| KM2B |
Rating:      |
2020-12-22 | |
| Profoundly capable antenna design/optimization package |
Time Owned: more than 12 months. |
Arie Voors, creator of 4NEC2 wrapper
4NEC2 was created by Arie Voors in his long pursuit (early-90's to 2008) to put an extremely powerful wrapper around the public domain NEC2 Fortran code made public by the Lawrence Livermore National Lab in the USA. Mr. Voors donation of his nearly 20 year investment in developing 4NEC2 is one of the great gifts of all time to the Amateur Radio community, and, to the academic RF community.
The current wrapper is C++ based with incredible capability of which I will document part of that capability here.
I) Far Field Pattern
There are three ways to view the far field antenna pattern upon creating a model and running it.
a) using the 2D default view that pops up after the run. This enables rotating through user specified slices around 360 degrees and pulling detailed gain data as a function of angle and azimuth angle.
b) using the 3D, color coded view available from the drop down 3D selection after running a model. In this view one can show the antenna, and, its full 3D, color coded in terms of dBi gain, visual view. This view enables detailed 3D analysis of antenna pattern, in the context of the antenna placement, for understanding peaks, nulls, and all aspects of complex far field propagation. I used this view to understand my 80m OCF on 40m, 20m, 17m, 12m, and 10m. Very complex patterns appear at all bands < 80m on this antenna.
c) 3D wire view that is selectable on the antenna view that shows as default. You have to select View, then, select far field pattern. A 3D wire view sliced at the angle decrement specified in the simulation setup shows. IF you toggle over to the 2D view of dbi Gain, then, rotate through the 3D view, a colored traced on the slice of the 2D view follows the 3D view. You can rotate through each far field gain slice at each angle around the antenna.
The above three separate graphical capabilities for far field pattern analysis are benchmark analysis tools. Nothing in EZNEC comes even close to the combined viewing ability in 4NEC2.
II) Symbolic Representation Enabling Mathematics and Optimization and automation of optimization.
Symbolic antenna construction within the context of the original FORTRAN card construct for NEC2. This ability to build a model by defining the software equivalent of a variable by using the SY declaration, coupled with Mr. Voors integration of two optimization approaches, combined, offer an incredibly powerful ability to find optimal outcomes with software automation, not manual tuning. All capability is all built into 4NEC2 package. EZNEC only recently added a partly capable Excel add on to do optimization. Before that addition, optimization only by single run manual trial.
4NEC2 enables the antenna model designer to design the antenna in a symbolic (variable) format. This enables the designer to specify, say, for an OCF, both of the lengths of the segments, and the angle from vertical, as variables. In complex antenna models, the variable name can be re-used in multiple locations. Then, if that variable (length for example) needs changing, it can be changed in one place in the model much like a software program. No need to manually change the variable multiple times due to "hardcoding".
Two aspects of 4NEC2 symbolic modeling capability greatly exceed EZNEC capability:
a) The ability to build in, and, use, the original FORTRAN text "card" commands that are documented for 4NEC2 (EZNEC shows nothing of those original commands). Why is this important? The original NEC2 documentation can be read, and, leveraged (unlike EZNEC which does NOT elevate this capability).
b) The ability to use variables (this capability does not exist in either original NEC2 or EZNEC). Arie Voors introduced this power into the 4NEC2 wrapper.
III) Non-Linear Optimization of Selected Symbolic Variables
Having built a model in symbolic format, in text, in the original NEC2 Fortran like data file, the resultant model can then be optimized using the built-in non-linear optimizer or a separate genetic evolver that enables more broadly checking different options. This optimization code was built into 4NEC2 C++ wrapper and does not exist in NEC2 or EZNEC. As an Example: I built my OCF model with insulated wire (also enabled by 4NEC2). I used symbolic format. Then, I selected the variable name for the two lengths of the wire antenna (short and long) and ran the optimizer with SWR at 100 (minimize SWR). The program ran and returned the optimal lengths for minimum SWR. It took 24 iterations in about 14 seconds on my new Win 10 platform with an SSD NvME hard drive. No manual change and rerun needed.
