Question Need help with lots of crashed on my PC - Dark Hero / 5900x

Toggle sidebar Toggle sidebar
A

Asryan

Reputable
Feb 20, 2019
141
2
4,585
Hi, I'm having lots of issues with my rig since few weeks :

To provide all infos :
The pc was mounted more than a year ago and after lots of hours of testing and stressing it was all fine for a year. No issue whatsoever. Only thing i've changed recently is the AIO, went for a Ryujin 2. Temps are good, 75c on CB20/23 and between 50 and 70c while gaming depending on how CPU intense the game is.

System Configuration/Specifications -
1.CPU : 5900x
2.Motherboard : Dark Hero VIII
3.UEFI/BIOS version : 3801
4.DRAM/RAM : Gskill trident neo 3600 / 32gb (8x4)
Graphics Card : Asus strix 3090
Power Supply Seasonic Prime Ultra 1000 titanium
Storage device ( SSD/HDD ) : NVME : Sabrent SSD 1TB Rocket Nvme
SSD: Samsung EVO 850 - 250Go + Samsung EVO 850 - 500Go

My issues started few weeks ago :

I remember I had a small issue when installing a fan I pressed the "safe boot" button by error. I booted into bios but mouse wasn't reponding so I forced shutdown and i had some issues then, when using retry button it booted normally.

Few days after, at f irst it was QCode 00 two times, one time during gaming, one time on a spreadsheet.
I read that an overtight cpu cooler could cause this, i've loosen it a bit and no Qcode anymore.

But :

Last week while gaming , WHEA error 18
A fatal hardware error has occurred.
reported by component processor core
Error source : Machine Check Exception
Error type : Cache Hierarchy Error
ID APIC du processeur : 4

What i did:
I tweaked curve optimiser on core 2, went from -30 to -20 thinking some windows update or maybe just time made my OC a bit less stable

And now yesterday

First crash : Screen went black, sounds was choped off but still working f but pc was running as nothing happened. I had to shutdown it by the psu.
Nothing in event manager or in reliability monitor except than an expected restart.

Second crash

Few moments later PC restarted when i plugued a external drive to a usb c extender which is plugued in the back of the tower. Screen went black and one of the PCIE of the GPU was flashing. I had to force shutdown

What i did at this moment :
Reinstall GPU with DDU
Disabled all CPU overclock in bios
went back to stock power plan (i was using custom by usmus)

All seemed fine but few hours later while working, Pc became very slow. Start menu was poping but nothing was responding in it, web page were not loading..
I forced the shut down, restarted and the pc turned itself off when it booted to windows. I rebooted again but it was fine .

What I could do?
use the clear Cmos button?
Upgrade Bios to 4006 (heard 4201 wasn't good?)
reinstall windows ?
go back to stock cables instead of cablemods?
memtest?

It's annoying me like hell i'm just losing all will to play, it's also a pain to work on my laptop when i'm remote..

Thanks for the help
 
Sort by date Sort by votes
The hardware combo looks perfect and your problems started when you changed the cooler.

You resolved the over tight cooler but it might've caused damage to the socket?

So nudging the safeboot caused it to switch on while you were installing a fan. Well that shouldn't have hurt anything unless you shorted something while the power was on by touching it but I would've thought you'll feel an electric shock or tingle if that were the case. I would probably rule out any electrostatic discharge.

Then the next problem was simply inserting a usb drive. It could've just been a random glitch.

but there are 3 symptoms suggesting the problem is with the motherboard.


There is more than one problem here though. Why do you have windows on the sata 3 samsung evo ssd's when you could run it faster on the n.v.m.e?

That doesn't strike me as someone who is particularly enthusiastic or knowledgeable about hardware. A nerd would definitely have windows on the n.v.m.e drive.

So you're in a bit of a tangle don't do anything hasty.


I would suggest moving the data from your N.v.m.e drive to the 500mb evo for the time being if it will fit.


So you've got more work to do to clear the data off the 'storage drive' so windows can be installed on it and the evo's aren't strictly speaking ideal storage drives. They have 5 year warranties but they will probably wear out before the n.v.m.e drive.

You can use them for now to stash any data you want to save off your nvme, supposing it will fit and you aren't using a full 1tb of storage.

Then disconnect both EVO drives.

Run a clean install of windows

Windows 10

forums.tomshardware.com forums.tomshardware.com

with the target drive being the n.v.m.e drive and that should be the only drive connected to the system

So now the problem becomes a bit clearer. If you simply cannot install windows because inserting the USB install drive doesn't work, then the problem is likely the mobo.

If there's a hardware problem windows setup is likely to fail but it's useful as kind of a diagnostic sometimes. If you can't simply install windows then there is a high likelihood of a hardware problem.

So in the event of any struggles to install windows I'd suspect a mobo problem at this juncture.
 
Upvote 0 Downvote
I don't understand where you get that windows was in the sata, it's on the nvme of course maybe the "storage device" wasn't a good choice of word but my os is indeed on the nvme.

For the cooler I don't think I have forced it too much but who knows

Should I just reinstall windows without clearing bios?
 
Upvote 0 Downvote
Well clearing the bios can't hurt anything only sets it default. So more simply removing the samsung evo drives might resolve it. Depends how old they are they might be near the 5 year warranty expiration and glitching or they might be the source of the glitch too.
 
Upvote 0 Downvote
The system does access the drives even when you don't but if they're going out then it would be best to copy the data off them to your nvme.

