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.)
Broad band, dual polarized probes are becoming increasingly popular options for use in near-field antenna measurements. These probes allow one to reduce cost and setup time by replacing several narrowband probes like open-ended waveguides (OEWG) with a single device covering multiple waveguide bands. These probes are also ideal for production environments, where chamber throughput should be maximized. Unfortunately, these broadband probes have some disadvantages that must be quantified and corrected for in order to make them viable for high accuracy near-field measurements. Most of these broadband probes do not have low cross polarization levels across their full operating bandwidths and may also have undesirable artifacts in the main component of their patterns at some frequencies. Both of these factors will result in measurement errors when used as probes. Furthermore, the use of a dual port RF switch adds an additional level of uncertainty in the form of port-to-port channel balance errors that must be accounted for. This paper will describe procedures to calibrate the pattern and polarization properties of broad band, dual polarized probes with an emphasis on a newly developed polarization correction algorithm. A simple procedure to measure and correct for amplitude and phase imbalance entering the two ports of the near-field probe will also be presented. Measured results of the three calibration procedures (pattern, polarization, channel balance) will be presented for a dual-polarized, broad band quad-ridged horn antenna. Once calibrated, this probe was used to measure a standard gain horn (SGH) and will be compared to baseline measurements acquired using a good polarization standard open-ended waveguide (OEWG). Results with and without the various calibration algorithms will illustrate the advantage to using all three routines to yield high accuracy far-field pattern data.
The US NEXRAD weather surveillance Doppler radar (WSR-88D) was recently upgraded to polarimetric capability. This upgrade permits identification of precipitation characteristics and type, thus providing the potential to significantly enhance the accuracy of radar estimated rainfall, or water equivalent in the case of frozen hydrometeors. However, optimal benefits are only achieved if errors induced by the radar hardware are properly accounted for through calibration. Hardware calibration is a critical element in delivering accurate meteorological information to the forecast and warning community. The calibration process must precisely measure the gain of the antenna, the Polarimetric bias of the antenna, and the overall gain and bias of the receive path. The absolute power measurement must be accurate to within 1 dB and the bias between the Polarimetric channels must be known to within 0.1 dB. These requirements drive a need for precise measurement of antenna characteristics. Engineers and scientists with the NEXRAD program employ solar scanning techniques to ascertain the absolute gain and bias of the 8.53 m parabolic center fed reflector antenna enclosed within a radome. They are also implementing use of daily serendipitous interference strobes from the sun to monitor system calibration. The sun is also used to adjust antenna gain and pedestal pointing accuracy. This paper reviews the methods in place and under development and identifies some of the challenges in achieving the necessary calibration accuracies.
If a planar near-field (PNF) scanner is large and there is insufficient temperature regulation in the chamber to keep ordinary thermal expansion/contraction from causing unacceptable position errors, then consideration must be given to compensation techniques that can adjust for the changes. Thermal expansion/contraction will affect almost everything in the chamber including the floor, the scanner structure, the encoder or position tapes, the AUT support, and the mount for any extra instrument(s) used to measure and correct for position error. Since the temperature will generally cycle several times during a lengthy acquisition, error-correction solutions must account for the dynamic nature of the temperature effects. This paper describes a new automated tracking-laser compensation subsystem that has been designed and developed for very large horizontal PNF systems. The subsystem is active during the acquisition to account for both static and dynamic errors and compensates for those errors in all three dimensions. The compensation involves both open-loop corrections for repeatable errors with high spatial frequency and closed-loop corrections for dynamic errors with low spatial frequency. To close the loop, laser data are measured at a user-defined interval between scans and each scan that follows the laser measurements is fully compensated. The laser measurements are fully automated with no user interaction required during the acquisition. The challenges, goals, and assumptions for this development are listed, high-level implementation considerations are provided, and resulting measured data are presented.
In this work, we propose a new method for measuring the gain of a circularly polarized antenna at millimeter wave (mm-wave) and terahertz (THz) frequencies. Traditional methods of CP-gain measurement require measurement of linearly polarized gain in two orthogonal planes. However, for mm-Wave frequencies, this method is not practical since rotation of antenna under test can cause errors in measured phase. To avoid this, we present a novel phase-less method. For validation, circularly polarized slot array antenna under test is measured at 106 GHz. The antenna design and fabrications process is also presented.
