From Barntech,while BTCtalk was down:

Hey guys.

So we believe we have discovered the source of the problem. Basically we are looking at a bad inductor at LP1 which in turn in causing the regulator at U20 to fail and in some cases burn up. The fix is to replace the inductor, and if the regulator chip has failed, replace it also. All the boards out there that are experiencing the ?no green light? symptom have had the regulator fail and will need both components replaced. The good news is that this issue is isolated and doesn?t affect the rest of the boards. Everything else is solid.

We?re really sorry for the inconvenience this causes, and we?re determined to make it right. The bad components weren?t present in our original prototypes, and as you know the boards often mine successfully for hours or days before failing so the problem didn?t show up when the factory tested the pre-production boards or when we tested them before sending them out. In the interest of getting everyone mining quickly we chose to only test the boards for about 15 minutes and in that 15 minute none of them failed.

Now though, with the boards being run for longer, under different conditions, we are seeing it. Sigh. At least we know what it is and how to fix it, so that?s the next step.

First things first, stop running your boards. If they haven?t failed now, there is a pretty good chance they will at some point. Swapping out the LP1 inductor is a reasonably simple task. Swapping out the U20 regulator after it fails is not so simple. The more of these boards we can get to before the regulator goes the better. Its not worth getting a few extra days of hashing to potentially make the boards harder to repair and lose more time in the long run. STOP RUNNING THEM UNTIL THIS PART IS REPLACED.

We are coming up with a couple alternative solutions for replacing these parts.

Replacing just the inductor (green light on)

This is the good news option - you?ve powered off the board while the green light was still coming on.

Option 1 - DIY Inductor Replacement

Once we have selected and tested a suitable replacement inductor we?ll publish it for you all. Those of you we feel confident with a soldering iron, or have a friend who is handy with soldering can replace it yourselves.  We will publish a detailed instructional video on how to do this. This will be the quickest and cheapest way, but please only choose to do this if you are confident that you know what you?re doing. We can?t take responsibility for your boards once you start tinkering with them.

Option 2 - Send the board back

For those of you who don?t feel confident replacing the inductor yourselves, we will get them shipped either back here to us for repair or to a local repairer closer to you. We?re currently looking into overseas repair options so boards won?t necesarily have to come all the way back to Australia, though it may be easier in the long run. I?ll see what i can come up with.

Replacing inductor and regulator (no green light)

If your boards have already lost the regulator chip (no green light) then you will need to send them in so both the inductor and regulator can be replaced, unless you are VERY capable replacing chips on PCB boards, if you believe you are, then you are welcome to have a go at this too, but just know that the regulator chiup has 16 tiny little feet that all need to be soldered.

Source:
http://drillbitsystem.com/forum/index.php?topic=196.45

Hey guys.

Have updated the first page. It looks a little something like this: ... (in short, good news!   )

**** Update 10-12-13****

Hey guys. OK i'm back, had a great weekend away and came back to good news! The team have implemented the fix into 30 boards and they have been running successfully for two days. Looks like we're in business. I am shipping out those 30 boards now and the rest will be ready and tested tomorrow. We have passed the info on to our repairers in the states and the U.K. so we are ready to go into full repair mode.

Can everyone who is keen on doing the fix themselves please write to me so i can make note of who you are and pass on the relevant information.

Thanks everyone for your patience in this. We are truly almost out of it now. Can't wait to get this done so we can move onto future plans!!

I've noticed tensions have been running high and i understand it completely. It certainly has been a frustrating situation. But please don't take it out on the people helping make all this happen. This has been an amazing community effort and has been so wonderful to see everyone band together and solve this situation as a group. Massive thanks to TK1337 and Milkbottlec for stepping up to help out with repairs and going way above and beyond. Massive thanks to Lajz for keeping the forum running smoothly. And to everyone for chucking in your thoughts to help get to the bottom of it all.

As for what actually went wrong here, unfortunately we don't really know. Could have been bad or fake parts, could have been bad workmanship putting the parts on. As for the design, we are almost 100% sure that the chosen parts were right and the design sound, but i guess its possible that they just weren't up for the job. On paper there was plenty of overhead in there. Its kind of impossible to say really. But hey, we've found a way around and the boards now work. Thats the main thing.

Now i imagine things are going to get a bit messy with packages all over the place so please make it easier for me and let me know the status of your stuff. First thing i need to know is everyone doing their own repairs.

Oh and there is a new firmware that is running very smoothly indeed. More on that soon.  

Thanks guys.

Barntech

Source: http://drillbitsystem.com/forum/index.php?topic=206.msg1332#msg1332

Did you add the seed nodes in the sample config found on the first page of this thread?