No description available.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: Worldwide, more large antenna arrays are being deployed in areas of science previously dominated by other antenna geometries. Applications for large arrays include Radar, Satellite Communications and Radio Astronomy. Even though the use of large arrays solve some of the difficulties posed by more traditional antennas, new challenges are also faced. One of these challenges is the problem of noise coupling, and how the overall system performance is affected by it. The Focal Plane Array (FPA), which is a new example of a large antenna array, is currently being researched at a number of institutions worldwide for use in Radio Astronomy. As a result, FPA’s were used as an example element to demonstrate the practical importance of this research. In this study, the effect of mutual coupling on the noise performance of FPA’s was illustrated. This was done by calculating the mutual coupling between the elements of the array, and then calculating the noise power received by each element as a result of the mutual coupling. Next, the Active Noise Figure and Active Noise Temperature were calculated. These parameters were introduced to visualise the effect of mutual coupling on the overall noise performance of the array. Since FPA’s are by definition large, conventional brute-force analysis techniques are very resource intensive. Solving the coupling terms using these methods therefore requires the use of computer clusters even during the design phase of the antenna, which is very expensive. A method was therefore developed to calculate the coupling terms of a large array using Periodic Boundary Conditions. The method uses infinite array analysis, which resulted in an improvement in memory usage in orders of magnitude. This improvement comfortably places the memory requirements for the analysis of large arrays within the range of current personal computers. The results also displayed a reasonable amount of accuracy for use during the design phase of an array. The additional noise power on each element as a result of mutual coupling were also calculated. This was achieved by developing an equivalent circuit diagram that represents the system in terms of the noise and transmission parameters of the LNA of each receiver channel, and the coupling terms of the antenna array. Lastly, the active noise temperature and active noise figure are calculated. The theory was implemented by means of a script with a graphical user interface, to provide easy-to-use access to the theory. A quick reference table of estimated noise coupling penalty versus first term coupling and LNA noise temperature was also compiled. The results of an example calculation showed a significant amount of noise coupling in an 8×8 Vivaldi array. The noise coupling resulted in an increase in system noise temperature, Tsys, in the order of 9% of the LNA noise temperature, TLNA. According to the SKA Tsys budget, this results in an approximate Tsys increase of 1.3 Kelvin. In the context of Radio Astronomy, this additional source of noise cannot be ignored, as it can greatly affect the usebility of the telescope for certain areas of research. / AFRIKAANSE OPSOMMING: Groot antennaskikkings word deesdae al hoe meer ingespan in plek van ander tradisionele antennamodelle. Toepassings vir groot antennaskikkings sluit Radar, Satellietkommunikasie en Radioastronomie in. Alhoewel die gebruik van groot antennaskikkings baie van die probleme wat deur ander tradisionele antennamodelle veroorsaak word oplos, word nuwe uitdagings terselfdertyd geskep. Een van hierdie nuwe uitdagins is ruiskoppelling en hoe dit die ruisgedrag van die stelsel as ’n geheel affekteer. ’n Beeldvlakskikking (FPA), is ’n opwindende nuwe voorbeeld van ’n groot antennaskikking en die moontlikheid vir die gebruik daarvan in radioastronomie word tans wêreldwyd nagevors. Om hierdie rede is die FPA gekies as voorbeeldelement om die bruikbaarheid van hierdie navorsing in die praktyk te beklemtoon. In hierdie studie word die effek van wedersydse koppelling op die ruisgedrag van FPA’s geïllustreer. Dit word gedoen deur eers die wedersydse koppelling tussen die elemente van die antennaskikking te bereken en dan die ruisdrywing wat deur elke element ontvang word as gevolg van wedersydse koppelling. Daarna word die Aktiewe Ruistal en die Aktiewe Ruistemperatuur bereken. Hierdie nuwe parameters word bekendgestel om die gevolge van wedersydse koppelling op die ruisgedrag van die stelsel as ’n geheel te visualiseer. Omdat FPA’s per definisie groot is, vereis die analise daarvan deur middel van konvensionele metodes baie rekenaar hulpbronne. Hierdie metodes vereis dus die gebruik van rekenaarbondels of superrekenaars selfs gedurende die ontwerpfase van die antenna, wat baie duur en onprakties is. Daar is dus ’n metode ontwikkel wat gebruik maak van periodiese randvoorwaardes om groot antennaskikkings te analiseer. Die metode benader ’n groot antennaskikking as ’n eindig-opgewekte oneindige skikking van antennas. As gevolg hiervan, word die geheueverbruik met ordegroottes verbeter. Hierdie verbetering plaas dus die analise van groot antennaskikkings binne die vermoëns van huidige persoonlike rekenaars. Die resultate wys ook ’n aanvaarbare graad van akkuraatheid vir gebruik gedurende die ontwerpfase van die skikking. Die bykomende ruisdrwying op elke element as gevolg van wedersydse koppelling is ook bereken. Om dit te vermag, is daar ’n ekwivalente stroombaandiagram ontwikkel wat die gekoppelde stelsel in terme van die ruis- en transmissieparameters van die laeruisversterker (LNA) aan elke ontvangerkanaal en die koppelterme van die antenna skikking voorstel. Laastens word die aktiewe ruistal en die aktiewe ruistermperatuur ook bereken. Die teorie is geïmplimenteer deur gebruik te maak van ’n grafiesegebruikerskoppelvlak (GUI). Die GUI verskaf aan die gebruiker maklike toegang tot die teorie wat onwikkel is in hierdie navorsing. Daar is ook ’n snelnaslaantabel geskep met benaderde waardes van ruiskoppelling vir ’n verskeidenheid waardes van LNA ruistemperature en eerste element koppelling. Die resultate van ’n 8×8 Vivaldiskikking voorbeeld, het ’n beduidende hoeveelheid ruiskoppelling getoon. Die ruiskoppelling het ’n maksimum toename in stelsel ruistemperatuur, Tsys, van ongeveer 9% van die LNA ruistemperatuur tot gevolg gehad. Volgens die huidige Tsys begroting van die SKA, kom dit neer op ’n Tsys toename van byna 1.3 Kelvin. In die konteks van die radioastronomie, kan hierdie toename in ruistemperatuur nie geïgnoreer word nie aangesien dit die bruikbaarheid van die teleskoop vir sekere velde van navorsing nadelig kan beïnvloed.

