Health & Medicine

A Graphical Aid for the Complex Permittivity Measurement at Microwave and Millimeter Wavelengths

Description
We introduce a novel procedure to retrieve the complex permittivity of dielectric materials. It is a variant of the well-known waveguide method, and uses as input the one-port reflection data from a vector network analyzer connected to a short
Published
of 3
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  IEEE MICROWAVE AND WIRELESS COMPONENTS LETTERS, VOL. 24, NO. 6, JUNE 2014 421 A Graphical Aid for the Complex PermittivityMeasurement at Microwave andMillimeter Wavelengths Mário G. Silveirinha  , Senior Member, IEEE  , Carlos A. Fernandes  , Senior Member, IEEE  , andJorge R. Costa  , Senior Member, IEEE   Abstract—  We introduce a novel procedure to retrieve the com-plex permittivity of dielectric materials. It is a variant of the well-known waveguide method, and uses as input the one-portre fl ectiondatafromavectornetworkanalyzerconnectedtoashort-circuited rectangular waveguide  fi lled with a dielectric sample of known length. Here, it is shown that for low to moderate loss ma-terials, the locus of the re fl ection coef  fi cient in the complex planeversusfrequencyisapproximatelyacircumferencearcwith curva-tureradiusthatdependsmainlyon andsuchthatthesweptangledependsmostlyon .Itisproventhat fi ttingthetheoreticalcircum-ference arc with the measured data not only allows identifying pos-sible measurement errors but also enables estimating the complexpermittivitywithgoodaccuracy.Agraphicalbasedimplementationof the method is described and validated experimentally.  Index Terms—  Measurement of complex dielectric permittivity,microwave and millimeter wave measurements, waveguide graph-ical method. I. I  NTRODUCTION T HEknowledgeofthecomplexpermittivity of adielectricmaterialisof keyimportanceforthedesignof microwaveandmillimeter-wavecomponents,suchasprintedcir-cuits, fi lters, and antennas. For the past two decades, the authorshavebeenworkingondielectriclensantennas[1].Anaccuratean-tennadesignandcharacterizationrequiresapreciseknowledgeof thematerialcomplexpermittivitywithintheoperatingfrequency band.Forlensantennasapplications,itisalsoessentialtocon fi rmthat the dielectric material is to a good approximation isotropicand homogeneous before the lens fabrication. Typical materialsusedindielectriclenseshavelowtomoderatepermittivity,, and loss values of the order of .There are several well-known techniques for the measure-ment of the complex permittivity at microwaves and mil-limeter-waves.Averycompleterevisionispresentedin[2].Thewaveguidebasedmethodallowsusingsmallersizesamplesthanopen air methods. This is quite useful because it permits cutting Manuscript received September 23, 2013; revised November 29, 2013; ac-cepted February 26, 2014. Date of publication March 24, 2014; date of currentversion June 03, 2014. This work was supported in part by the Fundação para aCiência e Tecnologia under Project mm-SatComPTDC/EEI-TEL/0805/2012.M. G. Silveirinha is with the Department of Electrical Engineering, Institutode Telecomunicações, University of Coimbra, Coimbra 3030, Portugal (e-mail:mario.silveirinha@co.it.pt).C. A. Fernandes is with the Instituto de Telecomunicações, IST, Lisboa 1049-001, Portugal.J. R. Costa is with the Instituto de Telecomunicações, IST, Lisboa 1049-001,Portugal and also with the Departamento de Ciências e Tecnologias da Infor-mação, Instituto Universitário de Lisboa (ISCTE-IUL), Lisboa 1649-026, Por-tugal.Color versions of one or more of the  fi gures in this letter are available onlineat http://ieeexplore.ieee.org.Digital Object Identi fi er 10.1109/LMWC.2014.2310470Fig. 1. Experimental setup (open view). samples from different parts of the bulk material, eventuallywith different orientations, and investigate the homogeneityand isotropy of the material. Here we propose a graphical basedapproach that allows measuring the complex permittivity of a material and quickly diagnoses the occurrence of commonsources of error such as relatively wide air gaps or a misplace-ment of the sample. These imperfections can be detected duringthe course of measurements, allowing immediate correctiveactions. The method allows for an unambiguous determinationof the complex permittivity from a single material sample.II. F ORMULATION The experimental setup is shown in Fig. 1. The singlemode waveguide operation is assumed. The dimensions of the waveguide cross-section are in the H-plane, and in theE-plane. The sample under-test is non-magnetic and assumednon-dispersive within the frequency band of interest. For now,it is supposed that the sample with length completely  fi llsthe waveguide cross-section, without air-gaps. The rectangular waveguide is terminated with a short circuit. A vector network analyzer (VNA) is used to obtain the re fl ection coef  fi cientreferred to the plane, as the frequency is swept in theinterval , where is the centralfrequency and is the frequency span.The re fl ection coef  fi cient at is calculated with transmis-sion line theory. The waveguide transverse impedance in air (subscript ) or in the dielectric is given by(1)where is the longitudinal wave number in each medium(2)The re fl ection coef  fi cient at the dielectric side of theinterface is(3) 1531-1309 © 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.  