Solar Powered Lead Acid Battery Charger
Contents
1. 5 to 15 volts Applying the MAX232 bridges these differences with the assistance of a few external 10 microfarad capacitors Another chip called the MAX233 13 can be used with no external part but the size of the device is larger and there is a higher price tag 3 Figure 10 Buck Regulators 15V P8 mou The buck regulators are the one of the workhorses of our project and the involve converting high voltage low current into low voltage high current for charging the battery We have been able to replace the previous buck regulator with one that has been implemented on a chip IR2110SPbf which is opto isolated so there is no need for external iso couplers More about the buck regulator will be discussed in the power train section of this report Figure 11 New Microprocessor Pic18F2620 28 Pin SPDIP SOIC t RB7 KBIXPGD t RBSI KBIZ PGC RB5 KBI1 PGM RBA KBIOIAN1 1 w RBWVANSICCP2 RBZIINTZIANS RB1 INT1 AN10 t RBOINTO FLTO AN12 DD MCLRiverP RE3 LJ 1 RAO ANO LJ RAV AN1 LJ RAZ AN2 VREF CVREF LI RAWAN3 VREF RA4 TOCKIIC1OUT _ RAS AN4 SS HLVDIN G20UT Vss OSC1 CLKI RA7 L OSC2 CLKO RAG e L 10 RCO TIOSO T13CKI 11 RC1 T10SI CCP21 1 w 12 RC2 CCP1 113 RC3 SCK SCL LJ 14 PIC1 8F 2525 PIC1 8F 2620 a Vss RC7 RX DT RC6 TX C2. microprocessor we chose and why we choose it writing to the LCD display and the challenges we encountered in successfully compiling and downloading to the microprocessor Chapter IV will be a brief discussion on the budget of our project and finally in chapter V we will talk about the future work that needs to be done to achieve our expected goals Il REVIEW OF PREVIOUS WORK What follows is an excerpt from an e mail from Jack Gilmore at Advanced Energy who spear headed this project previously and helped develop the original prototype The design started off several years ago with just a relay driven voltage regulation type of charger like in a car As soon as the first maximum power point charger appeared on the market I realized that the charger could be improved by adding a buck regulator and regulating the voltage on the panels to the optimum for the panels 16 5V An improvement of tracking the maximum power point would take an additional investment of a microprocessor control to vary the duty cycle of the buck regulator This was tried two years ago The failure of the project was due to inadequate processor memory and lack of stack depth in the processor There was not enough effort placed on the software either Only one person worked on it 3 With our prototype we believe we have corrected the memory insufficiencies and lack of stack depth by changing our microprocessor to the more powerful PIC18F2620 Also by spli
3. Current Meter Circuit C29 1 R39 Sav 9 9K 1x t 4W Ras l DAN_C mamam gt oe VN BP R33 1 41 4W zgi omet i2 reas 24 Uo B Lezs 028 1K ADASTIR 121 4W By a 182 12 174W 53V ay 7 D ook e 1 Poul c27 C18 8 4 R29 5 a 12 1 4W R24 auue_ Wy aap R3 1X IZEN NWY T ca ae pe BB 1zivaw 021 Ae 80 iz 1746 AD8572R I A sy The next implementation step was to add in an additional buck regulator which will be explained in the Power Train Design Section The last step to allow for an additional power source was to take the voltage current and pulse width modulation Buck Regulator pins from the additional parts and connect them to the new processor However we did not have enough empty pins to start with so we first had to see if we could remove any unnecessary circuit components The first part we decided to remove was the external cooling fan With correct implementation of the circuit via software we would not need a fan to cool it This actually gave us 2 pins to use since the fan also had a temperature sensor output to let it know when to turn on Here is the fan and temperature sensor circuit 10 Figure 6 Fan and Temperature Sensor Circuit 15 5V P5 2 R41 1 K 1K ie 1 10K 12 1 8W 1 178W 1 174W P6 a Q1 D r oe amp NOS7 G2A A rae o S v7 R42 10K 12 174W The second part we removed was the unsafe boost voltage circu
4. diSelenide CIGS Online Document 2006 December 2 Available at HTTP http www daystartech com technology cfm 3 J Gilmore RE Final Report Online Document 2006 December 1 E mail to Andrew Giddings Andrew engr colostate edu 4 K Ross Project RS 232 to TTL cable Encoder The Newsletter of the Seattle Robotics Society Online Available at HTTP http www seattlerobotics org encoder aug97 cable html Accessed Dec 2nd 2006 5 Microchip PIC18F2620 Data Sheet Online Document 2006 December 1 Available at HTTP http ww1 microchip com downloads en DeviceDoc 39626b pdf 6 Switching Regulators Online Document 34 2006 November 29 Available at HTTP http www national com appinfo power files f5 pdf 7 International Rectifier IR2110 Data Sheet Online Document 2006 December 1 Available at HTTP http www alliedelec com Images Products Datasheets BM INTERNATIONAL_RECTIFIER International Rectifier_Actives and Passives_2737836 pdf 8 Dot Matrix Character LCD Module User s Manual Online Document 2006 December 2 Available at HTTP http www optrex com pdf Dmcman_full pdf Bibliography Solar Cell in Wikipedia The Free Encyclopedia 2006 Online Available Wikipedia Online http en wikipedia org wiki Solar_cell Solar_cell_efficiency_factors W Kester and B Erisman Section 3 Switching Regulators Online Document 3 10 2006
