I did some playing around with the Vector in 4NEC2. I haven't done the collinear design that was referenced in the previous two posts yet. I did learn something about how NEC, or at least 4NEC2, works...
Goal 1: To see if I can get NEC to properly show phasing data for the Vector antenna.
Goal 2: To see if I can get NEC's gain figures for the Vector design to be more in line with real world measurements.
Every model below is 1/2 wavelength above real/average ground with the exception of the dipole reference. For each model I have a picture that shows both phase and the vertical radiation pattern. Also included for each model is the NEC file I used in case anyone wishes to experiment themselves, or find faults with my methods or findings.
First the dipole reference. It has been mounted at the same tip height as the other models instead of the same base or feedpoint height. This adjustment was necessary for an accurate measure of unity gain for a dipole (dBd reference).
Code:
CE
GW 1 50 0 0 8.25 0 0 13.75 .0001
GE -1
GN 2 0 0 0 4 0.003
EK
EX 0 1 25 0 1 0 0 'Voltage source
FR 0 0 0 0 27.185 0
EN
Next I show the base vector model. This is the very same model that was used in my last post in this thread.
Code:
CE
GW 1 150 0 0 5.5 0 0 13.75 .0001
GW 3 50 .87 0 8.109 0 0 5.5 .0001
GW 4 50 -.87 0 8.109 0 0 5.5 .0001
GW 5 50 0 -.87 8.109 0 0 5.5 .0001
GW 6 50 0 .87 8.109 0 0 5.5 .0001
GE -1
GN 2 0 0 0 4 0.003
EK
EX 0 1 1 0 1 0 0 'Voltage source
FR 0 0 0 0 27.185 0
EN
As we can see, the bottom 1/4 wavelength of the 3/4 wavelength radiator and the radials that make up the "basket" area are in phase with each other and out of phase with the upper 1/2 wavelength of the 3/4 wavelength radiator.
In the next model there is one and only one difference from the model above, I moved the feedpoint up to a higher location. Instead of being at segment 1 it is now in segment 55 of 150 total segments.
Code:
CE
GW 1 150 0 0 5.5 0 0 13.75 .0001
GW 3 50 .87 0 8.109 0 0 5.5 .0001
GW 4 50 -.87 0 8.109 0 0 5.5 .0001
GW 5 50 0 -.87 8.109 0 0 5.5 .0001
GW 6 50 0 .87 8.109 0 0 5.5 .0001
GE -1
GN 2 0 0 0 4 0.003
EK
EX 0 1 55 0 1 0 0 'Voltage source
FR 0 0 0 0 27.185 0
EN
As we can see that according to 4NEC2 the entire antenna is now in phase, and we see a higher gain as a result. Perhaps this can be part of a work-around to get more accurate results using NEC based programs and the vector design, but I am not satisfied as to the accuracy of this model. My current thought on this change is it needs more study.
In the model below I reset the feedpoint back to the original location and made another change to the model...
Code:
CE
GW 1 150 0 0 5.5 0 0 13.75 .0001
GW 3 50 0 0 5.5 .87 0 8.109 .0001
GW 4 50 0 0 5.5 -.87 0 8.109 .0001
GW 5 50 0 0 5.5 0 -.87 8.109 .0001
GW 6 50 0 0 5.5 0 .87 8.109 .0001
GE -1
GN 2 0 0 0 4 0.003
EK
EX 0 1 1 0 1 0 0 'Voltage source
FR 0 0 0 0 27.185 0
EN
What I did was change the orientation of the wires used for the radials. They are in the exact same physical location but the change allows me to control how 4NEC2 calculates their phasing. Using this method I have found that I can change the phase of any wire on any antenna in 4NEC2 to have whichever phase I want it to have.
This produced the correct phasing as demonstrated by the CST based time domain model that is available on these forums. Unfortunately when looking at the "Vertical plane" gain plot it looks exactly like the plot from the first Vector model above.
This suggests that the phase data has nothing to do with the gain data in NEC, or at least 4NEC2. I believe that this simply shows a limitation of this modeling program as in the real world the radial elements being in or out of phase with the upper radial section would make a huge difference to the radiation pattern from the antenna...
---
I seem to have achieved both goals that I initially set out to achieve, however Goal 1 was not achieved to my satisfaction, and Goal 2 was only apparently achieved as the visual change in phase does not change how 4NEC2 actually handles the radiation pattern or gain of the antenna.
I have shown that phasing data as shown by NEC, or at the very least 4NEC2, can be manipulated to show pretty much anything I wish, thus any model in the future that shows phasing in any NEC based program will be questioned and/or examined.
The DB