Computer won't stay on- keeps shutting down

craigdt

craigdt

[H]ard|Gawd
Joined
Oct 27, 2016
Messages
1,075
We were watching a movie on Plex and my computer, restarted after about 3 minutes.

Powered off, started to reboot then shut down and rebooted all the way.

Weird.

Restarted the movie.

Watched for a couple minutes then repeated the same process.

Triggered on and off about half a dozen times within 20 seconds.

So I unplugged it.

Never had any issues before.

Where should I start first?

Power supply is Seasonic Focus Gold 850 watt.
 
Update-

I removed all devices I wasn't willing to fry, like GPU's, extra drives etc.

I moved the computer to a new outlet and surge protector and power cord.

Same behavior.
Starts to boot, then clicks off.
Then tries to boot again, etc.
 
Unplug anything you can, if it doesn't boot then start swapping components (psu, motherboard, ram, gpu, cpu) < in around that order (or by whatever spares you can get you hands on).

Though first make sure to check that everything is securely attached.
 
Good plan.

I'll have to order a spare psu!

What on earth would cause this after years of working fine?
 
Got me :)
also have you tried with just a single chip of ram?

Could also look around for swollen caps I suppose? (looks like a fairly new system in your sig, so very likely not)
I assume you're not or you've turned off any overclocking?
 
ChoGGi said:
Got me :)
also have you tried with just a single chip of ram?

Could also look around for swollen caps I suppose? (looks like a fairly new system in your sig, so very likely not)
I assume you're not or you've turned off any overclocking?
Click to expand...


Alright so I turned off my very mild overclock.

Instead of shutting off after 15 seconds or so like it did previously, it boots up great and ran about 5 minutes without a load before shutting down again.

Does this mean the processor is overheating?

Like my AIO cooler crapped out?
 
could be temps, what do they show?

You should be able to feel the pump working if you put your finger on the tube. Does the software for it show pump speed?
If not grab siv and see if it supports it (search for aio on that page).
 
Alright, so I was able to run the Intel XTU stress test for 5 minutes and the benchmark.

Temps were high 50s to mid 60s.
No throttling, etc etc. Seemed great.

So I let the machine just sit idle.

Restarted after about 10 minutes.

This time I got a BSOD error message: PFN LIST CORRUPT.

Never done this error message before or since.

But I have since updated all the drivers to the most recent versions, and did a BIOS update to the most recent release.

Restarted again after about 10 minutes.

Im currently typing the from the machine in question.... so far so good, but I have no doubt its going to shut down again.

Other thoughts?
 
PFN LIST CORRUPT is an HDD error.
Try running this in a Administrator command prompt : sfc /scannow to see if there are any errors. If that doesn't work, try disabling One Drive if you have it enabled, I've seen this cause that error before.
 
CAD4466HK said:
PFN LIST CORRUPT is an HDD error.
Try running this in a Administrator command prompt : sfc /scannow to see if there are any errors. If that doesn't work, try disabling One Drive if you have it enabled, I've seen this cause that error before.
Click to expand...

PFN list corruption is a memory error.
 
CAD4466HK said:
Yes, a memory error that is caused by a corruption of the PFN file that resides on the HDD . Hence a HDD error.
Click to expand...

Sooo I have a year old 960 Evo.

Just a reformat or is it irreparably ruined?

Reload OS? what's the fix?
 
I would run sfc /scannow to see if there are any errors. That way you can rule out any device drivers that are bad that can also cause the error. Then go from there.
 
CAD4466HK said:
I would run sfc /scannow to see if there are any errors. That way you can rule out any device drivers that are bad that can also cause the error. Then go from there.
Click to expand...

Does it just run a test or does it include some graphical components?

Because I got massive artificing/screen glitching.
 
CAD4466HK said:
Yes, a memory error that is caused by a corruption of the PFN file that resides on the HDD . Hence a HDD error.
Click to expand...

No, the PFN list is the database that the kernel uses to quickly find free or available pages of memory to allocate for programs to use. Storing it on the hard drive would be comically slow. I think you're thinking of the page file, which stores pages of memory that aren't currently in use when there aren't enough free pages in RAM.


