Solved Computer hangs after POST but before OS fully loads

Hiya, it's me again! Unfortunately I have ran into a new issue, and after exhausting everything I know and then some, I'm still stuck.

First and most importantly, the PC in question is my old rig now used by my partner. It's the exact same PC I used when I posted here with GPU issues back in October (I did get it fixed in the end! It was a faulty GPU. RMA'd & new one works great) minus the GPU & PSU, which I took with me to my new PC and returned the previous ones into this PC where they have worked perfectly fine as they have for years before up until Sunday (2 days ago) when this issue first started and has continued since.

Specs:
Mobo - Msi x370 gaming pro carbon (MS-7A32, Bios ver E7A32AMS.1L0)
CPU - Ryzen 7 1700 @ 3.0GHz
RAM - 2x 8GB Corsair Vengeance LPX DDR4, in slot 2 and 4 (DIMMB2 & DIMMA2)
GPU - MSI GTX 980
PSU - Novatech Power Station 750w Black Edition
Boot drive: Kingston 120GB SSD running Windows 10 on latest version & updates

Issue:
As much as I wish I could just easily say "this is the problem exactly", it's not that simple, as it varies depending on different factors.

• Windows 10 on boot drive - POSTs fine, begins loading windows, and freezes at a random point between a few seconds in, up to displaying the login screen (before it's fully loaded in).
• Windows 10 Automatic Repair - POSTs fine, begins loading automatic repair, freezes before it can progress from that stage, so around 1 to 5 seconds in.
• Windows 10 Installation media (USB 1 and 2 tested) - POSTs fine, begins loading and freezes anywhere from 1 to 5 seconds.
• Windows 11 Installation media (USB 1) - See above.
• Ubuntu 24.04 Live CD (USB 2) - POSTs, loads into GRUB just fine, running the default option or safe mode (nomodeset) results in a freeze at the splash screen. I have ran it with "quiet" and "splash" off and will touch on the results of that later. Apart from a few times, I have almost always been using the safe mode (nomodeset), even though there's absolutely no difference in behaviour.
• SystemRescueCD 11 (USB 2) - POSTs, loads into GRUB fine, running any option which launches SystemRescue results in all text going away as if it's getting ready to load it and freezes while still displaying the blank sky blue GRUB background it uses.
• Memtest86 v7 (Included in SystemRescueCD 11) - Works fine. No, seriously. It loads in and performs the tests. Not an issue in sight. RAM passed with flying colours (default options: 4 passes, 10 tests)
• Memtest86 v10.7 (USB 1 and 2 tested) - Again, this works fine, even with its fancy user interface. And once again, RAM passes (default options: 4 passes, 10 tests)
• grml 2024.02 (USB 2) - Well would you look at that, it works fine too! Only when using the "debug" launch item, the "graphical" one does not work due to an issue with grml itself and not the PC.
• BIOS itself - Works fine. Spent a few hours in there at one point and did not encounter an issue.

The issue first started happening when my partner was mid-game. She initially thought all the USBs died, it wasn't until I had a look that I noticed the system time was still displaying 10:07 instead of 10:15 where I thought it was an entire PC freeze. This was further confirmed when, after waiting a few minutes, it still had not advanced. Upon doing a hard reboot we encountered the first freeze, where it has persisted since. She took me step by step on what she did up to that point - Turn PC on from hibernation, log in, close all open chrome tabs, play music on spotify, open the EA launcher and start The Sims 4, playing for no more than 15 minutes until the aforementioned freeze. The music also stopped as soon as the freeze happened.

When I say freeze, I mean a proper unrecoverable freeze which no amount of time waiting will resolve. The PC becomes completely unresponsive, including the power button, and a hard reboot (hold down power button until it dies) is the only solution.

Here's the steps I've taken so far to try and solve this:

