Originally Posted by rabbit73
Yes, it's worth a try. I got 2.5 dB more when I combined two 4221s with a splitter/combiner.
Make sure that the two pieces of coax from the ants to the combiner are exactly the same length, and maybe even from the same batch of coax because the velocity factor can vary. Also, it is very important that the two baluns at the 4221s are connected in phase. If you end up with less signal (or even more confusing, the main lobe divides in two), reverse the 300 ohm connections on ONE of the baluns. This hasn't been mentioned recently, so I thought I better emphasize it.
I ended up with a narrower horiz beamwidth, like you would get with a 4228, which helps with multipath. I also had more gain because the combiner approach is more efficient than the 4228 open wire phasing line system.
I like to use the QUALITY bar on my Apex DT502 to finalize the aim when I have a multipath problem. The angle of acceptance for minimum multpath is much narrower than the beamwidth for max gain when aiming the antenna. This is why aiming an antenna between two azimuths often doesn't work very well, because off-axis aim causes the multipath reflection to be stronger in relation to the direct signal; the reflection can end up at the peak of the main lobe and the direct signal is at a lower gain point on the main lobe. The result is higher BER.
You might still have a problem because optimizing aim for multipath works well for static multipath, but not for wind caused dynamic multipath. But, the additional gain might help with dynamic multipath, as it did for me with dynamic multipath from cars passing in front of my antenna. The reason for that was because less of the limited ability of the error correction system (FEC) was needed for weak signal correction so that more was available for multipath correction.
I have read Ken's notes (about the best site I found for the old 4228 for information all in one place) extensively. Also the part about combining two 4221As for double lobes. Given my radar plot actually I am going to try out of phase on purpose and in phase.
I added my radar plot if you are curious. Note I only care about UHF 16, 28, 31, 36. All others are too far away to be considered consistently watchable or are not really on the air, etc.
I have a 54 degree spread. The station with the worst wind fade is 31, with 36 being second behind it but not nearly as bad. 16, 28 despite 28 being the weakest most distance only wind fades in extreme conditions and 16 almost never breaks up. The two worse, 31 and 36 are also the ones on farthest apart at 54 degrees.
I am not sure how precise the pattern ends up. But it looks like if I went 23 inches out of phase I would target my two weakest stations. I guess the question and proably only known by experimenting is will the middle ones 16, and particular 28 have enough signal to come in. Any thoughts besides just try it?
I have also over the last couple years reading have learned that the point of least mulitpath is much tighter than than most lobes that have gain. Thanks for bringing it back up, nothing is ever too trivial to mention. And I do have a roll of good RG6 where I make identical jumpers, another good point being from the same piece.
I know it's 99% experimenting and 1% theory a lot of the time!