Regarding constraints for optimization: One can optimize for gain, f/b ratio, SWR, etc. Or balance all of these in weighted format. Really, commercial optimization programs are sometimes not this flexible with constraints.
EZNEC now offers a for purchase, Excel add on, to support optimization. However, the base EZNEC platform offers no optimization and all optimization for EZNEC must be done by laborious manual changes. 24 iterations would be an hour of laborious typing and re-running in EZNEC.
IV) Frequency Sweep
The ability to take a model and scan a frequency range, for the model system built, and return SWR, Impedance, gain, and F/B ratio as a plot as a function of frequency and also write out the data array for further analysis in platforms like Matlab. For example, for my OCF, I scan from 3.0 MHz to 29.0 MHz for SWR to test the impact of center to end relative heights on impedance. In this manner I can tell if I need a 4:1 balun, or 5:1 or 6:1. Saves me wasting time physically testing poor configurations. The default plotting routine is fully featured, and, the data array can be saved.
Optionally, the user can couple with GNU plot by installing that and the plotting routine is amazing. I have used this capability to compare and plot my actual antenna with the model antenna and you can see results on my qrz.com page.
V) Main Pane Information Return
The data information returns, after running the model, on the main page, antenna efficiency and radiated power relative to fed power. So, if you define a LC network in the model, the code will compute loss. The code also computes both Antenna efficiency (loss from network matching) and overall Radiation efficiency capturing ground loss information for real ground.
VI) Transmission Line Modeling
The ability to build and use the original NEC2 transmission line “cards”. This enables one to add real transmission line loss for matching SWR of a real system to simulation.
VII) Automated Network Matching for Multiple Network Types
A built in, fully automated, network matching algorithm with user selections for L network, T network and one other type, and series or parallel, matching sections at any point in then antenna or transmission line is built in. This coupled with the loss computation enable the approximate design of a matching network for said model system. Example: The designer builds an elevated vertical that is 5/8 wavelength for 10m. Then, by placing the RF point at the bottom of the vertical, and, using the network matching automation, after building the far field pattern, one can select the network matching type, click run, and, all parameters are returned in both L, C, R and also, in complex coordinates if you want to permanently add them to the model afterwards.
VIII) Ability to Simulate Transmission Line Impedance
The ability to specify characteristic impedance for SWR computation. Example, for my OCF, if I desire to know the 50 ohm impedance response, then, I specify a 200 ohm characteristic impedance for computing SWR and assume that I have a 4:1 transformer in place. This enables modeling input transformers.
IX) 4NEC2 does read .EZ files so that all of the antenna models in the ARRL Antenna Book, which, oddly, exclusively use EZNEC modeling instead of Arie Voors free modeling format, can be read and converted into symbol format for optimization with 4NEC2.
The ability to read .ez file types, from EZNEC, and use them and modify them is very useful given ARRL’s support for EZNEC. Hence, all of the files in the ARRL Antenna Book, generated by the ARRL supported EZNEC, are available for read, and, modification into 4NEC2. And, once they are translated, you can modify the card format, and, turn them into variable based models for optimization.
X) Completely Free of Monetary Cost to the User
4NEC2 and ALL OF THE ABOVE capability IS FREE not including the time necessary to teach oneself how to use the platform.
Summary and Thank you to Mr. Arie Voors:
The 4NEC2 is extremely powerful, and, for free, and exceeds the capability of the EZNEC platform for NEC2, even after all these years since Arie made it available for free. Using 4NEC2 is also, once learned, MUCH easier to use than EZNEC to learn about the antenna system more rapidly due to symbolic representation coupled with optimization.
Note, since both 4NEC2 and EZNEC are built around NEC2, neither platform can model buried radial systems. However, one can model slightly elevated radial systems.
However, both platforms can be used with NEC4 which does allow buried radial systems.
|
|
| N2UJN |
Rating:      |
2016-07-03 | |
| Best Software for Ham Antenna Simulation |
Time Owned: more than 12 months. |
4NEC2 was created by Arie Voors in his long pursuit (early-90's to 2008) to put a wrapper around the public domain NEC2 Fortran code made public by the Lawrence Livermore National Lab in the USA.
The current wrapper is C++ based with incredible capability of which I will document part of that capability here.
