PC is rebooting itself all of a sudden

[Soad]

As in the subject header. My PC keeps downpowering and rebooting itself after 10-40 minutes. So it sometimes crashes instantly, reboots on it's own, Windows loads. Other times it crashes and I must hold down the power button to turn it off and on again. It looks like it may be getting stuck when trying to recycle after the crash occurs. The Motherboard light is still lit up, I can hear the fans or something else drawing power but the system appears dead. GPU lights, mouse, keyboard etc don't power on when this happens.

I have no idea what the issue is. I tried running the software based run commands like chkdsk r, sfc scannow, DISM with no problems reported. Ran some cleaner software, did disk cleanup, did a system restore to a day or two earlier prior to the crashes. Yet it still randomly reboots. No recent app software installation changes either, so it has to be a problem with the hardware. I also ran some health checks on the WD Blue SN550's through the WD Dashboard and that reports nothing either. I also had MSI Afterburner and HW Monitor running and I see no obvious spikes in temps + usage, everything appears to be normal. That being said I don't understand the wattages and what not. The only change I have made in the BIOS is re-enabling XMP Profile1 after every bios flash. Everything is upto date when it comes to windows updates, bios, chipset from AMD website, audio, network etc with most of these updates being done a couple of months ago. The crashes / automatic reboots began happening a few days ago. All of these system reboots have occurred when idle or when browsing.

The only thing I have noticed is the chasse where my PSU is located seems to be making a popping noise upon every boot-up and continues making the same noise every 10 or 20 seconds. I can't say I have noticed this sound before until recently. It sounds like the sort of noise you would hear when a water cooler is pumping water or is on the verge of death and makes a faint noise. I can't be sure if that is the issue or not though, not the best person to ask when dealing with computers.

Current system specs

OS: Windows 10
GPU: Gigabyte Aorus 1080ti (purchased in 2017)
Display: MSI Optix MAG274QRF-od 27 inch 2560x1440p 165hz (purchased in 2022)
CPU: AMD Ryzen 9 5900x 12 core processor (purchased in 2021)
MOBO: MSI MAG B550 Tomahawk (purchased in 2021)
RAM: Team Group 32gb DDR4 2X16 (purchased in 2021)
Cooler: Alpenfohn Brocken 3 (purchased in 2021)
PSU: EVGA SuperNova G2 1000W (purchased in 2015)
Case: Lancool II Mesh Performance (purchased in 2021)
NVMe: WD Blue SN550 1TB installed on Windows 10. Secondary WD Blue SN550 installed too (purchased in 2021)
SSD: Samsung Evo 850 1TB

I have no idea how to identify the actual issue. I have no other spare parts and no secondary system to test things on. Even if I did I would still be woefully out of my depth. I even botched my Windows install a while back selecting CSM Legacy instead of UEFI because I couldn't find a selectable boot drive. I just don't have a clue.

If a part or two needs replacing what would be the best upgrade path? I mainly just browse and watch / listen to content but I do buy some of the latest gaming titles from time to time. Those storage drives for example are terrible when downloading or patching STEAM games. The download speeds continually fluctuate or halts the download process the second the disk usage hits 100%, happens almost every single time I download any large file through STEAM.

If it turns out to be the PSU or GPU at fault i will be forced into yet more replacements due to more dead hardware. I don't know what the standard warranty for purchases made through Overclockers is but if it's 2 years or more I might be able to RMA the purchases made in 2021 back to Overclockers.

So I may need a PSU or GPU replacement being specced for the system above.

 

[Soad]

I think I will begin this by replacing the PSU do you have any solid recommendations? As I believe the PSU may be at fault and the cause for the random reboots. The specs of my current rig are in the post above including the current PSU. I guess my questions regarding the PSU upgrade would be is another 1000W sufficient or will more soon be demanded? Also is a budget of around £100 enough for another quality PSU or is more expected in 2023?

I ran one of those built in Windows ram memory tests earlier on with no faults found by the way so I may be able to rule that out as a possible cause for the reboots.

I might be in the need for another NVMe as well by the way to replace both or one of the two existing NVMe slots.

