In recent years, it has become imperative that electromagnetic (EM) susceptibility and emissions of electronic equipment be strictly controlled. The primary emphasis for such controls comes from stringent government regulations restricting electromagnetic pollution.
In the U.S., the FCC has had such regulations since 1979. The European community will be moving to a pan-European standard for EM emissions beginning January 1996 under the directive 89/336/EC.
Industries throughout the world are now keenly interested in electromagnetic compliance (EMC) because of the cost-competitive nature of today’s markets. A product cannot be marketed if it fails the regulatory limits.
To meet the requirements, additional suppression components might be necessary or the physical configuration of the product might require modification. Such changes can add costs and involve additional schedule delays.
The high-speed and high-performance/high-speed systems of today are also more EMC-prone. They are more likely to be the source of EMC radiation. And in general, a noisy design from the EMC point of view may not be a good high-performance design in the first place. It pays to catch EMC problems early rather than do and fix. The old adage better prevent than cure later is particularly relevant to high-performance systems.
What is an Electromagnetically Compatible System?
A system is electromagnetically compatible if it satisfies these criteria:
o Does not cause interference with other systems.
o Is not susceptible to emissions from other systems.
o Does not cause interference with itself (crosstalk).
EMC is concerned with generation, transmission and reception of spurious EM energy. A source produces the emission, and a transfer or coupling path transfers the emission energy to a receptor. Interference occurs if the received energy causes the receptor to behave inappropriately.
EMC Mechanisms
A typical noise source, an IC drawing sharp pulses of current, has many efficient propagation paths available for its distribution and reradiation of high-frequency noise. The current loop path from the IC to the bypass capacitor forms a loop antenna. The magnetic field from this loop can also couple to adjacent wiring to cause further radiation.
Coupling can also result when different circuits share a common conductor, such as power and ground, that has a significant non-zero impedance at high frequencies. Examples include radiation from signals and power/ground leaving an enclosure, radiation coupled through enclosure apertures, and radiation from passive component resonance with noise sources.
The FCC and other regulatory agencies classify emissions as conducted or radiated. Figure 1 illustrates each of these types in relationship to nearby devices and enclosures.
Conducted emissions are low-frequency emissions. This is the noise conducted to the power and ground network. Since these networks are relatively large, they can act as efficient antennas and radiate the conducted energy.
FCC measurements for conducted emissions range from 450 kHz to 30 MHz. The conducted emission limits are specified in volts. For example, the FCC limit for class B digital devices is 250 uV for the 450-kHz to 30-MHz range.
Radiated emissions are the emissions at the higher frequencies. This is the noise directly radiated from the system. In digital electronics, where radiated emissions are of prime importance, FCC specifies standards for radiated emissions in the frequency range of 30 MHz to 1 GHz. The limits for this radiation are measured and specified as electric fields, and the units are volt/meter. Table 1 lists the FCC limits for radiated emissions measured at 3 meters from the device.
What Is the Problem…Solution?
Many companies rely on in-house compliance teams to assist designers in implementing basic guidelines and test for EMC at the PCB prototype stage. In turn, the compliance teams depend on years of experience and knowledge found in the company’s EMC handbooks.
In theory, this procedure works. But since most engineers and designers have neither the expertise nor the time to implement the documented guidelines, the information usually remains locked away until used by EMC specialists to conduct manual audits after a design is completed–a time-consuming, error-prone process.
The cause and effect of a latent, manual verification for EMC are simple. The cause: Only a fraction of development time, if any, is spent catching EMC problems upstream, during functional design, physical floor planning, component placement or interconnect routing. The effect: Companies are forced to resolve these problems through expensive workarounds and redesigns, resulting in higher product development costs and missed market windows.
Current EMC technology is directing possible solutions down two distinct paths. Electronic design automation (EDA) tools are being developed based on either an analytical or a knowledge-based approach.
Analytical Solutions: There are significant practical difficulties in developing analytical or numerical methods for solving EMC problems. In related situations such as signal integrity, crosstalk and ringing can be accurately and efficiently predicted using circuit simulators.
Unfortunately, the same techniques, based on well-understood circuit concepts, cannot be easily extended to predict EM radiation because a significant contributor to radiated emissions is the so-called common mode current (Ic). The quasistatic assumption inherent in circuit theory precludes the existence of common-mode current. So the current distributions obtained from circuit theory is only the so-called differential mode current (Id).
In circuit systems, Id is much larger than Ic. However, microamperes of common-mode current can cause radiation equivalent to milliamperes of differential-mode current.
In terms of analytical and numerical methods, this means that circuit theory is not adequate for EM prediction. Direct solution of Maxwell’s equations is necessary. On the down side, we lose the efficiency and familiarity of circuit formulation. In addition, EM simulation of systems as a whole is not practical. The best application of analytical methods is to generate EM rules and guidelines.
Knowledge-Based Solutions: In the absence of EM analysis programs capable of handling complex realistic structures, most companies have established EMI design guidelines to ensure that a product will pass regulations.
Rules-based solutions leverage the extensive validation of EMC rules and guidelines by industry experts and customer experiences. Starting from a prepackaged set of customizable, extendible rules, experts develop electronic EMC checklists to drive the process of designing digital, analog and mixed-signal PCBs. The production-proven knowledge base developed within a company also can be passed on from specialists to designers through automated rule development and verification tools that are easy to learn, set up and use.
Once rules are established, designers continually monitor compliance with push-button rule-checking throughout the design process, beginning at the functional design stage and continuing through to physical floor planning, placement and routing. When violations are identified, advisory feedback assists designers in correcting the problem. The tedious nature of classifying potentially critical signals is streamlined through integration with signal integrity analysis tools.
A Front-to-Back Design
The jury may be out on the overall viability of bringing automated EMC solutions into the mainstream engineering community. However, here are a few suggestions on what a design for EMC methodology should comprise:
o A formal EMC design considerations checklist should be established by the company’s EMC specialists.
o This checklist should be automated in software and seamlessly integrated with functional and physical design and analysis tools to be used by design engineers and layout designers.
o The checklist should be available during schematic design, component selection, board stack-up, component placement, plane structure and interconnect routing.
o The approach should ensure that EMC design solutions do not adversely affect manufacturability or cost.
References
1. Paul, C.R., Introduction to Electromagnetic Compatibility, John Wiley, 1992.
2. Goedbloed, J., Electromagnetic Compatibility, Prentice Hall, 1992.
3. Haus, H.A., and Melcher, J.R., Electromagnetic Fields and Energy, Prentice Hall, 1989.
About the Authors
C. Kumar is a senior member of the consulting staff at Cadence. Before joining the company, he was a post-doctoral fellow at M.I.T. Dr. Kumar holds a Ph.D. degree in chemical engineering from Clarkson University.
Sandeep Khanna is a Director of Product Marketing at Cadence. Previously, he was employed at Norden Systems and Hughes Aircraft as a design engineer. Mr. Khanna earned M.B.A. and M.S.E.E. degrees from the University of Bridgeport.
Cadence Design Systems, Inc., 270 Billerica Rd., Chelmsford, MA 01824, (508) 667-8811.
Copyright 1995 Nelson Publishing Inc.
February 1995