Well the drives can go to sleep when not in use in windows power plan? But if it was working normally it would be seamless and you wouldn't notice any activity or glitches.
 
Upvote 0 Downvote
They are quite easy to disconnect just power off for five minutes if you really want to drain the power, pressing the power button a couple of times while it's switched off at the socket, and detach the power and sata cables.

Then the windows setup will detect any hardware problem. The hurdle seems to be where to stash your data while running a clean install of windows.
 
Upvote 0 Downvote
Asryan said:
Hi, I'm having lots of issues with my rig since few weeks :

To provide all infos :
The pc was mounted more than a year ago and after lots of hours of testing and stressing it was all fine for a year. No issue whatsoever. Only thing i've changed recently is the AIO, went for a Ryujin 2. Temps are good, 75c on CB20/23 and between 50 and 70c while gaming depending on how CPU intense the game is.

System Configuration/Specifications -
1.CPU : 5900x
2.Motherboard : Dark Hero VIII
3.UEFI/BIOS version : 3801
4.DRAM/RAM : Gskill trident neo 3600 / 32gb (8x4)
Graphics Card : Asus strix 3090
Power Supply Seasonic Prime Ultra 1000 titanium
Storage device ( SSD/HDD ) : NVME : Sabrent SSD 1TB Rocket Nvme
SSD: Samsung EVO 850 - 250Go + Samsung EVO 850 - 500Go

My issues started few weeks ago :

I remember I had a small issue when installing a fan I pressed the "safe boot" button by error. I booted into bios but mouse wasn't reponding so I forced shutdown and i had some issues then, when using retry button it booted normally.

Few days after, at f irst it was QCode 00 two times, one time during gaming, one time on a spreadsheet.
I read that an overtight cpu cooler could cause this, i've loosen it a bit and no Qcode anymore.

But :

Last week while gaming , WHEA error 18
A fatal hardware error has occurred.
reported by component processor core
Error source : Machine Check Exception
Error type : Cache Hierarchy Error
ID APIC du processeur : 4

What i did:
I tweaked curve optimiser on core 2, went from -30 to -20 thinking some windows update or maybe just time made my OC a bit less stable

And now yesterday

First crash : Screen went black, sounds was choped off but still working f but pc was running as nothing happened. I had to shutdown it by the psu.
Nothing in event manager or in reliability monitor except than an expected restart.

Second crash

Few moments later PC restarted when i plugued a external drive to a usb c extender which is plugued in the back of the tower. Screen went black and one of the PCIE of the GPU was flashing. I had to force shutdown

What i did at this moment :
Reinstall GPU with DDU
Disabled all CPU overclock in bios
went back to stock power plan (i was using custom by usmus)

All seemed fine but few hours later while working, Pc became very slow. Start menu was poping but nothing was responding in it, web page were not loading..
I forced the shut down, restarted and the pc turned itself off when it booted to windows. I rebooted again but it was fine .

What I could do?
use the clear Cmos button?
Upgrade Bios to 4006 (heard 4201 wasn't good?)
reinstall windows ?
go back to stock cables instead of cablemods?
memtest?

It's annoying me like hell i'm just losing all will to play, it's also a pain to work on my laptop when i'm remote..

Thanks for the help
Click to expand...
Could be anything.

Just to get it out of the mix put a copy of memtest86 on a flash stick.

Boot the stick and let it run.....no errors allowed.
 
Upvote 0 Downvote
The bios was running for a year with no problem. Simply updating the bios doesn't seem like the resolution to me. It isn't the factor that has changed in the proceedings since the new cooler was installed. Well I haven't heard of windows updates causing incompatibility with old bioses before.

The 850 samsung drives however have more of a shelf life and show few discernible alerts or symptoms if they're going out. The complaint seems to be a slowdown which is a plausible symptom of something going on in the data transfer area of the operating system.

The bit about the 'pci-e flashing' is a hint that there might be a mobo problem so there's more than one problem but the 'insert usb drive into the usb extender and the system crashed and a pci-e light flashed' well that's another glitch that may or may not hint at a deeper problem or it might've been the extender, or just a co-incidental problem but since the system has been accessed with the over tight cooler problem and the safeboot button was pressed several things seem to be going on.



So simply disconnecting the drives first and booting windows seems like the easy thing to do to, but get's more complicated when somewhere needs to be found to stash the data. Disconnecting the usb extender will eliminate another possible glitch.
 
Upvote 0 Downvote
Yes, simply disconnect the power and sata cables while the system is off. That will test if the drives are ailing. If windows runs normally with the drives disconnected you can deduce that one or both of the drives is causing the slowdown.

The other glitches... insert usb drive and the system restarts... could be the usb extender

Since the mobo has been accessed and there was that tight socket problem and pci-e light flashed when insert usb... I don't quite understand the connection between the usb bus and pci-e lanes.

The accidental 'press safeboot while installing fan' shouldn't have hurt anything?
 
Upvote 0 Downvote
Will do. How long should I let it run?
qyfs.jpg
 
Upvote 0 Downvote
I've just flashed 4066, not sure if I reinstall windows but maybe I should.

Running bios stock, just tweaked boot options

Running a bit hot to my taste with vcore auto...

Not sure if I should bother with oc when everything is stable?

I had a profile at some point 46 all core at 1.28 which was quite stable but I heard that running fixed voltage on Ryzen isn't recommanded?
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom