Microwave Journal
www.microwavejournal.com/blogs/1-rog-blog/post/21052-teaming-software-tools-with-pcb-materials

Teaming Software Tools With PCB Materials

November 15, 2013

Choosing a printed-circuit-board (PCB) material is a critical step in the design of any new RF/microwave circuit, since the material’s characteristics will ultimately determine the performance and capabilities of the circuit. Picking an electronic-design-automation (EDA) software package to model any new design can go a long way towards improving the design experience (and the levels of performance possible) since the right software can help predict the performance of circuits fabricated on different PCB materials. In particular, electromagnetic (EM) simulation software tools are quite useful for simulating high-frequency designs, and RF/microwave design engineers currently enjoy a wide selection of available EM simulation software tools. Quite simply, such software tools can predict the propagation of EM waves when not in free space, such as through the conductors of a PCB material with a known dielectric-constant (Dk) value.Understanding how these EM simulation software tools differ and how they can be applied to model different PCB materials can greatly benefit the design process.

In general, EM simulation programs use closed-form equations or field-solving methods to calculate the electrical behavior of a circuit design. Closed-form equations, often used to compute the electrical properties of conventional RF/microwave transmission lines such as microstrip and stripline, can be solved quickly (in real time) in a computer. As a result, EM simulation software based on closed-form equations can provide an excellent “first look” at a circuit design in terms of impedance and loss. These fast results come with some compromise in accuracy since these closed-form equations represent generalizations of common microwave transmission lines and structures and may not precisely portray a particularly circuit structure.

Field-solving software can provide much greater accuracy than programs based on closed-form equations, but may require considerably longer times to arrive at an answer. Such software can model the transmission lines and basic circuit structures of simulation programs using closed-form equations, but can also tackle more exotic circuit structures and designs. Unfortunately, field-solving software tools have been known to require days to solve more complex circuit calculations, and the accuracy of these software tools greatly depends upon a user’s knowledge of the computational limits of different field solvers and different EM analysis approaches and numerical methods used with different field solvers.

Among the analysis approaches most often used in RF/microwave EM simulation software are the finite-difference method (FDM), the method of moments (MoM), the finite-element method (FEM), the transmission-line matrix (TLM) method, the method of lines (MoL), and the Monte Carlo (MC) method. For example, the FDM is an EM analysis approach that has been used for two-dimensional (2D) as well as three-dimensional (3D) circuits and designs. This numerical method employs approximations to replace complex differential equations with more “computationally manageable” algebraic expressions (to shorten calculation times). The method represents a circuit or electronic structure with a physical space grid and a matrix. The approach has been applied with time variations, as the finite-difference-time-domain (FDTD) method, to perform modeling over wide ranges of frequencies. Provided that essential sampling theory requirements are met, the full-wave FDTD method can analyze a wide frequency range with a single simulation run.

EM software based on an MoM approach relies on calculations of boundary values (rather than calculations of values throughout a space of interest) and probability theory to achieve numerical estimations while using less computing resources than some of the other EM analysis methods. EM software using the frequency-domain MoM approach is often used for planar or 2D analyses based on homogeneous circuit media and typically thin conductive layers, essentially constructing a mesh across a surface to model that surface. The approach can provide fast solutions to circuit simulations, but is typically not used when full 3D EM simulations are required.

The FEM divides a circuit or structure into many subregions or finite elements for analysis. Different elements are used, depending upon the requirements of a model. Although the FEM modeling approach is typically more complex than EM analyses based on FDM or MoM approaches, it can also provide an accurate and thorough 3D EM analysis of a PCB and its circuits. In addition to 3D EM analysis of high-frequency circuits, the versatile FEM has been used in software tools for thermal analysis, stress-strain analysis and other engineering issues.  

EM software using the time-domain-based TLM approach typically approaches a circuit design as a set of lumped circuit elements, modeling simulated EM fields by wave pulses propagating in a mesh of high-frequency transmission lines. Simulation software based on this approach is usually applied to lossy, dispersive, and even nonlinear PCB media. The TLM modeling approach is very flexible, and effective for analyzing arbitrarily shaped circuit structures.

The MoL approach is a variation on a FDM analysis, although it can provide higher accuracy than a standard FDM EM analysis and with considerably less computational time and resources required. EM software using the MoL method essentially divides a circuit or structure for EM analysis into cells with one or two dimensions, and the unknown dimension or dimensions of the cells are found by means of analytical methods. The MoL method is effective for performing high-frequency planar EM analysis but can have limitations for predicting complex behavior, such as spurious modes.

In some cases, a software simulator may be asked to model a circuit or system with a large number of possible configurations, and MC methods make it possible to apply random number and probability theory to analyze and optimize those circuits and systems. Simulation software based on MC methods samples a design in a number of random configurations, using those random configurations to represent a circuit or system as a whole. Mathematical software tools often use MC methods for analysis, making it possible to vary different parameters in a circuit or system and study the effects of those variations on performance.

These are some of the modeling techniques commonly used in commercial EM software simulation tools. These software programs must calculate how the EM wave propagation will be affected by the circuit material’s characteristics at different frequencies of interest—no trivial task. A choice of PCB material with known and reliable characteristics can help to reinforce those EM simulations.

Do you have a design or fabrication question? John Coonrod and Joe Davis are available to help. Log in to the Rogers Technology Support Hub and “Ask an Engineer” today.