PC Erratically shuts down, when starting up certain games.

[DiegoLavega77]

Hey everyone, my name's Diogo and since i was very young ive been building my PC piece by piece untill today, being that its sort of a frankenstein piece of hardware but one that has worked for me for long years, being modified every so often but allways with the help either of professionals or friends that knew how to work with the matter better than i did.

 

Long story short i allways had small issues here and there but this is a concerning one that recently started happening, its relevant to add that i bought a 3060TI OC from zotac (God bless the website which let me ship one for myself) about 2 to 3 months ago changing from a old 1050TI from asus.

That being said my current build is as follows: 
AMD Ryzen 5 2600X Six core

2 sticks of 8 GB G.Skill RAM DDR4 And 3000MHZ i believe 

I have a PSJET 700W PSU that i bought quite a long while ago, brand new.

And a SSD 480GB 

All mounted on a PRIME A320M-K motherboard, all of this bought brand new at a professional retailer from portugal. (Chip7) they did the mounting for the processor motherboard and ram also cleaning the PC, which is something i do regularly aswell (Cotton swabs and a brush),

 

My issue started recently about a month ago to now, where sometimes as i start certain game just as its reaching its menus shuts down my PC, and it doesnt turn back on unless i disconnect all the cables completely, press the power button a couple of times and then turn it back on, which i do by advice from a friend of mine since this can ocasionally happen from current issues, so much so that i have pondered buying a new case being afraid that this is the issue.. also because the case is horrible and old.. but thats besides the point, which is that money is tight as i lost my job due to covid.

 

The games that this have happened on boggle me if its a CPU issue since they arent the most intensive but they are:

Days Gone,

Deep Rock Galactic
It takes two

 

Every time this happens i open up other games to see if this also happens (For example death stranding) and everything works fine, but i go back and it turns off again. Then a couple days later i try again and the game that was shutting down works fine?

 

Attatched are CPUID files and 2 pictures from the case and the back in hopes to shed some light on other relevant information that i might not understand as well as other people.

 

Thank you in advance to everyone replying to this thread, as i wanted some advice before spending money on professional help.

 

DESKTOP-1T9AOIM.txt

[Pixelfie]

Is it overheating? Seeing how much cables it has there's probably no airflow. And it seems you have no fans in the rear. 

 

You also have a weird PSU, you should never buy one from an unknown brand like this. Buy from brands like Corsair, Seasonic or Be Quiet instead.

[DiegoLavega77]

2 minutes ago, Pixelfie said:

Is it overheating? Seeing how much cables it has there's probably no airflow. And it seems you have no fans in the rear. 

 

You also have a weird PSU, you should never buy one from an unknown brand like this. Buy from brands like Corsair, Seasonic or Be Quiet instead.

I always had concerns with my PC overheating since the case is old and has no fans, i left my PC pannel open and i had no issues, that being said im not really familiar with monitoring its temperature but being that the games shutdown my PC even before i start playing questions the fact that its not under (I believe) high pressure when it shuts down.

 

I definitely agree that i should have a better PSU, this one was bought by a friend of mine when i was younger since the budget wasnt very high, that being said it never really gave me any apparent issues.

[8-Bit Ninja]

1 minute ago, DiegoLavega77 said:

I always had concerns with my PC overheating since the case is old and has no fans, i left my PC pannel open and i had no issues, that being said im not really familiar with monitoring its temperature but being that the games shutdown my PC even before i start playing questions the fact that its not under (I believe) high pressure when it shuts down.

 

I definitely agree that i should have a better PSU, this one was bought by a friend of mine when i was younger since the budget wasnt very high, that being said it never really gave me any apparent issues.

I'd start with a new PSU. Do you have amazon in Portugal and can you return items up to 30 days for free? If so id do what @Pixelfiesaid and buy a new quality one and see if that solves your issues, if it doesn't you can always return it. 

[Pixelfie]

Download speccy and look at the temps of both the CPU and GPU. Important is that you look at it when playing games. The GPU should be under 80 and CPU under 70, while they can get hotter before shutting down, that still means there's some issue with cooling.

 

Check here which PSU would be best for you: 

 

[SorryBella]

3 minutes ago, DiegoLavega77 said:

I definitely agree that i should have a better PSU, this one was bought by a friend of mine when i was younger since the budget wasnt very high, that being said it never really gave me any apparent issues.

Yeah, definitely power delivery issue. 3000 series draws pretty high under load and any PSU that cant handle its power rating properly pretty much can knock its OCP really fast.

