Home

Self-sufficient sensors - Worcester Polytechnic Institute

Contents

1. 45 Figure 3 15 XCTU Interface with No Radio Modules Connected neee 47 Figure 3 16 XCTU Interface Adding a Device a 48 Figure 3 17 XCTU Interface Configurable Parameters eene 49 Figure 3 18 XCTU Interface Addressing Fields nnssssnsssnsenssnnsssennnven 50 Figure 3 19 XCTU Interface Data Visualization Console nnen nennen eenen eneen ennen 51 Figure 4 1 Indoor Test Environment for the PV Module nnee eenen 53 Figure 4 2 Power and Voltage Plotted Against Resistance oooooncnccccccnnnnncnnnonnnnnnnonnncncnnncccnnnnnnnnss 54 Figure 4 3 Power and Voltage Plotted Against Resistance Scaled Axis eee 54 Figure 4 4 Power Plotted with Trend Line Against Voltage ccccccccccnccnnnnononnnoonnnnnccnnnnnnnnnnnnnos 55 Figure 4 5 Evaluating the Local Maximum of the Power Curve eneen een eenen 56 Figure 4 6 Comparison of Power Vs Load Curves in Different Light Environments 57 I139u1e ders Cardboard Cover im POSIUOT Swes oxssetice setts sah cabeeacdvnads saaradeeaneusaueaabsaoucanacnatedaacaseeaseesaeeee 58 Figure 4 8 Outdoor Environment for Testing the PV Module nennen 59 Figure 4 9 Indoor Versus Outdoor Power Characteristics eese 60 Vill Table of Tables Table 2 1 Electrical Specification of the XBee Wireless Module eenen Table 2 2 Comparison of Battery Characteristics by Chemical Composition
2. Table 4 1 Data Regarding Road Simulation Testing IX 1 INTRODUCTION In this section we discuss our motivation to create a remote sensing unit that is powered using renewable energy We describe the flow of ideas in chronological order that gave us incite on our final design We then describe our project and detail the criteria we extrapolated from our background research Finally we mention existing products and patents that are relevant to our project 1 1 Motivation Many fields of science and technology are dedicated to monitoring our world Some of these fields are on the cutting edge of technology and use sophisticated equipment to remotely monitor dangerous environments For example the health of agricultural crops can be measured by autonomous drones equipped with infrared cameras 1 and devastating tsunamis can be detected hours ahead of landfall by deep sea underwater pressure sensors In both cases remote technology can dramatically improve the accuracy of data while simultaneously reducing maintenance costs and risks posed to humans by dangerous and demanding tasks Yet there are still environments in this modern world of ours that lack this important technology Most instances of wildfire detection are still conducted entirely by human observers and the structural health of many bridges is still monitored in person In these fields the accuracy of the data collected is still prone to human error and the environments a
3. 0 10000 20000 30000 40000 50000 60000 20000 BOC SOc 100000 0 100 200 300 400 500 600 700 800 900 1000 Load k Ohms Figure 4 9 Indoor Versus Outdoor Power Characteristics We were unable to collect measurements with a high enough resolution between resistance values to generate a power voltage graph that could be used to calculate an ideal load for maximum power however as the above figure shows that value is just shy of 1000 Furthermore the shown data was collected on a clear day As previously discussed the ideal load changes with the environmental conditions so load would need to change with the environment cloud cover sunrise sunset etc in order to maintain maximum power 4 2 Power Generation Based on Car Speed Most of our results involved using cars driving at or close to highway speeds Given our location and resources we were unable to safely test our product at highway speeds However we talked to Brian from CMSC Central Massachusetts Safety Council who allowed us to use a test track in West Boylston 60 MA At this track we were able to safely reach a top speed of 40mph not quite highway speed but sufficient to test whether or not our idea is viable The data we collected focused on the wind speed and power generated from purely wind produced by a car passing by We were unable to test wind speed from the car wake in ideal conditions As a result our data has factors that were out of our control such as n
