PC turning off in games

[calluM]

Hi, wondering if anyone can help with any further troubleshooting tips before looking at replacing PSU.

When opening any game my PC will just turn off. Originally this started with just CS2 when in lobby, if I could make it into the game it wouldn't do it at all until I returned back to the lobby, but has since started happening on all games I play.

Specs:

i7-9700k (No OC)
NZXT Kraken X62 280MM
ROG Strix OC RTX 2080 (No OC)
32GB DDR4 Corsair Vengeance RAM
MSI Z390 Gaming Pro Carbon AC mobo
SeaSonic Plus 750W Gold
3x Samsung 860 Evo 1TB
Windows 11 Pro 23H2 22631.3085

All of these parts are 6 years old, PC is regularly cleaned.

So far I've tried:

-Reinstalling drivers
-Disabling overlays
-BIOS Update
-Reseated CMOS just to reset BIOS to default
-Using 1 stick of RAM at a time to rule out RAM error
-Trying games installed on each SSD to see if it's SSD specific
-Checking temps (CPU and GPU temps are well below point of shut down)
-Reinstalled Windows & wiped files
-Ran Cinebench for 2 hours (all cores max temp 65 & no shutdown)
-Ran Unigene Heaven for 30 minutes on max settings, no throttling or overheating

Temps monitored with HWMonitor

Just wondering if theres anything else you'd recommend trying before I start buying new hardware?

[calluM]

Oprindeligt skrevet af smallcat:

i had over 200 shutdowns . It turned out the technician leaving the cables in the case made some of them touch the RAM sticks . i pulled the cables away from the memory sticks and issue is gone . Over 2 months everything is fine .

Thanks for your response, I'll reseat all of the power cables and tidy them up so theres nothing touching. Appreciate it

[_I_]

psu brand/model/age?

grab the correct drivers
https://www.msi.com/Motherboard/MPG-Z390-GAMING-PRO-CARBON-AC/support#driver

chipset, lan, audio

[calluM]

Oprindeligt skrevet af _I_:

specs?

psu brand/model/age?

All in original post:


i7-9700k (No OC)
NZXT Kraken X62 280MM
ROG Strix OC RTX 2080 (No OC)
32GB DDR4 Corsair Vengeance RAM
MSI Z390 Gaming Pro Carbon AC mobo
SeaSonic Plus 750W Gold
3x Samsung 860 Evo 1TB
Windows 11 Pro 23H2 22631.3085

All of these parts are 6 years old, PC is regularly cleaned.

[calluM]

Oprindeligt skrevet af _I_:

psu brand/model/age?

grab the correct drivers
https://www.msi.com/Motherboard/MPG-Z390-GAMING-PRO-CARBON-AC/support#driver

chipset, lan, audio

Thanks for the tip. All up to date drivers for chipset, lan, audio, bluetooth & gpu, probably should have put that in the OP, my bad.

[_I_]

from the mobo mfg site, not windows update?

[calluM]

Oprindeligt skrevet af _I_:

from the mobo mfg site, not windows update?

Hi, got them from the MSI z390 support page after reinstalling Windows a couple of days ago.

[calluM]

Not sure if relevant, I do see in Event Viewer a few of the same error shown below just before the critical error when PC shuts down.

ACPI thermal zone \_TZ.TZ00 has been enumerated.
_PSV = 0K
_TC1 = 0
_TC2 = 0
_TSP = 0ms
_AC0 = 344K
_AC1 = 328K
_AC2 = 323K
_AC3 = 318K
_AC4 = 313K
_AC5 = 0K
_AC6 = 0K
_AC7 = 0K
_AC8 = 0K
_AC9 = 0K
_CRT = 392K
_HOT = 0K
minimum throttle = 0
_CR3 = 0K

A quick Google search brings up a bunch of forums of people saying redo thermal paste, my paste is 6 years old so I'll be reapplying some anyways but my temps when running cinebench for 2 hours didn't pass 65 degrees whilst pinned at 100% on all cores so I don't believe it to be a temp issue?

[_I_]

the cpu can run without a cooler,
if its not telling windows how its throttling that can cause crashes, but should not turn off on its own

powering off is most likely psu problem
thats why i ask for psu brand/model and its age

[calluM]

Oh I see. Pretty sure my PSU has 10 year warranty so might see if theres anything I can do with that if I can put it down to the PSU. I’ve just noticed it works fine when fps capped to 60, as soon as I put it higher PC instantly turns off?

[_I_]

when it turns off, what do you need to to to turn it back on again?

[calluM]

Oprindeligt skrevet af _I_:

when it turns off, what do you need to to to turn it back on again?

Power button to turn it back on, it doesn't restart or anything just a click type noise my PSU's always done when turning off, all the RGB turns off and fans stop spinning.

[Phénomènes Mystiques]

Disable, or turn down to lowest setting, C-states in BIOS and see if it shuts down. Especially if using XMP/DOCP; in which is considered RAM OC (over clocking).

C-states and XMP/DOCP does not always get along as C-states is a CPU throttler.

It is not the thermal paste. If you do re-paste, which there should be no need unless you removed the heatsink/AIO, please use Arctic Silver or Graphene, best on the market - silver is top rated for longevity up to 10 years+.

You should also check the power saving modes and in Advanced settings check to see if its the power saving mode in Advanced Power Settings: Turn off HDD after xx amount of minutes as this was an issue for me with one certain Windows Update setting it to 5 minutes instead of 240 which caused games to crash while paused.

Also make sure TPM 2.0 is enabled as well as Secure Boot in the BIOS for Windows 11.

You also should be running BIOS revision 7B17v1C.

If you have BIOS revision 7B17v1C, 100% make sure TPM 2.0 and Secure Boot are enabled as this version fixes a few issues with both features.

