Question System shutting down.

Jun 29, 2020
10
1
15
Hi! I'll try to cut the long story short.

Specs: 16gb Ram ( Corsair Vengeance ddr3 kit);
CPU: intel core i7-4770k;
GPU: Asus Strix RX580 8gb;
Mobo: Maximus VI Extreme;
PSU: Corsair HX 1050;

Some background about the boot drive: Windows is installed on a 1TB Western Digital HDD, and I've had some trouble with it in the past, in the form of the system completely freezing and becoming unresponsive to any command straight after booting. (This went away after I formatted and reinstalled windows on it a couple of years ago.) I am giving context about it because every time the system forces a power off, without fail, I hear a sequence of 3 clicks and 3 faint beeps coming from the boot drive. (Click, beep, click, beep, click, beep) right before the system turns off. After some poking around, it makes the exact same sound whenever I shut down or restart normally. (But it's a single click-beep in that case).

The problem: I upgraded my GPU to the one listed above. It runs smoothly up until a point, BUT the system completely shuts down seemingly when it needs to go through higher work loads (such as benchmarks and demanding games). I checked all connections and later found out that the GPU wasn't plugged in all the way, so I went and shoved it all the way, which solved the problem for exactly a day. It benchmarked, gamed, (you name it) perfectly fine, so I don't think it's a dead GPU. Also, there are no visible artifacts, no weird colors, no dead pixels, etc. It doesn't slow down or lose performance before the shut down happens, or anything like that. It's on default settings, no overclock or special profile applied. (the CPU is under an OC profile from bios, but it's been running like that for years with no issues.)

My troubleshooting: -Right off the bat, I don't think the PSU is strangling it. I even went and unplugged it from the protection adapter and plugged it straight into the wall, with no success.

-I went and updated every single thing Windows was asking for.

-I downloaded DDU and I removed the old NVIDIA drivers and the current AMD driver (just so I can perform a clean driver installation).

-DDU'd the current driver and installed an older version.

-DDU'd all the drivers and just ran it with Microsoft Basic Display Adapter.

-I updated the BIOS. (Although I should mention, the last version available is from 2014.)

-Reset CMOS (multiple times)

-Changed PCI slot.

-Switched RAM sticks between them, ran it with only one RAM stick installed, and then the other.

-Checked device manager for errors; tried running benchmarks in Safe Mode (as suggested on a subreddit)

-Made sure it's not overheating, manually cranked the fans up to 100 and set the AC in my room to max. (And even it it were overheating, it would have probably throttled first instead of shutting down completely, methinks.)(I also installed MSI Afterburner, but I wasn't able to make it display over the benchmark)

Today, while troubleshooting, the system wouldn't post at all, and it gave me the A2 error, with the speaker giving off the 1 long, 4 short beeps error. (Which, after consulting the internet I assumed was a hard drive issue), but after unplugging everything and plugging it back in, it started working again. (I also disabled all the PCI slots from the MoBo, just to make sure it wasn't the GPU).

I am fairly convinced that it's a hardware issue, but I have no idea what else it could be. Any and all help would be greatly appreciated.
 
Jun 29, 2020
10
1
15
I forgot to mention, the MoBo has an extra 4-pin CPU power plug, in addition to the 8-pin one, but there's conflicting information on the internet about it needing to be powered. The manual says it's either one OR the other, but I've read about people asking Asus Tech support and getting conflicting answers, raging from what the manual says to "you need to plug both of them in if you're overclocking". There's also another 6-pin connector above the PCI slots for (quoting from the manual) "when installing multiple display cards to provide sufficient power supply." Should I be looking into those as well?
 
I think you should plug in all the power connectors to the motherboard. Buy a 240GB(or larger) SSD, remove your current HDD, do a fresh install of windows on the new SSD. See if this solves the problem. If not I would suspect a fault with the GPU or motherboard (or PSU)
 
Jun 29, 2020
10
1
15
Have you tried plugging in the additional 4 pin mobo power plug?
I don't have a 4-pin cable, so I tried plugging in part of a PCI 8-pin cable, which made the motherboard completely unresponsive to starting, so I quickly postponed that idea until I could go and get a 4-pin. (I left my PSU accessories and extra cables in my hometown, so I'll have to go get them.)
 
