A Brief Overview of the Zeland Software Suite

by MD0MDI
A Brief Overview of the Zeland Software Suite - Figure 8

A Brief Overview of the Zeland Software Suite

The antenna modeling software packages with which most casual antenna designers and analysts are familiar rely on either NEC or MININEC. However, as we move into the realm of “wireless,” we often outrun the ability of either type of core to do the jobs that we need.

For example:

1. We can quickly surpass the frequency for which NEC is adequate to our task. It is desirable to maintain a segment length-to-wire diameter ratio of 1:1 minimum and higher, if possible. We also need to maintain about 10 segments per wavelength and more if possible. A 1 GHz, a wavelength is 0.3 m (300 mm) long. The segments for a minimally segmented element would be 0.03m (30 mm) long, barely enough for a fat 1″ (25.4 mm) element–assuming that we have no odd geometries that require more segments per wavelength. Alternatively, a helix depends upon many short, straight wires for simulation, and hence the need for segments much shorter than 0.1 wavelength. The allowable wire diameter shrinks accordingly. Figure 1 shows a pair of shapes that–at upper UHF and high frequencies–often surpass NEC limitations. And these are but two of many.

A Brief Overview of the Zeland Software Suite - Figure 1
A Brief Overview of the Zeland Software Suite - Figure 1

2. NEC and MININEC calculations rest upon a thin wire foundation, a thin ROUND wire. Above about 400 MHz, it begins to make sense to fabricate antennas on circuit boards, using elements that are flat strips rather than round wires. Up to a point, we can model such antennas within NEC or MININEC using equivalent round wires. However, the more complex the structure, the less adequate the substitutes are. Parasitic relationships begin to depend upon how much surface area is facing the next element. NEC and MININEC cannot easily adjust this dimension and still sustain the overall cross-section area of the conductor.

3. As we move still higher into the UHF region, it becomes practical to integrate into an antenna design any required transmission lines to terminals, impedance transformation networks, phasing lines, and even a variety of filters. All can be fabricated using a single substrate or a coherent collection of substrates. Many facets of the design may make use of the substrate properties to create lumped components or their equivalents. We may further integrate these design elements with 3-dimensional shapes, such as horns. But each element on the original substrate also has 3-dimensions. The wire-grid potentials of NEC often fall short of being able to deal with this degree of integration.

These are only some of the reasons designers working in the wireless realm have turned in part to a different order of software to simulate their complex structures that include antennas. Some of the desirable features of such software would include the following:

1. The ability to work directly with surface areas and 3-dimensional objects, accounting for their conductive and non-conductive components;
2. Compatibility with the .DXF format of file transfer from such drawing programs as AutoCAD and TurboCAD;
3. Integration with or ready transfer between radiation analysis modules and circuit design modules, with SPICE capabilities associated with the latter;
4. Integration with or ready transfer between radiation analysis modules and filter design modules. . .

I leave this list without an end, because an engineer’s wish list and a programmer’s challenge are equally unending.

Software packages aimed at meeting these needs do exist. They are large, complex, and often require special training–not to mention lots of trial and error experience–to use effectively. One such program that I recently had the chance to review is Zeland Software’s suite. The suite is usually called IE3D, but actually consists of a collection of modules, each with specific specialties, but many overlapping functions:

IE3D is historically the first of the stand-alone modules offered by Zeland (1993). Now in release 8, it is also the best documented of the modules, with well over an inch-thick manual devoted to it. The name “IE3D” captures two of the main features of the modules. “IE” is short for an integral equation, method of moment, full-wave electromagnetic simulator. Unlike NEC programs that require very large models using wire-grid techniques, the “3D” portion of the IE3D title indicates that the program uses a 3-dimensional non-uniform triangular and rectangular mixed meshing scheme to capture complex shapes. The difference between a wire grid shape and the IE3D technique appears in Figure 2.

A Brief Overview of the Zeland Software Suite - Figure 2
A Brief Overview of the Zeland Software Suite - Figure 2

The program is geared toward microwave and millimeter-wave applications with more than antennas (both normal and patch) in mind. It is also aimed at uses in integrated circuit and filter design, EM scattering analysis, and other applications involving coaxial and digital circuits. The module solves for current distribution, network parameters, near fields, and radiation patterns. The program has the ability to model arbitrary shapes in multiple layers within a complex dielectric environment. Figure 3 shows a sample of a 3-dimensional shape set that includes metal thickness as well as area, terminals, and connection jumpers.

