[SOLVED] New build PC loses power, stays out for a certain amount of time, comes back online after that itself

Aug 7, 2020
5
0
10
Hello everyone,

I've saved and spent a lot of money on building a dream system for myself. Tuesday was the day I've had everything together and I was finally able to build it.
Everything was completly new, except the graphics card, that I already bought over a year ago for my old system and the old drives I used. It took quite a while but then everything worked kind of well at first.

After finishing the build i tried to boot it for the first time and it worked without any problems. I did some more cable managing after that, closed the case and it wasn't starting. I thought, maybe I must have touched the powerswitch cable and thats why it wasn't going on. So I opened it again, doublechecked the powerswitch cable and tried again, this time it worked. I thought that must have been it. I've then continued with installing drivers for the graphics card, browsers, etc. and did some stresstesting on GPU and CPU to look out for the temps and everything was turning out perfectly. I also changed the RAM MHZ speed in the bios to 3200 mhz to get the most out of it.

Now, seemingly at random times, my computer just turns completly off. I can't really recreate it but I think that most of the time I've had many programs that worked seperatly open at the same time. Like a game, Teamspeak and a Browser with a video. First thought was: Is it temperature? - Checked it, no its not. The graphic cards never really gets higher than 75 degrees and the CPU is below 62-63 degrees at all times.
It happens 1-3 times per day, just goes black, stays black. I can't turn it back on myself if it happens. And then takes like 1-15 seconds to restart itself without me doing anything. At one time it turned on till before the "Press F2 for Bios screen" and then went black again three times before finally booting correctly. At first I suspected that the CPU might be taking too much power because I only connected an 8 pin cable to the motherboard for it but now I'm not sure because other people said something would be happening then. I have no second PSU to check if it is faulty sadly.

Any helps, ideas or input would be highly appreciated, because I do not know how to fix this problem on my own. Thanks in advance for everything you can provide!

Specs:

Motherboard: Asus ROG STRIX Z-490-F GAMING
CPU: Intel Core i9-10900K
CPU Cooler: NZXT Kraken Z63 280mm
GPU: MSI GeForce RTX 2070 SUPER GAMING X TRIO
PSU: be quiet! STRAIGHT POWER P11 Platinum 1000W
RAM: Patriot DIMM 64 GB DDR4-3200 Kit
Case: be quiet! DARK BASE PRO 900 rev.2
 
Solution
I think GamersNexus (Youtube tech channel) commented this week on a prebuilt system by NZXT that also had the 20+4 pin the wrong way around, the 4-pin above the 20. It's never going to stay there. Quality control, so so.

I don't know what the 4 pins do but every ATX motherboard has them so they must be important.
Looked it up:

ATX12V 2.x

"The ATX motherboard connector was extended to 24 pins. The extra four pins provide one additional 3.3 V, 5 V and 12 V circuit. "

Source: https://en.wikipedia.org/wiki/ATX
Hello everyone,

I've saved and spent a lot of money on building a dream system for myself. Tuesday was the day I've had everything together and I was finally able to build it.
Everything was completly new, except the graphics card, that I already bought over a year ago for my old system and the old drives I used. It took quite a while but then everything worked kind of well at first.

After finishing the build i tried to boot it for the first time and it worked without any problems. I did some more cable managing after that, closed the case and it wasn't starting. I thought, maybe I must have touched the powerswitch cable and thats why it wasn't going on. So I opened it again, doublechecked the powerswitch cable and tried again, this time it worked. I thought that must have been it. I've then continued with installing drivers for the graphics card, browsers, etc. and did some stresstesting on GPU and CPU to look out for the temps and everything was turning out perfectly. I also changed the RAM MHZ speed in the bios to 3200 mhz to get the most out of it.

