This dissertation investigates a simple LC-matching network for the impedance bandwidth enhancement of microstrip patch antennas. Wideband impedance matching is a standard practice for active circuits. Simple impedance matching of antennas is also quite common, but data on wideband impedance matching of antennas is not found very much in the open literature. The matching circuit presented consists out of a resonant LC-circuit with a quarterwave matching line as part of the design. Results for a number or experimental antennas, on which the new technique was applied, are included in the report. A well-defined design procedure is also presented, and results in a relatively small circuit to implement. It is shown that the antenna VSWR bandwidth could be improved to more than double the original size in most of the antennas investigated. / Dissertation (M Eng (Electronic Engineering))--University of Pretoria, 2006. / Electrical, Electronic and Computer Engineering / unrestricted

La simulation numérique est un outil indispensable pour la résolution de problèmes électromagnétiques. Le temps de calcul ainsi que la précision des résultats dépendent du niveau de description utilisé et des dimensions électriques du problème analysé. Dans le cas de figure où il est nécessaire de modéliser un environnement électriquement grand, l'emploi de méthodes classiques comme la méthode des différences finies dans le domaine temporel (FDTD) se heurte à des difficultés en termes de temps de simulation. Cet aspect devient particulièrement problématique quand l'analyse doit être menée pour plusieurs valeurs d'un ou plusieurs paramètres définissant le problème étudié. Face à cette variabilité et afin d'éviter la multiplication de simulations coûteuses, Il est nécessaire d'opter pour des méthodes appropriées. Les travaux de cette thèse sont consacrés au développement d'une nouvelle approche permettant de résoudre de manière efficace l'aspect variable rencontré dans un problème multiéchelle. Cette nouvelle méthode issue de l'association de la FDTD à Grille Double (DG-FDTD) avec un modèle de substitution se nomme la MM-DG-FDTD (Macro-Modèle basé sur la DG-FDTD). L'emploi de la DG-FDTD vise à traiter l'aspect multi-échelle en divisant le problème étudié en sous-volumes FDTD possédant chacun ses propres paramètres de simulation. L'utilisation d'un modèle de substitution rapide construit en peu de simulations électromagnétiques permet de remplacer rigoureusement et rapidement la dernière simulation de la DGFDTD. Après avoir validé le bon fonctionnement du modèle de substitution sur des cas simples, une utilisation de la MM-DGFDTD est effectuée sur deux applications inscrites dans deux contextes : la dosimétrie numérique et le couplage entre antennes. La première propose d'estimer la valeur du champ électrique à l'intérieur de l'oeil gauche d'un modèle de corps humain. Ce dernier est soumis au rayonnement électromagnétique d'une antenne localisée sur un véhicule. Cette étude est effectuée pour un grand nombre de positions du corps autour du véhicule. La seconde application consiste à évaluer la valeur du coefficient de transmission entre deux antennes ULB (Ultra-Large Bande). La position de l'antenne de réception évolue au sein d'un grand environnement. Pour ces deux études, la MM-DG-FDTD montre sa capacité à donner des résultats rapides et précis en comparaison d'une utilisation classique de la DG-FDTD. / Numerical modelling is an essential tool for solving electromagnetic problems. The computation time and the accuracy of the results depend on the description level used and electrical dimensions of the analyzed problem. When it is necessary to model an electrically large environment, the use of conventional methods such as the finite difference time domain (FDTD) face difficulties in terms of simulation time. This becomes particularly problematic when the analysis must be carried out for several values of one or more parameters defining the studied problem. Considering this variability aspect and in order to avoid multiplication of costly simulations, it is necessary to choose suitable methods. This thesis aims to develop a new approach to solve effectively the variable feature encountered in some multi-scale problems. This new method called the MM-DG-FDTD (MacroModel based Dual Grid-FDTD) results from the association of the Dual Grid- FDTD (DG-FDTD) method with a substitution model. The use of the DG-FDTD permits to treat multiscale aspect dividing the studied problem into FDTD sub-volumes. Each step has its own simulation parameters. The fast substitution model built in few electromagnetic simulations replaces in a rigorous and efficient way the last expensive DG-FDTD simulation. The proper operation of the substitution model is firstly validated using simple cases. Next, a use of the MM-DG-FDTD is performed on two applications belonging to two contexts: numerical dosimetry and antennas coupling. The first one proposes to estimate the value of the electric field inside the left eye of a human body model. It receives electromagnetic radiation from an antenna located onboard a vehicle. This study is carried out for a large number of positions of the body around the vehicle. The second application is to assess the value of the transmission coefficient between two UWB (Utra-Wide Band) antennas. The position of the receiving antenna changes in a large environment. For both studies, MM-DG-FDTD shows its ability to give fast and accurate results compared to a classic use of the DG-FDTD.

