Pulse power system is widely used for medical, industrial and military applications. The operational principle of the pulse power system is that the energy from the input source is stored in the capacitor bank or superconducting inductive device through a dc-dc converter. Then, when a discharging signal exists, the stored energy is released to the load through pulse forming network (PFN) generating high peak power pulse up to gigawatts within several tens of or hundreds of microseconds. The pulse power system has been originally developed for the defense application. After the format of the voltage compression and voltage addition stages for the short-pulse high power acceleration has been established, it has been evolved to be common. Then, its application has been extended to food processing, medical equipment sterilization and wastewater treatment since many present environmental problems have been known in the early 70's or even earlier. In addition, the pulse power system is newly spotlighted due to the recent world events. The application examples are to treat anthrax-contaminated mail, and the use of accelerators to produce high power X-rays for security screening. Furthermore, the pulse power system has been applied for the tactical weapon system such as electrothermal-chemical (ETC) gun, coilgun and active armor system. Because the pulse power system applied for the tactical weapon system has the potential to be integrated in the military vehicle, a compact, lightweight pulse power system is strongly required for the future weapon system. In this thesis, a distributed power system (DPS) for the capacitor charger is introduced for the application of the active armor system. Furthermore, a design methodology is presented for the front-end converter to achieve the high efficiency as well as the high power density. Design parameters are identified and their impact on the design result is studied. the optimal operating point is determined based on the loss comparison between different operating points. In order to further improve the power density utilizing the unique operation mode i.e. pulse power operation, transformer design using amorphous-based core is provided and the design result is compared with that using ferrite-based core. A 5 kW prototype converter is built up and the experimentation is performed to verify the design. / Master of Science

A new approach is presented that relies upon 3D electromagnetic simulation results to characterize the complex permittivity of homogeneous dielectric materials. By modeling the test fixture and obtaining a set of simulated S-parameters through an iterative solution process, the dielectric constant and loss tangent can be found. With further development, the 3D simulation results may be used to replace the need for complex theoretical analysis of the measurement geometry. The method is applied to an X-band rectangular waveguide setup, for which the theoretical S-parameters can be readily calculated. A Teflon sample, for which the dielectric properties are well-known, is used for all measurements and calculations. After presenting a detailed derivation to obtain the theoretical S-parameters, the Teflon sample is measured and compared to the theoretical results, from which the comparison shows great promise. An inverse solution algorithm is used to solve for the material properties from the experimental S-parameters. Low-frequency measurement of the Teflon sheet was performed by using a dielectric capacitor test fixture. The results show the effect of an air gap between the electrode and sample, producing serious errors. / Master of Science

University of Macau / Faculty of Science and Technology / Department of Electrical and Electronics Engineering

The focus of this thesis has been to determine the usable voltage range of carbon-based supercapacitors (SC). Supercapacitors are a relatively new type of capacitors with a vast increase in capacitance compared to capacitors which utilize a dielectric as charge separator. A SC consists of two electrodes and an electrolyte separating the electrodes. The charges are stored by electrostatic forces in the interface between the electrode and the electrolyte, forming the so called electrochemical double-layer (EDL). With porous electrodes the effective surface area of the interfacial zone can be made very large, giving SCs a large storage capacity. The limiting factors of a SC is the decomposition potential of the electrolyte and the decomposition of the electrodes. For commercially manufactured SCs the electrolyte is usually an organic solvent, which has a decomposition potential of up to 2.7-2.8 V. Compared to aqueous electrolytes with a thermodynamic limit of 1.23 V. The drawback of using non-aqueous electrolytes is that they are not environmentally friendly, and they increase the production cost. It is claimed that the voltage range can be up to 1.9 V using aqueous electrolytes. Some researchers have focused on aqueous electrolytes for these reasons. In this thesis two different electrolytes were tested to determine if the voltage range could be extended. The experiments were conducted using a three electrode cell and performing cyclic voltammogram measurements (CV). The carbon electrodes were made of  two different sources of grahite, battery graphite or exfoliated graphite, and nano fibrilated cellulose was added to increase the mechanical stability. The results show that the oxidation potential of the carbon electrode was the positive limit. A usable potential of about 1 V was shown. However, when cycling the electrodes to potentials below the decomposition limit, for hydrogen evolution, interesting effects were seen. A decrease in reaction kinetics, indicating a type of conditioning of the electrode was observed. An increase in charge storage capacitance was also observed when comparing the initial measurements with the final, probably corresponding to an increase in porosity. / KEPS projekt Sundsvall Mitt Universitet