I) There are three ways to view the far field antenna pattern upon creating a model and running it.
a) using the 2D default view that pops up after the run. This enables rotating through user specified slices around 360 degrees and pulling detailed gain data as a function of angle and azimuth angle.
b) using the 3D, color coded view available from the drop down 3D selection after running a model. In this view one can show the antenna, and, its full 3D, color coded in terms of dBi gain, visual view. This view enables detailed 3D analysis of antenna pattern, in the context of the antenna placement, for understanding peaks, nulls, and all aspects of complex far field propagation. I used this view to understand my 80m OCF on 40m, 20m, 17m, 12m, and 10m. Very complex patterns appear at all bands < 80m on this antenna.
c) 3D wire view that is selectable on the antenna view that shows as default. You have to select View, then, select far field pattern. A 3D wire view sliced at the angle decrement specified in the simulation setup shows. IF you toggle over to the 2D view, then, rotate through the 2D view, a colored traced on the slice of the 2D view follows the 3D view. AMAZING.
The above three separate graphical capabilities for far field pattern analysis are benchmark analysis tools and NOTHING in EZNEC comes even close to the combined viewing ability in 4NEC2. NOTHING.
II) Symbolic antenna construction within the context of the original FORTRAN card construct for NEC2.
4NEC2 enables the model designer to design the antenna in a symbolic (variable) format. This enables the designer to specify, say, for an OCF, both of the lengths of the segments as variables. In complex antenna models, the variable name can be re-used in multiple locations. Then, if that variable (length for example) needs changing, it can be changed in one place in the model much like a software program. No need to manually change the variable multiple times due to "hardcoding".
Two aspects of 4NEC2 symbolic modeling capability greatly exceed EZNEC capability:
a) The ability to build in, and, use, the original FORTRAN text "card" commands that are documented for 4NEC2 (EZNEC shows nothing of those original commands). Why is this important? The original NEC2 documentation can be read, and, leveraged (unlike EZNEC which does NOT elevate this capability).
b) The ability to use variables (this capability does not exist in either original NEC2 or EZNEC). Arie Voors introduced this power into the 4NEC2 wrapper.
III) Having built a model in symbolic format, in text, in the original NEC2 Fortran like data file, the resultant model can then be optimized using the built in optimizer/genetic evolution code. This optimization code was built into 4NEC2 C++ wrapper and does not exist in NEC2 or EZNEC: Example: I built my OCF model with insulated wire (also enabled by 4NEC2). I used symbolic format. Then, I selected the two lengths of the wire antenna (short and long) and ran the optimizer with SWR at 100 (minimize SWR). The program ran and returned the optimal lengths for minimum SWR. It took 24 iterations in about 4 seconds.
One can also optimize for gain, of f/b ratio, etc. Or balance all of these in weighted format. Amazing.
NOTHING like optimization exists for EZNEC. All optimization for EZNEC must be done by laborious manual changes. 24 iterations would have been an hour of typing and re-running in EZNEC.
IV) The ability to take a model and scan a frequency range, for the model system built, and return SWR, Impedance, gain, and F/B ratio as a plot as a function of frequency and also write out the data array for further analysis in platforms like Matlab. For example, for my OCF, I scan from 3.0 MHz to 29.0 MHz for SWR to test the impact of center to end relative heights on impedance. In this manner I can tell if I need a 4:1 balun, or 5:1 or 6:1. Saves me wasting time physically testing poor configurations. The default plotting routine is fully featured, and, the data array can be saved.
Optionally, the user can couple with GNU plot by installing that and the plotting routine is amazing. I have used this capability to compare and plot my actual antenna with the model antenna and you can see results on my qrz.com page.
V) The data information returns, after running the model, on the main page, antenna efficiency and radiated power relative to fed power. So, if you define a LC network in the model, the code will compute loss.
VI) The ability to build and use the original NEC2 transmission line “cards”. This enables one to add real transmission line loss for matching SWR of a real system to simulation.
VII) A built in, fully automated, network matching algorithm with user selections for L network, T network and one other type, and series or parallel, matching sections at any point in then antenna or transmission line. This coupled with the loss computation enable the approximate design of a matching network for said model system.