Now, seemingly at random times, my computer just turns completly off. I can't really recreate it but I think that most of the time I've had many programs that worked seperatly open at the same time. Like a game, Teamspeak and a Browser with a video. First thought was: Is it temperature? - Checked it, no its not. The graphic cards never really gets higher than 75 degrees and the CPU is below 62-63 degrees at all times.
It happens 1-3 times per day, just goes black, stays black. I can't turn it back on myself if it happens. And then takes like 1-15 seconds to restart itself without me doing anything. At one time it turned on till before the "Press F2 for Bios screen" and then went black again three times before finally booting correctly. At first I suspected that the CPU might be taking too much power because I only connected an 8 pin cable to the motherboard for it but now I'm not sure because other people said something would be happening then. I have no second PSU to check if it is faulty sadly.

Any helps, ideas or input would be highly appreciated, because I do not know how to fix this problem on my own. Thanks in advance for everything you can provide!

Specs:

Motherboard: Asus ROG STRIX Z-490-F GAMING
CPU: Intel Core i9-10900K
CPU Cooler: NZXT Kraken Z63 280mm
GPU: MSI GeForce RTX 2070 SUPER GAMING X TRIO
PSU: be quiet! STRAIGHT POWER P11 Platinum 1000W
RAM: Patriot DIMM 64 GB DDR4-3200 Kit
Case: be quiet! DARK BASE PRO 900 rev.2
What drives are in the system? Did you do a clean install of windows? Did you try running the memory at stock speeds (2933) to see if it stabilized the system?
 
  • Like
Reactions: Forn
Aug 7, 2020
5
0
10
What drives are in the system? Did you do a clean install of windows? Did you try running the memory at stock speeds (2933) to see if it stabilized the system?

Thanks for your answer and time. I'll try to give as good information as I can.
  1. My drives are a Samsung SSD 860 EVO 500gb and a Samsung HD 103UJ 1tb HDD
  2. Yes the windows is a clean install on the SSD.
  3. I have not tried changing the memory speed, but will do this from now on and see if it changes anything.
It happens very frequently today, the third time now just before you answered. I was playing NBA 2k20 while it happened. Oddly the Motherboard still shows decorative light (The STRIX ROG logo) while the PC is out, same for the USB Xbox One Controller Adapter, which it wouldn't if power is shut off manually.

Edit: DRAM Frequency from 3200 mhz to -> Auto (maybe that will help)

Edit2: It did not change anything. I can recreate the problem. Everytime I start NBA 2k20 and go in to a game it goes out now. Often needs 2-3 times to restart then. Always does one "clicking" noise when going on or off. Just trying to tell everything that might help.
 
Last edited:

mamasan2000

Distinguished
BANNED
Sounds like a reboot because of RAM issues.

PC does not turn off completely unless you unplug it from the wall. It still gets power. LEDs and similar will be lit up. And PSU is just waiting for you to press the 'On' button. Then the PSU checks if the 'power is good?' If yes, then it fires up components, not all at the same time. It might seem like that but can you tell which component got turned on 10 ms earlier or similar? I can't.
 
  • Like
Reactions: Forn
Aug 7, 2020
5
0
10
Sounds like a reboot because of RAM issues.

PC does not turn off completely unless you unplug it from the wall. It still gets power. LEDs and similar will be lit up. And PSU is just waiting for you to press the 'On' button. Then the PSU checks if the 'power is good?' If yes, then it fires up components, not all at the same time. It might seem like that but can you tell which component got turned on 10 ms earlier or similar? I can't.

What kind of issues could the RAM have, and how could I test it?
The RAM is in the slots A2 and B2 (which are Slot 2 and 4 from left to right) as it was advised by the motherboard manual.

edit: Yeah tried it with a different heavy load game and instantly goes out again. This time though, the PC started a diagnosis before booting repairing the "drives" it said. It did so for like 10 seconds and then booted again without problems. This is weird, and I am scared now to test further. I've read that drives can malfunction if a system gets a lot of power outtages.

edit 2: Since the PC is wonky right now I did some tests with benchmarks. Prime95 and FurMark ran without any problems. Then I've found a video about someone having a similar problem. He ran the Heaven benchmark where it crashed. I ran that too, and it crashed for me too. So I would assume it has something to do with the GPU maybe.
 