Neste trabalho, investigou-se capacitores MOS fabricados sobre superfícies irregulares contendo formas retangulares periódicas com 100 nm de altura, obtidas a partir de corrosão por plasma localizadas. Os óxidos de porta com 4,5 nm de espessura foram crescidos em ambientes ultrapuros de O2 ou pirogênico a fim de comparar a uniformidade de cobertura sobre os degraus verticais dos perfis retangulares. Foi mostrado que a oxidação pirogênica ou convencional na temperatura de 850 ºC permite obter óxidos de porta sobre degraus com altura de 100nm com baixa corrente de fuga e alto campo de ruptura. Esse comportamento pode ser interpretado como óxidos de porta perfeitamente amoldados sobre os degraus de 100nm de altura. O impacto deste resultado é agora a possibilidade de implementar óxidos de porta para transistores de porta envolvente e FinFETs. / In this work, it was investigated MOS capacitors fabricated onto periodic rectangular shapes, 100 nm in height, obtained by localized plasma etching onto silicon wafer surfaces. 4.5-nm gate oxide growth was performed in ultrapure dry O2 or pyrogenic environments in order to compare the coverage uniformity at the step edges of rectangular shapes defined onto the silicon surfaces. It was shown that pyrogenic and conventional oxidation at 850 ºC allows one to obtain gate oxides on 100nm-stepped silicon surfaces with low leakage current and high dielectric breakdown field. This behavior can be understood as highly conformal gate oxides over silicon steps with height of 100 nm. The impact of this result is now the feasibility of implementing gate oxides for surrounding gate transistors (SGT\'s) and FinFETs.

Neste trabalho, investigou-se capacitores MOS fabricados sobre superfícies irregulares contendo formas retangulares periódicas com 100 nm de altura, obtidas a partir de corrosão por plasma localizadas. Os óxidos de porta com 4,5 nm de espessura foram crescidos em ambientes ultrapuros de O2 ou pirogênico a fim de comparar a uniformidade de cobertura sobre os degraus verticais dos perfis retangulares. Foi mostrado que a oxidação pirogênica ou convencional na temperatura de 850 ºC permite obter óxidos de porta sobre degraus com altura de 100nm com baixa corrente de fuga e alto campo de ruptura. Esse comportamento pode ser interpretado como óxidos de porta perfeitamente amoldados sobre os degraus de 100nm de altura. O impacto deste resultado é agora a possibilidade de implementar óxidos de porta para transistores de porta envolvente e FinFETs. / In this work, it was investigated MOS capacitors fabricated onto periodic rectangular shapes, 100 nm in height, obtained by localized plasma etching onto silicon wafer surfaces. 4.5-nm gate oxide growth was performed in ultrapure dry O2 or pyrogenic environments in order to compare the coverage uniformity at the step edges of rectangular shapes defined onto the silicon surfaces. It was shown that pyrogenic and conventional oxidation at 850 ºC allows one to obtain gate oxides on 100nm-stepped silicon surfaces with low leakage current and high dielectric breakdown field. This behavior can be understood as highly conformal gate oxides over silicon steps with height of 100 nm. The impact of this result is now the feasibility of implementing gate oxides for surrounding gate transistors (SGT\'s) and FinFETs.

No description available.

