Summer 2000

Project Archive

Lauren Berryman (Electrical Engineering) - University of Pennsylvania


Advisors: Dr. Nader Engheta, Dr. Edward N. Pugh
Collaborators: Claire Daniele, Lauren Daniele

ABSTRACT: This paper describes the electroencephalographic detection and analysis of scalp potentials utilizing fMRI brain volume data to design stimuli that activate specific visual areas and pathways. Our dual purpose was to re-evaluate current scientific models and to investigate the origin of neural activity in response to a specified stimulus. Through data collection and analysis, we tested the widely used three-shell conductivity model of the head. We concluded that this model is inappropriate for source analysis due to inherent model flaws and the disagreement of our experimental observations with model predictions. We approached the problem of dipole current source analysis by first pinpointing the source location within the visual cortex using collected metabolic fMRI data. Then, with our stimulus designed to activate that identified area of cortex, we recorded electromagnetic scalp distribution patterns. By examining the collected sensory evoked potentials with Fourier analysis, we found frequency components present in the scalp recording at twice the stimulus flicker frequency. By tracing frequency response patterns and identifying corresponding contrast response parameters, we were able to show that for our stimulus, the magnocellular pathway was dominant in the flow of visual information from retina to cortex. View Paper | View Slides

DeAnna Burns (MEAM) - University of Pennsylvania


Advisors: H. Bau (MEAM, Penn), J. Tanaka (Medicine, Temple), P. Mueller (BioPhysics, Penn), J. Van der Spiegel (EE, Penn)

ABSTRACT: Cells are electrically active; thus fabrication of a cell-based biosensor will contribute much information about how cells respond to different environments. With such a device, researchers can submit various types of cells to different forms of stimulus and observe the cells’ reactions. I designed and fabricated a biosensor out of DuPont Green Tape ceramics and thick film conductor materials, working with human embryonic kidney cells in order to test the compatibility of cells with various types of materials. I found that DuPont’s gold paste conductor is harmful to cells and that pure gold is needed in order for cells to attach and live on the sensor’s electrodes. Next, I turned to sea urchin eggs, which are larger and more robust than most cells. I constructed a device containing channels with two pure gold electrodes covered with a thin insulating layer of Teflon. A small portion of the Teflon removed over each electrode provides the necessary small area needed for an electrode that is used to measure the electrical activity of cells. Although I was unable to gain many meaningful electrical measurements, an improved sensor that is constructed out of more compatible materials should be constructed and used to test a cell’s electrical response to a change in environment. View Paper | View Slides

Frederick U. Diaz (Electrical Engineering) - University of Pennsylvania


NSF-AMP Undergraduate Research Program
Advisors: Dr. Dwight L. Jaggard and Aaron D. Jaggard

ABSTRACT: This research project is based on previous work done by Dwight L. Jaggard and Aaron D. Jaggard on fractal ring arrays. We utilize fractals in the fabrication of a class of Cantor ring arrays. Fractal descriptors such as dimension, stage of growth, and lacunarity are applied in designing and characterizing these Cantor rings. Using the continuous arrays developed by Jaggard and Jaggard as a point of reference, we examine the design of analogous discrete arrays. We examine various ways of periodically and randomly thinning and building up these arrays azimuthally. The arrays are compared to their periodic and random counterparts as well as the continuous case to evaluate performance. Mainbeam quality, sidelobe level, and visible range are used to rate the arrays. The Cantor ring arrays and the corresponding array factors are simulated using programs developed in MATLAB. Our goals include developing fractal arrays that have low sidelobes comparable to those of periodic arrays while maintaining the robustness of random arrays. We aim for superior fractal performance at an equal number of elements as the periodic and random cases, as well as comparable performance using fewer elements. View Paper | View Slides

Héctor E. Dimas (Electrical Engineering) - University of Pennsylvania


NSF-AMPS/SUNFEST Summer Undergraduate Research
Advisors: Dr. Dwight L. Jaggard, Aaron D. Jaggard

ABSTRACT: This paper describes the use of fractal arrangements for the design of planar antenna arrays. Spiral initiators were used to play the chaos game, which generated our fractal array. Once these antenna arrays were produced, the resulting radiated field was calculated using scripts written in Matlab. Variations on the number of arms and inner circles in our spiral initiators were tested to characterize which formation would allow for our fractal array to perform equivalently or better than a set of random elements placed on a given circular area. The compared random array held the same number of elements as our spiral array did which was usually 441. The resultant spiral array behaved much like tapered arrays in that the concentration of points tended towards the center. The overall behavior of the fractal array in comparison to the random array was comparable in terms of directivity and average sidelobe level. Additional attempts to make the spiral array less tapered allowed for possible characterization as a “multi-fractal” array by changing the number of points that made up inner circles of the initiator spiral. These results also proved to be comparable to random and spiral arrays. Implications of these results and further exploration needed are discussed. View Paper | View Slides

Xiomara Feliciano-Diaz (Mechanical Engineering) - Turabo University, Puerto Rico


Advisor: Dr. Jorge J. Santiago-Aviles

ABSTRACT: Hip fractures can be a life-threatening injury among the elderly. Annually, thirty million falls are reported, out of which two hundred and fifty thousand result in hip fractures. Medical costs for these fractures are estimated between seven and ten billion dollars annually. However, the number of hip fractures is expected to double in the next fifty years, which defines the current problem as critical and alarming. In order to provide a suitable solution to this problem, Dr. Jorge Santiago from the Electrical Engineering Department at the University of Pennsylvania has developed a project to build such device. The product will be a belt-like device that would prevent hip injuries by means of sensors that will recognize a potentially injurious fall and then trigger an inflatable cushion before impact. View Paper | View Slides