I would also like to apologise for the previous post as I struggle to explain / describe things properly.

 

[Journey]

Have you made a note of the temps being seen on the CPU/GPU etc.?

 

[Pulseammo]

I think I will begin this by replacing the PSU do you have any solid recommendations? As I believe the PSU may be at fault and the cause for the random reboots. The specs of my current rig are in the post above including the current PSU. I guess my questions regarding the PSU upgrade would be is another 1000W sufficient or will more soon be demanded? Also is a budget of around £100 enough for another quality PSU or is more expected in 2023?

I ran one of those built in Windows ram memory tests earlier on with no faults found by the way so I may be able to rule that out as a possible cause for the reboots.

I might be in the need for another NVMe as well by the way to replace both or one of the two existing NVMe slots.

I would also like to apologise for the previous post as I struggle to explain / describe things properly.

Don't be so hard on yourself, it sounds like you're doing a good job trying to investigate your issue.

Did you try looking in "eventvwr" to see if the system is logging any faults? To check that just press the windows key on your keyboard and then type "eventvwr" and press enter. Expand the bit where it says "Windows Logs" on the left and then click "System". See if there is anything with an error or a warning marked there (although note that there will be some warnings which are mostly just benign and not very useful).

Asides from that you system is quite new, I don't know what the warranty is on it but there is a customer service section on the forum and they should be able to help you there. PSUs tend to last longer than 2 years though!

As for your NVME drive, I'd take one thing at a time and get the system to stop crashing first. It's extremely unlikely that any NVME drive of any generation can't keep up with your internet connection and it's more likely to do with steam unpacking game files.

 

[Soad]

Don't be so hard on yourself, it sounds like you're doing a good job trying to investigate your issue.

Did you try looking in "eventvwr" to see if the system is logging any faults? To check that just press the windows key on your keyboard and then type "eventvwr" and press enter. Expand the bit where it says "Windows Logs" on the left and then click "System". See if there is anything with an error or a warning marked there (although note that there will be some warnings which are mostly just benign and not very useful).

Asides from that you system is quite new, I don't know what the warranty is on it but there is a customer service section on the forum and they should be able to help you there. PSUs tend to last longer than 2 years though!

As for your NVME drive, I'd take one thing at a time and get the system to stop crashing first. It's extremely unlikely that any NVME drive of any generation can't keep up with your internet connection and it's more likely to do with steam unpacking game files.

The Event Viewer highlighted some critical errors at the times of the reboots, but I don't believe there to be any useful information provided in the reports, it was just stating that the system shutdown unexpectedly. In terms of errors just the usual TPM 2.0 etc (CSM Legacy botched install so I can't enable those Windows 11 required features without reinstalling the OS and finding a way to get UEFI running)

All of the 2021 parts were purchased through Overclockers, The PSU (2015) and the GPU (2017) carried over from the previous build though.

The NVME situation is quite commonly reported on the Internet. The WD Blue SN 550 lacks a dram cache so STEAM downloads are definitely affected in both the downloading phase and during patching. Download speed consistently drops to 0 when the SN550 reaches 100%. I also found some information a while back that WD made some adjustments that reduced the efficiency performance of the NVME.

I still believe replacing the PSU is the best bet to begin with but I have no idea what would be deemed a good replacement for my system. Hopefully you or someone else can recommend a good one.

I just hope it doesn't turn out to be a faulty GPU, CPU or Motherboard, as they are much more costly to replace, especially the GPU.

Have you made a note of the temps being seen on the CPU/GPU etc.?

Yes, nothing out of the ordinary, at least idly. GPU temps remained at 30c. The rest of the system were between 32 and 36c, CPU steady at 37c with occasional spikes upwards in the 50's when the usage increased, I don't monitor the temps too often, but when I first got the AM4 build I read on the internet that the 5900x is constantly adjusting it's clocks so higher temps idly or under low loads can be normal. Both the CPU fans and the Pump RPM's remained consistent to the usual norms.

 

[Werewolf]

I'm tempted to say possibly PSU related especially at 8 years old, although I'd expect that to happen more under load.

