Computer stuck in a power cycle

MiniAl

Honorable
Jan 6, 2014
6
0
10,510
Before I post anything else, I assume my complete specs would be helpful:

  • ■My case is a Zalman Z11 Plus,
    ■Motherboard is a MSI Z77-MPower (LGA 1155 socket),
    ■I have an Intel Core i5-3570K CPU (this is not over-clocked in anyway),
    ■My RAM is two 4 GB sticks of Corsair's Vengeance (DDR3 1866 MHz) running in dual channel; again no modifications have been made to voltages,
    ■For graphics I have two NVIDIA GTX 770s with 2GB of dedicated memory from MSI (these are running in SLI),
    ■The system is powered by a Corsair AX860i 860W power supply, and
    ■I have a Creative Sound Blaster Z soundcard.

This is my first ever build which I put together over the weekend of 2013-11-15. Initially there were no issues.

However, soon after the build was complete (around early December) the system randomly turned off and turned back on. The issue disappeared and I thought no more about it.

Just before Christmas the system did it again but this time repeatedly power cycled and would not boot. I stripped the whole thing down and tested it as I put each component in and the system was fine. No issues.

I read somewhere that faulty q-connectors my have cause the problem so I took them out.

I thought the problem was gone until about an hour ago (of which I have been searching). The system now will not boot and is recursively stuck in a phase of powering on for around a second, dropping power completely, and then powering on again before just cutting out.

I've watched the debug LED output and the system always cuts after the value 0x15 is shown. The manual (which is avalble here) states that this represents "Early North Bridge Initialization" (page 1-37 of the manual).

I've Googled around but not been able to find anything. I'm sure after I go back to my machine after conducting this it'll likely power-on but that doesn't detract from the fact that there is an issue.

I'm new to hardware so I'm not sure what could be causing this. Anyone that can help?

UPDATE: After a power lost the system can only be restarted by removing the RAM, swapping the BIOS, and then turning on leaving the side panel off.

UPDATE #2: After my last discovery I have left the side panel fan unplugged. And this time my machine did not power down. Could a molex connector cause a short and shutdown the system? I'm very much a software person and decided to build a machine to learn these things.
 

MiniAl

Honorable
Jan 6, 2014
6
0
10,510

I ran the MSI update software most days and as of last week I didn't have an update so I assume I have the latest.

The PSU has a self-test button that when not hooked up reports its status. It gives a green light for as long as I hold the button (I've done it up to a minute at a time). Could this be incorrect?
 

MiniAl

Honorable
Jan 6, 2014
6
0
10,510


A friend just suggested swapping my BIOS (comepletely forgotten about this feature) and the machine instantly booted. I don't know if this a permant thing though.

I'll see if I can borrow another PSU and test that too.
 

Tradesman1

Legenda in Aeternum
Let's you know not to count on MSI auto - especially if BIOS was out of date, that should be identified even before any drivers, might want to check yourself in the future rather than count on them to say ;) BIOS was my first thought in the original post.
 

MiniAl

Honorable
Jan 6, 2014
6
0
10,510

The issue just happened again, but I swapped over the BIOS again and now everything is fine? I'm grabbing a BIOS update directly from MSI now and seeing if that helps in anyway.
 

MiniAl

Honorable
Jan 6, 2014
6
0
10,510

The BIOS downloaded from MSI and the version stated by my board is identical. Is it possible the "Live Update" software installed it incorrectly?

Also, I've just been running my computer without the side panel on because I'm been messing with it. My computer was on as I put the side on but just as the panel clicked into place the whole thing reset. The case is metal; could it be the cause?

I will try your suggestion about RAM voltage in the morning. I don't know how to do it and I'm kinda of tired; don't want to make anything worse. :'(.
 

MiniAl

Honorable
Jan 6, 2014
6
0
10,510
I apologise for pushing this thread back to the top but I find it annoying when I find a thread that describes my issue and then has no solution.

Anyway, I think I have fixed this issue and I must admit I feel pretty stupid...

My motherboard has another connector that is used to send extra power to the graphics cards. The manual describes this as "JPWR3: ATX 6-pn Power Connector" and it's detailed on page 1-15 of the previously linked manual.

Anyway I found it when I stripped down my build and did a load of research. Turns out it helps with stability of power hungry graphics cards and is basically a huge must when it comes to SLI. My other research about the Northbridge talked about it controlling a high-speed graphics bus. I'm still learning my to me those go hand in hand.

I've hooked up this port and had not issues since. Hopefully my complete stupidity and inability to read all of a manual means I'm the only one that had this issue...