Motherboard or PSU problem, not sure which or what to do

[Andarix]

Hello all, I recently rebuilt a large amount of my system (everything but RAM/PSU/HDD/peripherals) and it's been working fine for about a month. All of a sudden, however, when launching current gen games (Dragon Age Inquisition, Diablo 3, Metal Gear Solid Ground Zeroes) the computer will completely shut off. No warnings, no beeps, just power off. Sometimes it won't come back on immediately.

At first I thought it was my CPU, but the temps are running just fine, sometimes even cold. Now I'm torn between the motherboard being faulty and overheating or the PSU dying when it tries to load something demanding.

I ran OpenHardwareMonitor, AMD Overdrive and SpeedFan to check the temperatures of everything I could, idling the temperatures are 26C-34C on CPU, GPU and Mobo- but when launching a game one of the three mobo temps are reaching about 50C to 60C before the computer shuts off entirely.

The GPU is only about 6 mos old, the mobo and CPU are less than a month old, and the PSU is over two years old. So to me, I think it's a power supply issue and not a mobo one, but I want to make sure if that's the case before I go replacing things. The spike in temperatures on the motherboard is what's giving me pause.
What should I do?

Specs:
CPU: AMD FX-6300
GPU: XFX Radeon R9 280X
MB: Gigabyte 78LMT-USB3
Memory: 8GB Crucial
PSU: Thermaltake 600W

 

[MundoDragon]

Do you have a back-up PSU you can try. I would start there, based on the symptoms. Also, is there any way you can provide the actual model number of the Thermaltake PSU?

EDIT: Those temps are fine when under load. Nothing to be concerned about there.

 

[Andarix]

This is the power supply: http://www.thermaltakeusa.com/products-model.aspx?id=C_00001656

Unfortunately I don't have an extra PSU handy to test it.

 

[MundoDragon]

Make sure you have all the necessary power connectors connected to that graphics card. I would suggest to try running with the integrated graphics for a little while and see if the issue persists.

 

[Andarix]

I'm having issues trying to get my integrated graphics to work- any suggestions? Switching the plug to the integrated card gives me the POST long beep, 2 short beeps and no display. Otherwise, I reseated and replugged the other GPU in again and the problem still persists.

 

[MundoDragon]

Try resetting the CMOS on your motherboard to load fail safe defaults. Some boards have a jumper, some a button, yet others may require you to remove the battery for about 10-15 minutes.

 

[Andarix]

Try resetting the CMOS on your motherboard to load fail safe defaults. Some boards have a jumper, some a button, yet others may require you to remove the battery for about 10-15 minutes.

I got it to work with the integrated card- launching games no longer shuts down the computer, but there's no way they can run them, though.
Past that, I re-re-plugged in the Radeon card, made absolutely SURE the PCI-E plugs were fully in, correctly and it is still shutting down.

So that means the problem may be a snapped wire near/in the PCI-E power connectors? Which means new power supply? I may have an extra PSU available to me in a few hours so I have no way to check right now, is there anything else I can check in the meantime?

 

[MundoDragon]

Sounds to me like a bad video card. It would help if you could test it in another system or test this current system with a different discreet card, just to rule out a bad PCI-E slot or PSU power. Barring that, I would suggest RMAing it, if possible.

 

[Andarix]

Sounds to me like a bad video card. It would help if you could test it in another system or test this current system with a different discreet card, just to rule out a bad PCI-E slot or PSU power. Barring that, I would suggest RMAing it, if possible.

I got my hands on a friend's PSU- and everything worked with it. So it's definitely bad wires somewhere on the PSU.

Any suggestions for a good replacement?

 

[MundoDragon]

I would attribute it to the graphics card itself, rather than the PSU wires, unless you had another card to test the system with.

EDIT: Doh, just noticed you have tested the entire system with another PSU. I would simply replace it. I would suggest any 600-700 watt Antec or EVGA PSU that is in your budget.

 

[MundoDragon]

Also, if your PSU is modular, you may be able to try and order a new set of PCI-E cables from the manufacturer. They may even outright replace them if it's still under warranty. Might save you a few bucks.

 

[Andarix]

Unfortunately it's neither under warranty nor modular.

Thanks so much for your continued help.

 

[MundoDragon]

Then I would just grab a new PSU. You could probably get a good quality 600-700 watt for about $40-$50 USD, so it shouldn't break the bank.