Welcome to the AMTA paper archive. Select a category, publication date or search by author.
(Note: Papers will always be listed by categories. To see ALL of the papers meeting your search criteria select the "AMTA Paper Archive" category after performing your search.)
We report on the initial phase of our study to assess the electromagnetic interference and electromagnetic compatibility measurement and calibration procedures at the National Institute of Standards and Technology. We are developing a measurement-based uncertainty analysis of calibrations and measurements in the anechoic chamber. We intend to characterize all sources of uncertainty, which include power and probe-response measurements, noise, nonlinearity, polarization e.ects, multiple re.ections in the chamber, drift, and probe-position and probe-orientation errors. We present simple and repeatable measurement procedures that can be used to determine each individual source of uncertainty, which then are combined by means of root-sum-squares to state the overall measurement or calibration uncertainty in the anechoic chamber. We report on work in progress and future plans to characterize other EMI/EMC facilities at NIST.
This paper discusses a measurement technique for accurately characterizing low cross polarization level of antennas, and associated sensitivity and errors. The technique involves two-antenna transmission (S21) measurement that includes an AUT and a reference antenna that has low cross polarization level. This technique needs two far-field transmission data from two different relative roll angles. The cross-polarization sensitivity is determined by SNR of cross-polarization component and cross-polarization of the reference antenna. The cross-polarization error is related to roll angle uncertainty and receiver noise.
Spherical near-field to far-field transformation techniques allow for the reconstruction of the complete radiation pattern of the antenna under test (AUT) from the knowledge of the tangential electric field over a spherical surface [1-2]. However, in practical spherical near field measurements there are zones on the measurement sphere where data are either not available or less reliable. When the spherical wave coefficients (SWC) are calculated from incomplete near-field data by setting to zero the unknown samples, the abrupt discontinuity in the field values at the edge of the scan area may lead to erroneously large values of the higher-order spherical harmonic coefficients. Different solutions have been proposed to circumvent this problem [3-4] and have been demonstrated effective for small truncation areas [3]. In this paper a novel approach is proposed for the reduction of the truncation error in spherical near-field measurements. The method is based on a proper filtering of the SWC in accordance with the extent of the minimum sphere enclosing the AUT. More specifically, it consists in iteratively imposing the matching of the near-field with the measured samples and performing a spectral filtering in the spherical harmonics domain, based on the knowledge of the physical extent of the AUT [5-8]. The procedure has been tested on synthetic as well as measured near-field data and has proved to be effective and stable against measurement errors. The approach has shown to be effective even for increasing truncation areas.
Millimeter-wave measurements on spherical near-field scanning systems present a number of technical challenges to be overcome to guarantee accurate measurements are achieved. This paper will focus on the affect of mechanical alignment errors of the spherical rotator system on the antenna’s measured performance. Methods of precision alignment will be reviewed. Sensitivity to induced mechanical alignment errors and their affect on various antenna parameters will be shown and discussed. Correction methods for residual alignment errors will also be described. The study includes 38 and 48 GHz data on the Alphasat EM model offset reflector antenna measured by TeS in Tito, Italy on a NSI-700S-60 Spherical Nearfield system, as well as a 40 GHz waveguide array antenna measured by NSI on a similar NSI-700S-60 Spherical Nearfield System at its factory in Torrance, CA, USA.
This paper analyzes the applicability of near field scanners into existing compact test ranges. The analysis is motivated by creating multi-purpose test chambers having the advantages of both, near field systems and compact test ranges. This contribution comprises the discussion of near field scanners at several positions inside a typical compact test range. A ray tracing analysis is presented taking these positions into account in the assessment of near field errors due to multi-path reflections. It is presented how reflections from the absorbers and reflectors are differently impacting near field measurements of low, medium and high gain antennas. The impact is quantified in terms of error levels used in common near field error budgets. It is shown that the combined approach is realizable for specific configurations only.