A Brief Overview of the Zeland Software Suite - Figure 8
A Brief Overview of the Zeland Software Suite - Figure 8

IE3D is subdivided into several sub-programs. MGRID is the starting point, since it comprises the layout editor for constructing a geometry for analysis. Every shape to be simulated is composed of a set of polygons, and each polygon is described by a set of vertices (or corners). Vertex matching is a condition of ensuring the electrical connection between two adjacent polygons.

Much of the IE3D manual is devoted to mastering the art of object or circuit construction. Because of its power, MGRID requires a considerable learning curve for full mastery. Indeed, Zeland offers seminars for mastering basic techniques, and it is likely that these seminars are the best and quickest introduction to the art of geometry fabrication within the program. The review period for the program was not sufficiently long for me to fully acquire a decent mastery of the techniques. Hence, all illustrative graphics are downloads from Zeland. However, I was able to replicate a number of useful shapes to get a feel for what is possible. And what is possible is considerable for work in the 1-25 GHz range and beyond.

IE3D proper is the simulation core for numerical analysis of the model. (Note the growing naming problem that emerges due to the evolution of the Zeland offerings: IE3D names the numerical core, the program section–as distinct from Fidelity–and the overall package of programs.) A selection of benchmarks accompanies the package to confirm the accuracy of the IE3D application. Most of the bench marks are half a decade to a full decade old, with references to machines as slow as 66 MHz. It is likely that these benchmarks are due for updates, with reference to CPUs in the 1 GHz to 2 GHz category. As well, the entire manual suffers somewhat from evolutionary pains. Many sections are written in what appears to be “second-language” English. Running the manual through the hands of a competent technical editor with experience in writing instructions would likely smooth the user’s reading and possibly even ease entry into the program.

For nodal circuit simulation and parameter display, the program contains a module called MODUA. There is a post-processor called CURVIEW for the display and animation of current and field distribution. Equally important to antenna and EMC interests are PATTERNVIEW and FIELD, post processor modules for examining radiation patterns and calculating near fields, respectively. Radiation patterns are routinely displayed in 3-dimensional terms, with color indications for field strength, as illustrated in Figure 4. However, the vivid color graphics is backed up by an impressive array of rectangular and polar plots. Equally impressive are the number of ways in which the user can compare output data from models. The latest “powerpack” release of IE3D also contains a genetic optimizer capable of handling a considerable number of variables and optimizing goals.

A Brief Overview of the Zeland Software Suite - Figure 4
A Brief Overview of the Zeland Software Suite - Figure 4

The user in effect must custom design his purchase or licensing for this program according to a careful analysis of present and future needs. The pricing differential between the “limited” package and the “powerpack” version of IE3D is better than 2:1. The limited package restricts the user to 1,000 unknowns of less. The intermediate package removes this limitation, but does not include 3 features available only in the powerpack edition: the genetic optimizer, magnetic current modeling for slot structures, and the higher efficiency iterative matrix solver.

Although billed a complementary to IE3D, Zeland almost makes the case for replacing IE3D with Fidelity, the finite-difference time-domain (FDTD) program. Introduced in late 1997, Fidelity contains within its overall collection modules that are comparable to those in the IE3D package: a “Workshop” for the geometry construction phase of the work, Fidelity proper as the FDTD simulation core, MODUA as a schematic editor and parameter display unit, a current animator, and the same pattern-view module as used in IE3D.

Time-domain analysis offers several advantages over the method of moments analysis used in IE3D or other finite element methods. FDTD simulations produce wide-band results over a considerable frequency range, while typical method-of-moment (MOM) techniques require multiple runs called frequency sweeps. As well, FDTD memory requirements are only proportional to the number of elements in the construction, whereas MOM techniques require the square of the number of construct elements for their matrices. In addition, FDTD simulators are easier to implement mathematically for both electric and magnetic fields and handle more easily complex dielectric structures.
Indeed, the fidelity workshop or geometry creation module deals directly in 3-dimensional objects, rather than expanding on 2-dimensional objects, as is done in IE3D. At present, Fidelity uses non-uniform rectangular meshing as its basis, but a future release will introduce conformal meshing as the basic object-creation technique. Although we may represent an object with only a few cells in each of 2 cross-sectional dimensions, FDTD results begin to converge with analytical results (and the MOM results) only as we significantly increase the cell density. In a sample problem, an 8-fold increase in the number of cells across the outer diameter of a coaxial structure reduced a 15% error down to about 3% (relative to both analytic and to IE3D results). Convergence of results requires careful attention to both pre- and post-processing activities.