4. Figure 3 6 Safety Tubing for Power Cable from PV Module 35 Figure 3 7 Cable Connector to be Attached to DC Generator Because we are using two different sources of energy we needed to isolate each connection so that the PV module would not supply power to the generator thus turning it into a motor and the generator would not apply power to the PV module which could damage the PV module To achieve this each socket is connected to the battery through a Shockley diode which are used because they have a low forward voltage bias around 0 5 V This 1s important to us so that we do not lose power when the turbine or PV module is not producing a lot of power 36 e q e LE i Li r e A A r x g a J a e a le oocoonooo 250001 E 7 inooEDEDdE US OoOoO00 9 noooo _ OOOO ksa 40 o f Charge Controller and Circuitry E S o Peel eee SININNUISN gt ezel S cea svxaj fi sul era ba U Feat x Mesa gt weal FAT Je asil esa 8 veel y e l x o 2 ap aa EHE e A S aw aa Figure 3 8 6ZSG4 E dx3 Lee 37 Top View of Assembled Circuitry Figure 3 9 3 2 Software The next two sections will detail how the software for both the MSP430F5529LM and the XBee S6B were implemented 3 2 1 Microcontroller MSP430F5529LM Our Software Implementation was programming the MSP430F5529 using C in Code Composer Studio CCS CCS is a product d
5. TIMERO A0 VECTOR _ interrupt void TIMERO AO ISR void g_slowdown_loopl if g slowdown loopl1 gt SPEED 1 g slowdown loopl1 0 g slowdown loop2 4 71 if g slowdown loop gt SPEED2 1 g slowdown loop2 0 ADCIA2CTEO ADCL2SC Start sampling j j pragma vector ADC12 VECTOR _ interrupt void ADC12 ISR void g ADC samples g adc index ADC12MEMO g adc index ADC121FG amp 0x0000 if g ADC samples g adc index 1 gt THRESHOLD g emergency 1 trip the flag to send buffer at end of buffer j if g adc index gt MAX 1 I think this should be a multiple of 4 for most efficiency no useless data written g ade complete 1 g adc index 0 pragma vector USCI AO VECTOR _ interrupt void USCI AO ISR void j void write to mem void unsigned int ADC samples buffer MAX unsigned int 1 unsigned int TxFlag 0 for 1 0 1 lt MAX 1 ADC samples buffer i g ADC samples 1i j unsigned int j while FCTL3 0x0001 BUSY wait for not BUSY j for j 0 J lt MAX j j 2 Looping write blocks to store all values from ADC samples buffer if FCTL3 LOCKA FCTL3 FWPW LOCKA LOCK unlock the memory to write i else FCTL3 FWPW FCTL1 FWPW WRT BLKWRT set to write and do block writing int value ADC_samples_buffer j int value ADC samples buffer j 1 starting address for BANK A flash write pq 22 F5529 data sheet start 0019FF end 00
6. Transmit in 8 bit segments 1 start and 1 stop bit UCPEN UCAOCTL1 UCSWRST software reset UCAOCTLO UCMODE1 UCMODEO UCSYNC UCMSB UC BIT UCSPB 88 o UART Mode LSB First 9 bite one stop bit UCAOCTEL UCSSELO 7 7ACLK select UCAOBRO 0x03 UCBRx 3 UCAOMCTL UCBRSO UCBRS1 UCBRSx 3 UCAOMCTL amp UCOS16 No oversampling UCAOCTL1 amp UCSWRST clear bit to get transmitter ready UCAOIFG amp UCTXIFG buffer is ready to be written to P3SEL 0x08 set P3 3 for peripheral P3SEL amp Ox7F set P3 7 for I O This is for Sleeping request P3DIR 0x80 set P3 7 to output P3OUT amp Ox7F set P3 7 Output to be 0 to wake up XBee unsigned int 1 unsigned int firstFlag 1 unsigned int secondFlag 0 73 for i 0 1 lt TXMAX 1 send the Tx data TXMAX many times loop int current int 0x1800 the current flash memory address to Transmit while current lt int Ox19FE 1f UCAOSTAT UCBUSY 0x00 when done transmitting previous info if iirsiFlag int tmp read memory current gt gt 8 shift to the right by 8 to get the second half of the int UCAOTXBUF tmp Livete lag 0 secondFlag 1 1E UCAOSTAT amp UCBUSY Ox00 when done transmitting previous info if secondF lag UCAOTXBUF read memory current when we write to this the transmitter will automatically transmit after writing current tirstF
7. 4X Characterisucs oFthe ALTIOA2P pen ntt Nede 22 42 Power Generation Based on Cat 5peed edet Unde Ian Ru De qua edna deren 60 4 3 XBee Transmission Range Stationary ooccccccccnonnnocnnnnnnnononononnnonnnnnnnnnnnnnncnnnnnnnnnnnnnnicnnnos 62 44 XBee Transmission Range Drive By iui rei oe ore il 62 45 Summary or RESUMES cinerea caso iio cee a DU OE ce uiae t 63 I CONCESIONES 64 IE SOMDA EEND 64 D2 Project Recommenda non Sisar medie di dud 64 BDOS apeere rl D UTD 67 APPEAU MATO A A ee 70 Appendix B ALETIO I2P Indoor Test Data ioc iion doe predetto niue oh tnos aeu paulo Mala canes 75 vi Table of Figures Figure 1 1 Overview of US Energy Consumption by Source and Sector 2013 eeeeeeeeeeeeees 2 Figure 1 2 Individual servicing Tsunami sensor buoys 4 nnee eee nenenennennnenven 4 Figure gt Voltree Bigeneroy Harvester dete eei cu vous in ossi cutie abana ate mentees 5 Figure 2 1 Monocrystalline left and Polycrystalline right PV Modules 1 Figure 2 2 ALT10 12P Polycrystalline Photovoltaic Module eneen 12 Figure 2 3 Turbulence Models of Modern Horizontal and Giromill Wind Turbines 13 Figure 2 4 Visual Comparison of Vertical and Horizontal Axis Wind Turbines 14 Figure 2 5 Darrieus Wind Turbines With Savonius Type Start Up Rotors 15 Figure 2 6 Effects of Offset Blades in a Savonius Ro
