New build crashing with no blue screen

[Michael987]

Hello,

I have been having some problems with my PC in the past week. I built this PC about 4 months ago and I did not have any problems with it until now.

What is happening is that my PC just completely shuts itself down and then restarts itself. There are no blue screens or messages on the PC, it simply shuts down instantaneously and then reboots about 3-5 seconds later.

My current system specs are: http://i.imgur.com/KpUSquN.png

The PSU I use is: https://www.amazon.co.uk/gp/product/B0100QRVP0

I have used PCpartpicker and that shows the total wats at about 500W and since my PSU is 650W I should be fine

I do not think this is a heat related issue since the highest temps I am getting is about 50° for the cpu which is fine and about 55° for the GPU, of which I have had running before at about 75° with no issues.

I have done some google searching of this but the answers are divided, some saying PSU problems and some saying GPU problems and so on.

I have checked event viewer and the error messages I get are Kernel-Power (Telling me it did not shut down correctly which I already know) and the other error message is event ID 1101 - Audit events have been dropped by the transport. 0. When I did a search of this I found this was usually involved with the crashing, however there were many different answers that people gave that solved their issues; some of which included disabling audio drivers via Device Manager but that has not helped me.

After my latest crash (1 hour ago), I checked for any BIOS updates and have installed some new ones which have recently been released via MSI's website.

I have checked via Device Manager than there are no "!/?" next to anything which would suggest an update is needed.

My crashes do not really have any pattern, some are when Im on the internet, some on YouTube and some on different games. However, I have never gotten two crashes on the same day and are usually 2-3 days apart from each-other, no matter what I am doing.

What could be causing this problem?

Furthermore, what things can I do to try and find out which component is causing the problem ?

If there is anything else you need me to do and post let me know, thanks.

 

[Colif]

whats speccy show voltages on motherboard as? might want to use HWINFO for a second opinion as speccy lied to me yesterday and made a PSU look far worse than it actually is, so I don't believe it much. I just want to confirm its not the PSU before we move along

have you used a program like driver booster to make sure you up to date?

 

[Michael987]

whats speccy show voltages on motherboard as? might want to use HWINFO for a second opinion as speccy lied to me yesterday and made a PSU look far worse than it actually is, so I don't believe it much. I just want to confirm its not the PSU before we move along

have you used a program like driver booster to make sure you up to date?

I have no idea what I'm doing with HWINFO so you will need to tell me what you want a screenshot of.

I have not used a program like Driver Booster before as they sound like a version of PC Optimizer Pro telling me I have x amount of things wrong and I need to download lots to fix it. However, if it is actually a reputable software I will give it a run.

 

[Colif]

honestly, if you got HWINFO64, i never worked out how to get voltage info from it either. Doesn't help i have 4k and the program doesn't scale well. It seems to show everything except what i want. I will see if i can find where it shows.

I would have suggested Speedfan but it shows every register before you set it up and gives to much info

driver booster seems pretty good, doesn't dl things that don't match what you have, careful downloading ethernet drivers if you using them at time, it helps to disable firewall as it may freeze pc - it did it to me

 

[Colif]

when you start it up, don't tick summary only or sensors only and you will get all 3.

on the main interface there are 4 buttons
Summary which is screen i normally look at
Save report
Sensors
About

Choose sensors and scroll down the stupidly long list of sensors until you reach your motherboard
Under the temp sensors is a list of voltages
Just need +12 volt, +3.3volt & +5volt

and I will stop wasting your time.

 

[Michael987]

honestly, if you got HWINFO64, i never worked out how to get voltage info from it either. Doesn't help i have 4k and the program doesn't scale well. It seems to show everything except what i want. I will see if i can find where it shows.

