Computer randomly restarting without warning

Mr. CooL ICE

My desktop has recently started restarts at random. It first happened while gaming but has also happened while doing some benign things like browsing. Here are some troubleshooting steps I took.

• Nothing listed for the restart in event viewer, aside from the error that the computer had rebooted without being shut down after the restart. Nothing to indicate the cause of the restart. Turned off 'automatically restart' under System Failure from Advanced System Settings but no joy.
• After some of the restarts, I went into the BIOS and saw no signs of high temperatures anywhere. Considering CPU/GPU could be high/low prior to a restart, I concluded a CPU/GPU temperature cause is unlikely.
• I ran the mdsched.exe and it always restarted before the test completed. There were no results listed in the event viewer after the test. I checked my RAM for any visual faults and tried switching slots, but no difference.
• Tweaked various things, such as turning down RAM frequency in the BIOS and changing power mode from performance, etc. Again, no joy.

PC is a self build that has been added to over the years. My current specs are as follows:
AMD Ryzen 5 1500X 3.5Ghz
Asus ROG Strix X370-F Gaming Motherboard
Corsair 16GB DDR4 3000Mhz RAM
Asus ROG Strix GTX970
Corsair CX600M PSU

I'm veering towards the PSU being the cause of the restarts. The restarts have no discernible pattern. They even happen while I'm viewing the BIOS. The PSU is a bit old and after a bit of a google, I've read a few bad reviews of the Corsair CX series in general. Also, the utter lack of anything in the event viewer makes me think the CPU being the culprit makes sense.

Is there any other likely cause I haven't thought of?

mordeith

Have you run an actual RAM check? I know you have done a physical check but have you run a diagnostic or tried it with one stick only?

Charlie19

RAM was my firsts thoughts too.

Mr. CooL ICE

Yes, mdsched. It's the windows diagnostic tool. Tried a good few times and it never got a chance to complete as the restart kept happening during the run. And always at different times during the run, i.e. after 2 minutes or 20 minutes. And as it never completed, I never got results.

Tried the same thing with Memtest86 and exact same thing, I e. Never gets the chance to finish and restarts at any time.

FanadMan

Have you reseated/switched around the RAM? One module.....try it in another slot?

€10 or so for a tube of thermal paste and you can reseat the CPU.

Grab All Association

PSU. I’m fairly certain that’s the same/similar to the PSU I had in a build recently that exhibited the same behaviour. I took a PSU out of an older doner Dell XPS, replaced its and it stopped happening.

Mr. CooL ICE

FanadMan wrote: »

Have you reseated/switched around the RAM? One module.....try it in another slot?

€10 or so for a tube of thermal paste and you can reseat the CPU.

Just a single slot, unfortunately. Would have to find someone with a same spec ram piece to test it against.

Why thermal paste though? Nothing indicates overheating. It's low CPU utilisation prior to restart and BIOS temp levels are normal afterwards. It's not like it restarts exclusively when running a game or compiling code.

FanadMan

Mark Rippetoe wrote: »

Why thermal paste though? Nothing indicates overheating. It's low CPU utilisation prior to restart and BIOS temp levels are normal afterwards. It's not like it restarts exclusively when running a game or compiling code.

Very late night drunken diags Which is why I've stopped fixing phone problems in the pub.

Didn't properly read your prev posts. It does sound like a dodgy PSU. But would reseat ram/cables/cards before buying a new one.

Fingers crossed it'll get sorted. Sorry for drunken rambling.....I'm not normally like this, honest

Mr. CooL ICE

Ah jaysus, you're grand. All help is appreciated. I'm up with a teething baby and sleep deprivation makes figuring this stuff out difficult.

I'd normally have subbed out suspect hardware myself before asking for help, but I don't have any spare RAM or PSU so sourcing some is a bit of an ordeal.

However, I remembered I have my previous CPU/motherboard/RAM stored away here so I can hook that up to the PSU to try confirm if it's the issue.

Mr. CooL ICE

Hooked up my old mobo/CPU/RAM to the PSU and let it run yesterday evening... 2 hours of running Memtest86* and not a single restart. Thought I was back to the drawing board. Did the same thing this morning and 4 restarts within an hour. New PSU has been ordered.



*Used this for the sake of running something bootable and not cause any hassle with my current install of Windows.

K.O.Kiki

I'd order a dual-channel RAM kit just for the sake of it.

Not only can you then discount RAM being bad, but it's also a boon in games - some games can run as bad as 50% worse with a single stick.

Also RAM is dirt cheap right now (2x8Gb DDR4-3000 CL15 for €70), and due to rise later in the year.

TheBoyConor

Some psus have a built in self test with a procedure to follow to check if they are working. I found this with my Dell Optiplex which was giving me the "orange light of death" led error.
Check your PSU documentation to see if there is a test procedure.

Homelander

K.O.Kiki wrote: »

I'd order a dual-channel RAM kit just for the sake of it.

Not only can you then discount RAM being bad, but it's also a boon in games - some games can run as bad as 50% worse with a single stick.

Also RAM is dirt cheap right now (2x8Gb DDR4-3000 CL15 for €70), and due to rise later in the year.

I was also going to say this. A single stick of 16Gb will cripple you in some games.

Vicxas

Sounds like a bad rail in your PSU, i had the same problem as you with my desktop years ago. It would run fine then all of a sudden at random it would restart with no errors in the logs.