The fact you can hear something "popping" near the PSU is I'd say reasonably concerning, if you take the side panel off the machine can you definitely localise it to the PSU?

Do you have access to a spare PSU you can swap in for testing?

 

[Jamin280672]

I've got to ask, are you using PBO and what is you infinity fabric set at ? Did you build and setup the machine yourself of was it purchased and built by someone else, this kind of sounds like an idle crash to me, when people were pushing the curve optimiser and PBO hard, they would then stress test and pass all tests, but obviously that's full load, but then find that when the computer was seemingly doing nothing or next to nothing it would randomly crash.

The fix is simply to try for a while at stock settings or dial the settings back a bit in curve optimiser and PBO to see if it fixes things.

 

[Soad]

I've got to ask, are you using PBO and what is you infinity fabric set at ? Did you build and setup the machine yourself of was it purchased and built by someone else, this kind of sounds like an idle crash to me, when people were pushing the curve optimiser and PBO hard, they would then stress test and pass all tests, but obviously that's full load, but then find that when the computer was seemingly doing nothing or next to nothing it would randomly crash.

The fix is simply to try for a while at stock settings or dial the settings back a bit in curve optimiser and PBO to see if it fixes things.

If PBO is the game boosting feature that is selected by clicking on the CPU profile next to the XMP Profiles then no. Never touched any of the pre-overclocked profiles except for enabling XMP Profile1 for the 3600 advertised ram speed. If I did accidentally click something I shouldn't have in the past the Mflash BIOS updates should have reset everything back to the defaults anyway.

All of the parts were bought separately from Overclockers but put together by my Stepdad. He too thinks it could be the PSU at fault after taking a quick look through my case.

I played through Marvel's Spider-Man like 6 weeks ago and it was fine, never experienced any crashes, freezes or automatic reboots under load or idly until it began happening some days ago.

I'm tempted to say possibly PSU related especially at 8 years old, although I'd expect that to happen more under load.

The fact you can hear something "popping" near the PSU is I'd say reasonably concerning, if you take the side panel off the machine can you definitely localise it to the PSU?

Do you have access to a spare PSU you can swap in for testing?

I was able to isolate the sound back to the PSU but I can't be certain that is the cause for the reboots until I replace it. I will probably need some help with that, picking out the part that is,

 

[mickyflinn]

Mite as well check the memory to rule things out , download memtest free version.

MemTest86 - Download now!

Download your free copy of MemTest86 here for USB.

www.memtest86.com

 

[Mcnumpty2323]

if it wasnt for the noise you mentioned
i would always suggest looking for software causes first
by running windows in safe mode
or running a live cd/usb
actually even with the noise i probably would still do that
to see if it still happens or not
since trying stuff first that costs nothing is always preferable
could the noise be the psu fan? hard to know without
hearing it
i may be wrong but the supernova g2 may have 10 year warranty?
if you need to replace psu
my basic rule of thumb is look for 10 year warranty
its not uncommon nowadays even on reasonably priced psus
and i would say you dont need a 1000w psu
unless you suddenly want to add a much more powerful gpu
you shouldnt be needing 1kw
only using an 860w for my overclocked 2080ti and 5950x
plus full watercooling set up

 

[Soad]

if it wasnt for the noise you mentioned
i would always suggest looking for software causes first
by running windows in safe mode
or running a live cd/usb
actually even with the noise i probably would still do that
to see if it still happens or not
since trying stuff first that costs nothing is always preferable
could the noise be the psu fan? hard to know without
hearing it
i may be wrong but the supernova g2 may have 10 year warranty?
if you need to replace psu
my basic rule of thumb is look for 10 year warranty
its not uncommon nowadays even on reasonably priced psus
and i would say you dont need a 1000w psu
unless you suddenly want to add a much more powerful gpu
you shouldnt be needing 1kw
only using an 860w for my overclocked 2080ti and 5950x
plus full watercooling set up

I didn't try booting the system up into safe mode or let the system rest in the BIOS but I did run some of the built in Windows applications and software packages to see if any corruptions or faults could be found within the OS, NVME's, ram etc, all returned with no issues. My PC uptime was pushing 18 hours before the reboot occurred, since then it's rebooting at a tiny fraction of that uptime. Like I said in a previous post there have been no instances of system crashes, freezes or system reboots in the 2 years I have had this PC, also no recent software installations, system restore didn't solve it either, still rebooting constantly.