I would have suggested Speedfan but it shows every register before you set it up and gives to much info

driver booster seems pretty good, doesn't dl things that don't match what you have, careful downloading ethernet drivers if you using them at time, it helps to disable firewall as it may freeze pc - it did it to me

when you start it up, don't tick summary only or sensors only and you will get all 3.

on the main interface there are 4 buttons
Summary which is screen i normally look at
Save report
Sensors
About

Choose sensors and scroll down the stupidly long list of sensors until you reach your motherboard
Under the temp sensors is a list of voltages
Just need +12 volt, +3.3volt & +5volt

and I will stop wasting your time.

I made screenshots of everything on the sensors page that came up before you said what to choose, therefore Im going to just post all 4 of them.

1: http://i.imgur.com/bCO2fA5.png
2: http://i.imgur.com/pvVy6CB.png
3: http://i.imgur.com/ILcD1Rh.png
4: http://i.imgur.com/7W4knNp.png

The ones you want are specifically in the second screenshot at the bottom

 

[Colif]

okay, so its not power. as they look okay to me. sorry to waste your time

did you run Samsung Magician and let it set a performance scheme? I only wonder as i see you running at max speed on all 4 cores - it did that to me last year and I was already running overclocked at time.

Since its not blue screening its likely to be hardware of some kind, could run memtest86 to make sure its not ram. process of elimination I guess.

 

[Michael987]

okay, so its not power. as they look okay to me. sorry to waste your time

did you run Samsung Magician and let it set a performance scheme? I only wonder as i see you running at max speed on all 4 cores - it did that to me last year and I was already running overclocked at time.

Since its not blue screening its likely to be hardware of some kind, could run memtest86 to make sure its not ram. process of elimination I guess.

Just downloaded Samsung Magician now. What performance scheme would you recommend that I use? I will then set it and restart PC after the Battlefield premiere it done.

I will do a memtest86 soon to check if it is my RAM causing any issues and then probably report back the results here once that is done. I have some old RAM which I could use instead of running the test as that would be much quicker to see if it crashes, although as I said I have not found a way to make the system crash in a reliable way

 

[Colif]

i wasn't suggesting you get Magician, I was asking if you were using it already, are you running something like afterburner to run max speed or was that just what sensor was reporting at time. my CPU ruining at 779mhz right now as its doing nothing at all.

I don't have a power scheme set, i just let win 10 run its balanced power. I don't need CPU running 100% all the time.

if people suggest gpu as possible cause, run 3dmark or unengine benchmarks and see if it crashes.
Evo ssd have no moving parts, ssd not likely to be cause of crashes. Magician has tests for the ssd in it. Also some new firmware for your ssd
you can run seatools for windows and check out the seagate hdd
its about now you tell me you have some USB device thats causing the error

 

[Michael987]

i wasn't suggesting you get Magician, I was asking if you were using it already, are you running something like afterburner to run max speed or was that just what sensor was reporting at time. my CPU ruining at 779mhz right now as its doing nothing at all.

I don't have a power scheme set, i just let win 10 run its balanced power. I don't need CPU running 100% all the time.

if people suggest gpu as possible cause, run 3dmark or unengine benchmarks and see if it crashes.
Evo ssd have no moving parts, ssd not likely to be cause of crashes. Magician has tests for the ssd in it. Also some new firmware for your ssd
you can run seatools for windows and check out the seagate hdd
its about now you tell me you have some USB device thats causing the error

I was not running anything to change the SSD at all, the only thing I have on is the High performance power plan for the PC that could have an effect on the SSD I guess.

I will try running a gpu performance/stress test to see what happens then.

 

[Michael987]

Ran the Heaven Benchmark 4.0 from unigine and GPU hit 70° however there was no crash or problems, so I think the GPU may be fine.

Will try and find a CPU Benchmark tool to see if I can test that also.

I ran CPU on a stress test at 100% for about 7 minutes. It got up to about 64° and had no problems there either so I dont think it is a temperature problem otherwise it should have shut itself down

 

[Colif]

running out of things to test: https://downloadcenter.intel.com/download/19792/Intel-Processor-Diagnostic-Tool

 

[Colif]

is your ram running in an xmp profile? cause it shows as 933.4 whereas mine at normal default on my motherboard is showing as 666.3.

what speed ram are you running?

