The survey from US Department of Energy showed that about one-third of energy consumption in US is due to air conditioning and refrigeration systems. This significant usage of electricity in the HVAC industry has prompted researchers to develop dynamic models for the HVAC components, which leads to implementation of better control and optimization techniques. In this research, efforts are made to model a multi-evaporator system. A novel dynamic modeling technique is proposed based on moving boundary method, which can be generalized for any number of evaporators in a vapor compression cycle. The models were validated experimentally on a commercial supermarket refrigeration unit. Simulation results showed that the models capture the major dynamics of the system in both the steady state and transient external disturbances. Furthermore the use of MEMS (microelectromechanical) based Silicon Expansion Valves (SEVs) have reportedly shown power savings as compared to the Thermal Expansion Valves (TEVs). Experimental tests were conducted on a supermarket refrigeration unit fitted with the MEMS valves to explain the cause of these potential energy savings. In this study an advanced cascaded control algorithm was also designed to control the MEMS valves. The performance of the cascaded control architecture was compared with the standard Thermal Expansion Valves (TEVs) and a commercially available Microstaq (MS) Superheat Controller (SHC). The results reveal that the significant efficiency gains derived on the SEVs are due to better superheat regulation, tighter superheat control and superior cooling effects in shorter time period which reduces the total run-time of the compressor. It was also observed that the duty cycle was least for the cascaded control algorithm. The reduction in duty cycle indicates early shut-off for the compressor resulting in maximum power savings for the cascaded control, followed by the Microstaq controller and then the Thermal Expansion Valves.

Graduation date: 1990

Rapid progress over the last decade on thin film transistor (TFT) active matrix organic light emitting (AMOLED) displays led to the emergence of high-performance, low-power, low-cost flat panel displays. Despite the shortcomings of the active matrix that are associated with the instability and low mobility of TFTs, the amorphous silicon TFT technology still remains the primary solution for the AMOLED backplane. To take advantage of this technology, it is crucial to develop driving schemes and circuit techniques to compensate for the limitations of the TFTs. The driving schemes proposed in this thesis address these challenges, in which, the sensitivity of the OLED current to the transistor variations is reduced significantly. This is achieved by comparing the data signal with a feedback signal associated with the pixel current by means of an external driving circuit through a column feedback line. Depending on the nature of the feedback signal, (i.e. current or voltage) several pixel circuits and external drivers are proposed. New AMOLED pixel circuits with voltage and current feedback are designed, simulated, fabricated, and tested. The performance of these circuits is analyzed in terms of their stability, settling time, power efficiency, noise, and temperature-dependence. For the pixel circuits with current feedback, an operational transresistance amplifier is designed and implemented in a high-voltage CMOS process. Measurement results for both voltage and current feedback driving schemes indicate less than a 2%/V sensitivity to shifts in the threshold voltage of the TFTs. By using current feedback and an accelerating pulse, programming times less than 50 s are achieved.

Detailed material characterization of spherical silicon is conducted for the first time. Experimental results on crystallinity, impurities, and structural defects are presented to investigate the effect of growth mechanism and processing of the spheres. Based on the material properties, the spherical bulk is characterized from electronic point of view. A model is developed to interpret photoconductivity decay measurements and extract minority carrier lifetime, the most influential parameter in device performance. The model includes the spherical geometry as well as the radial profile of carrier lifetime and the measurement results are used to characterize the quality of the spheres and the effectiveness of the process steps. To analyze and predict the performance of spherical cells, a three dimensional opto-electric model is developed. The model separately treats the optical generation and carrier collection and is able to calculate the spectral response and the device characteristics. A simulation tool is created based on this model. The simulation results are of great importance in designing novel structures and optimizing fabrication processes. The necessary characterization methodologies are developed to measure the spectral response and I-Vcharacteristics of individual cells, as well as an array consisting of several cells. A new device design with passivated selective emitter is proposed based on simulation results. Plasma processes were developed to selectively etch back the deep emitter of the spherical cells using reactive ion etching, and then passivate the surface using plasma enhanced chemical vapor deposition of silicon nitride. The improvement in characteristics of the fabricated device is characterized using the developed measurement setups. Industrial up-scaling and manufacturability of the proposed devices and processes are also discussed.

The aim of this work was to study the effect of flow rate on the fuming rate/silica flux of liquid silicon in order to gain a better understanding of the silica fuming during industrial ladle refining during silicon production. The formation of silica fume is the results when the liquid silicon is exposed to air. This silica fume has been shown to be a health hazard when breathed by plant workers and increased environmental regulations call for its elimination. This work is being done as part of the FUME project (FUgitive emissions of Materials and Energy) with funding from the Norwegian Research Council and the Ferro-alloys Industrial Research Association (FFF). To examine the silica flux a series of experiments in a medium scale induction furnace were carried out. A graphite crucible with an inside diameter of 11.4 cm was charged with refined or unrefined metallurgical grade silicon. This silicon was melted and air was blown onto the surface at 3 different rates (1, 3.5, 5 LPM). The produced silica fume was collected in a filter and analyzed for particle size and compositions differences. The total mass of the fume produced was used to estimate the silica flux.The silica flux was found to vary with flow rate from 0.00156 moles/s-m2 at 1 LPM to 0.02788 moles/s-m2 at 5 LPM. With no significant difference between refined and unrefined. The particle size and composition were also found to vary with flow rate. The particle size distribution was similar to what was found in industry (75% under 0.01 µm). However, the size of the largest particles was found to decrease with increasing flow rate. Finally the amount of impurities in the fume decreased with increasing flow rate as well. Additionally, in continuation of previous work on small scale oxidation experiments with liquid silicon in a muffle furnace were done and a more in-depth analysis by EPMA was carried out. It was found that addition of calcium increases the oxidation rate, and that the addition of calcium and aluminum results in an oxide layer with multiple phases.

