Vertically Polarized Yagi Stack

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.

Modeling Results

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

Construction

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.

Antenna File

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

Sensitivity Analysis

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's Antenna

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 0.20.25 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.


December 16, 202188108 MHz