ABSTRACT A modern method is presented to measure small antenna efficiencies by implementing a variant of the WheelerCap method. As antennas become smaller in size, antenna efficiency typically decreases either because of matching functions or reductions in the antenna radiation resistance. It is important to know while designing small antennas, how much efficiency reduction can be tolerated before a particular design needs changing. Measuring antenna efficiency by integrating radiation functions is not a trivial task and prone to measurement errors. The modern method presented uses a plastic sphere which is internally coated with a highly conductive metallic paint, having low resistivity values (less than 0.1ohms per square), and useful to implement the Wheeler Cap measurements. The measurements are conducted with the use of a modern Agilent Vector Network Analyzer (VNA) which is calibrated to the antenna port (which includes any antenna matching networks). The apparatus was used to measure small “planar” antennas thus producing extremely good results. The paper presents the methodology used for the development of the apparatus and the measurement results.
A novel approach using Artificial Neural network (ANN) is proposed to identify the faulty elements present in a nonuniform linear array. The input to the neural network is amplitude of radiation pattern and output of neural network is the location of faulty elements. In this work, ANN is implemented with three algorithms; feed forward back propagation neural network, Radial Basis Function neural network (RBF) and Probabilistic neural network and their performance is compared. The network is trained with some of the possible faulty radiation patterns and tested with various measurement errors. It is proved that the method gives a high success rate.
The radiation characteristics of antennas can be deter-mined by measuring amplitude and phase data in the ra-diating near-field followed by a transformation to the far-field. Accurate phase measurements especially at high frequencies are very demanding in terms of the required measurement equipment and tolerances. Phaseless mea-surement techniques have been proposed, which often deal with a second set of amplitude only measurement data in order to compensate the lack of phase information. In this paper the concept of phaseless spherical near-field measurements will be addressed by presenting a phaseless near-field transformation algorithm for spherical antenna measurements, working with amplitude only data on two spheres. In particular the measurement of a patch antenna is considered to demonstrate the utility of the technique for low gain antennas. To address the issue of robustness, inaccurate measurement distances as well as spherical rotation angles are considered in order to evaluate the accuracy of the method against probe positioning errors. Furthermore noise contributions are introduced to emu-late measurement inaccuracies in general.
ABSTRACT In this work, a probe compensated near-field – far-field transformation technique with spherical spiral scanning suitable to deal with elongated antennas is developed by properly applying the unified theory of spiral scans for nonspherical antennas. A very flexible source modelling, formed by a cylinder ended in two half-spheres, is considered as surface enclosing the antenna under test. It is so possible to obtain a remarkable reduction of the number of data to be acquired, thus significantly reducing the required measurement time. Some numerical tests, assessing the accuracy of the technique and its stability with respect to random errors affecting the data, are reported.
The NIST 18 Term Error Analysis Technique uses a combination of mathematical analysis, computer simulation and near-field measurements to estimate the uncertainty for near-field range results on a given antenna and frequency range. A subset of these error terms is considered for alignment accuracy of an antenna’s RF main beam. Of the 18 terms, several have no applicable influence on determining the beam pointing or the terms have a minor effect and when an RSS estimate is performed they are rendered inconsequential. The remainder become the dominant terms for identifying the alignment accuracy. There are six terms that can be evaluated to determine the main beam pointing uncertainty of an antenna with respect to dual band performance. Analysis of the near-field measurements is performed to identify the alignment uncertainty of the main beam with respect to a specified mechanical position as well as to the main beam of the second band.
The Mathematical Absorber Reflection Suppression (MARS) technique is a method to reduce scattering errors in near-field and far-field antenna measurement systems. Previous tests by the authors had indicated that NSI's MARS technique was not as effective for directive antennas. A recent development of a scattering reduction technique for cylindrical near-field measurements has demonstrated that it can also work well for directive antennas. These measurements showed that the AUT shouldbeoffsetfromtheorigin byadistanceatleastequal to the largest dimension of the AUT rather than only 1-3 wavelengthswhich hadbeenusedfor smallerantennasin the earlier MARS measurements. Spherical near-field measurementshaverecently beenconcludedwhich confirm that with the larger offsets, the MARS technique can be applied to directive antennaswith excellent results. The MARS processing has recently been modified to produce significantly improved results. This improvement isespeciallyusefulfor antennaswherethephasecenterof the horns is located inside the horn and varies with frequency like pyramidal Standard Gain Horns (SGH). Fewermodesarerequired for thetranslatedpatternandthe filtering is more effective at reducing the effect of scattering. The improvement is very apparent for pyramidal horns.