Thesis (PhD)--Stellenbosch University, 1994. / ENGLISH ABSTRACT: An artificial chiral medium can be made by embedding metal helices (chiral or handed structures) with random orientation inside au achiral hnst medium. The electromagnetic behaviour of such artificial chiral media can be explained by an additional (to p. and t') constitutive parameter, the chirality parameter e. Chiral media have certain special properties of which optical activity is the most prominent. Optical activity is the term used to describe the rotation of the polarization plane of a linearly polarized wave as it travels through a chiral medium. The chirality parameter is directly linked with this rotation. The reflection and the transmission coefficients from a chiral slab are well known in terms of the constitutive parameters and the thickness of the sample. In the thesis a set of .inversion equations are derived that can be used to determine the constitutive parameters (p., £, e) of a chiral medium from the measured S-parameters 811, S21x and S2l'jJ (Le. the reflection and co- and cross-polarized transmission coefficients respectively). An accuracy analysis of the measurement method is made. This is done by using the first order partial derivatives of the inversion equations. A sensitivity analysis is done on the inversion equations and the results are in an analytical form. This makes it possible to determine the contribution of each of the measurement errors to the total expected error. A statistical root mean square method is used to predict the expected error in the measured constitutive parameters. Several artificial chiral samples are measured and an accuracy analysis done on the measurements. The inaccuracy of the method of inversion for low loss samples that are multiples of half a wavelength thick is illustrated by example and explained by the sensitivity analysis. The free-space system (11-17 GHz) used to measure the S-parameters is described in detail. The radiation patterns of the focused lens antennas are measured and compared to those obtained from two theoretical models. A free-space calibration procedure is developed and the dispersion in the focal region of the antennas incorporated into it. The results from different combinations of calibration standards are compared and related to the idea of minimum sensitivity calibration. The possible use of an artificial chiral medium as a microwave absorber is also discussed. / AFRIKAANSE OPSOMMING: Kunsmatige kirale media kan gemaak word deur metaalhelikse lukraak te orienteer in 'n mekirale (bv. diEHektriese) medium. Die mikrogolfeienskappe van kirale media kan toegeskryf word aan die anti-simmetrie van die mikroskopiese insluitings. 'n Voorwerp word as kiraal beskryf indien dit nie deur translasie of rotasie op die spieelbeeld daarvan gepas kan word nie. Die menslike hand is 'n goeie voorbeeld van so 'n kirale struktuur. Die elektromagnetise eienskappe van 'n kunsmatige kirale medium kan beskryf word deuT 'n addisionele (tot j.t en E) samestellingsparameter, die kiraliteitsparameter ~. Kirale media het verskeie ongewone eienskappe waarvan opti~se aktiwiteit waarskynlik die prominentste is. Optiese aktiwiteit beskryf die verskynsel waar die polarisasierigting van 'n lineer-gepolariseerde golf geroteer word terwyl dit deur 'n kirale medium beweeg. 'n Direkte verband bestaan tussen die rotasie en die kiraliteitsparameter. Die weerkaats- en transmissiekotHfisiente van 'n kirale plaat is bekend in terme van die samestellingsparameters (j.t, f. en {) en die dikte van die plaat. In die tesis word 'n ste1 inversievergelykings afgelei wat gebruik kan word om die samestellingsparameters van 'n kirale medium te bepaal vanuit die gemete weerkaats- en transmissiekoeffisiente, Sl1, S'lIz en SUg. (S213/ is die kruis-gepolariseerde komponent van die transmissiekoeffisient wat 'n aanduiding van die rotasie gee.) ,n Analise word gedoen op die akkuraatheid waarmee metings uitgevoer kan word. Dit word gedoen deur gebruik te maak van die eerste orde parsHHe afgeleide van die inversievergelykings. 'n Sensitiwiteitsanalise van die inversievergelykings word gedoen en in 'n analitiese vurm aangebied. Die vorm maak dit moontlik om die bydrae van elke meetfout in S11, S2l:c en S21'!J tot die totale meetfout in p., € en ete bepaal. 'n Statistiese metode (wortel gemiddelde kwadraat) word gebruik om die verwagte foutwaardes in die gemete samestellingsparameters te bepaal. Verskeie kunsmatige kirale media word gemeet en 'n analise word gedoen op die akkuraatheid waarmee elk gemeet is. Die onakkuraatheid van die inversiemetode vir plate met lae verliese en wat veelvoude van 'nhalwe golflengte dik is, word getoon deur middel van 'n sensitiwiteitsanalise. Die vryeruimte meetstelsel (11-17 GHz) wat gebruik word om die weerkaats- en transmissiekoeffisH~ nte te meet, word breedvoerig bespreek. Die gemete stralingspatrone van die gefokusseerde lens antennes word vergelyk met teoreties bepaalde waardes. 'n Vryeruimte kalibrasietegniek word ontwikkel en die gemete dispersie in die fokusgebied word daarby ingesluit. Die resultate, soos verkry van verskillende kalibrasiestandaarde, word vergelyk en in verband gebring met die tegniek van kalibrasie-met-minimum-sensitiwiteit. Ten slotte word die moontlike gebruik van kunsmatige kirale media as mikrogolfabsorbeerders ondersoek.

