EM.Picasso In A Nutshell

From Emagtech Wiki
Jump to: navigation, search

A Quick Introduction To Planar MoM Simulation

EM.Picasso (Printed Circuit and Antenna System Simulator and Optimizer) is a simulation software tool for analysis and design of planar microwave circuits and antennas. It features a computationally efficient simulation engine that is based on the 2.5D Method of Moments (MoM). The method of moments is a full-wave integral-equation-based numerical technique for solving Maxwell’s equations. Rather than discretizing the entire computational domain, MoM takes advantage of the analytical solutions that might be available for the background structure to reduce the mesh size significantly. This is embodied in the form of dyadic Green’s functions, which are indeed the solutions to the background boundary value problem with elementary sources. MoM’s highly reduced mesh size can lead to faster computation times and lesser memory requirements. This is particularly important for antennas and other open-boundary structures. In such cases, the appropriate Green’s functions take care of the unbounded background medium of the problem, which inherently has infinite extents.

Unlike FDTD, which has a very general nature, the applicability of the mehtod of moments depends on the availability of the Green’s functions for a given problem. For many planar microwave circuits and antennas, the background structure is a layered planar substrate. This consists of a vertically stacked configuration of one or more dielectric layers of infinite lateral extents. From the top and bottom, the stack-up may end in either the free space, or an unbounded dielectric half-space, or perfect electic conductor (PEC) ground or a perfect magnetic conductor (PMC) ground. There are Green’s function solutions for such multilayer planar structures, albeit not in closed form and involving recrusive algorithms. EM.Picasso carries out all these mathematically complex computations behind the scenes during a MoM simulation.

The first thing you do in EM.Picasso is to define the background structure of your project in the Stack-up Manager. Planar metal objects like signal traces, microstrip lines, patches, etc. are modeled as PEC traces. Slots carved out of infinite ground planes are modeled as PMC traces. These are used to couple energy from one side to a substrate layer to the other side of the ground. Vertical vias and interconnects, shorting pins, plated-through holes, etc. are modeled as embedded object sets. You can define both metal and dielectric embedded objects. In EM.Picasso, you draw only surface objects like rectangles, circles, triangles, polygons, etc. on horizontal planes. In the case of embedded objects, you draw their cross sections at their base plane. EM.Picasso automatically extrudes the base geometry and extends it across the host substrate layer. There is an important approximation that EM.Picasso makes when modeling embedded objects. It assumes that volumetric objects carry vertical currents or sustain vertical fields only. This is the underpinning assumption of the 2.5D MoM, which limits the legitimate embedded object types to either slender columns and posts, or thin films sandwiched between two metal plates.

A Planar MoM simulation requires a source and one or more observables if you want to see any results at the end of your simulation. Since it is quite fast, EM.Picasso can be effectivley used to compute the port characteristics of planar circuits like filters, matching networks, directional couplers, etc. or the radiation characteristics of planar antennas and arrays. De-embedded sources are specifically provided for accurate port characterization. A plane wave source can be used to compute the reflection and transmission characteristics of planar layered periodic structures.

EM.Picasso provides different simulation modes to suit your different modeling needs. The simplest mode is analysis, or a single-shot simulation. You click the Run button, and in a few seconds, the simulation results are generated. You can also run parametric sweeps where the value(s) of one or more project variables are varied successively. You can define intricate relationships tying your project variables to the geometrical and material properties of your physical structure. Or you can perform an optimization whereby the project variables are varied in a systematic way to achieve one or more design objectives defined based on your project observables.


Back to EM.Cube Wiki Main Page