422 IEEE MICROWAVE AND WIRELESS COMPONENTS LETTERS, VOL. 24, NO. 6, JUNE 2014 Fig. 2. Locus of the re fl ection coef  fi cient from to 65 GHz on theSmithChart(orcomplexplane)fordifferentpermittivityvaluesofasamplewithin a waveguide with width : (a) re fl ection coef  fi cientat the dielectric side of the interface; (b) re fl ection coef  fi cient at theair side of the interface. On the other hand, the re fl ection coef  fi cient at the air sideof interface , can be expressed in terms of as follows:(4)This can be rewritten in a more convenient way as(5)where(6)For weak material loss, , and a small fre-quency span , (3) represents approximately a circumferencearc in the Smith Chart (or complex plane). This is con fi rmed inFig. 2, for a set of simulated curves in the frequency interval. Next, we note that the derivative of the parameter with frequency is(7)But typically we have(8)hence, may be considered ap proximately constant within theswept frequency band. Within this assumption the re fl ection co-ef  fi cient given by (5) is a bilinear (Möbius) transformation(9)with and complex constants. It iswell known that this transformation maps circles and lines intocircles and lines [3]. Thus, because in case of low loss liesin a circumference centered at the srcin, we conclude that ismapped by (5) into another circumference with radius(10)and whose center is shifted away from the srcin to(11)Calculated examples are presented in Fig. 2.Consistent with these results it can be checked that can bewritten as(12)where the phase is given by(13)The swept arc associated with , , is related to the sweptfrequency rangeas(14)Thus, from the length (in radians) of the swept arc of circum-ference associated with the locus of , one can obtain a  fi rstestimate for the unknown permittivity by solving (14). Gen-erally this equation has several solutions. The physical solutionis extracted by comparing with the arc initial point.Once is known one can compute from (6), and then, fromthe radius , we can  fi nd(15)This result is easily obtained from (10). Finally a  fi rst approxi-mation for is found using(16)III. E XAMPLE OF  A PPLICATION This section describes the application of the method to thedetermination of the complex permittivity of Polyethylene inthe interval .A V-band rectangular waveguide receptacle with nominal di-mensions , and lengthwas fabricated and terminated with a short-circuit. A calibra-tion procedure is used to accurately determine the actual valuesof and . This calibration involves the measurement of the re- fl ection coef  fi cient of the empty receptacle at the interfacefor the frequency span of interest. An optimization procedure isused to fi t the measured data (circumference in the Smith Chart)with (3). In the present example, this yields and.The Polyethylene sample is then placed inside the sampleholder, and is measured for this setup. The locus of the ex- perimental data in the complex plane is shown in Fig. 3. Using astandard least square minimization one can  fi nd the circumfer-encethatbest fi tsthemeasureddata.Itisfoundthatitiscenteredat , that the radius isand that the swept arc is radians. Desirably, thelength of the sample and the frequency span must be suchthat the measured arc amplitude is larger than 180 in order toreduce the estimation error in the  fi tting process.Using(14) and (16), it is found that .This  fi rst guess value is then fed to an optimizing routine that fi nds the best  fi t between the experimental data and the theo-retical model from (5). The optimization goal is to minimizethe squared distance between measured data and the theoreticalmodel (5). This re fi nement gives .The use  SILVEIRINHA  et al. : A GRAPHICAL AID FOR THE COMPLEX PERMITTIVITY MEASUREMENT AT MICROWAVE AND MILLIMETER WAVELENGTHS 423 Fig. 3. Locus of the measured re fl ection coef  fi cient of a Polyethylene samplewith length from to 65 GHz on the Smith Chart.Fig. 4. Locus of the simulated re fl ection coef  fi cient of a MACOR™ samplewith length from to 65 GHz on the Smith Chart, with andwithout a air gap adjacent