Jason Gillman (EE) - University of Pennsylvania


Advisors: Dr. J.J. Santiago-Aviles, P. Espinoza

ABSTRACT: This research attempted to fabricate a gas-diverting valve using LTCC technology for the emerging field of meso-scale devices. A photolithographic process for patterning Low Temperature Co-Fired Ceramic (LTCC) tapes utilizing DuPont Green Tape? tape coupled with DuPont Riston? 9015, a dry photoresist, is described in detail. Based on the device design, calculations were made determining the theoretical resistance and magnetic flux were made. These numbers will determine the voltage necessary to run the device once built. Although the device was not constructed since the resolution of the dry photoresist system was being pushed led to difficulty in etching the tape, lamination and development techniques of the Riston? were perfected, and insights into the composition of the tape prior to and after etching were made. Suggestions for improvements in the process are given, as well as thoughts on what should be the next step in this research. View Paper | View Slides

Tamara Knutsen (EE) - Harvard University


Advisors: Professor Jim Ostrowski (GRASP Lab), and Kenneth McIsaac (GRASP Lab)

ABSTRACT: In this report, I investigate the design and development of an eel-like amphibious robot. After reviewing research on dynamic mobile robots and control and motion planning for biomimetic robot systems, I analyzed the prototype created by Kenneth McIsaac and Jim Ostrowski and redesigned its key mechanical, electrical and communication components. The significant changes I implemented affected the number of links of the robot, its waterproof characteristics and the communication design (both hardware and software). I designed and manufactured the mechanical components using the state-of-the-art CAD/CAM facilities at the University of Pennsylvania and assembled and waterproofed the prototype at the GRASP laboratory. McIsaac and I made the necessary modifications to the open-loop control of the robot and at the time of writing this report, the prototype seems to be functioning as expected. The closed loop control of the robot is still being developed at this time. View Paper | View Slides

Heather Marandola (Electrical Engineering) - Swarthmore College


Advisor: Jim Ostrowski

ABSTRACT: This paper describes the design of an accelerometer system to be used for the transmission of collision forces during the course of a contact sporting event. Discussed in this paper are basic accelerometer and transmission principles, the development and results of data acquisition experiments and the design of three evaluation boards: accelerometer, transmission and reception. My work has led to the design of preliminary calibration and display techniques for dual axis accelerometer outputs, and I have produced evaluation boards which can be utilized in further experimentation. Though a secure transmission link was not established, progress has been made in the development of the transmission scheme to be implemented on the device. Measurement along three axes was not implemented; however, with minor revisions, the evaluation boards are capable of handling one, two, three or four axes. The device is not ready to be installed into the sporting arenas, but some of the principles and design ideas presented in this paper might be of use in future implementation. View Paper | View Slides

Charlotte Martinez (Electrical Engineering) - University of Pennsylvania


Advisors: Dr. J.J. Santiago-Aviles, P. Espinoza

ABSTRACT: A glucose sensor using silicon and platinum was designed and fabricated. The interaction of glucose oxidaise and glucose produces proteins, which have dielectric properties that produce an electrical output measured through the bonding pads of a sensor. The sensor consists of two layers of platinum: one porous layer that serves as an area for the reaction to take place, and a second layer, that consists of two bonding pads. The sensors were placed in solutions of different glucose concentration. Their resistance was measured. It was found that in a solution of less concentration, the resistance was greater. We also researched non-invasive ways of measuring glucose. Technologies used included Raman spectroscopy, near-infrared spectroscopy, photo acoustic spectroscopy and mid-infrared spectroscopy. View Paper | View Slides

Julie Neiling (Computer Engineering) - University of Evansville, Indiana


Advisor: Dr. James Ostrowski

ABSTRACT: This paper describes the investigation of the accelerometers, gyroscopes, speaker, and microphones on the Sony AIBO, which is a robotic dog. The purpose is to use these sensors to determine the position of the robot during a soccer game at RoboCup. The gyroscope generates data used to tell rotation; accelerometer data is used to calculate distance traveled; audio communication can be used to pass messages that relay to its position on the playing field to the other robots. The sensor data will be incorporated into the robot AI through individual software modules. The primary emphasis of the paper is the research involved in development of the audio communication. To accomplish the communication we used various frequency analysis and signal processing tools to make the robot produce tones and correctly identify them. View Paper | View Slides

Shiva Portonovo (Electrical Engineering) - University of Pennsylvania


Advisors: Dr. Jan Van der Spiegel and Dr. Nader Engheta

ABSTRACT: The problem of perceiving objects that are suspended in scattering media limits the ability of humans to derive information about their environment. This can be the cause of several important dangers to personal safety, while it can also prevent people from learning about their surroundings. >From studying the visual systems of animals that have a better sense of perception in such media, the concept of polarization-difference imaging was born. A system that employs this technique uses the polarization properties of light in novel ways to produce images that detect objects in scattering media and make surface features discernable. Such a system is a polarization-difference camera. A polarization-difference camera can consist of a polarization analyzer, a solid-state imager, and an image display system. The polarization analyzer separates detected light into orthogonally polarized beams. The solid-state imager transduces the light signals into electronic signals and processes the data to produce the polarization-difference output and other necessary operations. The image display system transforms the data so that images can be optimally displayed. This report focuses on the design aspects of the solid-state imager for a polarization-difference camera. The solid-state imager may be constructed with complementary metal-oxide semiconductor (CMOS) devices, currently the most widely used variety of devices in integrated electronic systems. The pixels used in the imager, which employ active transistors, are known as active pixel sensors. A set of several designs of components that may be used in the solid-state imager are described and analyzed. Suggestions for optimization for an integrated design follow. View Paper | View Slides