Question A devilish problem with Maximus XIII Hero and i9-11900K Q code d4 and its reincarnation

[cemster]

Hello,
I rebuilt my workstation from 10 years old P9X79 WS to a couple of months old ROG MAXIMUS XIII HERO coupled with an 'Intel i9-11900K', 'Corsair VENGEANCE® LPX 128GB (4 x 32GB) DDR4 DRAM 3200MHz C16 ' and an 'Asus GTX 1660Ti...

The reason I have a new build is that towards its last days the old build it started to freeze... It was a hardware problem.
The first thing I replaced was my GPU card with ASUS DUAL GTX1660TI O6G EVO. Which is also my GPU card of the new build. And also a quite new purchase.

The problem I am having is 'White light' sometimes followed by "1 long 3 short Beeps"...
And Q code displays "d4" but the post continues after a while and it settles on "A0" And as the hard drive light blinks away, as the "White" light indicates a GPU error remains and no signal is received by the monitor.

This Q code 'd4' has started right away with my very first boot attempt. However, as I rebooted, it did not give any error on my second or on the third try. Which I kept the system on, almost for three days at that time. Then next day on the fourth booth, it stuck on GPU card error again. This is how it continued for six days.
And on that sixth day, after a successful boot, the system and the monitors freeze after a while! You know as my old build did! So, I rebooted again, it freezes again in a few minutes. And after that, I couldn't pass the GPU error; the "d4" and "white light" and the "beeps". Yet the post continues and the Q code finally displays "A0". Meantime, the GPU card shows a "White" light as oppose to a "Red" light and both mobo and GPU card illuminate RGB lights.
When the system was working I had many HDDs and an SSD installed, and a few USB-connected devices plus fans, all working fine.

Now, I just have an SSD and 8 fans. I have an i9-11900K cooled by a Noctua NH-14 with two fans. That's it. I even unplugged anything coming from the case such as USB connections and whatnot.
I power the whole thing with a Corsair AX1200i PSU, and it is the only thing that I haven't replaced from my old system...
It shows the green light and seems working fine also monitored with its "iCue" app when I could boot. When the system was working I checked every component.
Rams were showing 64 GB instead of 128GB. I realize that two of them weren't properly placed so I corrected that and got 128 GB. I checked each Ram individually and they were ready for OC and they were showing proper voltage and everything. Same with the CPU. Each core seemed to be functioning normally. At one point it even OC'ed %10 with auto OC.

Even though I do suspect the PSU, I can't be a hundred percent sure if the PSU is really the problem...
I couldn't figure out which device was the faulty one with my previous system either. I took my system to three different shops for testing and none of them couldn't find anything wrong! Even with all the stress tests and whatnot, they couldn't even manage to duplicate the system freeze... And as soon as I get home, I get the "frozen screens" which they couldn't produce. Once even for a consecutive three days, they fished for the error and got nowhere...

So, as you can see it is weird luck I am having. It feels like this is a curse really...
I replaced the GPU card first and that didn't help. Nevertheless, I realized I replaced the wrong device. But maybe as the previous GPU card (Zotac) was faulty and with my bad luck, the new GPU card that I bought was/is also bad. Then, I end up replacing everything except the PSU.
I just don't want to buy a new PSU and find out that in fact, it was the GPU card that was the problem all along...

So I did two more tests today. First I removed the GPU card. Since I have the integrated CPU graphics. But I get all the same errors. White mobo light and beeps...
Then I managed to borrow a 600 watt Thermaltake PSU. I hooked up with high hopes that it would work and I can finally pinpoint the faulty device after all the drama.
But no such luck! I still am getting the same errors. And now the white light is accompanied by a Yellow light as well, indicating that something wrong with Rams too...
When the Corsair AX1200i was connected, and the white light came up, the yellow light also came up for a very short while, and as the post continues the yellow light went away as the white light remained every time.
However, with this 600 Watt PSU, the yellow light also remains and the post does not continue... So 'd4' is the final Q code there...