Bakgrunden till detta projekt är att Vattenfall Eldistribution AB (nedan kallat Vattenfall) kommer att införa ett avgiftssystem för inmatning av reaktiv effekt till sitt elnät. Avgiften införs till följd av problem på elnätet som orsakas av reaktiv effekt. Umeå Energi Elnät AB (nedan kallat Umeå Energi) har historiskt matat in reaktiv effekt vilket motiverade detta arbete som utreder den reaktiva effekten på Umeå Energis elnät samt bidrar med information om två alternativa tillvägagångssätt att bemöta avgiften. Alternativ 0 är att kompensationsutrustning inte installeras, utan att ett abonnemang på inmatning av reaktiv effekt upprättas. Alternativ 1 är att kompensationsutrustning installeras. För att utreda den reaktiva effekten erhölls och behandlades data på reaktiv effekt i Umeå Energis nät. Historisk kondensatordrift togs fram för år 2016 ur händelsehistoriken hos Umeå Energis driftcentral. Kondensatordriften år 2015 kunde enbart erhållas från ett tidigare arbete på Umeå Energi eftersom ett begränsat antal händelser lagras i händelsehistoriken. Genom att subtrahera kondensatorernas produktion från den reaktiva effekten i Umeå Energis anslutningspunkter som uppmätts av Vattenfall erhölls data som mer representerade underliggande fenomen på nätet. Utan kondensatordrift beräknades inmatningen enligt Vattenfalls definition uppgå till cirka 34 MVAr utifrån data från 2015 och 2016. För åren 2018 till och med 2023 beräknades ändringar i reaktiv effekt till följd av förändringar på Umeå Energis nät. Vid beräkningarna försummades ledningarnas induktiva karaktär, vilket gav ett tomgångsscenario med maximal produktion av reaktiv effekt. År 2023 beräknades inmatningen ska ha ökat till 59 MVAr till följd av förändringar på Umeå Energis nät. Med antagandet att Umeå Energi inte kommer att drifta kondensatorbatterierna så att inmatningen höjs föreslogs för alternativ 0 val av abonnemang på inmatning av reaktiv effekt för åren 2018 till och med 2023 utifrån de 34 MVAr som nämnts ovan och inverkan från förändringarna på nätet. År 2019 föreslogs ett abonnemang på 41 MVAr, och 2023 föreslogs ett på 59 MVAr. Kostnaderna för dessa beräknades enligt Vattenfalls tariff till 820 000 kr respektive 1 187 000 kr. Kostnaden för eventuell överinmatning av reaktiv effekt beräknades med tariffen för överinmatning årligen uppgå till maximalt 76 000 kr med 95 % sannolikhet enligt den korrigerade standardavvikelsen hos inmatningen utan kondensatordrift åren 2015 och 2016. Optimal kondensatordrift beräknades för åren 2015 och 2016 genom att addera den produktion av reaktiv effekt från befintliga kondensatorbatterier som gav minst absolutvärde i reaktiv effekt. Beroende på hur ofta kondensatordriften justerades erhölls olika resultat. En undersökning av störningar till följd av kondensatorkopplingar rekommenderas för att få en förståelse för förutsättningarna för optimal kondensatordrift. Det bedömdes inte ekonomiskt motiverbart med mer avancerad kompensationsteknik såsom statiska VAr-kompensatorer då variabla reaktorer kan kompensera dygns- och säsongsvariationer i reaktiv effekt. Den reaktiva effektproduktionen i ledningar är störst på 145 kV-nivån och kommer öka i framtiden på denna nivå. Det är därför sannolikt här kompensationsutrustning såsom reaktorer först bör installeras. För att kunna ta så bra beslut som möjligt angående den reaktiva effekten rekommenderas att snarast möjligt ingå ett arbetssätt som om avgiftssystemet redan tagits i bruk och utöka ett representativt dataunderlag. / The background of this project is that Vattenfall Eldistribution AB (hereinafter referred to as Vattenfall) will establish a system of fees for input of reactive power. This will be done due to problems in the grid caused by reactive power. Umeå Energi Elnät AB (hereinafter referred to as Umeå Energi) has historically input reactive power, motivating this work which investigates the reactive power in the grid of Umeå Energi and provides information on two alternative approaches to responding to the fee. Alternative 0 entails no installation of compensation technology, and that a subscription for reactive power input is established instead. Alternative 1 entails that compensation technology is installed. To investigate the reactive power, data on reactive power in the grid of Umeå Energi were obtained and processed. Historical operations of capacitor banks for the year 2016 were obtained from the history of events of the control center at Umeå Energi. The operations of the capacitor banks during 2015 could only be obtained from an earlier work at Umeå Energi since the number of events stored in the history is limited. By subtracting the capacitor banks’ production from the reactive power measured by Vattenfall in the connections of Umeå Energi, data more representative of underlying phenomena were obtained. Without capacitor production of reactive power, the input was calculated according to the definition of Vattenfall to about 34 MVAr, by using data from 2015 and 2016. For the years 2018 through 2023, changes in reactive power due to changes in the grid of Umeå Energi were calculated. These calculations did not consider inductances, and thus yielded zero-load scenarios with maximum reactive power production. By the year of 2023, the input was calculated to have increased to 59 MVAr due to changes in the grid of Umeå Energi. Assuming that Umeå Energi will not operate the capacitors so that the input is increased, for alternative 0 subscriptions for input of reactive power were suggested for the years 2018 through 2023 by considering the abovementioned 34 MVAr and the changes in the grid. Subscriptions of 41 MVAr and 59 MVAr were suggested for the years 2019 and 2023 respectively. The costs of these were calculated with the fee specified by Vattenfall to SEK 820,000 and SEK 1,187,000 respectively. Calculations with the applicable fee yielded that the yearly cost of possible over-input could amount to a maximum of SEK 76,000 with a 95 % probability, using the corrected standard deviation of the input without capacitor production of reactive power for the years 2015 and 2016. Optimal capacitor bank operations were calculated for the years 2015 and 2016 by adding the production of reactive power from existing capacitor banks which yielded the minimum absolute reactive power. Depending on how often the capacitors were operated different results were obtained. An investigation of power quality disturbances due to capacitor bank operations is recommended to achieve an understanding of the conditions for optimal capacitor bank operations. It was not deemed economically justifiable to install more advanced compensation technologies such as static VAr compensators since variable reactors are able to compensate daily and seasonal variations in reactive power. The production of reactive power in cables is the largest on the 145 kV level and will increase in the future on this level. It is therefore likely here compensation technologies such as reactors should be installed first. To be able to make as good decisions as possible concerning the reactive power, it is recommended to as soon as possible commence a working method as if the fee system had already come into effect; thus increasing the amount of representative data.

