PC freezes after 5-15 minutes while gaming

[dhaan]

Hi, I have had this problem for a while now, but i can't find a solution. maybe you guys can help?

My PC freezes after 5-15 minutes with all the games I play.
The screen will go black and the last played second of sound will go in a loop.
When i press the reset button the pc won't post, it has to be completely off before it boots normal.

System specs:

CPU: I7 4790K
GPU R9 290X
PSU Corsair AX 760
RAM 16GB corsair vengeance
MOBO: MSI Z97M
Cooler: Corsair H75

Temps while gaming:

CPU 48°
GPU 80°

 

[damric]

How old is that AX 760?

 

[dhaan]

How old is that AX 760?

It's only one year old.

 

[damric]

Alright this might be a drawn out process to troubleshoot. You have some instability somewhere. I doubt it is the PSU, but I'm not ruling it out yet. But lets run some tests on your RAM (Memtest 86+) and Prime95 custom blend with 80% of RAM used. If that's fine then we will stress the CPU, then the GPU, then both at the same time. What drive are you using? I ask that because freezing can also mean a bad HDD or SSD. And of course you also have motherboard symptoms with the resetting and POST not being quite right. You might end up having to breadboard this too, but all of this hassle will save much money compared to taking it to a shop. I'll be back in a few hours.

 

[dhaan]

Alright this might be a drawn out process to troubleshoot. You have some instability somewhere. I doubt it is the PSU, but I'm not ruling it out yet. But lets run some tests on your RAM (Memtest 86+) and Prime95 custom blend with 80% of RAM used. If that's fine then we will stress the CPU, then the GPU, then both at the same time. What drive are you using? I ask that because freezing can also mean a bad HDD or SSD. And of course you also have motherboard symptoms with the resetting and POST not being quite right. You might end up having to breadboard this too, but all of this hassle will save much money compared to taking it to a shop. I'll be back in a few hours.

I have a Samsung 840 evo 120GB and a seagate barracuda 2TB

prime95 seems to be running fine

memtest86 just finished with no problems

pc crashed after 20 minutes of furmark

 

[damric]

Alright this might be a drawn out process to troubleshoot. You have some instability somewhere. I doubt it is the PSU, but I'm not ruling it out yet. But lets run some tests on your RAM (Memtest 86+) and Prime95 custom blend with 80% of RAM used. If that's fine then we will stress the CPU, then the GPU, then both at the same time. What drive are you using? I ask that because freezing can also mean a bad HDD or SSD. And of course you also have motherboard symptoms with the resetting and POST not being quite right. You might end up having to breadboard this too, but all of this hassle will save much money compared to taking it to a shop. I'll be back in a few hours.

I have a Samsung 840 evo 120GB and a seagate barracuda 2TB

prime95 seems to be running fine

memtest86 just finished with no problems

pc crashed after 20 minutes of furmark

Furmark crashing indicates it would be either power supply or graphics card...however it is much more likely to be power supply. This could be a rare instance that a good quality PSU has gone bad. Unless you have a spare available to test, which one will you RMA first? I'd probably RMA the PSU first, myself.

Furmark quickly reveals power delivery weaknesses. Those could be on the card (bad VRMs, caps, or inductors), but more likely to be caused by the PSU. It definitely sounds like there is a defective component. A multimeter would help find it, and an o-scope would help even more, but most people don't have those.

 

[dhaan]

Alright this might be a drawn out process to troubleshoot. You have some instability somewhere. I doubt it is the PSU, but I'm not ruling it out yet. But lets run some tests on your RAM (Memtest 86+) and Prime95 custom blend with 80% of RAM used. If that's fine then we will stress the CPU, then the GPU, then both at the same time. What drive are you using? I ask that because freezing can also mean a bad HDD or SSD. And of course you also have motherboard symptoms with the resetting and POST not being quite right. You might end up having to breadboard this too, but all of this hassle will save much money compared to taking it to a shop. I'll be back in a few hours.

I have a Samsung 840 evo 120GB and a seagate barracuda 2TB

prime95 seems to be running fine

memtest86 just finished with no problems

pc crashed after 20 minutes of furmark

Furmark crashing indicates it would be either power supply or graphics card...however it is much more likely to be power supply. This could be a rare instance that a good quality PSU has gone bad. Unless you have a spare available to test, which one will you RMA first? I'd probably RMA the PSU first, myself.

Furmark quickly reveals power delivery weaknesses. Those could be on the card (bad VRMs, caps, or inductors), but more likely to be caused by the PSU. It definitely sounds like there is a defective component. A multimeter would help find it, and an o-scope would help even more, but most people don't have those.

I have a multimeter what do I need to measure?

 

[damric]

Measure the voltage drop at the PCIE power cable pins while the GPU is under load. There may also be various test points on the card depending on the model. DON'T cause a short or you will be sorry.

 

[dhaan]

I don't want to take the risk, I have send a rma request to the store i bought it from.

Thanks for the help!

 

[damric]

I don't want to take the risk, I have send a rma request to the store i bought it from.

Thanks for the help!

Yeah I would feel uncomfortable with someone who is not an electronics technician probing around. Unfortunately this needs to be done at load. But anyways, hopefully the RMA works out. If not, RMA the card. If that still doesn't fix it...we start over.