Help Needed with PC Restarting at Random

Hi Ozbargain, calling all the PC experts to get some assistance. I have this issue where my PC will just randomly restart. It can happen within a few hours or last up to 6 days but at some stage it will just restart. I run the PC 24/7 running some VMs. No resource issues and CPU usually sits at 20% and utilising about 32 GB of ram. I don’t run any games and I’m confident there are no overheating issues.

The error is always the same and that is: Event ID 41 – Kernal Power - The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

Generally no other real errors in the logs that point to any issues, this has been ongoing for over 6 months.
Specs are as follows:
Windows 10 Pro 20H2 build 19042.746
Motherboard: Asus Prime X570-P with the latest stable Bios 3001 (no overclocking for anything)
CPU: AMD Ryzen 9 3900X
64 GB RAM Corsair CMW32GX4M2C3200C16 Vengeance RGB PRO
Video Card: Radeon RX570
PSU: Corsair AX1200i
C:\ Crucial 500GB NVME CT500P1SSd8 and have another 3 drives installed

What have I done to troubleshoot and resolve the issues and in no particular order:
• Always update Windows to the latest version
• Always have the latest chipset and drivers for all devices
• Updated bios many times and am on the latest stable release
• Set all the power settings in windows to high performance and don’t allow any device to turn off etc.
• Turn off fast boot
• Originally had RX580 video card and then installed a Radeon RX570 to confirm that there were no issues with the video card. I have also flashed both
cards with the original BIOS.
• I have played around with the BIOS settings for the RAM and turned on DOCP and then back off again. All other settings are stock
• Run memory diagnostics in Win 10 and no errors.
• Updated the firmware for all SSD drives I have installed.
• Run diagnostics for hard drives with no errors.
• Run Check disk and windows SFC
• Have re-seated the RAM modules
• Have re-seated all the power connectors from the motherboard
• All devices are running from a UPS
• Have changed UPS
• Have changed the power cables
• Have run OCCT and at times it will run fine with no errors
• Sometimes OCCT will crash the computer and it will become unresponsive whereby there is no display on the monitor and I can’t RDP to it so only option is to hard reset. Cannot get logs when this happens.

I’m sure I’ve done a bunch of other things that just don’t come to memory at the moment.
Please help in resolving this issue which has been plaguing me for some time.

Thanks in advance for all your help.

Comments

  • +3

    Run a stresstest with AIDA64 https://www.aida64.com/ - I bet you it’s probably the PSU on it’s way out if you get a sudden reboot especially when if it suddenly reboots when it gets up to the graphics card testing part.

    • I had a problem that was pretty much the same, drove me insane, lucky I had a spare PSU, problem solved

      • +1

        I will try the stress test tomorrow to see if it reveals any issues.

        • +1

          I ran the stress test for around 90 minutes and all worked fine. I think I am going to bite the bullet and replace the PSU this afternoon.

          • @Ford351: I have replaced the PSU and will post an update on how it goes. Downgraded from a $400 Platinum PSU to a $150 Gold PSU.

  • +4

    Is the PSU stable? Do you have any brownout issues, as in, does this happen maybe when an appliance in the house starts, fridge starts pumping cold air washing machine starts, oven?

  • +2

    As above, the PSU dude. Replace it. You've pretty much ruled out everything else. If it's still happening with a brand-new PSU, then the motherboard would be the only other explanation.

  • +1

    Power supply faulty.

    That error code is literally "someone switched me off".

  • I have a spare PSU I can take out from an old PC to test it. I don't think I have any brownout issues as it is running from a UPS and I have 3 phase power and there is nothing I can pinpoint when this happens. The current PSU which is platinum rated by Corsair says it has a 7 year warranty on them, is there anyway to prove it is faulty to hopefully get an exchange?

    • AXi Series PSUs have a 10-year warranty taken straight from their warranty page

      https://help.corsair.com/hc/en-us/articles/360033067832-Cors…

      When you warranty the psu, you have to be very careful with your language or wording of the issue, if indeed it is the psu, then you should say its an intermittent fault.
      Intermittent faults are hard to prove, so they generally just take your word for it.

      Explain the problem exactly, i would say computer constantly reboots unexpectedly usually within 1-3 days, replaced power supply and stopped having the unexpected reboots.

      btw OCCT crashing actually points to the cpu being the problem, however you can be in the unfortunate situation of having more than 1 part being faulty.

