Vertically polarized directional antennas often have poor azimuth patterns. Side nulls don't invariably occur as for horizontal antennas and the main beam usually is broad. Stacking two antennas side-by-side can reduce beamwidth and side response, but secondary forward lobes may appear and high mutual coupling can obliterate low backlobes. To address these issues, I jointly optimized a stacked pair of eight-element Yagis with AO 9.67.
Each Yagi has two reflectors, a bent driven element, and five directors. The reflectors have unequal but symmetrical lengths and offsets. Boom length is 2.2 meters and boom spacing is 1.8 meters. Red dots mark the 75Ω feedpoints.
Calculated performance is for 28 analysis segments per element halfwave. Forward gain includes mismatch and conductor losses. Subtract 0.3 dB to account for the loss of a ferrite power combiner. F/R is the ratio of forward power to that of the worst backlobe in the rear half-plane.
Frequency Impedance SWR Mismatch Conductor Forward F/R
MHz ohms Loss dB Loss dB Gain dBd dB
87.5 87.3-j25.6 1.42 0.13 0.01 8.61 23.91
88 89.2-j26.1 1.44 0.14 0.01 8.63 25.03
89 91.3-j27.0 1.46 0.15 0.01 8.68 24.55
90 91.7-j27.1 1.46 0.16 0.01 8.74 24.28
91 91.1-j25.9 1.44 0.15 0.01 8.83 24.01
92 89.7-j23.9 1.41 0.13 0.01 8.93 23.90
93 87.9-j20.9 1.35 0.10 0.01 9.04 23.78
94 86.1-j17.1 1.29 0.07 0.01 9.17 23.84
95 84.3-j12.7 1.22 0.04 0.01 9.30 23.91
96 82.9-j7.7 1.15 0.02 0.01 9.44 24.14
97 81.7-j2.4 1.09 0.01 0.01 9.57 24.25
98 80.9+j3.2 1.09 0.01 0.01 9.70 23.97
99 80.3+j8.6 1.14 0.02 0.01 9.82 23.85
100 80.0+j13.6 1.21 0.04 0.02 9.94 23.91
101 79.6+j18.1 1.27 0.06 0.02 10.05 24.16
102 78.8+j22.2 1.34 0.09 0.02 10.15 24.10
103 77.6+j25.9 1.40 0.12 0.02 10.24 23.94
104 75.8+j29.5 1.48 0.16 0.03 10.29 23.95
105 75.4+j33.3 1.55 0.21 0.04 10.28 24.20
106 77.3+j32.3 1.53 0.19 0.06 10.23 24.70
107 68.3+j17.5 1.30 0.07 0.11 10.08 26.29
108 45.0+j9.0 1.70 0.30 0.36 9.35 23.92
Use 10 mm aluminum tubing supported by insulated mounts. Mount the driven element halves so that each tilts 12° toward the rear. The crossboom can be metallic, but for best pattern use a nonconductive mast section near the antenna. Connect the feedpoints to a power combiner using 75Ω cables of equal length. Install a current choke at each feedpoint. Install another immediately after the power combiner and one more 30″ down the feedline to help isolate it from vertical fields. Read these notes before building anything.
Vertically Polarized Yagi Stack Free Space Symmetrix 87.5 90 92 95 98 100 102 104 105 106 107 108 MHz 9 6063-T832 wires, meters ang = 11.7705 ; driven element angle hi = .2508416 ; inner reflector offset ho = .1046344 ; outer reflector offset rp = 0 ; element positions dep = .4881432 d1p = .5708491 d2p = .6464628 d3p = .926394 d4p = 1.538641 d5p = 2.236308 ri = .8934421 ; inner reflector half-length ro = .8565239 ; outer reflector half-length de = .794646 ; driven element half-length d1 = .6825598 ; director half-lengths d2 = .6753074 d3 = .6511202 d4 = .6293857 d5 = .5870633 s = .8878059 ; half of stacking distance shift y s 1 rp ho -ro rp ho ro .01 1 rp -hi -ri rp -hi ri .01 rotate end1 y ang 1 dep 0 0 dep 0 de .01 rotate end1 y -ang 1 dep 0 0 dep 0 -de .01 rotate end 1 d1p 0 -d1 d1p 0 d1 .01 1 d2p 0 -d2 d2p 0 d2 .01 1 d3p 0 -d3 d3p 0 d3 .01 1 d4p 0 -d4 d4p 0 d4 .01 1 d5p 0 -d5 d5p 0 d5 .01 1 source Wire 3, end1
The following table shows the largest performance degradation over the antenna file frequencies in dB when altering a symbol value by Tol.
Symbol Tol Gain F/R
ang 1.0000 0.03 0.17
hi 0.0010 0.00 0.05
ho 0.0010 0.00 0.02
rp 0.0010 0.00 0.02
dep 0.0010 0.03 0.08
d1p 0.0010 0.02 0.17
d2p 0.0010 0.06 0.20
d3p 0.0010 0.01 0.04
d4p 0.0010 0.01 0.06
d5p 0.0010 0.00 0.04
ri 0.0005 0.00 0.07
ro 0.0005 0.00 0.14
de 0.0010 0.01 0.01
d1 0.0005 0.17 0.23
d2 0.0005 0.12 1.15
d3 0.0005 0.07 0.25
d4 0.0005 0.01 0.35
d5 0.0005 0.01 0.09
s 0.0010 0.00 0.07
Konrad Kosmatka built this stack in Plock, Poland. His writeup is here.
Konrad plugged the element ends with Teflon inserts. This method is much less likely to alter element tuning or add loss than using plastic end caps. Styrofoam also should work well.
Very short leads at the feedpoint minimize stray inductance and bimetallic washers prevent corrosion. Ferrite chokes inhibit unwanted shield current.
To combine the two antennas, Konrad used a simple T junction followed by a quarter wavelength of 50Ω cable to match the resulting 37.5Ω to 75Ω. Loss is 0.2–0.25 dB lower than for a ferrite power combiner.
Calculated and measured patterns match well except for the bulge at 40° due to a taller mast 5 meters away.
SWR at the combiner with the antenna wet and dry. Measured SWR was quite close to that calculated.
Heavy icing caused severe performance degradation. This shows SWR as the temperature rose above freezing.
Worst degradation occurred at this partial thaw where slipping ice tubes enclosed the sensitive element ends.
88–108 MHz