Thesis (MEng)--University of Stellenbosch, 2001. / ENGLISH ABSTRACT: Balanced loads such as antennas normally require baluns when they are measured. For wide band applications, designing and building a balun complicates the measurement and introduces errors. A simple model for load impedances was developed, together with a novel measurement procedure. The procedure enables the measurement of balanced loads using a network analyzer with no balun. Measured and simulated results are presented. / AFRIKAANSE OPSOMMING: Gebalanseerde laste soos antennas benodig gewoonlik balons om korrek gemeet te word. Die ontwerp van 'n balon vir wye band toepassings bemoeilik die metings en veroorsaak foute. 'n Eenvoudige model vir die lasimpedansies is ontwikkel sowel as 'n eenvoudige meetmetode. Die metode word gebruik om die gebalanseerde laste te meet met 'n netwerk analiseerder sonder die gebruik van 'n balon. Gemete en gesimuleerde resultate word getoon.

A novel modification to the standard cascode amplifier architecture is presented in SiGe which allows for an optimal separation of gain and breakdown functions through the mixed breakdown cascade architecture, opening the door for moderate power amplifiers in SiGe. Utilizing this technique, a two-stage, high-gain amplifier operating at X-Band is fabricated and measured. The 20 dB of gain per stage represents the highest gain at X-Band at the time of publication. Additionally, a near one Watt power amplifier is designed and fabricated at X-Band, which represents the highest output power in SiGe at X-Band at time of publication. Related to the power amplifier design, thermal considerations are also investigated. The validity of utilizing lumped mutual thermal coupling in SiGe devices is presented. Using this finding, a thermal coupling model and network which are compliant for use with commonly available HBT models and circuit simulators is presented. This model and network is used to thermally optimize SiGe PA cells based upon layout spacing.

Thesis (M. App. Sc.)--Carleton University, 2004. / Includes bibliographical references (p. 110-113). Also available in electronic format on the Internet.