[DiegoLavega77]

3 minutes ago, 8-Bit Ninja said:

I'd start with a new PSU. Do you have amazon in Portugal and can you return items up to 30 days for free? If so id do what @Pixelfiesaid and buy a new quality one and see if that solves your issues, if it doesn't you can always return it. 

 

I looked into amazon portugal but it doesnt exhist in portugal only spain, so the return policy applied if we shipped the product back to spain through national mail, which would come at a loss, though i was hoping this would be a good solution, thank you anyways for the reply!

[SamClan]

it is crashing time, you need

Hwmonitor - for temps

Prime95 - stress testing CPU and RAM

Furmark - Stress testing GPU

 

 

you let run HWMonitor and start with Prime95 let it run 15 - 30 mins have an eye on the Temps 

after that turn the programm off via icon, the x just minimize it

Furmark, let it also run for 15-30 min, check your temps

 

after that run both at the same time 

 

 

figure out where it crashes. if it only crashes if both of the programs are running try and get a new PSU. 

install the new one do the test again and see if it also crashes if not good. if it crashes you might send the PSU back and get your money 

[DiegoLavega77]

3 minutes ago, Pixelfie said:

Download speccy and look at the temps of both the CPU and GPU. Important is that you look at it when playing games. The GPU should be under 80 and CPU under 70, while they can get hotter before shutting down, that still means there's some issue with cooling.

Will definitely try that but how would this work if the PC shuts down completely even before getting ingame? the only thing i do monitor is my GPU under stressfull games with MSI and it never goes above 71 even running cyberpunk on max settings with ray tracing on the best setting

 

[DiegoLavega77]

5 minutes ago, SorryClaire said:

Yeah, definitely power delivery issue. 3000 series draws pretty high under load and any PSU that cant handle its power rating properly pretty much can knock its OCP really fast.

My best guess is PSU so far from the answers so ill try the fix another fellow left and test it, thanks for the reply.

[DiegoLavega77]

2 minutes ago, SamClan said:

it is crashing time, you need

Hwmonitor - for temps

Prime95 - stress testing CPU and RAM

Furmark - Stress testing GPU

 

 

you let run HWMonitor and start with Prime95 let it run 15 - 30 mins have an eye on the Temps 

after that turn the programm off via icon, the x just minimize it

Furmark, let it also run for 15-30 min, check your temps

 

after that run both at the same time 

 

 

figure out where it crashes. if it only crashes if both of the programs are running try and get a new PSU. 

install the new one do the test again and see if it also crashes if not good. if it crashes you might send the PSU back and get your money 

Gonna test this right now to see if i can get a pinpoint on which one would make it go down, thank you

[DiegoLavega77]

18 minutes ago, SamClan said:

it is crashing time, you need

Hwmonitor - for temps

Prime95 - stress testing CPU and RAM

Furmark - Stress testing GPU

 

 

you let run HWMonitor and start with Prime95 let it run 15 - 30 mins have an eye on the Temps 

after that turn the programm off via icon, the x just minimize it

Furmark, let it also run for 15-30 min, check your temps

 

after that run both at the same time 

 

 

figure out where it crashes. if it only crashes if both of the programs are running try and get a new PSU. 

install the new one do the test again and see if it also crashes if not good. if it crashes you might send the PSU back and get your money 

Found the culprit!

So after stressing my CPU and RAM just fine, with a 45 minimum tempo and max 83, 

 

Just as i started furmark it went on for 2 to 3 seconds and shutdown again, 

 

So im guessing this means this might be a PSU not handling the GPU issue? since the GPU is pretty recent?

[SamClan]

Just now, DiegoLavega77 said:

Found the culprit!

So after stressing my CPU and RAM just fine, with a 45 minimum tempo and max 83, 

 

Just as i started furmark it went on for 2 to 3 seconds and shutdown again, 

 

So im guessing this means this might be a PSU not handling the GPU issue? since the GPU is pretty recent?

was it shutting down while you were running Prime95? or did you run Furmark on its own?  

[DiegoLavega77]

Just now, SamClan said:

was it shutting down while you were running Prime95? or did you run Furmark on its own?  

Ran furmark on its own after running prime95 on its own and i made sure to shutdown prime 95 after closing it, and watched the prime95 test temperature go back to normal levels before starting the other one!

[SamClan]

could be still faulty GPU or Motherboard but go with the PSU, don't cheap out on your PSU AND buy it where you can give it back and get your money. 

[DiegoLavega77]

Just now, SamClan said:

could be still faulty GPU or Motherboard but go with the PSU, don't cheap out on your PSU AND buy it where you can give it back and get your money. 

