It's a pain to adjust the lengths of a multi-element beam (a Yagi-Uda) when the antenna is up on the tower, so many times, the strategy is to do the adjusting by measuring the resonant frequency of the elements with the antenna sitting on saw horses or a ladder, where it is reachable. In fact, this might even be the manufacturer's recommended procedure. It's well known that putting an antenna in close proximity to a dielectric will change the resonant frequencies, and hopefully, the manufacturer has accounted for this.
The question arises, though, of the magnitude of the errors and the sensitivity to ground parameters (if the manufacturer's facility were in the Mojave desert, and you're assembling the antenna standing in a swamp, you KNOW there's going to be differences). A strategy that has been recommended by the author and others is to do the adjustments with the antenna pointing straight up (at least some of the elements are farther away). But, just how big a difference does it make?
A parametric investigation was done using LB Cebik's 3 element 20m yagi model, as provided with 4nec2. A copy of the NEC deck is at the end of this page. The ground was set to the Sommerfield Norton using the 13 and 0.005 for dielectric constant and conductivity of the soil, respectively. The excitation was put in the middle of each of the 3 elements in turn, and the frequency was swept in 10 kHz steps to determine the resonant point for that feedpoint (when X was closest to zero). The results are summarized below. By changing the rotation and height parameters, it's pretty easy to "move" the antenna around.
Here's the data (Z rounded to roughly 1%)
|70ft||6 ft (5,0.001, poor)||6ft (13,0.005, avg)|
|Driven (33')||14.18 MHz (24.9-j0.03)||14.07 MHz (31.8+j0.15)||14.09 MHz (33.0-j0.02)|
|Reflector (34.56')||13.61 MHz (43.4-j0.07)||13.61 MHz (45.7+j0.20)||13.60 MHz (40.8+j0.01)|
|Director (31.05')||14.64 MHz (46.8+j0.27)||14.69 MHz (49.0-j0.19)||14.77 MHz (41.3-j0.04)|
Note that the resonant frequency of the reflector doesn't change all that much (although the Z changes by 10%), but the driven and director resonant frequencies change quite a bit. And, with poor soil, the change is less. However, do this over your freshly watered lawn, and it might change even more.
Tipping the antenna vertically, with the reflector 6 feet off the ground (13,0.005) gives the following data:
|Driven||14.18 MHz (25.8-j0.18)|
|Reflector||13.61 MHz (36.9+j0.03)|
|Director||14.61 MHz (43.0-j0.17)|
Comparing to the 70ft data, the driven and reflector elements are resonant at the same frequency, and the director is slightly lower.
Interestingly, if one breaks the driven element (by putting a 1k resistor at the feedpoint), the resonance of the director is now at ), representing leaving the feedpoint open, we get some different data. The director has almost exactly the same resonance (as expected, because the interaction with the driven element is gone, and it's almost 30 feet in the air). The reflector resonance is lower than in the complete antenna (probably because the coupling with the driven element normally brings it higher)
|Element (Driven broken with 1k resistor at feedpoint)||Vertical, w reflector 6ft high||70ft horizontal|
|Director||15.23 MHz (81.3+j0.20)||15.23 MHz (75.7+j0.04)|
|Reflector||13.45 MHz (44.1+j0.22)||13.5 MHz (61.02+j0.18)|
Adjusting a multi-element beam with the antenna sitting 6 feet off the ground and looking at the resonant frequencies may not be a good plan. The soil properties will change the resonance. If you have to adjust by resonant frequency (as opposed to by mechanical dimensions), then tipping the antenna so it points straight up is a good idea.
The resonance of the other elements changes when the driven element is open or shorted, so if you're adjusting to match a manufacturer's recommendation, make sure the test condition is the same.
CM 3-element Yagi, 20 meters, 1" Al CM L. B. Cebik, W4RNL CM Modified by Jim Lux, W6RMK, for investigation of height above ground effect on Z CE SY len=16.501 '16.617 SY ht=70 '6 or 6+11.272 GW 1 41 -len -.817 0 len -.817 0 .0415 GW 2 41 -17.28 -11.272 0 17.28 -11.272 0 .0415 GW 3 41 -15.525 11.272 0 15.525 11.272 0 .0415 'GM 0 0 90 0 0 0 0 0 'rotate GM 0 0 0 0 0 0 0 ht GS 0 0 0.3048 GE 1 GN 2 0 0 0 13 0.005 LD 0 1 21 21 1000 'break driven element EK EX 0 3 21 0 1 0 FR 0 0 0 0 14.175 0 EN