M.Ing. / In this thesis, a new isolated DC-DC converter topology is designed which implements a high frequency primary tap changing transformer. This converter is designed to be implemented into renewable energy applications. Renewable or alternative energy resources are becoming more popular by necessity. The voltage generated by renewable energy resources is dependent on the weather conditions, thus the voltage may vary. This novel topology can maintain a regulated DC output voltage for a wide input voltage range by implementing a high frequency primary tap changing transformer. This converter is thus designed to accommodate the voltage changing conditions of renewable energy resources. This converter only requires output bus capacitance for DC bus stiffening. The overall required bus capacitance is therefore effectively reduced compared to other converter topologies. This isolated topology also improves the fundamental power quality. Other power quality components of this new converter are also analysed.

D.Ing. / The necessity of frequency selective interconnects in future integrated power electronic systems are highlighted. A brief historical background illustrates that such interconnects have been utilized successfully in other fields of electrical engineering, although based on different high frequency effects, and configurations. Based on high frequency characteristics of typical interconnects, it is hypothesized that the required frequency selectivity could be obtained through utilization of the skin- and proximity effect, and low conductivity materials, to increase high frequency resistance significantly. Finite element simulation results for a large number of interconnects, and reference structures, are presented in an effort to identify relevant parameters and mechanisms. A hybrid lumped / distributed parameter impedance model is proposed. Parametric analysis is conducted to determine limitations and constraints of the proposed technique. Frequency selective damping of turn-off related power electronic switch and interconnect inductance resonance, is investigated as a possible application of such higher resistance. A simplified analytical model is proposed, and utilized to calculate turn-off waveforms and percentages of damping. An approximation of maximum damping possible is presented. Utilization of enhanced high frequency resistance to realize interconnect based low pass filters for medium power integrated power electronic modules, is investigated as a second application. Based on typical parameter influence, a number of structures are evaluated with finite element simulations. An analytical, lossy transmission line model is developed. Parametric analysis for a chosen structure is conducted, followed by discussion of maximum attenuation, and relative effectiveness. As a third application, the above concepts are applied to 1.5kA nominal current interconnects. A number of structures are evaluated. Application of consecutive impedance mismatches to increase attenuation is investigated. Current and voltage capacity constraints are discussed. Experimental verification of the presented concepts in general, are presented. Technical difficulties and limitations are identified. An objective oriented discussion completes the thesis, with the conclusion that the original hypothesis has been validated.

M.Ing. / The product of switching frequency and output power of solid state converters has increased dramatically during the past few years. This progress can mainly be attributed to the recent developments in power electronic switching devices. These developments are mainly responsible for a reduction in the physical size of the converter, especially that of the reactive components, and an increase in the dynamic response of the converter ...

M.Ing. (Electrical & Electronic Engineering) / The technology of the use of electronic devices as power switches is still being developed. Because of the increasing demand for low mass and cost, the switching frequency must be as high as possible. The limiting factor for the frequency is the energy loss which is dissipated in the semiconductor crystal during each switching-transient. If tllis switching loss can be reduced, a higher frequency can be obtained. The switching loss is due to the non-instantaneous switching process of the semiconductor device. For the reduction of this switching loss, the device must be switched as fast as possible via its control-electrode. Further improvement can be obtained through the use of snubber networks. This thesis considers the bipolar high-voltage transistor as a power switch. The unfavourable switching loci of the transistor- switch is discussed in chapter 1. In chapter 2 the basedrive during transistor turn-off is investigated, in order to reduce the turn-off time. The following chapters de;al with the use of snubber circuits, which relieve the transistor during turn-on and turn-off. The advantages, as well as the disadvantages and limitations of snubbers are discussed. After the linear turn-off and turn-on snubbers have been treated, the non-linear turn-on snubber, which uses a saturable inductor, will be considered from an experimental point of view. The results show that this snubber is very effective in relieving the transistor during turn-on. In chapter 7 the non-linear turn-off snubber is treated on an experimental as well as on a mathematical bas e . Cri teria, which must be satisfied for realisation of a practical nonlinear capacitor, is also laid down. Because of the fact that a first order snubber is always loading the transistor during the complimentary switchin,]-transient, chapter 8 deals with a second order snubber, which consists of both a capacitor and an inductor. Experimentally it was found that the turn-on time is reduced with the use of a turn-on snubber, while the turn-off time is increased with the use of a turn-off snubber. In chapter 9 the physical behaviour of the switching processes is investigated, after which the influences of the snubbers on the switching times is explained qualitatively...

