Random Full Power Off Crashes

[Big_Rich]

A couple of weeks after taking delivery of my computer, I had a few random crashes. Not BSODs or windows errors, full power off crashes. Cant reboot using the power button, I have to power cycle the PSU. The keyboard, mouse, motherboard and RAM RGB, AIO screen and power button all remain illuminated, the fans stop, but nothing responds. Only a power cycle will allow me to reboot. Sometimes if I don't wait a few minutes, it will immediately crash again when POSTing, but not always. Then sometimes it will stay on for 3-4 days at a time, I think it's cured, and it does it again. I'm really getting sick of it.

I've tried updating the BIOS, swapping the RAM round, using only 1 stick of RAM, unplugging literally everything except keyboard mouse and a single monitor, disabling EXPO and running the RAM at 4800... All to no avail. Event viewer shows nothing because whatever is causing it doesn't even give the OS chance to write an error log. It's just like someone flicked a switch.

It's definitely not heat related as I can run Furmark and Cinebench on a loop for hours on end, only for it to crash the day after whilst just sat on the Windows desktop. Also, the temps are fine when I've watched HWMonitor during benchmarking. There's absolutely no predicting it. I can be in a game, watching YouTube, browsing the web, even doing nothing, then bang, complete power loss with no warning. Other times I can leave it on for 2-3 days consecutively and it's fine. Totally unpredictable instability, and it's starting to get on my nerves. Not what I expect from a handbuilt system costing nearly 4 grand.

Here's the spec:

Case
CORSAIR 5000D AIRFLOW TEMPERED GLASS GAMING CASE
Promotional Item
Get a discount code for 20% off select peripherals at Corsair.com
Processor (CPU)
AMD Ryzen 9 7950X3D 16 Core CPU (4.2GHz-5.7GHz/144MB w/3D V-CACHE/AM5)
Motherboard
ASUS® ROG STRIX X670E-A GAMING WIFI (AM5, DDR5, PCIe 5.0, Wi-Fi 6E)
Memory (RAM)
64GB Corsair VENGEANCE RGB DDR5 5600MHz (2 x 32GB) AMD
Graphics Card
24GB NVIDIA GEFORCE RTX 4090 - HDMI, DP
1st M.2 SSD Drive
1TB SOLIDIGM P44 PRO GEN 4 M.2 NVMe PCIe SSD (up to 7000MB/sR, 6500MB/sW)
1st M.2 SSD Drive
2TB SOLIDIGM P44 PRO GEN 4 M.2 NVMe PCIe SSD (up to 7000MB/sR, 6500MB/sW)
1st Storage Drive
4TB SEAGATE BARRACUDA SATA-III 3.5" HDD, 6GB/s, 5400RPM, 256MB CACHE
Power Supply
CORSAIR 1500W HXi SERIES™ MODULAR 80 PLUS® PLATINUM, ULTRA QUIET
Power Cable
1 x 1.5 Metre IEC C19 UK Power Cable
Processor Cooling
CORSAIR iCUE H150i ELITE LCD XT RGB CPU Cooler
Thermal Paste
STANDARD THERMAL PASTE FOR SUFFICIENT COOLING
Extra Case Fans
3 x Corsair AF120 RGB ELITE PWM Fan + Controller Kit
Sound Card
ONBOARD 6 CHANNEL (5.1) HIGH DEF AUDIO (AS STANDARD)
Network Card
ONBOARD 2.5Gbe LAN PORT
USB/Thunderbolt Options
MIN. 2 x USB 3.0 & 2 x USB 2.0 PORTS @ BACK PANEL + MIN. 2 FRONT PORTS
Operating System
NO OPERATING SYSTEM REQUIRED
Operating System Language
United Kingdom - English Language
Windows Recovery Media
NO RECOVERY MEDIA REQUIRED
Office Software
FREE 30 Day Trial of Microsoft 365® (Operating System Required)
Anti-Virus
NO ANTI-VIRUS SOFTWARE
Browser
Google Chrome™
Warranty
3 Year Silver Warranty (1 Year Collect & Return, 1 Year Parts, 3 Year Labour)
Delivery
STANDARD INSURED DELIVERY TO UK MAINLAND (MON-FRI)
Build Time
Standard Build - Approximately 5 to 7 working days
Welcome Book
PCSpecialist Welcome Book
Logo Branding
PCSpecialist Logo
Packaging
PCS Corsair Carton Upright

