Random Black screens on 1 year old build, absolute headache during diagnosis

[Musiz]

SPECS:
cpu: 7950x
cooler: arctic freezer ii
mb: asrock x670e pro rs
ram: corsair vengeance ddr5 32gb 4800MT
gpu: msi suprim 4090 (watercooled ver)
psu: corsair rm1000x

Hey everyone, im coming to you guys with a very weird problem to see if anyone has any advice.

i built this system over a year ago and its been running perfect and cool the entire time. about two weeks ago my pc started randomly shutting off completely and rebooting. as i did not have the components to test i sent it to a pc repair shop to diagnose and they said it was due to a bad cpu cooler. so i bought a new cooler and am having the exact same problems. and when i am using the system it does not get above 60C. after that i bought a new motherboard (MSI X670 AORUS ELITE AX) thinking it could be the only issue and the new motherboard does not boot at all. i tried with and without a gpu and in both cases the "vga" light turns on which would indicate a gpu fault, but without the gpu? the final thing i tried was updating the bios of the new motherboard, so i took out the cpu, ram, and gpu before trying to qflash the new bios and when i hit the qflash button the system would only turn on for half a second before turning off again. so i need youre guys help determining if its a bad cpu, or two bad motherboards.


facts/observations:

system crashes happen in *almost* any condition, happens in windows and in linux, does not happen when system is "idle" but if youre moving the mouse around or browsing chrome or gaming it can crash anytime
system has not crashed in bios settings
system crashes observed on different wall outlets. even in different buildings
system crashes are INCREDIBLY random, sometimes you dont even have enough time to finish loading windows, other times i can play video games for literal hours.
issue is observed with different ram
issue is observed with different psu
issue is observed with no gpu
issue is observed with two different coolers
tried clearing CMOS of both motherboards, to no avail
different issues are observed with two different motherboards, one does not boot while the other does not stay on
no overclocking profile applied to cpu, gpu, or ram

im on the verge of just ordering another motherboard and a new cpu but if anyone can think of anything that would point to or the other or if anyone can give me any ideas to test id appreciate it greatly

 

[Bill_Bright]

Too bad you didn't come here first as I am certain it would have been suggested you swap in a known good PSU to see what happens - especially before spending money on anything else.

That advice still stands. Since EVERYTHING inside the computer case depends on good, clean, stable power, you need to ensure you are providing it.

 

[SL2]

A wild guess, have you tried with all SSD's removed? Like Linux live USB? Only reason I'm asking is that you haven't mentioned any SSD, and one board hints at PCIE error, or rather GPU error, although when there's no GPU present.

It never crashes during OS install?

 

[Beginner Macro Device]

Either for some weird reason your SSD/cable is toast or yes, you should replace your CPU.

 

[Mister300]

I did every swap of HW possible and software uninstalls it came down to the Cablemod custom PCI-E 6 pin cable breaking contact upon thermal expansion, when I used the supplied EVGA cable it solved the random boot and shutdown issues. Find another cable and try it. I do custom PC builds for a living. The reboots and issues were totally random even occurred when making a new boot SSD or when gaming. See my pic the white cables to the GPU were the issue.

Mine would as crash in BIOS or idle desktop, I figured it out after seeing a customer touch his six pin PCI cable and it caused his build to reboot also. Cable mod are nice aesthetically but are flimsy in their lightweight design. The ATX pins wobble inside the Cablemod male connector resulting in a broken connection which will cause the GPU to lose power and reboot the PC. Have not had a 41 error or power reboot error in 2 yrs since replacing .

 

[Musiz]

thank you all for writing your suggestions. let me clarify on some other details;
@Bill_Bright the pc shop i took it too said they still observed the issue with a new power supply, but i might buy a psu anyways because they were wrong about it being thermal
@SL2 although i never physically removed the ssd i DID use a usb portable boot of linux for my testing, and still observed crashes
@Mister300 the only custom psu cables i use is a custom 4x pcie to 12vhpwr cable for my 4090, and like i said issue is still observed when gpu is physically removed

 

[Bill_Bright]

the pc shop i took it too said they still observed the issue with a new power supply, but i might buy a psu anyways because they were wrong about it being thermal

Oops! My bad. Somehow I missed where you said above, it happened with a different PSU. Sorry. While possible 2 PSUs were faulty, that is unlikely and even more unlikely they both would give the same symptoms.

What do you mean they were wrong about thermal? I note it is unlikely a cable, or its connection would be affected by thermal expansion. This is because these cables should never increase in temperature enough for the expansion and contraction of matter to affect the connection UNLESS too small a gauge wire (not likely), or the connector or connection was faulty, damaged, loose and not tightly secured in the first place. Then even a slight expansion might "open" the connection between one or more contacts. One other exception might be if there was a short somewhere in a circuit causing excessive current through the cable, causing it to get too hot. But if that were the case, it seems more likely you would be experiencing other issues too, and not so intermittently.

That leads me to this. Being a frequent building then you are aware that cases are designed to support 1000s of different motherboards of different sizes. So, it is common for cases to have more motherboard mounting points than some boards have mounting holes.

A common mistake by the less experienced and distracted pros alike is to insert one or more extra standoff in the case under the motherboard. Any extra standoff creates the potential for an electrical “short” in one or more circuits. The results range from "nothing" (everything works perfectly) to a wide assortment of odd problems, to "nothing" (as in nothing works at all ). To add to the confusion, these issues may be intermittent, depending on heat, expansion/contraction of materials, as well as continuity/resistance through the contact point. Therefore, you need to ensure the case only has an inserted standoff where there is a corresponding motherboard mounting hole.

Note the latest version of the ATX Form Factor standard hopes to eliminate these issues by dictating where standoffs will go, not just where they may go. But not all existing boards or cases comply with those latest standards - yet. So, you still should verify you only inserted a standoff where there is a corresponding motherboard mounting hole.

I recommend taking everything out of the case and assembling the computer on a large, unfinished bread/cutting board to see if it boots there. Then inspect the case and verify again, only the necessary standoffs have been installed in the correct places before reassembling back inside the case.

I also would normally recommend all homes and computer user have access to AC Outlet Tester to ensure the wall outlet is properly wired and grounded to Earth ground. I still recommend that for you (and everyone) however, I note you said you tried this computer in different outlets and different buildings too. So unless the different buildings are serviced by the same drop and service panel, facility wiring is not likely the problem either.

Other than that, I'm afraid I am grasping at straws.

 

[SL2]

@SL2 although i never physically removed the ssd i DID use a usb portable boot of linux for my testing, and still observed crashes

Well, removing it will cost you nothing.

If running without SSD doesn't help, then remove the board from the case and try again.

CPU, board, 1xRAM, PSU. Nothing else.

Next step is to underclock the CPU, by a lot. Half the speed. Disable PCIE (select IGP, don't just yank the GPU). Maybe underclock RAM/IF? Check out other things to disable or change in BIOS.