ROG Crosshair VIII Hero Wi-Fi Intermittently not starting

[AsusExplorer]

Has anyone had this issue where the ROG Crosshair VIII Hero Wi-Fi sometimes wont start?
Its only happening intermittently and randomly, so far twice in a month. No pattern to it, but its only the second time so far. I must have booted this machine about 60 times since I built it so its happening like once every 30 boots.
The rig is new.

Symptoms:
I see power to the board; i.e.: the start button is lit and the Hero lighting is on and pulsing, which is normal.
Room temp is around 17c, i.e.: cool.
Pressing the case button and nothing happens. No fans, no HSF, no LED's, no video card and no start.
Pressing the on-board start button and nothing happens either.
Power cycle the board by turning off and on the PS and then press the case button and it starts normally.
There is no q-code displayed while its off. Once its logged into Windows the q-code settles to AA.

I'm running with a Ryzen 5 3600XT with stock HSF.
Its running the RAM at 1066 MHz which is a G.Skill 32GB kit on the QVL. I did run it for a while using the XMP 1800 MHz profile with no issues after running hours of Prime95 as a test and Memtest86 for hours, but decided to go back to 1066 to eliminate that as a possible factor.
BIOS has everything on AUTO pretty much.
Has a hefty power supply, 1000W
CPU-Z shows normal vcore, ranging from ~0.432 to ~1.224 ish at idle. Stressing CPU pushes vcore to ~1.352
Temps seem normal given stock HSF.
All new components except for the video card which is a 1080. Video card has had zero issues in previous system.
BIOS was flashed to 3401 after the last time it failed to start.

Now today its done it again just out of the blue.
Otherwise the system runs beautifully, theres no pattern or obvious reason for this behavior.
Any advice or help will be appreciated, TIA

[AsusExplorer]

So this seems to still be an issue.
Happened again today after weeks of being fine.
It also locked up in the BIOS while I was checking things, which may or may not be related.
Looks like this issue happens very infrequently without any pattern I can see.

The conditions are:
1. I use the system without issue and shutdown normally.
2. Come back later (sometimes minutes later) and the system will not start.
3. Case power button has no effect.
4. Board start button is lit, but pressing it has no effect.
5. I then power cycle the PS and then it starts fine after that.
6. No relevent events in Windows, but then this happens outside of Windows anyway.
7. Q code was A9 when it locked up in the BIOS, which is a normal code.
8. Temps are excellent.
9. 32GB of G.Skill RAM running at stock, which is on the QVL.
10. I had fast boot set off, soft state power to USB set off, rebar set off and everything else on auto or default (except for that annoying auto download of Armoury software which is disabled of course).

Since the last time this was happening I've made a few changes to the system, so I've:
1. updated the CPU to a Ryzen 9
2. added a new Noctua HSF
3. Removed the board from the case, checked the standoffs and re-installed board,
4. changed the CMOS battery for good measure,
5. Updated the BIOS to the version that was out in May,
6. Checked all connections and devices were plugged properly,
7. Swapped out the cable loom to the case power button and USB ports.

I noticed that PBO was enabled so I've now disabled it, but I cant see how that would make any difference.
Only other info about this system is that I do have a couple of HDD's installed, but that shouldnt be an issue!

Anyone else had this issue with this board?

[RedSector73]

Firstly, wow, you have done a lot of good troubleshooting and investigation.

So what PSU are you using ?

Have you also checked the connectors for power are all home, so the 24pin is clipped properly inclusive of the EATX above the cpu, left of motherboard are both connected/clipped on. (long shot given all the work already done) On this vain, the 24pin power make sure none of the pins are home not loosened in the harness (again long shot stuff)

I'd stick XMP profile for the RAM that is natively supported by Ryzen, this isn't the source of the issue.

Point (5) in second post make me wonder if dead short issue maybe randomly present, check you have not overtightening the HSF to the CPU and that no cables are able to earth out from air movement, etc something random or as things heat (normally) up that may cause dead short condition.

Cold comfort is I've had this board since launch, no problems like you currently have.

[AsusExplorer]

The PS is a Seasonic Prime Ultra, 1kW. Its an awesome PS.

I've rechecked the board mounting stand-offs (actually removed the board and re-seated it) and completely unplugged and re-plugged everything.
Actually swapped the CPU over (from Ryzen 5 to Ryzen 9) and swapped out the HSF (not because of this issue, it was a planned upgrade).
Re-seated everything, re-routed cables.