Nonetheless, FDTD techniques, as used in the Fidelity package, can produce impressive results with objects consisting of highly complex combinations of conductive and dielectric materials. Figure 5 shows the E-field and the H-field that result from a plastic covered object–in this case, a typical hand-held transceiver.

A Brief Overview of the Zeland Software Suite - Figure 5
A Brief Overview of the Zeland Software Suite - Figure 5
As Fidelity has rapidly evolved, it has improved its automation and ability to handle objects that are not immediately amenable to FDTD cell treatment. The current release (3.0) includes provisions for gap ports, thin-wire models, helix models, and wire-mesh objects. In addition, the program includes a growing number of objects in its library to ease the process of creating complex structures.

Perhaps one of the most complex objects–and yet one of the most important in an age of cell phones–is the human head. As Figure 6 shows, Fidelity contains a rotatable model of the human head that can form the basis for any number of more complex assemblages. The figure adds a transceiver with a projecting antenna as a case in point. Just how finely one might detail the interior layers of the head to correspond with the variable density and material that actually compose the human head is not completely clear at first sight. However, it is a foundation upon which to build truly significant models of the interaction of RF energy with the most critical part of human anatomy.

A Brief Overview of the Zeland Software Suite - Figure 6
A Brief Overview of the Zeland Software Suite - Figure 6
Interestingly, the Fidelity portion of the manual is not structured in quite the same way as the IE3D portion. For example, the menu structure, given an entire early chapter in IE3D, is consigned in Fidelity to an appendix. For applications in the GHz and upward range, FDTD techniques will eventually supplant MOM techniques. It is likely that with some future release of the overall program package, the thinner Fidelity manual will grow to match or even surpass the size and fine detail built into the IE3D manual.
In 1999, Zeland introduced MDSPICE, a mixed (time and frequency) domain version of SPICE, the standard for virtually all circuit simulation work. MODUA is the module included with both IE3D and Fidelity to extract lumped-component equivalent circuits from the s-parameters produced by model analysis. In the IE3D portion of the manual (Chapter 9), Zeland discusses the relative advantages and disadvantages of lumped-component analysis and makes a case for its own extension of SPICE to include mixed-frequency domain and time-domain simulations. Perhaps the key element to MDSPICE is that it does not require conversion of the s-parameters into RLC-equivalents, thus avoiding accuracy problems associated with the conversion process. At the same time, MDSPICE can make use of standard SPICE netlists.
At present, the MDSPICE portion of the Zeland manual is relatively thin compared to the Fidelity section–which–as we have noted–is thin compared to the IE3D segment. (Neither the Fidelity nor the MDSPICE segments of the manual have indices.) Independent SPICE familiarity is advisable for the user.

Not included in the manual are two interesting modules offered by Zeland. One is of their own making: COCAFIL: a suite to design and analyze waveguide-coupled cavity filters so widely used in all upper UHF/SHF applications. Figure 7 shows a sample screen for the design of a specific filter.

A Brief Overview of the Zeland Software Suite - Figure 7
A Brief Overview of the Zeland Software Suite - Figure 7
Zeland also offers MTRAN, a third-party format converter, for which one end is .DXF. .DXF is the standard CAD format used in such programs as AutoCAD and TurboCAD for file transport to other media. Zeland is currently working on making the other end more directly compatible with the IE3D file format.

Still further into the Zeland offerings is a transmission line analysis and synthesis module called LineGauge. The basic program is a free download from the Zeland site (http://www.zeland.com). A “professional” version is also available, but is not free. Figure 8 shows a sample LineGauge screen for a coaxial cable. However, the program–as indicated by the entries at the right of the figure–covers a wide range of transmission line types.

