UHF Horizontally Stacked 2xCM4221HD (Hacked), connected Out-Of-Phase for a NULL on-axis, with 2 or more Beams on either side, described on HDTVPrimer as the "Two-Antenna Trick": http://www.hdtvprimer.com/antennas/ganging.html
Since the original CM4221's are no longer available, the replacement CM4221HD was analyzed.
As shown in the charts for five different Horizontal Stacking Distances, a fairly deep Null can be formed directly on-axis, with Narrow Beams formed on either side of the Null. For some combinations of Frequency and Stacking Distance, an additional pair of usable Beams may or may not be formed at angles even further away. BTW: Horiz. Stacking Distance is between the Centers of each 4-Bay.
Note that for this CM4221HD example (which has a fairly wide Beamwidth), the third pair of Beams has very little usable Gain. Different Antennas will likely exhibit different performance.
The Chart at the bottom plots the Angle between the forward Null and the Adjacent Beams, as well as the Secondary Beams (if usable Gain) versus the Horizontal Stacking Distance.
Using CM4221HD's, (without modifications), the Reflector Rods limit how close the 4-Bays can be mounted. With modifications, the Null-to-Beam Angle could be increased a bit more.
To receive a weak station separated by a very small Angle from a strong local station, the two Antennas need to have a fairly large Stacking Distance...far more than would be feasible mounted on a typical Rotator, which is limited in it's ability to tolerate strong twisting in the wind. If using fixed locations for the two Antennas, their location(s) would need to be carefully adjusted while monitoring how well the strong local station is nulled out. In similar Antenna sidelobe nulling adjustments, I had to add a Variable RF Attenuator to reduce the tuner signals levels closer to threshold to "see" the null.
Per the Azimuthal Pattern Charts, strong signals should be attenuated at least 10-15 dB relative to the weak desired signal. Actual On-Air tests would need to be conducted to verify this prediction and see if even deeper Nulls are possible given "real world" inaccuracies in pointing the Array and matching the Antennas [doublecheck dimensions].
Date(s): 20 Dec 2013. Album by holl_ands. 1 - 76 of 76 Total. 2986 Visits.
enlarge 214KB, 1024x910 1 H-Stk CM4221HD NULL BEAM D = 24-in 3D View
enlarge 264KB, 1024x737 2 H-Stk CM4221HD NULL BEAM D = 24-in Front View [1 large square = 12.5 inches]
enlarge 125KB, 461x1024 3 H-Stk CM4221HD NULL BEAM D = 24-in Top View [1 large square = 12.5 inches]
enlarge 179KB, 1152x714 4 H-Stk CM4221HD NULL BEAM D = 24-in UHF Raw Gain = 11.4 to 15.0 dBi for Azimuth = 71 & 109-deg
enlarge 202KB, 1152x714 5 H-Stk CM4221HD NULL BEAM D = 24-in UHF SWR (300-ohms) Under 4.3 is a Bit High