Quick LED flash, nothing happens

[slip228]

I just put together my first build, but it won't turn on. No fans, nothing. Basically when I hit the power button the power light flashes for that quick second and then nothing else happens. If I try to do it again without turning off and back on the psu then nothing happens at all. I tried re-seating all the components, trying with just one stick of ram, and making sure all the power connectors were in properly. I just found the case speaker hiding under a box and stuck that in but I didn't get any beeps or anything.
this is what I have:

gigabyte P67X-UD3-B3 motherboard
intel i52300 processor
corsair 2x2gb ddr3 ram
evga geforce gtx460 video card
antec neo HE550 watt psu
ocz solid3 60g ssd (for now)
lancool pc-k58 case

I did get the psu and the gpu from my brother who decided to buy a macbook (lame) but they were in perfect working condition in his computer.

I can't figure out if this is a motherboard problem, a cpu problem, or a psu problem. Any help would be greatly appreciated!

[ARchamps]

Could be a grounding issue in the case. Did you use standoffs to mount the motherboard? Try putting together everything outside of the case (carefully) and see it it turns on. Try clearing the CMOS too (check the motherboard manual). Best way is to just take out the battery and leave it for about 20 minutes.

See if that helps and let us know.

[slip228]

Alright, so I tried putting everything together outside of the case and cleared the cmos but still nothing.

**Also forgot to mention that as soon as I turn on the power supply I have power to the mouse and monitor, just nothing else.

[zburns]

Also forgot to mention that as soon as I turn on the power supply I have power to the mouse and monitor, just nothing else.

In doing the following, please only touch the fibreglass 'edges' of any circuit board you handle.

1. Do the mouse and monitor stay on continously when you press the case 'On' button and the power light blinks on then off?

2. Does any led on the mobo light up, for a moment, and then go out (sort of in sync with the power light)?

3. Is the power light you refer too on the front of your case and is it a part of the same sub chassis that contains front panel USB ports, audio ports and IEEE 1394 port ?

4. Did you use and are you still using a 'static wrist strap' as you handle the components?

I would suggest you try the following. (be sure to wear the wrist strap for static and connect it to the case) Run a test with just the motherboard and RAM connected plugged up to the psu. Before running the test just mentioned, take out your video card and the SSD. If the power light still goes off, then remove the RAM and try again.

If you have any questions on any of the above, do a post and ask the questions before doing anything? This is an effort to prove the mobo or the psu are not the problem (leave the cpu alone for now).

[slip228]

I did use an anti-static wrist strap. The monitor and mouse do stay on even after the case led flashes on then off. No mobo lights flash at all. And yes the power light I am referring to is the front panel.

I'll try that and see what happens.

Thanks!

[zburns]

Side question. Are your components back in the case or still 'out of the case'. Problem with out of the case is lack of the mobo being grounded to the case as well as the psu being grounded to the case. I do not know if it makes a difference, but it would be best if when you do the test the components are inside the cabinet. Also, look carefully in your mobo manual and see what leds on the mobo light up (also look at your mobo instructions for the same question) -- what leds light when power turned on! -- something (preferably a led) has to light up when mobo power is on, otherwise we may not know if the mobo is 'powered on' -- led is the easiest way to tell.

You need for 'something' to light up on the mobo when power is on. What you are looking for right now is a defective component -- SSD, Ram stick, video card, etc.. If a component has a full short in it, your psu protective circuit will kick in and shut the shorted 'rail' down -- something like that.

Side comment. If you happen to be in a very cold and snow climate right now, the air is cold and very dry. These are the worse conditions for static so please be careful, even if the conditions are not as cold as I describe.

---

EDIT: Almost immediately after doing the last post regards having to have a mobo led lit up, I realized that is not necessary. The front panel led will serve the same purpose as the mobo led would. So in running the test by taking out the components, the test will be successful if the front panel led stays on when you power up the cabinet. ( I assume you have double checked the wiring from the front panel to the mobo and any place besides the mobo.)

Again, if there is any question or confusion, do a post and ask!
-----------------------

[zburns]

I am doing a separate post regards what I just found out. Your HE550 is fairly old from a design viewpoint. I just saw reviews dating back to 2006 and 2008. Go to this link: http://www.hardwaresecrets.com/fullimage.php?image=4836 This is a picture of the label on the power supply. It has three rails for +12 volts, each rail with a output of 18 amps.