This is even more confusing. I am baffled. Is it the new Motherboard that is the problem?
'D4' is explained as either integrated GPU, or if a GPU card is installed the GPU card is not detected or has a problem. So they suggest checking the CPU! But the CPU was working just fine a week ago. As all the Rams were...

Also, all my parts are compatible. I did my research! Matter of fact, both Asus and Corsair gave me their blessings personally. Asus put some 128 GB Corsair rams that higher than 2333 MHz into their compatibility list after my request. The video card is also listed as a compatible device.

I also did reset the CMOS a couple of times.

I am running the latest BIOS. I updated the unit before running the system through Asus's USB port. The drivers were installed through the driver disk that came with the unit.

Nothing is automatically updated in my system.

I don't know where to go from here?

And if you read the whole thing I appreciate it, I know it's a lengthy one but so is my curse...

 

[VirtualLarry]

Even though I do suspect the PSU, I can't be a hundred percent sure if the PSU is really the problem...
I couldn't figure out which device was the faulty one with my previous system either. I took my system to three different shops for testing and none of them couldn't find anything wrong! Even with all the stress tests and whatnot, they couldn't even manage to duplicate the system freeze... And as soon as I get home, I get the "frozen screens" which they couldn't produce. Once even for a consecutive three days, they fished for the error and got nowhere...
So, as you can see it is weird luck I am having. It feels like this is a curse really...

MOVE!!!

Edit: Or have your house wiring checked out by someone really competent.

 

[cemster]

Can you be more specific?

 

[VirtualLarry]

When you brought your PC to the three shops, did you bring any of your peripherals or monitor? Most ordinary PC shops wouldn't want those things, but this sounds like the type of extended problem that might require further holistic analysis.

Are you connecting with DisplayPort or HDMI? Have you tried a new, fresh, display cable?

 

[blckgrffn]

Have you tried a different power cord? Do you have a power conditioner/ups between the computer and the wall?

In my experience hard freezes like that are often power related but there is more to power than the PSU.

 

[cemster]

It is not a cable or type of cable issue, I assure you. I have three monitors and many cables.

The shops I am talking about were not vendor-related or authorized services, they were PC repair shops. So yes they had the whole machine. As I mentioned in my long and boring story...

 

[aigomorla]

1. can you try the video card on another board / system?
(this will completely rule out the video card)

2. try to put the videocard in the other x16 slot and see if you have issues.
(this will rule out the pci-e slot and board)

3. Disable XMP on ram but increase voltage to about 1.35v and see if this helps anything temp.
(this is xmp voltage but your ram will be running at base clock and should rule out for ram)

4. You already trouble shooted the PSU, but can you try to put the psu in another PC and see if the problem is repeatable?
(this will rule out the PSU)

 

[DAPUNISHER]

have your house wiring checked out by someone really competent.

it has to be this. The fact it works for 3 solid days of testing at the shop, without issue is indicative that the machine is fine. The power delivery between the grid and the system is nerfed somewhere. Especially since OP says he gave them monitor, m&kb, the works. If a peripheral, cable, etc. were at fault, it would have turned up at the shop.

 

[aigomorla]

I would honestly say at times like this, a good Sine wave UPS is a godsend.
If it has a LCD, it will tell you what voltage its pulling from the wall, and go on battery power when you hit a brown (too low voltage).

 

[Iron Woode]

Can you be more specific?

Get a multi-meter and check the voltage at a receptacle. Maybe try a few just to see. If you have a dedicated receptacle tester, use it to make sure they are wired correctly.

Then let us know what you saw.

 

[cemster]

My wiring is done by an engineer. I have 2.0 ohms ground provided by a fat copper rod buried under the ground 4 feet deep, connected to my system with a thick copper cable. The audio equipment is on a dedicated circuit. I have a superb power conditioner. (Which just in case I tried bypassing the UPS as well)

Also, I remind you that my previous built worked flawlessly for a decade in the same circumstances... So, I assure you, I have a clean and balanced power in my studio.