  • If not any of the above, perhaps check CPU temps (does not seem to be on the troubleshoot list)? Or does the random restart also happen when not under load?

    • CPU temp is at 53 degrees at the moment, the PC is never under heavy load. It just restarts at random times.

      • Sorry. Posted before seeing your post above. Thought you might have already checked the PSU (based on kernal power message). If you have time, you might want to try swapping out with the spare PSU. For me it was the power cable rather than the PSU that was the problem but see you have already swapped that.

  • You got some really good parts. imagine if you had to use your PC for work stuff every day, rendering stuff, and it kept randomly restarting. What a pain in the ass. The 20 hours you spend scouring the internet for answers could have been 20 hours spent working instead. Your troubleshooting time could easily end up costing over a thousand dollars in lost productivity, just to get your PC to turn on properly.

    • As someone who works in this field, your emergency recovery plan is too poor for your scenario then.

      If 20 hours of downtime can cost you "thousands" then a second machine at the cost of, say, $3k is extremely cheap insurance.

      Also, the parts are powerful, but not extremely redundant, single psu, non ECC ram, etc.

      It's a very good machine for sure, but your scenario has someone with no business sense behind the wheel, or someone who neglected their emergency plan when forming their business.

  • +1

    My brother's computer did this for a while too, turns out it was the restart button. Try disconnecting, see if that helps?

  • +1

    I have / had a similar problem with my desktop PC except it was even more frustrating.

    The PC would progressivly freeze. First the mouse would stop working. You could use the keyboard to CTRL+S, enter text, ALT+TAB to other programs or CTRL+SHIFT+ESC to Task Manager. Then after a minute or so the keyboard would stop working. Task Manager would show activity for another minute and then nothing. The screen would stay on frozen. After 5 minutes PC would reboot itself.

    I tried every test I could think of - after 30 years in User Support I can think of quite a lot - but I could never find anything wrong.

    Soon after Covid started I was helping a friend who owns several office buildings clear out one of the offices that had been abandoned by an ES company that decided to close down their WA outpost. They said he could have everything in it for $4000 which he accepted.

    There were:

    • Partitions and built-in desks
    • Conference room with large table and chairs
    • Office chairs
    • 2 very large central colour lasers
    • Kitchen equipment
    • Half a dozen laptop and desktop PCs
    • Monitors and risers
    • Keyboards and mouses
    • USBC to HDMI adaptors
    • Pocket audio recorders
    • 3 data projectors

    I assessed the specs of the PCs.

    I booted the PCs off a USB drive. Fortunately, (because we didn't have to worry about what to do with the data) they had all been wiped before the office was abandoned.

    My friend said I could have whatever I wanted. I was tempted by the full size A3 laser but I decided I didn't really need it.

    I took the highest performance laptop, a monitor, an office chair and a data projector.

    I never solved the re-boot problem. I started using my new PC and added the problematic desktop to the dozen or so PCs I have in the office (what my partner calls the junk room). I have PCs starting with ones with 5.25" and 3.5" floppy drives - you never know when you might want to read one of those. I even have an IBM MCA server running OS/2 - it hosted a software (DisplayWrite4, Lotus 1-2-3 and dBase III) sharing menu system I wrote in the late 80s. I fire it up every couple of years to marvel at what I created.

    When I setup new PCs for people they sometimes offer me the old one. My partner says "They're not being nice to you. They're using you as a rubbish disposal service."

    My response is that you can never have too many PCs or other computing devices.

    From where I'm sitting I can see:

    • 2 laptops on the dining table - hers and mine.
    • Her recipe Android tablet on kitchen bench
    • Desk with my "new" PC with 2 monitors - my main work PC.
    • 1 laptop on side table by TV chair
    • 2 abandoned Windows tablets and an iPad behind TV.
    • 2 Android phones
    • 2 Google Nest devices

    The PCs upstairs in the office run a variety of OSs of various vintages.

    The real point of this story is to ask if anyone has any suggestions as to what could be causing a PC to progressively freeze?

    • Sounds like the motherboard is toast. Another thing to check is the smart status on the hard drive which holds the boot OS. Vintage PCs are awesome, I shudder at all the ones I put out for garbage collection and got scrapped, but what can you do. At one stage our garage had 20 computers with their crts and everything, all spec’d out for windows 98 and a couple of iMacs and a Macintosh SE

      • +1

        At one stage our garage had 20 computers with their crts and everything, all spec’d out for windows 98 and a couple of iMacs and a Macintosh SE

        Damn man. Pressing 'F' for your retro PC collection.

