ELECTRICAL & COMPUTER ENGINEERING DEPARTMENT
UNIVERSITY OF UTAH
ABSTRACTS
OF
STUDENT
PRESENTATIONS
March 29, 2007
SESSION 9 12:05 p.m. Conference Room A
Evaluation of Microwave Office and design of a Quad Band Patch Antenna
Daniel G. Chilinski, Seme Anoka, Chung HimYuen, (Cynthia Furse), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
Providing an evaluation of how well Microwave Office performs and how easy Microwave Office is to use. This is so that it can later be determined if Microwave Office is a good choice for future use at the University of Utah, using accuracy, cost, ease of use and learning curve for the decision factors. The performance of Microwave Office will then be compared with that of CST (Computer Simulation Technologies) and ADS (Advance Design System) to hopefully provide the best fit for the University of Utah. Modeling multiple antennas in each of the different software packages and comparing the simulated results with measured results will be the basis of this comparison. The final stage of this project to design a Quad-band patch antenna with operational frequencies in the GSM cell phone network and the Bluetooth network.
SESSION 9 12:25 p.m. Guest House A
CST MICROWAVE STUDIO EVALUATION FOR ANTENNA DESIGN
Seme L. Anoka (Richard Grow, Cynthia Furse), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112.
In modern microwave education, many pedagogical issues have to be addressed. Generally speaking, a large amount of material is covered in a one-semester course. Therefore, appropriate microwave tools are required to demonstrate microwave circuits with as much ease as possible. Needless to say, software tools are indispensable to microwave engineering and related courses and there is a need to address the use of these tools in the teaching process. Furthermore, with the onset of the wireless revolution, mostly through cellular radio technology, microwave applications now constitute a significant portion of the electronics industry. Cost, time to market, and manufacturing capacity are some of the strongest influences in microwave engineering. The factor of cost versus performance can greatly be enhanced by the use of software tools for Computer-Aided Design (CAD). Hence, researchers and microwave engineers rely on these software tools. As a consequence, the software tools should be used as supplements to microwave courses covered by standard textbooks.
The design and simulation of several patch antennas is performed using CST Microwave Studio for validation purposes. Special emphasis is placed on the design and simulation of an inset-fed rectangular patch antenna. While CST Microwave Studio is the software of choice, the inset-fed rectangular patch antenna will also be analyzed using AWR Microwave Office and/or Ansoft Designer (HFSS). To further appreciate the significance of software tools in the design process of antennas, the inset-fed antenna will be fabricated and its measurements compared with the simulated results.
SESSION 9 12:25 p.m. Guest House A
SIMULATION OF A WIFI SYSTEM USING ADS
Chung Him Yuen (Cynthia Furse), Department of Electrical and Computer Engineering, University of Utah, Slat Lake City, UT 84112
When we design antennas, we would like to simulate our designs before the fabrication of the prototype. There are many different commercial software in the market and we would like to compare them. One of the objectives in this project is to find out the best software packages for Microwave and RF Design for the university. We will rate the software in terms of the ease of usability, user interface and functionality. We are going to use ADS to simulate a WiFi system working at around the frequency of 5GHz and we would like to plot the radiation pattern of the electric field of the transmitter antenna and the receiver antenna. The S-parameters of the system can be found using the software and we can use this to plot the frequency response of the antenna system. We will expect that the value of S11, which is related to the reflection coefficient, will be at its minimum at the desired frequency.
SESSION 9 1:05 p.m. Guest House A
FRACTAL ANTENNA STUDY
Cory M. Miyatake (Richard Grow), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
In recent years, fractal antennas have gained popularity in the antenna and communication systems community. What initially started out as an antenna idea based on Mandlebrot geometry, has become the up and coming solution to multi-band antennas. Fractal antennas have been advertised to have a frequency response that is both broadband and multi-band. In order to investigate the relevance of using fractalization to improve the performance of an antenna, several antennas have been designed, fabricated, and tested. A corner-fed square loop antenna was used as a baseline for performance metrics. Additionally, the square loop antenna was fractalized using the Minkowski Island geometry to create 1st and 2nd iteration antennas. Upon the completion of the design and fabrication, the antennas were compared to see what advantages fractal antennas offer. Furthermore, the results from the simulated and measured antennas were also compared to assess the accuracy of software simulation in antenna design.
