This question describes the background context of the problem and then asks a question. I want to learn about the NEC2 simulation program and the method I chose was to take the collection of C++ files and build projects using the Xcode C++ and Swift languages so it will run on a Mac. The C++ version takes a vector of strings that is the card deck, submits that to the C++ files, which then return elevation and azimuth and radius values that are saved to a .csv file. The Swift version does the same thing but has an additional step that is in bridge files called NECBridge.cpp and NECBridge.h. That’s how any Swift program allows C++ to become a seamless part of the Swift program. Both the C++ and Swift/C++ versions return the same eField values that are then graphed using the Swift Foundation library. It all works well. I’ve been able to generate 3D images of various single tower radiators, (typically a ¼ wavelength vertical antenna), a yagi and a log periodic beam. When I construct ¼ wavelength arrays that are in phase but 90 electrical degrees separation, or 180 degrees separation, I get nice patterns that are right-angles to the plane of the two antennas. But when I try to construct an array of ¼ wavelength antennas, of two or more based on known AM radio stations, and based on the FCC information about the antennas, the results have not shown the expected patterns. For example, an array of two that where the two are in close proximity, and set up to be directional off the end of the array in one direction but suppressed in the other direction, returns a near perfect circle. I’ve looked at a number of variations (too many to include here) and none have shown the expected result. AI inquiries have been helpful because they search the literature and return suggested solutions, but they have not returned anything that leads to a solution. I understand that one has to account for mutual coupling, and one suggestion is to increase the power in one of the antennas to overcome the coupling. That hasn’t worked either. Based on what I see of the card decks that seem to work, I believe the programs are working correctly. There’s just this problem with some directional antennas. I wonder if the problem is in the card deck. I wonder what others might think of this. Below is a deck for a two-antenna array. The desired pattern is cardioid, but the result is circular.
GW 1 14 0.0 0.0 0.0 0.0 0.0 46.69 0.0254
GW 2 14 -15.615 27.046 0.0 -15.615 27.046 46.69 0.0254
GE 1
GN 1 0 0 0 0 0 0 0
EX 0 1 1 0 1.0 0.0
EX 0 2 1 0 2.4113 -3.0904
FR 0 1 0 0 1.60 0
RP 0 19 73 0000 0.0 0.0 5.0 5.0 0.0 0.0
EN
This question describes the background context of the problem and then asks a question. I want to learn about the NEC2 simulation program and the method I chose was to take the collection of C++ files and build projects using the Xcode C++ and Swift languages so it will run on a Mac. The C++ version takes a vector of strings that is the card deck, submits that to the C++ files, which then return elevation and azimuth and radius values that are saved to a .csv file. The Swift version does the same thing but has an additional step that is in bridge files called NECBridge.cpp and NECBridge.h. That’s how any Swift program allows C++ to become a seamless part of the Swift program. Both the C++ and Swift/C++ versions return the same eField values that are then graphed using the Swift Foundation library. It all works well. I’ve been able to generate 3D images of various single tower radiators, (typically a ¼ wavelength vertical antenna), a yagi and a log periodic beam. When I construct ¼ wavelength arrays that are in phase but 90 electrical degrees separation, or 180 degrees separation, I get nice patterns that are right-angles to the plane of the two antennas. But when I try to construct an array of ¼ wavelength antennas, of two or more based on known AM radio stations, and based on the FCC information about the antennas, the results have not shown the expected patterns. For example, an array of two that where the two are in close proximity, and set up to be directional off the end of the array in one direction but suppressed in the other direction, returns a near perfect circle. I’ve looked at a number of variations (too many to include here) and none have shown the expected result. AI inquiries have been helpful because they search the literature and return suggested solutions, but they have not returned anything that leads to a solution. I understand that one has to account for mutual coupling, and one suggestion is to increase the power in one of the antennas to overcome the coupling. That hasn’t worked either. Based on what I see of the card decks that seem to work, I believe the programs are working correctly. There’s just this problem with some directional antennas. I wonder if the problem is in the card deck. I wonder what others might think of this. Below is a deck for a two-antenna array. The desired pattern is cardioid, but the result is circular.
GW 1 14 0.0 0.0 0.0 0.0 0.0 46.69 0.0254
GW 2 14 -15.615 27.046 0.0 -15.615 27.046 46.69 0.0254
GE 1
GN 1 0 0 0 0 0 0 0
EX 0 1 1 0 1.0 0.0
EX 0 2 1 0 2.4113 -3.0904
FR 0 1 0 0 1.60 0
RP 0 19 73 0000 0.0 0.0 5.0 5.0 0.0 0.0
EN