In this thesis, noise coupling simulation is introduced into the behavioral level. Methods and models for simulating on-chip noise coupling at a behavioral level in a design flow are presented and verified for accuracy and validity. Today, designs of electronic systems are becoming denser and more and more mixed-signal systems such as System-on-Chip (SoC) are being devised. This raises problems when the electronics components start to interfere with each other. Often, digital components disturb analog components, introducing noise into the system causing degradation of the performance or even introducing errors into the functionality of the system. Today, these effects can only be simulated at a very late stage in the design process, causing large design iterations and increased costs if the designers are required to return and make alterations, which may have occurred at a very early stage in the process. This is why the focus of this work is centered on extracting noise coupling simulation models that can be used at a very early design stage such as the behavioral level and then follow the design through the various design stages. To realize this, SystemC is selected as a platform and implementation example for the behavioral level models. SystemC supports design refinement, which means that when designs are being refined and are crossing the design levels, the noise coupling models can also be refined to suit the current design. This new way of thinking in primarily mixed-signal designs is called Behavioral level Noise Coupling (BeNoC) simulation and shows great promise in enabling a reduction in the costs of design iterations due to component cross-talk and simplifies the work for mixed-signal system designers. / Electronics Design Division

Delta-Sigma (ΔΣ) analog-to-digital converters (ADCs) are traditionally used in high quality audio systems, instrumentation and measurement (I&M) and biomedical devices. With the continued downscaling of CMOS technology, they are becoming popular in wideband applications such as wireless and wired communication systems,high-definition television and radar systems. There are two general realizations of a ΔΣ modulator. One is based on the discrete-time (DT) switched-capacitor (SC) circuitry and the other employs continuous-time (CT) circuitry. Compared to a CT structure, the DT ΔΣ ADC is easier to analyze and design, is more robust to process variations and jitter noise, and is more flexible in the multi-mode applications. On the other hand, the CT ΔΣ ADC does not suffer from the strict settling accuracy requirement for the loop filter and thus can achieve lower power dissipation and higher sampling frequency than its DT counterpart. In this thesis, both DT and CT ΔΣ ADCs are investigated. Several design innovations, in both system-level and circuit-level, are proposed to achieve lower power consumption and wider signal bandwidth. For DT ΔΣ ADCs, a new dynamic-biasing scheme is proposed to reduce opamp bias current and the associated signal-dependent harmonic distortion is minimized by using the low-distortion architecture. The technique was verified in a 2.5MHz BW and 13bit dynamic range DT ΔΣ ADC. In addition, a second-order noise coupling technique is presented to save two integrators for the loop filter, and to achieve low power dissipation. Also, a direct-charge-transfer (DCT) technique is suggested to reduce the speed requirements of the adder, which is also preferable in wideband low-power applications. For CT ΔΣ ADCs, a wideband low power CT 2-2 MASH has been designed. High linearity performance was achieved by using a modified low-distortion technique, and the modulator achieves higher noise-shaping ability than the single stage structure due to the inter-stage gain. Also, the quantization noise leakage due to analog circuit non-idealities can be adaptively compensated by a designed digital calibration filter. Using a 90nm process, simulation of the modulator predicts a 12bit resolution within 20MHz BW and consumes only 25mW for analog circuitry. In addition, the noise-coupling technique is investigated and proposed for the design of CT ΔΣ ADCs and it is promising to achieve low power dissipation for wideband applications. Finally, the application of noise-coupling technique is extended and introduced to high-accuracy incremental data converters. Low power dissipation can be expected. / Graduation date: 2010

The ever increasing diversity in communication systems has created a demand for constant improvements in receiver components. This thesis describes the design and characterisation of front-end receiver components for various challenging applications, including characterisation of low noise foundry processes, LNA design and multi-band antenna design. It also includes a new theoretical analysis of noise coupling in low noise phased array receivers.In LNA design much depends on the choice of the optimum active devices. A comprehensive survey of the performance of low noise transistors is therefore extremely beneficial. To this end a comparison of the DC, small-signal and noise behaviours of 10 state-of-the-art GaAs and InP based pHEMT and mHEMT low noise processes has been carried out. Their suitability in LNA designs has been determined, with emphasis on the SKA project. This work is part of the first known detailed investigation of this kind. Results indicate the superiority of mature GaAs-based pHEMT processes, and highlight problems associated with the studied mHEMT processes. Two of the more promising processes have then been used to design C-band and UHF-band MMIC LNAs. A new theoretical analysis of coupled noise between antenna elements of a low noise phased array receiver has been carried out. Results of the noise wave analysis, based on fundamental principles of noisy networks, suggest that the coupled noise contribution to system noise temperatures should be smaller than had previously been suggested for systems like the SKA. The principles are applicable to any phased array receiver. Finally, a multi-band antenna has been designed and fabricated for a severe operating environment, covering the three extremely crowded frequency bands, the 2.1 GHz UMTS, the 2.4 GHz ISM and the 5.8 GHz ISM bands. Measurements have demonstrated excellent performance, exceeding that of equivalent commercial antennas aimed at similar applications.