ABSTRACT
Electromagnetic interference (EMI) measurements may be divided into those imposed by EMI requirements, which also impose a test method, and measurements useful in nding the source of EMI problems. These problems include lack of self-compatibility or interference with other systems. Equipment manufacturers may not possess specialized EMI test equipment and therefore wish to make measurements with standard electronic test equipment, such as the oscilloscope. This chapter deals with the use of standard and specialized test equipment and perhaps most importantly with common sources of error in their use. It has been said that one measurement is worth a hundred predictions; however, EMI measurements may signicantly affect the parameter measured, resulting in error. The location of antennas may affect the calibration curve, for example, errors greater than 30 dB are common in undamped shielded-room measurements of radiated emissions. A measurement with an oscilloscope may corrupt the measured signal as a result of either the additional ground connection or by radiated pickup on the oscilloscope test leads. Thus, great care has to be exercised in the selection and application of the measurement technique; otherwise, the prediction may be more accurate than the measurement. A design should not be based solely on measurements of performance, such as lter attenuation and shielding effectiveness; instead, the measurement should be used to validate the design. When large discrepancies exist between theory and measurement, both should be examined for error. It is incorrect to assume that measurement data are intrinsically more accurate than predicted values. The measured parameter may not be the intended; for example, small loop antennas may intercept the radial magnetic eld and not the intended horizontal when the loop orientation is incorrect. The incorrect test equipment may be used for the measurements, leading to error. For example, a manufacturer of equipment that failed a commercial radiated emission test was using an amplitude modulated (AM) radio with a rod antenna to measure a relative reduction in eld after EMI xes were implemented. The criterion used was the very subjective one of reduction in audio noise. A potential source of error was the radio’s automatic gain control (AGC), which may have a 20 dB dynamic range. Thus, a reduction of 20 dB or greater in eld strength may be achieved with no perceptible change in measured noise level. Had the AGC voltage been measured by a DVM or small panel meter, a more accurate measurement might have been achieved. However, even this technique has the limitation that the AGC is operative only over a limited range of signal strength.