SESSION 9 1:25 p.m. Guest House A
EFFECTS OF MINKOWSKI FRACTAL MODIFICATION OF A UHF SQUARE-LOOP PARASITIC ARRAY
Douglas Wetzel (Richard Grow) Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
In the realm of high frequency communications antenna systems with relatively large gain and directivity usually have significant physical dimensions. Techniques for reducing the size of an array typically use reactive loading or some modification of radiator geometry.
This study was undertaken to investigate geometric modification of a square loop parasitic array at 432 MHz, popularly known as a ‘cubical quad’, through the use of Minkowski fractal generation applied to the radiating elements.
Simulation of a standard and two separate Minkowski modifications were done with the public domain antenna analysis product MMANA-GAL. Comparisons of forward gain, front to back ratio, and feedpoint impedance were made. Finally, prototypes of the standard and a modified 2nd order Minkowski generation treatment were built and tested to experimentally confirm the simulation results.
SESSION 9 2:05 p.m. Guest House A
NORTHWEST PIPELINE VERNAL COMMUNICATIONS SYSTEM DESIGN & LINK BUDGET
ANALYSIS
Rob Armstrong (Richard Grow), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
Northwest Pipeline is a natural gas pipeline that stretches from the Canadian border in Washington down through the northwestern states to the New Mexico – Colorado border. A microwave backbone is in place along the pipeline that transmits and receives critical gas control data to and from Salt Lake City where the flow of gas in the pipeline is monitored and controlled. Northwest Pipeline provides natural gas to many different customers. The gas that is purchased by their customers is monitored and controlled at meter stations. These meter stations are controlled remotely in Salt Lake City in the Northwest Pipeline office.
A new meter station has been built in the Vernal Utah area to deliver gas to Questar. This new meter station needs to communicate with the existing microwave backbone so that its data can be monitored in Salt Lake City. The new meter station is located about five miles away from the communications backbone and has a direct line-of-sight path. A communications system has been designed to transfer data from the station to the existing system. The system consists of an FSK radio that modulates data from the meter station. The data is then transmitted using a yagi-uda antenna and received by an omni-directional antenna. The data is removed from the received signal using another FSK radio and sent to Salt Lake along the microwave backbone. Link budget analysis of the system has been performed using both mathematical modeling and software.
SESSION 9 2:25 p.m. Guest House A
REMOTE OPTICAL FREQUENCY DETECTION OF VIBRATING OBJECTS
Brian Randall (Richard Grow), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
A device capable of remotely detecting the vibrating frequencies of objects such as engines, generators, boilers, utility pipes, etc. may be very helpful for determining proper operation of these objects under certain conditions. A device capable of remotely detecting the vibrating frequency of objects was designed, constructed, and tested. By illuminating the objects with laser light and using a photoelectric transistor and designed circuitry to detect and analyze the reflected laser light signal, the operating frequencies of the objects may be determined. Through the use of National Instruments data acquisition (DAQ) and LAbView software, digital signal processing techniques are implemented to filter noise and monitor the received signals. The methodology of this project and results of testing are presented.
SESSION 9 2:45 p.m. Guest House A
FILTERING PULSATILE DATA ACQUIRED WITH A CRITSCAN UNIT
J. Michael Doot (Richard
Grow), Department of Electrical and Computer Engineering, University of Utah,
Salt Lake City, UT, 84112
Current Methods for determining hematocrit in a patient are limited to invasive, time consuming methods. Critscan is a device which determines a patients hematocrit non-invasively and in a timely manner. The current rev. of the critscan machine does not include any signal processing capabilities, the next generation is expected to have this capability, therefore, research must be done to determine the filtering techniques to be employed by the device to more accurately calculate hematocrit. The purpose of this presentation is to discuss the design, and implementation of a filtering system to be combined with the current critscan unit with the use of a DSP board. Included are discussions of the types of filters implemented, the choice of an appropriate domain for the filtering to take place, adaptive techniques employed and design of signals for the purpose of testing. It is anticipated that the addition of a signal processing step in the analysis will increase the accuracy of the device by as much as 5%.
SESSION 9 3:05 p.m. Guest House A
AUTOMATED SYSTEM HARDWARE IMPLEMENTATION
Mike L. Bentley (Richard Grow), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
GE OEC is an industry leader in the medical imaging field. With its mobile C-arm x-ray system, it dominates the market due to quality and reliability. As the technology in this industry grows, so does the potential for the introduction of a wide array of systematic problems. Some of these problems are straightforward with a relatively simple correction, while some are very intermittent and troublesome to find. Many of these problem areas are due to timing or just simply random occurrences without statistical data on the rate of occurrence. This paper's focus is on implementing an automation test device that can significantly reduce the ever-increasing complexity of finding and reproducing these more advanced system issues. This automated system has the functionality of five basic system features; cycling power, creating patient profiles, populating text fields, emulating the keyboard/mouse, and x-ray acquisition. LabView Software simulates the user's input and interfaces with the system via a National Instrument PXI Chassis and mock-ups of various system components. At this time the automated system can create and populate patient data text fields in a third of the time as a human tester. Thus meeting the intent of this automated system to reduce man-hours of product verification and increase the throughput of reliability testing to ensure a more stable and safe product.