8. wanted to use the large code memory banks to store our information but we found that the banks held important data for our code and would cause the program to crash if erased As an alternative we decided to use the information memory The information memory is segmented into four banks the size of 128B giving us 512B total The total size of the bank is sufficient enough for us to store a value every 6 minutes as mentioned previously in section 2 2 4 1 13 The flash memory default mode is Read mode In read mode writing and erasing the memory are disabled In order to write to the memory four steps must be followed in order To write a segment to the memory the watchdog timer must first be disabled Next the flash controller must be configured in the way you want it and the WTR bit must be high After that we can write the data into any memory segment by providing a pointer to the address Finally once the data is written the WRT bit is set low and the LOCK bit is enabled Although it is not illustrated the LOCK bit must be 0 during the write to memory 13 The steps are illustrated below Disable watchdog Setup flash controller and set WRT 1 Write byte or word Set WRT 0 LOCK 1 reenable watchdog Page 347 from the MSP430x5xx Family Users Guide 21 In order to erase data from the flash memory five steps must be followed in order First the watchdog timer must be disabled Next a busy bit needs to be polled unti
9. Asynchronous means that the USCI and the peripheral do not need a shared clock in order to operate Synchronous on the other hand does need a synced clock Due to our interface with the XBee we did not need a shared clock thus choosing UART mode for our USCI 13 27 UART mode on the MSP430F5529LM operates using two pins RxD and TxD for receiving data and transmitting data We were only concerned with the transmission pin sending the data to the XBee module UART mode can operate in 7 bit or 8 bit with odd even or non parity We chose to operate the UART in 8 bit mode because the data we are transmitting is a factor of 8 16 bit integers which drastically simplifies the data transmission and receiving Odd and even parity are very simple ways to check for data error when using asynchronous communication In order for it to work both the transmitting MSP430 and the receiving XBee need to have the same parity For simplicity we decided to choose non parity and transmit errors if any arise 13 Another initialization for the UART mode was the order with which we sent the data Least Significant Bit LSB first or Most Significant Bit MSB first We opted to use LSB because the XBee sends receives in LSB Having them both be the same would keep our data in the same order thus making it easier to interpret and receive The final initialization we did for the UART mode was the programmable baud rate The baud rate is how fast we are
10. FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF 0606 C735 2955 1217 0A08 2659 4711 0023 002A FEF8 1011 8004 FFFF 086A 9DC 0657 076D 0506 05E3 0612 7F0A 7E81 7FA7 6302 8A08 860C 2A0E 2E12 9622 2100 2300 0900 0F0G0 0500 FC 4110 3002 3802 3C01 3D00 4400 4000 4801 4202 A003 1001 5104 5202 5302 5402 SFAC 6202 6104 6704 6104 680A 8502 4704 900C 9004 D110 A81C 9804 64A8 4065 9190 60DO 9861 6246 5063 9392 6766 6851 FFOO FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF 0120 A500 012E 1100 0110 96E2 011C 2600 011