 

[cemster]

1-2-3-4

I will try these steps. I can't seem to find any decent PSU to experiment.
Unfortunately, I don't have anybody with a high-end system around me...

And thank you all for the support

 

[bigboxes]

it has to be this. The fact it works for 3 solid days of testing at the shop, without issue is indicative that the machine is fine. The power delivery between the grid and the system is nerfed somewhere. Especially since OP says he gave them monitor, m&kb, the works. If a peripheral, cable, etc. were at fault, it would have turned up at the shop.

I've run through a UPS AVR for years. Take the dirty power out of the equation.

 

[blckgrffn]

I mean you might just have a bum motherboard. At some point that's the simplest answer.

 

[VirtualLarry]

I mean you might just have a bum motherboard. At some point that's the simplest answer.

"Three shops" couldn't diagnose that? Hmm.

 

[blckgrffn]

"Three shops" couldn't diagnose that? Hmm.

Well something is broken. It only happens after a while of use. How long are the shops really interacting with a pc to troubleshoot it? I don’t know what those shops actually did, either.

At some point you just have start replacing components. If he has a power conditioner and high end UPS and stuff got weirder with a new PSU 🤷‍♂️ Swapping a board takes less time for me than taking it somewhere, dropping it off, trusting they won’t break it, going back, picking it up, etc.

Besides the PSU the motherboard is the next most complicated (ie prone to failure) component I can think of.

Even a cheaper board would be fine for testing to see if things go sideways.

 

[aigomorla]

Besides the PSU the motherboard is the next most complicated (ie prone to failure) component I can think of.

actually no.. the PSU is usually the last thing to go unless you majorly undershot your wattage requirement, and ran that unit hot all the time.
The first to go is usually the RAM, depending mostly on the XMP setttings, how long it was run at XMP, and how hot they ran.
XMP is vendors "factory overclocking", meaning its not the JEDEC of ram.

Then its usually the board, as the power regulators and caps are more prone to fail on the board then the PSU, unless again, your wattage calculations are really bad, and you ran the psu very hot all the time.

 

[bigboxes]

actually no.. the PSU is usually the last thing to go unless you majorly undershot your wattage requirement, and ran that unit hot all the time.
The first to go is usually the RAM, depending mostly on the XMP setttings, how long it was run at XMP, and how hot they ran.
XMP is vendors "factory overclocking", meaning its not the JEDEC of ram.

Then its usually the board, as the power regulators and caps are more prone to fail on the board then the PSU, unless again, your wattage calculations are really bad, and you ran the psu very hot all the time.

I agree with you, but I have had plenty of PSU failures. Usually, it's the CPU that is the last to go. But then again, I'm not going for max overclock like some on here.

 

[Iron Woode]

OK, time for some basic troubleshooting:

take the motherboard out of the case. Have only the cpu with HSF, 1 stick of ram and the PSU. Nothing else.

Does it post?

 

[cemster]

"Three shops" couldn't diagnose that? Hmm.

Well, all they did was changing cables, replacing the thermal paste, performing the usual onboard tests for CPU and GPU, except Rams, they were tested outside the system,
and just wait for the problem to occur... So, it wasn't like tones of lab gadgets and whatnot...

Okay, I will go to the next phase and start from the basics as Iron Woode
suggested... I suppose if everything I do doesn't produce a post, then it gotta be either the PSU or the Mobo... Right? I gave the 600W PSU back so then I have to find a PSU that is even a little better, in case I have to try with the GPU card plugged in.

But... What if during the barebone tests, at some point the machine boots up, and I have the display and everything... Yeay? No! Not really! I would be at my start point!
These last four years with my old build and now with this, these errors are not constant!
It works! And then it doesn't work! Then it works again! Then it stops working!
I can't tell you how horrifying that feeling is. Not knowing when it's going to stop working and you will be stuck in a middle of a project and not knowing what is the cause of this to fix it...

