1

There are many resources online regarding the issue when the PC shutsdown a few seconds after pressing the power button. I have a variation of the problem which is reproducible under certain conditions. I would like to hear the community's thoughts on what may be causing it. Found a solution, scroll to end of post, I would still like some more information to rule out a problem with my motherboard.

Scenario

When the computer is powered off for a fair period of time and then powered on maually:

  • Computer powers on fans spin for ~3 seconds (No POST beep)
  • Computer shuts down
  • Computer powers itself on immediately afterwards
  • Computer abpruptly shutsdown (<1 second after it turns itself on) - computer remains off

After the above, pressing the power on manually resumes normal startup, POST beeps once. Computer operates as normal.

Other power on/off combos tried that do not show the above issue:

  • Restarting the PC using either the software reset or manual works fine, Computer boots normally
  • Shutting down the PC either by software/hardware and then immediately powering on works fine.

Things I've tried:

  1. Clearing the CMOS - did not help
  2. RAM is fine (tried bench testing with one module at a time)

Unfortunately I do not have another PSU to test with. Here are my system components:

  1. ASUS VI GENE MB
  2. INTEL i7-4770k CPU
  3. 32GB GSKILL 2133MHz running at 1.5V
  4. SEASONIC Platinum Series 520W PSU 520FL
  5. ASUS GeForce 670GTX

All components are less than a year old so luckily under warrantee.

Guidance appreciated.

EDIT: whenever the issue occurs I'm prompted to enter BIOS by pressing F1 as the overclocking presets are lost. but entering BIOS shows my config as it should be and I exit without making any changes and the PC starts normally.

EDIT: CPUID HWMonitor Output

CPUID Output

EDIT & SOLUTION: I played around with a fresh start on BIOS and slowly enabling one by one the options I had set. Doing this I enabled the ASUS motherboards VRM Spread Spectrum, its description saying it improves system stability... WRONG!! this is the thing that caused my power on problems. I suspect on some occasions the motherboard fails to provide the CPU the voltage it requires and fails the POST. Also I did some reading here -> http://www.overclock.net/t/1401976/the-gigabyte-z87-haswell-overclocking-oc-guide and they say the Haswell processor has inbuilt VRM's so it is not needed to be done by the motherboard. Having ASUS VRM SS disabled in BIOS fixes my power on problems.

I'd like to ask the community to rule out motherboard issue, Does anyone run an Haswell setup i7-4770k with VRM SS enabled in BIOS and it is working for them?

  • Is this a new issue? Had the computer been working previously and then started doing this, or did you just build it? – slow_excellence Jul 7 '14 at 20:59
  • @slow_excellence, Its a new issue when powering on, when I purchased the PC I did notice this when upgrading the BIOS firmware. It would power on then off and I would need to re-configure the bios parameters, such as overclocking, memory profiles etc. I did forget to mention this but every time the issue occurs BIOS prompt me to re-enter BIOS as the Overclocking presets are lost. Will update the question, also shot a short video and now it did something I have not experienced before, it automatically started up after it shutdown normally. – tekiwibird Jul 7 '14 at 21:05
  • @slow_excellence, problem was not there when I initially built the PC. PC was a 24/7 rig so I can't pinpoint when this occurred. PC has only recently been required to reset power on/off as I'm now dual booting between windows and Linux. I've only noticed this now – tekiwibird Jul 7 '14 at 21:19
  • I could be mistaken but I don't believe your power supply can effectively handle the strain of your system. I did a ballpark estimate using the newegg calculator (no clue how accurate that is) but it put your drain at nearly 500W. The PSU can handle this but it puts quite a bit of strain on it and the rest of your system. You typically want at least a 30% gap between your system load and your PSU's rating to work with from my experience. This article may be of some help to you. – slow_excellence Jul 7 '14 at 21:20
  • @slow_excellence - Thanks for the info, I did the power calc's before building with only one graphics card in mind and the selected PSU was adequate and hence why I chose the platinum series because of its performance at load. – tekiwibird Jul 7 '14 at 21:33
2

A key test I use for Power Supllies:

First and easiest, get a multimeter and read the volts out from the power supply. Put the negative lead in any ground on the PSU. Then move the red or positive lead from rail to rail. Do this with the machine on. The most common issue that I see is wild variation on the 3.3v rail. When I can watch a rail vary from something like 3.0 to 3.9v often and quickly it is a bad PSU.

If this passes, do the old bare bones boot and see if a device can be ruled out or as the culprit. If this makes no difference try checking the mains for a brownout.

If all still seems to point to no issues, it is probably the motherboard.

| improve this answer | |
  • Thanks for the info, where would you get access to the voltage rails of the PSU? would this be on the motherboard or poking into the molex connectors at the back of the PSU? – tekiwibird Jul 7 '14 at 21:34
  • Yep, you can just wedge it into the 20/24/20+4 ATX connector (Whichever style you have). The pinout is standard by PSU type, but you don't really need it, you have 5, 3.3 and 12v. As soon as you plug the meter in to each hot you will know what it should be, a 3.26 is going to be a 3.3 and not a 5 for example, and nothing should ever fall dead in the middle at 4.15 for example. If you need it though, just find a pinout online i930.photobucket.com/albums/ad146/ocztech/ATX24-1.jpg – Austin T French Jul 7 '14 at 21:39
  • should I do this test under load or just the PSU on the bench? – tekiwibird Jul 7 '14 at 21:46
  • I generally do it under load. Not necessarily running a game or anything. Once you find a bad PSU this way you will realize it is pretty obvious with a multimeter when that is the issue. – Austin T French Jul 7 '14 at 21:51
  • 1
    Thanks for this idea, I tested the PSU and there is no worrying variation on the 3.3V line, it hit 3.4V as a max. This is a nice guide to test a PSU when you don't have a spare, +1 – tekiwibird Jul 8 '14 at 9:21
0

Problem is BIOS's VRM Spread Spectrum causing boot instabilities, disabling this on the ASUS motherboard fixes the problem and the PC POSTs every time.

Waiting on more information from ASUS as to why it says that the option is recommended and it will improve system stability in BIOS.

I enabled it because it is said to reduce EM emissions as the PC I work on is part of my radio lab.

Here is some info on Spread Spectrum and why you would use it -> http://www.ni.com/white-paper/4154/en/

Why it should be disabled if you are OC'ing -> http://www.techarp.com/showFreeBOG.aspx?lang=0&bogno=266

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.