        Every time I watch LGR's videos I always think why didn't I hang onto those old DOS/Win 95/98 boxes of mine from back in the day; it would have been great to restore and upgrade those with some beefier period-accurate parts and make a vintage gaming machine to run older games on.

        I looked into ordering some 90s PCs that matched the ones I used to have from eBay and other places, but the ones I was looking at were mostly overseas and shipping absolutely kills it.

        • The going rate seems to be $250 now on eBay for local stock, and that’s not even with 3DFX cards. I’ve never seen a 3DFX card ever, only the pretty shithouse S3 Virge cards, then the upgrade to AGP and nvidia coming on the scene

          • @DemocracyManifest:

            The going rate seems to be $250 now on eBay for local stock, and that’s not even with 3DFX cards.

            What kind of specs and OEMs are we talking? Sub-100Mhz Pentiums with 8MB of RAM, 1GB HDDs and all the peripherals included (CRT, mouse, keyboard, speakers, etc)? IBMs, Compaqs, Apple?

            I’ve never seen a 3DFX card ever, only the pretty shithouse S3 Virge cards, then the upgrade to AGP and nvidia coming on the scene

            They never caught on much in Australia compared to the US; plus they were only around for 6 years before Nvidia gobbled them up, not much time for them to really market much outside the US or establish wider distribution channels. My only exposure to 3dfx Voodoo cards were their ads gaming/computing magazines back in the day, I don't think I ever actually came across a PC running one during the entirety of the 90s.

            The GPUs I went through in the 90s were an ATI 3D Rage Pro, some horrible SiS integrated graphics chipset, a Diamond S3 Virge and an Nvidia Riva TNT of some kind. I think I had a laptop with some kind of Realtek GPU as well, back when they used to make 2D/3D video cards.

            • @Miami Mall Alien: Have a look yourself it’s not very specific tbh, win 98 PCs with SSDs are pretty popular. I forgot about those ATI and NVIDIA models

  • Unfortunately last night the PC restarted again with the same error. I have disconnected the restart switch as recommended here as well and will see how it goes. Does this point to the Motherboard being faulty?

    • So you've got a brand-new PSU in there now?

      It has to be the mobo.

      • It’s not brand new as I removed it from an old PC that I have but have never had any issues with it so I’m confident to rule out the power supplies now.

  • What power level psu do you have. Mine had an issue where it was not high enough and when my pc wanted to draw larger power load. Ie.. play a game or have all hard drives active it would shut down. I went up to a more powerful psu

    • The corsair PSU is 1200 watts so it had plenty of power and the other one I am testing is 800 watts which also has more than enough power for the system.

  • It's restarted again. I will pursue a warranty claim on the motherboard. I purchased the motherboard from Futu online on ebay, should I pursue them or go directly to Asus? Anyone have experience with dealing directly with Asus? Thanks for your help.

    • Don't bother chasing up an RMA from the manufacturer directly; always let the retailer handle it if the product is within warranty. They should be able to refund or replace the mobo upon receipt and after performing testing to confirm it's faulty.

      That's how ACCC consumer protections and statutory warranty is supposed to work and it's far less hassle, generally speaking.

      OEMs will generally want you to pay shipping both ways and will take 2-3 months at best to fulfill an RMA, especially if there's no Australian-based repair centre/authorised service partner. They may also deny your RMA claim for a variety of reasons or only offer a refurbished part as replacement.

      • Thanks - I've contacted Futu so hopefully it goes well.

    • +1

      Anyone have experience with dealing directly with Asus?

      Yes. ASUS Australia doesn't want to know you. Contacting your retailer is the way to go.

  • Hi all, I wanted to give everyone an update. I sent the motherboard back to Futu Online and after 3 weeks I received the MB back on Friday. I can firstly say that Futu Online have been very response and have provided updates throughout the process and can't fault them. The MB was shipped to me with a resolution of "Update BIOS can solve the problem" and this was from ASUS. So Friday evening I assembled the PC back again. The BIOS was the same version as I had sent to them which was 3001. Again this morning Sunday 14/02 the PC restarted with the same error. It appears to me that nothing was really done by Asus. Asus have released a new version of the BIOS and Chipset drivers so I have updated and will see how this goes. I wrote back to Futu Online saying what a disappointing outcome and if this continues I will request a full refund or replacement MB.

Login or Join to leave a comment