Chipset driver version 10.1.18508.8239

Could also try increasing Virtual Memory a bit. Raise VM to 16,384MB - 32,768MB with an initial range of 4983MB (also make sure you have that much space on said drive). I highly doubt this is an issue, but worth it to test/rule out.

I am leaning towards culprits TPM and Secure Boot possibly. If so, these errors would appear in the Event Viewer as TPM failed to engage or similar.

Seeing your BIOS updates have re-established these two feature defaults, it is a possibility they turned them off/on and without one, the other will fail as per my own personal TPM research has suggested.

Seeing you installed/re-installed the BIOS, all settings were defaulted and this also can cause issues if you have custom settings, such has XMP and its much needed OC Power Profile unless you saved/loaded a previous BIOS settings custom profile.

And one last thing to check, GPU Hot Spot although I am sure this already has been checked.

I hope this helps a bit.

Oprindeligt skrevet af _I_:

the cpu can run without a cooler

Never run a CPU, or GPU, without a cooling solution!

[calluM]

Oprindeligt skrevet af Wynters:

Disable, or turn down to lowest setting, C-states in BIOS and see if it shuts down. Especially if using XMP/DOCP; in which is considered RAM OC (over clocking).

C-states and XMP/DOCP does not always get along as C-states is a CPU throttler.

It is not the thermal paste. If you do re-paste, which there should be no need unless you removed the heatsink/AIO, please use Arctic Silver or Graphene, best on the market - silver is top rated for longevity up to 10 years+.

You should also check the power saving modes and in Advanced settings check to see if its the power saving mode in Advanced Power Settings: Turn off HDD after xx amount of minutes as this was an issue for me with one certain Windows Update setting it to 5 minutes instead of 240 which caused games to crash while paused.

Also make sure TPM 2.0 is enabled as well as Secure Boot in the BIOS for Windows 11.

You also should be running BIOS revision 7B17v1C.

If you have BIOS revision 7B17v1C, 100% make sure TPM 2.0 and Secure Boot are enabled as this version fixes a few issues with both features.

Chipset driver version 10.1.18508.8239

Could also try increasing Virtual Memory a bit. Raise VM to 16,384MB - 32,768MB with an initial range of 4983MB (also make sure you have that much space on said drive). I highly doubt this is an issue, but worth it to test/rule out.

I am leaning towards culprits TPM and Secure Boot possibly. If so, these errors would appear in the Event Viewer as TPM failed to engage or similar.

Seeing your BIOS updates have re-established these two feature defaults, it is a possibility they turned them off/on and without one, the other will fail as per my own personal TPM research has suggested.

Seeing you installed/re-installed the BIOS, all settings were defaulted and this also can cause issues if you have custom settings, such has XMP and its much needed OC Power Profile unless you saved/loaded a previous BIOS settings custom profile.

And one last thing to check, GPU Hot Spot although I am sure this already has been checked.

I hope this helps a bit. :csdsmile:

Oprindeligt skrevet af _I_:

the cpu can run without a cooler

Never run a CPU, or GPU, without a cooling solution!

Thanks for the tips, just had a look at all of the above, HDD turn off time was set to never, i presume thats default on ultimate performance mode? I'm running BIOS ver 7B17v1C, chipset ver 10.1.18508.8239, TPM 2.0 & Secure Boot were enabled.

C-State was enabled, I've disabled that and will test, if that fails I'll try the page failing just to see if either resolve so I'll know which fix it was!

My bios settings before flashing included my CPU OC inc. voltage, loadline etc. but not fussed about applying any overclocks so I've only enabled XMP since flashing.

Whilst running the heaven benchmark for 2 hours, my highest GPU hotspot was 81.8c. It did hit power limit, reliable voltage limit and operational voltage limit though, not sure if thats significant.

[Phénomènes Mystiques]

Oprindeligt skrevet af calluM:

Thanks for the tips

You are welcome.

Oprindeligt skrevet af calluM:

just had a look at all of the above, HDD turn off time was set to never, i presume thats default on ultimate performance mode?

That would be correct.

Oprindeligt skrevet af calluM:

I'm running BIOS ver 7B17v1C, chipset ver 10.1.18508.8239, TPM 2.0 & Secure Boot were enabled.

Good.

Oprindeligt skrevet af calluM:

C-State was enabled, I've disabled that and will test, if that fails I'll try the page failing just to see if either resolve so I'll know which fix it was!

It may be for the best, or at it's lowest setting besides off.

I also forgot to mention that you may even find an extra C-state called C6 (or similar) somewhere on its own and that is the highest C-state I have seen on newer boards.

Oprindeligt skrevet af calluM:

My bios settings before flashing included my CPU OC inc. voltage, loadline etc. but not fussed about applying any overclocks so I've only enabled XMP since flashing.

XMP/DOCP is considered an overclock. The thing you should check if the VCore voltage. I seen with my AMD, they had the VCore at 0.9v when proper it should have been 1.35v.

Just something to check, Intel usually is good with core voltage though.

And if the BIOS has any power settings relating to OC, please turn those to On when using XMP.

Oprindeligt skrevet af calluM:

Whilst running the heaven benchmark for 2 hours, my highest GPU hotspot was 81.8c. It did hit power limit, reliable voltage limit and operational voltage limit though, not sure if thats significant.

Like I said, I highly doubt it is a thermal issue. 81,8c at that state is normal.

[_I_]

Oprindeligt skrevet af Wynters:

Oprindeligt skrevet af _I_:

the cpu can run without a cooler

Never run a CPU, or GPU, without a cooling solution!

what?
https://www.youtube.com/watch?v=ycIF1NDkW6M

thats a pentium g, but still functional with no cooler
the i7 will do the same thing with proper drivers installed