I observed something today on a new B5c link. The signal is good and strong when I run it on U-NII-3 at full power. It’s a 7-mile hop and needs all the speed it can get, so I set it at 2x40 and get MCS9 and a -60 dBm signal. Seeing such a good path, I did a little experiment. I moved one of the two channels down to DFS (U-NII-2). The power had to fall from +23 to +3 (27 dB antenna). And the signal on that carrier got much weaker. I expected its MCS to fall below 9 too. But both fell. After a little while it got up to MCS5 on both.
I then looked at a couple of older links, and noticed that they had one MCS across the board, even with one DFS and one non-DFS. But those were shorter urban hops, and the higher noise on non-DFS channels made DFS work as well. Now I’m building a big network (over 70 B5cs) in the sticks and there isn’t much noise, just trees and mountains to contend with. But on a short hop DFS would be nice to get one channel out of the way of unsync’d sectors.
So is this a hardware limitation or what? Can the MCS of each frequency be what that frequency’s SNR supports, or do both frequencies have to throttle down to the lower of the two? That sort of limits the flexibility of 2-channel operation. If it’s a software limitation it would be nice to see it fixed.