In most implementations, there is no 'summing', rather the "best" of the N (typically 2) antennas is fed to the baseband.
So there are 2 IF stages, with independent AGC-type detectors, and the one with the stronger signal gets demodulated? > Another approach is to use a power splitter and put two antennas on > the primary connector. This would bring the gain of the omni down by > 3 dB (plus loss of splitter and extra coax/connectors involved) and > then you could put a low-gain antenna on the other port. I'm not sure there is any system gain with this approach. I didn't mean to claim there was. This is simply constructing an antenna pattern which is half of the high-gain pattern and half of a lower-gain pattern, which might or might not have greater utility. My understanding of Fred's problem is that the signal level below the antenna is too low, and that the antenna-centric solution is to fill in the sidelobes a little bit. -- general wireless list, a bawug thing <http://www.bawug.org/> [un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless
