Coin Identification Using Eddy Currents

Transcription

1 Coin Identification Using Eddy Currents First Semester Report Full Report By Frank Lindauer Mohammad Alhusainan Momin Garrouch Prepared to partially fulfill the requirements for ECE401 Department of Electrical and Computer Engineering Colorado State University Fort Collins, Colorado Project advisor(s): Dr. George Collins Mr. Garrett Durland Approved by:

2 Abstract Coin fraud continues to be a problem for any industry that relies on automatic coin identification and processing such as vending machines, change machines, car washes, and public laundries. For example, acceptance of foreign currency, which may have a lower exchange rate, or worthless tokens or slugs, which have no value, directly impacts the bottom line of these business owners. Replacing the coin slots with debit or credit card readers may help to eliminate the coin fraud but also introduces additional security problems and cost (need access to data line or Wi-Fi connection) while limiting machine access to customers who have these cards. This project will demonstrate one of the many practical uses of eddy currents by designing, building, and testing a system that can identify the denomination of United States (US) coins based upon measured changes in the magnetic field caused by induced eddy current. These circular electric currents are induced within the metallic coins as they roll into the magnetic field produced by a permanent magnet. The induced eddy currents in turn produce a magnetic field that is opposite in direction to the field produced by the permanent magnets. These opposing forces work to slow the momentum of the coin as they continue to roll through the magnetic fields. By timing the coin as it passes through these magnetic fields, we can measure the effect that the induced eddy current has on the coin and in turn identify the denomination of the coin. We were expecting to see a distinct time separation between the coin denominations that would make the identification process easier; however, the actual data that we collected does not bear that out. There are overlaps in the timing intervals between all four coin types. Based upon the data results, our system would not be able to consistently identify the denomination of the coins correctly. As a result, we have proposed the following items to explore as solutions to this issue in the second semester: Reducing the friction of the acrylic guide rails to produce more consistent coin rolls and record more precise times. Replace the permanent magnets with electromagnets and integrate Hall Effect sensor and potentiometer to allow us to measure and control strength of magnetic field. A precisely tuned magnetic field would help us to measure the impact of the field on each coin denomination and find the best field strength to increase the separation intervals. Alter the angle of the demonstration ramp to determine the best angle that balances the influence of the electromagnetic field with the effect of gravity. We strongly believe that combining two or more of these proposed solutions will definitely resolve the issue.

3 Table of Contents Abstract... 2 Table of Contents... 3 List of Tables & Figures... 4 Chapter 1 Introduction... 5 Chapter 2 Summary of Previous Work... 7 Chapter 3 Project Details... 7 Objectives... 7 Constraints... 7 Coin Ramp... 7 Circuit Design... 9 Microcontroller Programming Testing Decisions, Procedures, and Results Chapter 4 Conclusions & Future Work Conclusion Next Semester Plans References Appendix A Abbreviations Appendix B Budget What was planned Donations Estimate for labor Total cost Funding Appendix C Project Plan Evolution Initial project timeline, submitted September 17, Revised project timeline, submitted October 8 th, Revised project timeline, submitted November 6 th, Acknowledgments... 28

4 List of Tables & Figures Table 1: Coin Timing Data by Denomination Table 2: Project Plans for Next Semester Table 3: Inexpensive Budget Table 4: Brand Name Budget Table 5: Realistic Budget Table 6: Project Budget Thus Far Table 7: Project Timeline Sept. 17, Table 8: Project Timeline Oct. 8, Table 9: Project Timeline Nov. 6, Figure 1: Coin Ramp with Circuit... 8 Figure 2: Close Up of Acrylic Guide Rails... 9 Figure 3: Connection from single photo interrupter to Arduino board Figure 4: Connections from two photo interrupters to Arduino board Figure 5: Photo interrupter Schematic [3] Figure 6: Quarter Timing Data Scatter Plot Figure 7: Dime Timing Data Scatter Plot Figure 8: Nickel Timing Runs Scatter Plot Figure 9: Penny Timing Data Scatter Plot Figure 10: All Denomination Timing Data Scatter Plot Figure 11: Standard Normal Curve for Quarter Figure 12: Pie Chart of Total Project Hours... 22

5 Chapter 1 Introduction Coin fraud continues to be a problem for any industry that relies on automatic coin identification and processing such as vending machines, change machines, car washes, and public laundries. For example, acceptance of foreign currency, which may have a lower exchange rate, or worthless tokens or slugs, which have no value, directly impacts the bottom line of these business owners. Replacing the coin slots with debit or credit card readers may help to eliminate the coin fraud but also introduces additional security problems and cost (need access to data line or Wi-Fi connection) while limiting machine access to customers who have these cards. The main theory behind our project is based on electromagnetics, specifically Faraday s Law of Induction and Lenz Law. Faraday s Law governs the interaction between a conductor and a magnetic field that produces an electromotive force otherwise known as electromagnetic induction [1]. The induction is equal to the negative rate of change in the magnetic flux relative to time. The equation for Faraday s Law is:. Lenz s Law states that an induced electromotive force always gives rise to a current whose magnetic field opposes the original change in magnetic flux [2]. His law basically added the negative sign in front of Faraday s Law of Induction. Eddy currents are used in a wide variety of applications throughout industry such as metal detectors, conductivity meters for non-magnetic metals, non-destructive testing, coating thickness measurements, proximity sensing and many others. This project will demonstrate one of the many practical uses of eddy currents by designing, building, and testing a system that can identify the denomination of United States (US) coins based upon measured changes in the magnetic field caused by induced eddy current. These circular electric currents are induced within the metallic coins as they roll into the magnetic field produced by a permanent magnet. The induced eddy currents in turn produce a magnetic field that is opposite in direction to the field produced by the permanent magnets. These opposing forces work to slow the momentum of the coin as they continue to roll through the magnetic fields. By timing the coin as it passes through these magnetic fields, we can measure the effect that the induced eddy current has on the coin and in turn identify the denomination of the coin. Our system design will consist of a display ramp that will channel the coin into a narrow chute that will guide the coin past two optical sensors, also known as photo-interrupters, and a magnetic field.

6 As the coin enters the ramp it will pass through a photo-interrupter. A photo-interrupter consists of a diode that emits a ray of light and a sensor that receives the ray of light. When the coin passes through the photo-interrupter it blocks the ray of light and records the time via the microcontroller. As the coin rolls down the ramp, it will pass through a static magnetic field generated by permanent magnets harvested from a hard disk drive. As the coin rolls through the magnetic field, the rotational motion of the metallic coin induces eddy currents on the surface of the coin. The eddy currents in turn induce a magnetic field that opposes the field of the permanent magnets thus slowing the momentum of the rolling coin. After the coin leaves the magnetic field, it will then pass through a second photo-interrupter which will record the stop time again via the micro-controller. The micro-controller measures how long the coin takes to pass through both photo-interrupters. Different types and amount of metals disturb the magnetic field in different ways, for instance, a larger coin contains more metal and therefore will disturb the field more than a smaller one. A quarter will disturb the magnetic field differently than a dime and will therefore take longer to roll down the ramp. By using the timing and magnetic field data, we can define an algorithm that will determine the denomination of the coin and will send that information to a liquid crystal display (LCD) screen displaying the identity of the coin as well as the time the coin took to roll the length of the ramp. We are expecting to see a distinct time separation between the coin denominations The use of different sensor types and different measurement parameters will make our detection and identification system more robust and less prone to error and reduces the risk of fraud. Another benefit of this system is its robustness, as it can successfully operate in harsh environments. Chapter 2 contains the summary of our research and interest into the background of our project. Chapter 3 contains the details of the work performed on the project broken down into major project components such as circuit design and data analysis. Chapter 3 also contains additional project information such as our overall project goals and constraints that we are working within. Chapter 4 contains our summary of the project for the first semester as well as our project plans for the second semester.

7 Chapter 2 Summary of Previous Work Vending machines have been around for a long time, lots of research has been done to improve the quality, size, and the inner working mechanisms of these machines. We are not expecting our project design to revolutionize the industry but we were intrigued by the operation of the coin identification mechanism and the engineering technology behind it. In particular, this video piqued our interest, How Stuff Works: Deconstructed Vending Machine. In short, we are using our senior design project as a learning experience versus an entrepreneurial endeavor. Chapter 3 Project Details Objectives Design, test, and build a coin ramp and use it to detect and identify different coins. Program microcontroller to linearize, offset, and scale data received from sensor. Integrate the sensor, LCD screen, and the micro-controller with the circuit. Build ramp display to demonstrate operation of coin identification system. Integrate the microcontroller to display type of the coin on the LCD. Constraints Our coin identification system will be designed to work only with US currency. Minimizing the power consumption of the electronics. Minimize physical size of electronics and circuitry in anticipation of small available space in commercial application. Overall project budget limited to $450.00, minimize per unit cost of electronics and circuitry. Coin Ramp The coin ramp is one of the primary parts of the design as it is used to demonstrate the operation of our coin identification system. We built it so we could perform consistently repeatable tests on the coin by rolling it down the ramp. The coin ramp is built from wooden materials with two parts of acrylic that work as a guide for the rolling coins. On the ramp we placed two rare earth permanent magnets on the outside middle section of the acrylic track to produce the magnetic field and two photo-interrupters to calculate the time it took the coin to pass through the ramp. Through the process of building the ramp we faced a few difficulties related to the physical operation of the ramp.

8 One such difficulty was finding the correct rolling angle and ramp height, because the steeper the ramp gets the less effect the magnetic field will have because the gravity effect will make the coin roll faster through the field. Through numerous field tests, we determined the best angle for the ramp that balanced the influence of the magnetic field while maintaining an angle that ensured the coins rolled the entire way to the bottom of the ramp. Magnets Photointerrupter Figure 1: Coin Ramp with Circuit Another difficulty that arose was with the acrylic guide rails on the ramp. The quarter, dime, nickel and penny are all different widths, with the nickel being the widest and the dime the narrowest. We were trying to provide as narrow of a guide as possible in order to limit the side to side movement of the coin as it rolled down the ramp. The less side to side movement the coin had, the more consistent the timing results would be. We encountered tests where the nickel would get stuck in the guide rails at various spots on its way down. We detached and reattached the guide rails many times to remove imperfections in the acrylic surfaces and to provide the proper spacing in between the guides to best accommodate all of the coins. As of now, the guide rails are wide enough to not inhibit the free rolling motion of the nickel but are not tight enough to keep the dime from moving from side to side. We are going to brainstorm about ways to improve the guide rails to produce more consistent timing results.

9 Acrylic Guide Rails for Coins Figure 2: Close Up of Acrylic Guide Rails A third difficulty was trying to physically attach the photo-interrupters to the ramp. The initial ramp was built with accommodations to mount the photo-interrupters on the extreme outer edges of the ramp. However, this would have required us to place or drop the coins into the guide track and we felt we needed a release mechanism that would provide us with consistently repeatable results and the overall design of the ramp was changed. The ramp was increased in length from about 2 feet to 2.5 feet to allow for placement of a release mechanism. The photo-interrupter that was to be placed on the upper end of the track was then placed a few inches from the end. A short section of guide track was added to the very top of the ramp, holes were drilled in the acrylic to allow a metal wire to be inserted which will act as a release gate. Circuit Design Circuit design in our project has passed by two phases. First of all, we had our Arduino board connected to one photo interrupter with two pull down resistors to get better readings as shown in the figure below:

10 Figure 3: Connection from single photo-interrupter to Arduino board In the second phase, we duplicated the components we had from the first circuit to get better readings and make it easier to hook it up to the Arduino board. Example of the circuit is in the figure below. Figure 4: Connections from two photo-interrupters to Arduino board Microcontroller Programming The microcontroller programming started with initializing the photo-interrupters and making them interact with the microcontroller. We used the 5v pin on the Arduino as the power pin for the photodiode and connected the other side to one of the digital pins, and the same for the transistor on the other side as shown in figure 5. Then we wrote some code that allowed the sensor to capture the time in milliseconds that when an object passed through the sensor. We were able to observe the photo-interrupter communicating with the controller through the serial monitor. When there is nothing passing through the photo interrupter we get a reading of a 1024 and when the coin passes through the sensor the reading changes to very low values.