 

[Michael987]

is your ram running in an xmp profile? cause it shows as 933.4 whereas mine at normal default on my motherboard is showing as 666.3.

what speed ram are you running?

I have no idea what an xmp profile is.

The RAM I have is: https://www.amazon.co.uk/gp/product/B00J8E8ZOC/ - I have not changed anything about this speed wise

I will download the other link you sent me now also and report back.

Finished running the link you sent me, I passed.

Is there anything from that test that you wanted posted here or did you just want to know if it passed?

 

[Colif]

okay, your speed looks about right for those modules, i don't understand my speed now... i think i just running my ram slow, yours fits specs of your board,

xmp just overclocked ram profile but you have to set it so i doubt its anything to do with this.

 

[Colif]

just wanted to pass... not much we haven't tested now, not sure what to look at next

 

[Colif]

so far we checked psu, CPU, ram, ssd (I assume), possibly hdd, whats left?

motherboard tests all cost money i think unless u can find a trial.

 

[Michael987]

just wanted to pass... not much we haven't tested now, not sure what to look at next

I still need to test RAM with memtest so I will do that at some point.

So far we have tested a lot which has somewhat narrowed it down but what to at this point?

If I cant replicate the crash then I cant find the source of it really as it seems to happen at random times

 

[Colif]

i will wait for ram tests before guessing what else it might be. I have to go out for an hour but will return.

 

[Michael987]

so far we checked psu, CPU, ram, ssd (I assume), possibly hdd, whats left?

motherboard tests all cost money i think unless u can find a trial.

Did not do HDD as that was somewhat confusing on what it wanted me to do haha.

Will post at a later date for RAM tests, until then I dont think there is much else to be done unless someone else comments with something to try or you do.

 

[turkey3_scratch]

Is there a pattern to when it shuts off?

 

[Darkbreeze]

when you start it up, don't tick summary only or sensors only and you will get all 3.

on the main interface there are 4 buttons
Summary which is screen i normally look at
Save report
Sensors
About

Choose sensors and scroll down the stupidly long list of sensors until you reach your motherboard
Under the temp sensors is a list of voltages
Just need +12 volt, +3.3volt & +5volt

and I will stop wasting your time.

Actually, if you're running or using HWinfo, which I recommend way more than I ever would HWmonitor or Open hardware monitor, you WANT to check the box for "sensors only". Everything useful you can get from HWinfo will be included in the sensors only window. The rest is just mostly cosmetic redundant sensor data. There are a few useful features in the main window, but most anything you need to know for diagnosis or troubleshooting will be in the sensors window. Voltages will look like this. +12v, +5v and +3.3v are the system voltages you would generally be looking at if the PSU is your target or suspect:

 

[Colif]

Given how little i have used HWINFO I am not going to argue with that

 

[Michael987]

Is there a pattern to when it shuts off?

No pattern that I have found as yesterday is crashed 3 times on me whereas the other crashes have always been 2-3 days apart regardless of what I did.

Sometimes it would be an a reddit thread, other times on a YouTube video and then most of the times on different games.

The game which crashes the most for me right now is Overwatch

 

[Darkbreeze]

Since hardware does not seem to be the problem, at least on the surface, I'd start with doing a CLEAN driver installation, using the latest Crimson 16.5.1 hotfix driver, and using the DDU, as follows:

http://www.tomshardware.com/faq/id-2767677/clean-graphics-driver-install-windows.html


This driver: http://support.amd.com/en-us/download/desktop?os=Windows%2010%20-%2064


Heaven benchmark may not use the same aspects of the driver or API as any given game does, so the fact that it doesn't crash during the stress test may not indicate that drivers are not the problem. If doing a clean GPU card driver installation doesn't solve the problem, then a clean Windows installation may be the next step.