 

[Big_Rich]

Forgot to mention, the only error messages I've ever seen was something briefly in the iCue software about mixing EXPO and XMP RAM sticks, which obviously they aren't. They show up as a matched pair of EXPO sticks in the BIOS, HWINFO and CPUID. Plus I've only seen this message once or twice. Also, very occasionally the iCUE software doesn't "see" the RAM and the RGB on the two stick stays in it's default rainbow display mode instead of matching the fans, AIO, keyboard, mouse etc. Could the RAM be bad? Wouldn't bad RAM just cause system errors and BSODs, not full power off crashes?

 

[TonyCarter]

I notice you ordered it without Windows, so how did you activate/source your copy?

 

[Big_Rich]

I notice you ordered it without Windows, so how did you activate/source your copy?

Already had a key, got the ISO from here: https://www.microsoft.com/software-download/windows11

 

[ubuysa]

That really doesn't sound like any kind of software problem to me. If I had to guess I'd say the PSU was the most likely cause, but that is just a guess.

You need PCS here and I fear you'll need to RMA it. It would help you, and PCS, if you were able to make it fail to order. Either by loading it up, or by leaving it idle. PCS have to reproduce this on their bench before they can fix it. You don't want it back after a week with 'no fault found', so do try and find a pattern that will make it fail.

 

[Big_Rich]

It's just done it again. Been gaming on it for 5 hours this morning no problems whatsoever, then *poof*, powered off.

Took a few attempts to boot back up, and I had the following errors:

Needless to say, I've never touched any of the AIO cables, it's definitely plugged in to the CPU_FAN header (and always has been) and the RAM isn't mismatched XMP/EXPO.

 

[ubuysa]

Download HWMonitor (free), install it and run it. Expand the CPU temperatures section so we can see every core.

Post a screenshot of that display when the PC is completely idle.

Post another screenshot of that display when the PC is running the biggest load you can throw at it.

 

[Big_Rich]

Download HWMonitor (free), install it and run it. Expand the CPU temperatures section so we can see every core.

Post a screenshot of that display when the PC is completely idle.

Post another screenshot of that display when the PC is running the biggest load you can throw at it.

Idle on desktop:

During 30 Minute looped Cinebench 2024 Run:

 

[ubuysa]

It's not CPU overheating that's shutting it down then. Now post two screenshots (idle and full load) showing the GPU temps and all the motherboard temperatures. We need to establish whether it's shutting down for temperature or for some other reason.

 

[Big_Rich]

It's not CPU overheating that's shutting it down then. Now post two screenshots (idle and full load) showing the GPU temps and all the motherboard temperatures. We need to establish whether it's shutting down for temperature or for some other reason.

As I've already mentioned, I'm 110% positive it's nothing thermal related. I can game intensively for hours, only for it to power off sat on the desktop doing nothing or just watching a film or YouTube. Sometimes it powers off a few minutes after bootup before I've even had chance to do anything.

GPU Temps at Idle:

GPU Temps after ~15 Minutes of Furmark:

 

[SpyderTracks]

As I've already mentioned, I'm 110% positive it's nothing thermal related. I can game intensively for hours, only for it to power off sat on the desktop doing nothing or just watching a film or YouTube. Sometimes it powers off a few minutes after bootup before I've even had chance to do anything.