SESSION 10 2:25 p.m. Officer’s Club West
IMPLEMENTATION OF STABILIZATION AND NAVIGATION CONTROL SYSTEM FOR AUTONOMOUS R/C AIRCRAFT
Nicholas L. Smith, (Marc Bodson), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
A small R/C model airplane is engineered to host sensors that provide the necessary information to an on board computer for autonomous flight. The design utilizes four thermopile's and a 3 axis accelerometer redundantly to detect pitch and roll angles of the aircraft. In addition a barometric pressure altimeter that provides relative altitude with 4” inch precision. The thermopile's are set for two differential measurements (front to back, side to side) of Earths black body radiation. The accelerometer is used with axis x and y in the complementary configuration to that of the thermopiles. All sensor and GPS navigation data is coded and broadcast to the ground via 1200 baud RF link. The data is both displayed and written to file in real time via Matlab GUI. The planes servo control signals are simultaneously coming from the ground R/C unit and the onboard computer. A custom analog ASIC multiplexes the two sets of servo controls allowing rapid manual takeover of the aircrafts flight. The ASIC also listens for loss of signal from the ground controller in which case flight control is given to the computer until the ground signals are detected again.
SESSION 10 2:45 p.m. Officer’s Club West
IMPLEMENTATION OF STABILIZATION AND NAVIGATION CONTROL SYSTEM FOR AUTONOMOUS R/C AIRCRAFT
Aaron J. Greer, (Marc Bodson), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84117
An embedded computer system was designed and implemented to autonomously control a small airplane. A custom-designed printed circuit board (PCB) was utilized to interface computers and a host of sensors including a GPS receiver, infra-red thermopiles, three-axis accelerometer, and barometric altimeter. The system model was created using Simulink and data collected from the airplane while it was in flight was used to approximate feedback gain values for stability and speed of response. After gathering data and configuring the feedback control loop, the system was tested and verified as operational for short-term autonomous flights.
SESSION 10 3:05 p.m. Officer’s Club West
INTRODUCTION AND SYSTEM MODELING
Adam Gilligan, Charles P Zacher, Vahid Nassersharif, (Marc Bodson), Department of Electrical & Computer Engineering, University of Utah Salt Lake City, UT 84112
Fully articulated suspension vehicles are increasingly being proposed for high-risk, rough train missions and safety, such as planetary exploration, hazardous site clean up, robotic, and military applications. Future planetary exploration missions will require vehicles with adequate mobility, which controlled by computerized processors will be able to perform difficult mobility tasks in rough terrain. Such tasks can normally result in loss of wheel traction, leading to entrapment, loss of stability, and even tip-over. In general, articulated suspension consists of fully automated actuators compensating for disturbances arising in the terrain the vehicle travels over. This project consists of implementing computer-controlled suspension on a RC model car 1/10 that of actual size. Sensors will be mounted and used to measure pitch, twist and roll angles in respect to the body of the car. These measurements will serve as inputs to a DSPIC microcontroller containing the control algorithms for maintaining pitch, roll and twist angles.
SESSION 10 3:25 p.m. Officer’s Club West
ACTIVE SUSPENSION OF A R/C CAR: HARDWARE & SOFTWARE INTEGRATION
Adam Gilligan, Charles P Zacher, Vahid Nassersharif, (Marc
Bodson), Department of Electrical & Computer Engineering, University
of Utah Salt Lake City, UT 84112
The hardware used in order to convert the passive suspension of a R/C car into an active suspension are the following: an accelerometer, current sensors, a PIC controller, servo motors, and various signal filters. The servo motors are the active element, replacing the passive shocks that were initially installed at each of the control arms of the vehicle. The accelerometer is used to measure any type of body roll or pitch. The current sensors measure the change of current being drawn by the servo motors where the measured current is being used as representation of the gravitational force being applied at each of the four control arms. The outputs of the accelerometer and the current sensors are sent through low pass RC filters to filter out some of the noise inherent in the accelerometer and current sensors. Using the conditioned signals from the accelerometer and current sensors, the PIC controller is able to calculate a PWM signal that is outputted to the servo motors and effectively moves the control arms to cancel any disturbances caused to the body of the vehicle.