11 Figure 5: Photo-interrupter Schematic [3] The second part of the programming was calculating the time difference between the two interrupts. We implemented it using a built-in function inherent to most of the Arduino boards called millis() [4], as this function starts a timer when we upload the code to the board. We used interrupt service routines (ISRs) to ensure that the time recorded by the microcontroller was as accurate as possible. It works by recording the time when the coin passes through the first interrupter and records the time through the second interrupter and then calculates the time difference between both the recorded times. For example, if the coin passed through the first photo-interrupter after 200 seconds of running the program and went through the second after 202 seconds which means that the coin took 2 seconds to pass through the ramp. Testing Decisions, Procedures, and Results The way we intend to program the controller to identify a rolled coin is by defining a unique roll time interval for each coin type within the code so that if a random coin is rolled and its roll time happens to fall in the range of one of the four coin types then the microcontroller will identify it as the coin it matched the range of. In order to define the time ranges for the four different coins, we tested ten different rolls for each coin and recorded the maximum (max) and the minimum (min) values for all coins which should define the time interval for each coin type. To further analyze our data, we calculated the mean and standard deviation values for the data of the different coins. Table 1 below lists all the data gathered for the four coin types and also shows the existence of time interval overlap between the following coins: Quarter & Dime Dime & Nickel Dime & Penny

12 Penny & Nickel Run # Quarter Time (s) Nickel Time (s) Dime Time (s) Penny Time (s) Min Max Mean Stand. Dev Time Range [MIN,MAX] [ ] [ ] [ ] [ ] Table 1: Coin Timing Data by Denomination This time overlap will create confusion within the microcontroller leading to false coin detection in a sense that it would be possible for the microcontroller to read a Quarter as a Dime and vice versa. Figures 6 through 9 below show the scatter plots for each of the four coin types.

13 Figure 6: Quarter Timing Data Scatter Plot Figure 7: Dime Timing Data Scatter Plot

14 Nickel Figure 8: Nickel Timing Runs Scatter Plot Figure 9: Penny Timing Data Scatter Plot We finally plot all gathered data for the four different coin types on one graph to check whether the min and max interval of the coins overlap or not. Unfortunately it turns out that an overlap exists between multiple coin types. You can see the results in Figure 10 below.

15 Nickel Figure 10: All Denomination Timing Data Scatter Plot We believe that this time overlap issue is a result of the high standard deviation values for the different coins. Lowering these standard deviation values of each coin is the key to tackling this issue especially knowing that each coin type has a unique mean value as seen in Table 1. Fig 11 below shows the standard normal curve of the Quarter data. Since the probability of having a reading within one standard deviation from the mean is 68.3%, decreasing the standard deviation will force readings to have greater probability of lying closer to the mean value, therefore leading to more accurate results. Improving the accuracy of our readings will result in smaller differences in the Min and Max values of each coin. Throughout the data gathering process, we noticed that when the coins roll down in the ramp they randomly get slowed down at some points. After closely examining our ramp structure, we notice that the acrylic surface is actually rougher than we thought it would be. This eventually results in a friction force exerted on the coin whenever it comes into contact with the acrylic s surface. We believe that smoothing out the acrylic s surface would significantly reduce the standard deviation for each coin to the point where the min and max values would be close enough to the mean values to provide us with unique time ranges for each coin.

