I replaced everything except the case, PSU and fans. The Ryzen 5600X, 16GB of Corsair Vengeance, RTX3060, 980Pro SSD, HDD, Gigabyte B560M Aorus Elite are all new.
After building it I noticed that often the power button wouldn’t power on the system first time. It needed multiple presses or long presses before it would actually start. I replaced the battery and now usually two or three presses were enough. Recently I gave up after 20/30 tries. Yes I did try to do this manually by shorting the mobo pins. Exact same situation. So I redid all the wiring. Being a Corsair HX1200 the cables are bit stiff so now they’re all pretty relaxed. The least bent possible. Didn’t fix.
So I borrowed my neighbor’s RM850. Disconnected the HX1200 PSU cables from the motherboard. (All other cables as well, except CPU fan) Put the RM850 on the table, used it’s cables (not the HX) and connected just the 24pin and CPU. Flipped the switch and popped the breakers. Tried again, PSU again shut off but this time didn’t take the breakers with it.
At this point I thought it might be shorting somehow. I took the motherboard out of the case. Connected only cpu and 24 to the HX1200 shorted the pins. Nothing. Did it again a few times, booted up. When it was booted I hit the clear CMOS. From that point I could get it to power on every time first try. However, after flipping the PSU switch it took me multiple tried again for it to start.
At this point I replaced all the motherboard standoffs and removed the middle one since that one didn’t really fit causing it to be ever so slightly misaligned. Now its perfectly flush.
Connected every single connection, connected GPU and all storage. Again needed multiple tried which is at least an improvement.
When it runs, like it’s doing whilst typing, it runs perfect. I can stress test for hours without it even hitching.
Ready for this? A: It’s your power switch in your PC. Take it out, clean it up. Make sure there’s nothing connecting the terminals. Try it again. I’m not talking about the power switch for your PSU. I’m referring to the switch on the front of your case. I remember working on an old HP PC for a client and was practically tearing my hair out because of this. When I tried hooking up a different switch outside of the case it ran like a charm. It’s such a rare thing and it only happened to me once, but this is very likely your issue. If not then you have a short somewhere because only a short like that could throw your breaker. Check your switch by using a spare switch and check for pinched wires or places where contacts or terminals could be inadvertently contacting the metal in the case. I hope this helps.
for test booting
make sure you have the 24pin and atx leads attached… (no custom cables or ones from another psu)
toggle between the single and multi rail. and try booting from each
if you can get it to display the bios, go into the power settings and disable the c states.
this will stop the psu trying to use low power modes.
save and restart. then shut down and see if you can get it to boot first time.
When you press the power button you can hear the PSU clicking on for a split second or it’s just dead, totally silent?
Cables are not interchangeable so you really risked frying everything in your system doing that!
Clearing the CMOS while the system is booted into an OS is not idea, you should really avoid that to not brick your motherboard.
Are you changing any settings into the BIOS when this happens?
Yeah, I’ve heard of different issues where the CPU wasn’t well-mounted, and it would cause this kind of non-start unless it was ~already~ warm, thermal expansion had to happen to make it fully contact and work. Or the inverse, computers starting fine and becoming unstable due to bad pin contact after warming up.
That said, typically this is attributed more to LGA style processors, but it’s still theoretically plausible here. Wouldn’t say it’s a dead-ringer but I’d at least eliminate that variable.
The power switch on the case seems more plausible to me, but not if the switch has already been disconnected. The HP I mentioned was an old LGA 775. I don’t know if there was a seating issue on that CPU but replacing the case switch solved the problem. Another thing I considered was perhaps if there’s a hairline crack in the board somewhere. Thermal expansion can be a factor in this as well. Just seems odd that it would work perfectly fine for hours without BSOD? This really is a mystery. Time to break out the meter and start looking for shorts.