5. load current By varying the duty cycle of the clock we can control how much of the 380 V the load sees which will be 48V if 4 series 12V batteries are to be charged The capacitor acts as a filter on the 380 volt pulses minimizing the decay between pulses inductor current ripple and generating an approximate DC output voltage 6 19 Figure 16 Simplified Buck Regulator SWITCH gt SWITCH SWITCH oN lt _ In the previous year s effort see Figure 1 the MIC4420BM component served as the switching transistor and the LM78L05 acted as the diode resistor current shunt in the Buck regulator just discussed In our design however we used the APT2X60D60J dual rectifier diode isotop package for the diode and the APT5010JPR isotop package was our transistor switch Both packages were rated for power systems use at 500V We also replaced other components that tackled opto isolation of the dangerous power levels of the buck with an International Rectifier IR2110 chip see Figures 17 18 7 This freed up some space on the board and allowed us to include a second IR2110 for an auxiliary set of expansion panels relatively easily The inductor which is located off board was rated for approximately a hundred amperes in order to support such massive swings in voltage The capacitor and resistor you see in parallel at the back end of the buck regulator represent the battery We also realized the need for sha
6. to change and set battery charging constants verify voltage and current readouts and other general calibrations on the system Our first plan of action was to output text to the LCD display which is done by the microprocessor sending data streams to the LCD in order for it to display its respective character This turned out to be a more difficult task than it appeared One of the main difficulties we had in starting the software interfacing was acquiring the necessary software to communicate and program the microprocessor Microchip has a software developers Integrated Development Environment IDE called MPLAB This IDE is a powerful programming tool that supplies large libraries of tutorials to make programming easier Microchip also provides a C18 C compiler that we found for educational purposes was unfortunately a trial version which is now limiting our program size and certain compile time optimizations The retail value of this software is 495 per license which is clearly out of our budget range We will however continue to use this complier until it isn t sufficient for our purposes We also required an 22 In Circuit Debugger to program and debug our microprocessor For this we choose the Microchip ICD 2 The MPLAB ICD 2 is a low cost real time debugger and programmer which will connect to our PIC MCU The ICD 2 connects to the computer using a USB port and connects to the microprocessor using a 5 wire connection to hook the cloc
7. we F2 satel A 4 eu 8 ya sene AR m j lee w 3 N me R23 sv V S q gg ivaw B ipin Los Y Ww rms efine a fen ata Led ca M TR ww 0 swasr2R ET s fla lg y ay 45V Vv aw ei R27 Raai v 3 TS isv EMTA ue o Tame Sie ty 4 Japas72R _ a Tetsane SS aia a 05 MANAGIT ff XXXK ADVANCED ENERGY isd bv oe dood neuer Giese pee x Wk 14 ea H E Proreerany mrormarion Now that the design was complete at the end of the first semester we will now cover the process of building and debugging our circuit We will include both our successes and our failures in building the circuit At the start we had to take our circuit which was already in the necessary CAD program covered in the original budget section and make sure everything was acceptable before printing the circuit To do this we mainly had to make sure the software had decided to run all the traces with correct spacing for higher voltages Once this was correct we proceeded to actually print the circuit The blank boards used in this prototyping machine are a piece of dielectric with solid copper on each side To print the board we place the blank in the machine both of which are covered in the original budget and set the etching bits to match the necessary 17 spacing for our traces Once it was set up we simply loaded our design into its memory and pushed start To create the board it simply used the etching bits to clear away the unused part of the copper Since this boa
8. 2 in circuit debugger module to the chip and successfully flashed executed a menu system and duty cycle PWM selector program The next step would require reading in the four measurements battery voltage battery current panel voltage and panel current Once these value have been A D converted we can manipulate theses hexadecimal representations by coding C equivalent subroutines for P IV and Vout dutycycle Vin etc equations Then the PIC would need to constantly monitor these measurements and make corrective decisions autonomously We also expect next year s students to have modes of charger operation that address the stages of battery charging including a CC mode CV mode and boiling mode which evenly disperses the charges periodically to increase battery life and performance Eventually we would even like to see the software 30 in place for the addition of a second array of solar panels from a different manufacturer and the second buck regulator as well as variable battery array size to make things more adaptable to a customer s particular equipment We feel this project has taught us the value of planning ahead the value of a finding and using an industry mentor and the difficulties unique to multidisciplinary embedded system designs utilizes power software hardware optics active passive large signal small signal digital control filters and system design We have learned that the most influence you can have o
