I have to throw this out there. I must have a strange case in my house, but I get the best RF results with the transceiver in my basement. This never made sense to me, but has been proven by trying many different placements. (I also notice that the center isn't the best in my case but I attribute that to a specific set of heater ducts (Metal, of course) that run through the floor.
Has anyone else run into this situation? The higher I put the RF unit, the better it works from outside, but the less reliable it is from inside. With it in the Attic (Call this 3rd floor...) I have to play with pointing the Palmpad to get a signal through. On the second floor it works 50% of the time, first try. Strangely, in the same room there seem to be spots that are good and others that are terrible. With the unit in the basement (Actually, in a plug on a basement "ceiling" bean, which would be 6 inches below the floor.) I have no dead spots anywhere in the house (I even tried the Palmpad while in the attic for a test.) and it usually works 100% of the time.
From outside, however, it seems to follow what I consider normal. The higher, the better. In the attic, I can get 75% or so from inside my garage, which is detached and about 60 feet away. With the RF in basement, forget most of outside.
I'm thinking of trying a "Smart" repeater in the attic to get the best of both worlds, but I dislike doing things that make no logical sense to me, which this doesn't. I've honestly been reading many posts just to see if anyone had the same crazy results, and have never heard (read?) of anyone saying this. Seeing as this was a transmitter placement thread, sounded like a good place to try.
(Note: I have done a basic "Mapping" of all metal, wire, coax, cat5, etc. and don't think that's what is causing this upside down result, but I think these items do cause some of the "Dead spots" with certain placements. Just FYI.)