A Brief Overview of the Zeland Software Suite - Figure 8
A Brief Overview of the Zeland Software Suite - Figure 8
I had no difficulties installing the Fidelity suite of programs or its master control module for switching among the elements into an XP machine at 1.7 GHz. Operation of the program requires a key (a physical key, called a “dongle”). However, in the brief period that I had available for evaluation, I was only able to sample the capabilities. The key must be returned to Zeland following evaluation.
Sample run times appear to be quite reasonable and related to the complexity of the geometry of the model being run. Screens are relatively intuitive once one becomes acclimated to what is required to build a geometric structure for analysis. Indeed, model construction requires the most attention and time to ensure that what will emerge is sensible. For those used only to modeling in NEC, where one specifies a collection of wires as the geometry of the model, the transition to IE3D and/or Fidelity will require patience and close attention to all details. The ability to specify multiple layers to a structure, each having different properties, means that considerable pre-planning must occur before the first step of geometry specification within these programs. In short, effective modeling for the GHz and up range, where the structures do not necessarily resemble those of wire and tubular element antennas, demands a change in the way we think about structures.
The Zeland offerings are multi-faceted, competent, and full of creative as well as analytical opportunities. But there are two prices to pay for these very positive features. One price is in dollars. A full non-network suite might cost between $40,000 and $50,000, with the smallest IE3D module priced at about $10,000 and the basic Fidelity unit costing about $20,000. Needless to say, the investment is suited perhaps only to a corporate enterprise.
The second investment is the time and energy required by the learning curve toward mastery of the program’s possibilities. Although Zeland literature suggests that some structures can be created in only a few minutes, the beginning user should not expect such results instantly. With patience, the manual can be used to develop considerable facility with the software. However, attending the introductory workshops offered by Zeland is likely to be far more efficient. Such guided experience will likely avoid the thousand missteps that attended my own introductory meanderings through the programs.
It is likely that all of the programs in the Zeland suite will continue on their evolutionary paths towards every greater capabilities and efficiencies. At what point Fidelity may supplant IE3D as the foundational element in the suite is uncertain, but it remains likely that FDTD techniques will dominate microwave and millimeter-wave simulations before very long. However, it is not clear at the present moment whether one can do all that is needed with Fidelity without IE3D along as a comparator and as a supplement.
Given the likely change in dominance of programs within the Zeland suite, perhaps it is time for the software maker to re-evaluate and remake its manual(s). Sundry comments along the way of this program overview might also form sufficient reason for the review.
I do not cast these notes as a review of Zeland’s offerings, because that step would require that I have sufficient standards and comparators by which to register a judgment. Unfortunately, almost all offerings in the FDTD and related domains of UHF/SHF modeling and simulation bear similar costs. Were it not for a specially arranged evaluation period courtesy of antenneX. I might not even have had the chance to work briefly with IE3D and Fidelity.
So my bottom line is not directed toward potential purchasers. It takes far more experience with UHF/SHF simulation to guide such possible users than I can bring to bear. My notes are largely directed toward those interested in antennas for the wireless and related region of the spectrum whose experience is limited to NEC or MININEC. The requirements for refined analysis of highly integrated systems of components, including antennas as simply one of those components, include very large increments of software sophistication beyond what we find in introductory level MININEC and NEC programs. As a crude example, even at VHF, our antenna models tend to ignore exposed fastener heads and similar small bumps in the otherwise smooth antenna element structure. Actually, from about 300 MHz upward, we should not ignore these bumps, although we routinely correct for them in field adjustments. Above about 600-800 MHz, we may no longer have the luxury of field adjustment. Put another way, field adjustment may consist of making a whole new prototype, just the task that simulation is supposed to eliminate. Hence, the input to the simulation package becomes ever more detailed, as does our ability to translate in minute detail the simulation into a working prototype.
Packages such as IE3D and Fidelity offer not only simulation engines for calculating with accuracy the likely performance of our creations. As well, they offer ever-increasing abilities to create the detailed geometries that close the gap between reality and model–or between model and reality, as you prefer.
Originally posted on the AntennaX Online Magazine by L. B. Cebik, W4RNL

Last Updated : 29th January 2025

error: Content is protected !!

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More

free counters