A New Standard Radiator For Shielding Effectiveness Measurements

Abstract

The lack of a standard emission source make shielding effectiveness measurements difficult to perform with any degree of repeatability. This lack of repeatability tends to make EM1 engineers reluctant to place too much reliance on shielding effectiveness measurements. This paper describes a new standard radiator that can be used in making shielding effectiveness measurements that are repeatable. This source is physically small, battery powered, and operates over a frequency range of 10 MHz to >1 GHz. The radiator produces a standard dipole radiation pattern, and radiates sufficient energy to test most shielded enclosures. Introduction The need to predict the performance of an EM1 shielded device has become more important than ever. As the speed of computers and other devices increase, existing 'rule-of-thumb' shielding designs are no longer sufficient. Traditional methods to quantify the effectiveness of designs to control EM1 are becoming outmoded. One of the major contributors to this problem is measuring the performance of shielded enclosures or boxes. It is becoming accepted throughout the industry that shielding effectiveness measurements have limited use. However, shielding effectiveness measurements continue to be used by many people as a measure of 'goodness' of a particular shielded enclosure or box. This paper will describe a new standard source that may be used in evaluating the shielding effectiveness of enclosures. It will also discuss some of the problems associated with shielding effectiveness measurements and show how this new radiator improves the the repeatability of measurements. This new standard radiator was developed 'ointly between the National Institute of Standards and Tecknology (NIST), the U.S. Navy, and Digital Equipment Corporation. This new radiator is physically small (IO cm diameter), battery operated, operates over a frequency range of 10 MHz to above 1 GHz, and has no metal connection between the radiator element and the control unit. Di ita1 has successfully used this radiator to correlate the &AI performance changes with enclosure modifications, for both large and small enclosures. Shieldina Effectiveness The term 'Shielding Effectiveness' is commonly misused by most engineers and managers. Shielding effectiveness is really a relative term, having meaning only when Mark Seth Seth Industries, Inc. two similar constructions are compared. Unfortunately, they have become very familiar with the term, and they feel very comfortable using shielding effectiveness values without considering the difference between the tested configuration and the target configuration. It is not uncommon for an engineer to measure the radiation from a particular circuit board, note the difference between the received level and the desired limit, and decide on the number of dB of shielding effectiveness needed to meet the desired limit. Since the presence of the shield will change the radiation source characteristics, and since the shielding effectiveness value was determined from a different type test, the expected result may not be realized. Most shielding effectiveness values are created using a far field Electromagnetic (EM) source. This is fine for radar applications, but typical computer shielding applications have EM sources in the near field. The interaction between the source and shielding panel (when in the near field) cannot be ignored if reasonably accurate results are desired. All equations, graphs, etc. are also created from far field configurations, a ain resulting in inaccurate results when applied to reayistic shielding.problems. Since the EM source will be changed due to the presence of the shield, the shielding effectiveness test must include this effect, or any test results will be merely a comparison of 'apples and oranges'. Naturally, it is valid to compare various shielding options when the actual source is used, became then the final result is based upon the true configuration. Comparison between 'option A and 'option B will show a relative difference between the shielding effectiveness of the two options. A better term for this comparison test might be "Shielding Performance." In spite of all the problems associated with shielding effectiveness measurements, they still remain useful to the design engineer. When a particular shielded enclosure will be used for a given computer, etc., then design decisions must often be made concerning design trade-offs. However, extreme care must be taken when making these design decisions, since often the test results are more a measure of the test setup than the true shielding effectiveness. Shieldina Effectiveness Tests The problems inherent in shielding effectiveness tests of enclosures, cabinets, and boxes are well known throughout the industry. A few of the most critical problems will be repeated here. 1. The test set up is critical, and often a slight change in source position will cause the results to vary wildly. CH3169-0/92/0000-0052 $3.00 01992 IEEE 256 2. If the radiating antenna is too large, then the interaction between the shield and the antenna (effectively loading the antenna) will cause large differences between the a.mount of energy delivered to the antenna and the expected value. 3. The coax cable traveling from the enclosure to the shielded test room wall (or generator) can 'leak', resulting in reduced effective Shielding effectiveness. 41. Typically, the wrong radiating source is used. Some configurations would use a horn antenna as the source, but the actual source of EM1 is seldom as well behaved as i i plane wave. Some configurations used a wire strung within the enclosure as the soiurce, but no concern was given the the 'antenna' impedance, and so the actual energy launched by this antenna would vary dramatically as a function of frequency. E;. Some configurations required a large RF generator be placed within the enclosure, making this setup impractical f'or anything smaller than a full cabinet. Many configurations were so sensitive to the physical setup within the enc:losure, that results could not be duplicated. All these factors made EM1 engineers reluctant to use the shielding effectiveness test results with a high degree of confidence. These problems can be overcome, but the source must be physically small, electrically isolated, and electromagnetically well behaved. Such a source exists, and should be used as the general shielding effectiveness source when testing shielded cabinets, enclosures, and boxes. !standard Dipole Radiator Description The Standard Dipole Radiator (SDR) is a physically small (1 0 cm) s herical source that is electrically isolated. Many of t l e problems associated with typical radiating sources are eliminated when using the SDR. The SDR !System is depicted in Figures 1 & 2 and consists of: 41 The Sphere (antenna) 4) The Control Unit o Interconnecting Fiber Optic Cables 'The Sphere is the radiating element or antenna. Circuitry ;and power sources for developing the standard field are contained within the sphere's core. The control unit supplies the sphere with a modulated laser signal and provides monitor readouts. The control unit is connected to the sphere by fiber optic cables which provides electrical iisolation between the sphere and control unit. Figure 3 shows the block diagram of the SDR System.