SESSION 10 3:45 p.m. Officer’s Club West
ACTIVE SUSPENSION OF AN R/C CAR: CONTROL DESIGN AND EXPERIMENTAL RESULTS
Adam Gilligan, Charles P. Zacher, Vahid Nassersharif, (Marc Bodson),
Department of Electrical and Computer Engineering, University of Utah, Salt
Lake City, UT 84112
Considering the goal of zero body inclination by means of an active suspension, three orientation input variables exist that must be controlled. An experimental open-loop response was found for the pitch and roll actuation. From this, the system constants predicted by the modeling can be calibrated, and the nonlinear region of actuation can be identified. In addition, the open-loop response was used to identify the limitations of the system. The controller within the closed-loop system consists of individual proportional and integral terms for each of the orientation variables. An approximation of the controller gains was made from the open-loop response, however final tuning was conducted using the criteria of the shortest rise time with limited overshoot and zero steady state error by trial and error. Final comparison of the system with regard to a passive suspension is given, further demonstrating the physical limits of the implementation but also the success of the system at low vehicle velocity.
SESSION 10 4:05 p.m. Officer’s Club West
DESIGNING AND CONSTRUTION OF PLC TRAINER
Shawn Allred (Dr Mark Bodson), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112.
PLCs (Programmable Logic Controller) are replacing the old hardwire relays, which will take care of many flaws that are seen in relay controlled machines, and are also the first step in achieving automated architecture.
This project is to design and construct a PLC Trainer (software and hardware). A PLC Trainer is a trainer that will assist in training individuals on Ladder Logic code as well as help them understand how Programmable Logic Controllers work. In the designing of the trainer, many aspects were taken into account, to name a few: what is commonly used in a production plant that may use high volume of PLC’s; ruggedness, longevity as well as simplicity (so the design could be easily followed).
As for the code, the challenges may consist of momentary inputs, timed inputs, combination of multiple inputs, latching, unlatching, and several other practical scenarios that are commonly used. The code is to help individuals that have little or no knowledge achieve a good understanding of Ladder Logic, for troubleshooting purposes.
SESSION 11 2:25 p.m. Guest House B
DEVICE VARIATION IN SUB-DESIGN RULE TRANSISTORS
W. Scott Lee, Anthony
Thatcher (Kenneth S. Stevens), Department of Electrical and Computer
Engineering, University of Utah, Salt Lake City, UT 84112
In order to examine the amount of variation as semiconductor devices scale, many NMOSFETs were designed and characterized. These devices had widths ranging in size from 5λ (the minimum active width permitted by the process design rules) to 0.5λ. Using these devices, inverters were built. The PMOSFETs had widths that were twice as large as the NMOSFET widths. The inverters were placed into ring oscillators ranging in size from 3 to 51 inverters. The circuitry was specifically designed to measure transistor parameters such as power consumption, variation, and switching time. The ring oscillators were fabricated using the AMI 0.5 micron process. The oscillators will be tested using different supply voltages, and the results will then be analyzed in order to measure the tradeoffs between variation, reliability, power, and performance.
SESSION 11 2:45 p.m. Guest House B
MEASURED EFFECTS OF NEUTRON FLUENCE ON SEMICONDUCTORS
Ryan Roush, (Angela Rasmussen), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
In order to demonstrate the effects of neutron fluence on semiconductors six National Semiconductor LF353 Operational Amplifiers were tested prior to neutron exposure. These tests included frequency and gain measurements, the recording of output waveforms for 250 kHz & 500 kHz sine waves, and the recording of forward and reverse bias break down voltages for each of the input and output pins. The LF353 Operational Amplifiers were then separated into three groups and subject to neutron fluence, from a TRIGA Nuclear Fission Reactor, at 10kW of power for 10, 20, or 30 minutes. The LF353 Operational Amplifiers were then cooled over a period of two weeks and subjected to the same testing as described above. Subsequent to neutron exposure a decrease in gain and higher bias breakdown voltages were measured. It was demonstrated that the longer the integrated circuit was exposed to neutron fluence the greater amount of defect was measured. These effects are attributed to the chemical changes that occurred inside the Bipolar Junction Transistors junctions, during the neutron fluence. These findings demonstrate the increasing need for radiation hardened component integration into defense and aerospace electronics.