Three-level diode-clamped inverter is being widely used these days. Extensive research has been carried out on pulse width modulation (PWM) strategies for a three-level inverter. The most widely used PWM strategies are sine-triangle pulse width modulation (SPWM) and centered space vector pulse width modulation (CSVPWM). The influence of these PWM strategies on the DC-link capacitor current and voltage ripple is studied in this thesis. The sizing of the DC capacitor depends on value of the maximum RMS current flowing through it. In this work, an analytical expression for capacitor RMS current is derived as a function of operating conditions like modulation index, power factor angle of the load and peak load current. The worst case current stress on the capacitor is evaluated using the analytical expression. The capacitor RMS current is found to be the same in SPWM and CSVPWM schemes. The analytical expression is validated through simulations and experiments on a 3kVA MOSFET based three-level inverter. Harmonic analysis of the capacitor current is helpful in better evaluation of capacitor power loss. Therefore, harmonic analysis of the capacitor current is carried out, using the techniques of geometric wall model and double Fourier integral for SPWM and CSVPWM schemes. The theoretical predictions are validated through experiments. The capacitor RMS current is divided into low-frequency RMS current (where low frequency component is defined as a component whose frequency is less than half the switching frequency) and high-frequency RMS current. The capacitor voltage ripple is estimated analytically for SPWM and CSVPWM schemes, using the low-frequency and high-frequency capacitor RMS current. The voltage ripples due to SPWM and CSVPWM schemes are compared. It is found that the voltage ripple with SPWM is higher than that with CSVPWM. A simplified method to estimate the capacitor power loss, without the requirement of FFT analysis of capacitor current, is proposed. The results from this simplified method agree reasonably well with the results from the detailed method. A space vector based modulation scheme is proposed, which reduces the capacitor RMS current at high power factor angles. However, the proposed method leads to higher total harmonic distortion (THD) than CSVPWM. Simulation and experimental results, comparing CSVPWM and the proposed PWM, are presented.

