System auto off mid use, power and reset buttons unresponsive.

[thyroid]

Sys Specs
=========
Ryzen 1800x w/ hyper 212 using noctua in push pull
Asus x370 Prime - Pro
G.SKILL Ripjaws V Series 32GB (2x16GB) DDR4 3000MHz C16
Corsair TX 750
EVGA Superclocked GTX 1080

Hi all,

So I'm having a little bit of an issue and I'm trying to nail down possible causes. This is a new AMD build (just over a week old) and also my first amd build.

Basically when I'm running Photoshop, and sometimes Photoshop + Unity, the system will just turn off. No BSOD, just one second on, the next off. The power and reset buttons will usually be unresponsive but will sometimes work. While those buttons are unresponsive the case fans and led lights are still getting power. When the buttons are unresponsive I have to recycle the psu and everything will boot up fine afterwards.

No major OC happening. Docp set system ram to 2933. Cpu was at 3.85. Was stable during prime95 tests.

Currently I'm testing mem to make sure no issues there. First pass at stock settings, second pass at docp settings. So far no errors (memtest86 @ 9+hrs).

I'm thinking it could be the MB because of the power/reset buttons becoming unresponsive while the case fans and leds are powered. Another indicator is there seems to be issues with the board in general based on user experiences I'm reading online.

I cannot rule out the PSU however the PSU has been in 3 previous builds. It's definitely long in the tooth but has been a rock solid performer since first use, never an issue. I just find it hard to believe it would suddenly decide to selectively fail after a new build. Not saying it's impossible just improbable. I don't have a spare PSU to rule that out.

I still have time to bring back the mobo and swap it for another. I'm debating just paying the difference and grabbing a an asus hero 6.

What does it sound like to you? What steps would you take to diagnose the issue? What x370 MB would you recommend?

Thanks for the help!

 

[James Mason]

It would sound like an overheating issue.
Uhh i don't see a GPU listed there, and your CPU doesn't have integrated graphics, so what's your GPU?
In theory PSUs can overheat as well, is the PSU fan still spinning?
Overheating Mobo components can cause a shutdown like that, but you'd have to have a completely useless computer case that blocks all air flow for that to happen.

The Case fans still working are because the computer is trying to cool itself down.

 

[thyroid]

Added GPU. Sry about that.

That's pretty interesting. Could the cpu be contributing then? I was planning on re-seating the cooler to make sure.

I've been running the GPU in this case for a year now (previous build, i7 2600k, corsair h80 cooled), and never had issue with temp or perf.

The case is a Corsair case and is well ventilated. The front fans intake cool air, the 212 pushes the air through the heat sink, and a rear noctua case fan pushes hot air out. There are side and top grills for fans as well, and they're unblocked. I plan on adding two more case fans.

I'll have to double check the spinning psu fan. Originally the psu fan was pointing up into the case but I've reversed it to point through the bottom grill of the case.

I should also mention the pc is in the coldest room of the house. Idle cpu temps sit ~30c and load ~65c to ~70c. Not sure the actual mb temps tho.

The temp issue seems a bit odd since I can run prime95 for hrs with no heat issues but photoshop will crash within minutes. Still tho I'll double check cooling systems are working. Thanks for the suggestions!

 

[James Mason]

Hmm, well first off, disable automatic restarts, and make sure the system is generating a kernel memory dump:
https://support.hp.com/us-en/document/c03413399

Then make it crash, open up Event Viewer and look for recent Red X Errors. (DO NOT COPY AND PASTE THEM WHOLE INTO THE THIS THREAD)
Also get BlueScreenView to read those kernel memory dumps:
https://www.nirsoft.net/utils/blue_screen_view.html

And then you want to google the file names + extension and with their BugCheckCodes for the red highlighted items to see what associated problems come up.

NTOSKRNL.EXE is basically Windows itself, so that will always be there, won't necessarily be the culprit, there can be a problem here, but the bugcheck code will indicate if there's a problem besides "it crashed".
hal.dll is basically how your hardware talks with your software, so that also will likely be there always, there can be a problem here, but the bugcheck code will indicate if there's a problem besides "it crashed".

 

[thyroid]

Thanks for the post!

So I've followed your steps and the only critical error is kernel-power with bug code 0.

According to ms it could be a bad oc, heat, PSU, or ram.

Memtest96 8 pass 0 errors.

Hwmonitor cpu max temp 46c, mb max temps 36c and 34c.

Clocked back the OC to default and still kernel-power error happened.

PSU is 750watts powering ~550watt system. Never any issues before this build so it's hard to suddenly lay blame on it. I'll try and borrow a PSU from work to test out.

Aside from that I'm not sure. Cheapest path for me is to exchange the board anyways and if the problem still occurs then will grab new PSU.

 

[James Mason]

Yeah PSU watts don't mean anything to whether it's a good or working PSU.
And no blue screen dumps turning up anything different?

 

[thyroid]

There's no blue screen. It's the equivalent of someone switching the PSU off. One second it's there, next off. I could check for men bsod dump tho, just in case.

 

[James Mason]

yeah, that certainly sounds like a mobo or psu issue.
if you can replace the mobo for free, try that. if still not fixed you'll have to replace the PSU as well.

Though it could still be like, a shorting out GPU for example, there's not much you can really do to diagnose something like this except stress specific parts to see if you can reliably reproduce the error

 

[thyroid]

Yeah very true. It was a core component upgrade so only cpu, ram, and mobo are new. Never had an issue with any of the other parts (gpu, psu, etc...) so pretty confident at this point its something related to the mobo.

Asus support suggested a CMOS reset which didn't solve the issue. Also I just noticed an AI SUITE update on asus' site so im trying that but I'll probably bring the board back tomorrow. The prime-pro left a bitter taste in my mouth so im thinking of going msi carbon or maybe even hero 6. I'm open to suggestions if you have any.

 

[thyroid]

I just had one of the crashes where the system will successfully reboot and it gave some more info.

A second Kernel-Power Error:

Kernel-Power, EventID 172, Task 203

Event Data > State 2, Reason 6

Connectivity state in standby: Disconnected, Reason: NIC compliance.

Google-fu brings up nothing aside from psu or other component failure.
Two previous 'Information' logs precede each critical error in the event viewer.

First is:
File System Filter 'FileCrypt' (...) has successfully loaded and registered with Filter Manager. Event ID 6

Second is:
File System Filter 'npsvctrig' (...) has successfully loaded and registered with Filter Manager. Event ID 6

Now I need to figure out what npsvctrig is.

 

[James Mason]

Did you install your LAN drivers for your mobo from your mobo's website? (not the CD that came with it)
https://www.asus.com/us/Motherboards/PRIME-X370-PRO/HelpDesk_Download/

The first one, "NIC" means Network Interface Card i.e. your ethernet port/controller/drivers.
That could be a motherboard issue, or a driver issue if for whatever reason you don't have them.

npsvctrig is another network related thing, but appears to be working.

 

[thyroid]

It turns out it was a bad windows install. Speaking to the tech at the store, his recommendation was to format and reinstall windows, and so far it's worked. Up time is over 24hrs at this point. Thanks for your assistance James! I appreciate it.

 

[scottbdrury]

I am having the exact same problem, with very similar specs. Has the windows refresh held up?.