SESSION 11 3:05 p.m. Guest House B
GENERAL PURPOSE INTERFACE BUS (GPIB) DESIGN
Greg Miner, Rick Walker, (Angela Rasmussen), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
In order to avoid discontinuance of the Hart Scientific Model 3560, due to the obsolescence of the GPIB controller previously used, the 3560 was redesigned with a National Instruments GPIB controller. Redesign of the hardware and circuits required the change from a mostly through-hole printed circuit board design to a mostly surface mount design. All existing components were researched and converted to surface mount counterparts, which also required finding vendors for parts that are not already used by Hart Scientific. All schematics, printed circuit board design parameters, circuit designs (create a schematics and layout) were designed for the new components. With the circuit design complete, a bill of materials for all parts on the board was created and sent to the circuit board loading house for quotation and purchase. The printed circuit board design was completed, manufacturing documentation created, loading documentation created, raw board CAD files (Gerber files) created, and the board was sent out for manufacture. Once the hardware design was complete, the new code to interface the microcontroller with the new GPIB controller was developed. Writing the code for this interface required knowing the requirements of the IEEE-488 standards for GPIB communication. Once implemented, the entire design was validated and found to comply with all requirements.
SESSION 11 3:25 p.m. Guest House B
DESIGN OF AN INTELLIGENT MULTIPROTOCOL SERIAL COMMUNICATION MULTIPLEXER
Nick A. Nielson (Angela Rasmussen), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
In order to qualify human machine interfaces (HMI) for use in extreme environmental conditions, a 24-to-1 multiprotocol serial multiplexer is designed. The multiplexer, nicknamed the "mux box," is capable of testing 24 HMI terminals at a time. It controls power and communication with each HMI while the HMI is being tested in an environmental chamber. Two mux boxes can be connected together to facilitate the testing of up to 48 HMI units simultaneously. A wide range of product design concepts were practiced including schematic design, PCB layout, Verilog HDL development, and final system testing. The control system of the mux box consists of an Altera Cyclone FPGA which listens to commands from the host over the I2C bus and performs actions based on the data received. After debugging both hardware and software, a reliable testing platform for human machine interfaces is in constant use.
SESSION 11 3:45 p.m. Guest House B
ELECTRICAL DESIGN OF THE LDS PHILANTHROPIES BUILDING
Christopher K. Kobayashi, Ken G. Garner, (Angela Rasmussen), Department of Electrical Engineering, University of Utah, Salt Lake City, UT 84112
The complete design of electrical systems for the new construction of the LDS Philanthropies building located in Provo, Utah was successfully engineered. This included the engineering of: power service, power distribution, telecom service and distribution, lighting and lighting controls, fire alarm system, and security system. Integration and coordination was required between service providers such as Provo City Power, Qwest, and BYU. Careful attention was used to design a properly sized electrical system with backup power for egress lighting. The design was done under all applicable codes including but not limited to: NEC, IBC, IMC, IFC, IECC, NFPA 72 and all applicable state and local codes. The building was designed around a three-phase, 4-wire, 120/208 Volt, 60 Hz power system and the main service was sized at 2000 Amps. After redesign of the mechanical system for the building it may have been advantageous to change the electrical system to a 277/480 volt service, although this was not done.
SESSION 11 4:25 p.m. Guest House B
PROFESSIONAL AUDIO MULTI-CHANNEL DECODER
Shawn R. Nageli (Angela Rasmussen), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
Multi-channel audio signals such as Dolby Digital (DD) and Digital Theater Systems (DTS) which use the IEC61937 digital standard for transmission are commonly found in decoders which do not easily conform to proper digital and analog connections required by professional audio (PA) equipment. The purpose of this project was to develop a decoder that accepts these multi-channel data audio signals and then converts them for use in the professional audio application. The decoder accepts these multi-channel audio data streams through either an optical, coaxial, or AES/EBU digital connection. This data stream is then sent through a proprietary Digital Signal Processor (DSP) which decodes the audio signal into four (I2S) audio data streams. These streams are then sent to four digital transmitters and four D/A converters providing eight channels of audio in both a digital and analog format. The digital output of the decoder conforms to the IEC60958 AES/EBU digital standard while the analog outputs provide a differential balanced signal that contains the common voltage and impedance characteristics found in a PA application. Both of these outputs use a standard XLR connection for external connectivity. This type of decoder will allow these multi-channel signals to be more easily used in a wider variety of applications.
