Asus dark hero startup issue ?

[karlosfandango]

Hi all
I have a Asus Crosshair VIII DarkHhero ...with an AMD 5900X , 32gb of Gskils ram

Had a few times over last few months since recieving board where the system wont power on.
You press power switch on case and nothing .
You press power button on motherboard ...nothing ....and its lit up white ...all rgb is working ( on when powered off )
If i turn off PSU and back then power it back on ...system boots fine.
It can then be fine for say 4 weeks ...then happen out the blue again.
Was ok with old motherboard....so only things that have changed is MB and CPU.

Tried reflashing bios back to earlier version ...no different.
Had new graphics card since last time so rules that out.
Tried taking ram sticks out ...no difference.

Seems starnge how can run for weeks without issue ( but dont use it every day )

Im using an EVGA supernova 850W PSU.

Not major issue as all have to do is power off PSU and all good again ...but got to be something wrong.
Asus say sounds like something tripping PSU protection ...will have to try unpluging 24 pin and 8 pin cables next time , then plug back in to see if motherboard or psu .

Any help most appreciated.
Karl

[von_nihil]

[pokotaan]

When I shut down pc after gameing or editing videos, this start up issue happens 100%.
So I just checked below.

Case1:Start up pc -> Logon to windows -> Just shutdown ->[result] Start up issue is not happened.

Case2:Start up pc -> Run 3D mark ->After the 3D mark finished quick shut down ->[result] Start up issue is happened. 5/5

Case3:Start up pc -> Run 3D mark -> after the 3D mark finished 5min cool down time then shut down ->[result] Start up issue is not happened.


I guess
just before shutdown -> mb BIOS hold a high temperature on some parts and just keep holing while pc is off. *
then after the shutdown, when you try to boot pc, the bios check the held temp.
PC fans are not running and the held temp is high.
So it is not starting up?

Could someone confirm this with your pc?

My PC spec
Ryzen7 5800X
CORSAIR DDR4-3600MHz 16GBx4
Asus Strix-LC-3080Ti-O12G-gaming

And I'm using Hwinfo64 pro, every sensors show under 80 degree temp when Im playing game and editing vid.
So it most be ok.*

[outofdate]

pokotaan wrote:
When I shut down pc after gameing or editing videos, this start up issue happens 100%.
So I just checked below.

Case1:Start up pc -> Logon to windows -> Just shutdown ->[result] Start up issue is not happened.

Case2:Start up pc -> Run 3D mark ->After the 3D mark finished quick shut down ->[result] Start up issue is happened. 5/5

Case3:Start up pc -> Run 3D mark -> after the 3D mark finished 5min cool down time then shut down ->[result] Start up issue is not happened.


I guess
just before shutdown -> mb BIOS hold a high temperature on some parts and just keep holing while pc is off. *
then after the shutdown, when you try to boot pc, the bios check the held temp.
PC fans are not running and the held temp is high.
So it is not starting up?

Could someone confirm this with your pc?

My PC spec
Ryzen7 5800X
CORSAIR DDR4-3600MHz 16GBx4
Asus Strix-LC-3080Ti-O12G-gaming

And I'm using Hwinfo64 pro, every sensors show under 80 degree temp when Im playing game and editing vid.
So it most be ok.*

This is a pretty helpful info, thanks for testing it. I'll take a look at similar behavior on my setup.

[Cutsman]

This may seem drastic, but can you take the motherboard out and see if it boots and stays up with only the essentials plugged into it? I'm wondering if the board is fully secured to the case, all motherboard holes lined up with standoffs. Or maybe there's an extra standoff shorting the board.

Edit: Literally just noticed this thread was 37 pages long. Lol oops!

[BillBittel]

I think the part number difference has to do with what country the mobo was manufactured in. I assume you have heard that Asus just announced a recall of the Z690 Hero due to a capacitor that was installed backwards in some boards, resulting in failure to detect DDR5 DIMMs in some cases and catastrophic failure of 2 MOSFETs that provide power to the DIMMs in some cases - actual fire and burned components. If you have not heard about this, Google it. Here are a few links about the Z690 Hero Recall announced last week.

https://www.asus.com/us/news/u5866wtjwmbj1fue/
https://www.tomshardware.com/news/asus-owns-up-to-fiery-z690-motherboard-flaw-starts-recall-program
https://www.techspot.com/news/92796-manufacturing-fault-rog-maximus-z690-hero-motherboards-prompts.h...
https://videocardz.com/newz/asus-to-recall-rog-z690-hero-motherboards-affected-by-reversed-memory-ca...

In the case of the Z690 Hero, the part number with a "V" in it means made in Viet Nam. The part number with a "0" in it means made in China. They usually provide US distributors with boards made in Viet Nam to avoid US import tariffs on Chinese products. I think this may be the difference in the Dark Hero part numbers but I don't know for sure. FWIW my Dark Hero has never had a problem starting up. The part number on my box is 90MB1760-M0AAY0 so apparently it was made in China even though I bought it through a US distributor (New Egg).

