 PLUS Bronze

 

The old PSU was a 500W cheap one 4 years ago.