9. 232 link and the final assembly B Power Train The power train portion of our project is arguably one of the least time consuming components but nevertheless is most integral to the controller s performance Here we have one fundamental goal to take the relatively high voltage low current being generated on the panels and convert it into a high current low voltage configuration in order to charge the lead acid batteries The mechanism by which this is accomplished is a device known as a buck regulator A buck regulator consists of an inductor a capacitor a diode and a transistor that is used as a switch The switch is controlled by the PWM see Appendix A which is generated by a clock on the microprocessor When the square pulses of the PWM are high the transistor is turned on and the high voltage from the panels is applied to the inductor generating a current through it see Figure 16 This current is delivered to the load and charges the capacitor Then when the square pulse is low the transistor turns off and the voltage across the inductor is removed However current in an inductor can t change instantaneously so the high voltage reduces to maintain this current Eventually the negative voltage at the input of the inductor drops below the threshold of the diode turning it on and discharging the inductor current through the load During this off time the capacitor also discharges through the load contributing some to the total
10. After this he allowed us to make decisions about where each component should go We knew that the wires that connected the diode and switch attached to the back of the heat sink to the capacitors should be as short as possible to reduce stray inductance So we tried to make the capacitors as central to all the components and as close to the diode and switch as possible We needed two inductors for each solar panel arrays only have one right now but may have two in the future We needed three shunts for the batteries and for the two solar arrays We also needed a RS232 port for communication from the computer to the microchip used only for troubleshooting Figure 21 shows are initial design Figure 21 Chassis with major components Inductors Capacitors Shunts Power Supply RS232 Port We also decided that we would like to have the circuit board attached a clear plastic cover panel The LCD and the toggle and enter buttons would be attached to the panel This would allow for safe viewing of all the components at E Days If this were to be placed in a home we would make the cover panel opaque for aesthetic reasons Figure 22 shows our design for the clear plastic cover panel Figure 22 Clear Plastic Pannel Cover Clear Plastic oggle Enter Circuit Board After we designed both the box and the cover panel we began the fabrication All of the components for this portion of the project were donated by Advanced Ene
11. Appendix A NI N AUA RA U Nel wW Ww UQ Ww Q N N N e Aa OU N N O O N N Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Table 1 Table 2 LIST OF FIGURES Old Schematic Original Voltage Meter New Voltage Meter Original Current Meter New Current Meter Circuit Fan and Temperature Sensor Circuit Boost Voltage Circuit New Pin Layout RS232 with Isolation Couplers Buck Regulators New Microprocessor Pic 18F2620 Switches and Flash Circuit Battery Current Meter Voltage Sources New Schematic Simplified Buck Regulator Typical Connection of the IR2110 regulator chip Schematic diagram of the IR2110 regulator chip MPLAB ICD 2 interface connector 4 bit Initialization Chassis with major components Clear Plastic Pannel Cover LIST OF TABLES Character Bit Stream Table Budget Costs O Oo N 10 11 12 13 14 14 15 15 16 17 20 21 ai 23 24 27 28 25 29 I INTRODUCTION Harnessing solar energy has been the holy grail of renewable energy research for some time now Photovoltaic solar cells convert light energy to high voltage and low current which can be manipulated to provide power to our modern electrical devices and homes With the ever rising costs of fossil fuels the need for an efficient and affordable solar
12. December 2 Available HTTP http www analog com UploadedFiles Associated_Docs 84245885Power_sect3 PDF 32 Acknowledgements Special thanks go out to Jack Gilmore and Advanced Energy for their significant contributions to our senior design project from design considerations to component donations We would also like to express our gratitude to Microchip for their PIC microcontroller sample 33 Appendix A A D Analog to Digital CIGS Copper Indium Gallium diSelenide fabrication process for solar foil EEPROM Electrically Erasable Programmable Read Only Memory HPVEE Hewlett Packard Visual Engineering Environment ro Input and Output ICD2 In Circuit Debugger V 2 00 IDE Integrated Development Environment IR2110 International Rectifier chip for buck regulator LCD Liquid Crystal Display MPLAB Integrated Development Environment for interfacing PIC microcontrollers with bit streams created from assembly or c projects PIC Peripheral Interface Controller PWM Pulse Width Modulator RS232 Radio Electronics Television Manufacturers Association Standard serial communication port 34
