Question PC Freezes with new GPU and PSU ?

[Dajsus]

So I recently upgraded my system with a new GPU and PSU. First few days everything was fine and then random freezes started, like complete freeze where windows would not respond. I completely reinstalled Win11 and went back now to Win10 but still having issues. I do not get a BSOD or any kind of an error in the Event Viewer. There is not exact time when this
My specs are:
Ryzen 5600x
Asus TUF B450-Plus Gaming
32 (4x8GB) Kingston 3200Mhz
Gigabyte RX 6900 XT (Brand new)
NZXT C850 PSU (Brand new)
2x Sata 240GB SSDs
1x HDD 1TB


So far what I have tried as I am unable to reproduce the issue:

• Ran multiple benchmark and stress tools to see if the issue is going to happen (AIDA64, OCCT, FurMark and etc) no freezes.
• Tried 2x8 GB RAM only in slots 2 and 4 aswell as 1 and 3. same issue would happen.
• Ran Windows Memory diagnostics no errors.
• Updated BIOS on the Motherboard and the GPU.
• I am using 3 separate PSU cables to power the GPU.
• Tried Recommended and Optional Drivers,

As I have read there are some issues with AMD and dual monitor, so I tried a single monitor and still it would freeze.
(I googled alot of this and tried numerous things from other tomshardware posts)

I am out of Ideas I suspect there is an issue with one of the new components, but I am unable to currently obtain an additional PSU to test. As for the GPU I can test a 1050 TI but that one does not use external power so I do not know if it would be a reliable test.

 

[Ralston18]

Look in Reliability History.

Much more user friendly than Event Viewer and the timeline format may prove revealing.

How were the new GPU drivers installed? Manually or via some third party tool or installer?

"I am using 3 separate PSU cables to power the GPU. "

Did you mix and match PSU cables from other PSUs?

 

[Dajsus]

1. I did check it I forgot to add:

Problem Event Name: LiveKernelEvent
Code: 1a1
Parameter 1: ffffc40568bd0040
Parameter 2: 0
Parameter 3: 0
Parameter 4: 0
OS version: 10_0_19045
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19045.2.0.0.256.48
Locale ID: 1033


2 From the AMDs website.
3 NZXT PSU cables were used only

Just to add I serviced a lot of PCs over the years since I used to work in PC part sales just never came across something this random.
I mean reading the reliability history I totally missed the Hardware error part since the summary was hidden like this for some reason.
https://prnt.sc/eqvjTtG_G-24

 

[scout_03]

what happens if you use only new psu and old gpu or new gpu and old psu also event show windows error .

 

[Roland Of Gilead]

Hey there,

I'd add to the other advice, by suggesting you clear CMOS after the bios update. This could be the issue.

You need to be running whatever the latest bios is that includes the 1.2.0.7 Agesa build.

 

[Dajsus]

@Roland Of Gilead I did clear CMOS, just forgot to add that to multiple things that I performed, also to add the BIOS update was done after the issues started.

@scout_03 Sadly the old GPU is sold already, only one I have is a 1050 Ti which does not require any external power. As stated above would not be a good measurement, as for the old PSU it is too weak to test the new GPU as it is a 500W supply with unknown efficiency from a generic brand.

 

[Dajsus]

I mean everything points me to one of the components being bad but just wanted to make sure.

 

[scout_03]

do you have a friend that could made test for in is system and have a powerfull psu to do it .

 

[Roland Of Gilead]

Have you tested the ram outside of Windows, with Memtest86+? If not, do that, and let us know.

Freezes like that can be ram related, and/or conflicts.

Presuming all system drivers, including chipset drivers are up to date? (Edit- I see you've done this)What power plan are you running? Ryzen balanced?

 

[Dajsus]

I am waiting for a power supply to arrive on Monday to do a test on that one.

 

[Roland Of Gilead]

Updated BIOS on the Motherboard and the GPU.

Wait, you did a vbios update?

 

[Dajsus]

Wait, you did a vbios update?

Yes, and both are post-issue updates.
And sorry I said etc in the benchmark/stress-test tools section, but I also used Memtest86 and 64.

 

[Roland Of Gilead]

Is it possible to row back on the vbios to the previous one!? Did you back up the original vbios? With checking for so many things on the fly, introducing a vbios update when not required, might have boinked things.

 

[Dajsus]

Is it possible to row back on the vbios to the previous one!? Did you back up the original vbios? With checking for so many things on the fly, introducing a vbios update when not required, might have boinked things.

It is a dual bios card, I tired using the "Silent" one which was not flashed and the same issue happens.

 

[Dajsus]

As said all of the updates were done one by one not on the same day this has been going on for 3 weeks.

 

[Roland Of Gilead]

It is a dual bios card, I tired using the "Silent" one which was not flashed and the same issue happens.

Ah, understood.

 

[Dajsus]

Have you tested the ram outside of Windows, with Memtest86+? If not, do that, and let us know.

Freezes like that can be ram related, and/or conflicts.

Presuming all system drivers, including chipset drivers are up to date? (Edit- I see you've done this)What power plan are you running? Ryzen balanced?

Also to add I tried all of power plans but for some reason, after the last Windows reinstall, I don't see the Ryzen power plan. And yeah chipset drivers are installed.

 

[Ralston18]

Reliability History: What, if known, was done or happened November 18th & 19th?