The Georgia Tech Research Institute (GTRI) performed modeling and measurements for the problem of using multiple antennas on a platform with crew locations. GTRI personnel analyzed the electromagnetic compatibility/interference between multiple antenna systems on the platform by modeling the electromagnetic (EM) fields with Method-of-Moments (MOM) and Physical Optics/Uniform Theory of Diffraction (PO/UTD) modeling methodologies. Power densities were generated with the model at various crew locations on the platform and compared with the appropriate radiation hazard standard. Following the modeling effort, power density measurements were performed on the multiple antennas at various crew locations. The measured results were compared with the modeled results and the radiation hazard standard, and samples of both results are presented. In cases where measured and modeled data results do not agree to within the measured data error budget, the model and modeled data results were re-analyzed for errors. Updated modeled data results were generated and compared with measured data results, with the updated results presented.
Prior to modern computer-aided measurement techniques all measurements were made with analog procedures that required the personal attention of measurement professionals. Modern techniques rely on careful set-up involving standards, and the test equipment applies error correction based on these standards. How-ever, there is an over-reliance on computer-based measurement equipment to do all of the thinking, leading to inappropriate use of these techniques, in turn leading to large, unsuspected measurement errors. This paper analyzes a situation wherein a network analyzer was used to isolate radome insertion loss from a combination measurement. The procedure used led to gross errors indicating gain in a passive device. The source of these errors is identified as an incorrect referencing procedure used to isolate radome characteristics from the combined antenna-radome characteristics. This error is common and applies to an entire class of measurement problems.
Phased arrays antennas have desirable features in terms of simplicity, compact dimensions and low weight for low frequency applications requiring dual polarization and medium gain such as RCS measurements. However, a fundamental problem with phased arrays technology in wide band applications is grating lobe limitations due to the grid topology of the phased array elements. The spacing of the array elements cannot be to close in order to limit element coupling and not to large to avoid grating lobes. Consequently, conventional phase array antenna applications are generally limited to a useable frequency bandwidth of 1:2. A unique grid topology has recently been developed to overcome this problem [1, 2]. By interleaving three separate phased arrays, each dedicated to a different subband with close to 1:2 bandwidth, the useable bandwidth of the combined phased array antenna can be extended to as much as 1:7 while maintaining the nice performance features of the basic phase array technology. Based on this technology a large dual polarized phase array antenna has been designed for indoor RCS testing in the frequency range from 140MHz to 1000MHz. The operational bandwidth of the array is split into three subbands: 140-260 MHz, 260-520 MHz and 520-1000 MHz. The array is 6.34 x 6m and weighs less than 250Kg. Due to the element spacing and topology the phased array is sensitive to excitation errors so the beam forming network (BFN) feeding the elements must be wellbalanced. A uniform amplitude and phase distribution for the array excitation coefficients has been selected to simplify the BFN design and minimize possible excitation errors throughout the bandwidth. This paper describe the antenna electrical design and performance trade-off activity, the manufacturing details and discuss the comprehensive validation/testing activity prior to delivery to the final customer.
An effective near-field – far-field transformation technique with spherical spiral scanning tailored for antennas having two dimensions very different from the third one is here proposed. To this end, an antenna with one or two predominant dimensions (as, e.g., an elongated or quasi-planar antenna) is no longer considered as enclosed in a sphere, but in a prolate or oblate ellipsoid, respectively, thus allowing one to remarkably reduce the number of required data. Moreover these source modellings remain quite general and contain the spherical one as particular case. Numerical tests are reported for demonstrating the accuracy of the far-field reconstruction process and its stability with respect to random errors affecting the data.
