My custom built PC crashes very often

Rudmer15

Prominent
Jul 22, 2017
19
0
510
I built a PC with a friend and all the parts that we bought are new. When i'm playing a graphically intense game for around 10 minutes on the highest settings, my pc just stops and shows a random color (depending on what i saw before the crash) then i hear a few sounds, and it freezes and i have to hold the power button down and reboot. This is very frustrating for me and i've already asked for some advice here on Tom's Hardware and many users said it could be the CPU-cooler, which i replaced. then they said it could also be my PSU (fairly low quality). I just bought a brand-new corsair CX550, but still, it crashes after around 10 minutes 🙁
I don't really like the idea of having to spend more money on stuff i don't need, so help is really appreciated.

here are my pc's specs:
CPU: AMD FX-6300
GPU: asus Radeon ROG RX-460 STRIX
MOBO: AsRock 970m Pro3
PSU: Corsair CX550
COOLER: HYPER 212 EVO
OS: Windows 10 Professional

I hope you could help me resolve my problems, and you can also take a look at my other posts. Thank you in advance!
 
Solution
By selecting all, it pushes all 6 cores of my FX 6300 to 100%. If I just select 1 or a few tests, the load bounces around the 6 cores and fluctuates, but doesn't reach 100%.

Edit: It is not like Passmark performance test that runs one test, then the next. Overdrive runs all of the tests simultaneously. There are 3 tests per core, CPU Test (integer calculation and stack operation), FPU test (floating point calculation) and Calculation Test.


my temps are very stable when playing, no overclocks, and i just updated my graphics card drivers!
 


The CX550 isn't necessarily a high quality PSU.
Are you getting any crash log messages when you reboot into windows?
What was your old PSU?

I don't know why someone suggested your CPU cooler. The description you're giving indicates something is happening to the graphics card - not saying the graphics card needs to be replaced, but just that the error is related to it. The graphics card takes power from the PSU and motherboard, runs the software, and displays an image (simplified for this context). I assume you've already updated your drivers for the card, lacking other details I don't think it's the motherboard. The graphics card is recent. It leads me to believe you're pulling power that doesn't exist or your card is overheating or it's a simple driver issue (make sure it's up to date)

What is the model of your case that all of this is inside?
 


I may be reading more into this than I am allowed but after reading your description I then exclaimed NO WAY!!!(sarcasm before coffee)when I saw the Radeon card. AMD and Nvidia cards seem to crash much differently than each other. They will also both go black and then back to normal with a Windows(I think) lending a hand by disabling and enabling the devices.

So, in my uninformed opinion it is your GFX card that is the issue.

What led you to the PSU decision?
 


my case is a sharkoon VG4-W
 


I already made a tread about my problem and someone said my PSU was of very low quality and since he probably knows more about psu's than i do i thought i should indeed replace it. maybe my gpu\cpu was consuming too much electricity than my PSU can handle? i'll try to update the drivers for my gpu now.
thanks :)
 


What does "my temps are very stable" mean? I dont care if they are stable I want to know if they are too hot. That cpu for example, amd recommends it not exceed 62c package temp, it will throttle at 70c and thermal shutdown and/or instability will occur if it manages to get any hotter than that. My guess is something is unstable most likely the cpu or there is an issue with your gpu drivers. I doubt its the psu, the cx550 is not very good but if this same thing happened with another power supply I would say its very unlikely the issue is related to the power supply. The cx550 just uses some sub standard parts like cheap Chinese caps that hurt its logenvity, but it performs alright and should be more than capable of powering that pc.
 


i did a furmark stresstest\benchmark and my gpu didn't fail. i'm still going to update my GPU drivers and i'll keep you informed!
 


I'll admit I haven't heard about the LC-Power LC6460GP3 because it's a German product. However it doesn't look like a bad PSU. Sure, it's a dual rail design with the rails rated at 16A and 19A. That isn't all that high but more than enough for the 460. It is also 80+ Silver which means it's an efficient PSU.

While I agree with SR-71 Blackbird's suggestion for a new PSU I disagree with DSzymborski's assertion that it's a garbage tier. He didn't offer any sort of evidence to support his assumption.

What would help is the aforementioned Event Viewer info. When it crashes what does the Event Viewer say?

I also suggest er wait a minute. You said your CPU was 48. Wait one

At first i thought it was the processor, so i bought the Hyper 212 but it still happens. I stress tested my pc with AIDA64 Extreme. After 12 minutes of stress-testing the processor it happened again. Weird thing is: when i bought my Hyper 212 (2-05-2017) i immediately stress tested and it could go on for 2,5 hours without crashing and now it happens after 12 mins even though the temperatures dont go higher than 48.

Where did you get that temperature?

http://www.amd.com/en-us/innovations/software-technologies/technologies-gaming/over-drive is the only utility I would use for an AMD CPU. It also has an integrated CPU stress test that I would run while you have some coffee, tea, hoice or just sit back and relax until it's finished.

Don't understand the way AMD offers thermal margins? http://www.tomshardware.com/faq/id-2122665/understanding-temperature-amd-cpus-apus.html
 


Downloaded it, how do i start the stress-test? and somehow it isnt showing any information in the GPU section.
EDIT: I did find a section where it says Benchmark so i ran that. System performance shows: 9925
Also found a section that says Stability test but it has a lot of boxes i can tick so i need to know which one
 


http://www.amd.com/Documents/AMD_Dragon_AM3_AM2_Performance_Tuning_Guide.pdf says something about

The AMD OverDrive™ utility features an impressive system stability assessment. The
“Stability Test” tab under “Performance Control” incorporates AMD‟s CPU test algorithms to
stress the CPUs worst-case scenario.

I don't have an AMD CPU so I am not familiar with the stress test.

You could also run Prime95
SMALL FFTs only

Watch your thermal margins while prime runs. How low do they get? I am looking for a max temp er margin.
 


Sounds logical. Can't say if selecting all is best or not because I am not familiar with OD.
 
By selecting all, it pushes all 6 cores of my FX 6300 to 100%. If I just select 1 or a few tests, the load bounces around the 6 cores and fluctuates, but doesn't reach 100%.

Edit: It is not like Passmark performance test that runs one test, then the next. Overdrive runs all of the tests simultaneously. There are 3 tests per core, CPU Test (integer calculation and stack operation), FPU test (floating point calculation) and Calculation Test.
 
Solution


Merci beaucoup! Your insight is appreciated. I was lost.