Power On, Power Off, Power On

Antonio Java

Distinguished
Mar 20, 2013
12
0
18,510
Alright folks, I have stumper for you all.

I pretty much just finished building the setup you see below. While I still intend to do a few more upgrades in the coming months (more RAM, second graphics card for CrossFire, etc.), what you see is pretty much it for now, and it's running like a dream.

Except for one thing that has me a bit confuzzled.

If the unit is shut off for long periods (that includes shutting off the UPS to which it is connected), the next time I switch it on, the power-up sequence tends to be a little funny.

Immediately after I push the power button, the entire thing comes on. Power light comes on, the fans spin up, everything pretty much comes to life. But after a grand total of two seconds of doing all that (but before POST, I think), the entire thing powers back down, as if I never pushed the power button at all.

This sort of problem isn't new as far as I could read in the forums, but get this: After around 3 seconds of the unit powering itself back down, it turns itself ON all on its own without intervention, proceeding to POST and OS bootup like nothing happened.

So that is: Push power button, computer turns on for 2 seconds, powers down for around 3 seconds, then powers back up again to normal operation.

As noted above, it seems to only happen when the unit has been shut down and completely devoid of power for a long time (note that I turn my UPS off as well while I'm not using the rig, so pretty much no power gets into it while off). If I use the unit for a while, shut it down (including the UPS), and turn it back on again, it starts up just fine.

I have my theories, but they're just that: theories. I want to run this problem by you folks first in case someone far wiser than I knows something that I don't which might be helpful in this scenario without having to taint you with my noob theories. :p

And before you ask, yes, the CPU is overclocked (it normally runs at 3.6GHz, 3.8GHz on turbo, but currently runs at 4.6GHz). It IS quite stable during normal operation and I have never seen it go beyond 50-degrees Centigrade under my normal use (gaming).

Apart from the queer startup sequence, I've had absolutely no hiccups once the computer is already running.

Ideas?


System Specs
CPU: Intel Core i7-3820, 4.6 GHz – 125GHz Strap / 37x Multiplier
CPU Cooling: Antec H2O 620 Liquid Cooling (push-pull exhaust)
Motherboard: Sabertooth X79
Power Supply: Aerocool Strike-X 80+ Gold 1100-Watt Modular Power Supply
RAM: 2x G.Skill Ripjaws-X 8-GB 1600Mhz DDR3 (16-GB total)
Hard Drives: Patriot Pyro SE 120-GB SATA 3 Solid State Drive, 80-GB 2.5" Western Digital Scorpio Black
GPU: PowerColor PCS+ Radeon HD7870 Myst. Edition
Chassis: Cooler Master Storm Trooper Full Tower Chassis
Chassis cooling: 5x 120-mm intake fans, 1x 200-mm exhaust fan
UPS: Apollo Blast 550va
 

Orlean

Distinguished
Nov 28, 2011
340
0
18,860
That Aerocool PSU you are using is most likely the culprit from the description of your problem, it's built by Anyson and they have yet to release a decent PSU to my knowledge. You need to get a better quality PSU also 1100w is overkill for that system a good quality 550w unit will do, 650-700w since you plan on crossfiring later on.
 

Antonio Java

Distinguished
Mar 20, 2013
12
0
18,510


That did cross my mind. Before buying this PSU, I read as much as I could about it and I'm aware of the relationship between Anyson and Aerocool. While I am aware of the criticism of Anyson's products, I had read little regarding the Strike-X line itself. And since I'm one who likes giving underdogs the benefit of the doubt, I went ahead and tried it. So yes, once this queer startup issue came along, the PSU was among my first suspects. However, as I do not have an extra PSU with which to troubleshoot, I'm hoping to get input on other angles of attack. :)

Also, I purposely went with a 1,100-W PSU for efficiency purposes. As many computer users already know, PSUs are inherently inefficient as their loads reach their maximum points, while being at their most efficient at around half load. Since my "complete" setup in the near-future is going to hit the 500-W mark more or less, I decided to give myself not just a lot of headroom, but also hit the most efficient spot possible, just to save on the power bill. :p
 

Antonio Java

Distinguished
Mar 20, 2013
12
0
18,510


Now there's an idea. It did occur to me as well that it might be the mobo, though I dreaded the thought. I'm not absolutely sure, but the problem did kinda start showing up after I UPDATED the stock UEFI with a newer one from the website. I'll try reinstalling it and/or downgrading it and see what comes of it. Thanks!
 

Antonio Java

Distinguished
Mar 20, 2013
12
0
18,510
Alright, update:

I went to the ASUS website intending to download an older version of the UEFI when I saw that they posted a new one just yesterday. So I went ahead and went with the latest BIOS driver (BIOS 3408; prior to that, mine was BIOS 3305, which was the latest until yesterday and the one I was using for most of my build process).

I'm happy to say that after the UEFI was flashed and the system restarted, it restarted NORMALLY. Just to make sure it was consistent though, I turned it off completely (including the UPS), left it alone for a long while, then powered it back up again. The brief shutdown at startup was no longer present. :)

It bears noting though that when the new UEFI was flashed, it reset my CPU settings to default -- no overclocking. Therefore the issue was definitely with the UEFI. Whether or not my overclocking to 4.6GHz had anything to do with the issue, it all boils down to the UEFI, since that is what controls the overclocking in the first place.

I intend to re-overclock the system and will try to note if the problem crops up again. But for now, I'd say this issue has been solved! :D Or narrowed down, at the very least. Glad to know that if the problem does happen again, it won't pose any real problems.

Thanks for the idea and reassurance, mcnumpty23! ^_^

Thank you also, Orlean, for raising the theory about the PSU. I was already contemplating having it changed at the store, but I'm glad I didn't have to. :D