[SOLVED] PC constantly restarts

[HolyWhatNow]

Hi, I'm hoping someone can help me with diagnosis of a pc that restarts after around two minutes every time it is turned on. It does get as far as logging on, but always resets a minute or two later. Any help would be greatly appreciated.

For a little background info the pc is about nine months old and has always been fully functional until about two weeks ago. I wasn't doing anything strenuous with it at the time, and hadn't downloaded or updates any programs, but there was a windows update a day or two before the problem started.

I'm about 90% certain it is a hardware issue of some sort, but one thing is throwing me off.

The only exception to this is that restart cycle is that for some reason the problem does not happen in safe mode. This initially made me think it was a software problem, but all my attempts to fix have achieved nothing.

I'll explain what I've already tried...


Software:

Performed system restore, and then when that didn't work, formatted drives and reinstalled windows onto a completely blank drive.



Motherboard:

Unsure how to check, but no visual damage, no lights on to indicate fault.

Processor light flashes around time of reset, but unsure if it's causing the restart, or if it's caused BY the restart.

HDD:

Fully formatted, and then tested in an old PC. Confirmed to work without issue. Confident in saying I can rule this part out.


CPU

More or less the only part I can't visually check. I have no spare thermal paste to reinstall, so once I remove it I cant do any further testing. So leaving this until ruling out other components.

Fans

All of the case fans as well as the CPU cooler fan are working, seem to be a good speed and generate a good breeze.


RAM

Has been removed, dusted, visually inspected, and then tested one stick at a time. No improvement.


PSU

Following advice and against my better judgement, I fully replaced the power supply, which had zero impact. Also confident this is not the cause.

GPU

Fans are not spinning, but as far as I know this is normal, as they won't activate until the GPU heats up more.

Peripherals:

All peripherals have been disconnected, and the restarting still occurs with all the usb slots empty.

Case:

Disconnecting the reset switch was the only suggestion I could find here, and it changed nothing.

Any suggestions on what I can try next?

 

[Deleted member 1272431]

Do you have a motherboard video out so to try and restart without the video card? If so please try.
I don't know which procedure you have followed to system restore but try booting in safe mode and uninstalling updates in Control Panel -> programs and features -> Uninstall a program if you find any updates starting with KB-XXXXXXX. If so, start from newest update and reboot each time to find the culprit.
Finally, I would try to disconnect power, remove the button cell battery in the motherboard and wait 24 hours then plug the battery and power back: this resets the BIOS.

 

[jay32267]

It sounds like you don't have enough time to run any programs correct?

I'd go in the BIOS and check whatever temps are there.

I'd also check the snugness of the cooler and make sure the fan is spinning.

...and please list your system specs.

 

[Schlachtwolf]

I would fresh install windows on a clean drive, my experience is if windows has a gremlin then it has it, start again. I think if you can boot in safe mode then it is a software problem not hardware. If you can get into BIOS disable fast start as that can cause problems after Win 10 updates, wortha try anyway.

 

[rgd1101]

malware scan the system?

 

[HolyWhatNow]

Thank you for your suggestions about the mobo reset, I will absolutely try that later today. I don't think it has a video out, but I will double check.

I seem to recall reading that the processer requires a discreet GPU in any case.

I don't have time to run, install, or download any programs before the pc resets, you're correct on that one. I did check temps in bios sorry, forgot to mention that. Nothing was above 27c.

I'm away from the pc at the moment, but parts are:

Win 10 64 home
Mobo: MSI b450 A max pro
CPU: Ryzen 5 3600x
GPU: powercolor Radeon rx 5700xt
Ram: patriot viper steel ddr4 c17
Psu 650w, seasonic, can't remember model off top of head.

And for those suggesting a windows reinstall... I did that already as mentioned. I completely formatted the hard drive, (100% empty) and created a fresh install media and reinstalled. I assume that would also account for malware since everything was deleted?

 

[Schlachtwolf]

Just a thought, is it plugged directly into a power socket or from a power brick, if it isn't then plug it into the wall socket ..... basic but just a thought...

 

[HolyWhatNow]

Basic thoughts are not bad thoughts! Unfortunately that was something I had already considered though.

Initially a brick, but it has also been tested direct into wall socket as well. Different rooms in the house too, in case there was a fault with the actual socket.

 

[Schlachtwolf]

Since PSU has been changed out I wonder if it is a Mobo issue, but it is really wierd that Safemode works fine but normal boot then crash. Have you updated the driver on your GPU before the issue started? Can you test it in another PC?
The reason I ask is in Safemode it may run on the generic Win 10 driver but in normal mode will load Adrealine or whatever AMD driver is installed. A long shot but sometimes they even hit. A faulty GPU can bring a plethera of issues not just no-display.

 

[HolyWhatNow]

That's why I was only 90% sure about it being hardware related. I don't understand either why a software issue would persist after a fully emptying the hard drive, or why a hardware issue wouldn't also occur in safe mode.

I didn't update the drivers manually recently, and as far as I'm aware they don't automatically update. Sorry if this sounds like a stupid question but in any case would the drivers not be stored on the hard drive and also wiped off during a format?

 

[Schlachtwolf]

Reasons fpr failure....

1. PSU , not your issue
2. Mobo, is a real possiblity
3. RAM, has been discounted
4. GPU, not likely but possible-test in another system
5. CPU, not likely but possible-test in another system
6. OS corruption, discounted as a clean install fail to reoslve said issue.

That is my summary on what is going on, is it correct?

Test CPU, GPU in another system from a mate or a local PC shop (they are often more reasonable than one would think), if they are fine then the Mobo maybe dead. My gut feeling is your VRM's on the Mobo are having an issue, overheating and shutting the system down, it is a good but also budget board and might have an issue.....

 

[HolyWhatNow]

Yes, I'd agree with your summary there. Unless of course there is any likelihood that both of the ram sticks died simultaneously, but that seems less likely.

I do have some updates now, but sadly no fix.

I borrowed an old compatible GPU from a relative and installed it into my own pc, and the exact same issue continued. So in a way it's good and bad news.

GPU doesn't seem to be the problem (so I don't have to replace the most expensive part of my pc) however no solution found as of yet.

Equally I tried the suggestion of resetting the CMOS (admittedly I used a jumper cap on the relevant pins of the mobo rather than actually removing the battery as described, but that was the instruction given in the manual) and found that this too did not change anything.

So realistically this means it is probably either the motherboard itself, or the CPU. As I lack the means to test them, that means ordering new ones.

Which brings me to what I hope will be my final question. What order should I replace them in?

That's both in terms of which is most likely to be a fix for the problem, but also risk.

i.e. can putting a new CPU on a faulty mobo damage the new CPU? Or vice versa, a faulty CPU damaging a new mobo?

 

[Schlachtwolf]

Mobo is more likely than CPU itself. If it was the CPU dead, Temps would be wierd and most likely not turn on at all. As i said above that it will run in safemode is a little give-away, Barebones loaded it is OK but under moreload the Virtual Memory controller or the Chipset is doing something it should'nt.

Did you update the Chipset driver recently, if not it is worth a try.... but my money is on the board.