Thesis (MScEng)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: An electro-optic sensor system was developed which uses the modulated scatterer technique [1] to measure the fine structure-amplitude and phase-of electric fields with the minimum of perturbation. The theory for the technique is discussed in detail, with the Lorentz Reciprocity theorem as foundation. The system is also treated as a general two port and solved in terms of the impedances present. This comprises self impedances and mutual impedances. The signals scattered by the dipole scatterer are modeled as amplitude modulated signals and a detailed detection scheme is presented. The detection of the signals is done by a digital correlation process. This is quite similar to synchronous detection. A detailed sensitivity analysis is presented that was used to determine the quality of the system. This analysis is backed by a practical sensitivity measurement. A few antennas and scatterers are measured in the X-band using the system, and results are compared with theory. These systems include a pyramidal horn, an E-plane sectoral horn and a simple waveguide. Half plane diffraction by a metal sheet is also presented. The practical implementation of the theory is emphasized. / AFRIKAANSE OPSOMMING: 'n Elektro-optiese meetsisteem is ontwikkel wat die gemoduleerde strooier tegniek [11 gebruik om die fyn struktuur-amplitude en fase-van elektriese velde te meet. Hierdie tegniek versteur die velde minimaal. 'n Volledige ondersoek word gedoen na die teorie van die stelsel. Die basis van die metode Ie in die Lorentz Resiprook teorie. Alhoewcl daar gekonsentreer word op hierdie teorie, is die stelsel ook beskou as 'n tweepoort. Die wedersydse en ei impedansies van die verskillende komponente word gebruik om die velde op te los. Die gestrooide seine, as gevolg van die dipool strooier, word gemodelleer met b hulp van amplitude modulasie. Die deteksie van hierdie seine word deeglik bespreek. In plaas van 'n analoogstelsel, word 'n digitale korrellasie proses gebruik om die velde te meet. Di sensitiwiteit van die stelsel word geanaliseer. Verder word metings met di stelsel gen em om die teoretiese analise te staaf. 'n Paar X-band sisteme word gebruik om metings te neem. Hicrdi metings word ook vergelyk met teor ti se resultate. Stelsels wat gemeet is sluit die volgencle in: 'n piramidale horing antenna, 'n E-vlak sektorale horing antenna, 'n X-band golfgeleier en die diffraksi op die rand van 'n plaat. Die praktiese implemcntering van die teorie word beklemtoon.

nÃo hà / A expansÃo das redes de telecomunicaÃÃes sem fio e o fenÃmeno da convergÃncia digital trazem a inerente necessidade da pesquisa de novos componentes que assegurem a sustentabilidade e a evoluÃÃo dos sistemas. Novos tipos de antenas, menores e mais eficientes, sÃo exigidas à medida que novos dispositivos vÃo surgindo. Neste contexto, as antenas ressoadoras dielÃtricas, construÃdas com novos materiais, aparecem como excelente opÃÃo para substituir as antenas metÃlicas tradicionais. Este trabalho apresenta uma proposta de antena ressoadora dielÃtrica circularmente polarizada, operando na frequÃncia central de 2,25 GHz, na qual um esquema de alimentaÃÃo por sonda Ãnica à empregado para excitar dois modos ressonantes em um dielÃtrico em forma de quarto de cilindro. Este leiaute permite a ativaÃÃo de modos de baixa ordem, com distribuiÃÃo ortogonal dos campos eletromagnÃticos, ressoando em frequÃncias prÃximas com uma diferenÃa de fase de 90Â. SÃo introduzidos conceitos da teoria eletromagnÃtica envolvendo cavidades ressonantes e caracterÃsticas dos materiais cerÃmicos que compÃem o dielÃtrico. AlÃm disso, os processos de modelagem por computador e de construÃÃo de um protÃtipo sÃo explicados. Os resultados sÃo discutidos comparativamente entre o modelo computacional e as medidas experimentais executadas em laboratÃrio. O estudo demonstra uma boa concordÃncia entre os resultados simulados e os experimentais e evidencia a viabilidade da antena para aplicaÃÃes que necessitem de polarizaÃÃo circular na regiÃo do espectro de frequÃncias prÃximas a 2,25 GHz. / The expansion of wireless telecommunications networks and the phenomenon of digital convergence bring the inherent need for research of new components to ensure the sustainability and evolution of the systems. New types of antenna, smaller and more efficient, are required as new devices emerge. In this context, the dielectric resonator antennas, built with new materials, appear as an excellent option to replace the conventional metallic antennas. This work presents a proposal for a circularly polarized dielectric resonator antenna to operate at the center frequency of 2.25 GHz in which a single probe feeding scheme is used to excite two resonant modes in a quarter-cylinder-shaped dielectric. This layout allows the activation of low-order modes with orthogonal distribution of electromagnetic fields, resonating at near frequencies with a 90 phase difference. The concepts of electromagnetic theory related to resonant cavities and the characteristics of dielectric ceramic materials are introduced. Furthermore, the processes of computer modeling and constructing of a prototype are explained. The results are discussed by comparison between the computational model and experimental measurements performed in the laboratory. The study shows a good agreement between the simulated and experimental results and demonstrates the feasibility of the antenna for applications requiring circular polarization for operating at the region of the frequency spectrum close to 2.25 GHz.