16 Figure 11: Standard Normal Curve for Quarter Chapter 4 Conclusions & Future Work Conclusion Although the coin ramp mechanism we built is not fully functioning as it should due to the time overlap between the coins, we re highly confident that a combination of the following proposals would help to solve the problem The first alternative solution would be placing a stronger magnet on the ramp to possibly to further slowdown the coins. Each coin has a different metal composition therefore each coin would be slowed down at a different rate. Other factors might affect this as well such as weight and size of the coin. The second alternative solution suggests changing the ramp s angle to rely on the coins different sizes and weights to possibly solve the problem. We strongly believe that combining two or more of these suggested solutions would definitely solve the issue and would leave no chance for doubt. Next Semester Plans

17 At this point, we consider the project to be half way done. As for next semester, we intend to work on solving the coin overlapping issue and adding programming the LCD to display output on the screen. We will also use next semester to integrate an electromagnet to our mechanism which will involve a potentiometer to vary the magnetic field strength. Further programming will be made to the microcontroller to display the magnetic field strength on the LCD screen. Our team will be more than happy to work on adding more features to our mechanism if we manage to finish the required work ahead of time. Table 2 briefly summarizes the required work for next semester as well as possible additions if time allows. Required work Get unique time interval for each coin type Program Arduino board to display coin type and total roll down time Use an electromagnet and Integrate potentiometer to vary the magnetic field Integrate a Hall Effect sensor Possible Additives Etch circuit on circuit board Mechanically sort coins into different bins Make system compatible with other foreign coins Involve a high proximity sensor to possibly detect coin composition Further Program the microcontroller to display magnetic field strength on the LCD Table 2: Project Plans for Next Semester Despite the hard work and challenges we faced with the project up till now, our team has never been more eager and enthusiastic about making this project fully functioning by next semester. We are fully aware of the challenges that await us for next semester and we re looking forward to facing them.

18 References [1] "Faraday's Law of Induction," [Online]. Available: [2] "Lenz's Law," [Online]. Available: [3] "Utopia Mechanicus," [Online]. Available: interruptor slotted optical switch/. [4] "Arduino Millis Function," [Online]. Available:

19 Appendix A Abbreviations Wi-Fi US LCD USB PC ECE CSU Local area wireless internet technology United States Liquid Crystal Display Universal Serial Bus Personal Computer Electrical and Computer Engineering Colorado State University

20 Appendix B Budget What was planned The preliminary studies made on the budget at the beginning of the semester ended up being more overpriced comparing to what we ended up paying to this point. When it comes to buying any product, there is always a tradeoff between price versus quality. The common idea that priced products usually provide better quality than cheaper ones is mostly true. Electrical appliances is one of the best examples to prove the idea right where more expensive products are usually more durable, reliable, and even more efficient when comparing them to cheaper ones. We believe that finding the right balance between price and quality is the key when it comes to making the right product decisions and potentially avoid extra costs due to failures that will increase the overall project cost. In order to maintain a low project cost while maintaining good project quality, we made sure to extensively study each component individually before we actually proceeded in buying it. As part of our planning our budget, we went ahead and compared our expected budget while using the cheapest components off the market versus the expected budget while using very expensive component as shown in table A.1 and A.2 respectively. Results for both budgets we re very different as expected and the two budget models are not what we want to be ending up with. As mentioned earlier we looked for the best balance that would fit our needs. The initial project budget we came up with at the beginning of the semester is shown in Table A.3. We all agreed to the fact that this budget plan is tentative and is flexible to any changes that we might face throughout the project. As we initially expected, changes did indeed come up during the course of the project which ended up shaping our budget as well as the total cost of the project differently. Some of these changes were due to project decision changes and others were a result of part failure. Donations also had some minor impact on our budget which will be discussed later on. The only decision related changes made were regarding the LCD where we initially intended to buy a Chinese manufactured 16x2 LCD which was Arduino compatible, until our industry supervisor suggested buying the Sparkfun 16x2 LCD which he thought would be more reliable based on his experience. As stated earlier, part failure also had an impact on our total cost due to bad decision making. The decision of buying a fairly cheap generic Arduino board turned out to be not very successful since the board s USB port failed to work after 2 weeks of use only. We ended up buying another Nano board which was almost $10 more expensive.

