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.)
Compact antenna test ranges intended for low cross polarization antenna measurements require the use of feeds with polarization ratios typically greater than 40 dB across the included angle of the quiet zone as well as across the frequency band of interest. The design for a series of circular corrugated aperture feeds to meet these requirements is presented. The feeds are based on a circular waveguide OMT covering a full waveguide frequency band with interchangeable corrugated apertures to cover three sub-bands. In order to validate the design of this series of scalar feeds, high accuracy cross-polarization data was collected. The primary limiting factor in the measurement of the polarization ratio was the finite polarization ratio of the source antennas. A technique for correcting for the polarization ratio of the source is presented along with measured data on the feeds. The technique begins with the accurate characterization of the linear polarization ratio of the standard gain horns using a three antenna technique, followed by pattern measurements of the feeds, and ends with the removal of the polarization error due to the source antenna from the measured data. Measured data on these feeds is presented before and after data correction along with data predicted using the CHAMP® moment method software.
Compact Antenna Test Ranges are eminently suitable for obtaining the far-field patterns of various types of antennas provided that the frequency is not too low. Typically, a low frequency limit of 1 or 2 GHz is realizable. There are, however a number of important applications between 500 and 1000 MHz for antenna diameters between 1 and 3 meters. The far field distance R = 2D2/l is just too large for an indoor far field range. It is generally accepted at present that a good solution for an indoor range for these kind of measurements is very difficult to realize. In this paper the low frequency performance of single and dual-reflector Compact Antenna Test Ranges will be investigated. It will be shown that with carefully designed serrations and feeds, excellent antenna measurements can be carried out at frequencies below 1.0 GHz for a large number of applications. For purposes of comparison, low frequency performance of a compact range with so called blended rolled edges will be presented as well.
Proper operation of a planar NFR (near field range) includes probe correction as part of the processing of the measured data to result in accurate far field angle patterns, particularly for low cross polarized patterns. The far field transform of the near field data produces the angular spectrum which is the product of the plane wave transmission coefficient pattern of the AUT (antenna under test) with the plane wave receiving coefficient pattern of the probe. Probe correction consists of dividing the angular spectrum by the complex probe angle pattern resulting in the pure far field pattern of the AUT [1]. For best accuracy of co and cross polarized AUT patterns one needs to use accurately measured probe complex co and cross polarized patterns in probe correction for each NFR test frequency. The most accurate probe measurements are usually obtained from specialized test laboratories. However, if the number of frequencies is large, this may create problems due to cost or schedule. Because of this it is typical to procure probe calibration at only a few frequencies spanning the test band for each AUT even though pattern measurements are needed at several additional frequencies falling between the calibration frequencies. A typical strategy at any given test frequency is to perform probe correction using the nearest-neighbor-frequency probe calibration data. This strategy produces some unknown error in the processed probe corrected far field patterns of the AUT at each non-calibrated frequency. Inthis paper we will show a method for estimating the non-calibrated frequency probe correction error for co and cross polarized patterns with examples.
A bias error in planar near-field measurement data comes from receiver crosstalk or leakage effects [1, 2, 3]. The bias error is a complex constant added to every near-field data sample. After transformation from the near-field to the far-field, the bias error becomes an easily identifiable spike located at the center of k-space. If one is measuring a horn, then the bias error produces a small bump or spike at the center of the far-zone pattern (i.e. at the center of k-space). If one is measuring a highgain antenna with the antenna beam pointed away from the center of k-space, then the bias error causes an erroneous sidelobe spike at the center of k-space. The bias constant is difficult to estimate be cause it may be more than 60 dB below the peak near-field level. Nevertheless, if the effect of the bias error can be seen in the far zone pattern of the test antenna, then it can be estimated and removed from the measured data. An algorithm is presented that is used to estimate the bias constant directly from the near-field data, then the bias constant is simply subtracted from the data. Examples using measured data are used to illustrate how the algorithm works and to show its effectiveness.
Traditional measurement methods assume that very accurate antenna to range alignment of the antenna under test (AUT) is convenient or possible. It has recently been shown that the use of non-rectilinear co-ordinate systems are of particular use for the purpose of correcting antenna to range misalignment. Additionally, this misalignment correction can be used to construct an extended composite measurement plane from a series of mis-aligned scans that themselves can be considered as constituting a polyhedral measurement surface. This paper describes the additional processing that is required to yield corrected near and far field data from an acquisition of a mis-aligned AUT. This technique is then illustrated with example results. The agreement of the corrected results is determined via the application of image classification techniques which correlate antenna patterns in a reduced vector pattern space in terms of their overall global features.
A compact cassegrain antenna has been designed for dual-band satellite communications, operating at 20GHz and 30GHz. The antenna consists of a parabolic reflector, a hyperbolic sub-reflector, and a dual-band choke feed. The cassegrain structure has been optimized considering theoretical and measured feed patterns using different software packages, for maximum antenna efficiency with minimum sidelobe levels for a compact design objective. Experimental studies have been carried out in the near-field chamber of the University of Manitoba. The knowledgenof the near-field is helpful in order to adjust different components of the cassegrain antenna. After adjustment, results in terms of gain and radiation patterns are computed by Fourier transform using near-field data, and compared to the measurements realized in the compact range of the University of Manitoba. Comparisons are also made with the results obtained by simulation.
A near-field antenna range will often utilize a flexible microwave cable assembly as a means to transport the sampled signal from the moving sample antenna to a receiver as part of the measurement system. The performance of that cable directly impacts the quality of the final far-field pattern. It has been observed that the cable had been exhibiting a flex life much shorter than anticipated. Analysis of a failed cable revealed that the problem was the result of non-uniformities in the extruded jacket, which produced sites of high stress. These sites ultimately caused the cable conductors to work harden and fracture. A cable which utiized a woven expanded Polytetrafluoroethylene (ePTFE) fiber as an outer jacket was substituted, resulting in a threefold improvement in flex life to date, with the cable still in operation at this writing.
Two types of cellular handset testing are presented. The first studies models of a cellular handset near the human head. A comparative analysis is done between simulation and measurement of an inexpensive head mockup compared to a more expensive head mockup. Peak gain values have good agreement within about 1 dB. The second type of cellular handset testing is for a PCS band PIFA antenna integrated to a cellular handset. This paper describes the design and experimental study of the radiation patterns of a PCS band (1850-1990MHz) cellular handset with an internal PIFA. The PIFA described in this paper has good gain, impedance matching, and reduced sensitivity to human body interaction. This PIFA is a good cellular internal antenna.
SATIMO has recently installed a spherical near-field antenna measurement system for ALLGON MOBILE COMMUNICATIONS, the market leader in the field of antennas for mobile telephones. This spherical near-field system, as shown in Figure 1, allows for real-time measurements of antennas and will among other be used for the measurements of the radiation characteristics of mobile telephones and satellite terminals in the presence of the human operator. The system consists of a circular of 4m diameter containing 64 dual polarized measurement which are electronically scanned giving a real-time near-field pattern cut over 310° in elevation. A full sphere measurement including near-field to far-field transformation is available in seconds with a single +/- 90° azimuth rotation. The paper will present the measurement system and the results of the final acceptance tests. The acceptance tests are based on both range inter comparisons and also by measurement of key terms in the overall error budget.
NPL has been providing antenna gain standards since the late 1970's, initially to service internal needs for microwave field strength standards. To meet the increasing industrial demand for the calibration of microwave antennas in areas such as satellite communications and radar, NPL has developed an antenna extrapolation range. The current facility, which is due to be replaced by the end of the year, is used to measure the gain of microwave antennas in the frequency range 1 to 60 GHz, often with a gain uncertainty as low as ± 0.04 dB. Axial ratio, tilt, sense of polarisation and pattern measurements can also be made in the same facility, while for larger antennas a planar near-field scanner is used. Of the many measurement techniques for determining the gain of an antenna, the most accurate is the three antenna extrapolation technique [1,2] which was developed at the National Institute of Standards and Technology (NIST) at Boulder, Colorado, and is the method used at NPL. This is an absolute method as it does not require a prior knowledge of the gain of any of the antennas used. Since calibration data is often required across a wide frequency band, the measurement techniques and software have been developed to allow measurements to be performed at a large number of frequencies simultaneously. This reduces the turn round time, the cost and the need for interpolation between measurement points.
