[T.S.O.M.]

Hello,

5 days ago, I wanted to check/test my old GPU for a last time for functionality, before finally selling it - at first, I wanted to do that immediately after procuring my (then) new GPU almost 3 years ago, but then decided to preserve it as a backup solution for the case my (then) new GPU goes defunct or I decide to sell it. So it lay in my room untouched for about 2 years straight. Now, I have yet again a new GPU (a RTX 3080 Ti) which I didn't even unbox yet and so, I began to disassemble my current GPU (a Turing-based Nvidia-GPU), but BEFORE inserting the 3080 Ti I wanted to do a final checkup of the old GPU laying in my room since years like I mentioned at the beginning of this posting...

This is how the procedure went on:

• Power off Computer by shutting down Windows
• Turn off PSU switch (PSU cable still connected to PSU)
• Turn off power source switch (PSU cable still connected to power outlet of power source)
• Discharge myself by touching the radiator of a heater
• Lay the computer case flat
• Screw off GPU screwdriver
• Remove power supply cables from my currently used GPU
• Release GPU by pressing the PCIE lock
• Remove current GPU and place it on a safe spot in my room
• Insert old GPU (that laid in my room for years)
• Connect PSU power supply cables to GPU
• Screw up GPU screwdriver
• Connect GPU to Monitor
• Power on power source switch
• Power on PSU switch
• Power on Computer by pressing chassis power button

Result: The computer immediately turns on (all fans start spinning: CPU fans, chassis fans, PSU fans) and within a second completely turns off again.

Now, I'm familiar to this phenomenon of multiple On's & Off's automatically in a row (or loop, if OC isn't stable) before finally successfully POSTing, after the motherboard was deprived of power for a longer time, but THIS TIME there were no consecutive/repeated Turn-Ons in a row - and not even a loop! The system just turns off within a second of pressing either the chassis power button or the onboard power button - and that's it! I would have to press the button again myself to initiate another Turn-On and again within a second, it turns off completely without further activity!

I need to mention that the motherboard itself appears to be fine, though - the RGB lighting flows through the board as usual and the onboard power button is glowing white as normal...

I have tried various solutions to fix this:

(for each solution, I performed the first 3 steps and eventually the last 3 steps of the above listed procedure)

• removing old GPU and re-inserting my currently used GPU -> problem still persists!
• powering up with NO GPU at all (both PCI-E x16/x8 slots empty/free), monitor connected to integrated GPU on MB backpanel via DisplayPort -> problem still persists!
• checking cables and connections on motherboard, PSU and chassis
• pressing and holding the "Clear CMOS" button on motherboard backpanel for 10 seconds and then Re-Try
• removing CMOS battery for over 5 minutes and re-insert, then Re-Try
• pressing and holding the chassis power button and onboard power button each one for 30 seconds to remove any residual current
• testing PSU for functionality by unplugging all cables and then plugging the 10+18 pin motherboard power cable to PSU with the other end of the cable attached to Seasonic's PSU Tester device. -> PSU has no issues.

None of them worked and slow, but surely I'm starting to become desperate

Please help!

T.S.O.M.

 

[D-EJ915]

Try to reseat CPU, sometimes clear cmos does not fully reset the board.

 

[T.S.O.M.]

Try to reseat CPU, sometimes clear cmos does not fully reset the board.

Thanks for helping me out and sorry for my late response!

I did what you told me (removing CPU and then re-inserting it on the CPU socket again) and it didn't fix the problem, but it revealed something else: apparently, a dust particle was in the CPU socket (on the pins) after I removed the CPU!

How is that possible!? I didn't do ANYTHING to the CPU the whole time I was operating on the motherboard - and before starting to operate/touch the motherboard, the CPU (and the whole system) worked flawless. I only removed my currently used GPU after shutting down Windows and then inserted my old GPU - nothing else, except setting the computer/chassis flat and then up again multiple times! So, how did that dust particle find it's way into the socket despite the CPU being firmly sealed!? I've no explanation for this... (except the dust particle got into the socket while I was removing/re-inserting the CPU - I don't recall accurately...)

The main revelation though, is even weirder: in order to remove the motherboard from the chassis (to start the system on a free surface), I unscrewed the mounting screws and then started to unplug all cables and connections to the MB. After unplugging the CPU power cable from it's motherboard connector, immediately a red LED on the MB just next to the connector started to glow. Just out of curiosity, I tried to power on the computer in this state (WITHOUT the CPU power cables connected) and surprise: the system started up properly (no immediate turn-off anymore), but no keyboard initialization, no screen and hence, no access to BIOS. And it got even weirder: Once I reconnected the CPU power cables to their MB connector, the immediate turn-off phenomenon occured again! Removing those cables from the MB connector yet again attenuated this phenomenon!

I'm starting to think the CPU is dead (the Q-Code LED always displays a "00" code, when turning on the computer - since the very beginning of my GPU changing session!), but how is that possible despite only having serviced the GPU without touching anything else!? And what about this weird behavior when unplugging and reconnecting the CPU power cables?

I'm really out of explanations...

T.S.O.M.

 

[D-EJ915]

It is possible it could be one of the cpu vrms died too, hard to say really since I think dead cpu would exhibit similar behaviour like you said.

 

[1devomer]

It is possible it could be one of the cpu vrms died too, hard to say really since I think dead cpu would exhibit similar behaviour like you said.

Dunno to be honest, but this is usually the sign that something is shorting out or is not connected properly.

So sanity check, start over by the following.
-Please provide the whole config specs.
-Disassemble everything for troubleshooting out of the case.
-Inspect visually the motherboard, psu, gpu, smell around for the magic smoke.
-Check the cpu, the motherboard socket, the memory slots, the pci-e slots and other connectors.
-Short a black lead with the green lead of the 24pin psu connector, to soft start the psu (check google), is the psu booting and running fine???
-Clear the bios CMOS, by getting out the bios battery for some time, at least 10/15min.
-Put together only the cpu + mobo + 1ram + psu, did the computer at least start, bip, postcodes, try different memory slots?

==>If yes, and you are confident that everything is now fine, plug back a working gpu, checking for video signal.

-If the rig is not powering, try mobo + psu + 1ram, did you get the power led and fan spin?
-Same thing with mobo + psu + cpu?
-Same thing with mobo + cpu + 1ram + psu, but with the motherboard 8pin cpu power disconnected.
-Post the result of your tests.

Hope you can manage to resurrect the rig!

 

[T.S.O.M.]

Update:

It looks like, I've found the dead computer part: it's the motherboard (Maximus XI Hero, Z390) ! I've plugged my 9900K, the memory sticks and the PSU to my OLD motherboard (Maximus X Hero, Z370) which I intend to sell - and bingo! The system finally boots up successfully to BIOS menu! (I didn't plug anything else on the old board, though... only the essential parts - the monitor was connected to the iGPU via DisplayPort on MB backpanel)

So both the CPU and it's iGPU appear to be fine!

But I still don't know what the hell I've done wrong in the first place - if I did ANYTHING wrong at all. I still don't know, how this could happen!

Any comments?

Important Question: Is ANY boot data (installed operating systems, bootloaders, boot entries, boot order/priorities) kept in the motherboard ROM? I ask this, because it is obvious that I'll have to RMA this board for a replacement...

T.S.O.M.

 

[o1dschoo1]

I

Update:

It looks like, I've found the dead computer part: it's the motherboard (Maximus XI Hero, Z390) ! I've plugged my 9900K, the memory sticks and the PSU to my OLD motherboard (Maximus X Hero, Z370) which I intend to sell - and bingo! The system finally boots up successfully to BIOS menu! (I didn't plug anything else on the old board, though... only the essential parts - the monitor was connected to the iGPU via DisplayPort on MB backpanel)

So both the CPU and it's iGPU appear to be fine!

But I still don't know what the hell I've done wrong in the first place - if I did ANYTHING wrong at all. I still don't know, how this could happen!

Any comments?

Important Question: Is ANY boot data (installed operating systems, bootloaders, boot entries, boot order/priorities) kept in the motherboard ROM? I ask this, because it is obvious that I'll have to RMA this board for a replacement...

T.S.O.M.

It happens sometimes... Its rare for a board to die randomly but it does.

 

[T.S.O.M.]

It happens sometimes... Its rare for a board to die randomly but it does.

How can I rule out, that a short circuit (by whatever means it happened...) was the cause? What other causes can cause a motherboard to die?

And I have the urge to bump this question again, because it's very important:

Important Question: Is ANY boot data (installed operating systems, bootloaders, boot entries, boot order/priorities) kept in the motherboard ROM? I ask this, because it is obvious that I'll have to RMA this board for a replacement...

T.S.O.M.

T.S.O.M.

 

[o1dschoo1]

How can I rule out, that a short circuit (by whatever means it happened...) was the cause? What other causes can cause a motherboard to die?

And I have the urge to bump this question again, because it's very important:



T.S.O.M.

With windows 10 you literally just plug up whatever your storage means is and it will boot.
What could cause it? Failed vrm too high vcore vrm overheat from too high vcore/bad case flow psu is on its way out atatic shock from your hands etc etc... Theres a a tone of ways honestly it couldve just been a freak accident. Test your psu with a multimeter.

 

[1devomer]

How can I rule out, that a short circuit (by whatever means it happened...) was the cause? What other causes can cause a motherboard to die?

And I have the urge to bump this question again, because it's very important:



T.S.O.M.

Well, maybe the old gpu is faulty, and killed the pci-e or what not, on the motherboard.
And at this point, i would leave aside the old gpu, for the moment.

As said above, you could test yourself the motherboard, looking for shorts.
Inspecting carefully the pcb, connectors, for any sign of damage.
You could check, for example, if the input 24/8pin is shorted, if the cpu vrm is shorted, the pci-e slot is shorted to ground.

There is nothing special stored in the bios, most of the setting values are stored in the NVRAM portion of the bios.
Stuff like your overclock, boot order, bios settings in general.

 

[T.S.O.M.]

Another Update:

I'm about to RMA my dead (?) M11 Hero; probably tomorrow. Now, I have a new and VERY important question:

Even if the Maximus XI Hero (a Z390 board released in Fall 2018) is probably discontinued: ASUS should have thousands or even hundreds of thousands of other samples of the exact same model in stock NOT intended for sale, but dedicated as replacement samples ONLY - for the scenario to provide replacement, if a RMAed board cannot be revived/repaired. Is that so?

I will not accept a different model (e.g. a Z490-board or a Z390-board, but different model like e.g. "Maximus XI Code") as refund or a simple money refund - because that would mean, I'll have to everything all over again (reinstalling and reconfiguring Windows/the entire System because the drivers are different), which would take MONTHS, since I do a Windows (re-)installation and (re-)configuration very meticulous and extremely detailed...

T.S.O.M.