GPU upgrade troubles, possible power failure?

[Sentient6]

Hi.

I've upgraded my GPU from an NVIDIA GTX 750TI to a Radeon RX 590, and I've experiencing weird crashes whenever I play games. 5-10 minutes into a game, the PC seemingly crashes, the sounds gets stuck on that weird feedback loop-like noise like you get from a BSoD, and both my monitors flash "No Signal" and go black as if the PC is off. But the PC isn't off - all the fans and all the lights are on. This persists until I shut it down or restart.

I assume it's power-related? Task Manager shows GPU loads of ~70-90% depending on the game (I've tried several), and MSI Afterbutner shows temperatures of barely over 60C. Afterburner doesn't show GPU voltage for whatever reason (says 0 mV), but it's been like that wit the 750TI as well.

My PSU is a Thermaltake Toughpower 650W, which should be plenty, right? It's not new, but I've never had problems with it. Both the 6-pin and the 8-pin power sockets connected to the PSU with its native cables, no adapters of molex or anything. I assume that everything is connected properly, because, well, it works until I start a game.

Specs:
Windows 10 64-bit
MSI Z97 GAMING 7
I5 4650k
8GB RAM
Adrenaline drivers 19.1.1 (NVIDIA drivers removed with DDU in safe mode)

Also, don't know if it's related (I'm guessing yes), but as I type this I have a video playing on the second monitor (in VLC), and whenever I switch active windows, I get weird glitches where the screen goes black for a second or the video lags for a few moments.

Is in the GPU? Or is it the power supply? Any way to tell without just getting a new power supply?

EDIT: Just to make sure, I ran DDU again (in safe mode), removed the AMD drivers, restarted, ran it again to remove NVIDIA drivers, restarted, installed AMD drivers version 18.12.2, and restarted again. That seeming fixed the VLC/video flicker/glitches on active window switch, but I still got the same crash in a game, 5 minutes into playing

 

[Dark Lord of Tech]

What PSU?

 

[Sentient6]

What PSU?

Thermaltake ToughPower QFan 650W. Like I said in the post, it's not new, but never any problems. Of course the 590 is a much more power hungry card than the 750TI...

 

[Dark Lord of Tech]

Had the same issue with my RX 580'S three of them Nitro+ SE and LE, had to RMA all 3 of mine.

 

[Sentient6]

So a GPU problem. Ok, I'll check the stores's return/exchange policy..

 

[Sentient6]

Bit of an update, I placed the GPU into a different PCIe slot on the motherboard and used a different cable to connect it to the PSU. It didn\t work, and now I'm getting those crashes when not in a game, just browsing. One time I didn't restart the PC immediately after a crash and after 15-20 seconds it shut off completely (then turned on fine).

So should I buy a new PSU or just try and return the GPU?

 

[Sentient6]

An update:

The issue had to do with the card's HDMI ports.

I took the GPU to the store I got it, and rather than exchange it outright they ran an overnight stress test. It ran fine. So I took the card back and bought a new PSU (SeaSonic Prime Gold 650W). Encountered the exact same issue (after ~5 minute of gameplay both monitors flash "No signal" and turn off, buzzing sound form the speakers, PC lights/fans still on). On a whim I decided to disconnect the monitors one by one. One of them is has a DVI connector, the other is a VGA connector, which I connect to the new GPU with an adapter (https://www.amazon.com/TechRise-High-Speed-adapter-Converter-Charging/dp/B01G1DL68E Not this exact one, but just like it).

The DVI monitor by itself worked without a problem for over 30 minutes. The moment I connected the second monitor (while still running a game), I got the crash. When I tried the VGA/HDMI monitor by itself, sure as shit I got the usual 5 minute crash. Same thing for both HDMI ports.

I also tried to see if it's the AMD audio drivers that were causing the issue (never had the problem with the 750TI), but I'm still getting the crash after disabling the "AMD High Definition Audio Device" in the Device Manager. The drivers are a bitch to remove, though, as the "AMD HDAD" keeps showing up shortly after I uninstall that driver from the Device Manager. But once I disabled it, it stays disabled.

So I'm guessing it's not the drivers (device disabled), and I can't fathom it being the adapter, so I'm taking the GPU back to testing.

 

[Dark Lord of Tech]

On 2 of mine none of the ports would detect a monitor as being attached.

Replacements are running fine so far.

 

[Sentient6]

An update to close the tread, in case anyone has similar problems:

It was the adapter after all. Or, rather, the combination of the adapter and the GPU. They worked fine separately (GPU with an HDMI monitor, the adapter with the old GPU), but put together it causes the crash. No fucking clue why, it just does. Confirmed this by replicating the crash on a system that runs the GPU fine with an HDMI monitor.

 

[Dark Lord of Tech]

Good to hear , I wasn't so lucky , never used an adapter all 3 still failed. Replacements are running fine so far all ports work.