13. K te RCS SDO 4 RC4A SDI SDA 14 The new microprocessor is more capable of performing the actions needed to make this project become a success This processor will be discussed more in depth within the software section Figure 12 Switches and Flash Circuit 5V sd oe UNSAFE MINUS HIGH VOLTAGE 4 5V ny Bo E 1174W P14 E TE aw lA g i z P13 ss R34 x Switches ICSP conn These components have kept from the original design the switch circuit is for outside control of the device and is comprised of three options up down and select for use of monitoring the voltages and currents on the liquid crystal display as well as using a new feature which is to strike the battery with a higher voltage than the charging voltage but only when it is fully charged which has been found out to prolong the battery life This method of charging will also be discussed in the software section as well The Flash circuit is the component needed for programming the processor Figure 13 Battery Current Meter Boti side oe RE A E 1 174W Gls wi C2 s zF 100 63V v7 15 Like in the original design we have a circuit to measure the value of current coming from the battery The remaining figures are other smaller but essential parts of out device the component A is a voltage source that will be used for each of the operational amplifiers that are being used to measure the currents coming off ei
14. Solar Powered Lead Acid Battery Charger Final Report Spring Semester 2007 By Andrew Giddings Scott Johnson Jake Shelton Alex Wolke Steve Searcy Doug Antioco Prepared to partially fulfill the requirements for EE402 Department of Electrical and Computer Engineering Colorado State University Fort Collins Colorado 80521 Report Approved Project Advisor Senior Design Coordinator ABSTRACT Our problem is the ever rising costs of non renewable energy from natural resources and the resulting energy crisis facing our planet Solar energy is not a revolutionary technology by any means and has long been criticized for it s inefficiency and other shortcomings Modern advances by companies such as DayStar Technologies have made great gains towards reducing the power conversion inefficiencies and have succeeded in creating a solar cell foil that is close to 20 efficient as opposed to the traditional 10 and have greatly reduced costs with CIGS technology 1 Our product addresses the other implementation difficulties related to solar energy primarily What do we do when the sun goes away every night We ought to store the energy right And how do we propose to do that Well with batteries of course But its not that simple the battery currents and voltages need to be regulated to provide optimum power transfer to the client because efficiency reduces cost and high cost is the last major road block in making solar a feasible
15. V RS R W 0 0 No data should be transferred to or from the display during this time DB 0 DB 0 Wait more than 4 1ms RS RW DB 0 0 0 DB DB 1 1 Function Set Command 8 Bit interface No data should be transferred to or from the display during this time DBs DB 0 1 DB i Function Set Command 8 Bit interface No data should be transferred to or from the Wait more than 100us display during this time RS RW 0 0 DB 0 DB DB 0 1 DB i Function Set Command 8 Bit interface After this command is written BF can be checked RS RW 0 0 DB 0 DB 0 DB DB 1 0 Function Set Sets interface to 4 bit Function Set Interface 4 bit Set N and F for number of lines and character font Display OFF Clear Display Entry Mode Set Display ON Set C and B for cursor Blink options Note BF should be checked before each of the instructions starting with Display OFF Initialization Complete Display Ready This initialization will need to be a subroutine so that every time the program needs to write to the LCD this subroutine will be summoned In order to program not just one letter but strings of letters a character array method will have to be created The font table in Table 1 is the data streams that need to be sent to the LCD in order for it to display its respectiv
16. aced We had been storing the board with Jack Gilmore and through relocation of offices our board was lost and never found So another board was etched and soldered The replacement prototype was completed the week before engineering days hardware and software debugging was necessary A meeting was set up at AE to debug the circuit and we were able to debug the circuit using current voltage flow techniques Our advisor at AE was going to be unavailable for the following week so we obtained the board for the software group as we were fairly sure the hardware was debugged After realizing that the other group was not getting expected results we had a full meeting to look at the circuit which turned out to still have software and hardware incompatibilities The new LCD screen did not work with the fresh circuit and it also did not work with the old circuit so we used the other LCD which worked with the older circuit and it also did not register on the new circuit either Therefore we understood that there were still at 18 least hardware problems involved We think the LCD problem might be involved with the LCD connection but we were not confident if that was the case or how to fix it We were at least about to find out that the pulse width modulator was functioning correctly and that we were able to communicate to the processor Future work that will be needed will be a more thorough debugging of the circuit and LCD connections establish the RS