The European Space Agency (ESA) began serious investigations into the implementation and exploitation of near field antenna testing techniques already in the early 1970s where all three near field measurement geometries were considered (1). Spherical near field scanning was selected by ESA as being the most promising alternative to even larger conventional outdoor ranges. In the meantime, work was underway at the Technical University of Denmark (TUD) on spherical wave theory and its application to near field antenna measurements (2,3). As work began under ESA contract to demonstrate the technique, the most important aspect, the transformation algorithm and software was developed allowing dual polarized probe pattern and polarization corrected spherical near field measurements to be implemented (4).
Antennas characteristics can be measured in two ways. lfrequency Domain Method (FDM) is more widely known. The main measuring instruments: Microwave Generator and Receiver. In Time Domain Method (TDM) measurements are fulfilled by using superwide band pulses. The main measuring instruments: Pulse Generator and Sampling Oscilloscope. TDM shows a number of advantages but for narrow-band antennas TDM is difficult to apply and FDM is required. At the testing polygons aimed to measure various antennas we set equipment allowing to use both measurement methods. For TDM we used a two channel sampling converter SD200 of Geozondas production with bandwidth 0-18 GHz. To unify measurements we developed a 3-channel sampling converter SD303 allowing besides pulse to measure sine wave amplitude and phase difference in dynamic range 100 dB. The third channel is used for synchronization. Thus the same instrument assures antenna measurements both in TDM and FDM. At 100 m distance the following characteristics are obtained in Time and Frequency Domains Measurements: Bandwidth 1- 18 GHz. Antenna pattern dynamic range 60 dB Gain measurement accuracy 0.5 dB Phase difference between 2 antennas error 0.5 - 3° (depends on frequency). Hardware, software and digital signal processing algorithms are considered.
Design optimization and measurement of the Luneburg lens antennas are the focus of this paper. One of the important design aspects of an optimal Luneburg lens antenna is to construct a high performance lens with as low number of spherical shells as possible. In a uniform Luneburg lens, the gain is decreased and unwanted grating lobes are generated by reducing the number of shells. This deficiency in the radiation performance of the uniform lens may be overcomed by designing a non uniform lens antenna. The optimized non-uniform spherical lens antenna is designed utilizing the dyadic Green's function of the multi-layered dielectric sphere integrated with a Genetic Algorithm (GA)/Adaptive Cost Function optimizer. Additionally, a novel 2-shell lens antenna is studied and its performance is compared to the Luneburg lens. Finally, measured results for far field patterns and holographic images are shown for the Luneburg lens antenna using the UCLA's bi-polar near field facility.
A simulation tool used during the design of near-field ranges for phased array antenna testing is presented. This tool allows the accurate determination of scanner size for testing phased array antennas under steered beam conditions. Estimates can be formed of measured antenna pointing accuracy, side lobe levels, polarization purity, and pattern performance for a chosen rectangular phased array of specified size and aperture distribution. This tool further allows for the accurate testing of software holographic capabilities.
Antenna measurement techniques historically have been dominated by an assumption that an antenna is a discrete component of the overall electronic system into which it is built. Under this assumption, the measurement technique is to remove the antenna from its host electronic system and place it in a generic test system to measure the gain, pattern, etc. Although this technique still applies to many antenna measurements, it does not work well in cellular/PCS handset measurement applications because cellular/PCS handsets exhibit significant electromagnetic coupling to the human holding the phone. Therefore, the antenna should be measured in situ with a person holding the phone or, for practical reasons, with a mannequin arranged such that it can hold the phone. The mannequin is placed on an azimuth positioner and a near-field probe is moved on a very accurate circular arch from zenith to a significant angle below the mobile phone horizon plane. A description of the chamber and system, and measured results are provided.