Abstract— Antenna arrays works at their peak performance when they are well calibrated at the factory. Once they are employed in a real environment, they might be subject to unpredictable disturbances. That’s why recalibration after operational deployment is required but is usually not done due to practical difficulties. In some applications such as Direction Finding (DF), direction of arrival estimation is susceptible to the antenna model errors. However, the evolution of Direction finding antenna, as the strong integration of an antenna array mounted on a vehicle and the use of more efficient antennas tend to increase this type of disturbances. This paper proposes to evaluate the benefit of an in-situ measurement system for detecting and compensating the disturbance of antenna radiation. The influence of permanent scatters on one hand and variables (open door…) on the other hand in the vicinity of antenna array is investigated. We present a quantitative study of a biased calibration using a model combining 3D electromagnetic simulation, a complete receiver model and a MUSIC direction of arrival algorithm characterization. Two antennas arrays with same height are compared: a standard dipole array and an electrically small UWB antenna array.
This paper will describe newly developed mechanical and electrical alignment techniques for use with plane-polar near-field test systems. A simulation of common plane-polar alignment errors will illustrate, and quantify, the alignment accuracy tolerances required to yield high quality far-field data, as well as bounding the impact of highly repeatable systematic alignment errors. The new plane-polar electrical alignment technique comprises an adaptation of the existing, widely used, spherical near-field electrical alignment procedure [8] and can be used on small, and large, plane-polar near-field antenna test systems.
Compact range measurement facilities have been used successfully for many years to characterize antenna performance as well as radar signature. This paper investigates strategies for improving compact range measurement accuracy by mitigating errors associated with ground reflections inherent in most range designs. A methodology is developed for strategically modifying, or patterning, the surface between the range source antenna and the reflector to reduce error terms, thereby increasing measurement accuracy. Candidate patterns were evaluated using a full-wave computational finite-difference time-domain (FDTD) model at VHF/UHF frequencies to determine baseline performance and develop trade rules for more advanced designs. Physical optics (PO) models were used to analyze the final design at the frequencies of interest.
ABSTRACT When the mechanical requirements are established for a spherical near-field scanner, it is desirable to estimate what effects the expected mechanical errors will have on the determination of the far field of potential antennas that will be measured on the proposed range. The National Institute of Standards and Technology (NIST) has investigated the effects of mechanical errors for a proposed outdoor spherical near-field range to be located at Ft. Huachuca, AZ. This investigation was performed by use of theoretical far-field patterns and introducing position errors into simulated spherical near-field measurements using software developed at NIST. Periodic and random radial and angular position errors were investigated. Far-field patterns were then calculated with and without probe-position correction to determine the effects of mechanical position errors. Periodic errors were found to have a larger effect than random errors. This paper reports the results of these investigations.
Numerous applications for antenna, radome and RCS measurements require a very accurate positioning capability to properly characterize the product being tested. Testing of weapons (missiles), guidance systems, and satellites, among other applications, require multi-axis position accuracies of a few thousandths of an inch or degree. For global positioning, spherical error volumes can be extremely small having diameters of .002 inches to .005 inches. This paper addresses the issues that must be resolved when highly accurate mechanical positioning is required. Many factors such as thermal stability, axis configuration, bearing runout and mechanical alignment can adversely affect the overall system accuracy. Additionally, when examined from a global positioning system perspective, the accuracy of the entire system is further degraded as the number of axes increases. Successful system implementation requires carefully examining and addressing the most dominant error factors. The paper will cover current tools and techniques available to characterize and correct the contributing errors in order to achieve the highest possible system level accuracy. A recently delivered 4 ft radius SNF arch scanner, which achieved ± .0043° global positioning accuracy, will provide insight into these methods and show how the dominant factors were addressed.
Some antenna measurement applications require the precise positioning of an antenna along a prescribed path which may be realized by a combination of several, independent physical axes. Coordinated motion allows for emulation of a more complex and/or precise positioning system by utilizing axes which are mechanically less complex or precise and are correspondingly more easily realizable. An ideal coordinated motion system should 1) Allow for the description of coordinated paths as parametric mathematical functions and/or interpolated look-up tables 2) Support control variable parameters which affect the trajectory 3) Compute a feasible trajectory within given kinematical constraints 4) Generate measurement trigger signals along the trajectory 5) Minimize control-induced vibration 6) Compensate for multivariate positioning errors. This paper will describe a novel approach to virtual-axis coordinated motion which offers significant improvements over existing motion control systems. This advancement can be applied to many antenna measurement problems such as Helicoidal Near-Field Scanning and Radome Characterization.