21 Table 3: Inexpensive Budget Table 4: Brand Name Budget Table 5: Realistic Budget Donations Run PC Computers and Repair was very kind to donate one of their unused hard drive magnets to our project. They were very cooperative and they also offered to donate more parts if needed. The donated magnet was worth $7. We used the donated hard drive to replace the need of buying magnetic cores which helped in lowering the total project cost in a small factor. Estimate for labor The total hours per person spent on this project up to this point are expected to be around 156 hours. The breakdown of the total hours is as follows: 70 hours were spent on research 16 hours were spent on building and reconfiguring the ramp 18 hours were spent on building the circuit and connecting all hardware (i.e. sensors, LCD, microcontroller, magnets) 17 hours spent on hardware testing 35 Hours were spent on programming the microprocessor

22 Hours Research 22% 45% Building Ramp Circuit and Hardware connections 11% Hardware testing 12% 10% Programming Figure 12: Pie Chart of Total Project Hours The total project cost was not affected by the labor hours since we re the project owners and we re serving our own project without hiring anyone else to do any labor or non-labor task. Total cost After taking the incurred losses as well as donations and plan changes into consideration, the new project total can be calculated as shown in the spreadsheet in Table 4 below. The total amount paid at this point totaled $123 including incurred losses from the first microcontroller. The spreadsheet also shows the second semester as well as the total expected project cost which turned out to be $29 and $ respectively.

23 Table 6: Project Budget Thus Far Funding We re currently funding our project through the ECE student project fund which assigns $150 for each student working on ECE senior design project. Payments were initially made out of our pockets but we re currently in the process of getting our refund through the ECE department at CSU.

24 Appendix C Project Plan Evolution Initial project timeline, submitted September 17, 2014 Phases Start End Group Member Duration Date date (Fall Semester Plan) 9/17/14 10/8/14 Mohammad 3weeks Phase 1: Develop design parameters, and design the eddy current sensor. Model it using Pspice or cadence. Phase 2(a): Build the self-oscillating circuit 10/8/14 10/29/14 Momin 3weeks Phase 2(b): Build a prototype of the eddy current sensor. 10/29/14 11/12/14 Frank 2week Phase 2(c): measure the prototypes 11/12/14 11/19/14 Mohammad 1weeks electrical properties Phase 3(a): build an improved prototype and 11/19/14 12/3/14 Momin 2weeks measure its properties Phase 3(b): Preparing final report and 11/26/14 12/10/14 Mohammad, 2weeks presentation. Frank, Momin (Spring Semester Plan) Phase 4: connect oscillating circuit to a micro controller, and LCD screen 1/20/15 1/27/15 Frank 1week Phase 5: program the micro-controller to 1/27/15 2/17/15 Mohammad 3weeks display the distance of the metallic object on the LCD or on the PC screen Phase 6: measure power consumption and 2/17/15 3/3/15 Momin 2weeks the power of noise observed on the system by using the oscilloscope. Phase 7:Gather data and make table summarize the performance of the eddy current sensor 3/3/15 3/24/15 Frank 3weeks Phase 8: Preparing final report and 3/24/15 4/14/15 Mohammad, 3weeks presentation, E-day preparation. Frank, Momin Table 7: Project Timeline - Sept. 17, 2014

