I am also having issues with this board.  I can put any cards in PCIE1 and PCIE2 slots and they all work with rxboost and straps.  I can remove the card from PCIE1 slot and place it into PCIE3 - 6 and I get blue screen.  This board only seems to run two cards.  Maybe a MB bios setting?  Anyway spent hours restoring bios to stock.  What fun.  Hope this gets addressed soon.  I only have one of these boards.  The rest are H81 Pro BTC and they are work great!  Ty

Update- I got it to work by using some of the white PCIE slots - I used PCIE1-1, PCIE1, PCIE2-1,PCIE2 (basically the first 4 slots and had to put a spacer to keep from shorting) then skipped a white one for space and used PCIE4-1 and PCIE5-1  May try all white slots to space things out tomorrow.  Hope this helps someone else.  No more blue screen or lockups.  so strange.
Update #2 - Moved all but the black pciex16 to the white slots for spacing and working - (until my charter internet died!)

I just finished a boxing match with this board myself.  I tried to use a processor from a previous build and that TB 250 board had shorted out in a puff of smoke.  I had thought the CPU was undamaged.  The CPU worked in this board for about 2 weeks.  After that, the symptoms were just as you described.
I ordered a new CPU G4400 and when I installed it;  at first, same problem.  After hitting myself on the head, I noticed the CPU fan plug was not fully inserted and the CPU fan was not even spinning.  So double check the CPU and Fan with another one if possible.  Since you replaced the Mobo already, I would say your CPU may be next on the list to replace.
Also, don't use a DVI to HDMI adapter on the Mobo.  I found that I couldn't get into the bios with the adapter and had to find an old DVI to DVI cable that would work.
After setup, I am able to use a DVI adapter with HDMI dummy plug, but can't see anything until it hits windows loading screen.  This could be a simple setup in the bios fix, but just don't feel like hanging out in the 100 degree mining room messing around with it.  If it works...I don't mess with it!  Good Luck

Based on this website you too late PoW is over :-(  - http://bitconnectpool.co/

I am the mining buddy crypto mentioned... I purchased the program from amazon already.  Just don't have time to play with it yet (had the header pins from a previous project and just need to solder them on).  Currently using one with my xmas tree and google home and I just say "ok google" turn on the dam tree and it repeats it back.  Kids love it!  Something to play with over the holidays for sure.  If i was going to use it full time for a 2400w server psu (240v), i would pull it out of the plastic case and use a separate approved electrical box.  I would think getting 12awg stranded wire to fit would be impossible without making some solid 12 AWG pigtails to splice the cut power cord to.   

I use the breakout board to power the motherboard, not the pico's supply.  I have a special cable with a 6 pin that plugs into the breakout board and the other end is an 8 pin which plugs into the motherboard (8 pin cpu power connector).  There a two wires that branch out from the 8 pin to attach the 12v dc barrel plug from the Pico.  This take the load of the CPU off the Pico and puts in onto the Server supply.  Also if you are running OS from flash drive, the breakout cable of the Pico is not even needed.
You can also just use a 6pin to 8 pin cable to plug from breakout board into the motherboard (which you may be doing already).  Good luck.

Of course I just staked 500 ion for 90 days - oh well.  Will wait for the next pump :-)

As long as people still keep overpaying for graphic cards on the second hand market (fleebay), i don't see this changing anytime soon.  Run the numbers, measure your risk tolerance, purchase wisely and most importantly, have some fun.

With the GPU shortage, all I have on my mining racks are shoes!!  LOL

With current GPU's only producing 80 cents each per day after electric/fees, don't think this thing is in my future.

Server power supplies with breakout boards that are 93% rated and kept under 80 percent load are not that loud.  I replaced all my 90% rated server supplies with the 93% and they are much less noisy.  I use the pico power supplies with a special built cable ( 6 pin to 8 pin ESP & barrel plug) that connects to the pico.  But if your looking to resell the PSU's when done then i prefer the Corsair I series because they monitor power usage (as I use 240v) quite easily. Once I know I have a stable rig, I convert it over to the 1200w server psu's.

Take a look at this video - https://youtu.be/SB7T94PaNDo  Similar riser card concept.
The key is using the PLX Technologies chips that you can see in the video - http://www.mouser.com/ds/2/38/PEX8712_Product_Brief_v1.0_20Oct10-514315.pdf
It can be done, but at what cost and software will need to evolve to support all the cards on one OS, and the multitude of other reasons (like one motherboard dies and lots of hash goes down).
I already use the -wd 0 command in claymore and if I see the hash rate drop in the monitor, then I know a GPU needs some attention (overclocked too high or dead).

Some of you guys are already using a second power supply and should test (on the 12+ rigs) if the second PSU is shut off - does the miner continue hashing away, and just the GPU's attached to the turned off second power supply stop working, or does it take the whole miner down?  If not, then I think eventually we will see 4u rack enclosures with just the GPU's and power supply in each, with long USB 3 cables routed down the rack to the motherboard in the bottom 4u enclosure that has the motherboard and multiple pci-e switching cards.  To big for my budget, but something to envision.  