D.Ing. / This thesis treats aspects of electromagnetic integration in planar structures. These structures are for use in power electronic converters. The electromagnetically integrated structures fulfil the same electromagnetic function as conventional component structures, but have manufacturing advantages. In order to reduce the cost of power electronic systems, a new cost effective manufacturing technology is needed. In this work a planar manufacturing approach is considered. This manufacturing approach lends itself well to the concept of integrated electromagnetics, where one physical structure fulfils more than one electromagnetic function. As not all the electromagnetic functions of a given converter may be integrated into a single structure, the approach is one of hybridisation. Some aspects remain conventional and some aspects are totally electromagnetically integrated. A proposal for a new schematic representation of converter topologies is put forward. These new schematics bridge the gap between conventional circuit diagrams and full three dimensional figures of the physical structures. These schematics form the basis of the design and analysis procedure and are necessary in order to convey information on the structures in a concise manner. A complex electromagnetic function, that of a resonant circuit with transformer coupling, is described and analysed in detail. The design equations governing the structure are derived and their implications on the volume of the structure is discussed. In the analysis only first order effects are considered. Two case studies are carried out. The two converters are described in detail with respect to their construction. Several aspects of general planar construction are identified and discussed.

This thesis presents the development of a Hybrid Multilevel Switching Converter (HMSC) for Ring-Magnet Power Supplies (RMPS). The thesis includes the analysis, design and control of the proposed converter. It introduces multilevel switching converters to the field of ring-magnet power supplies. RMPS feed the electromagnets that produce the magnetic field required to energize and guide subatomic particles in a synchrotron. The accuracy and extreme precision of this magnetic field imposes stringent restrictions on the design and performance of the RMPS used. Study of conventional power supplies highlights the need for modern power supply solutions which can meet the specifications of RMPS. The complete frequency-domain analysis of the conventional resonant-type RMPS along with the Energy Make-up Unit (EMU) is presented. The resonant frequency drift is identified as the main factor in the design of the EMU. The analysis of the input filter network is presented for development of design criteria for input filter components. The principle advantages and disadvantages of the resonant-type RMPS are summarized before identifying multilevel converters as a viable option among switching converters for a non-resonant type of RMPS. The Hybrid Multilevel Switching Converter (HMSC) is proposed as a non-resonant type RMPS to overcome the disadvantages of the resonant-type RMPS. The operational features of the HMSC are explained and the simplification of the general HMSC configuration for positive output currents is identified. The steady-state analysis of the HMSC develops comprehensive design criteria for the device ratings and component stresses, including the methods for reducing the switching losses in the HMSC. Multilevel converters encounter voltage balancing problem among the DC-link capacitors. It is shown that the HMSC configuration is versatile in minimizing this problem. Harmonic spectrum of the output voltage of the HMSC is derived and the effect of number of output voltage levels in reducing the harmonic contents is established. A detailed survey of different current control techniques is presented to form the background for developing an effective current control algorithm for multilevel converters. A dead-beat current control strategy is chosen as an appropriate control technique to suit the needs of RMPS. The control scheme is extended to the control of multilevel converters in general. The control algorithm is developed to track a given arbitrary current reference signal for both single-variable and multi-variable systems. It is also shown that the output dead-beat control is a special case of the pole placement technique. The transient behaviour of the system has been studied and stability considerations of the system are examined. Extensive computer simulation studies have been performed using SABER to study the reference tracking nature of the proposed control scheme. The output current of the HMSC using the modified dead-beat control scheme is shown to follow a given arbitrary-reference with very small tracking error. The reference tracking nature has been simulated for a simple RL magnet load and a magnet load with LCR filter. Experimental results obtained from a laboratory prototype of the HMSC with an RL load, have been presented to substantiate the analytical results. Criteria for improvement in the reference tracking properties of the proposed system have been identified. / Graduate