Antenna measurement systems employ mechanical positioners to spatially orient antennas, vehicles, and a variety of other test articles. These mechanical devices exhibit native positioning accuracy in varying degrees based on their design and position feedback technology. Even the most precise positioning systems have insufficient native accuracy for some specific applications. As the limits of economical positioning accuracy are approached, a new error correction technique developed by MI Technologies satisfies these higher accuracy requirements without resorting to extreme measures in positioner design. The new technique allows real-time correction of repeatable positioning errors. This is accomplished by (1) performing a finely grained measurement of positioner accuracy, (2) creating a map of the errors in both spatial and spatial frequency domains, (3) separating the errors into their various components, and (4) applying correction filters to algorithmically perform error correction within the positioner control system. The technique may be used to achieve extreme positioning accuracy with positioners of high native accuracy. It may also be applied to conventional (synchro feedback) positioners to achieve impressive results with no modifications at all to the positioner. The following paper discusses the new error correction technique in detail.
The improved calibration model proposed in this paper is based on the traditional capacitance model which suffers from errors caused by the assumption that the capacitance is independent of frequency and the permittivity of the ambient medium under test. By analyzing the near-zone field of the coaxial opening, we introduce the new near-field capacitance to account for the dependency on the external permittivity. Simulation results show that the calibration error is significant reduced for low and moderate loss medium. And the calibration of the unknown coefficients simply requires the premeasurement of three known material including air, which provides convenience for the real field measurement. Measurement results obtained by a novel wideband in-situ coaxial probe are included to prove the accuracy improvement improved calibration model. by using this
A desired feature of modern field probes is that the useable bandwidth should exceed that of the Antenna Under Test (AUT) [1]. Recent developments in probe and orthomode junctions (OMJ) technology has shown that bandwidths of up to 4:1 are achievable [2-5]. The probes are based on inverted ridge technology capable of maintaining the same high performance standards of traditional probes However, in typical Spherical Near Field (SNF) measurement scenarios, the applicable frequency range of the single probe can also be limited by the content of µ.1 spherical modes in the probe pattern [6-7]. This is because the traditional NFtoFF software applies probe correction under the assumption that the probe pattern is fully specified from knowledge of the E-and H-plane patterns only [8]. While this condition is guaranteed for virtually any type of probe for small illumination angles of the AUT and/or a long probe/AUT distance this assumption may lead to unacceptable errors in special cases. This paper describes the design and experimental verification of a Ka-band probe based on the inverted ridge technology. The probe is intended for high precision SNF measurements in special conditions that require less than -45dB higher order spherical mode content. This performance level has been accomplished through careful design of the probe and meticulous selection of the components used in the external balanced feeding scheme. The paper reports on the electrical and mechanical design considerations and the experimental verification of the modal content.
There are a number of near-field measurement situations where it is desirable to use a broad band probe to avoid the need to change the probe a number of times during a measurement. But most of the broad band probes do not have low cross polarization patterns over their full operating frequency range and this can cause large uncertainties in the AUT results. Calibration of the probe and the use of probe pattern data to perform probe correction can in principle reduce the uncertainties. This paper reports on a series of measurements that have been performed to demonstrate and quantify the cross polarization levels and associated uncertainties that can be measured with typical log periodic (LP) probes. Two different log periodic antennas were calibrated on a spherical near-field range using open ended waveguides (OEWG) as probes. Since the OEWG has an on-axis cross polarization that is typically at least 50 dB below the main component, and efforts were made to reduce measurement errors, the LP calibration should be very accurate. After the calibration, a series of standard gain horns (SGH) that covered the operating band of the LP probe were then installed on the spherical near-field range in the AUT position and measurements were made using both the LP probes and the OEWG in the probe position. The cross polarization results from measurements using the OEWG probes where then used as the standard to evaluate the results using the LP probes. Principal plane patterns, axial ratio and tilt angles across the full frequency range were compared to establish estimates of uncertainties. Examples of these results over frequency ranges from 300 MHz to 12 GHz will be presented.