Seeing that he's gotten this error once, and other times it just shuts down at random times. SFC /scanow will not likely produce anything useful. In fact, if there is a memory issue, then it most certainly cause a false hit.

OP, check the memory for errors. Windows has a tool built in, there's also bootable tools like Memtest.
 
CAD4466HK said:
I would run sfc /scannow to see if there are any errors. That way you can rule out any device drivers that are bad that can also cause the error. Then go from there.
Click to expand...

Nothing on the SFC scan I guess

"Windows Resource protection did not find any integrity violations"
 
Currently running the Windows memory diagnostic tool

Again.

It shut itself off halfway through the second pass
 
ryan_975 said:
No, the PFN list is the database that the kernel uses to quickly find free or available pages of memory to allocate for programs to use. Storing it on the hard drive would be comically slow. I think you're thinking of the page file, which stores pages of memory that aren't currently in use when there aren't enough free pages in RAM.


Seeing that he's gotten this error once, and other times it just shuts down at random times. SFC /scanow will not likely produce anything useful. In fact, if there is a memory issue, then it most certainly cause a false hit.

OP, check the memory for errors. Windows has a tool built in, there's also bootable tools like Memtest.
Click to expand...

I am not talking about the page file.

https://www.thewindowsclub.com/pfn_list_corrupt-windows

https://win10faq.com/pfn_list_corrupt-error-in-windows-10-how-to-fix/

The PFN table does reside on the HDD. If the OP would have gotten the PFN_LIST_CORRUPT STOP, than I would say it's a memory error. But I guess it wouldn't hurt to try memtest either at this point.
 
craigdt said:
Currently running the Windows memory diagnostic tool

Again.

It shut itself off halfway through the second pass
Click to expand...

Hmmm, if you can't get it to run a complete test, try pulling one stick and see if it runs, or pull one stick, boot it up and see if it crashes, if it doesn't crash then you know you had a bad stick. try one at a time till you through your whole set.
 
Okay memory diagnostic says it detected no errors.

What is the next logical step?
 
Do you have another PSU you could swap in?
Double check all your connections as well.
 
I do have a cheap refurb evga $9 special I could try.

Any connection in particular to double check?
 
CAD4466HK said:
. If the OP would have gotten the PFN_LIST_CORRUPT STOP, than I would say it's a memory error. But I guess it wouldn't hurt to try memtest either at this point.
Click to expand...
Ummm... what? He specifically stated that he got one.


CAD4466HK said:
I am not talking about the page file.

https://www.thewindowsclub.com/pfn_list_corrupt-windows

https://win10faq.com/pfn_list_corrupt-error-in-windows-10-how-to-fix/

The PFN table does reside on the HDD
Click to expand...

Both of those links are incorrect about what the PFN list is, and neither make any mention of where it is.

Here's what MS says about what it is:
The Page Frame Number database contains lists that represent the physical memory pages of the system. The kernel uses the lists to track which pages are “in use” (allocated to working sets), free, available, and so on. This allows the kernel to quickly know which pages of memory are best to use or reuse. For example, if a page is needed for a process working set, then the kernel can get a page from its free list towards the allocation.
Click to expand...

and where it resides
It’s important to note that the PFN database resides in kernel virtual address space and the more physical memory [RAM] that a system has, the larger the PFN database must be.
Click to expand...
source: https://social.technet.microsoft.co...les/15259.page-frame-number-pfn-database.aspx

If you look further into it, you'll find that the PFN database resides in non-pageable memory because it has to be reachable when the system cannot handle a page fault (usually during ISR and DPCs). At this point, you're free to believe me or not. I won't address it again since it's of little value to the OP in solving his problem.



OP, open the Event Viewer and see if you can find any errors being logged there. Some help on how to do that can be found here.
 
craigdt said:
I do have a cheap refurb evga $9 special I could try.

Any connection in particular to double check?
Click to expand...

Not sure about that PSU, I would say check all your GPU and PSU connections to your drives and mobo are secure, reseat your memory as well.
Plus you could check your event viewer like ryan suggested and you could also try and run chkdsk to rule out the HDD.
 
ryan_975 said:
Ummm... what? He specifically stated that he got one.




Both of those links are incorrect about what the PFN list is, and neither make any mention of where it is.

Here's what MS says about what it is:


and where it resides

source: https://social.technet.microsoft.co...les/15259.page-frame-number-pfn-database.aspx

If you look further into it, you'll find that the PFN database resides in non-pageable memory because it has to be reachable when the system cannot handle a page fault (usually during ISR and DPCs). At this point, you're free to believe me or not. I won't address it again since it's of little value to the OP in solving his problem.



OP, open the Event Viewer and see if you can find any errors being logged there. Some help on how to do that can be found here.
Click to expand...

It's not that I don't believe you, but I have seen this error before and it was always the HDD.
Now granted, the errors I saw happened every 5-10min like clockwork. The OP's are weird as it's only happened once.
 
So I hooked up another power supply. Same result

Swapped memory sticks and in different slots.

Moved the machine to a different outlet and new power cord.

Removed the m.2 ssd and reseated it.

Removed the CPU and reseated it.

Its actually worse now.

Now it doesn't even stay on long enough to get to the bios screen.

Now what?
 
craigdt said:
So I hooked up another power supply. Same result

Swapped memory sticks and in different slots.

Moved the machine to a different outlet and new power cord.

Removed the m.2 ssd and reseated it.

Removed the CPU and reseated it.

Its actually worse now.

Now it doesn't even stay on long enough to get to the bios screen.

Now what?
Click to expand...

Did you run chkdsk? What does your event viewer show for errors?
 
CAD4466HK said:
Did you run chkdsk? What does your event viewer show for errors?
Click to expand...

I did not run chkdsk before my most recent update.

It does not stay on long enough to even boot.

Clicks on/clicks off within 5 seconds.

I did review event viewer previously... didn't see much. I did snap a few pics of a few things but really not much
 
craigdt said:
I did not run chkdsk before my most recent update.

It does not stay on long enough to even boot.

Clicks on/clicks off within 5 seconds.

I did review event viewer previously... didn't see much. I did snap a few pics of a few things but really not much
Click to expand...

I hate to say it, but I think you need to tear it completely down, set it up outside the case with just CPU, 1 stick of RAM and HDD.
Are you sure it's not overheating, causing a thermal shut down?
 
CAD4466HK said:
I hate to say it, but I think you need to tear it completely down, set it up outside the case with just CPU, 1 stick of RAM and HDD.
Are you sure it's not overheating, causing a thermal shut down?
Click to expand...

I was watching temps before swapping out power supply today and after running a light stress test, it never hit higher than 65.

I don't think it's overheating in 10 seconds or less, right?

Sadly, I think you are right about the complete tear down.
 
craigdt said:
Are these 2 seperats separate things?

I set the bios back to default a couple days ago.
Click to expand...

Are you sure it reset? Pull the battery and PSU off of it (unplug) and power cycle the power switch a few times than wait about an hour, then pop the battery back in and reconnect the PSU. Do this before you gut the rig.
 
Does this all look okay?

Are the pins on the motherboard supposed to be laying down somewhat or strait up and down?


L5PgJyK.jpg

3R9KgWk.jpg
 
The pins look good and should be laying like they do. The CPU looks fine except for 2 circular ones by your thumb. Not sure if it was the lighting or the way you're holding the chip but it looks like there is something on them.
 
deaedius said:
The pins look good and should be laying like they do. The CPU looks fine except for 2 circular ones by your thumb. Not sure if it was the lighting or the way you're holding the chip but it looks like there is something on them.
Click to expand...

Yeah- I saw that too.
It was like that when I removed it from the socket.

Looks like adhesive or something. Not like thermal paste.

Do I clean it off?

How?

Also-
I removed all the guts from the case.

Just the mobo, processor and ssd hooked to the power supply.

Still the same behavior. Rinse for a minute or less and shuts down.

Is the gook on the processor to blame?
 
You must log in or register to reply here.
Back
Top