1. Tried normal boot up a few times
2. Tried forcing Automatic Repair to appear, which it did, only to also freeze as already described
3. Switched monitors to make sure it wasn't a dying display
4. Tried running Windows 11 Installation media a couple of times
5. Tried running Windows 10 Installation media a few times
6. Unplugged all drives, including the boot drive. From here on out, even right now, they have not been reconnected again.
7. Tried running Windows 10 & 11 Installation media again a few times
8. Ran Ubuntu 24.04 Live CD from USB twice
9. Ran Memtest86 v10.7 and allowed it to complete all 4 tests successfully
10. Ubuntu again
11. Switched to using Displayport instead of HDMI to make sure it's not just a dying HDMI port
12. Tried SystemRescueCD from USB a few times
13. Tried Memtest86 v7 that's included in SystemRescueCD and once again allowed it to complete all 4 tests successfully
14. Ran grml 2024.02 debug successfully a few times to ensure it's not a fluke.
15. Tried grml graphical and failed to boot. Found out it's an issue with grml itself.
16. Reset BIOS
17. Tried windows 10 installation media again, and when that failed...
18. ...I reset bios by removing CMOS battery, hitting the power button a few times, and leaving it for overnight + most of the next day. I hit the button a few more times before reinserting the battery and plugging back the power
19. Left bios at default settings
20. Tried ubuntu again
21. Tried Windows 10 media again
22. Switched to a different dp port on the GPU does in case
23. Windows 10 media a few more times
24. Verbally abused the PC, followed by begging it for forgiveness... Was worth a try!
25. Tried Ubuntu, but this time removing the "quiet" and "splash" options. More on this later. Tried this ~20 or so times
26. On the advice of a computer repair shop who said "it's either RAM or Mobo", I tried different ram configs with Ubuntu & Windows 10 media
    1. Swapping RAM around (from R1 in DIMMB2 and R2 in DIMMA2 to R2 in DIMMB2 and R1 in DIMMA2)
    2. Running in 1-3 instead of 2-4 (R1 in DIMMB1 and R2 in DIMMA1)
    3. Running 1 stick (R1) in DIMMB2
    4. Swapping which stick is in DIMMB2 (R2)
    5. Trying the same thing with DIMMB1
    6. 1-3 again, but R1 and R2 are swapped
    7. Tested single stick in the 3 other slots as well, as well as switching sticks and doing it again
27. Out of defeat, I tried windows 10 media a few more times.

Quick Edit: I have also attempted to use the USBs in different USB ports several times throughout the process listed above. It did not make a difference

Now about the Ubuntu with different launch options - When running with "quiet" and "splash" disabled, I found that there are only 2 messages on which the system does it's usual freeze. "nomodeset" does not affect this in any way whatsoever. The messages are:

EFI stub: Loaded initrd from LINUX_EFI_INITRD_MEDIA_GUID device path
OR
Freeing initrd memory: 73352K

The "EFI stub" message is the only message displayed. The other one has a bunch of stuff before it, which would take too long to write out there. I've included a picture of the monitor with the text to this post. There doesn't seem to be a rhyme or reason for why it sometimes freezes on the first message and at others allows it to progress a bit before freezing again on the second message. I've been able to get both results to happen with the different RAM configs I tried. I also tried waiting a different amount of time on GRUB a few times before booting to see if that has an effect and nope. Like I said already, nomodeset does not make a difference.

I've inspected every inch of the motherboard & the GPU and found nothing odd - everything looks fine from that aspect. If we ignore the freezing, the PC works just as it has done forever, with nothing out of the ordinary or any hints that something is not right.

I'm still baffled about how memtest86 or grml work just fine yet everything else seems to have a problem, and just how out of the blue this issue has come on. If I could get some help It'd be much appreciated!

Bruce said:

In you step list, at #6 you unplugged all drives, at which point did you then reconnect the boot drive?
Have you tried replacing the boot drive and reloading Windows on a new drive?

Click to expand...

At no point whatsoever - The drives have stayed disconnected, including the boot drive. Everything after step 6 was ran off of 2 different USBs onto which I loaded images as needed in that moment with Rufus.

Edit: And I have not tried that yet as the PC currently struggles with things like the windows installation media and an ubuntu live cd, so I thought it useless to go through all that effort to end up at the same result. I'd also have to find a spare drive big enough to house windows, which I don't think I have, but can check.

Bruce said:

Let’s show the complete PC specs using Speccy by Piriform. Get Speccy
Is Speccy safe?

In Speccy, click File > Publish Snapshot (hit Yes if prompted) > Copy to Clipboard > Close.
Paste that link into your post.

Let’s also show installed programs using MiniToolBox by FarBar. Get MTB
In MTB, tick List Installed Programs, click Go then close the program.
A file MTB.txt is created in the same folder, attach that in a post.

Click to expand...

Completely unable - Those tools are windows only. That is a problem as the only 2 things that have worked without issue are memtest86, and grml (https://grml.org/) which is a heavily stripped down linux distro, and I can only run it as a Live CD. Nothing else currently works without freezing in some way.

Bruce said:

That Novatech PSU, is that the brand or the mob you bought it from?