Is there any way to rule out the GPU or Motherboard?

[SamClan]

1 minute ago, DiegoLavega77 said:

Is there any way to rule out the GPU or Motherboard?

GPU well you could swap it into another system that runs stable. and see if you have the same problems. 

 

Motherboard only if you have another Motherboard. swap it and everything works fine, Motherboard 

 

BTW: the 12 pin Power for GPU is it on cables that are also plugged in into something else? 

 

what you also could do is get GPU-Z, Log to file active left button corner, and look what your GPU is drawing on power

[Pixelfie]

9 minutes ago, DiegoLavega77 said:

Is there any way to rule out the GPU or Motherboard?

Try in a different system. If you don't have one ask a friend if they wanna help. You can also check your PSU that way, but you should replace that anyway, before it fries all parts.

[DiegoLavega77]

7 minutes ago, SamClan said:

GPU well you could swap it into another system that runs stable. and see if you have the same problems. 

 

Motherboard only if you have another Motherboard. swap it and everything works fine, Motherboard 

 

BTW: the 12 pin Power for GPU is it on cables that are also plugged in into something else? 

 

what you also could do is get GPU-Z, Log to file active left button corner, and look what your GPU is drawing on power

So i ran the GPUZ sensor,opened up frostpunk ran the game and closed it again, im not sure what the yielded info means though, i attatched both the sensor log and a screenshot of what the app looked like, 

 

I believe by the 12 pin you mean the power cable from the GPU, its running into a adapter because my PSU didnt have enough cables, i hope this is what you were asking.

 

GPU-Z Sensor Log.txt

[DiegoLavega77]

6 minutes ago, Pixelfie said:

Try in a different system. If you don't have one ask a friend if they wanna help. You can also check your PSU that way, but you should replace that anyway, before it fries all parts.

 I know someone who can test the GPU and will definitely try that out, thanks for the suggestion ill definitely look into a better PSU as fast as possible

[SamClan]

Just now, DiegoLavega77 said:

So i ran the GPUZ sensor,opened up frostpunk ran the game and closed it again, im not sure what the yielded info means though, i attatched both the sensor log and a screenshot of what the app looked like, 

 

I believe by the 12 pin you mean the power cable from the GPU, its running into a adapter because my PSU didnt have enough cables, i hope this is what you were asking.

 

GPU-Z Sensor Log.txt 101.93 kB · 0 downloads

I hoped your system would crash and would have a crazy power spike in it. 

 

 

About the 12 pin: it is about if you have only the 12 pin adapter connected to the cable, GPU - Power Adapter - PSU OR GPU - Power Adapter - HDD - SSD - PSU. Some kind of overload on that cable 

so you could just use another cable to plug the adapter in

 

 

[DiegoLavega77]

41 minutes ago, SamClan said:

I hoped your system would crash and would have a crazy power spike in it. 

 

 

About the 12 pin: it is about if you have only the 12 pin adapter connected to the cable, GPU - Power Adapter - PSU OR GPU - Power Adapter - HDD - SSD - PSU. Some kind of overload on that cable 

so you could just use another cable to plug the adapter in

 

 

yes the adapter goes directly into it, GPU-> ADAPTER->PSU, i dont think i have any other adapters on my PC connected to anything else to test another cable,

 

So should i log this test but with the app that made it crash the last time to pinpoint the spike?

I dont usually mess with this, apologies for the weird info

 

PS: Not sure if its relevant but frostpunk was at the maximum possible specs.

[SamClan]

32 minutes ago, DiegoLavega77 said:

yes the adapter goes direcelt yas GPU-> ADAPTER->PSU, i dont think i have any other adapters on my PC connected to anything else to test another cable,

 

So should i log this test but with the app that made it crash the last time to pinpoint the spike?

I dont usually mess with this, apologies for the weird info

 

PS: Not sure if its relevant but frostpunk was at the maximum possible specs.

all good, no need to apologies. 

 

if you want you could do that. but no need yet. 

 

 

[DiegoLavega77]

19 minutes ago, SamClan said:

all good, no need to apologies. 

 

if you want you could do that. but no need yet. 

 

 

Alright so im headed to test the rest of the options thank you for all the help!

[DiegoLavega77]

48 minutes ago, SamClan said:

all good, no need to apologies. 

 

if you want you could do that. but no need yet. 

 

 

So in the latest reading i launched Days gone which is what consistently shuts down my PC and i got GPUZ to log up untill shutdown, can you recognize any spike at the end from this file?

GPU-Z Sensor Log.txt