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.)
A novel probe design for measuring complex permittivity of soils in-situ from 10 to 1000 MHz without taking soil samples is presented. The dielectric constant and conductivity of soil is derived from step-frequency reflection taken inside a small freshly bored hole. As a result, permittivity at various depths with in-situ moisture content and soil texture can be obtained in the fields. A novel calibration method was developed to account for the frequency- and material-dependent geometrical factor which causes bias errors in conventional calibration methods. Experimental measurement results and simulation results are used to prove the efficiency and accuracy of this method.
Far field measurements of ground vehicle antennas in anechoic chambers often require the creation of a plane wave by near field hemispherical probing with associated mathematical transformations to the far field/plane wave result. Direct far field measurements can be done to save time when the frequency is low enough. This paper discusses a method of extending the frequency band where direct measurements can be done by synthesizing a plane wave using a small array of antennas. The use of an array to create a plane wave in an anechoic chamber usually results in errors due to the reflections from the walls of the chamber. The technique to be described in this paper is to model the wall reflections and the array antenna characteristics and to use optimization techniques to derive an antenna placement and power distribution scheme to optimize the plane wave. Several optimization techniques will be described and results from testing in a 1.2 meter long sub-scale chamber model will be shown. Improvements in the far field measurements will be discussed.
Highly accurate spherical near-field measurement systems require precise alignment of the probe antenna to the measurement surface. MI Technologies has designed and constructed a new spherical near field arch positioner with a 1.5 meter radius to support measurements requiring accurate knowledge of the probe phase center to within .0064 cm throughout its range of travel. To achieve this level of accuracy, several key design elements were considered. First, a highly robust mechanical design was considered and implemented. Second, a tracking laser interferometer system was included in the system for characterization of residual errors in the position of the probe. Third, a position control system was implemented that would automatically correct for the residual errors. The scanner includes a two position automated probe changer for automated measurements of multi-band antennas and a high accuracy azimuth axis. The azimuth axis includes an algorithm for correcting residual, repeatable positioning errors. This paper defines the spherical near-field system and relation of each axis to the global coordinate system, discusses their associated error sources and the effect on global positioning and presents achieved highly accurate results.
New developments in ortho-mode junctions (OMJ) and probe technologies has enabled near field probes with up to 1:4 bandwidth, while maintaining the high performance standards of traditional narrow band probes [1–3]. The new probe technologies are based on inverted ridge structures providing four symmetrical feeding points for external balanced feeding. The inverted ridge structure stabilizes the frequency dependence of the OMJ while the external balanced feeding is a crucial feature to achieve the desired high performances. This paper briefly review the theory of balanced feeding and derive performance guide lines on the external beam forming network for achieving high port-to-port isolation and matching on a wide bandwidth with the inverted ridge probe technology. The relationship between excitation errors in the balanced feeding scheme and the spherical mode index µ.1 content of the probe is also investigated and upper bounds on acceptable excitation errors are derived.
G/T is a standard figure of merit describing the sensitivity of receive, active, microwave antenna systems. It is therefore critical to have an accurate test technique for measuring the G/T of systems with varying levels of sensitivity. The Yfactor method is a commonly used and wellunderstood technique for characterizing receive microwave systems. While the concepts behind the Yfactor measurement techniques are generally understood, little quantitative work has been done establishing limits and rules regarding selection of measurement hardware and design of the measurement system. This paper seeks to provide examples of test systems and quantify the error induced into the measurement due to mismatched standard hardware. The variables and techniques involved in a G/T test are examined using simulated results of measurement systems performed in VSS (Visual System Simulator).
In this paper, a beam-steering computer design is explored for a large space-fed phased-array antenna. GTRI previously developed a beam-steering computer for a smaller phased-array antenna which accomplished spherical propagation focusing and multiple phase-only beam-broadening modes. In a subsequent effort, the beam-steering computer design was scaled for a large phased-array antenna to accomplish similar tasks. To verify the design, a series of far-field measurements was initiated to characterize the performance of the antenna by comparing with past measured near-field data and modeled results. One of the primary responsibilities of the beam-steering computer was the focusing of the spherical propagation wave front. A measurement technique is discussed to accomplish this focusing for the large space-fed phased-array antenna by correcting measurement errors in the spherical propagation routine of the beam-steering computer. Additional patterns were taken using the updated feed horn focal point for spherical propagation correction. By correcting the phase errors caused by spherical propagation defocusing in the original beam-steering computer, significantly better antenna performance was obtained, including higher peak gain, reduced nearby sidelobe levels, and removal of beam-pointing errors. Another important responsibility of the beam-steering computer was the ability to realize multiple antenna modes, including a focused pencil beam as well as defocused broadened-beam modes. A stochastic gradient descent algorithm was utilized to obtain several phase tapers to accomplish beam-broadening for the antenna modes. These modes were implemented in the beam-steering computer and tested on a far-field range. The antenna patterns were compared with modeled results and with previous measured data to ensure validity of the implementation.