Everything I say here in this para is hypothetical; I am just guessing right now. Your 460 video card may require more amps than 18; without reading the review, I cannot tell if two of the rails can be run in parallel to jack the amps up to 36 amps (2 x 18), the video card plug might be a problem.

I will check the video card right now to the degree that I can find good info. But the three rails for +12v; each rail 18 amps just make me suspicious that the video card needs more than 18 amps and it (video card) is just pulling down your psu.

------------------------

OK, I just went to Evga site. Here is the spec on your card: Requirements: Minimum of a 450 Watt power supply. (Minimum recommended power supply with +12 Volt current rating of 24 Amps.) Two available 6-pin PCI-E power dongles.

So you do need two of the three outputs from the psu just for the video card and the remaining output for the mobo, cpu and other things.

Again, it is the 'old design in terms of years' and how Antec did it compared to today that is bugging me -- so all I can do is look into it tomorrow morning. But it absolutely has to be looked at as the potential cause of the problem.

It may be that you can parallel all three of the so called rails together and get the availability of the max amps from all the outputs. I will take a look in the morning and see what I can find.

[zburns]

New Post, new comments (just after the one above).

As you run the tests I first described, because you are only testing one component at a time, you may not be pulling enough amps from the psu to cause a problem -- so it may look to you as everything is ok.

But when you put it back together again, the original prob comes back.

So let me know what you find with the tests -- but the old design of the psu and its impact or compatibility with your new components and their wiring has to be 'ruled in as the problem' or 'ruled out as the problem' -- NO CHOICE!

EDIT 10:20 pm est: The test I first asked you to run, assumed one component was damaged and had a short circuit or partial short circuit that caused the PSU to shut down. With the new information on the older PSU, the problem is assumed to be a 'demand' for 24 amps, yet 18 amps are only available on each psu rail -- and the assumption now is that only one 18 amp rail is servicing the video card but the video card needs 24 amps; therefore, the particular 18 amp psu 'rail' is being overloaded and the psu short circuit protection is kicking in. We just have to prove it by looking at the specs and wiring.

[slip228]

Thanks for all the comments!

My first guess at to what was happening was either the power supply or gpu just because they are used parts. While my brother used them together successfully in his computer I guess doesn't mean that it will work in mine, right? I was also a little confused because the psu has the 24pin, 8pin and a 4pin connector, but I have nowhere for the 4pin to go and I just assumed it was for 4pin mobos whereas mine is an 8pin. That's okay right? Or am I missing something really important there?

I'm going to test a newer 650w psu in it today and see if it makes any difference, and try everything else you guys suggested and I'll see what happens.

Thanks!

[zburns]

Did you run the individual component tests as I described? If not, I would hold off until you try the 650 watt psu. Is the 650 watt psu a normal off the shelf brand or is OEM brand made for a specific manufacturer such as Apple or HP?


My first guess at to what was happening was either the power supply or gpu just because they are used parts. While my brother used them together successfully in his computer I guess doesn't mean that it will work in mine, right?

As long as your brothers parts are ATX form factor, they theorectically will work in your new build. Because of the age, the Industry ATX Specification may have been slightly altered during the time the parts were made and today. For example, if you ran the test on the individual components and they all passed, ie. front cabinet light stayed for all tests, that would say that the psu was ok, and probably the problem is what I describe with the three 18 amp rails.

You just need to know if the individual components are ok; that being the case, the NE550 is being overloaded.

[slip228]

Alright, I did all the individual tests and everything was okay. So I went ahead and put in a Cooler Master eXtreme power plus 650w psu and didn't hold my breath. I hit the button, noticed that the led stayed, looking in to the case and saw fans spinning and mobo lights on! Then the "everything is okay" beep happened and I will be honest I squealed with excitement. So it was the psu all along, thank god.

Thank you so much for all the help, zburns. It's greatly appreciated!

It's amazing how much you learn when you put together your first computer.

[zburns]

Great news. Congratulations. I am really glad the psu was the problem. What I mean is that your overall assembly process was a success from the gitgo. The psu problem was a 'fluke' -- an age related fluke!

[ARchamps]

Glad to hear!