17. e character 7 Table 1 Character Bit Stream Table eee el le ee ep Ei a a SE Se ee RBBB XXX X0001 2 H E ev a a oe BRB CKY ee ee ae ee ee eee i Te e pe LESTE SLD e ee ee Re EEE ae alr T 4 j ae xem oF 4 B T fot t S T b PR ye a TETEM eta wwa 3 ELI eja JAY FD a Se a ROE XXXXOLLO De oy z ee REE ee a a a ee s o m r mw ol BIH 8 x a128 rz EF ae ee er ee a a Er mn Lol T a aT e g EE SS PSS J Z zl lalnlel jl Cane a FEREN EAN EPS DE E F a jee oT Trew ECEE TE EE STE OF Tr ine Ge ee ead Soe a er xem olt 6 LF IE e319 ee See es Pi Jira Eos false Pin area tr a ee ote Character of high order bat 1110 and 1111 may be madequate 25 Once we successfully wrote a character or string to the display we implemented a menu system This menu system will be navigated by 3 buttons that are attached to our microcontroller I O RC3 RC5 Two buttons will navigate up and down and the third button will be a select This will allow for browsing and calibration of the duty cycle to the pulse width modulator After we implemented the LCD we began to work with the pulse width modulator on the microcontroller in order to change the power from the buck regulator Pin RC1 is going to be hooked to the buck regulator which will change the pwm to produce the appropriate voltage that has to go to the battery Lead Acid batteries charge in 3 stages so in order t
18. energy alternative The heart of our product then is microprocessor control of the stored energy The approach taken by our team of engineers was essentially an embedded systems approach We needed to interface the microprocessor which we programmed with some current voltage limiting hardware a buck regulator opto isolators and various other necessary components to effectively read the currents coming off the panels and entering the batteries and those currents leaving the batteries to power some desired application It s true our project as described is not breaking any real ground however most products meant to regulate a battery voltage are completely analog and offer very little control to the user They are also not digitally self optimizing which is a possibility for us As far as future improvements and evolutionary capability we are offering an extra buck regulator component and a few pins for a future expansion solar array as technologies in this sector improve to generate even more power We suspect we can be competitively priced if not cheaper than our closest competitors who don t offer expansion options and digital flexibility Il M IV VI TABLE OF CONTENTS Title Abstract Table of Contents List of Figures and Tables Introduction Summary of Previous Work Design A Hardware B Power train C Software Fabrication Budget Conclusions and Future Work References Bibliography Acknowledgements
19. energy system has never been greater This is why our market of interest focuses on the residential community to help offset the cost of utilities while striving to be environmentally conscientious One group Energy R evolution claims The sunlight which reaches the earth s surface is enough to provide 2 850 times as much energy as we can currently use On a global average each square meter of land is exposed to enough sunlight to produce 1 700 kWh of power every year 1 These are encouraging statistics when coupled with advances in our field including the conversion efficiency of the solar cells with companies like DayStar Technologies and ENREL developing a solar foil which is nearly 20 efficient up from 10 in the previous generation 2 However the energy from this foil would be useless if it could not be stored in some manner for use at another time when the sun may not be as readily available i e batteries The field of solar energy engineering requires products like ours to maximize this power storage and monitor the power levels Our project is essentially all of the technology that must be in place in order to interface the power generating solar cells to the power storing batteries In order to accomplish this we must choose a microprocessor generate its necessary control signals read in the voltages and currents coming off both the batteries and panels and use a buck regulator to manipulate the voltages and currents accordin
20. g to an optimization scheme regulated by the microprocessor With the guidance of our industry expert Jack Gilmore we plan to demonstrate this controller at scaled down operating levels at Engineering Days in the spring of 2007 We intend to have this project demonstrated outside with bona fide solar panels and not simulated ones Although this is what we plan to accomplish for the end of the year the overall goal for this assignment is to implement this technology into a household environment The following is a mid project review of what we have accomplished so far the problems we have encountered the possible solutions and our plans for next semester In chapter II we ll start with a recap of the previous prototype its successes and ultimate failures Then in chapter III we will elaborate on the technical aspects of the hardware power train and software of our design We ll discuss hardware design in detail explaining what we added from the previous year s effort and what we removed the options available with the extra pins on our microprocessor isolation of different sections of the circuit to provide protection to the more sensitive components and explanation of the circuit functions Next we ll discuss the hardware involved in the buck regulator circuit in more detail as well as discuss our decision to utilize the rectifier chips in place of some other components Next we ll focus on the software side of our project what