11. Launchpad development board both made by Texas Instruments We chose this component for the following reasons First all three members of the project team have experience working with the MSP430 family of microprocessors A big step in working with a microcontroller is becoming familiar with the architecture Initializing a new board can take time especially if the engineer is unfamiliar with the board Secondly the chips in the F55xx class of the MSP430 are specifically designed to be used in ultra low power applications With a rated power consumption of 290 u A MHz at 8MHz when fully active and 2 1 u A MH in Low Power Mode 3 the MSP430F5529LP easily confines to our strict power budget in our design 12 18 Finally the MSP430F5529LP provided a suitable analog to digital converter ADC programmable flash memory and user asynchronous receiver transmitter UART communication all of which are needed in our design The following sections will detail those features 2 2 4 1 Analog to Digital Converter ADC The MSP430F5529 boasts both a 10 bit and a 12 bit ADC at a speed of 200 kilo samples per second ksps We decided to go with the 12 bit ADC because it gives us a resolution of 610uV per bit as compared to the 10 bit ADC resolution of 2441 V The 12 bit resolution calculation is shown below 212bit 4096 Divisions Full scale range 2 5V OV 2 5V OV Resolution In other words 610UV is the smallest
12. a sort of emergency signal iii Acknowledgements The authors would like to acknowledge the following individuals and thank them profusely for their support of this project Bob Boisse ECE Shop Bill Appleyard ECE Shop Lifeng Lai Advisor Yehia Massoud ECE Department Head 1V Table of Contents PRO SUA A E ees ii Executives UIDI eo cde da 111 ACA VA Ts o ers DR iS INS E RD HT iv Table ODE PITS vaandel stede Ret eda mato shawnee vii jr ERI PT Tr rr ix t INTRODUCTION or l E PE Costes tenes ie cox du Diete cot EROS Onda ie cioe roe ecol enu O case Cio c d opo oel mop UA c EO SIO DdacE I PME STEM e ii Daad 3 Za sun Dec dte orae nennen keae tds 3 ho Wilde DT 4 AA o o II O doede Maddens 6 LR GS sl Ree 7 Ir O 7 LA Existe Products and Patents ins 8 2 METHODOLOGY eyes deedeetee iva o ibo ue vadens 9 21 SET Regue MENS ai da Losses aa Iu Lob R ue aeos pepe Donee tete 9 DONNE Gero AS IS I UU eines eae 10 ZI SOLI PA a M MEE DRM AR AH UE MEM MD SR 10 2c AD DEDIT es 12 PAPAS MEO TA DCINDC 17 22A Mierocontrolleb unite eea A 18 Ld 0 p 23 220 Di m A A 26 24 Battery and Charge Controller iia aaa ens 27 3 JZMPEEMENDPDATION rn 2 2 20820 62 piste tat te and a 31 Sx MEME 6 E euer ee Pr re ne eee erence eee 31 PE ROOD Wale eeen 38 24 1 Nicrocontroller MSPA30F 55291 Maio 38 AO Wareless NiO GUNG sd ep Costin dev PE e AU PE ARAS 46 A SIRES Se B 52
13. able to transmit receive the data The baud rate on the MSP430 can be set to a variety of different preset values We chose the value 9600 because that was the default value on the XBee module Having the same baud rate will help minimize errors in the Tx Rx process 13 2 2 5 XBee To transmit the data we chose an XBee wireless module The XBee is a popular component among wireless engineers because it is relatively easy to program and operate When two or more XBees are in a system they will automatically connect and communicate with each other making wireless communication relatively easy The interface and protocols between these two XBees transmitting and receiving is not important because two XBees connected in the same system will recognize and connect to a similarly designed XBee This means the protocols and interface could be anything as long as the two XBees are the same making the data we see on the output of the microcontroller s UART pin the same data as what the receiving XBee displays In this way the XBee is modular for all sensors and the data the sensors produce 23 The XBee consumes up to 309mA of current when transmitting far more than the microcontroller This is the main power consumption of the entire system as wireless transmission usually requires large amounts of power However the XBee consumes very little power when not actually transmitting as there are two sleep modes that can be activated by setting a slee
14. amp Add devices or Discover devices to add Change between Configuration radio modules to the list EJ Consoles amp Network and amp Device Cloud working modes to display their functionality in the working area Figure 3 15 XCTU Interface with No Radio Modules Connected When the XBee is plugged into a USB port in a laptop clicking the Add Devices button in the top right of the screen will allow the user to choose a port with a device connected as seen in Figure 3 16 47 XCTU B LA J Add radio device cg A Add a radio module A A A A j Select and configure the Serial USB port where the radio 4 module is connected to 9 f Click on amp Add devices or Change between Configuration Select the Serial USB port Discover devices to add TUNE E m Consoles d Network and A COM5 USB Serial Port radio modules to the list D Device Cloud working modes to display their functionality in the working area Refresh ports Baud Rate 9600 v Data Bits 8 v Parity None v Stop Bits 1 v Flow Control None v The radio module is programmable nera mos rio Set defaults d Os all Figure 3 16 XCTU Interface Adding a Device The only available port here is COMS5 where the XBee is plugged in Clicking Finish will tell XCTU to look for an XBee and add the resulting XBee to the list S
