New build -- won't boot

[Gustavus]

I have assembled many computers, but this one has me stumped. All new components. Gigabyte GA-H55M-UD2H motherboard, i3-540 CPU, 700watt OCZ PS, 4 GB of GSkill Ripjaws ddr3 RAM and only one card, a Gigabyte GC- N 66T 128 D PCI-Express. Checked and rechecked everything. When it powers on the fans start, the onboard phase LED's lightup and all appears to be going as it should. In a few seconds the fans stop, the LED's go out and it appears to be completely turned off. Then after about three seconds the entire sequence repeats again. In other words, pressing the start button is not required once it goes into this cycle. I can only stop the cycle by turning off the PS.

Any help or suggestions would be appreciated.

And the assembly is on the bench on an insulating pad. I do all assemblies this way and only transfer to a case once everything is checked out and working.

 

[C1]

Peel the system back even more. Minimum RAM, no video. no KB or mouse, no external fans except CPU. Are there any beep codes during POST attempt? We are looking to achieve any kind of sustained operation then work out from there. If this cannot be achieved then in all likelihood, the MB should be swapped out.

 

[Kenmitch]

Being stuck in the reboot loop sucks. Did you try clearing cmos? Your memory is in the brown slots? Did you check the cpu socket for damaged/bent pins before you installed the cpu?

 

[denis280]

Hi so i guest you d'ont have windows install yet.i would start the mobo simple way.take the( GC- N 66T 128 D PCI-Express) out.then restart to see.you should have to clear Cmos if you d'ont have anything install yet.and if that d'ont work start with one stick of ram.

 

[Gustavus]

This motherboard has four memory slots -- two blue and two white. I have tried the memory in both sets, and even tried one stick. Same behavior.

I never get to boot. A few seconds after power is applied it powers off as I explained.

I have tried with the video card out same behavior. Of course it never gets far enough for an audio code. The motherboard is on an insulated pad, so there is no way anything can be shorting on the back.

As I said, all components are new -- used here for the very first time.

One further thing I have noticed. If I turn the power off, at the end of the two second cycle it will attempt to cycle, meaning the LED's lightup and the CPU fan makes an effort to start, but as soon as the charge left in the PS capacitors is gone it really does turn off. So it doesn't act like the turn off is a signal to the PS but something internal to the MB.

Of course I have no way to get to the BIOS.

 

[Kenmitch]

Of course I have no way to get to the BIOS.

Did you try to clear the cmos with the jumper? Or at least make sure it didn't fall off or possibly is in the wrong spot already?

Have you let the board run thru the reboot cycle a couple of times to see what happens. Pretty sure the board is trying to configure itself when it does this. The same way alot of newer boards do this when trying to overclock with improper settings.

It's possible your MB has a bent pin. Did you look at them?

How you starting it up anyways. Cables from case? Jumping the pins? Reason I ask is I had a case that had a sticky reset switch once which wouldn't let me boot up.

Not calling you a dummy or anything but you did connect the 4 or 8 pin cpu power connector? I've seen this overlooked a couple of times before.

 

[Gustavus]

Thanks to all who have responded.

This motherboard, unlike most that have three pins for the clear CMOS, two for parking the jumper and two that clear CMOS, has only two so no jumper is normally there. I have jumpered it to clear CMOS, but the problem is occurring so quickly I doubt CMOS is even involved. At any rate clearing CMOS has no effect.

The system recycles until I kill the power. Comes on for about five seconds, CPU fan turns, LED's lightup etc. goes off, then in two seconds does all of this again. As far as I can tell this is an endless cycle. At least it has always continued until I gave up and turned off the PS.

I examined the pins under a 10x jewelers loupe and they look fine to me.

For bench assembly I have a cable pair with a momentary SDST switch that causes the PS to turn on. Have used it for years. The PS does turn on so the momentary closure is working.

The MB has a 24 pin power connector and a 4pin CPU. Both are plugged in --securely. Just in case I swapped the PS to another computer -- on the bench -- and it works perfectly. And it is an OCZ 700w.

And since it has never gone through a boot cycle, everything is at default.

 

[denis280]

Gigabyte GA-H55M-UD2H (Bring it Back defective)

 

[Kenmitch]

Did you try a diff set of ram? G.Skill makes some nice ram but your MB might not like the default spd settings. Memory compatability is hit and miss. Sometimes you won't even be able to boot. Sometimes you can limp into bios and set it up. Most of the time it's not an issue at all.

The reboot loop of death is just your MB trying to configure itself. About the only thing it can stumble on is memory timings....Unless the MB is dead

Seems like your troubleshooting skills are adequate

 

[Gustavus]

Kenmitch
Unfortunately I do not have any other DDR3 memory. I will check Newegg and Tiger Direct for a cheap set I could use to test if that is the problem.