This paper describes an application of well known microwave holography to the practical case of the space antenna for the European Navigation System GALILEO. The antenna consists in an array of 45 patch elements, divided into six sectors, fed by a two level beam forming network. In fact, the procedure described in this paper has been used in the frame of the development of the GALILEO Navigation antenna to identify element feeding errors. A planar hologram on the aperture plane of the array has been obtained by a set of spherical near field measurements. Sampling the resulting aperture field distribution (in amplitude and phase) allowed reconstructing the excitation law and identifying errors. The developed procedure was validated with a number of test cases assessing numerical errors introduced by the process. Applying the back-projection to the measured far-field led to discover that some sectors of the array were overfed and that errors were present in the central power divider responsible of the first power distribution in the antenna. A new power divider was then manufactured and integrated into the array leading to a well performing antenna.
In this paper a method to identify faulty elements in a planar array using Artificial Neural Networks (ANN) is presented. The input to the neural network is amplitude of deviation pattern and output of neural network is the location of faulty elements. A planar array of 5×5 number of isotropic elements with uniform excitation and spacing ?/2 is considered. Either one faulty element or two faulty elements can exist in the array. The network is trained with some of the possible faulty deviation patterns and tested with various measurement errors. ANN is implemented with Radial Basis Function neural network (RBF) and Probabilistic neural network and their performance is compared.
Measurement of radome beam deflection and/or Boresight shift in a compact range generally requires a complicated set of positioner axes. One set of axes usually moves the radome about its system antenna while the system antenna remains aligned close to the range axis. Another set of axes is normally required to scan the system antenna through its main beam (or track the monopulse null) in each plane so the beam pointing angle can be determined. The fidelity required for the beam pointing angle, combined with the limited space inside the radome, usually make this antenna positioner difficult and expensive to build. With a far-field range, a common approach to the measurement of beam deflection or Boresight shift uses a down-range X-Y scanner under the range antenna. By translating the range antenna, the incident field's angle of arrival is changed slightly. Because the X-Y position errors are approximately divided by the range length to yield errors in angle of arrival, the fidelity required of the X-Y scanner is not nearly as difficult to achieve as that of a gimbal positioner for the system antenna. This paper discusses a compact-range positioner geometry that approximates the simplicity of the down-range-scanner approach commonly used on far-field radome ranges. The compact-range feed is mounted on a small X-Y scanner so that the feed aperture moves in a plane containing the reflector's focal point. Translation in this 'focal plane' has an effect very similar to the X-Y translation on a far-field range, altering the direction of arrival of the incident plane wave. Measured and modeled data are both presented.
The measurement accuracy of state-of-the-art RF test facilities like near-field or compact test ranges is influenced due to applied system hardware as well as operational facts which are influenced by human errors. The measurement errors of near-field test facilities were analyzed and published in the past times and are based on the 18-term error model of Newell [1]. For compact test ranges and especially for the cross-polar free compensated compact range a similar error model was established at Astrium GmbH within a study for the satellite service provider INTELSAT [2] in order to define possible facility performance improvements and maximum achievable values for the measurement accuracy. It has to be remarked, that test programs for space applications require very stringent adherence to procedures and documentation of process steps during a test campaign. Within this paper, recommendations for process optimizations and procedures will be presented to guarantee the adherence to the valid error budgets and to minimize the Human Factor. A description of main error contributions in the Compensated Compact Range (CCR) of Astrium GmbH will be performed. Furthermore, the error budgets for pattern and gain measurements and achievable performance improvements will be given.
In this paper, three possible approaches for definition of a highly accurate reference pattern of a reference antenna are described and their pros and contras are discussed. Following the most reliable approach, a dedicated measurement campaign was planned and carried out in 2007-2008 for definition of the highly accurate reference pattern of the VAST12 antenna. In planning the campaign, conclusions from the first comparison campaign with the VAST12 carried out within the ACE network in 2004-2005 were taken into account and these are also presented and discussed. Some typical measurement errors and uncertainties are listed and briefly discussed.