University of Manitoba Department of Electrical & Computer Engineering. ECE 4600 Group Design Project. Progress Report. Microwave Imaging.

Size: px

Start display at page:

Download "University of Manitoba Department of Electrical & Computer Engineering. ECE 4600 Group Design Project. Progress Report. Microwave Imaging."

Dylan Young
5 years ago
Views:

1 University of Manitoba Department of Electrical & Computer Engineering ECE 4600 Group Design Project Progress Report Microwave Imaging by Group 12 Steven Brown Trevor Ingelbeen Brett Trombo Bryce O Donnel Steve Demedash Devon Hudson Academic Supervisor(s) Dr. Joe LoVetri Industry Supervisors Ian Jeffrey 151 Industries Colin Gilmore 151 Industries Paul Card 151 Industries Date of Submission January 12, 2015 Copyright 2015 Steven Brown, Bryce O Donnel, Trevor Ingelbeen, Steve Demedash, Brett Trombo, Devon Hudson

2 TABLE OF CONTENTS Table of Contents 1 Introduction Project Progress and Future Work Field Coil Antennas Software Conclusions Appendix A Updated Gantt Chart Appendix B Updated Budget Appendix C System Diagram i

3 1 Introduction 1 Introduction The Electromagnetic Imaging Lab (EIL) at the University of Manitoba is researching breast imaging systems for the purpose of non-invasive breast cancer detection. The imaging systems researched in the EIL use microwave imaging (MWI) technology to detect breast cancer as an alternative to MRI and mammogram technology. The first MWI system contains a field coil capable of producing 0.2 Tesla magnetic field and Group 12 is designing the hardware for this project. The material costs of this project have driven a scope change therefore this project is no longer progressing according to the original plan. The second MWI system requires a dual polarized antenna array to produce a 3D image of the healthy and cancerous tissue within a patient s breast. During the design phase of the dual polarized antenna, Group 12 encountered significant project scope change and is no longer progressing according to the original plan. Lastly, the EIL requires control and monitoring software for its devices. Group 12 is developing new software for the devices mentioned above as well as improving existing software within the EIL. The software project is progressing as originally planned. 2 Project Progress and Future Work The progress for group 12 from the past four months and planned future work is described below for each of the three project components, with our new tasks outlined in Appendix A. To view a system block diagram refer to Appendix C. 2.1 Field Coil In the past four months, Group 12 has divided the magnetic field coil design into three sections: field coil, cooling system and power supply. Bryce and Trevor simulated the magnetic field with Matlab to design the coil size and electric current required to produce the magnetic field. The 1

4 2 Project Progress and Future Work simulation determined that a magnetic field of 0.2T could be achieved by feeding 500 Amps of DC current though a field coil. Bryce, Joe LoVetri and Dan Card determined that from an economic standpoint, using copper refrigeration tubing was the best material for the copper conductor and cooling system. Refrigeration copper is malleable, readily available, and can be manually coiled without outsourcing to a manufacturer. The cost of the copper conductor required for this project created a scope change for project. Group 12 and its supervisor removed the manufacturing of the entire field coil from the project scope. Instead Group 12 will now produce a 1/16th scaled proof of concept test and if time and money permits, the remainder 15/16th of the project will be built. The cooling system design was completed in parallel with the field coil design by Bryce and Trevor. Originally the cooling system was proposed to be completed after the coil design but due to the high importance of the cooling system it was incorporated into the field coil design. The high amperage required to produce the specified magnetic field will generate a large amount of heat. The heat will affect the resistivity of the coil and potentially damage the coils insulation. The cooling system maintains a safe operating temperature by circulating a coolant through the hollow interior of the field coil conductor. The power supply will provide the field coil with 500A which will produce the magnetic field. Trevor designed the power source using lithium batteries because Group 12 had access to these batteries. Trevor and Bryce designed the power source to have a freewheeling diode to allow a gradual current decay across the inductive coil when the device is turned off. A properly rated controllable switch was also used in the power supply so that the GUI control software can turn the source on or off. In the next two months, Trevor and Bryce will assemble the individual components required to build and test the entire field coil, power supply and cooling system. A complete list of parts and their expected delivery date is located in the budget section of Appendix B. The first component that Trevor and Bryce will assemble is the power supply. Jumper cables, capable of withstanding 500A DC are required to connect the field coil, freewheeling diode and switch to the batteries. 2