I believe the EVGA SuperNova G2 did come with a 10 year warranty but I don't believe I still have the receipts nor do I think I could wait 2 weeks to 2 months for an RMA return, + if that isn't the issue I have wasted a bunch of time. I would sooner just buy a replacement

EDITED SECTION

The historical history for my 2015 build or the 2016 SSD purchase through Overclockers no longer shows in my Order History View Orders so I can't even RMA the PSU even if I wanted to. The only history stored is my 2021 purchases.

I appreciate everybody doing the best they possibly can to save me some money and bearing with me with my badly detailed descriptions and lack of knowledge overall, but I do want to get a replacement PSU to rule it out once and for all, I would still need some help with deciding what to go with because I have no idea what's suitable for my rig. I just know I would like another good one.

I still think I would like another NVME too to replace the SN550.

Mite as well check the memory to rule things out , download memtest free version.

MemTest86 - Download now!

Download your free copy of MemTest86 here for USB.

www.memtest86.com

I already ran a Windows based memtest that rebooted the PC to test the modules, came back with no faults found. I did however try downloading the app you mentioned but the PC rebooted twice within an 11 minute period before I could run an actual test.

 

[Werewolf]

Re the order history IIRC overclockers changed their system in 2021, they still have a record of the sales etc and I think if you contact them via the stores webnotes or phone they should be able to get it (I believe the store staff have access to the archived orders from the old system*).

However given the age they probably won't be able to arrange the rma themselves, but would need you to do it, and for that that should be able to reissue you proof of purchase if needed.


*I don't know the details, but I suspect the old system was incompatible with the new one in terms of allowing customers access to its records.

 

[Mcnumpty2323]

you could probably try rma by serial number
If you can't get receipts
done that before
technically youre not legally required to have a receipt
though yes probably take some time For an rma
Worth trying even if you buy a new psu and turns out the psu is the issue


only worry is you go buy a psu
find out didnt fix it
so buy a new motherboard etc etc
ideally you need someone with some compatible spare parts

but basically any psu offering 10 year warranty
around 850/860w should be fine for what you need
example

Corsair RMx Series RM850x 80 PLUS Gold Fully Modular ATX Power Supply (CP-9020200-UK)

Order Corsair RMx Series RM850x 80 PLUS Gold Fully Modular ATX Power Supply (CP-9020200-UK) now online and benefit from fast delivery.

www.overclockers.co.uk

or if you want to future proof more then get another 1000w one
Same idea look for 10 year warranty
Corsair, be quiet,seasonic etc

For further future proofing
you would be looking for an atx v3 standard psu
though you need to clearly see it has all the required parts
as pretty sure they can label it atx v3 but still not have a 12vhpwr connector/cable
due to a loophole in the specs for atx v3 standards

 

[Soad]

I will probably order a MSI MPG A850G PSU over the weekend. Hopefully it's more than sufficient to replace my existing EVGA SuperNova G2 1000W.

MSI MPG A850G PCIE5 UK PSU 850W 80 Plus Gold Modular Power Supply

Order MSI MPG A850G PCIE5 UK PSU 850W 80 Plus Gold Modular Power Supply now online and benefit from fast delivery.

www.overclockers.co.uk

I might order a 2tb WD Black SN850X NVME aswell. Two things I would like to know before-hand though is does it have a dram cache? and is the lacking of the encryption a deal changer with or without it? To be completely honest I could probably get a 1TB instead, my current storage drives have over 500gb+ of free space on each one. The only things I install onto them is software applications and game installs from time to time. Would the Samsung 990 Pro be a better buy at 1 or 2tb? One last thing if I do purchase another drive would I reinstall the os onto it and use it for the vast majority of things including game installs? I am also fully aware that my Motherboard only supports 2 NVME's so one of the SN550's will have to be taken out.

