Sign in with
Sign up | Sign in
Your question
Solved

GA-H55M-UD2H MOBO New Build Power Looping

Last response: in Systems
Share
October 19, 2011 7:43:35 PM

I thought I posted this in another thread but can't seem to find it. I apologize if this is a duplicate...

My new build is stuck in the dreaded power loop cycle. I read everything I can find here and a few other places but am stuck on what to do next.

Here is my configuration:

Gigabyte GA-H55M-UD2H Motherboard
Intel Core i3-540 Clarkdale 3.06GHz
G.Skill Ripjaws X Series 8GB (2 x 4GB) 240-pin DDR3 SDRAM DDR3-1600 (PC3 12800)
Seagate ST31000524AS 1TB 6.0GB/s SATA HDD
Samsung Blu-ray SATA optical drive
Rosewill Stallion Series RD450-2-DB 450W ATX V2.2 PSU
nMediaPC HTPC 5000B Micro ATX Case

This was my first build in a long time. I did the usual steps

1. Installed CPU and heat sink/fan on mobo
2. Installed memory modules in DDR3-1 and DDR3-3 slots
3. Mounted PSU in case (I know, should have bread boarded it, but I didn't)
4. Installed the HDD and Optical drives in case
5. Installed backplate
6. Installed mobo into case
7. Wired stuff together -- double, then triple checked the connections to assure proper orientation and seating...including the 1 x 4pin ATX12V (the one this mobo uses), and chassis/CPU fans
8. plugged it in, turned on the PSU switch (re-checked to make sure it was set for 115v) and pressed the case power button...

I had no keyboard, mouse or monitor attached at this point.

Fans whirled, the 4 Phase LED's on the mobo lite, the case power light came on, the HDD and optical diive clicked and whirled...all for about 3 to 5 seconds then it all powered down. Before I had time to pull the plug, the unit tried to start again and the result was the same...about 3 to 5 seconds on, then off for about 5 seconds. It did this until I switched off the PSU. Once off it tried to re-start but "flickered" out...I assume it was just drawing the last bit of power from the PSU capacitors.

Frustrated, here's what I did...

1. Removed all peripherals...HDD, Optical Drive, chassis fans, chassis connections (USB/Card Reader Ports, Audio connections and Reset button). I kept the CPU Fan connected and the power-on switch only. Tried it again...same results.

2. Removed the memory module from the DDR3-3 Slot...same results...

3. Removed the memory module from the DDR3-1 Slot...same results...

4. Inserted the memory module that was in DDR3-3 Slot into DDR3-1 Slot...same results...

5. In this configuration I then got out my trusty voltmeter and measured the power at each of the pins on the 1 x 4pin ATX12v connector and the 1 x 20pin + 4pin Main connector. All voltages read correct (measuring one at a time) during the period when the unit was trying to POST.

6. I then removed the PSU and mobo from the case and set it on a non-conductive surface (the mobo box) and reconnected the 24pin and 4pin power connectors...same results

By the way, the only mounting posts in the case were those that lined up with the mobo mounting holes with the excpetion of two that were .75 inches away from the board. Somewhere along the way, and I can't remember when (there were several) I removed the coin battery and shorted the CLR_CMOS pins.

7. While sitting in the mobo box, I re-installed one of the memory modules in DDR3-1...same results

8. Tried the other memory module in DDR3-1...same reults

9. Tried DDR3-2, then DDR3-4 sockets with each of the memory modules...all 4 tried...same results.

10. Removed the heat sink/fan and CPU...looked at the socket -- looked fine (but my eyes are not great and I had no loop)...I did not try to power on w/o the CPU or heatsink/fan.

I suspect a couple of things.

1. The PSU may not be providing sufficient power even though the voltages are proper during the short time it's trying to start up.

2. The memory modules may not be compatible or are bad.

3. The socket or CPU may have been damaged during the install (though I don't think so)

4. Since I have no other "known good" components to swap in/out, anything in the mix could just be defective.

So, what did I do wrong, am doing wrong, or more important, should try next? I would appreciate help...

Thanks in advance,

-Bill

Best solution

October 19, 2011 9:02:11 PM
Share

Havent read your post after the bit where you described the fault .

Sounds like a short from the mb to earth so check the stand offs are in the right places

If they are and the problem continues then try disconnecting the power and reset switches from the case , and see if it will start normally by touching the tip of a screw driver , momentarily, over the two pins that go to the power button .
The problem might be a faulty switch in the case
October 20, 2011 4:53:51 AM

Problem solved! It ended up being user error and/or poor design. As I mentioned in my original message I may have damaged some pins in the LGA 1156 Socket. Tough to see, but with enough light and a magnifier I was able to see 4 pins that did not look like they were right. I carefully took a needle and "nudged" them to a closer to normal position.

Re-seated the CPU, re-installed the HSF and powered it up...it looked much better. I slowly added cables, then the drives, then put it all in the case.

It may have problems down the road, but for right now it seems very robust! I'm using it now to send this, and at the same time am running 100 passes of IntelBurnTest v2.52...so far 24 complete passes, no failures.

Keep your finger's crossed for me...
Related resources
October 30, 2011 9:20:56 AM

Outlander_04 said:
Havent read your post after the bit where you described the fault .

Sounds like a short from the mb to earth so check the stand offs are in the right places

If they are and the problem continues then try disconnecting the power and reset switches from the case , and see if it will start normally by touching the tip of a screw driver , momentarily, over the two pins that go to the power button .
The problem might be a faulty switch in the case


As I said in my earlier post, it was indeed 4 bent and shorting pins on the 1156 CPU socket. I hope the needle fix is reliable...
October 30, 2011 9:21:20 AM

Best answer selected by bdinnocenzo.
!