 

[Kenmitch]

I have jumpered it to clear CMOS, but the problem is occurring so quickly I doubt CMOS is even involved. At any rate clearing CMOS has no effect.

Once the power button is pushed it's CMOS that makes the system work.

The reboot cycle is sometimes a bug in bios....But other times it's just the MB is trying to configure everything and is having trouble doing it.

Kenmitch
Unfortunately I do not have any other DDR3 memory. I will check Newegg and Tiger Direct for a cheap set I could use to test if that is the problem.

I kinda jumped to that conclusion already as it looked like the only thing you didn't eliminate in your trouble shooting.

It would be best to eliminate the ram before you waist a week or two on a MB rma. I've seen many a post with similar issues with multiple MB rma's turn out to be ram compatability in the first place.

Having xtra parts laying around sucks money wise....But they sure do come in handy at times like this

 

[Gustavus]

As luck would have it, the i3 540 has an onboard video so the external video card is not essential to testing. Doesn't matter since the bootup never got that far. I will buy a couple of sticks of DDR3 memory to see if memory incompatibility is the problem. Corsair shows up pretty regularly in Newegg's and Tiger Direct's daily specials.

This was intended to be my first system to break the 4GHz barrier. The best I have stably done -- with P4 CPU's -- is 3.9 GHz. Motherboard, CPU, memory, CPU cooler were all bought because of detailed overclocking reports I had read. Seemed a sure thing.

 

[Gustavus]

Having failed in all of my troubleshooting, I spent several hours searching on net and found quite a few people who had experienced exactly the same failure I am seeing. CPU fan spins for maybe five seconds, then powers down but recycles over and over. Some of the more experienced had swapped memory, replaced the CPU with another one (one case), changed power supplies, tried with no memory to try and get a beep code etc. One person had gotten his board RMA'ed by Newegg only to have the same problem. I do not have another CPU, nor spare memory, but those persons that did still had the problem. Apparently a lot of the H55M-UD2H motherboards have some very strange fault in them. I have sent a detailed report to Gigabyte support, but a common thread to all of the posts i read was that Gigabyte's tech support is virtually non existent. Guess I am out a $100 on this build. Will order a different motherboard.

 

[postmark]

If you only have the CPU in the board, and no memory, do you get a memory beep sequence? Also, you might want to try a much smaller power supply as you could be running into an issue where that power supply might not be able to support a small load (that power supply seems way to big for that system in your first post).

 

[Gustavus]

postmark

Thanks for the reply. It hadn't occurred to me that going to a smaller PS could make a difference. I had a new OCZ500SXS in the lab, so switched to it. Made no difference.

I get no audio codes, no matter what I do. I have tried with memory in different slots, no memory at all, nothing connected etc. No audio codes. LED's light up (all four of them) and the CPU fan turns for about five seconds etc. I have a CoolerMaster 212 so swapped it in instead of the Intel stock cooler -- still the same.

Like I said in my previous post, I found quite a few people who have experienced exactly the same problem with the Gigabyte H55M-UD2H motherboard, and none of those threads ended with a report of success. So far Gigabyte hasn't answered my tech support request -- but that is consistent with what other people were reporting too.

I would think the MB might be incompatible with the i3 540 CPU, but I began by extensive searching on overclocking that CPU and found very detailed posts on success with the pair.

Can't imagine what the problem could be -- but looks like I just lost on this one.

 

[MagnusTheBrewer]

Check the HSF for good coverage with thermal paste. You may be overheating. The CPU gets hot in a hurry. May not be the problem but, it can't hurt to check.

 

[Brainonska511]

I get no audio codes

Just to be sure, you did hook up the case speaker to the motherboard header (probably near where you plug in the power/LED/reset/HDD light... (or it has one built-in), right?

 

[Gustavus]

Thanks for the replies. Appreciate the help.

I have lots of experience spreading a uniform thin layer of Artic Silver TC -- use a single edge razor blade and get a nice thin uniform cover. Besides I have switched back and forth between the Intel stock cooler and the 212 I planned to use once I got it running. Same problem in all trials.

For bench testing I have a momentary SPST switch on the end of a twisted pair that I plug into the pins on the MB connector. Similiarly I have a speaker I cannibalized from some long gone computer, also on the end of a twisted pair, that I plug into the appropriate pins on the MB connector.

Do this sort of thing often so these are just a part of the bench setup. And I don't depend on memory, I check the pinout in the manual every time I set up a new system.

Still no response from Gigabyte tech support (?)