Test problems radiate below 30 MHz
Richard A. Quinnell, Contributing Technical Editor- September 1, 2004
The standard requirement for EMC testing below 30 MHz concentrates on the conducted emissions of devices. Growing interest in power-line communications (PLC), also called broadband on power-line (BPL) systems, is now forcing adoption of radiated emissions requirements for this band. What those requirements—and their related test procedures—will cover, however, remains uncertain.
Amateur radio operators have long known that the practice of guarding against only conducted emissions for frequencies below 30MHz is inadequate for protecting radio communications from unintended radiators. Organizations such as the ARRL (American Radio Relay League; www.arrl.org) have worked with manufacturers to correct interference caused by products such as switching power supplies and plasma-screen displays. Despite these efforts, changes to regulatory requirements have not been forthcoming.
That has changed with the advent of networking communications using the power distribution grid as the carrier. Individual products create isolated problems; PLC would cause global emissions.
As a result of the potential for widespread interference, organizations as diverse as the BBC (British Broadcasting Corp.; www.bbc.co.uk), Boeing Co. (Chicago, IL; www.boeing.com), and the NTIA (National Telecommunications and Information Administration; www.ntia.doc.gov) have all filed opinions with regulatory agencies. Each is arguing for strict control of PLC-radiated emissions to prevent interference with commercial, avionics, and other essential radio communications systems.
|Worldwide radiated emission standards are far from standard regarding low-frequency operation, but efforts to consolidate them are underway. Courtesy of University of Leige.|
A few countries, such as Germany, have implemented their own national standards. But, unfortunately, there is still disagreement over what the limits for emissions should be and how to measure them (see figure).
The international standard closest to completion comes from the European CENELEC/ETSI Joint Working Group on EMC of Conducted Transmission Networks (www.cenelec.org; www.etsi.org). The group has offered a draft standard that is now receiving comment and is expected to be adopted as part of the CISPR 22 regulations by mid-2005.
In the US, the NTIA recommended standards to the FCC in a report this past April (Ref. 1).
Draft standards all call for testing of emissions by using a calibrated loop antenna to detect magnetic fields, then converting the results to equivalent electric fields for comparison to specification limits. A measurement distance of 10 m is typically quoted.
Even so, there is still debate, particularly regarding the enforcement of the specifications. Jonathan Stott at BBC Research and Development points out that a 10-m measurement does not adequately represent the home environment where interference is most likely to occur, saying that a 3-m distance is more appropriate. The German government wants to eliminate any requirement to test above 30 MHz for PLC in order to simplify inspection, noting that PLC implementations use only lower frequencies.
Field tests or chamber tests?
Some engineers are also concerned about the ability to gain reliable field test measurements given the effects of local architecture and cable topology and given the presence of conducting elements inside walls. Yet, measurements in a test chamber may not be any better.
James McLean of TDK RF Solutions (Cedar Park, TX; www.tdkrfsolutions.com) explained that most chambers, even ferrite-lined ones, exhibit high-Q modes at frequencies below 30 MHz, complicating testing.
There is even disagreement on the relationship between the E- and H-fields close to cables. John Powers, who is conducting power communications research at Open University (Manchester, UK; www.open.ac.uk), pointed out that the assumption of a 1/R field strength variation near a cable is valid only for infinite straight wires. He noted that because power-line systems are finite line segments, the assumptions do not apply.
While the debates rage on, the standards process is slowly grinding toward resolution. When the standards are complete, EMC test engineers will have to begin handling a whole new range of radiated frequencies