21. it We felt that this component was unnecessary as we could use the power train to effectively control the voltage Here is the unsafe boost voltage circuit Figure 7 Unsafe Boost Voltage Circuit 15V 1 Ue UNSAFE BOOST VOLTAGE ASE P4 T TEE R39 R43 Taw 2 saw 521 4W 11 Now that we had three free pins to use on the processor we just relabeled the pins putting the duplicates next to the originals like this Figure 8 New Pin Layout 12 Figure 9 RS232 with Isolation Couplers 51 R17 L A C6 k O a g 55 1 1 6W MAX232 R16 us Sik SOOO ix 1 8W In order to communicate with our processor we will need some hardware safety measures to accompany the components needed to connect the RS232 port for an outside computer in which to monitor the status of operation For this situation iso couplers were used so we can transfer signals from the processor to the computer and back while keeping both circuits electronically isolated A separate isolated power source is connected to MAX232 chip which is also needed because the voltage standards used in the processor differ from that of the computer for example transistor to transistor logic coming off the processor uses a logic 1 specify a line voltage of 5 volts and a logic 0 is a line voltage of 0 volts whereas in the RS232 the logic 1 refers to a voltage of 5 to 15 volts and logic 0 refers to
22. k the data the ground and the voltages as seen in Figure 19 Figure 19 MPLAB ICD 2 interface connector VDD PICXXXX VPP MCLR Connector Once the ICD 2 was in place on the prototype board we tested to see if we could program the microcontroller The results of the testing were that the ICD2 debugger would not connect to our microcontroller when we were using any version of our IDE newer than v7 00 Once this problem was solved we soon realized that our C18 compiler would not compile with any IDE versions older than 7 22 The solution to this problem with the advice from the people at microchip was to take out a few resisters on our board that were needed in the v7 00 but not needed in v7 5 We then wrote a program to display text on our LCD At first we simply initialized the appropriate registers and tried a print Hello World For our testing phase we used a simulated UART which displayed the correct print statement However our microcontroller is hooked up to our LCD using 4 I O pins and not a serial 23 connection The ports RBO RB3 are used to communicate to the LCD display We initialized these I O ports using our C18 Manual which came with the compiler and tried following the steps in Figure 20 to initialize the LCD in order to transfer the letters to it Figure 20 4 bit Initialization Figure 20 4 Bit Initialization 2 2 2 2 4 Bit Initialization Wait more than 15ms after Vec 4 5
23. le in two semesters Although we made great gains this semester in the hardware components board fabrication and software interfacing we didn t actually get all of the code in place necessary to make the measurements calculate the required feedback duty cycle and allow the unit to self optimize its power curves unattended For next semester we are recommending that two software students very familiar with C embedded systems PIC programming finish up the software necessary for this project s ultimate success Our problem was in not having the sufficient tools or experience necessary with these kinds of devices to be able to isolate the debugging issues of the software to one localized arena For example if we received an error during execution we weren t sure if it was a Microchip MPLab software run time error whether we had failed to initialize a register on the PIC properly for program download and execution or whether the biasing components on the PIC processor pins had a ground fault or other hardware related problem This ambiguity led to a lot of headaches and probably the failure to meet our goals As we turn the project over to fresh hands the current team has finished all of the components necessary for the hardware and power train to function properly All of the wires chips R s L s C s have been chosen and connected and we assume will function properly We also have successfully interfaced with the computer through the ICD
24. n LEVEL PULSE HO ited B VouNos goer FILTER z HIN eg 5 LEVEL A pa SHIFT 4v COM 21 C Software Our first task was to choose the microchip that we were going to use for this project We chose Microchip microprocessors as they are free to sample and if sent into production the microprocessor is cheap and powerful Our initial design specifications were to have a 10 12 bit A D converter several I O ports a pulse width modulator at least 3K Bytes of RAM and 1K of EEPROM We chose a Microchip PIC18F2620 which is a 28 pin package with 10 bit A D conversion C compiler optimized architecture and 3K of RAM and 1K of EEPROM We could have used a PIC24 which is a 16 bit microprocessor but it was a 64 pin package and the project that we were provided with from previous work is also a 28 pin package The reason we chose not to go with the microprocessor from the previous work was due to the fact that it had insufficient memory and only an 8 bit A D converter The previous project required more decimal accuracy so we went with the 10 bit A D converter The software team s job is to interface the circuit and microprocessor with the batteries and solar panels The main goal is to measure the maximum power distribution in order to distribute power to and from the batteries from the solar panel and to the home or electrical device that will be used In doing this there will be an LCD display and menu system in order