VIII) The ability to specify characteristic impedance for SWR computation. Example, for my OCF, if I desire to know the 50 ohm impedance response, then, I specify a 200 ohm characteristic impedance for computing SWR and assume that I have a 4:1 transformer in place. This enables modeling input transformers.
IX) The ability to read .ez file types, from EZNEC, and use them and modify them. Hence, all of the files in the ARRL Antenna Book, generated by the ARRL supported EZNEC, are available for read, and, modification into 4NEC2. And, once they are translated, you can modify the card format, and, turn them into variable based models for optimization.
X) 4NEC2 and ALL OF THE ABOVE capability IS FREE not including the time necessary to teach oneself how to use the platform.
Summary:
The 4NEC2 is so much more powerful, and, for free, than the expensive EZNEC platform for NEC2. Using 4NEC2 is also, once learned, is MUCH easier to use than EZNEC to learn about the antenna system more rapidly due to symbolic representation coupled with optimization.
Note, since both 4NEC2 and EZNEC are built around NEC2, neither platform can model buried radial systems. However, one can model slightly elevated radial systems.
|
|
| KK4OBI |
Rating:      |
2016-01-29 | |
| A phenominally good product |
Time Owned: more than 12 months. |
After many attempts at building antennas and not understanding what was going on, a wise ham suggested 4NEC2. Using it is like a drug, you can't quit. Idea after idea. Problem after problem. The enormous scope and deep potential of this program gradually sinks in. In particular is the Evolve/Optimize function that automatically finds the conditions you are searching for.
Putzing with it has filled many binders and files. So much so that much of the results from these studies, plus 30+ related antenna models are now available on the "Bent Dipoles" website.
www.qsl.net/kk4obi
The incredible author, Arie Voors, has now turned to other endeavors due to the developments in 64-bit and multi-processors. His work in 4NEC2 will non-the-less continue to serve hams indefinitely.
|
|
| N4ES |
Rating:      |
2015-04-13 | |
| What a powerful tool! |
Time Owned: 0 to 3 months. |
| I finally discovered 4NEC2, thanks to the eHam reviews. It is so good, it should be called 4NEC8! It's the most fun you can have without laughing! Having an optimizer makes it possible to design multiple element arrays that would be extremely time consuming or impossible otherwise. I spent a couple of days installing the software and reading the tutorials. As a user of EZNEC 3, 4, and 5 PRO, I had never even seen raw NEC code before. EZNEC is very easy to use and gives great results, especially for simple verticals and dipoles. However, scripting the 4NEC2 code in Notepad wasn't bad at all. Today, I designed and optimized a 3-element Moxon-like wire beam to hang vertically from the trees. After thousands of optimizer iterations, It looks pretty good. I can't wait to build it. Thanks Arie! |
|
| N1DAY |
Rating:      |
2014-04-26 | |
| Truly educational and useful |
Time Owned: 3 to 6 months. |
| Be prepared to take some time learning how to take advantage of everything this modeling program has to offer. Once you have ascended the learning curve, this piece of software can be fully appreciated for how amazing it is. The optimizer is just plain awesome and has saved me so much time and generates significant improvements on all my designs. In addition to being a really useful tool, it is fun to watch the 3-D display morph as the optimizer steps through the design possibilities. |
|
| N5EG |
Rating:      |
2013-03-14 | |
| Awesome |
Time Owned: more than 12 months. |
The 4NEC2 software is now hosted at http://www.qsl.net/4nec2/
4NEC2 is an extremely powerful antenna modeling program based on NEC2 code. 4NEC2 is my tool of choice for antenna modeling.
The main attraction of 4NEC2 is the graphical entry of antenna models. I find it far faster for initial entry than X-Y-Z entry of coordinates. A spreadsheet also allows initial entry or adjustments after initial entry. It has tools to auto generate grids, radials, and some quite complex geometries, provides scaling, translation, and rotation of antennas and subsets, replications of antenna subsets, and many other aids far too numerous to list.
It provides graphical display of antenna currents and phase, and allows selecting segments for numerical current display in complex format. This is useful for computing mutual impedance for example.