FundaÃÃo Cearense de Apoio ao Desenvolvimento Cientifico e TecnolÃgico / With global need to reduce energy consumption, the search for more efficient technologies has become the focus of many studies. Among these technologies, it can mention the photovoltaic solar energy and LEDs, which have shown an expansion in recent decades. Photovoltaic generation is shown as an attractive energy source because it is renewable and its raw material is practically inexhaustible. While LEDs have a promising advance in lighting and is used in several applications. To integrate these technologies, this paper proposes the study of a DC-DC step-up switched-capacitor (SC) converter for LEDs applied to photovoltaic stand-alone systems. The proposed circuit differs from other topologies SC to insert an inductor in series with the input source, which can operate in discontinuous conduction mode (DCM), reducing losses switching, or continuous conduction mode (CCM), allowing the reduction of conduction losses in the circuit. The converter is driven by the frequency modulation, which is obtained as a function of input voltage. For this reason, the current in the LEDs can be stabilized without the need for sensor or feedback. The prototype developed in the laboratory was designed for a lamp of 54 W and operating at high frequency (up to 165 kHz), allowing the reduction of the circuit volume. Experimental results of the circuit in DCM and CCM show that the converter has a high yield, validating the proposal. / Tendo em vista a tendÃncia da reduÃÃo do consumo de energia no mundo, a busca por tecnologias mais eficientes tem se tornado o foco de muitos estudos. Dentre estas tecnologias, pode-se citar a energia solar fotovoltaica e os LEDs, que vem apresentando uma expansÃo nas ultimas dÃcadas. A geraÃÃo fotovoltaica se mostra como uma atrativa fonte de energia, por ser renovÃvel e sua matÃria-prima ser praticamente inesgotÃvel. Enquanto os LEDs apresentam um avanÃo promissor na iluminaÃÃo, sendo utilizado nas mais diversas aplicaÃÃes. Visando a integraÃÃo destas tecnologias, este trabalho propÃe o estudo de um conversor CC-CC elevador com capacitor comutado (Switched Capacitor - SC) para LEDs, aplicado a sistemas fotovoltaicos autÃnomos. O circuito proposto se difere de outras topologias SC por inserir um indutor em sÃrie com a fonte de entrada, o qual pode operar no modo de conduÃÃo descontÃnua (MCD), reduzindo as perdas por comutaÃÃo, ou no modo de conduÃÃo contÃnua (MCC), possibilitando a reduÃÃo das perdas por conduÃÃo do circuito. O conversor à acionado atravÃs da modulaÃÃo por frequÃncia, a qual à obtida em funÃÃo da tensÃo de entrada. Por este motivo, a corrente nos LEDs pode ser estabilizada sem a necessidade de sensores ou de realimentaÃÃo. O protÃtipo desenvolvido em laboratÃrio foi projetado para uma luminÃria de 54 W e operando em alta frequÃncia (atà 165 kHz), possibilitando a reduÃÃo do volume do circuito. Os resultados experimentais obtidos do circuito MCD e MCC sÃo analisados e validam a proposta, mostrando que o conversor apresenta rendimento elevado