Hi Peter and Ian,
Thanks for your clarifications.
Yes, maybe I missed the point a bit, but unfortunately as you say Peter, without proper test gear and time to experiment, it's going to be a rather hit and miss or is it "empirical" affair. Combining the signals even through a proper splitter/combiner will effectively lose 3db although that loss could be negated with aerials with 3db gain. Wherever the aerials are placed, unless they are optimally spaced, the polar pattern is going to be very hit and miss and not provide the solution. Just thinking about it, every such setup would be unique and dependent on the distance between the aerials which depends on the aircraft.
My suggestion was to put the aerial where it will have a clear view in the most required direction, i.e. directly in front. OK, there may be some obscuration from the airframe, but it would be even on both sides. A colinear would improve the gain in the horizontal plane, so the signal from the obscured direction would hopefully also be improved. With the warning range reduced in the software setup, maybe it would give near equal coverage all around.
In this situation and apart from mounting the aerial externally, I don't think there is a simple solution other than a true diversity setup, but that's another story.
Best regards,
Frank.