SESSION 11 4:45 p.m. Guest House B
NICKEL SILICIDE CHARACTERIZATION ON
SCHOTTKY DIODES
Michael Gonsalves (Angela Rasmussen) Department of Electrical and Computer
Engineering, University of Utah, Salt Lake City, UT 84112
The effect of pre-Nickel deposition ion implantation is investigated to
determine if a more thorough phase transformation will occur during
silicidation. Ion implantation on test wafers of four different species
is performed. After ion implantation the test wafers undergo a nickel
deposition; a pre sheet resistance measurement is then taken. The wafers
are subsequently thermally annealed and a post measurement is taken. The
difference of these measurements is examined to see if there has been a
decrease in the test wafers overall sheet resistance. Following this
initially test run a DOE or design of experiments is explored to determine what
variables from the initial test run are statistically significant and which
variables are statistically insignificant. The results from the DOE lead
to a second test run in which test wafers follow a similar path as stated
above. Following various evaluations to interpret the results the experiment
is performed on a Schottky diode product lot and the reverse leakage current is
measured.
SESSION 11 5:05 p.m. Guest House B
TITAN 40W CELL REDESIGN
Tyler Squire, DJ Burton, Jason England, Justin McKinnon, Justin Prince (Angela Rasmussen, A. K. Balaji) Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
The Titan 40W testing cell redesign project was to automate the rigorous testing procedure that each unit undergoes. The Titan 40W is a high-voltage three-phase, surge suppression device. It is designed and manufactured at EFI Electronics, which is a part of Square D. A voltage surge of 1300 V is used to test whether the unit suppresses the voltage correctly. Another test is a full power test in order to verify that the unit functions properly and that the LEDs on the front of the unit light up when the device is working. Previously this test was done by hand and required the full attention of the employee at the test station. The automation was done by using a programmable logic controller (PLC) to run the test and by designing a device that holds the unit and shears the wires when finished. The main goals of the redesign were to simplify the testing process in order to minimize human error, increase productivity and improve the quality of the product that is being manufactured.
SESSION 11 5:25 p.m. Guest House B
FABRICATION AND CHARACTERIZATION OF A POLYSILICON MOSFET AS A SPINFET BASE
Daniel T. Watrous (Mark Miller, Justin Jackson, Divesh Kapoor), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, Utah 84112
Interest in developing nano-scale devices that manipulate both the charge and spin of electrons has fueled much research in recent years. To be commercially viable these devices will require proven and reliable means of fabrication. One solution has been proposed in which a traditional MOSFET device is modified to create a periodic potential in the inversion layer when actively biased. The period of this potential is on the order of 10nm. The present research works to establish the basis of the spin device by fabricating a suitable, small-scale NMOSFET device and characterizing the thickness of the polysilicon, and oxide layers. Layer thicknesses of less than 5nm were obtained. Very high doping was applied to both the initial substrate on the order of 2x1017 cm^-3, and in the polysilicon on the order of 2x1019 cm^-3. Both aluminum spiking due to annealing and depletion of the polysilicon layer during characterization were observed. Fabrication and characterization are presented along with some background NMOSFET and SpinFET theory.
SESSION 11 5:45 p.m. Guest House B
Demonstrations on the Concepts of Semiconductor Physics
Arezo Jamaleddin (Angela Rasmussen), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
This project is an introduction to semiconductor devices for undergraduate electrical engineers, other interested students, practicing engineers and scientists whose understanding of modern electronics needs updating. It is organized to bring students with a background in sophomore physics to a level of understanding which allows them to read of the current literature on new devices and applications. I am responsible to develop a suite of visualization demonstrations for explaining concepts of semiconductor physics. These demonstrations will be compiled and made available on the web. Each demonstration would be appropriate for a sophomore level. There are ten demonstrations that I am responsible for: Schrondinger Equations; Band structure 3- Fermi-Dirac Function; Carriers( holes and electrons within a structure) in intrinsic and doped material; Drift velocity and applied field; Drift and Diffusion; PN junction under bias, BJT operation; MOSFET operation; and JFET/MESFET operation. The main purpose of the project is to explain basic concepts on the semiconductor physics and to be utilized in the semiconductor courses for future students in the Department of Electrical Engineering at University Of Utah.