25. n your project is at the very beginning so a strong understanding of your design goals and a pathway to success are vital to project completion Although we had met over the summer and established a general understanding of how the unit should work we did not get a technical understanding of the project until it was much too late to meet most of our goals we had outlined Also scope of an engineering project is a very important boundary to get nailed down Along the way we added the second panel array components and capability before we had even prototyped the first array which was a distraction On the topic of prototyping we discovered too late that a breadboard implementation of the PIC microprocessor and associated electronics would have facilitated easier debugging and should be extremely valuable to a software team next time To sum up the progress we made on the project this year though considerable was ultimately hampered by scope broadening time scheduling and insufficient tools knowledge for debugging a PCB recommending breadboard Hopefully next semester s students will benefit from our extensive documentation and complete the software in time for E days 08 31 References 1 Energy R evolution energy technologies solar power photovoltaics Online Document 2007 April 8 Available at HTTP http www energyblueprint info 30 0 html 2 DayStar Technologies Products Technology Copper Indium Gallium
26. o do this correctly with a microcontroller we will need to monitor both time using the clock and voltages using our op amps Stage 1 takes about 5 hours and the battery is gradually charged to 70 Stage 2 takes about 5 hours and the charge current is gradually reduced as the cell is being saturated Stage 3 is the float charge which compensates for the self discharge In order to implement this we will use a method with a case statement in order to judge what state the battery is in currently Stage 1 will simply use constant current to charge Stage 2 will use constant voltage with around 2 4V to charge Stage 3 will use a float charge or constant voltage of 2 25V Lead acid batteries are also temperature sensitive so a 2 25V charge will need around 25 C A temperature gauge will not be implemented so the owner of the battery charger will need to take precaution as to not overheat or freeze the batteries The stages of charging will be implemented using global variables as for later calibration if needed using the LCD menu system 26 IV Fabrication After the power train group had finished their design of the power system they focused their attention on building the prototype The specification for this portion of the project again came from our industry advisor Jack Gilmore He told us that he wanted all the components contained within a single chassis with inputs for the solar panels and outputs for the battery array and the circuit panel
27. pability to the original circuit to be able to work with an additional power source such as an additional set of solar panels The first step in doing this was to add in a second voltage meter to detect the additional voltage from the second power source Here is the original voltage meter circuit Figure 2 Original Voltage Meter 5VREF 01 To do this we used the extra external pin on P9 and added in a duplicate voltage detecting circuit line in with the two existing ones However as you can see D1 only allows for two zener diode connections from the meters to 5VREF Since we knew we already had multiple two zener components we decide to add in an additional two zener HSMS 2804 part to accommodate the third meter Here is the new set of voltage sensing circuits Figure 3 New Voltage Meter 5VREF 09 4 2 HSMS5 7604 peas Be 2 3 HSMS 2804 R1 R3 R19 K K 12 4W121 4 The next part to add was an additional current meter for an additional power source Here is the original Current meter Figure 4 Original Current Meter Solar Bott 01 i IOW 6 5 R32 CLARE R31 ON 8 U6 8 u 174W Ig aL tea ne ADB572R 7a l 63V Z R28 To implement the second current meter we simply had to add in an additional external pin set and a duplicate meter circuit to the schematic Here is the new current meter circuit Figure 5 New
28. ping capacitors on the front end of our buck regulator where the panels connect to the gate of the switch We discovered that a large foil capacitor could smooth out our ripple current and withstand the 30A max current while an array of electrolytic caps about 60 uF would smooth out the large swings in voltage 0 380 V stemming from alternately connecting and disconnecting the panels Our research showed no single capacitor that would satisfy these two criterium so we connected both types of caps from the panel Vin node to ground in parallel When it came to assembling this buck regulator circuit we used a large heat sink donated from Advanced Energy to mount the FET and diode on and we cut up rolls of sheet copper in 20 wide strips to make the connections between the caps FET and diode because we wanted to minimize stray inductance in these time constant critical switching components The other connections to the inductor and external components were made with regular larger gauge wire where stray inductance wasn t an issue Figure 17 Typical Connection of the IR2110 regulator chip Typical Connection up to 500V or 600V Refer to Lead Assignments for correct pin configuration This These diagram s show electrical connections only Please refer to our Application Notes and DesignTips for proper circuit board layout 7 Figure 18 Schematic diagram of the IR2110 regulator chip ee s g HV