Photovoltaic (PV) systems are a rapidly growing segment in the renewable energy industry.  Though they have humble origins and an uncertain future, the commercial viability of PV has significantly increased, especially in the past decade.  In order to make PV useful, however, significant effort has to go into the power conditioning systems that take the low-voltage dc from the panel and create utility compatible ac output.  Popular architectures for this process include the centralized inverter and the distributed micro-inverter, each with its own advantages and disadvantages.  One attempt to retain the advantages of both architectures is to centralize the inverter function but construct PV panel-level micro-converters which optimize the panel output and condition the power for the inverter.  The main focus of this work is to explore the technical challenges that face the evolution of the dc-dc micro-converter and to use them as a template for a vertically integrated design procedure. The individual chapters focus on different levels of the process:  topology, modulation and control, transient mitigation, and steady-state optimization.  Chapter 2 introduces a new dc-dc topology, the Integrated Boost Resonant (IBR) converter, born out of the natural design requirements for the micro-converter, such as high CEC efficiency, simple structure, and inherent Galvanic isolation.  The circuit is a combination of a traditional PWM boost converter and a discontinuous conduction mode (DCM), series resonant circuit.  The DCM operation of the high-frequency transformer possesses much lower circulating energy when compared to the traditional CCM behavior.  When combined with  zero-current-switching (ZCS) for the output diode, it results in a circuit with a high weighted efficiency of 96.8%.  Chapter 3 improves upon that topology by adding an optimized modulation scheme to the control strategy.  This improves the power stage efficiency at nominal input and enhances the available operating range.  The new, hybrid-frequency method utilizes areas where the modulator operates in constant-on, constant-off, and fixed-frequency conditions depending on duty cycle, the resonant period length, and the desired input range.  The method extends the operating range as wide as 12-48V and improves the CEC efficiency to 97.2% in the 250-W prototype.  Chapter 4 considers the soft-start of the proposed system, which can have a very large capacitive load from the inverter.  A new capacitor-transient limited (CTL) soft-start method senses the ac transient across the resonant capacitor, prematurely ending the lower switch on-time in order to prevent an excessive current spike.  A prototype design is then applied to the IBR system, allowing safe system startup with a range of capacitive loads from 2μF to 500μF and a consistent peak current without the need for current sensing.  Chapter 5 further investigates the impact of voltage ripple on the PV output power.  A new method for analyzing the maximum power point tracking (MPPT) efficiency is proposed based on panel-derived models.  From the panel model, an expression demonstrating the MPPT efficiency is derived, along with a ripple "budget" for the harmonic sources.  These ripple sources are then analyzed and suggestions for controlling their contributions are proposed that enable circuit designers to make informed and cost-effective design decisions.  Chapter 6 illustrates how results from a previous iteration can provide a basis for the next generation's design.  A zero-voltage-switching (ZVS) version of the circuit in Chapter 2 is proposed, requiring only two additional MOSFETs and one inductor on the low-voltage side.  The maximum switching frequency is then increased from 70kHz to 170kHz, allowing for a 46% reduction in converter volume (from 430cm³ to 230cm³) while retaining greater than 97% weighted efficiency. / Ph. D.

Integrated architectures for power electronic circuits have been a subject of recent interest. Integration offers several benefits such as reliability, control on parasitic elements related to discrete components, and ease of manufacture. The main objective of this particular research has been to contribute towards effective modelling of integrated passive circuits operating in power electronic circuits. Integrating passive components in one distributed space can be difficult to understand, and hence to design. Field electromagnetics is often unwieldy for a power electronics circuit designer, so a SPICE-like circuit simulator is often an effective design environment. This dissertation closely examines both lumped and distributed SPICE-compatible models. Four SPICE-compatible models have been investigated by comparing them with an analytical distributed solution. This analytical solution is used to thoroughly derive the causes of all resonance points, as well as impedances at low/high frequencies; which are the important factors that characterize the integrated passive. This analytical solution is only implemented in a narrow range of boundary conditions; hence the SPICE-compatible methods must be developed, since SPICE then handles the algorithmic work of handling the more complicated boundary conditions found in power electronics.

Power electronics is a diverse and multi-disciplinary field with constant opportunities to grow, change, and try new techniques to push the envelope. Immediately following the importance of hardware improvement is working on the algorithms controlling the switches. There are many controllers used in the field, all with their own unique benefits and drawbacks. This document will study the feasibility of adding the "Take-Back-Half" algorithm to the ranks of controllers utilized. This algorithm is a non-linear modification to the PI controller with potential benefits in speed and stability.

The future of microprocessors is unknown. Over the past 40 years, their historical trend has been for adopting smaller and more powerful designs that drive the world that we live in today. The state of the microprocessor business today faces a crossroad, wishing to continue on the historical trend of doubling the number of transistors on a chip every 18 months (Moore’s Law) but also facing the realistic task of needing to power these sophisticated devices. With the low voltages and high currents that are required for these microprocessors to operate, it poses a difficult task for the future designers of the voltage regulators that are used to power these microprocessors. The technique that has been widely adopted as the preferred method to power these devices is called a multiphase buck converter, or multiphase voltage regulator. This thesis is a continuation of and is aimed to improve previous work done by two former Cal Poly students, Kay Ohn and Ian Waters. A new design that uses an interleaving control scheme, careful component selection, an input LC filter, and a reduction in board size seeks to improve the efficiency, input current noise, and increase the current density of the original design. Research was first conducted to determine how to best make such improvements. The design phase ensued, which used design calculations and simulations to test if the proposed multiphase topology was plausible. Once the theory was fully proven, a real hardware circuit was created and tested to confirm the results. The results yield a multiphase design with improved input noise filtering, greater efficiency, more equal current sharing, and higher current density as compared to previous topologies in this field. Parameters such as output voltage ripple, load and line regulation, and transient response remained excellent, as they were with the previous work.