Rapid progress over the last decade on thin film transistor (TFT) active matrix organic light emitting (AMOLED) displays led to the emergence of high-performance, low-power, low-cost flat panel displays. Despite the shortcomings of the active matrix that are associated with the instability and low mobility of TFTs, the amorphous silicon TFT technology still remains the primary solution for the AMOLED backplane. To take advantage of this technology, it is crucial to develop driving schemes and circuit techniques to compensate for the limitations of the TFTs. The driving schemes proposed in this thesis address these challenges, in which, the sensitivity of the OLED current to the transistor variations is reduced significantly. This is achieved by comparing the data signal with a feedback signal associated with the pixel current by means of an external driving circuit through a column feedback line. Depending on the nature of the feedback signal, (i.e. current or voltage) several pixel circuits and external drivers are proposed. New AMOLED pixel circuits with voltage and current feedback are designed, simulated, fabricated, and tested. The performance of these circuits is analyzed in terms of their stability, settling time, power efficiency, noise, and temperature-dependence. For the pixel circuits with current feedback, an operational transresistance amplifier is designed and implemented in a high-voltage CMOS process. Measurement results for both voltage and current feedback driving schemes indicate less than a 2%/V sensitivity to shifts in the threshold voltage of the TFTs. By using current feedback and an accelerating pulse, programming times less than 50 s are achieved.

Detailed material characterization of spherical silicon is conducted for the first time. Experimental results on crystallinity, impurities, and structural defects are presented to investigate the effect of growth mechanism and processing of the spheres. Based on the material properties, the spherical bulk is characterized from electronic point of view. A model is developed to interpret photoconductivity decay measurements and extract minority carrier lifetime, the most influential parameter in device performance. The model includes the spherical geometry as well as the radial profile of carrier lifetime and the measurement results are used to characterize the quality of the spheres and the effectiveness of the process steps. To analyze and predict the performance of spherical cells, a three dimensional opto-electric model is developed. The model separately treats the optical generation and carrier collection and is able to calculate the spectral response and the device characteristics. A simulation tool is created based on this model. The simulation results are of great importance in designing novel structures and optimizing fabrication processes. The necessary characterization methodologies are developed to measure the spectral response and I-Vcharacteristics of individual cells, as well as an array consisting of several cells. A new device design with passivated selective emitter is proposed based on simulation results. Plasma processes were developed to selectively etch back the deep emitter of the spherical cells using reactive ion etching, and then passivate the surface using plasma enhanced chemical vapor deposition of silicon nitride. The improvement in characteristics of the fabricated device is characterized using the developed measurement setups. Industrial up-scaling and manufacturability of the proposed devices and processes are also discussed.

Induction vacuum refining testwork has been carried out for the removal of phosphorus from silicon melts. This work is of interest for the production of solar-grade silicon, since phosphorus is hard to remove from silicon and an important impurity in solar cells.

Active electronically scanned array (AESA) radar systems for military and commercial applications have fueled interest in low-cost, high-performance technologies capable of delivering integrated circuits for transmit-receive (T/R) modules and monolithic radar systems. Silicon-Germanium (SiGe) Heterojunction Bipolar Transistor (HBT) technology has been flagged as a strong candidate for such applications because of its high-speed low-noise devices, high integration capabilities, and relatively low cost. This work investigates integration issues associated with monolithic silicon-germanium radar systems for military (8-12 GHz) and automotive (24 GHz) applications. The design and implementation of critical circuits, such as phase shifters, power amplifiers, up-conversion mixers, down-conversion mixers, and voltage-controlled oscillators will be investigated, along with the system level considerations associated with these components. These building blocks have been fabricated and tested at wafer level, utilizing commercially available SiGe HBT BiCMOS technologies, demonstrating acceptable performance for these applications. Preliminary research into substrate coupling associated with these BiCMOS technologies will also be presented, demonstrating the potential for circuit-to-circuit substrate coupling to occur at these microwave frequencies.

The objective of this work was to design high-speed circuits using silicon-germanium (SiGe) heterojunction bipolar transistors (HBTs) and complementary SiGe (C-SiGe) HBTs, as well as silicon (Si) complementary metal oxide semiconductor (CMOS) devices, for next-generation ultra-wideband (UWB) transceivers. The advantages of using UWB systems over conventional narrowband transceivers include their lower power requirements, higher data rate, more efficient spectrum usage, precise positioning capability, lower complexity, and lower cost. The two major components in a UWB transceiver IC are the radio frequency (RF) circuit and the analog-to-digital converter (ADC). In this work, circuit-level solutions to improve the speed and performance of critical building blocks in both the RF front-end and the ADC are presented. Device-related issues affecting SiGe HBTs for potential applications in UWB systems intended for use in extreme environments will also be investigated. This research envisions to realize various circuit blocks in a UWB transceiver including, a 3-10 GHz UWB low noise amplifiers (LNAs) in both the second (120 GHz) and third (200 GHz) SiGe technologies, an 8-bit 12 GSample/sec SiGe BiCMOS track-and-hold amplifier (THA), and a fifth order elliptic gm-c low-pass filter in C-SiGe HBT technology. This research will also focus on characterizing SiGe HBTs for UWB electronics for operation in extreme environments by investigating the proton radiation effects in the third generation SiGe HBTs.