[laeddis]

Thank you, BillBittel! You beat me to the punch! I just came back to this thread to update on that finding. I have been following along with the Z690 topic and saw a Reddit thread in which they discussed the part number convention. Playing around in the ASUS online store to inspect motherboards in different regions it seems everything they laid is accurate for how the latter half of the Part Number is used:

"90MB18E0 = sku/model, AUS ROG Z690 Maximus Hero

MV = seems to be manufacturer. MV = vietnam, M0 = china

A = probably destination market, A = north america., E = europe/middle east , J = japan, U = Australia & NZ

AYO = common across all."

So... all this to say my Part Number theory is debunked, although now I wonder how ASUS would go about identifying revisions in Part Numbers if there was some revision at the component level - maybe that info isn't actually captured at the listed Part Number level and is issued by some marking to the PCB itself?

[UPDATE 1/3/21]: ASUS does appear to etch the PCB Revision onto the motherboard itself - however, to view it on the Dark Hero you have to remove the x570 chipset heatsink (the big block between the M2 slots). You can remove screws on the back side of the board and it pops off easily enough. You will see a "CROSSHAIR VII DARK HERO" etching underneath the PCI x16.0 slot, and right under the 'HERO' word is a 'Revision'. On my replacement board it is R1.01.

- Laeddis

[BillBittel]

laeddis wrote:
Thank you, BillBittel! You beat me to the punch! I just came back to this thread to update on that finding. I have been following along with the Z690 topic and saw a Reddit thread in which they discussed the part number convention. Playing around in the ASUS online store to inspect motherboards in different regions it seems everything they laid is accurate for how the latter half of the Part Number is used:

"90MB18E0 = sku/model, AUS ROG Z690 Maximus Hero

MV = seems to be manufacturer. MV = vietnam, M0 = china

A = probably destination market, A = north america., E = europe/middle east , J = japan, U = Australia & NZ

AYO = common across all."

So... all this to say my Part Number theory is debunked, although now I wonder how ASUS would go about identifying revisions in Part Numbers if there was some revision at the component level - maybe that info isn't actually captured at the listed Part Number level and is issued by some marking to the PCB itself?

[UPDATE 1/3/21]: ASUS does appear to etch the PCB Revision onto the motherboard itself - however, to view it on the Dark Hero you have to remove the x570 chipset heatsink (the big block between the M2 slots). You can remove screws on the back side of the board and it pops off easily enough. You will see a "CROSSHAIR VII DARK HERO" etching underneath the PCI x16.0 slot, and right under the 'HERO' word is a 'Revision'. On my replacement board it is R1.01.

- Laeddis

Thank you for confirming this. Asus does revise their boards on occasion. I have a Maximus Hero Wifi which has 2 versions that I am aware of. There are also at least 2 versions of the Maximus Hero (not WiFI). The original version had a totally different VRM design than the later rev 1.01. The original model had an ASP1400BR PWM controller and a (4+2) arrangement of BSG0812ND FETs. The rev 1,1 version has an ASP1405I, 4 x IR3599 doublers, and ZF906 FETs. I never knew this until I realized that my Hero Wifi did not have a VRM temp sensor that others did. Asus quietly removed that temp sensor from their online marketing pictures. Deleting a temp sensor is not a huge deal but many people seriously consider VRM design when they select a mobo. With Asus at least, the board you get could have a completely different design that was never mentioned by Asus. Kind of scary when you think about it.

[STARRAIN_ROG]

Hi all,
I'm afraid the random startup issue is still under analysis.

Hi laeddis,
It seems the issue you mentioned is different from the random startup issue.
Could you please create a new thread in the Service-Related Inquiries for RMA related discussion?
https://rog.asus.com/forum/forumdisplay.php?295
Thank you.

[cosanoma]

OK, now it's time for me to get in contact with my lawyer to get back my money.

[Romikemi]

Adding another name to the pile. Same situation. Built a Dark Hero system for my wife on 27 Dec. It starts fine if it has only been in sleep for around 3-4 hours, it never (as in 0%) starts when it has been left overnight. Every morning, she has to toggle the PSU.

Some anomalies in my case that I don't see in other posts:
1. When I initially built the system, after a couple of reboots one of the g.skill Trident Neo 3600 sticks didn't light up. Shut down, re-seated, lit up fine and remains lit. When it didn't light up, Windows detected the memory successfully.
2. When the board gets into this power-protected state, in my case the ROG logo on the back panel hood flickers in a dark room.
3. I have hibernate and hybrid sleep disabled, and all sleep options are set to never. However, and this is a Windows issue I need to dig up, the system sleeps when I've told it to never sleep.

I have both Armoury Crate and iCue installed (5000x case), but that doesn't matter because the same problem happens for people who don't have iCue installed. Armoury Crate BIOS settings are on the default. Posters on this thread have tried every possible option with no success, so there's no point.

I recently did a Crosshair VIII Formula build for my system and I have zero problems with it, but Formula is overkill for what she needs and I'm not doing custom loop.