25 Revised project timeline, submitted October 8 th, 2014 Phases Start Date End date Group Member (Fall Semester Plan) 8/25/14 9/17/14 Frank, Phase 0: Project research and background. Mohammad, Phase 1: Develop design parameters, and design the eddy current sensor. Model it using Pspice or cadence. Duration 3 weeks Momin 9/17/14 10/8/14 Mohammad 3weeks Phase 2(a): Build the self-oscillating circuit 10/8/14 10/29/14 Momin 3weeks Phase 2(b): Build a prototype of the eddy current sensor. 10/29/14 11/12/14 Frank 2week Phase 2(c): measure the prototypes 11/12/14 11/19/14 Mohammad 1weeks electrical properties Phase 3(a): build an improved prototype and 11/19/14 12/3/14 Momin 2weeks measure its properties Phase 3(b): Preparing final report and 11/26/14 12/10/14 Mohammad, 2weeks presentation. Frank, Momin (Spring Semester Plan) Phase 4: connect oscillating circuit to a micro controller, and LCD screen 1/20/15 1/27/15 Frank 1week Phase 5: program the micro-controller to display the distance of the metallic object on the LCD or on the PC screen Phase 6: measure power consumption and the power of noise observed on the system by using the oscilloscope. Phase 7:Gather data and make table summarize the performance of the eddy current sensor Phase 8: Preparing final report and presentation, E-day preparation. Table 8: Project Timeline - Oct. 8, /27/15 2/17/15 Mohammad 3weeks 2/17/15 3/3/15 Momin 2weeks 3/3/15 3/24/15 Frank 3weeks 3/24/15 4/14/15 Mohammad, Frank, Momin 3weeks

26 Revised project timeline, submitted November 6 th, 2014 Phases Start Date End date Group Member (Fall Semester Plan) 8/25/14 9/17/14 Frank, Mohammad, Phase 0: Project research and Momin background. Duration 3 weeks Phase 1: Develop design parameters for coin id system and necessary sensors. Phase 2(a): Build the selfoscillating circuit. Phase 2(b): Build a physical coin ramp. Phase 2(c): Build circuit and establish software connection between photo interrupter and micro-controller. Phase 3(a): Coding interrupt service routine to implement timers with photo interrupts. Phase 3(b): Initial testing of entire system with implementation of sensors and permanent magnet. (Note: This may carry over to the second semester depending on the time we have) Phase 4: Preparing end of semester report and presentation. 9/17/14 10/8/14 Mohammad, Momin, Frank 3 weeks 10/8/14 10/29/14 Momin 3 weeks 10/29/14 11/12/14 Frank 2 weeks 10/29/14 11/9/14 Mohammad,Momin 1.5 weeks 11/9/14 11/23/14 Mohammad,Momin,Frank 2 weeks 11/23/14 12/10/14 Mohammad, Frank, Momin 12/1/14 12/12/14 Mohammad, Frank, Momin 2.5 weeks 1.5 weeks (Spring Semester Plan) Phase 5: Program micro-controller to display coin rolling time on LCD screen. Phase 6: Integrate Hall Effect sensor with system. Phase 7: Program micro-controller to display magnetic field strength on LCD. Phase 8: Test system with Hall Effect sensor in place. Collect data on field strength. Phase 9: Modify circuit to integrate potentiometer, use to vary the magnetic field. 1/20/15 2/3/15 Frank, Mohammad 2 weeks 1/26/15 2/16/15 Momin 3 weeks 2/3/15 2/24/15 Frank, Mohammad 3 weeks 2/17/15 2/24/15 Mohammad, Frank, Momin 1 week 2/24/15 3/10/15 Mohammad, Frank 2 weeks Phase 10: Test system with 3/3/15 3/10/15 Momin 1 week

27 potentiometer in place. Collect data with variable field strength. Measure power consumption and the noise observed on the system. Phase 11: Gather data ; summarize 3/10/15 3/31/15 Frank, Mohammad, the performance of our system. Momin Phase 12: Prepare final report and 3/31/15 4/21/15 Mohammad, Frank, presentation, E-day preparation. Momin Table 9: Project Timeline - Nov. 6, weeks 3 weeks

28 Acknowledgments We would like to extend our thanks to our Faculty Sponsor, Dr. George Collins, our Industry Advisor, Mr. Garrett Durland, and our Senior Design Professor, Olivera Notaros for their advice, guidance, support, and constructive criticism throughout the course of this project. We would also like to thank Josh Olson for lending us his Arduino Uno board after our Nano board burned out.