I don't have a 4-pin cable, so I tried plugging in part of a PCI 8-pin cable, which made the motherboard completely unresponsive to starting, so I quickly postponed that idea until I could go and get a 4-pin. (I left my PSU accessories and extra cables in my hometown, so I'll have to go get them.)
You don't have a 4 pin PCI-E power connector on your PSU? How old is this PSU? I think you have fried the motherboard
 
Jun 29, 2020
10
1
15
You don't have a 4 pin PCI-E power connector on your PSU? How old is this PSU? I think you have fried the motherboard
It's a modular PSU, it probably came with the cables in its box, but as I said, it's a long drive away. Also, the motherboard is fine, I did this a couple of days ago and it just refused to start, but after I removed it it worked fine.
 
Jun 29, 2020
10
1
15
So what is the problem with the system right now? It shuts itself down during gaming or benchmarks?(An actual shutdown or does the power just turn off?)
It boots with integrated graphics, and it benchmarks with them (just with bad graphics, since no GPU.) Also, the system just acts like it loses power (it just turns off), but only when the GPU needs to work on more complex loads. I think the GPU is good, since I got it to work just fine yesterday.
 
Jun 29, 2020
10
1
15
Sounds like either a fault with the GPU or PSU. Next step would be to try a different power supply I guess.
That's what I also thought, but I find it very, very weird. The PSU was powering an older GPU with a higher TDP (that required two 8-pin connectors) up until last week, and the GPU was working flawlessly just yesterday. (And it still does, to a certain point.) I'm gonna try and pull the SSD from my laptop tomorrow and boot the Desktop from it, and see how it goes, I guess.
 
Jun 29, 2020
10
1
15
Yes I think it's the graphics card starting to fail, I had a similar experience with a vega 56 I replaced the PSU and the graphics card died shortly afterwards.
I tried with a different RAM kit, I tried booting from my laptop SSD (a Samsung Evo 860 500gb), I plugged a 6-pin into the "extra power for the PCI slots" slot on the motherboard, nothing. I'm starting to near my wit's end.

What I know for sure until now: It is not the RAM, boot drive, or CPU. I'm pretty sure it isn't the MoBo either. The only things left are the PSU and the GPU itself. I have no other PSU that can safely power this system, but I can test the GPU into another system next week.

As we speak, I enabled full bandwidth for the first PCI slot from the BIOS and I'm about to test that too, wish me luck.
 
I tried with a different RAM kit, I tried booting from my laptop SSD (a Samsung Evo 860 500gb), I plugged a 6-pin into the "extra power for the PCI slots" slot on the motherboard, nothing. I'm starting to near my wit's end.

What I know for sure until now: It is not the RAM, boot drive, or CPU. I'm pretty sure it isn't the MoBo either. The only things left are the PSU and the GPU itself. I have no other PSU that can safely power this system, but I can test the GPU into another system next week.

As we speak, I enabled full bandwidth for the first PCI slot from the BIOS and I'm about to test that too, wish me luck.
Good luck
 
Jun 29, 2020
10
1
15
Hi! I'm coming back to close this thread. Long story short, it was overheating because of a combination of stupid software settings and pigeon poop factory TIM. I poked around in the Radeon software, and I noticed something I didn't before. The fan profile was set to "zero RPM" and to never go above 50% fan speed. Needless to say, I wasn't expecting this, and that's why I assumed it wasn't overheating, since I was expecting them to crank up to full throttle if the GPU was starting to overheat. I applied a proper manual fan profile, and noticed that every time it was shutting down, the GPU fans were screaming for mercy.

TL;DR I took it apart, wiped off the (now fossilised) factory TIM, applied some Kryonaut, and it works like a charm.
 
  • Like
Reactions: Flayed