There was some sort of problem that did not appear again during the next 4- 5 days. What were those errors? Will surmise that there errors may have been "Windows not properly shutdown" because you were forced to power down via the power switch or unplugging the system. Which, in turn, can and does corrupt files and configuration settings.

Look at the informational events that occurred on those dates. Any dates or informational events that do not "fit" within the computer's environment and configuration settings?

Remember: you can use the vertical divider bars between the details columns to expand the column to full width.

Also in the Action column you can click View technical details to obtain more information. Unfortunately that information may or may not be helpful.

 

[Dajsus]

@Ralston18
17th is the Date of the fifth fresh Windows reinstall, 18th is going to be a power loss in the house, 19th is a Blue screen from my Focusrite Sound card since I installed the wrong driver by mistake. Then there were 4 days of nothing happening and 24th the issue happened again.

 

[Ralston18]

Expand the 24th to show all entries and include the Technical Details when the blue links are clicked.

How are you doing the Windows reinstalls?

FYI (from this Forum):

https://forums.tomshardware.com/faq/windows-10-clean-install-tutorial.3170366

 

[Dajsus]

@Ralston18 Compared to the Forum guide I do it exactly the same except I use Pro version.
19th Image

And the memory.dmp file from that day.BSOD was triggered then, and I went to check the drivers realizing it was a wrong driver I installed by mistake.

PROCESS_NAME: System

STACK_TEXT:
ffff80013703d158 fffff80077cf848e : 000000000000001e ffffffff80000003 fffff8009d61005d 0000000000000000 : nt!KeBugCheckEx
ffff80013703d160 fffff80077c039c2 : fffff80077cf846c 0000000000000000 0000000000000000 0000000000000000 : nt!HvlpVtlCallExceptionHandler+0x22
ffff80013703d1a0 fffff80077a77a77 : ffff80013703d710 0000000000000000 fffffd057e42fc20 fffff80077bfdbce : nt!RtlpExecuteHandlerForException+0x12
ffff80013703d1d0 fffff80077a76676 : fffffd057e42ee38 ffff80013703de20 fffffd057e42ee38 000000000000ffff : nt!RtlDispatchException+0x297
ffff80013703d8f0 fffff80077bfa3e2 : 0000000000000000 0000000000000000 0000000000000000 0000000000000000 : nt!KiDispatchException+0x186
ffff80013703dfb0 fffff80077bfa3b0 : fffff80077c0d465 0000000000000005 fffff80077a22816 ffff800100000000 : nt!KxExceptionDispatchOnExceptionStack+0x12
fffffd057e42ecf8 fffff80077c0d465 : 0000000000000005 fffff80077a22816 ffff800100000000 ffffc509c7cdd480 : nt!KiExceptionDispatchOnExceptionStackContinue
fffffd057e42ed00 fffff80077c064d0 : ffffc509c7cdd480 0000000000000000 0000000000000246 fffff80077a21ff1 : nt!KiExceptionDispatch+0x125
fffffd057e42eee0 fffff8009d61005e : ffffc509ca547600 0000000000000000 0000000000000018 0000000000000000 : nt!KiBreakpointTrap+0x310
fffffd057e42f070 ffffc509ca547600 : 0000000000000000 0000000000000018 0000000000000000 ffffc509ca4aef00 : FocusriteUsb+0x2005e
fffffd057e42f078 0000000000000000 : 0000000000000018 0000000000000000 ffffc509ca4aef00 fffff8009d5fef78 : 0xffffc509`ca547600


SYMBOL_NAME: FocusriteUsb+2005d

MODULE_NAME: FocusriteUsb

IMAGE_NAME: FocusriteUsb.sys

STACK_COMMAND: .cxr; .ecxr ; kb

BUCKET_ID_FUNC_OFFSET: 2005d

FAILURE_BUCKET_ID: 0x1E_80000003_FocusriteUsb!unknown_function

OS_VERSION: 10.0.19041.1

BUILDLAB_STR: vb_release

OSPLATFORM_TYPE: x64

OSNAME: Windows 10

FAILURE_ID_HASH: {f2ed2671-538d-3884-a031-60a1401fac66}

Followup: MachineOwner

 

[Ralston18]

So Windows 11 is installed - correct?

However the memory dump is showing Windows 10 information: OS_VERSION and OSNAME

So the correct FocusriteUsb drivers are now installed and all is well?

Reference:

FocusriteUsb:

This device?

https://focusrite.com/en/usb-interfaces

 

[Dajsus]

Its on Win10 now.
And thats correct one of them.

 

[Dajsus]

I think I maybe found the culprit, seems like the 24 pin was like 1mm out on one of the sides and was not fully in. I re-did all cabling since I think the CM TD500 Mesh case I have has a bit small PSU compartment and required me to push the PSU in a bit so which most probably popped the cable out a bit. Will update in a few days.

 

[Shesroruy]

1. I did check it I forgot to add:

Problem Event Name: LiveKernelEvent
Code: 1a1
Parameter 1: ffffc40568bd0040
Parameter 2: 0
Parameter 3: 0
Parameter 4: 0
OS version: 10_0_19045
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19045.2.0.0.256.48
Locale ID: 1033


2 From the AMDs website.
3 NZXT PSU cables were used only

Just to add I serviced a lot of PCs over the years since I used to work in PC part sales just never came across something this random.
I mean reading the reliability history I totally missed the Hardware error part since the summary was hidden like this for some reason.
https://prnt.sc/eqvjTtG_G-24/snow rider

do you know someone who could test it in their system and has a strong power supply?