Last edited:

mamasan2000

Distinguished
BANNED
1-5% of all hardware sold is faulty. Generally for a GPU to fail it has to be overclocked other than being faulty from the factory.

If it is your GPU, try to lower clocks on memory and core. See if that does anything. Drop both clocks by 500 Mhz. Then increase by like 100 Mhz at a time initially. Then do 50 Mhz steps, then 25 Mhz. You should end up with stable clocks.

If Heaven still crashes when dropping clocks by 500 Mhz, I doubt it's really the GPU.

For memory:
Run a couple instances of that, 2000 megs per instance til you almost have no memory left, Leave like 1 gig free after whatever Windows eats..
Run to 100% at least on every instance. 400% should be almost bombproof.
 
Last edited:
  • Like
Reactions: Forn
Aug 7, 2020
5
0
10
1-5% of all hardware sold is faulty. Generally for a GPU to fail it has to be overclocked other than being faulty from the factory.

If it is your GPU, try to lower clocks on memory and core. See if that does anything. Drop both clocks by 500 Mhz. Then increase by like 100 Mhz at a time initially. Then do 50 Mhz steps, then 25 Mhz. You should end up with stable clocks.

If Heaven still crashes when dropping clocks by 500 Mhz, I doubt it's really the GPU.

For memory:
Run a couple instances of that, 2000 megs per instance til you almost have no memory left, Leave like 1 gig free after whatever Windows eats..
Run to 100% at least on every instance. 400% should be almost bombproof.

I'll try both once the Computer has his "problems" again. Today I've had no issues yet and if so, I can not reproduce it. It does not crash at any of the tests aswell.

What I did yesterday was change the RAM to my older DDR4 16gb RAM and it worked with that. I've then noticed that through the moving of the Case, the Motherboard 20+4 Pin Cable was loose on the +4 cable part. It almost fell out on its own. Not sure if that happend because of moving the PC and laying it on the side, or if it was a mistake by me since the beginning. Does a Motherboard boot if only 20 of the 24 pins are powered?

I somehow can't believe that this is what is wrong though. I'm back on the 64gb ram today and its working aswell without problems. I'll update this, if the problem occurs again.
 

mamasan2000

Distinguished
BANNED
I think GamersNexus (Youtube tech channel) commented this week on a prebuilt system by NZXT that also had the 20+4 pin the wrong way around, the 4-pin above the 20. It's never going to stay there. Quality control, so so.

I don't know what the 4 pins do but every ATX motherboard has them so they must be important.
Looked it up:

ATX12V 2.x

"The ATX motherboard connector was extended to 24 pins. The extra four pins provide one additional 3.3 V, 5 V and 12 V circuit. "

Source: https://en.wikipedia.org/wiki/ATX
 
  • Like
Reactions: Forn
Solution
Aug 7, 2020
5
0
10
I think GamersNexus (Youtube tech channel) commented this week on a prebuilt system by NZXT that also had the 20+4 pin the wrong way around, the 4-pin above the 20. It's never going to stay there. Quality control, so so.

I don't know what the 4 pins do but every ATX motherboard has them so they must be important.
Looked it up:

ATX12V 2.x

"The ATX motherboard connector was extended to 24 pins. The extra four pins provide one additional 3.3 V, 5 V and 12 V circuit. "

Source: https://en.wikipedia.org/wiki/ATX

I've watched the video and what he showed and described looked exactly like the situation with my 20+4 cable. I double checked again now, and this time the 4pins are under the 20pins and it should be holding out fine. I don't really know if that was the issue, but it could be. Shame on me for making a mistake like that but I surely won't make it again.

I will investigate for the next days if this was the issue, but for now: Thank you!