WD Black SN850X 2TB SSD M.2 2280 NVME PCI-E Gen4 Solid State Drive (WDS200T2X0E)

Order WD Black SN850X 2TB SSD M.2 2280 NVME PCI-E Gen4 Solid State Drive (WDS200T2X0E) now online and benefit from fast delivery.

www.overclockers.co.uk

 

[Soad]

Just thought I would provide an update.

PSU replaced with the MSI MPG A850G but the rebooting still occurs at an alarming rate. PC is rebooting itself within a minute now. Powers down briefly then comes back on. My Stepdad initially had some trouble getting the PSU to power on as it appears one of the fan hub wires or something was short circuiting it if I recall what was said properly. That was replaced and the PSU powered up, the old EVGA Supernova G2 booted up under both conditions. The popping noise also disappeared with the MSI A850G PSU.

The GTX 1080TI was also removed from the system temporarily with an AMD Radeon 290 but the reboots still persist.

I honestly have no idea what the issue is. If the bulk of my system specs weren't recent I'd just abandon it and leave the PC behind, it stresses me the hell out whenever a problem presents itself.

I checked the standard warranties for the CPU and MOBO and it says 3 years. The RAM says 10 years manufacturer guarantee but I have no idea how to to actually go about identifying the faulty product.

I also went ahead and purchased the SN850X 2TB but attempting a fresh install of the OS is pointless at the moment.

 

[Mcnumpty2323]

Either take it out the case
And breadboard it
Connecting only the absolute essentials
Cpu fan,ram,gpu
24 pin and 8 pin cpu power and gpu power
No usb headers,no case fans,no controllers etc

Or do the same inside the case
Though that way won't rule out some sort of short
Against the case

Boot it immediately into bios and leave it a while
Check cpu temperature
If it doesn't crash in bios try booting into windows
By connecting only the windows drive
If that works gradually add 1 thing at a time like front usb etc

Edited
Set bios to defaults
Don't enable xmp either to begin with

 

[Journey]

Happy to look at it for you, should you live close by.

 

[Mcnumpty2323]

Yeah if the op doesn't mind saying
Where they are
More than likely people would be willing
To get hands on to help them
Especially useful if they also have compatible parts
To double check stuff
As I mentioned earlier you can go down the rabbit hole
And end up buying part after part otherwise

 

[Soad]

I would rather not disclose my location. I'm nowhere near the Midlands area anyway, but thanks for the offer.

I or well my Stepdad followed your suggestions by the way Mcnumpty, something interesting happened actually.

Boot disk error prompt screen suffered from regular reboots. The actual BIOS was on for over an hour with no reboots. BIOS was then reset to the default configuration, still no reboots in the BIOS. Re-organised the boot drives prioritising the OS drive and the reboots reappeared when loading into Windows, crashes and screen freezing also occurred. Drives formatted and removed from the system, New SN850X NVME added with a fresh install on UEFI, rebooted once and crashed once during the Windows install process. We eventually got into Windows but the regular reboots and now hard crashes have continued. Windows is upto date and I downloaded the essential drivers such as Nvidia GPU drivers, AMD chipset etc. The only peripherals connected were a monitor and keyboard. Then the mouse and an unused ethernet cable were added as it was tested in the living room. Enabling or disabling XMP didn't make any difference. The CPU temperature in the BIOS screen was 38c. It's definitely not a temp issue.

This was tested on Sunday. The PC has just been dumped there ever since. My Stepdad suggested RMA'ing or buying a new MOBO but he can't guarantee that's the fault so it's either time wasted or money lost....

 

[Mcnumpty2323]

Did you do it
With the components still in the case?
If so next suggestion would be remove from the case
And breadboard it and test again
Bit of a PITA but if next step is new motherboard
Will need to remove most of it anyway

Yeah might be motherboard
More likely than cpu to be faulty
But yeah your step dad is correct can't guarantee it

And yeah 38c after sitting in bios for a while is fine

Does that motherboard have
Any sort of post code display or leds for ram,gpu,cpu,storage faults?