5 2 Project Progress and Future Work Custom jumper cables will be built by Trevor and Bryce because they are less expensive than purchasing them from a supplier. Next, the batteries, freewheeling diode and switch will be installed, completing the assembly of the power supply. Prior to installing the switch, the interface between the GUI and switch will be assembled and tested by Trevor, Devon and Bryce. The GUI will control the on/off status of the magnetic field. During the assembly of the power supply, quality control measures will be implemented to determine the circuit properties and verify the functionality of the control switch. The next component that Trevor and Bryce will build is the interface between the power supply and field coil. The preferred method to electrically connect the field coil and power supply will require a soldered connection arm between the field coil and the cables from the power supply. If the soldered connection arm fails, a ground clamp will be used to connect the field coil to the power supply cable cables. Next, the cooling system will be assembled by Trevor and Bryce. The plumbing from the pump outlet to field coil will be assembled and tested to verify the flow rate of the cooling system. Finally, Trevor and Bryce will assemble the entire field coil system which consists of the field coil, power supply, and cooling system as shown in our updated Gantt Chart in Appendix A. All electrical connections will be terminated and the entire system will be tested to verify that the system is operating as intended. If required, troubleshooting will occur at this point. 2.2 Antennas The antenna portion of the project can be separated into three areas, the cavity resonator, the printed circuit board (PCB), and the control system. Thus far Group 12 has researched and preformed various simulations of time electric (TE10) mode resonating cavities. From these simulations, Brett and Steven Brown were able to create the polarization of an antenna in both orientations. The number of slots are not increasing in the antenna cavity as proposed due to the orthogonal orientation of the cavities. Brett and Steven Brown explored higher order modes 3

6 2 Project Progress and Future Work of waveguides, which does increase the physical size and complexity of the feed arrangement. Although more challenging, we have determined that a waveguide using higher order modes is better as it allowed both polarizations of antenna slots to be on one resonating waveguide cavity. Brett and Steven Brown have designed multiple resonating waveguide cavities using high frequency simulation software (HFSS). The HFSS simulations show a simulated TE11 resonating cavity with the reflection coefficient or return loss. The1 HFSS simulations have calculated the gain loss coefficient to dB, which exceeds our project specifications. Brett and Steven Brown are currently optimizing the antenna design by adjusting the input feed position and slot length with a method called optometrics. This is a process in which the model is simulated in the HFSS CAD Tool and the position of various variables are adjusted. The position and length of the coaxial feed that satisfies the most requirements are saved for the user. Brett and Steven Brown have decided to minimize the return loss of the resonating cavity to ensure that the maximum radiation would propagate from the resonating cavity as well as, maximize the surface current density on the PCB ground plane. Brett and Steven Brown are actively working on slot schemes (positioning of slots in the PCB) that cut the surface current density to maximize the radiation in the desired polarization. By selecting to cut the slots on one resonating cavity as opposed to making two, saves space in the antenna chamber because the resonating cavities with one polarization take up horizontal and vertical space. The TE11 resonating cavity occupies mainly one orientation which we can select. A singular resonating cavity also enables us to modify the existing slot selection system by simply running through all combinations of receiver and transmitter in one orientation followed by the other orientation. For the remainder of the course, Group 12 will concentrate on three antenna system components: the cavity resonator, PCB, and control system. The TE11 mode resonating cavity must adapt the design to the specific application. Manufacturing, mounting, and mechanical limitations must now be balanced with the best theoretical design. After final mechanical dimensions are selected the appropriate slot schemes will be selected and simulated to verify the new 4