Click to expand...

Both - That PSU is a transplant from a different PC I have that was built by Novatech, who supplied their own PSU for it as part of the Black Edition Barebones DIY Kit. The original PSU in this computer released it's magic smoke a few years back and it's been the replacement since. I did have the PSU tested by a shop not long after as I was a bit worried considering the different PC had it's mobo fried, but it came back as being in perfect health (mobo died as a result of being too cheap for its own good, not the PSU)

Bruce said:

In that case, it sounds like you have a motherboard and/or a processor issue.

Click to expand...

Is there any way to know which one it is? I'm not exactly in a position to easily throw down on a new mobo to find it's the CPU and have to splash on that too and vice versa. Being able to know which one it is means getting it replaced is an easier pill to swallow.

Bruce said:

And that PSU sounds dodgy as hell, no-name brand, rebranded by the PC supplier - could easily be some Chinese rip-off.
What warranty did the PSU come with?

Click to expand...

I just did some digging and found out that yes, it's a rebranded PSU. Specifically, it's a rebranded Sirtec PSU, using their High Power OEM brand. Novatech apparently isn't alone in this as ThermalTake seems to exclusively rebrand their PSUs as their own, and I seem to have found a few ebay listing's of their PSUs which look almost exactly identical to mine!

That whole system came with a 5 year warranty. That was for both the full system as one, and for each individual component as it was a DIY kit.

Bruce said:

As to testing the mobo or CPU, no such easy "yes/no" option. Best way is to get your hands on another unit - friend, family, neighbour, 2nd-hand, etc.

Click to expand...

Unfortunately that's not really possible for me. I have a second PC, my own, but nothing would be compatible as its an AM5 build, not AM4. The only other person I know that had a desktop gave it back to me, which just so happens to be the Novatech PC. Even if it was working, it's a LGA1150 socket.

Unless there is another way, it seems like I'll have to take it into a shop and see if they can check it over. The shop I already spoke to before making this post said they don't have the parts and most shops around only do mobile repairs, so it's going to be a challenge.

Thanks anyway.

Bruce said:

@AGDeveloper - any luck yet?

Click to expand...

Yep!

Just before I sent my last reply I got in contact with a repair shop about 40 minutes away in Edinburgh. I gave him the readers digest of what's happening and linked back to here for the full info, but before he came here he more or less agreed that it's either a CPU or mobo issue. After reading my post on here and the replies we arranged a date to have the PC picked up and checked over as he had the spare parts, which happened on Monday the 13th.

Next day (today if we ignore the fact it's 1am), we have the verdict: he is 90% sure it's the mobo. He said he tested the board with a different CPU, different ram, different psu and gpu and replicated the exact same issue. The motherboard was the only thing not different and the issue happened. Case closed? Not quite.

He tested the PSU, RAM and GPU seperatley and they where all okay, no issues whatsoever, so I guess I won't be allowed to rebrand the psu to the "Novatech Electronic Molotov" this time! However, when he tried the CPU he got the same issue... The exact same issue that the motherboard has. This, quite obviously, threw him for a loop.

He said the CPU definitely works, he did extensive testing to make sure of it, but he could not get any systems to boot without the same issue... Until they did... And didn't again. No rhyme or reason. It's quite odd and neither me or him seem to have a clue as to why or how this is possible, but unfortunately it's reality. This is how we've been left with a 90% certainty of it being the mobo, and a 10% on CPU.

The most likely cause of the mobos issue is a possible short due to excessive dust buildup. The PC was last cleaned back in late October when I replaced my new PSU and GPU with its original components and built my new rig. It was a proper deep clean with no corner left untouched, and when it was done I gave it to my partner and officially washed my hands of all responsibility as I had my own brand new blast furnace to look after.

For the CPU I might as well go get a bowl of alphabet soup and have more luck finding a reason there.

So that's that. The next steps we're unsure of. When dropping the pc off he recommended we upgrade to AM5, which is something we're considering, however we're also considering just replacing the mobo with a different one on AM4. We'll see what our budget allows and what we find. If the CPU turns out to also be cooked it's not the end of the world as we still have the latest AM4 chips like 5800x3d, and AM5 would need a new one anyway. Plus, if we go the AM4 route, it will be the first to go. Don't worry, a PSU will be right after, as Samsung is interested in buying this one to turn into and revive the note series.

It may not have been the resolution we where looking for but at least we now have an idea on how to proceed. Thank you very much for all your help