I don't see any fan-related errors in the first kernel log you posted. The issue lies in the three lines I quoted. The hashboards are receiving power for the initial chip-count test, but the signal isn't returning. In other words, the control board is not sensing even the first ASIC chip in the first domain. The '0 chip' problem is a bit tricky.

it could be a bad chip somewhere or some other issue on the hashboard. If you want to start fixing your own hash boards, I believe you need to obtain at least one working board to compare it against the non-functional ones. The fixture tool won't be of much help in this case; it will still point you to the first chip. Again, this isn't a simple task, so good luck.