Edit - The card in the video is selling on Ebay for $125 US FYI - http://www.ebay.com/itm/142479750737

Suggestion for an episode would be to take an older rx 480 rig and do the "cradle to grave" dag fix on a windows 10 machine (6 or 8 card rig).
The learning curve to get the power down using bios mod is not as easy as using the watttool which most of us "non pro's" had used.  Using MSI AB beta 16 is new to a lot of (i think) us.  I already had the settings for the sweet spot of GPU voltage from wattool at (1150mhz), I just needed to figure out which power stage using wattman was close to that value, then used the +/- 100mv swing that MSI AB allowed to fine tune it.  It was a massive pain, but the only way I know how to reduce the power consumption was to bios mod which is not intuitive for GPU voltage pointers.  I had a mixed rig with 4 x rx470's and 2 x rx570's, with Simpleminer and i was at 160Mh/s at 1050w at wall (no dag fix), but with the method above, was able to go back to win 10 and get the hash up to 175Mh/s with 850w at wall.  The key for me is going to be keeping good notes on each rig for MSI AB settings for each card, so if I ever have to rebuilt the OS, it will not take so long to fix. 

Thanks for the data collection - watched it live!  I received my Vega 64 today (powercolor reference) and tried many different combinations.  What I found worked best was to just to use wattman (not many choices) and play around with the combinations like you did.  I ended up tuning until the activity bar at the top of wattman was nice and smooth with no bouncing around (I am so technical).  For me this was 1401 MHz GPU - Memory 945MHz - Fan speed set to maximum (4561 RPM), temp was hovering around 61C and power limit -25%.  This gave me stable 37Mh/s (-dcr 10) at 276W at the wall.  Will see if it's still running in the morning only 25 min. into just letting it hash. 

Tytanick's last post on this forum was August 6th!  Interesting    Anyone hear anything about him taking a 2 week vacation or moving on to something new?

I had similar trouble with the last board i purchased and only had one stick of ram available.  If the computer wouldn't boot, I would just shut down the power supply, pull the ram and reset it in the other socket and it would boot (no idea why).  While testing for best overclock, if the miner crashed then it wouldn't boot until swapping ram slots yet again.   I have other boards running with only one stick, so I am not sure what is going on. I am using the board, and it has been hashing for 6 weeks no problems.  Good luck.

I have the same issue with the mining edition Sapphire (no dvi port and has quad bios).  I had to move these GPU's to a windows 10 machine.

I am having similar difficulties.  Used 1625 straps as well but not much different from what you got at 1750.  I think we are going to need some custom straps, but I can not get the cgi wrapper to work from @eliovp - even using the direct IP address he gave us. 

I purchased and use this one - search ebay - 72A-130BMLEN-02 240V L6-30 12xC13 outlet PDU Zero U NetApp 150-00027+A0 - Tested

Just don't setup a two power supply setup on one rig like I did and plugged in one power supply to one circuit protected side and the other power supply to the other circuit protected side.  fried 4 cards and think its because the ground difference caused higher voltage on 12v rail powering cards.  Not sure, but the other rigs that have two power supplies are fine, but they are plugged into the same side of the PDU above.

I learned a valuable lesson today and wanted to share.  I have a 240v PDU which has two 16 amp breakers built into it (6 plugs on each side).  I was installing a server PSU -1200w and using an standard PSU 850w to run the board and the risers.  What my mistake was that smoked 4 cards was plugging the 240v server psu into one side of the PDU at 240v and the atx psu 850w 204v into the other side of the PDU.  Apparently the two grounds (i think) one is A-side of panel and other is B-side.  I had successfully done this with my first rig, but was lucky enough to have both power supplies(Server & ATX)  plugged into one side of my PDU.  I guess this is lesson learned and if I had done the pico like I  was thinking the power would have all come from the server 1200 watt power supply.  Well I am going to try and RMA them, but they smell bad.  I just picked up 2 more sapphire rx 470's and was planing on building a new rig.  Guess I have room for them know.  Sometimes a lesson is learned the hard way and hopefully it just costs me time and some lost hash. :-(   Don't you just hate when you do stupid stuff?  I lost my test rig last week to a power supply that I had unplugged and it still had charge i guess.  When I went to plug it into an old test motherboard with 2 pci-e slots, it discharged and i heard it.  Put a new supply into the board and nothing.  I guess its just not my week!

They are having issues yet again! :-(  This is sad.  They must be making enough coin to upgrade servers, but it seems they are just milking the cow.  Time to invest in more robust servers or what ever it takes to keep the green rolling in.  They have always made things right by me, but enough is enough!  Time to start looking for alternatives.