Calibration of probes for planer near field range measurements is generally required to obtain accurate cross-polarization (xpol) data; however, probe calibration is costly and time consuming. Using analytical models in place of calibration is generally much more cost effective, but may result in larger measurement errors. In a previous paper [1], we showed that simple models of copol probe patterns with zero xpol can give accurate measured results, provided that the probe xpol is much better, generally 10-15 dB better, than the Antenna Under Test (AUT). The next question is “Can a lower performing (and cheaper) probe be used if both the copol and xpol probe patterns are modeled?” In this paper, we compute AUT xpol measurement errors that result from probe xpol errors, and we compare far field AUT patterns processed using probe models with patterns processed with calibrated probe files.
This paper elaborates on certain aspects of a new measurement process that permits an assessment of spherical near-field (SNF) measurement errors based on a set of practical tests that can be done as part of any SNF measurement. It provides error bars for a measured radiation pattern in an automated fashion.
– far-field transformation with cylindrical scanning are efficiently determined by using an optimal sampling interpolation algorithm. The comparison of the far-field patterns reconstructed from the acquired irregularly distributed measurements with those obtained from the data directly measured on the classical cylindrical grid assesses the effectiveness of the approach.
A stand-alone commercial program, performing advanced electromagnetic processing of measured data, is being developed by TICRA. The program reads the measured field and computes the extreme near field or the currents on the antenna surface. From the inspection of the extreme near field or currents, the program will solve typical antenna diagnostics problems, such as identification of array element failure and antenna surface errors, but also allow artificial removal of undesired contributions, such as currents on cables and fixtures, thereby saving valuable time and resources in the antenna design and validation process. The program will be based on two field reconstruction techniques, the SWE-PWE presented at AMTA in 2007, and a new and more accurate inverse higher-order Method of Moments (INV-MoM). The paper will illustrate the theory behind the two techniques and present numerical cases with simulated data.
The calibration and measurement of bistatic reflectivity at short range (3.3 m) presents challenges that are significantly different from the usual test range measurements (typically monostatic at 100 to 150 m). In order to overcome this an object-free calibration procedure has been applied, eliminating crosstalk, reducing other interferences and removing errors associated with the RCS and alignment of calibration objects. It is based on calibrating the transmitter and receiver antennas as a pair by directing the antennas toward each other. The method thus requires that the antennas can be separated. Furthermore the signal level needs to be handled e.g. by the separation distance or attenuators. The bistatic reflectivity measurements were performed by placing a clutter sample on a turntable which is located at the centre of a bistatic arc. This configuration enables us to do ISAR images. Background contributions were discriminated using a combination of synthetic resolution and zero-doppler filtration. The sensitivity variation across the antenna footprint was handled by calculating an equivalent area from measured off-axis antenna sensitivities. Reflectivities have been measured for a metallic test surface and for grass. The metallic test surface had been manufactured to correspond to typical theoretical bistatic clutter models.
A widely used time-gating technique can be effectively implemented in near-field (NF) antenna measurements to significantly improve the measurement accuracy. In particular, it can be implemented to reduce or remove the effects of the following measurement errors [1]: -multiple environmental reflections and leakage in outdoor or indoor NF ranges -edge diffraction effects on measurement accuracy of low gain antennas on a ground plane [3] In addition, reflectivity in the range can be precisely localized, separated and quantified by using the time – gating procedure with only one addition (a subtraction operation) added to the standard near-field to far-field (NF – FF) transformation algorithms. In this paper a step by step procedure is described which includes acquisition of near-field data, transformation of the raw near-field data from the frequency to the time domain, definition of the correct time gate, transformation of the gated time domain data back to the frequency domain, and the transformation of the time gated near-field data to the far-field. The time gated results, as already shown in [2], provides for more accurate far-field patterns. In this paper it is shown how the 3D reflectivity/multiple reflections in the measurement chamber or outdoor range can be determined by subtracting the time gated results from the un-gated data. This technique is illustrated through use of several measurement examples. It is demonstrated that the time gated method has a clear physical explanation, and, in contrast with other techniques [4,5] is less consuming (does not require mechanical AUT precise offset installation, additional measurement and processing time) and allows for a better localization and quantization of the sources of unwanted radiation. Therefore, this technique is a straightforward one and is much easier to implement. The main disadvantage cited by critics regarding use of the time gating technique is the narrow frequency bandwidth used in many NF measurements. However, it is shown, and illustrated by the examples, that the technique can be effectively implemented in NF systems with a standard probe bandwidth of 1.5:1 and an AUT having a bandwidth as low as 5% to 10%.
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.