15. changed The most important of these is the SSID field which contains XBee 004095DB7CA the name of the other XBee that is transmitting our data From that field and the Channel and Association Indication fields under MAC PHY section we can determine we are connected to the other XBee If the XBees were not connected then Channel would be FF and Association Indication would be some value besides O 0 indicates successful connection Both of these fields are read only given our configuration 16 49 XCTU jJ amp a x Beo Radio Modules Radio Configuration 000000409D5EB5EE Tes MESAPA EN NE Lo LARA A Port COMS 9600 8 N 1 N AT MAC 000000409DSEBSEE is A t MAC PHY v gt Network v Addressing e Change Addressing Settings y SH Serial Number High 40 La i SL Serial Number Low 9DSEBSEE F oz DJ DL Destination IP Address 255 255 255 255 OLO E D NI Node Identifier IS El l KP Device Cloud Description SA ke KC Device Cloud Contact amp 9 a 1 KL Device Cloud Location 9 CO Source Port 2616 SA Q DE Destination Port 2616 SA On eel aa Jel MI Omur een OR gt Serial Interfacing s 1 0 Settings kb MANC r Figure 3 18 XCTU Interface Addressing Fields The four modified values visible here are from the Addressing field section These are automatically modified for us by the DHCP Dynamic Host Configuration Protocol option previously s
16. for in a statistical energy balance Conventional hydroelectric power geothermal solar photovoltaic wind and biomass before any transformation to secondary or tertiary forms of energy for example coal is used to 5 Includes industrial combined heat and power CHP and industrial electricity only plants generate electricity Sum of components may not equal total due to independent rounding Includes commercial combined heat and power CHP and commercial electricity only Sources U S Energy Information Administration Monthly Energy Review May 2014 plants Tables 1 3 2 1 2 6 Figure 1 1 Overview of US Energy Consumption by Source and Sector 2013 In recent years social awareness of non renewable fuel sources and their harmful effects on our environment have driven the development of energy efficient and environmentally friendly systems However without a suitable replacement for non renewable fuel sources energy efficiency only prolongs the inevitable end of the earth s supply of coal oil natural gas and nuclear fuels For this reason we were motivated to make our system rely on renewable energy sources To further reduce our impact on the environment and the economic costs associated with system maintenance and replacement we strived to create a product with a theoretically infinite lifespan Unfortunately many factors play into the actual longevity of any design Every component of any design has a practical lifespan T
17. in a row void init Timer void void init ADC 12 void void init Ports void void write to mem void void erase mem int address void Tx void int read memory int address unsigned int g adc index 0 unsigned volatile int g adc complete 0 unsigned int g slowdown loopl1 0 unsigned int g slowdown loop2 0 unsigned int g ADC samples MAX keep the past 100 samples from the ADC so we can process them unsigned int g emergency 0 0 off 1 on Used to send the buffer if there is a threshold met int g start addr int 0x001800 global start address PXDIR O0 is input 1 is output PXSEL O is I O 1 is peripheral module other PXOUT gt 0 is low and 1 is high PXIN 0 is low and 1 is high Read Only void main void THIS IS THE MAIN FOR Tx WDTCTL WDTPW WDTHOLD Stop watchdog timer inil Ports init Timer TLE ADG L23 erase mem int 0x1800 erase mem int 0x1880 erase mem int 0x1900 erase mem int 0x1980 unsigned int 1 for i MAX i gt 0 i g ADC samples i 0 initialization of buffer I 70 enable interrupts while 1 if g_emergency 1 disable interrupts write to mem 3560 g emergency 0 reset emergency flag g adc index 0 reset adc index enable interrupts if g_adc_complete disable interrupts write to mem enable interrupts j end while loop end main void i