Additionally, there is a hand-optimized multithreaded version of the NEC2 engine for 4NEC2 called NEC2/MP that speeds up the execution by 2x to 10x depending on what kind of SSE instruction set your computer has. It's available from http://users.otenet.gr/~jmsp/
it just replaces the various nec2dxs???.exe files and leaves the rest of the 4NEC2 front-end alone.
The optimizer is very powerful, but it takes a little experimentation to become familiar. By proper definition of dimension variables, it is even possible to optimize while enforcing symmetry for example.
Highly recommended. Expect a little learning curve; once you get the hang of it the interface is actually quite intuitive. The 3d display engine is hands-down superior to anything else I've seen.
Hats off to Arie Voors for this wonderful tool!
-- Tom, N5EG
|
|
| VE3PBE |
Rating:      |
2012-01-05 | |
| Tough Job made Easy! |
Time Owned: 6 to 12 months. |
Designing an antenna system with tough geomertrical and environmental constraints is tough. Even if y ou know what you are doing.
Arie Voors has done a superb job in providing different ways to approach the incremental design process. Any modelling software has some learning curve. By following the instructions and examples provided it took only a few days.
My problem was to get a reasonable, low-cost 40-10m attic antenna in a crowded urban setting with metal surfaces everywhere. 4NEC2 enabled me to find the key elements impacting the overall performance from the 3D display of elements and current and phase. I was able to just fit a beam into a 32'x 33' attic with a pyramidal roof and truss members everywhere.
With the optimization functions I could find the best SWR and launch angles for the geometries.
I ran hundreds of combinations of antenna designs, shapes, feed points, etc.
Finally, one design came through as the best compromise.
I built it, installed it and using the AIM 4170C analyzer was able to verify the design. The SWR spectral response was very close to the 4NEC2 theoretical plots. By getting the ratio between threory and real, I could compensate for things like roof attenuation, velocity propogartion of insulated wire, complex surronding objects, etc.
But thanks to this product I now have a very good antenna that took me less time and money than any other antenna project to build, install and tune.
The power of computers and great sofware indeed saves time and money!
Thanks Arie Voors! |
|
| W9AC |
Rating:      |
2011-11-11 | |
| Impressive |
Time Owned: more than 12 months. |
At first, antenna modeling software is difficult to understand, primarily because of the Cartesian coordinate method used.
Two recommendations: (1) read the 4-part QST series on antenna modeling by L.B. Cebik; and (2) start with the pre-configured "Back Yard Dipole" example included in the 4Nec2 package -- and carefully study the input page that shows wires in X, Y, and Z coordinates, segments, and source. You can read a thousand pages about segmenting and just one input chart speaks a thousand words. Once you start with just a basic dipole, moving on to base-fed verticals, Yagis, quads and other geometrically complex antennas becomes much easier to model.
4Nec2 offers an unbelievable suite of modeling tools and features: Super-high segment count, built-in detailed Smith Chart, Geometry designer, and ability to view many complex output formats is on the short list.
|
|
| IZ3ATV |
Rating:      |
2010-11-02 | |
| Great, and for free. Hats off to Arie Voors! |
Time Owned: more than 12 months. |
4nec2 had quickly become one of my favourite software.
It's a valuable tool for most antenna's facts understanding. Its 3D feature really make you see how an antenna works.
Most of its precious features have already been said by the preceeding reviewers, what I can truly say is that 4nec2 had turned out particularly useful to me when I needed to evaluate the near-field value in my shack as produced by a quite close end-fed inverted-L antenna running right over my head.
All the words in the world don't suffice to say: thank You so much Arie Voors for your wonderful work. |
|
| K8RWH |
Rating:      |
2010-03-14 | |
| Superb + Cebik, ARRL |
Time Owned: 6 to 12 months. |
| Arie Voors deserves a zillion THANKS from all of us. This free software does everything and more the commercial versions do! I got LB Cebik's Antenna Modeling book and have begun to learn this software in an incredibly exciting voyage thru these dreary winter months! My spring antenna building plans are pretty well set, thanks to Arie and LB(sk). Every ham now has the opportunity to understand antennas like never before!! |
|