EMC Prediction, Analysis, and Modeling

It's always good to know what to expect before it too late to make changes

There are various approaches used for prediction and analysis. The approach used depends on the situation. Electromagnetic (EM) coupling from emitters to receptors (such as antennas, wires, cables, and circuit elements) is extremely complex, and involves multiple simultaneous coupling mechanisms. The advanced analytical techniques that are available such as method of moments (MOM), Finite Element, FDTD, etc , can be used to predict coupling levels with reasonable accuracy, provided the technique fits the geometry of the system. Some of these techniques are only applicable to bounded spaces or specific electrical configurations.

The prediction and analysis that is called out for the MIL-STD-461 EMC Control Plan is required so early in the design phase that the circuit geometry has not been established. In the absence of detailed design information regarding the hardware layout/configuration the EMC Systems Engineer is forced to use simple approximations to assess the EM coupling. This simplified approach is just a tool used when a quick response is required to determine the likelihood of low level devices and receivers being upset or damaged by high level RF emitters instead of waiting to perform a computational electromagnetic simulation.

The approach is based on the supposition that the EM coupled levels into the receptor wiring, cabling, and circuit elements can be modeled using the antenna characteristics of a non-resonant loop or dipole operating at the same frequency. That does not mean that it is impossible to have greater coupling than a dipole using a linear wire antenna but that requires the antenna to be specifically designed to do so, such as a linear array, and in most cases, we are dealing with ordinary shielded wires. Skip to end of discussion.

To use this approach to determine subsystem/equipment susceptibility requires an understanding of the susceptibility threshold of similar equipment. This information is generally determined during EMC testing, but in a new design, susceptibility data may not be available. In some cases analytical models have been previously developed and can be used. Receivers and EEDís are examples. However in most cases analytical data does not exist.

Even though the RF susceptibility levels that are used for an analysis may be the same as the actual systems levels, the voltage or current coupled into the system, will typically be much less than predicted. This is because worst case assumptions do not consider the geometry of the circuits involved, their location with respect to other grounded surfaces, any intentional or inherent shielding, EMI filters, and the circuit impedances at the susceptibility frequencies. Additionally the coupling may be near-field or far-field depending on the emitter to receptor separation distance.

There are three primary elements that make up an RF coupling situation, i.e. an emitter, a coupling process, and a receptor. Since the intent is to determine problem areas, the analysis needs only to address the worst case coupling which is generally intentional radiated RF energy coupled into wires, circuit loops, boxes, and antennas, and often does not consider conducted coupling processes. Each of these coupling elements is briefly described in the following paragraphs.

Emitter. The source of RF energy can be either Narrowband or Broadband and in the form of a changing current or voltage. These sources can be intentional such as a signal generator or broadcast transmitter, or unintentional such as an arc welder, lightning, or ESD. IN many cases this information is unknown and the susceptibility critera from the standard (MIL-STD-461, EU, etc) is used. There are three emitter types based on their modulation characteristics. The first is a carrier with narrow band modulation such as AM, FM, or Phase; the second is a carrier with broadband pulse modulation such as a RADAR system or computer clock; and the third is a single (non-repetitive) transient such as lightning or ESD.

Because the bandwidth for each emission type is significantly different the effect on the receptor will be different and the worst case analysis has to consider these differences. For example, a narrow band emission has an energy bandwidth that is less than the bandwidth of the receptor circuit, and increasing the circuit bandwidth will not result in capturing more energy. Not so with a broad band emitter. The captured signal amplitude will increase as the ratio of the recptor bandwidths, i.e. 20*log (BW2/BW1). A circuit with a really wide bandwidth (PRF < BW < 1/ t ) will respond to the energy in one spectral lobe at a time.

Coupling. The EM energy coupling process can be Conductive or Radiative. In the case of ESD/Lightning, it is generally referred to as direct injection if conducted and indirect if radiated. In the near field, radiated energy is coupled by electric (E) or magnetic (H) fields. The dominant field is determined by the emitter impedance. In the far field, radiated energy is coupled by plane waves in which the electric and magnetic components have a well defined relationship, i.e E/H = 377 ohms/sq. Regardless of the coupling process the energy always begins and ends as a conducted current. Most simplified analysis is restricted to worst case far field radiative coupling.

Receptor. The victim circuit element is generally a voltage detector with some bandwidth limiting. It can be an IC, a receiver via its antenna, PCB, EED, cable or any other device whose operation is upset by electromagnetic energy being picked up by a conductor associated with the device. The circuit impedance determines the type field to which the device is sensitive. High impedance circuits (Z>377 ohms) are more susceptible to electric fields whereas low impedance circuits (Z<377 ohms) are more susceptible to magnetic fields.

One of the shortcomings of Worst case analysis is that even though it does a good job proving a system will meet requirements -- because of the assumption -- it does a poor job determining failures. Even if a failure is indicated the system may still actually meet its requirements. But it does point out the problem areas early in the design.

The following analysis and modeling capabilities are available.

Problem Simulation and Modeling of Conducted and Radiated Coupling

Training is also available on the use of these techniques.