18. putting it back into read only mode The following pseudocode shows the process of writing to the flash memory for ADC sample size copy values into the DataBuffer Check to see if the flash controller is busy For the Size or the Databurrer 4 unlock the memory store a DataBuffer value into an integer store the next DataBuffer value into an integer write both values to the memory location increment the memory pointer wait until all the data is written re lock the memory clear the DataBuffer When all available memory is full the write memory function trips a transmission flag TxFlag which starts the transmission process of all the data in the flash memory TxFlag is tripped when the current 42 address pointer reaches or passes the memory address Ox 19DE which is 10 memory locations away from the last information memory location We chose that address value as to give a buffer zone where our information would not over write the next section of memory which could contain important code data The following figure shows one buffer of ADC samples stored into the information flash memory exeolsoo x 183A x 01874 x 1SAE x 18E8 x 01922 x 195C x 01996 exeolope 1D 01D3 01D3 1D 01D1 1CB 1CE 1CE 1CE 1CF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFF
19. results of stationary connection of the XBee However most of this extra distance can be attributed to the speed of the car causing propogation delay of the transmitted signal In addition to this delay the delay XCTU takes to collect the data and visually display it could also account for most of this delay We were successful in connecting ranging from speeds of 10 25 mph We were unable to collect data at additional speeds due to time constraints and safety considerations However this data supports further investigation into collecting data from the XBee at highway speeds If this is indeed successful data can be collected as a car moves by the device and there will be no need to stop to collect data or implement a permanent structure to relay collect this data This is a huge potential of this product 4 5 Summary of Results Overall the results of our testing were generally inconclusive mostly as a result of insufficient testing The nature of our system that is being meant to operate in the middle of a busy highway did not lend itself to field testing so we were forced to conduct our tests on a smaller scale Additionally our project budget limited our ability to purchase testing equipment We believe that better results could be achieved using better equipment in a more controlled setting 63 5 CONCLUSION 5 1 Summary Our project started as an idea to create a sensor application to eliminate human interaction in dangero
20. to compensate for this and one such feature is a secondary battery In this situation the battery would perform several functions the most important of which would be to store excess harvested power and distribute it to the system when the primary generation is low The battery would additionally behave as a signal buffer between the wind turbine and the rest of the system Because the turbine generates electricity by moving a magnet in the presence of a coil the power it generates has an AC component to it unlike the solar panel which outputs a constant DC source a large battery would act as a buffer that would protect the sensitive circuitry of the microelectronics and supply a DC voltage While they come in many different varieties almost all batteries can be considered primary or secondary Primary batteries are most often used as the main source of power in a system and are not rechargeable Secondary batteries can be recharged and usually serve as back up sources of power but can also be used as the main power source in some applications Batteries has many more characteristics beyond their ability to be recharged We were specifically interested in operating temperatures charge tolerances and self discharge ratings while considering secondary batteries for our application because we knew that the system would be exposed to a wide range of temperatures inconsistent charging voltages and infrequently tens of hours without a decent charg
