[SOLVED] Computer randomly shuts down and requires me to turn off the socket until the lights die and then turn it on again to power it up again.

Aug 27, 2019
8
1
15
I have a custom build:
CPU: Ryzen 2700X with the Wraith Max (no overclock)
GPU: EVGA 980Ti SC (no overclock
RAM: HyperX Predator 16GB(2x8GB) 3200MHz (Only clocked to 3000MHz)
MOBO: ASUS ROG STRIX B450-F
PSU: Riotoro Enigma 850WG2

I've checked temperatures and the CPU idles around 30-40 and same with the GPU. It was only till yesterday it started to randomly shut down. I thought it was because of overheating but I left the side panel off and my room's temperature was 18ºC so it was cool in the room.

At first, I checked the fans and they were all spinning, the GPU wasn't for some absurd reason so I went into EVGA's precision tool and made it follow a cooling curve and it was spinning as it should. The PSU had some weird thing where off of hybrid mode the fan wouldn't spin, but on hybrid mode, it would(?) The CPU cooler is cooling as it should. I had realised the backplate of the GPU was overly warm on Rainbow 6 Siege.

I've updated BIOS to the latest version. All motherboard drivers are up to date. All graphics drivers are up to date.

Nothing on my build is overclocked except the RAM because it was the default clock at 2400MHz and I kept it at 3000MHz just to be safe.

Before yesterday it had never shut down. I was just on Minecraft when it randomly shut down, and it wasn't a subtle shutdown, it was as if the power cable was disconnected. Today it has happened 3-4 times within 2 hours of me trying to play R6S.

I haven't checked anything else as I am not sure about what to do so any advice would be really helpful.
 
Solution
Welcome to the forums my friend!

Does the issue only occur under load?
Being as you have already covered BIOS / drivers etc.

Usually the running order of this kind of stuff can be:
PSU > GPU > RAM > MB > CPU.

Have you got any components you are able to swap and retest?
I would also disable the RAM OC and see if the issue persists. It may also be worth running 1 stick at a time and seeing if you can recreate the shut down.

PC Tailor

Illustrious
Ambassador
Welcome to the forums my friend!

Does the issue only occur under load?
Being as you have already covered BIOS / drivers etc.

Usually the running order of this kind of stuff can be:
PSU > GPU > RAM > MB > CPU.

Have you got any components you are able to swap and retest?
I would also disable the RAM OC and see if the issue persists. It may also be worth running 1 stick at a time and seeing if you can recreate the shut down.
 
  • Like
Reactions: m4jestic
Solution
Aug 27, 2019
8
1
15
Welcome to the forums my friend!

Does the issue only occur under load?
Being as you have already covered BIOS / drivers etc.

Usually the running order of this kind of stuff can be:
PSU > GPU > RAM > MB > CPU.

Have you got any components you are able to swap and retest?
I would also disable the RAM OC and see if the issue persists. It may also be worth running 1 stick at a time and seeing if you can recreate the shut down.

Thanks for the welcome!

It seems to happen under load as it's only been seen on Minecraft and Rainbow 6 Siege. I don't have any components to swap. I will attempt to see if it happens with no RAM OC, if it still does I'll attempt it with the individual sticks

But generally I load into a game in R6S and it'll be fine in the observation phase, but when it's taking action, within 2-5 minutes it'll just shut down.
 

PC Tailor

Illustrious
Ambassador
Sounds like a plan - it's an easy way of testing.

Following that it may be worth running Furmark to see if you can recreate the error when inducing load on the GPU. I'd still go by my running order of the above - just based on what you've covered already, I'd be testing what you can easily as far as possible.

PSUs and MB are hardest to test. Although in plenty of cases random shutdowns are indicative of failing PSUs.
 
Aug 27, 2019
8
1
15
Sounds like a plan - it's an easy way of testing.

Following that it may be worth running Furmark to see if you can recreate the error when inducing load on the GPU. I'd still go by my running order of the above - just based on what you've covered already, I'd be testing what you can easily as far as possible.

PSUs and MB are hardest to test. Although in plenty of cases random shutdowns are indicative of failing PSUs.

Update: I was going to change the RAM OC back to Auto but now the PC won't boot up at all. There's only the RGB light on the top left of the MOBO

Update 2: Also should I have only one connected with a splitter head go to my GPU, or have 2 separate cables that goes to the GPU. It's a 8pin+6pin connection.

Update 3: I took out the RAM individually to see and it still didn't boot. Decided to take out the GPU since it was the next easiest thing and the main concern of issue. The PC booted up but I can't check anything without a graphics card since my CPU doesn't have integrated graphics. Do I test it with another GPU if I can get my hands on one, or do I attempt to power the GPU with 2 separate cables from the PSU since it's modular?
 
Last edited:

PC Tailor

Illustrious
Ambassador
Update 2: Also should I have only one connected with a splitter head go to my GPU, or have 2 separate cables that goes to the GPU. It's a 8pin+6pin connection.
All GPU power ports have to be connected if that's what you're asking.


Update: I was going to change the RAM OC back to Auto but now the PC won't boot up at all. There's only the RGB light on the top left of the MOBO
Which RGB?
View: https://imgur.com/a/foQYV6P


Update 3: I took out the RAM individually to see and it still didn't boot. Decided to take out the GPU since it was the next easiest thing and the main concern of issue. The PC booted up but I can't check anything without a graphics card since my CPU doesn't have integrated graphics. Do I test it with another GPU if I can get my hands on one, or do I attempt to power the GPU with 2 separate cables from the PSU since it's modular?
When you say 2 seperate cables, is the GPU not fully populated with power connections?
 
Aug 27, 2019
8
1
15

PC Tailor

Illustrious
Ambassador
The case has the ROG logo with a rainbow RGB effect
Well it's you said a debug light lit up, I was wondering which one from the image I posted above?

All of them are connected but it's only using 1 wire with a splitter end.
In which case that should be fine. Apologies for misunderstanding initially.

So removing the GPU allowed the PC to boot up normally? This would go back to my first point that it could be PSU > GPU. Only way to really test with guarantee is to try a different GPU in your system and retest or try it in someone elses system with a good quality PSU.
 
Aug 27, 2019
8
1
15
Well it's you said a debug light lit up, I was wondering which one from the image I posted above?
Oh no it's not a debug LED it's this that was lighting up:
View: https://imgur.com/gallery/7pG8v8G


In which case that should be fine. Apologies for misunderstanding initially.

So removing the GPU allowed the PC to boot up normally? This would go back to my first point that it could be PSU > GPU. Only way to really test with guarantee is to try a different GPU in your system and retest or try it in someone elses system with a good quality PSU.

Yeh, removing the GPU did allow the PC to boot, the fans were all spinning.
My mate has decided he'll lend his GPU for testing and I'll test my GPU on his system but the earliest date he's available is Friday so hopefully I can work out if it's GPU or PSU.

Thank you so much for the help.
 
  • Like
Reactions: PC Tailor
Aug 27, 2019
8
1
15
So removing the GPU allowed the PC to boot up normally? This would go back to my first point that it could be PSU > GPU. Only way to really test with guarantee is to try a different GPU in your system and retest or try it in someone elses system with a good quality PSU.

So I tried another graphics card and my system works now, but it's a 1050Ti so it doesn't require additional power from the PSU. I have yet to test my GPU on another build so hopefully, I'll find out sometime soon.
 
Aug 27, 2019
8
1
15
Welcome to the forums my friend!

Does the issue only occur under load?
Being as you have already covered BIOS / drivers etc.

Usually the running order of this kind of stuff can be:
PSU > GPU > RAM > MB > CPU.

Have you got any components you are able to swap and retest?
I would also disable the RAM OC and see if the issue persists. It may also be worth running 1 stick at a time and seeing if you can recreate the shut down.

It's definitely the GPU as I have tested it on another build with similar specs to me and that PC wasn't booting up.
 

TRENDING THREADS