A conventional near-field measurement system can be considered to be a form of a two dimensional CW SAR radar. The grid of sampling points form a two dimensional synthetic aperture antenna array. The uniformly illuminated aperture used in near-field antenna measurements must be modified into a low side lobe design by tapering the illumination at the aperture edges. This is handled by the additional step of applying a Kaiser-Bessel or similar window function (Harris, 1978) as a function of spatial position to the DUT measurements.
The scan pattern can be virtually any type including planar raster, plane polar, cylindrical and spherical. The sample positions correspond to element positions in a synthetic aperture phased array antenna.
The test procedure is very similar to near-field processing with a few modifications. First, an XY planar grid of points is acquired. This raw data can be plotted in any form for use in the traditional quiet zone quality evaluation. The plots are typically scaled to a full scale range of a few dB as opposed to the typical 40 or 50 dB ranges when used for conventional antenna near-field measurements. The phase range also tends to be quite small (+/-15 degrees). X and Y cuts can be used to measure the phase curvature and amplitude ripple in the test region.
The far-field transform is then used to convert the measured phasefront into an angle or K space representation. As a significant difference from conventional near-field processing, the data into the Fourier transform must be tapered by a window function to suppress side lobes. The window function used typically is a Kaiser-Bessel window with a relatively high coefficient value.
The output of the transform is the received energy in either angle or K space. An ideal anechoic chamber should show a small illuminated region slightly off boresight (see figure 8). A beam precisely on axis corresponds to a coherent on axis leakage source, receiver quadrature unbalance or often a LO leakage in a remote mixer configuration. Beams at other angles are caused by reflections from lights, supports, walls, etc.
If the transform output is in angle space, one can read the direction of an interference path and transfer the coordinates into a theodolite. The theodolite, when positioned at the scanner position and aligned with the scanner reference frame, will point along the interfering path. The K-space mapping is somewhat like a fisheye lens image of the same thing. There should be no energy outside a K space circle of radius one. If energy is present, it is created by a multipath mechanism.
The leakage signal was removed by adding appropriate isolation in the RF network. Figures 5 and 6 represent the performance of the chamber after the leakage signal was removed. The amplitude ripple has been significantly improved, however there is still clear evidence of additional undesirable reflections at the -25 and -30 db levels. Through use of the angular spectrum information, it was relatively easy to isolate the cause the several of these reflections to be a support structure for the positioner.
Figures 7 and 8 show the final results after eliminating the support structure reflections and also now reorienting the XY scanner to be normal to the range axis. Reflections from the sidewalls, floor and ceiling are evident, but these are down at about the -60 db level. The amplitude ripple is now observed to be down to about +0.2 db. Further work on absorber placement couls likely have reduced the ripple lower.
The resluts of the imaging can be quite useful in diagnosing problems in the chamber configuration and can lead to improvements in the chamber's antenna measurement accuracy.
Hindman, G., Applications of Portable Near-Field Measurement Systems, AMTA symposium, Antenna Measurement Techniques Association, Boulder, CO, 1991. Shows examples of several types of portable near- field scan systems used for various applications including chamber imaging. Includes information on typical setup times, ease of use, and accuracies to be expected. Mensa, D., High Resolution Radar Imaging, Artech House, Norwood, MA, 1981. Describes pulse compression, synthetic aperture concepts and the related processing algorithms. A classic in the field. Slater,D.,Near-Field Antenna Measurements, Artech House, Norwood, Ma,1991. A comprehensive overview of near-field measurement techniques. Includes chapters on small, portable near-field measurement systems. Discusses similarity between near-field measurement concepts and terminology used in other fields such as synthetic aperture radar (SAR).
Figure 1 - NSI model 233T portable scanning system
Figure 2 - Portable near-field scanner to detect multipath and leakage effects in an anechoic chamber
Figure 3 - Amplitude ripple due to large leakage signal
Figure 4 - CW SAR image of "Dirty" chamber
Figure 5 - Amplitude ripple due to reflections
Figure 6 - CW SAR image showing reflections from floor and ceiling
Figure 7 - Amplitude ripple in "clean" chamber
Figure 8 - CW SAR image of "clean" chamber