This is my first time posting around here, although I've been hitting up TH for years via my Googly searches for the answers to life's electrical problems.
I've finally run into a problem that I can't figure out an answer to without asking for further help though. I'm building a new HTPC system from scratch. Currently, when I hit the power button the system fans spin for about 5 seconds then die off, then spin up for 5 seconds then die off... The cycle repeats.
Now let's see if I can recount what I've done and know so far:
- The system was booting fine into BIOS, for a while.
- I was having problems installing Windows; it gave me a strange error about not finding a necessary component (that seemed to normally happen to people who had accidentally deleted a partition important to the OS.)
- I tried some utilities found on "The Ultimate Bootup Disc". I can't recall exactly which utilities I used as I was in play mode at the time. They consisted of general system information tools for the most part though. One was a system stress test though, which brings me to the next point...
- The wires for the CPU fan were wrapped too tightly and prevented the fan from spinning. (Oops!! )
- After the stress test, I attempted to install Windows using an external DVD drive. Upon doing this, the computer began to do the reset thing, towards the end of POST. From here the time to restart shrank until leveling out at around the 5 second mark. (The CPU fan was still stuck at this point. Clearly could be related...)
- At this point I finally opened up the case and started removing non-essential components, starting with the HDD's and optical drive. The symptoms did not change (and don't from here on out, unless noted).
- I tried each RAM module in every available socket. I also tried with no RAM installed at all.
- I tried booting with the CPU removed.
- I tried removing the motherboard from the case with only power, RAM, CPU and power switch attached.
- I then tried removing the power switch and manually powering the system (via my handy paperclip). This resulted in the system fans spinning as long as I held the connection closed, but dying off as soon as I removed the connection (with no 5 second delay, as happens when using the power switch). With monitor connected here, no video signal is output.
- I tested all the pins on both of the motherboard ATX power connectors. All voltages read correctly. I have yet to figure out a method of being coordinated enough to manually keep the system powered and test the pins at the same time, in order to test under load. I can probably get the missus involved if need be, but she gets nervous about such matters.
I've got no systems beeps or handy LED's to help point me in the right direction here. Just silence and spinning fans. I've looked into getting a motherboard diagnosis card, but all my peripheral slots are PCIe and I have no printer port. This seems to eliminate all options for these cards.
So, my question to you all (after all that), is what else can I try in order to isolate the problem? Does anyone see any signs from my tests that I missed? I can always just return the mobo/cpu/psu for exchanges, but that really takes all the fun out of everything. Also, I'm fairly sure return shipping is not free.
Here are my components:
Intel i3 2105 3.1GHz
Gigabyte GA-Z68MA-D2H-B3 LBA 1155
G.Skill Ripjaws X Series 8GB (2 x 4GB) DDR3
SilverStone Strider Plus ST50F-P 500W ATX Modular PSU
I think I've got everything there, but it's alot so who knows. Any help or advice is greatly appreciated.
Did you by chance remove the CMOS jumper or have it somehow come off the board? That can cause problems like this. If you NEED to keep the pins jumped for it to keep running, it's likely a BAD board. Power buttons/switches are momentary switches thus, they only make contact when pused in, otherwise they are 'closed'. If the board in only running while jumped, it's a problem with the board. It wouldn't be impossible that the chipset is toast or the proc. to be dead. If the board was not set to have a temp. auto-shut down, that would be the problem. A proc. can only 'slow-down' soo much and, if it's not being cooled, it can get fried. If you want to go ahead with a challenge, pick up a CPU socket tester, that will let you know if the board is still good. If it is, then the proc. is probably done for.
Come back with responses/questions to this and we can go from there. Also, maybe try a new PSU. That also could be the problem.
I believe that the CMOS jumper is off normally for this board, and a jumper/connector is applied in order to do the reset. I will give a quick go with the flashlight to make sure there isn't anything fishy/shorted there.
I should also mention that I tried using the reset switch instead of the power switch, just to rule out a bad power switch. I did also jump the CMOS a couple times during my testing to rule out weird settings.
The only thing that confuses me about the switches is why the fans spin for ~5 seconds after I press them, but will die immediately when I "unjump" the power on pins manually. Is there some sort of mechanical delay built into both the reset and power switches? I performed the experiment a few times just to make sure I wasn't imagining the (admittedly slight) difference in behavior.
I'll look into the CPU socket tester tomorrow. Thanks for the insights!
Any tips on where to find an LGA 1155 socket tester? Or a rough idea on their price? In my search it looked like socket testers run in the $15 range, but those were for older socket types and given the places I was seeing the prices listed I wouldn't be shocked if they were completely wrong anyway.
eBay would probably be the best (cheapest) bet on a socket tester. Last I chacked an 1156 was about $20 so an 1155 would probably be close. If you "believe the CMOS jumper is off" on this board, try sticking one on, that should clear it, then give it another try. Also see if there are any other jumpers on the board and read up on them to see if any damage can be done with those.
Maybe try reseating everything and see if that helps. Otherwise, set it up to bench test and see what you find. Make sure there is nothing shorting it out, in any slots/bays and follow the others suggestions.
Hope this solves it! If not, someone will probably have some new, better suggestions but it seems the board is probably bad.