SESSION 12 12:05 p.m. Guest House B
STREAMING AUDIO OVER WIRELESS CONNECTION
Javier A. Najar (Gil Shamir), Electrical and Computer Engineering Department, University of Utah, Salt Lake City, UT 84112
A personal computer streams digital audio data to a remote audio receiver by using Bluetooth modules that conform to specification version 1.1. The modules are set to transparent mode so that the link behaves like a wireless UART serial port; this simplifies the system and makes it more flexible since it doesn’t have to adhere to other Bluetooth profile requirements. Because the data rate of the module is limited to less than 721 kbps, it is necessary to compress the high-quality audio data bit rate of 1411 kbps to less than 320 kbps using a suitable codec like MP3, AAC, Ogg Vorbis, etc. An embedded processor with the decoders implemented recognizes the compression format used and decodes the stream to PCM data. After the processing, the digital audio data is sent to the external audio device by means of SPDIF in digital form and/or converted to analog by DACs.
SESSION 12 12:25 p.m. Guest House B
ROCKBOX ON THE TOSHIBA GIGABEAT
Luis F. Gomez (Gil Shamir), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
In order to realize the full potential of
commerical digital audio
players, a new, open-source, replacement firmware project (Rockbox) spanning
numerous different models has been started. Preliminary work with low-level
device drivers pertaining to the Toshiba Gigabeat F-60 branch of the project is
outlined. This is a small part of the higher goal of porting the full
replacement firmware onto the device.
SESSION 12 12:45 p.m. Guest House B
VOIP NETWORK SIMULATION
J. Russel Lindsay (Gil Shamir), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
This project describes detailed research of Internet protocol network simulation algorithms, and the integration of those algorithms into an actual network simulation tool. The network simulation tool focuses providing a test environment for Voice over Internet Protocol (VoIP) trouble shooting. The main VoIP challenges that are addressed are network delay, packet loss, and jitter.
SESSION 12 1:25 p.m. Guest House B
Robustness of VoIP Codecs to Network Packet Loss
Jake E. Hunter (Gil Shamir), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
This project examines the effect of network packet loss on the voice quality of a Voice over Internet Protocol (VoIP) system. The strategy utilized was to test various VoIP coders/decoders (codecs) against levels of packet loss varying from 0-50% loss to find the codec that produces the clearest voice quality over a wide range of packet loss. The voice quality was determined using Mean Opinion Scoring (MOS) by a small group (10-20) of people. Logic suggests that the codec that produces the smallest packets will be the least affected by packet loss. The results of this experiment will serve as a proving ground for this concept.
SESSION 12 1:45 p.m. Guest House B
FILTERING DENTIST DRILL FREQUENCIES
Bryant L. Jewkes, (Gil Shamir), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
Many dentists are losing their hearing due to the high frequencies produced by the dentist drill. In order to overcome this problem a device that will filter out the frequency ranges of the drill needed to be designed. This was implemented using a circuit that can be micro fabricated in the future, as an earpiece or some other type of hearing device that a dentist would wear. This circuit, however, will still allow the frequency ranges of the human voice to pass through it, which will provide uninterrupted conversation between dentist and patient. To accomplish this task, many sound samples were taken and analyzed to find the range of frequencies that each sample covered. First, the sound of the dentist drill was recorded, as well as, the voice samples of speech for adult males, females, and children. Through Fourier Transform, performed in Matlab, the drill frequency was found to be 2.15 kHz. It was also determined that most human speech is less than 1.5 kHz. The final circuit implemented an amplifier stage with a second order notch filter to filter out the frequency of 2.1 kHz, while clearly passing frequency of the human voice.
SESSION 12 2:05 p.m. Guest House B
Resolution of the Gibbs Phenomenon
Andrew Nelson, (Gil Shamir), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
In some scientific and engineering applications such as data compression, tomography, and fluid flow, one often wants to reconstruct a signal for a set of Fourier coefficients or its orthogonal polynomial expansion. For physical systems that are well-behaved, this isn't a problem--but what if the signal one wants to reconstruct isn't periodic, or even completely continuous? This talk will discuss some methods for quickly resolving these signals, including Cesaro sums and robust reprojection bases for Gegenbauer and Freud polynomials, and how current results may be extended to other orthogonal systems like Legendre and Chebychev polynomials. A comparison of the accuracy of each method, and numerical evidence in support of the robust reprojection method, will also be included and discussed.