to the top wall of the waveguide. of curve  fi tting methods to determine dielectric properties withwaveguide measurements has been considered in other works[4], but the novelty of our approach is that it allows obtaining aquite accurate  fi rst estimate for the permittivity from the sweptarc length and curvature radius in the Smith chart. In this ex-ample the error in the initial estimate of the real part of the per-mittivity is less than 0.2%.In order to validate the proposed method, a disk sample fromthe same Polyethylene batch was cut, with a diameter andthickness anditscomplexpermittivitywas measured using the open resonator method Fabry-Perot[5]. The obtained complex permittivity value is[6]. The mismatch in the imaginar y parts of the mea-sured permittivities is larger than for the real parts, because it isdif  fi cult to measure very precisely (with any method) in caseof very low loss materials.IV. I  NFLUENCE OF  A IR   G APS It is well known that air-gaps, mainly in the E-plane wherethe E- fi eld is the stronger, can introduce considerable error inthe complex permittivity estimation [2]. This error decreaseswith the permittivity value of the material [7]. A solution for this problem is to apply a conducting paste to the edges of thesample [7]. Alternatively, if the height of the gap can bemeasured and is known, it is possible to relate the “correct”complex permittivity with the measured value as [2](17)where . Equation (17) can benumerically solved with respect to .A particularly interesting feature of our graphical methodis that the presence of air gaps or other perturbations (e.g., amisplaced sample or an imperfect short-circuit termination) can be detected during the measurements. In fact, air gaps srcinatehybrid or high order modes in the sample holder. Theses modesintroduceperturbationsinthe measured re fl ection coef  fi cientthat are manifested as ripples, curls or deviations from the ex- pected circumference representation. In order to exemplify this behavior two simulations were performed in CST MicrowaveStudio[8](Fig.4)foraMACORsample()withandwithoutanairgapintheE-planeof .For the case without air gap the proposed graphical methodgivesthepermittivity ,whileinthepresenceof the gap the method gives .Using (17) one can eliminate the in fl uence of the air gap, andthis yields .V. D ISCUSSION The proposed method is a simple graphic  fi tting procedureand yet it has several advantages. Since it is a waveguide basedmethod it only requires small samples. The samples can be cutfrom different locations of the material batch and used to eval-uate itshomogeneity and anisotropy. The method isnot plaguedwithmultiplesolutionsthatrequirethemeasurementofmultiplesamples. Most importantly, during the measurement procedureit is simple to identify errors and perturbations such as the mis- placement of the sample or air gaps. Therefore, a simple visualinspection of the data representation in the Smith Chart allowsfora fi rstvalidationand,ifneeded,anadequatecorrectionofthemeasurement procedure, even before the post-processing of themeasured data and the calculation of the complex permittivity.A CKNOWLEDGMENT The authors wish to thank V. Fred for prototype construction,A. Almeida for prototype measurements, and J. Almeida for initial developments.R  EFERENCES[1] C. A. Fernandes, E. B. Lima, and J. R. Costa, “Broadband inte-grated lens for illuminating re fl ector antenna with constant apertureef  fi ciency,”  IEEE Trans. Antennas Propag. , vol. 58, no. 12, pp.3805–3813, Dec. 2010.[2] L. F. Chen, C. K. Ong, C. P. Neo, V. V. Varadan, and V. K. Varadan  , Microwave electronics: Measurement and materials characteriza-tion . New York: Wiley, 2004, ch. 4.[3] R. Churchill  , Complex variables and applications . New York: Mc-Graw-Hill, 1984.[4] S. Khanal, T. Kiuru, J. Mallat, O. Luukkonen, and A. V. Räisänen,“Measurement of dielectric properties at 75–325 ghz using a vector network analyzer and full-wave simulator,”  Radioeng. J. , vol. 21, no.2, pp. 551–556, 2012.[5] T. M. Hirvonen, P. Vainikainen, A. Lozowski, and A. V. Raisanen,“Measurement of dielectrics at 100 ghz with an open resonator con-nected to a network analyzer,”  IEEE Trans. Instrum. Meas. , vol. 45,no. 4, pp. 780–786, Aug. 1996.[6] C. A. Fernandes and J. R. Costa, “Permittivity measurement andanisotropy evaluation of dielectric materials at millimeter-waves,”in  Proc IMEKO World Congress , Lisboa, Portugal, Sep. 2009, pp.673–677.[7] S. B. Wilson, “Modal analysis of the ‘Gap effect’ in waveguide dielec-tric measurements,”  IEEE Trans. Microw. Theory Tech. , vol. 36, no. 4, pp. 752–756, Apr. 1988.[8] CST [Online]. Available: www.cst.com
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks