Steve,
I wish I had the room to put up a 2 wave length loop as 'RNL' describes. For me, that would mean using my neighbor's lots and trees, which just isn't going to happen. So, the best I can manage is an 80 meter 1 wave length horizontal loop. Since everybody is always trying to make things just a little bit better, I've done the changing the feed point location thingy. And then made comparisons with an 'R-7' vertical I had at the time. There's a 'catch' in this, that horizontal loop wasn't exactly a 'regular' shape, so my findings aren't exactly dependable for all loops by any means. (It sort of tilted up at one corner, was sort of 'pointy' at one end, and I would be very optimistic if any part of it was close to 25 feet off the ground. The tilted corner was maybe 10 feet (?) higher than the rest of it, sort of.)
When fed from one corner, the loop behaved (heard things) very closely with that 'R-7' vertical. When fed from one side, it was almost the opposite of 'how' the vertical heard things. One would go 'up' with propagation changes, the other would go 'down' at the same time. It may not be a 'true' indication of polarity, but I have to say that if it isn't, it's certainly 'close enough' for me. One of those, "If it walks like a duck..." thingys, you know?
I also modeled the thing using an old program, "NEC4WIN95", which wasn't 'old' at the time, and was 'simple' enough for me to enjoy using. Did some very exacting 'eye-ball' guessing as to height, angles, etc, (including wire insulation color). Couldn't have been 'off' by more than maybe 50%, sort of. That was for 80 meters and several bands higher than that. On 80 meters, the radiation pattern was a very lovely 'ball'. The higher you went in frequency, the more that pattern changed, and the radiation patterns got really weird looking (no idea how to describe them). Somewhere around 14 Mhz it really got directional to the east. It happened to 'beat the snot' out a 3 element tribander at times (MARS and the 'Shield/Storm' time period and phone patches). I seriously doubt if I could duplicate that again, and didn't document it (no time), but it worked so who cares, right?
So, if you sort of discount all the variables, allow for a subjective opinion, I think the feed point can make a difference that corresponds to a change in polarization. Works for me!
- 'Doc