Anyways... Thank you, everyone...I feel the support...

 

[blckgrffn]

Well, all they did was changing cables, replacing the thermal paste, performing the usual onboard tests for CPU and GPU, except Rams, they were tested outside the system,
and just wait for the problem to occur... So, it wasn't like tones of lab gadgets and whatnot...

Okay, I will go to the next phase and start from the basics as Iron Woode
suggested... I suppose if everything I do doesn't produce a post, then it gotta be either the PSU or the Mobo... Right? I gave the 600W PSU back so then I have to find a PSU that is even a little better, in case I have to try with the GPU card plugged in.

But... What if during the barebone tests, at some point the machine boots up, and I have the display and everything... Yeay? No! Not really! I would be at my start point!
These last four years with my old build and now with this, these errors are not constant!
It works! And then it doesn't work! Then it works again! Then it stops working!
I can't tell you how horrifying that feeling is. Not knowing when it's going to stop working and you will be stuck in a middle of a project and not knowing what is the cause of this to fix it...

Anyways... Thank you, everyone...I feel the support...

I completely hear what you’re saying. When the trust is gone and you have dread in the pit of your stomach when you work with it? Ugh. Been there too many times. And some computers seem like they need an exorcism as much as anything technical.

Way more computers easily run fantastic out of the box/boxes to put up with this.

I’ve been know you part out known good parts and then junk the suspect bits (put in a bin until they are too old to be useful because I can’t part with things easily) and start over from clean a clean slate. 🤷‍♂️

Good luck!

 

[Iron Woode]

Well, all they did was changing cables, replacing the thermal paste, performing the usual onboard tests for CPU and GPU, except Rams, they were tested outside the system,
and just wait for the problem to occur... So, it wasn't like tones of lab gadgets and whatnot...

Okay, I will go to the next phase and start from the basics as Iron Woode
suggested... I suppose if everything I do doesn't produce a post, then it gotta be either the PSU or the Mobo... Right? I gave the 600W PSU back so then I have to find a PSU that is even a little better, in case I have to try with the GPU card plugged in.

But... What if during the barebone tests, at some point the machine boots up, and I have the display and everything... Yeay? No! Not really! I would be at my start point!
These last four years with my old build and now with this, these errors are not constant!
It works! And then it doesn't work! Then it works again! Then it stops working!
I can't tell you how horrifying that feeling is. Not knowing when it's going to stop working and you will be stuck in a middle of a project and not knowing what is the cause of this to fix it...

Anyways... Thank you, everyone...I feel the support...

sounds like a good plan.

if it does post, try a different ram module in the same slot to see if the ram is an issue. if they all work then I would run memtest86+ to make sure. You could even try XMP to see if there is a problem.

I just thought of something: did you do a close up inspection of the socket? Use a magnifying glass and look for any bent pins or anything out of the ordinary.

 

[cemster]

I think I made a wrong impression about the three shops and the replaced thermal paste and all... These are all related to the 'Previous Build'. not this one.

The new system has not left my place yet...

I checked the CPU as I took it out of its container very closely...

The new Ram sticks tested one by one with the iQUE monitoring and configuring app from corsair. They all work as they suppose to...

 

[Iron Woode]

I think I made a wrong impression about the three shops and the replaced thermal paste and all... These are all related to the 'Previous Build'. not this one.

The new system has not left my place yet...

I checked the CPU as I took it out of its container very closely...

The new Ram sticks tested one by one with the iQUE monitoring and configuring app from corsair. They all work as they suppose to...

do you mean you checked the socket or just the cpu itself?

Maybe post some pics of the socket up close so we can eliminate that as a cause.

 

[cemster]

Both Motherboard and CPU, PASSed the visual audit as I took them out of their respected boxes... The CPU then placed in its housing with ease and precision...