My gut feeling is that this is likely to be a fault on the board or, less likely, a BIOS issue.
No POST indicates power related issue.
I've just flashed to the BIOS file released last month and set everything back to default.
Monitoring to see if/when the issue recurs. If it happens again I will likely swap the board out.

As to what it could be, maybe the VRM, either a design issue or a faulty cap or something.
I note that ASUS did redesign the VRM for the Dark version- although if this was the root cause then I'd expect a lot more people to be complaining than I can see.
I noticed an interesting settting in the BIOS, "F1 to resume after a fault is enabled". Next time this happens I will try F1 out of curiosity.

I've tried rebooting, cold starts, resume from sleep about 20 times now since the BIOS update and all good, but thats always been the case. Whatever the issue is I cant see a pattern yet.

[RedSector73]

AsusExplorer wrote:
I note that ASUS did redesign the VRM for the Dark version- although if this was the root cause then I'd expect a lot more people to be complaining than I can see.

The redesign is to take advantage of automatic overclocking/underclocking through the board (on dark hero). Not because there was any VRM issue, the VRM's are complete overkill for the job.

Can I suggest backing off the CPU heatsink screws 1/4 turn and see if the issue continues (monitor temps but see if this helps).

[AsusExplorer]

RedSector73 wrote:
The redesign is to take advantage of automatic overclocking/underclocking through the board (on dark hero). Not because there was any VRM issue, the VRM's are complete overkill for the job.

Can I suggest backing off the CPU heatsink screws 1/4 turn and see if the issue continues (monitor temps but see if this helps).

Yes, I feel that if the VRM was an issue then many more peeps would be reporting this problem.
Hence why I suspect the board is faulty or the PSU has a fault.

I appreciate suggestions, but can you let me know why this might help, I dont know about this particular issue?
The Noctua NH-D15S is doing a great job, temps are well managed and the CPU benches and stress tests without any issues.

I think I bottomed out the screws on installation, but didnt over tighten them. It didnt "feel" like they were over tightened.
Also, I swapped out the Ryzen 5 with its stock HSF for a Ryzen 9 and the Noctua HSF and the issue persists. *Could whatever this CPU mount issue is still be the cause after the CPU/HSF swap? Is it about the HSF contacting a component on the board?

[RedSector73]

The idea is to deal with intermittently failure to start and eliminate another possible cause. 1/4 turn back of the screw will not affect temps unless you already don't do them up snug, still you can monitor and adjust back if this does not resolve the issue.

[AsusExplorer]

RedSector73 wrote:
The idea is to deal with intermittently failure to start and eliminate another possible cause. 1/4 turn back of the screw will not affect temps unless you already don't do them up snug, still you can monitor and adjust back if this does not resolve the issue.

Sure, but what I mean is why would this make any difference? Is there a known issue with HSF screws?

[AsusExplorer]

Just an update, still experiencing this issue. Its very random and infrequent. Doesnt happen for weeks, regardless of system use and then out of nowhere, wont start.
I'm fairly convinced its either the board or the PSU, most likely the board.
Will swap it out as soon as lockdown ends here - fingers crossed its not the PSU.
PSU would be cheaper, but is far less likely to be the issue.

Question for anyone, have you ever had an issue with a device such as a HDD causing this board to have startup issues?
I read similar issues somewhere while searching, but not sure if I believe it.
I do have a couple of HDD's connected to the system SATA ports. They seem to run ok, but they're old drives.
Otherwise I have a couple of SSD's in the M.2 slots and a few SSD's on the SATA ports.
I dont see any reason the system would have an issue with HDD's and power, but just putting it out there at this point.
My PSI is a Seasonic Prime Ultra Platinum 1000kw. New a few months ago when I built the rig. I dont believe it has any problems with this load.

[RedSector73]

Actually, I have had issues with HDD and M2 before various systems over the years. Just going to summary some of the signs I remember.

Windows BSoD with totally random causes.
Freezing in BIOS with 'A0" dispalyed.
Smart Status errors being recorded (when Sata cable was the actual root cause).
Large transfers going to 0 bytes for long periods of time.

lockdown, new word for winter !

Maybe this worth seeing if related (for nvidia cards)
https://www.nvidia.com/en-us/drivers/nv-uefi-update-x64/