Question Power loss under load

[Elvang]

I'm currently experiencing power loss randomly during gaming on high settings. The PC completely loses power, all lights out and fans stop; I have to flip the power on the PSU off and on before it will power up again. No signs before hand such as graphical defects. No BSOD, windows event log reports Event ID 41 Kernel Power. System is new, all drivers and bios updated, default settings for bios. Memtest passed. Stress testing power via OCCT causes the issue immediately, other tests run fine. Testing via Aida64 with graphics card, cpu, and memory all selected will also cause it immediately. Aida64 reports highest recorded temperature in the 68-75 C range for the tests with power loss. I've checked the cable seating, and am using a 12VHPWR connector for the GPU. I've also tried the power cord directly into a wall outlet, battery UPS outlet, and surge UPS outlet in case it was a UPS issue. Since it was easy to test, I've also tried stress testing the CPU with ECO settings and memory at a lower speed to same effect. The estimated power requirement is 700-800w depending on where I check.

System:

• Windows 11 Home 22H2
• 1x AMD Ryzen 9 7950X - 16-Core 4.5 GHz - Socket AM5 - 170W Desktop Processor
• 1x ARCTIC Liquid Freezer II 360 A-RGB 48.8 CFM Liquid CPU Cooler
• 1x ASRock X670E PG Lightning ATX AM5 Motherboard, Bios v1.11.AS03
• 2x G.Skill Flare X5 32 GB (2 x 16 GB) DDR5-6000 CL36 Memory
• 2x Kingston KC3000 2.048 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive
• 1x NVIDIA Founders Edition GeForce RTX 4080 16 GB Video Card
• 1x Lian Li O11 Dynamic EVO ATX Mid Tower Case
• 1x Corsair HX1000i (2022)
• 1x TP-Link WiFi 6E AX5400 PCIe WiFi Card
• 7x Noctua S12B redux-1200 PWM
• 1x CyberPower CP1500PFCLCD UPS

Is it safe to say its a PSU issue? I'm not sure what the next step would be to narrow the issue down aside from getting hold of another sufficient PSU to test swapping them.

 

[Lutfij]

Welcome to the forums, newcomer!

Troubleshooting is a process of elimination. Did you try working with a donor(borrowed) PSU that has a little more wattage at the entire system's disposal? This should help rule out your PSU being the root of the issue.

 

[Elvang]

I should be able to pick up a Corsair HX1500i on Wednesday to test swapping with, otherwise the only thing I have available is a 7 year old 750w.

 

[bill001g]

If you have to turn the power switch off and back on it many times means that some kind of protection circuit breaker in the power supply tripped.

It is not easy to say if that was because the protection circuit is defective or if your machine had a large power spike that tripped it. This tends to be a interesting video but 1000 watt supply should be enough.

View: https://www.youtube.com/watch?v=wnRyyCsuHFQ

 

[Elvang]

Interesting video, the demo at 26 seconds in is exactly what I experience. I'll be able to swap test a different, more powerful PSU on Wednesday, but if it is the GPU transient spiking then 1500w should be enough even for the 4080 spiking 2-2.5x draw in this system I'd think. The video mentions some motherboards don't handle it gracefully either, so hopefully its just the PSU causing problems.

 

[--SID--]

Is the PSU set in 12v single rail or 12v multi rail mode? You have to check this in iCUE.

 

[Elvang]

iCUE lists it as Multi-Rail OCP.

EDIT: Switched it to Single Rail OCP, ran the same Aida64 test as before that would instantly power loss. It ran ~35 seconds before power loss. Switching the power switch on the PSU and attempting reboot causes the case lights and fan to come on for half a second before shutting off again. Opened case up, checked cables again. Unplugged GPU first and attempted to reboot, was successful. Plugged GPU back in and rebooted fine. Power loss occurred a few minutes after reboot while idling, same issue where it does not want to power on again.

 

[Elvang]

Didn't mess with it after it started having trouble powering on yesterday. Got a hold of the HX1500i this afternoon. Re-seated the RAM, the GPU, all the cables, inspected the board again while I had the case open. Swapped out the HX1000i for the HX1500i.

I ran the computer through all the non-patreon tests in OCCT for a couple minutes and it passed where as before it would instantly power loss on the Power test. Ran Aidia64 stability test with all options selected and did not experience issues there either. Fortunately, it seems to have fixed the power loss. This is the only problem I've experienced with this build so far, is there anything I should check that could've been affected by the power losses?

That leaves the HX1000i either being under power for this build or defective (or I could be just that bad at seating connectors and 3rd try was the charm), and I'm not sure what I should do with it or how to test which it is. Just a heads up for anyone else planning on using the Lian Li O11D case and a similar size PSU as the HX1500i: it left very little room for plugging cables in with this case's layout.

 

[rcald2000]

FYI: BIOS version 1.11.AS03 is a beta BIOS. In your circumstance, I would reach out to Asrock support and inquire if the problem could stem from the BIOS version. If so, then I believe that your board can be BIOS flashbacked to the latest non-beta bios. If that inquiry and action failed, then I'm going to assume that the problem is a bad HX1000i. But again, the fact that you're using a beta bios adds a layer of complexity to troubleshooting this problem.

 

[Elvang]

Contacted ASRock, downgraded the bios and swapped in the HX1000i; same issues as before were happening so I'll get the process started with Corsair to check it out.