EMC in a High-Frequency World

The new year brings numerous EMC challenges, such as increasing clock speeds and conformity to worldwide standards.

Q. As clock speeds on electronic devices continue to rise, what critical technical challenges await the engineer aiming for electromagnetic compatibility (EMC) compliance?

A. Because it is only recently that clock speeds are exceeding 1 GHz, there is little public information on the subject. Common thinking is that, moving to frequencies above 1 GHz, all structures, including cabinets, enclosures, cabling, etc., can act like efficient radiators. From an antenna perspective, an effective radiator is a quarter- or a half- wavelength. At frequencies in the gigahertz region, a half-wavelength is only 24 cm (about 6 in.), meaning an antenna can be a ventilation slot, a chassis door without correctly installed radio-frequency-interference (RFI) gasketing, or even a card cage. If doors or other openings have gaps (e.g., improper fitting between the door and its mating surface) this long or longer, radio-frequency (RF) energy ingress and egress takes place—hence the need for better design for emission suppression and immunity. In other words, there is no shielding attenuation of a slot or gap when it is half a wavelength long; and once an emission from the product gets through a slot, it acts like any other wave in both its near- and far-field effects on surrounding devices. Similarly, the slot is a receiving aperture or antenna and lets in RFI environmental noise. This in effect moves EMC design down to the printed-circuit-board level, in addition to the elimination of RFI on attached external cabling via filtering.

Q. How do increasing clock speeds affect EMC measurement?

A. EMC measurement difficulties increase as the frequencies exceed 1 GHz for several reasons:

Product emissions tend to be in narrow beamwidths and so are difficult to find.
Measuring antennas tend to have narrower beamwidths, making it difficult to aim and search at the limit distance away from the product and still capture the radiated signal.
Instrumentation such as spectrum analyzers and receivers must pick up low-level signals at typical limit distances of 3 and 10 m between the antenna and the product under test.
Antenna-to-receiver cables become much more lossy, affecting the ability to measure signals above the noise floor of the spectrum analyzer or receiver.
More time is required to find the maximum emissions, due to the necessity of manual searching as the product cycles through its operating modes.
Product immunity to incident radiated fields is more difficult to assess due to the narrow beamwidth and the product's apertures, such as cable ports and slots.

Q. How does IEEE address needed standards work in areas such as the influence of increasing clock speeds on EMC?

A. IEEE is comprised of over 30 technical societies, including the IEEE EMC Society (EMCS), in turn made up of technical committees that deal with technology in a specific subject area. When standards are needed, the appropriate technical committee (TC), in this case the EMCS TC-4, Electromagnetic Interference Control, works with the society's Standards Development Committee to develop a standard. The writing and revising of standards, including studies of possible areas requiring standards, is a continual process.

Higher clock frequencies are also currently being discussed in the EMCS annual symposia. The next symposium is set to take place in Montreal in August 2001, and TC-4 is sponsoring several technical sessions on EMC design. The symposium in Washington, DC, this past August is another resource for the latest papers on the subject.

Q. How is the proliferation of portable electronic devices and wireless communications devices affecting EMC compliance?

A. Any time you have more sources of RF energy, the EMC design must accommodate with greater immunity. One area of growth is the use of the 2.45 GHz band, where such activity as Bluetooth, cordless phones, HomeRF, new RF lighting, and other systems are all vying for use and must work with each other's ambients. The proponents of these devices are thus concerned with interference from other devices using the same band. There have been claims of interference, but this situation is still coming to a boil. It is only a matter of time before products with lesser immunity in this band will not work together at all user locations.

The 2.45 GHz band is a very active area and one where the boundary between system design and EMC is unclear. A current study within the EMCS deals with the potential for interference in this band and possible remedies for that interference. The study is being conducted jointly by TC-6 (Spectrum Management) and the Standards Development Committee. IEEE Working Group 802.15 is conducting a similar study on interference between wireless local-area-network equipment, which falls under IEEE 802.11, and Bluetooth, which 802.15 is considering standardizing. These efforts are sharing information, though at this point the facts have not been clearly established, so it is not clear that there is a need to fix anything.

This is also partially an RF spectrum allocation issue, in which the Federal Communications Commission (FCC) allows only certain services in a band as a preference. All other uses of the band, when allowed, are usually low power and are always of a secondary priority level and therefore must live with any interference they encounter, with no relief from regulators. One such allocated service is industrial, scientific, and medical (ISM) equipment, which partially falls under the 2.45 GHz band and has unlimited radiated energy at this frequency (±50 MHz) as a priority. If manufacturers want their products to operate in this part of the spectrum, they must build more immunity or signal discrimination into those products, to create a system which does not respond to fields such as those from high-power ISM equipment.

Q. What changes in test site qualifications and testing methodologies can test houses and compliance laboratories look forward to in the next year?

A. The biggest challenge is how to qualify test sites used for measurement of radiated emissions above 1 GHz. The effect of the reflecting ground plane used below 1 GHz is under study for use above 1 GHz and may not be needed. If product emissions are narrowly focused and the measuring antenna also has a narrow beamwidth, does the emitted signal ever see the ground plane? If not seen, should the limit be changed because the ground reflection is not added to the direct signal? As a result of these questions, there is some suggestion that to ensure that the ground plane reflection is not considered, RF-absorbant materials should be placed on the ground plane to make a nearfree-space measurement with only the direct path between the product and antenna.

American National Standards Institute (ANSI) Accredited Standards Committee (ASC) C63, Subcommittee 1, Techniques and Developments, is pursuing analytical and experimental work to determine the beamwidth peculiarities of common antennas used above 1 GHz, how to qualify a test site above 1 GHz, and how to measure for maximum emissions other than by manually scanning all product surfaces to find the narrow emissions beam, which requires an enormous amount of time. Such study has led to the discussion of alternate techniques, such as the use of reverberation chambers or gigahertz transverse-electromagnetic-mode (TEM) waveguides. These chambers and waveguides in essence measure the energy or total power from the device in free space before applying those results to determine suitable limits that might better show conformance with regulations. However, these techniques are still not fully subscribed to by regulators, but they are gaining acceptance for certain types of products, such as those that can fit inside a TEM waveguide.

The next area of concern for testing laboratories is the growing requirement to be accredited for conformity work worldwide. The new international standard ISO/IEC 17025 on test and calibration laboratory competency will help test houses meet this requirement. The standard, published in December 1999 with a 2000 date, is already seeing use by accrediting bodies, and widespread use of the standard will occur by the start of 2002. The American Association for Laboratory Accreditation (A2LA) and the National Voluntary Laboratory Accreditation Program (NVLAP), two U.S. accrediting bodies in the area of EMC testing, have adopted ISO/IEC 17025 to replace ISO Guide 25. By the end of the 2-year transition period, only 17025 will be used. Test houses currently have the following options: to be audited and accredited to Guide 25; to be audited and accredited to Guide 25 with the addition of a gap analysis as to what is needed to meet 17025, in anticipation of having to meet it for the next audit; or to be audited and accredited directly to 17025.

Q. How does the International Electrotechnical Commission (IEC) affect U.S. EMC standards?

A. The IEC has two major technical committees which handle EMC: IEC TC77 and the International Special Committee on Radio Interference (CISPR). These committees have U.S. participation from a variety of sources, including product manufacturers, testing laboratories, academia, and regulators, all of which provide input that affects committee decisions. Especially affected are U.S. manufacturers that sell internationally, because IEC standards, translated into European norms, must be met before their products can be imported into Europe.

Since most companies are multinational and many countries are using IEC/CISPR standards as the basis for their regulatory systems, these standards must be met at some point. The ASC C63 committee is publishing pertinent IEC/CISPR standards with U.S. forwards, to state any U.S. differences or areas not applicable. CISPR 22, on emission measurements and limits of information technology equipment, is now available in the United States without going to a European source, and more such publications are on the way.

Q. How has the EMC Directive affected U.S. electronic-device manufacturers?

A. This affects U.S. manufacturers in that they not only have to meet RF emissions but also an array of immunity requirements not mandated in the United States. FCC has been more concerned with protecting the radio spectrum than with immunity, as immunity is considered a self-correcting product quality issue. In other words, if a manufacturer has a product that does not work in its RF environment, then that manufacturer will not stay in business or will get many complaints. However, it may take time for these issues to surface. For the European Union (EU), the EMC directive mandates that products have intrinsic immunity and not cause interference—no choice left to the manufacturer. The U.S. allows more latitude and lets the marketplace rid itself of products with inadequate immunity. What this means for U.S. manufacturers offering products in Europe is that in addition to meeting common emission requirements used worldwide and based on IEC/CISPR 22 information technology equipment limits (which are similar to FCC limits in most of the frequency ranges), they must meet EN 61000-4-2 (ESD), 3 (radiated immunity), 4 (fast transients/bursts), 5 (surge), 6 (conducted immunity), 8 (magnetic power-fields immunity), and 11 (voltage interruptions, dips, etc.), which are based on but not necessarily identical to IEC 61000-4-2, 3, 4, 5, 6, 8, and 11. Additionally, there are power line emission requirements in Europe based on IEC 61000-3-2 (harmonics) and 61000-3-3 (flicker).

Q. How is the IEEE Standards Association (IEEE-SA) pursu-ing the harmonization of EMC standards?

A. The IEEE EMCS Standards Development Committee's members are all IEEE-SA members and hence work in the IEEE-SA system to get their standards published. The IEEE-SA is working on arrangements for using its standards within the IEC standardization process in several technical areas, but EMC is not currently one of these areas. IEEE-SA and EMCS members are represented on technical advisory groups for TC77 and CISPR, and some are also on the TC77 and CISPR working groups, which helps facilitate the goals of harmonization. The IEEE-SA is an international organization, with members throughout the world active in their own country's TC77 and CISPR activities. In addition, the IEEE-SA holds the secretariat for ASC C63, which coordinates many EMC measurement standards in the U.S. and has Subcommittee 3, a technical advisory group for the U.S. for the CISPR plenary. It is through such international representation that the IEEE-SA supports key areas of EMC standardization in multiple countries.

Donald N. Heirman is a NARTE-certified EMC engineer and president of Don Heirman Consultants (Lincroft, NJ), a training and educational EMC consultation corporation. Bringing with him more than 30 years of experience in EMC, Heirman chairs, or is a principal contributor to, multiple national and international EMC standards organizations, including ANSI ASC C63 and CISPR. Heirman is a fellow of IEEE; is a member of its EMC Society board of directors and vice president for standards; chairs the IEEE Standards Association Standards Board; is president of NACLA; and is a member of the Technical Management Committee of the U.S. National Committee for IEC, among many other contributory positions in the field of EMC. Heirman took a few minutes out of his very busy schedule to speak with assistant editor Joshua Glover about his perspectives on the current state of EMC.

Donald N. Heirman can be reached at d.heirman@worldnet.att.net. For more information about ANSI, IEEE, or IEC, visit the organizations' Web sites at http://www.ansi.org, http:// www.ieee.org, and http://www.iec.ch, respectively.

Back to January/February Table of Contents