29. rd is for prototyping it is not the fastest of board making machines as it took nearly four hours to create our board Once the board was complete we checked it for errors and found only one where part of the connection pad for a capacitor had some copper missing Jack Gilmore discussed that this could be fixed just by realigning the capacitor so we did not need to recreate the board With a completed board we were now ready to start populating the board with parts To do this we went through all of AE s part bins to compile our list of necessary parts With the parts gathered we were ready to start soldering However neither of the hardware group members had ever soldered a part before so the second step in populating the board was to learn how to solder We learned to solder from Jack by de soldering and re soldering some of the motherboard resistors on an old board which took quite some time learning the correct amount of solder to be placed and learning some tips to increase the speed of soldering a part It would be good to note that overall the parts were fairly inexpensive under 10 not including the chip which is covered in the original budget and the soldering tools were about 40 total Once we knew how to solder the parts correctly we now proceeded to add parts to the board smallest first For new solderers this became a lengthy process taking about two weeks to get about half the board done At about this point our board was mispl
30. rgy When we received the chassis we began drilling holes and attaching components according to our design By E Days we had all the components attached and everything wired up 28 V BUDGET The budget for this project is based on fifty dollars per student per semester This makes a total of 600 00 for this project The following is a cost breakdown of our expenditures thus far and our planned expenditures for the following semester Table 2 Budget Costs Count Description Cost Total Cost Charge to Project Circuit board components o Microchip PIC18F2620 28 pin J f Sample All components 4065 Donated Eo a a e Components ofthe Charger 0 i Donated 1 Donated 1 Donated 1 Donated 3 Donated 1 43 00 Donated 1 Donated 1 23 94 Donated 1 Donated 1 19 62 Donated a a Donated a ey Pe ae 4 Donated i Donated i Donated a ne O ee ee In circuit Debugger i 16949 Trial offer Ca ae ree aoe ee Equipment used to make the board J Use of donated Use of donated ra Miscellaneous parts and tools 35 00 35 00 35 00 Miscellaneous administrative costs 6 00 6 00 6 00 Future Cost to the project 4 Aliuminum extrusion 1x1 x8ft 11 24 44 96 44 96 Total charged to project 21049 Left in project 389 51 29 VI CONCUSIONS AND FUTURE WORK In conclusion the solar powered lead acid battery charger project has proven to be more work than we could hand
31. ther the solar panels or the battery Component B is the voltage sources to be used within the layout Component C is a high precision 5 volt reference used to power the processor the circuit with components listed W1 through W4 are the grounds to be used for the layout Component D is the clock that will be used for the pulse width modulator Figure 14 Voltage Sources 5V 15V C12 5V 4 asy 5 wi 15 at lt we 3 t 63V is 7 O i Lo pa Tas 5V C VR2 AD5S86KR SV REF SVREF 5V id ee 16 Here is the final schematic we will be using for the remainder of the project Figure 15 New Schematic 4 Fev one DESORETON OF HANGE Rerne onone prone C E Pa eee w F Ed bs rel ne Re ate fe 3171 80 R28 P1 ay v Ba a By 12 178W c29 j EP aae ae maa 3 ae fa F ase wae frt Be c m os naw sol osx 59 A Nga B Ras 12 4W gt WY ove wor egt n PEE AG sive satan pers ee v Wik t Y f2 aw 7 ig le ca m we 6 we ra LE AK aad Se rd et 1214W zok 053 au Bow et tue 7 pe e Shy Gee wien Let Ge See A003727 Tes B Rae v B 7 Belay a gt _ c b cug ga S BWO eT Tew av w We 1 yea Al le
32. tting the bulk project into three smaller sections hardware power train and software each consisting of two engineers we hope to share the workload evenly and thus be able to place greater effort on the software to bring this project to fruition HI DESIGN A Hardware We would like to first cover the things from last semester At the beginning of the semester we started with the previous circuit that Jack Gilmore had attempted to implement First we would like to explain what we added to the original schematic Figure 1 Old Schematic tH EEES s Z 2d UNSAFE BUCK VOLTAGE cae UNSAFE BOOST VOLTAGE E ais ae ig tres E a IN OOOO O oy 4 Hags gta is SAFE VOLTAGE aa SVREF a meU UNSAFE MINUS HIGH VOLTAGE ay 1 page E B a TE F F ee sar SGAE Bownns Switches alk ICSP_conn A at ea gla Dose A bia dea Atma O Oa ges enana aans rwmeroormicarems LC fn LC Oran ini an ea aeRO R E PROPRIETARY INFORMATION The very first change we made was to put in a better microprocessor The details of which will be covered in the Software design section It then became desired of us that we should add the ca
Download Pdf Manuals
Related Search