this is a part of my support request to KnC. Is there anyone else in the same/similar situation? or possibly with a solution
it's regarding my 4VRM jupiter.
1) I rarely see the cgminer to run longer than 4 or 5 blocks. It restarts and is off for a few seconds. The down time lasts 10+ seconds, which means a loss of hashrate during the down time as well as during the "ramp up period" after restart.
2) One of VRMs on module 6 is offline according to bertmod. Or, to be more exact, the mod says, it's online and ok, but the current is just 0.125A. This loss of power is not compensated on other VRMs - those give out the same current as VRMs on the rest of borads. If I connect just this board to the controller, it produces just about 80GH ... although the bertmod says only 2 cores are disabled on the whole board.
3) High HW error rates. I am experiencing about 10-13% HW error rates. But the interesting thing is - the more boards I have connected to the controller, the higher HW error rate is. When I run just one board I usually get about 2-3%. This seems like an issue with the FPGA, but I have no way to tell for sure.