Conformity - November 2008 - (Page 19) Intuitively, it can be understood as follows - the full-wave simulation includes all effects which corrects for any errors in the Smith model. Actually, GSCF bypasses the Smith model. It employs the numerical model as the base instead. When one works out the algebra, any of the Edmax 1 calculations in the Smith model drop out, leaving only the free-space AFs and numerical geometry correction terms. The Smith formulation is not necessary in the GSCF method. It is included as a historical artifact. The latest edition of ANSI C63.5 provides a measurement approach to obtain GSCF when numerical values are unavailable. This is described in Annex H of C63.5. This method involves measuring the same pair of antennas on a reference site at different geometries. This boils down to a site-to-site comparison with some extra constants present. Unlike the earlier uncontrolled site-to-site comparison method, Annex H requires that the pair of antennas be measured at least five times on different parts of the reference site. The standard deviation from these 5 measurements needs to be within a certain range for the site to qualify as a reference site. In addition, the reference site has to meet several physical specifications including size and flatness. Log periodic dipole arrays, or log antennas, are also common for EMC applications. Unified correction factors are not feasible, as log antennas vary greatly by manufacturer make and model. A different approach is under consideration, which is modified from the Smith model. This is called complex fit NSA (CFNSA). It fits measured site attenuation data vs. receiving antenna height to a mathematical model in an effort to solve for log antenna phase center positions and antenna patterns. Varying phase center position and antenna pattern deviations from that of a point dipole are the dominant error sources for log antennas in the Smith model. CFNSA does not make assumptions on either parameter. Interested readers are referred to [9, 10] for more information. Changes, Present Vision and Future Plans The latest revision of ANSI C63.5 has a number of major changes from previous editions and hence the 2006 edition must be used solely. Most of these were added as normative annexes. The primary additions relate to free-space antenna factors (FSAF), geometry-specific correction factors (GSCF), the use of a standard antenna calibration site (SACS), and an informative annex on uncertainty. These topics improve the technical accuracy and repeatability of the measurement methods described, and address the compatibility with international standards. However, there still seems to be an occasional misunderstanding as to the proper use of ANSI C63.5 document, which we will now explore. From the latest version of C63.5, there are two sets of data needed when using EMC antennas. These are related to each other, as explained above. One data set is for product testing and, by international consensus, is the FSAF. This means that there is only one AF for one antenna. The other set of data is used when determining test site validation. FSAF and GSCF are combined to account for the effects of geometry in these NSA measurements. If the antenna is to be used for product testing, then the freespace AF is used to determine the emission levels for both horizontal and vertical polarizations. The SSM provides near free-space AFs that are acceptable as-is if no corrections are provided. C63.5-2006 [11] currently has corrections for biconical dipoles, and is working on application of free-space corrections for other types of antennas. The rationale for having a stricter NSA requirement than the usual ±4 dB is for the antenna calibration site to conform tighter to theoretical NSA, since any error that the site might introduce from the theoretical value would propagate through all of the calibration process. If the ±2 dB NSA requirement is not met, the site shall not be used for antenna calibrations. There are three requirements for validation of the antenna calibration site: 1. 2. The test site shall be constructed in accordance with C63.7 [12]. Measured NSA shall be within ±2 dB of an ideal site. november 2008 Conformity 19 http://www.pseinc.com http://www.pseinc.com
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.