In this paper, a near-field time domain scattering measurement technique is described. Near-field measurements are typically performed for radiation applications but not scattering applications. This time domain measurement approach borrows from many of the principles developed in the frequency domain and is ideally suited for broadband scattering characterization. The goal of determining the scattered far-fields of a structure is accomplished by the transformation of near-field data collected over a planar sampling surface. The scattered near-fields were generated with a probe excited by a fast rise time step. In particular, the near-fields were sampled with a second probe and digitized using a digital sampling oscilloscope. The bandwidth of the excitation pulse was approximately 15 GHz. The overall accuracy of this approach is examined through a comparison of the transformed far-field pattern to a numerical calculation.
In-situ pattern measurement of JHU/APL's 60-foot parabolic reflector antenna (S-band), using a low-earth orbit satellite as the source is described. The signal strength and X and Y tracking error voltages are measured as the antenna dish sweeps a matrix of points around the position of the moving satellite. The swept region is approximately ±0.30° from the antenna's boresight. This technique was evaluated during April 1998. This measurement was used to baseline the current performance of the ground station before the feed underwent significant modifications. Before the new feed assembly was installed, the position of the current feed was translated to the new feed assembly. Once installed the performance of the reflector was verified. Misalignment of the feed broadens the main beam and increases the sidelobes. More importantly, the inclusion of new components inside the feed also has the potential to introduce phase errors onto the tracking signals. These phase errors will be translated by the auto-track electronics into pointing errors causing the antenna system to inaccurately follow a target. This paper describes the measurement of the reflector antenna pattern and tracking pattern before the new assembly was installed. Results of pattern measurements with the new assembly will be presented at the conference
The use of dual polarization in meteorological radars offers significant advantages over single polarization. Recently a standard single-polarization Cuband radar was upgraded to operate in dual-polarization mode. The antenna has a 4.2m diameter parabolic reflector with a prime-focus feed. A spherical Fresnel-zone holographic technique was used to obtain the radiation pattern for the upgraded antenna. The sidelobes were higher than predicted and so the data was analyzed to identify the relative contributions of shadowing from the feed crook and surface errors in the dish. This paper describes practical considerations in the measurement of this antenna and the analysis of the results.
Two normalized pattern functions appropriate to Ultra-Wideband (UWB) antennas were discussed at the 1992 AMTA meeting [1]. The normalized pattern spectrum is an image showing radiated signal intensity as a function of azimuth or elevation angle and frequency. The spectrum is complex, and thus requires both an amplitude spectrum image and a phase spectrum image to be complete. It is normalized by dividing by the complex radiated signal at the defined boresite angle for the designated antenna. Therefore, on boresite, the normalized pattern spectrum is unity. The normalized impulse response pattern function is the Fourier Transform of the normalized pattern spectrum. This image plots intensity(and polarity) of the real impulse response of the antenna vs time and angle. On boresite, it is a band-limited impulse corresponding to the normalized pattern spectrum. This paper will discuss measurements of seven UWB antennas, and present normalized pattern results of these antennas. The antennas include both off-the-shelf products and experimental prototypes. Included are antennas which have been used for wide-angle UWB SAR imaging, a coherent UWB application where both signal attenuation and dispersion vs angle are important. The results show how pattern behavior can be separated from boresite transfer function information, and how antennas compare in this compact image format.
Three common methods of measuring circularly antennas on a far-zone range are: using a spinning linear source antenna (SPIN-LIN), measuring the magnitude and with a linearly polarized source antenna in two orthogonal positions (MAG-PHS), and using a circularly polarized source antenna (CIRC-SRC). The MAG-PHS and CIRC-SRC methods are also used in a near-field or com pact range. The SPIN-LIN method is useful because an accur te measurement of the axial ratio and gain can be made without the need to measure phase. The MAG-PHS method is the most general method and can also completely characterize the polarization of the test antenna. The CIRC-SRC method is the simplest and least time-consuming measurement if the antenna response to only one polarization is needed. The choice of measurement method is dictated by schedule, accuracy requirements, and budget. An analysis is presented that provides errors in the measured gain, relative gain pattern, and phase of the test antenna depending on the polarization characteristics of the source and test antennas. These results are useful for deciding which measurement method is the most appropriate to use for a particular job. These results are also useful when constructing more complete error budgets.