The numerical analysis used for efficient processing of spherical near-field data requires that the far-field pattern of the probe can be expressed using only azimuthal modes with indices of µ = ±1. (1) If the probe satisfies this symmetry requirement, near-field data is only required for the two angles of probe rotation about its axis of . = 0 and 90 degrees and numerical integration in . is not required. This reduces both measurement and computation time and so it is desirable to use probes that will satisfy the µ = ±1 criteria. Circularly symmetric probes can be constructed that reduce the higher order modes to very low levels and for probes like open ended rectangular waveguides (OEWG) the effect of the higher order modes can be reduced by using a measurement radius that reduces the subtended angle of the AUT. Some analysis and simulation have been done to estimate the effect of using a probe with the higher order modes (2) – (6) and the following study is another effort to develop guidelines for the properties of the probe and the measurement radius that will reduce the effect of higher order modes to minimal levels. This study is based on the observation that since the higher order probe azimuthal modes are directly related to the probe properties for rotation about its axis, the near-field data that should be most sensitive to these modes is a near-field polarization measurement. This measurement is taken with the probe at a fixed (x,y,z) or (.,f,r) position and the probe is rotated about its axis by the angle .. The amplitude and phase received by the probe is measured as a function of the . rotation angle. A direct measurement using different probes would be desirable, but since the effect of the higher order modes is very small, other measurement errors would likely obscure the desired information. This study uses the plane-wave transmission equation (7) to calculate the received signal for an AUT/probe combination where the probe is at any specified position and orientation in the near-field. The plane wave spectrum for both the AUT and the probe are derived from measured planar or spherical near-field data. The plane wave spectrum for the AUT is the same for all calculations and the receiving spectrum for the probe at each . orientation is determined from the far-field pattern of the probe after it has been rotated by the angle .. The far-field pattern of the probe as derived from spherical near-field measurements can be filtered to include or exclude the higher order spherical modes, and the near-field polarization data can therefore be calculated to show the sensitivity to these higher order modes. This approach focuses on the effect of the higher order spherical modes and completely excludes the effect of measurement errors. The results of these calculations for different AUT/probe/measurement radius combinations will be shown.
Measurement of E and H plane far field patterns for an open-ended rectangular waveguide in the free air operating between the frequencies of 16 and 19 GHz are shown and compared with the simulated patterns derived by several authors. Although the theoretical expressions give a broader pattern for the E-plane than for the H-plane, which is observed, measurements exhibit a sharper decay in the E-plane than the one obtained by simulation. In this work, we calculate the errors associated with the use of the different models that fail to correctly approximate the E-plane. Finally, we introduce a parameter in the best model to adjust the effective aperture dimensions in order to obtain a more realistic representation of the measured far field.
A Plane Wave Synthesis Approach for mitigating errors in antenna measurements caused by stray signals and imperfections in the measurement range illuminating fields has been demonstrated previously for a 2D scalar source [1]. This paper presents algorithms developed for the Range Transfer Function (RTF) method for a 2D vector source. Vector basis functions for both the field representation and the AUT representation are implemented to provide a robust numerical solution. The new algorithms are more stable because the plane wave angles and the antenna measurement angles may be completely general, provided that Nyquist rules of sampling are observed during both the field probing (to obtain the plane wave coefficients) and the antenna measurement (to obtain the raw pattern data).
Far-field uncertainty due to probe errors in planar near-field measurements is analyzed for the fast irregular antenna field transformation algorithm. Results are compared with the classical technique employing two dimensional Fast Fourier Transform (2D FFT). Errors involving probe's relative pattern, alignment, transverse and longitudinal position, interaction with AUT etc. have been considered for planar measurements. The multiple reflections error originating from the interaction of the probe and the AUT tends to deteriorate the radiation pattern to a greater extent. Therefore, a novel technique which utilizes near-field measurements on two partial planes is presented to reduce the multiple reflections between the probe and the AUT.
This paper describes the behavior of the antenna radiation pattern for different planar near-field measurement errors superimposed on the near-field data. The disturbed radiating near field is processed using multilevel plane wave based near-field far-field transformation to determine the far-field. Errors like scan area truncation, transverse and longitudinal position inaccuracy of measurement points, and irregular sample spacing are analyzed for an electrically large parabolic reflector at 40 GHz. The error behavior is then compared with the standard transformation technique employing 2D Fast Fourier Transform (FFT) using the same near-field data. In order to exclude the effect of any other measurement or environmental error, electric dipoles with appropriate magnitude profile and geometrical arrangement are used to model the test antenna.
Large phased arrays have stringent beam-pointing accuracy requirements over their scan volume. Measuring the beam-pointing accuracy of a phased array with a planar near-field scanner is convenient but can lead to erroneous results if the near-field sampling grid is not well controlled. This paper describes numerical experiments that were carried out to assess the impact of various types of grid errors on the measurement of beam-pointing accuracy. The types of grid errors considered include skewing and curvature in the plane of the grid. The numerical experiments use infinitesimal dipoles as the radiating elements and assume an ideal probe. It is shown that beam-pointing errors induced by grid errors that can be described by an affine transformation can be estimated in closed form. For more complicated grid errors, the model is shown to be a useful tool in estimating their impact on measuring beam-pointing error. Finally, the amount of over-scan required for accurate beam-pointing measurements over a large scan volume is examined.