Gigabyte EX58A-UD3R (rev:1)

[Dark Mantis]

Have you ever tried the following option: reset the BIOS? You may find how to reset the BIOS in the user manual.

Good point, onemilimeter, but also clear the CMOS first and then load the optimised defaults. It is also detailed in our FAQ section:
http://forum.giga-byte.co.uk/index.php/topic,2286.0.html

[cloneX]

G'day Folks......thanks for the suggestions.
I tried the clear cmos button on the back panel(held in for 10 seconds with the power off) same result! Then removed the cmos battery overnight...same result!
There is no indication that the H50 pump is operating(surely Corsair can add an LED to inform us if the pump is on/off) but the radiator cooling fan is functional
so I have replaced the Corsair H50 liquid cpu cooler with the stock Intel socket 1366 heatsink.....same result!
I have disconnected the system drive.....WD velociraptor, & connected the 2TB storage drive as master on sata0 using different cables......same result!
I have tried all these steps with 3 sticks of ram in triple channel array.....6GB instead of 12GB.
Thanks one milimeter but there is nothing in the manual or on Gigabytes web site that suggests how to accomplish resetting the bios....I always thought that clearing cmos resets the bios so if theres another way could somebody please give me a clue.
Thanks.

[onemilimeter]

Thanks one milimeter but there is nothing in the manual or on Gigabytes web site that suggests how to accomplish resetting the bios....I always thought that clearing cmos resets the bios so if theres another way could somebody please give me a clue.

I think you've done correctly to reset the BIOS.

By the way, do you think there is any chance that the graphic card or the monitor has problem?

[cloneX]

Definately not the monitor.....I have tried it on CRT & LCD ......I will try a different video card.....will try an nvidia GTX275 later today. I know this one works 'cause I'm using it now:).....will also try a different PCIe slot with that one.
Thanks
OK, I have changed the GPU to the GTX275...when I powered on. there was a momentary flash of the case fan LEDs then nothing.
Then I pulled it all down to a bare case & started again...nothing under the mobo causing shorts.
I changed the PSU to a Corsair HX750W Professional Series & left everything disconnected except case & CPU fans.
I still have the phase LEDs on...2 solid yellow then 2 solid green. The NB phase LEDs are solid yellow green yellow green...all as before.
SB voltage LEDs are not on......previously, I had one solid red here.
At what point of the trouble shooting process would you decide that the mobo is faulty?.....I can see nothing else that it could be.

[Dark Mantis]

I would have to concur that the motherboard is indeed the problem here from all the tests that you have carried out. RMA it and get a decent one asap.

[onemilimeter]

Hi cloneX...

You've tried all the possible troubleshooting. RMA may be your next option I think. You may consider to RMA the Rev 1.0 for Rev 2.0. Or, you may request for a newer batch than your current board (serial#1012XXXXXXX), which is an relatively older batch of the UD3R. I bought an UD3R in May 2010 and the batch number is "1015". Then, Gigabyte collected the UD3R and sent another UD3R with batch number "1022" in July 2010.

[bugtraq69]

  Sounds like my first UD3R (a Rev 1).

It ran Windows 7 for about 3 weeks then suddenly zilch. As soon as the Gigabyte splash screen was displayed, the damn thing continually rebooted. RMA'd back to Overclockers UK who confirmed the board was knackered, and GB UK replaced it with a Rev 2, which is working fine.

BTW, once you've stripped it down, have a look at the CPU pads. On mine I had around 20 or so pads that had failed to make contact with the socket pins. I never got confirmation from Gigabyte UK that insufficient pad contact was the root cause of my board failure, but I can't see how a board can function fully if not all 1366 socket pins are making contact, particularly if the pins concerned include ones for VCC/VSS power delivery. Whether there is a general X58 socket quality control issue (similar to that reported on Anandtech for the P55 at http://www.anandtech.com/show/2859) is anyones guess.
 
regards
Peter

[Dark Mantis]

Very interesting Peter, we have had quite a few motherboards with bent pins and the such recently, I think there may well be a problem of quality control. We even had one case, onemilimeter's board was returned from Gigabyte UK Technical Suport with the socket protector forced on the wrong way round. It had done some damage to the protector mainly but also a couple of pins. It does make you wonder what monkeys they are employing in the Technical Department. It really shouldn't be necessary for the consumer to have to check the pins in the socket before installation of the CPU.

[onemilimeter]

BTW, once you've stripped it down, have a look at the CPU pads. On mine I had around 20 or so pads that had failed to make contact with the socket pins. I never got confirmation from Gigabyte UK that insufficient pad contact was the root cause of my board failure, but I can't see how a board can function fully if not all 1366 socket pins are making contact, particularly if the pins concerned include ones for VCC/VSS power delivery. Whether there is a general X58 socket quality control issue (similar to that reported on Anandtech for the P55 at http://www.anandtech.com/show/2859) is anyones guess.
regards
Peter

In my opinion, the X58 socket is not a good solution. Intel is clever in the sense that it pushes the possibility of having "bent pins" (previously the pins are on the CPU) to motherboard manufacturer.

[Strahd]

There's one more thing you could try. I found a motherboard is the easiest to diagnose with a tester card like this http://www.elstonsystems.com/prod/pc_post_diagnostic_card.html
It gives you a HEX number and a manual that comes with the card tells you where the board stopped. You can get one on ebay for like 5 poundish. Very cheap and saves a lot of headaches specially if you build the computers yourself.

[Dark Mantis]

you are right that sometimes it does make a difference to know exactly where the boot process stops. A lot of the higher end boards feature these displays already built in.