New PC build - Various problems - Freezing and/or infinite on/off cycle

xALSQ

Reputable
Jun 2, 2015
23
0
4,510
TLDR: My new PC build turns on/off infinitely every 1 second. It manages to turn on successfully if I don't touch it for a few hours, but then after some time (sometimes minutes, but the more I leave it the more it lasts) it freezes and automatically shuts itself off after 5 seconds. Then it repeats the on/off cycle and I have to switch the PSU off to make it stop. Then it won't turn back on for a few hours.

Ok, so here's a [strike]short[/strike] history of how I proceeded with this and how the problem evolved.
All of the parts are brand new (I'll post the part list below). Once I assembled all the parts, I went to turn the PC on. After pressing the on switch, all of the lights would turn on, fans would start spinning, and then immediately after a second or two the PC would shut down. Then it would boot again on it's own. Then off, then on again, etc.. My motherboard is an ASRock Z97 Extreme 4, which has a debug LED on it. For the short second or two that the computer was on, the debug LED would at this time keep showing a code indicating a problem with the SSD drive.
I switched the PSU off (which was also the only way to stop the infinite loop of turning on and off, since neither the motherboard or the case power switched didn't shut it off for good). I unplugged all the cables from the back and so on, then I plugged the SSD into another SATA slot and I also tried a different SATA cable. After doing so, I used the button on my mobo to clear CMOS.
Now when trying to turn the computer on, it was still doing the same thing. It was late at night and I've already spent the whole day building this and trying to figure it out, so I called it a day and decided to try again tomorrow.
The following day, I tried turning the PC on again, and this time it actually booted. I got into the BIOS, checked around for a bit to see if all of the components were recognized, everything looked just fine. The temperatures were 20-25 degrees (both CPU and mobo). I went to the boot options and set it to boot from my USB flash drive so I can install Windows 10.
I restarted my PC and proceeded with installing Windows. I got up to the point where I am supposed to choose which version of Windows to install. Then it froze. 5 seconds after it froze, it shut down on its own. Then when I tried to turn it back on, the thing that was happening at the beginning started happening again. Infinite on/off cycle, debug LED showing the SSD to be the problem. I switched the SSD with the HDD (which works 100% fine) from my old computer, and the same thing was happening. I plugged the SSD back in and tried to turn it on again. This time it was also going on/off, but the debug LED showed a problem with the RAM. After stopping the on/off loop by switching the PSU off, I tried putting the RAM in slots 3 and 4 (previously they were in 2 and 4 (because of my CPU cooler)). This didn't work either. I tried only one stick in slot 2, didn't work. I tried all possible combinations and nothing worked. Each time I tried there was a different number on the debug LED, all of them indicating a problem with the memory. Then I took off my CPU cooler and turned it the other way around so that i can try out slot 1 for RAM. I tried out all possible configurations and it still didn't work. Also, when moving the CPU cooler I of course removed the "old" thermal paste and put new paste on. The paste was spread fine and that couldn't have been the problem. So now I know it's not the SSD and it isn't the paste or the cooler. I went to have lunch (so 30ish minutes later) and when I got back I turned the PC on and it worked. I went to the BIOS first and it was still recognizing each component. I booted from USB and tried installing Windows again. I got up to 30% or something, then it froze again and shut off on its own, repeating the on/off loop when I turned it on again. The freezing made me thing it might be my GPU, so I removed my dedicated GPU and plugged my monitor into the on board Intel graphics. It still kept going on/off. I also tried my old GPU, nothing.
The morning after that, I tried plugging my PSU and my monitor directly into the wall sockets (they were in an extension cable). The PC booted, I managed to install Windows completely. I downloaded some programs, installed Chrome, downloaded the drivers for my GPU. Everything seemed to finally work. It worked for a few hours. Then it froze again, in the middle of clicking through the desktop. And it shut down 5 seconds later. On/off loop again. I tried the integrated graphics again, and this time it worked. I got to my desktop and after about 1 minute it froze again, and shut down.
While all of this was happening, the debug LED would always show something different. It showed the SSD once, the memory was shown a lot of times, the GPU a few times. So it seemed to be completely random. After the most recent freezing, it again entered the on/off cycle when I tried turning it on. So I turned the PSU off and called it a day once again.
Tomorrow (after leaving it be for a few hours), I can turn it on again. It was on for a good few hours again. Then it casually froze (while browsing Chrome). And shut down, not wanting to turn back on. On/off. The motherboard stopped showing anything on the debug LED. I removed the SSD and the GPU and plugged the monitor into my integrated graphics. I used only 1 stick of RAM. I tried seeing if it will boot to BIOS like this. It didn't. I haven't touched it again for a few hours, and then when I turned it on it managed to go into the BIOS. Debug LED was showing only an SSD problem (since I didn't have one). I exited the BIOS on my own and shut the PC down. Installed the SSD again. Went into Windows normally, debug LED not showing anything. Then it froze again, shut down.

So my conclusions are:
It's not the SSD, GPU, thermal paste.
It's probably not the RAM, CPU or PSU?
Maybe the motherboard is faulty?

This was just an extreme source of frustration and sadness, since I was working so hard this summer to buy and build this, and then experiencing this has just been really devastating..
Any help would be really appreciated. My best guess is that it's the motherboard that has a defect, but here's my part list:

ASRock Z97 Extreme 4
Intel i5 4690k
Gigabyte GTX 980 G1 Gaming
HyperX Fury CL10 1600MHz
Samsung 850 EVO 250GB
beQuiet Dark Rock 3
EVGA G2 750W
NZXT s340 Case

EDIT: I've forgot to mention I have 4 case fans and LED lights which I plugged off at one point and the PC still wouldn't work, so it can't be that (2x Corsair 120 AF Quiet Edition, 2x NZXT (came with case)).

EDIT2: I just made sure that it's not my PSU. I've hooked up my new EVGA PSU to my old rig and it's running fine.
 

legion_of_cheese

Reputable
Sep 15, 2014
80
0
4,660
Heh, just as I wanted to write if you changed the PSU you edited the post. Well yeah, from your conclusions the culprit is the motherboard. But why is the bigger question. If it is something physical [a circuit or an electronic component] then you better RMA it. But if it is the BIOS's fault have you tried updating or downdating the bios?
 

xALSQ

Reputable
Jun 2, 2015
23
0
4,510


I haven't tried updating the BIOS, but the fact that at first it wouldn't even go into BIOS made me think it can't be that. I was planning on returning it, I just wanted to be 100% sure that it indeed is the motherboard, since I bought the parts in Germany online (and I live in Croatia), so just shipping it back is gonna cost me a bit, and they won't reimburse me if the motherboard isn't the problem. And they also don't have their own testing service or a technician for returns, they need to ship it to the manufacturer and so on..
But yeah then, I guess there really is nothing else that could be causing this.
Thanks for the help, now for the painful RMA process..