Session 13: Individual Projects
SESSION 13 1:25 p.m. Officer’s Club East
CHORUS EFFECT THROUGH AUDIO DSP
Leo S. Aksu (Ken Stevens), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
Chorus is a common audio effect which makes a single instrument sound like two or more identical instruments are being played at the same time. This is done by slightly delaying and pitch shifting the original input signal and playing the processed signal along with the original signal. This mimics what would happen physically if a second identical instrument was present a small distance apart from the original instrument since the sound waves from the second instrument would be slightly delayed and pitch shifted when they reach the listener. The chorus effect was implemented by routing the incoming audio signal through an Analog Devices ADZS-21364-EZLITE DSP evaluation board. The incoming signal was digitized (A/D) and the digitized signal was processed by the Analog Devices SHARC ADSP-21364 DSP chip. First, a delayed version of the original signal was created in the internal buffer of the DSP chip. The amount of delay was then modulated by a 1 Hz sine wave to create the desired pitch shift effect. The processed signal was finally mixed with the original incoming signal in appropriate proportion and the resulting signal was converted back to analog (D/A). The A/D and D/A conversions were made at 96 kHz with 24-bit precision with the Analog Devices AD1835 CODEC chip. A three minute audio demo will be given to demonstrate that the implemented chorus effect has created a richer and more pleasant sound.
SESSION 13 1:45 p.m. Officer’s Club East
A Pipelined Hardware Implementation of a Modified Tiny Encryption Algorithm
Anthony Thatcher, (Ken Stevens), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
The modified Tiny Encryption Algorithm (TEA) presented here is used as a "real world" implementation test of some experimental transistors. The algorithm is modified to use a sixteen bit rather than a sixty-four bit key, and operates on eight bit rather than thirty-two bit data inputs. This simplification was made in order to simplify the hardware, and allow it to be more easily testable. The hardware implementation uses a pipelined approach, in order to allow greater throughput at the expense of chip area. The test chip has two instances of the TEA hardware, one of those implementaed with the minimum recomended sizes of transistors, and the other composed of transistors with the minimum width allowable by the design rules of the .5 micron process used to fabricate the chip.
SESSION 15 3:45 p.m. Guest House D
REALIZATION OF COMPLEX HETEROGENEOUS MICROSYSTEMS: INTEGRATING DESIGN, COMPUTER MODELING, AND RAPID-PROTOTYPING
Ronald W. Boutte, Taylor M. Meacham, Nathaniel Gaskin, Brian Baker, Justin A. Horn, Kimball Ward (Ian R. Harvey), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
Efficient design of heterogeneous microsystems can be tricky and expensive because the design must account for on-chip mechanical transducers, actuation mechanisms, electronic circuitry, and occasionally optical components. Because the scale of these components can be 100's of microns to as small as 10's of nanometers, the system-level integration of these components relies heavily on several engineering disciplines and careful selection of design software. A visualization and design approach case study is presented for complex microelectromechanical system (MEMS) first realized as a 3D model in ProEngineer, rapid prototyped using standard 3D printing (macro-model), converted to individual mask layers using AutoCAD, and then fabricated in Sandia National Laboratories' SUMMiT-V(TM) process. A working closed-loop micro-deployable structure (hoberman-style iris) was designed in one semester and demonstrated in silicon just a few months later, owing to this increasingly important design methodology.
SESSION 15 4:05 p.m. Guest House D
CHARACTERIZATION AND TESTING OF A MICRO-DEPLOYABLE, POLYSILICON IRIS STRUCTURE
Taylor Merrit Meacham, (Ian Robert Harvey), Department of Electrical and Computer Engineering, University of Utah, Salt Lake City, UT 84112
During the Spring Semester of 2006, a micro-deployable iris structure was designed by a team of students from the University of Utah as part of an entry into the 2006 Sandia MEMS University Alliance Design Competition. Sandia’s CAD tools were used to design the structure to be fabricated used the SUMMiT V process, a process using five mechanical polysilicon layers, four sacrificial silicon oxide layers, and CMP planarization to create complex mechanical systems on a micrometer scale. The device was designed to be opened and closed by electrostatic comb-drive microengines. The fabricated iris’s diameter is slightly less than the thickness of a dime, or approximately 1.2mm. Construction analysis and characterization work has been performed on the microscale iris, using the facilities and equipment in the Utah Microfabrication Core Laboratory. This work has included visual examination with the use of optical microscopy, scanning electron microscopy, and optical profilometry. Some destructive analysis techniques, such as manual polishing, were used to cross-section into specific areas of interest, such as the iris’s unique polysilicon pin hinges. Functional testing was also conducted using the facilities of Sandia National Laboratories in Albuquerque, NM. Optical microscopy was again used to allow visual analysis during operation of the device and to capture video for later analysis. During testing, the iris was successfully actuated as designed multiple times. In addition, a few design problems were identified in order to improve the performance of the device for future applications.