Conformity - November 2008 - (Page 16) caused by the canceling of the ground-bounced and the direct rays, one of the antennas is scanned from 1 m to 4 m in height. Only the maximum response is kept as the base for calibration. NSA measurements for site validation tests are a reverse of the SSM for antenna calibrations. In this case, the antenna factors are known, and site performance can be compared with a theoretical value to gauge its fitness (within ±4 dB as indicated in ANSI C63.6 [6]). NSA is defined to be the site attenuation (path loss between the antennas over the ground plane in decibels) subtracted from the antenna factors of the two antennas involved. To make the Smith formulation work, several assumptions had to be made: • Recalling the formulation is based on the free-space farfield Friis equation, the antennas are assumed to be in the far field of each other. Near field terms are not considered. • To mathematically remove the ground bounce, the antenna pattern had to be known a priori. Since they are not known, all antennas are assumed to have the same pattern as a point dipole; i.e., with the well-known donut shape. • Mutual couplings among antennas and their ground images are not considered in the formula, and thus assumed to be negligible. • Antennas are considered to be physically small, so that the whole of the receiving antenna is considered immersed in a uniform field. • The separation distance is assumed to be known, which is the same as that of the two theoretical point dipoles. These assumptions work well for some geometries, such as in the case of biconical antennas at 10 m separation distance, horizontal polarization, and the transmit antenna at 2 m height. However, for site validation measurements over a volume occupied by the equipment under test, NSA as defined in ANSI C63.4 involves four geometries per separation distance (two antenna heights in horizontal and two heights in vertical polarization). It was quickly realized by many involved in site validation tests that errors due to these assumptions may be so large (they can be on the order of 3 or 4 dB in some instances) that they overwhelm and obfuscate the true site performance. An example of an OATS is shown in Figure 2. As shown in Figure 3, for example, a perfect site would fail the NSA measurement if only free-space AFs are allowed when measured with a pair of common biconical antennas. Test labs and anechoic chamber manufactures got around the limitations caused by these imperfect assumptions in the theoretical model by using the so called “geometry-specific antenna factors.” In practice, people first calibrated their antennas at all geometries required by NSA measurements, and generated antenna factors (using the SSM) for these specific setups. Of course, these antenna factors have the errors (caused by the above assumptions) built in. When the NSA measurements were performed, these same antenna factors were plugged back into the NSA formula. Realizing that NSA procedures are just the reverse of the antenna calibration process, the same errors cancel out. In reality, this is exactly the same as a site-to-site comparison method. Users end up comparing the site attenuation of the site-under-test to the site where the antenna calibrations were performed. Any of the mathematical calculations involving theoretical NSA and antenna factors are unnecessary (as they cancel out) when using the “geometry-specific antenna factors.” Figure 2: An example of an Open Area Test Site (OATS) used for calibrating antennas (photo courtesy of ETS-Lindgren) Figure 3: Measured NSA on an open area test site and numerically simulated NSA on a theoretical perfect site (vertical polarization, R=3 m, h1=1 m, and h2= 1~4 m), both by using free-space AFs 1 Conformity november 2008 In the late 1990s, a drive to use free-space antenna factors for product testing was gaining momentum, both domestically
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.