IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 10, 2011 1047 Study of Thermal Noise Generated in a Vivaldi Antenna Using the Improved Wheeler Cap Method Leonid Belostotski, Member, IEEE, Adrian Sutinjo, Member, IEEE, Ronald H. Johnston, Life Senior Member, IEEE, Michal Okoniewski, Fellow, IEEE, Michael A. Petursson, and Tom Burgess Abstract—We analyze sources of inefficiency in a single Vivaldi antenna. The analysis is based on simulations of antenna radia- tion efficiencies and measurement data obtained via the Improved Wheeler Cap (IWC) method. The accuracy of the IWC method is established in full-wave simulation (HFSS) by comparison to metal loss computation in method-of-moment simulation (FEKO) and ef- ficiency calculation of a lossless Vivaldi antenna. These results sug- gest reliability of the IWC method. The method is then used to in- vestigate sources of losses in the antenna. Index Terms—Antenna measurements, radio astronomy, Vivaldi antennas. I. INTRODUCTION H IGH-SENSITIVITY radio telescopes, such as the Square Kilometer Array (SKA), require a very low since (1) where is the total effective area of the telescope and is the system noise temperature [1]. Thermal noise generated by an antenna element in a phased array feed (PAF) of such a telescope contributes to . Reliable measurement of very high antenna efficiencies (roughly 95% or 15 K antenna noise temperature con- tribution) is very challenging [2]. This is also the case with numerical simulation of highly efficient antennas. Calculations of total radiated power in commercial simulation software using far-field Poynting vector integration often result in efficiencies exceeding 100% [3]. It has been demonstrated that greater Manuscript received July 26, 2011; accepted September 01, 2011. Date of publication September 26, 2011; date of current version October 10, 2011. This work was supported by the University of Calgary, the Natural Sciences and Engineering Research Council of Canada, the Alberta Ingenuity Fund, NSERC’s Special Research Opportunity, the Alberta Provincial Government’s iCORE program, the Dominion Radio Astrophysical Observatory, the National Research Council, the Herzberg Institute of Astrophysics, and CMC Microsys- tems. L. Belostotski, A. Sutinjo, R. H. Johnston, M. Okoniewski, and M. A. Petursson are with the Department of Electrical and Computer Engi- neering, University of Calgary, Calgary, AB T2N 1N4, Canada (e-mail: atsutinj@ucalgary.ca). T. Burgess is with Dominion Radio Astrophysical Observatory, Herzberg In- stitute of Astrophysics, National Research Council, Penticton, BC V2A 6J9, Canada. Color versions of one or more of the figures in this letter are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/LAWP.2011.2169769 Fig. 1. Two-port representation of an antenna. reliability can be achieved using dissipated power [3], [4] as opposed to radiated power calculations. Though very high efficiencies can be accurately calculated, it is less clear which measurement method is as reliable. One method that may be promising for our application is the Improved Wheeler Cap (IWC) method. As reported in [2], measurement of a very low-loss antenna (bowtie-like copper monopole) using the IWC resulted in 99.17 0.15% effi- ciency, which approaches 100% but does not exceed it. Thus, one aspect of this letter is to examine the reliability of the IWC method when a large antenna, such as a Vivaldi for PAF [5], is involved. Such a problem requires a modification to the IWC data post-processing to remove the effect of cavity resonances [6]. In addition, we present an estimate of accu- racy of the IWC via simulation of a lossless Vivaldi and by comparison to metal loss calculation as described in [3] using method-of-moments (MoM) FEKO software. Next, using the IWC method, thermal noise contributions from constituent parts (metal, dielectric, transition) of a single Vivaldi are dis- cussed. It should be noted that this work is limited to a single Vivaldi element, and efficiency measurement of a Vivaldi array is beyond the scope of our work. II. IWC MEASUREMENT OF VIVALDI ANTENNA EFFICIENCY A. Effective Noise Temperature Measurement Using the IWC Method: Overview The antenna is treated as a two-port network (Fig. 1), where represents the reflection coefficient of free space, is the antenna-port reflection coefficient, and the antenna electrical characteristics are modeled with a 2 2 matrix, , of -pa- rameters. For a passive network such as an antenna, the amount of effective input-referred noise temperature, , is related to the available gain of the network, and in the case of antenna under test (AUT), this relationship becomes [7] (2) 1536-1225/$26.00 © 2011 IEEE