7 2 Project Progress and Future Work dimensions. The best slot scheme from the simulations will be used for manufacturing. There three major system components that still have to be manufactured as seen on the updated Gantt Chart in Appendix A. After the resonating cavity design is finalized, the PCB design can be sized to meet physical constraints of the antenna chamber. Once the PCB size is determined it will be manufactured by an external vendor. The cavity resonator is purchased at the required dimensions and the one side is removed then replaced with the PCB at the University of Manitoba. The existing control system in the EIL is retrofitted to work with the new design. Next testing will be conducted on the input impedance, radiated power, and slot control of the antenna. Input impedance will be measured using a network analyzer and verified with the expected design results. If time permits the antenna will be tested in an antenna lab to determine radiation pattern and radiated power. The fall-back method will be to test the antennas in the prototype application system and results measured by the control system. Integration with the software control system will be the last testing procedure after the main components have been tested individually. 2.3 Software In the past four months, Group 12 has worked on the integration of the existing data acquisition and control software currently used by the EIL into a single software package. Individual VNA and LIA devices within the EIL are controlled by multiple versions of the control software. By consolidating these versions, a single version will exist that is capable of running a variety of different devices, allowing the user to select which device is to be used. To begin, Devon and Steve Demedash modified the existing versions of the separate software by removing extraneous comments and lines of code from the existing code. Devon and Steve Demedash removed large blocks of code that constituted individual sub-processes from the main loop, and created separate functions which increased the modularity and readability of the software. When consolidating multiple versions of the algorithm Devon and Steve Demedash maintained 5

8 2 Project Progress and Future Work functionality without reducing algorithm speed. However, consolidating the software versions added more decision-making steps to the main process, introducing significant lag. To maintain algorithm speed, and further reduce the code size, Devon and Steve Demedash designed a Finite State Machine for quick branching when device specific portions of code were encountered. In the next two months, Devon and Steve Demedash will follow the tasks outlined in the Gantt Chart located in Appendix A. They will test the software to ensure it is operating as intended. There are multiple hardware setups to be tested, this testing phase may be quick and present positive results from the outset. Alternatively it could last much longer than anticipated, and may take until the end of the semester to complete. After the testing phase, they will either need to reevaluate the completed software updates, or proceed with graphical user interface updates, and algorithm optimization. We introduced these as an optional task, to be completed if time permits. To update the graphical user interface, Devon and Steve Demedash will propose a new layout that does not take away any functionality that already exists. The majority of the user interface involves the selection of devices through which the user may configure the system for the devices used, and type of scan to be done. This layout is cumbersome, so moving the device selection panel over to a temporary tab or window would create a cleaner, more straight-forward interface. The method by which the antennas are selected also needs to be re-examined. Currently, this portion of the interface is small, and as such, it is difficult to select the desired antenna positioning. Optimizing the imaging algorithm involves two things. First, the time it takes to measure the data between antennas is the longest running time of the program, so they have the most to gain by focusing our attention on this area of the code. Secondly, the software currently sends out data to an excel file that is then loaded into Matlab where graphs and plots are generated. This transfer over to Matlab takes time and it would be nice to have this in the same program that performs the measurements. To do this Devon and Steve Demedash will need to get familiar with the preexisting Matlab code and create graphical elements in the imaging software that allow the 6

9 3 Conclusions program to display the relevant graphs after taking the measurements has been completed. 3 Conclusions Overall Group 12 has made significant progress in all areas of the project. Numerous setbacks in specific project areas has led to a change in scope for both the field coil and antenna portion of the project. Due to the sensitive nature of electromagnet devices more time was allotted to the design stages of the product to ensure of functionality of the final product. The field coil section will complete 1/16th of the original scope of the project due to expensive manufacturing and material cost. The antenna portion is now designing a TE11 resonating cavity to meet all of the project specification, while software is completing proposed tasks on schedule. With these new changes we intend on completing tasks as assigned in our revised project schedule found on our updated Gantt chart. 7

10 Appendix A Updated Gantt Chart Fig. A.1: Revised Gantt Chart for Microwave Imaging Project (1 of 2) 8

11 Fig. A.2: Revised Gantt Chart for Microwave Imaging Project (2 of 2) 9

Appendix B Updated Budget The budget for the our Microwave imaging project is $1346.48 and is covered but the Dept.

12 Appendix B Updated Budget The budget for the our Microwave imaging project is $ and is covered but the Dept. of Electrical and Computer Engineering with the remainder of the fund provided EIL. Fig. B.1: Up to Date Budget of the Project 10

13 Appendix C System Diagram Fig. C.1: System Overview 11

Improvements to a Microwave Imaging System for Breast Cancer Detection

University of Manitoba Department of Electrical & Computer Engineering ECE 4600 Group Design Project Final Project Report Improvements to a Microwave Imaging System for Breast Cancer Detection by Group