Numerical EM Modeling Questions

2. Which electromagnetic modeling software should we purchase to help solve our EMC problems?

This is truly a frequently asked question, since we receive regular emails and phone calls with this request. Unfortunately, it is a difficult question to answer without knowing the specific requirements of the person who is asking. The following points should be stressed and they apply to all numerical electromagnetic modeling codes:

• Regardless of the purchase price, numerical electromagnetic modeling codes require a significant investment of time and effort to learn to use properly. Expect to spend weeks learning to run the code and months before you can model the most basic configurations of interest. It generally requires years of experience with modeling and validation before you can be truly comfortable and confident with results obtained from numerical modeling software.

• There is no numerical modeling software that will help locate or anticipate problems with existing printed circuit board designs. Even laboratories that are heavy users of EM modeling software (like ours) do not try to use these tools to identify existing problems. Expert system software, design rule checkers, and troubleshooting experience are the best tools for identifying and solving EMC problems with existing designs.

• Numerical modeling software is good for studying known problems where the source, victim and coupling path have been identified. It answers the question, "What will happen under these conditions." If you can't provide all of the relevant boundary conditions, you won't get the answer you are looking for. You will always get a very specific answer to the very specific set of conditions you provided.

• There is no single technique that is good at modeling everything (or even most things). Different techniques such a methods of moments, finite element methods and finite difference time domain codes are all good for modeling some problems and inappropriate for modeling others.

Even with ideal codes and infinite computing resources, numerical modeling codes could not be used to predict the results of most EMC tests. This is partly because the tests themselves are well enough defined to ensure a repeatable outcome in every situation. However, it is primarily due to the fact that there are millions of parameters in a typical electronic system that may have a significant effect on the outcome of any given test. Product software, component transition times, material properties of insulators, and the RF impedance of metal-to-metal contacts are just a few examples. It is not possible to determine all of these properties and provide them to a numerical modeling code.

So if you're looking for software that will help you estimate radiated emissions from electronic products or troubleshoot existing designs, don't expect numerical electromagnetic modeling codes to be of much help. However if you're looking for tools that can help you to really understand what is happening when the relevant geometries and materials are clearly identified, then numerical modeling tools are invaluable. The real strength of numerical tools is apparent when both models and measurements are employed. Measurements can be used to confirm that the models have included all the relevant parameters. Models on the other hand can be used to verify that measurements results were not affected by operator or equipment errors and that the quantity being measured is in fact the quantity of interest. Models confirmed by measurements are also able to provide a wealth of information that is not directly measurable, such as the fields or currents in inaccessible places.