GPU Temps at Idle:
View attachment 41044

GPU Temps after ~15 Minutes of Furmark:
View attachment 41045

Out of interest, what peripherals do you have attached? I know it's a pain but can you list make and models? I'm wondering if perhaps there's a peripheral causing some kind of surge on the system?

And are you on a surge protector or socket splitter of any kind?

 

[Big_Rich]

Out of interest, what peripherals do you have attached? I know it's a pain but can you list make and models? I'm wondering if perhaps there's a peripheral causing some kind of surge on the system?

And are you on a surge protector or socket splitter of any kind?

PSU is plugged directly in to the wall. No splitter or extension.

Only peripherals I have plugged in now are Corsair K100 keyboard, Corsair Ironclaw mouse, and one of my two LG 34GP950G Ultrawides. Oh and a LAN cable.

Previously had another 34GP950G, a Samsung soundbar (can't remember the model number, I'm not currently at home) and my HyperX Cloud gaming headset.

 

[SpyderTracks]

PSU is plugged directly in to the wall. No splitter or extension.

Only peripherals I have plugged in now are Corsair K100 keyboard, Corsair Ironclaw mouse, and one of my two LG 34GP950G Ultrawides. Oh and a LAN cable.

Previously had another 34GP950G, a Samsung soundbar (can't remember the model number, I'm not currently at home) and my HyperX Cloud gaming headset.

It's certainly a very odd issue.

My though process was that error you're seeing about CPU_Fan detection error, plus the iCue errors, paired with PC randomly losing power, it sounds either motherboard or PSU related.

But those PSU's are quite frankly some of the best on the market, they really don't skimp on them in any way shape or form, I'd be really surprised if it was that.

Which is what led me to wondering if there's some sort of surge that's grabbing power away from critical parts of the motherboard leading to power loss.

This is a really strange one though, I'm wondering if it needs some monitoring with voltage meters on the board at the time of a power off to see where it's stemming from.

It may be worth removing all peripherals except monitors and just leave the PC on and see if it's at all possible to still create a power off, if it does, you know it's not peripheral related, if it doesn't, then perhaps reattach them one by one and test over a few days.

But for something like this, an RMA with PCS may be a far quicker result as they'll have the kit to be able to test the PSU and motherboard which are always difficult components to troubleshoot.

 

[Big_Rich]

Wait a minute!

When I google my RAM part number (cmh5x32g2b56z40a2), it comes up "Optimised for Intel mainboards" and is showing XMP Rev 3.0 in my BIOS.

VENGEANCE® RGB 32GB (2x16GB) DDR5 DRAM 5600MHz C40 Memory Kit — Black

CORSAIR VENGEANCE RGB DDR5 memory delivers DDR5 performance, higher frequencies, and greater capacities optimized for Intel® motherboards while lighting up your PC with dynamic, individually addressable ten-zone RGB lighting.

www.corsair.com

Have PCS put the wrong RAM in?!

 

[Big_Rich]

Right, spherical's to this, I've put an RMA in. It's crashed three times this afternoon. Absolutely sick to death of it.

 

[SpyderTracks]

Right, spherical's to this, I've put an RMA in. It's crashed three times this afternoon. Absolutely sick to death of it.

I would suggest an RMA

You did specify AMD RAM in the order, so it should be EXPO RAM rather than XMP, but that still shouldn't lead to stability issues, the board can run either XMP or EXPO, you can't enable EXPO on XMP RAM is all. But for me, I would personally want AMD EXPO RAM over Intel XMP, so it's worth highlighting to them IMHO

But I would still RMA it, these are odd issues that need quite in depth testing with specialist tools.

 

[Big_Rich]

It's definitely getting worse now. Powered off twice last night, and three times this morning already. Not managed to keep it on more than half an hour today yet.

 

[Scott]

This is a good sign for the RMA, it means it should be resolved quickly.

By the looks of it, the wrong RAM was installed. I would highlight this to them as it may be a very easy fix.