This thesis describes an electrically thin surface used for electromagnetic applications in the microwave regime. The surface is free-standing and its primary purpose is to modify the phase distribution, or the phase and amplitude distribution of electromagnetic fields propagating through it: it is called phase shifting surface (PSS) in the first case, and phase and amplitude shifting surface (PASS) in the second case. For practical applications, the surface typically comprises three or four layers of metallic patterns spaced by dielectric layers. The patterns of the metallic layers are designed to locally alter the phase (and amplitude in the case of the PASS) of an incoming wave to a prescribed set of desired values for the outgoing wave. The PSS/PASS takes advantage of the reactive coupling by closely spacing of the metallic layers, which results in a larger phase shift range while keeping the structure significantly thin. The PSS concept is used to design components such as gratings and lens antennas which are presented in this document. The components are designed for an operating frequency of 30 GHz. The PSS phase grating gives high diffraction efficiency, even higher than a dielectric phase grating. Several types of lens antennas are also presented, which show comparable performance to that of a conventional dielectric plano-hyperbolic lens antenna with similar parameters. The PASS concept is used in a beam shaping application in which a flat-topped beam antenna is designed. This work demonstrates the potential for realising thin, lightweight and low-cost antennas at Ka band, in particular for substituting higher-gain antenna technologies such as conventional dielectric shaped lens antennas.

This thesis describes an electrically thin surface used for electromagnetic applications in the microwave regime. The surface is free-standing and its primary purpose is to modify the phase distribution, or the phase and amplitude distribution of electromagnetic fields propagating through it: it is called phase shifting surface (PSS) in the first case, and phase and amplitude shifting surface (PASS) in the second case. For practical applications, the surface typically comprises three or four layers of metallic patterns spaced by dielectric layers. The patterns of the metallic layers are designed to locally alter the phase (and amplitude in the case of the PASS) of an incoming wave to a prescribed set of desired values for the outgoing wave. The PSS/PASS takes advantage of the reactive coupling by closely spacing of the metallic layers, which results in a larger phase shift range while keeping the structure significantly thin. The PSS concept is used to design components such as gratings and lens antennas which are presented in this document. The components are designed for an operating frequency of 30 GHz. The PSS phase grating gives high diffraction efficiency, even higher than a dielectric phase grating. Several types of lens antennas are also presented, which show comparable performance to that of a conventional dielectric plano-hyperbolic lens antenna with similar parameters. The PASS concept is used in a beam shaping application in which a flat-topped beam antenna is designed. This work demonstrates the potential for realising thin, lightweight and low-cost antennas at Ka band, in particular for substituting higher-gain antenna technologies such as conventional dielectric shaped lens antennas.