21. within the 2 412GHz to 2 472GHz with 13 channels each 22MHz wide This causes some overlapping and could be an issue if multiple transmitting XBees were within range Given the location of the XBees on the highway and the small radius of WLAN transmission there will be no interference caused by multiple XBees unless our wireless sensor platforms are within 100s of feet of each other that would be a lot of these products on a highway Perhaps the most important way to verify the XBee was working and prove that our code worked was seeing the data collected by the receiver XBee Digi International the makers of the XBee designed a 25 software interface for all of the XBee products called XCTU XBee Configuration and Testing Utility 15 Figure 2 8 shows the mainframe of XCTU XCTU TOR 6 B Radio Modules 000000409D5EB5EE EU x A A Function XBEE WI FI wants Port COMS 9600 N 1 N AT gt MAC 000000409D5EB5EE O eh leg DO B X A Send packets Q Send a single packet Pd Name Data Send selected packet an y Send sequence A iti 500 Transmit interval ms a 9 Repeattimes 1 g W Loop infinitely i B m E Figure 2 8 XCTU User Interface XCTU allows programming of the XBees and visualizes the data sent and received visualized data is in hex 16 In this way we would observe the memory bank on the microcontroller using CCS see the analog equivalent
22. 0 961C 011C COAS 0110 96A0 0112 6422 0110 9602 0112 431C 0114 2278 0194 OFDB 0158 0000 0182 0003 1AC2 0120 FFFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF Figure 3 13 Evidence that All Four Memory Banks Have Been Erased n In order to erase a section of memory the BUSY bit must be polled until the flash controller resets the bit The memory needs to be unlocked giving us the ability to modify the memory Once unlocked any 43 memory address within the section needs to be pointed at and written too using a dummy write The section of memory is reset to all 1 s and the memory can be relocked The following pseudocode shows our process Erase Memory poll on BUSY bit unlock the memory set controller to segment erase mode dummy write to memory location in segment relock the memory 3 2 1 4 UART Mode As previously mentioned in 2 2 4 3 Universal Serial Communication Interface UART Mode the transmission of the data from the memory will be sent via UART The Universal Serial Communication Interface USCI registers were configured to give us the following parameters UART mode LSB first 8 bits of data no parity and asynchronous mode Explanations for these parameters can be found in section 2 2 4 2 After the registers are set the ports for the Tx pin are set to use P3 3 The UCSWRST control bit must be cleared before transmission can begin and is easily
23. 1980 Bank B start 00197F end 001900 Bank C start 0018FF end 001880 Bank D start 00187F end 001800 g start addr value g start addr value2 while WAIT 0 wait until we can write the next bit 72 if g start addr int Ox19DE if it s the end of the info segment TRp leg sls butter is full transmit i kend of for loop while FCTL3 amp 0x0001 BUSY loop while BUSY j GEEN FWPW WRT BLKWRT FCTL3 FWPW LOCK reenable lock a duoc complete 0 enable interrupts for 1 MAX 1 gt 0 i g ADC samples i 0 initialization of buffer if TxFlag OE void erase mem int address while FCTL3 amp BUSY wait for not BUSY int dummy_addr dummy addr address starting address for BANK A pg 22 F5529 data sheet start 0019FF end 001980 Bank B start 00197F end 001900 liBank C start OUISEF end 001880 liBank D start 00lI57F end 001800 FCTL1 FWPW ERASE segment erase for our Banks MERAS BETE FWPW LOCKA unlock to erase dummy_addr 0 xFFFF if dummy_addr gt int Ox19DE if it s the end of the info segment dummy_addr address set it back into the info segment FCTL1 FWPW ERASE MERAS FCTL3 FWPW LOCK re enable lock int read_memory int address unsigned int read_value read value address return read value void Tx void
24. 2 Individual bits can only be programmed from 1 to 0 p 342 Each memory location can only be written to four times before an erase cycle is required p 339 The smallest memory size that can be written is one bit p 340 The smallest memory size that can be erased is 128 bytes p 340 The write to memory function writemem first copies the collected ADC samples into a buffer using a simple FOR loop The buffer is used to prevent any shared data issues Once complete the function polls the BUSY bit until it is ready The BUSY bit is high when the memory is writing or erasing and is 4 automatically reset when the flash controller is ready to write or erase The flash controller also has a safeguard LOCK bit which needs to be unlocked in order to write to a memory location Once the flash controller is not busy and is unlocked writing to the memory is as simple are pointing to a memory location and storing a value to it In order to make writing to the memory more energy efficient we write two values at the same time using block write Block write is 4 times as fast as as single byte write because the programming voltage remains high We poll a WAIT bit which is automatically set high once the writing is complete We loop through all the values in the data buffer storing them in sequence in memory Once all data is stored the BUSY bit is once again polled until it is reset by the flash controller The memory is relocked
25. 5 5 Volt positive Voltage Regulator R3 R4 R5 1K Ohm 1 8 Watt 10 IC2 NE555 Timer Chip R6 330 Ohm 1 8 Watt 1096 PB1 PB2 NO Momentary Contact Push Buttons R7 100 Ohm 1 8 Watt 1096 LED1 Green LED Q1 2N2222 Or Similar NPN Transistor LED2 Yellow LED Q2 IRF540 Or Similar Power MOSFET RLY1 40 Amp SPDT Automotive Relay C1 0 33uF 35V 1096 D1 1N4001 or similar C2 O 1uF 35V 10 R1 R2 10K Multi Turn Trim Pots R8 R9 Optional 330 Ohm 1 2 W Resistors Figure 2 9 Schematic for a Battery Charge Controller We first tested this circuit on a breadboard then assembled it on a perforated copper through hole board We did not used a printed circuit board in case we needed to modify the circuit at a later point in the design process 30 3 IMPLEMENTATION 3 1 Hardware All of the system components are mounted in a rectangular frame of wood and metal The frame measures 20 x20 x48 which are within the constraints of SR 4 This specific frame like the rest of our system is a prototype which can serve as the basis for future work Inside the frame the wind turbine is connected to the shaft of the generator by two rubber O rings inside a brass shaft collar A metal nail protrudes from the center of the base of the turbine this fits into one end of the rubber ring The other end of the collar fits snugly over the top of the generator shaft A small ring of ball bearings sits between the casing of the generator and the sh
26. BC O1 BC 61 BD 01 BC O1 BD O1 BC O1 BC 01 BC 01 BC w Send packets QD 154 Sond a single packet Name Data Send sequence Transmit interval ms 500 Figure 3 19 XCTU Interface Data Visualization Console This data comes from a photodiode which stores 16 bit unsigned integer values This console log shows the data being transmitted and displayed in order At the top left of the hexagonal representations there is 01 and BD which together represent one value 01BD the binary value of 445 In the current conditions we have our photodiode the data will be between 350 and 475 If the data went over 500 it would trigger the emergency transmission we can trigger this with a flashlight from our phone It should be noted that based upon the chosen sensors the values achievable would differ and the threshold may have to be modified in the microcontroller code by changing the macro threshold 51 4 RESULTS The following sections will detail our results of testing our roadside sensor platform 4 1 Characteristics of the ALT 10 12P Photovoltaic modules must be connected to a resistive load in order to produce power and although they have a rated maximum power in the case of the ALT10 12P 10 watts that output can only be realized under particular conditions The size of the ideal load for generating the most power varies with the amount of light that is hitting the photovoltaic cells so it is not practical to connect the mo
27. D3 1CC 1CD 1CE 1D2 1CE 1CD 1CE 1CC 1CD 1CD 1CE 1CF 1CD 1CD 1CC 1CE 1CC 1CC 1CF 1CE 1CE 1CE 1D4 1D4 1D5 01D2 01D1 01D4 1D 01D5 01D4 01D5 01D8 1F2 1FC 0000 0000 0000 0000 FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF 0606 C735 2955 1217 0A08 Figure 3 11 Example of an Emergency Buffer Stored in Memory 3 2 1 3 Flash Memory To be able to use Flash Memory properly we had to create functions that could write and erase specified sections of the memory The MSP403F5529LM data sheet states that any given memory location in the flash memory has the following parameters The memory s default mode is read mode p 34
28. F FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF 0606 C735 2955 1217 0A08 Figure 3 12 One Buffer of ADC Samples Stored into the Flash Memory The erase memory function behaves very similarly to the write memory function In reality we are not actually erasing the memory but instead writing all 1 s in the memory locations effectively destroying the data As mentioned previously the smallest segment of memory that can be erased is 128 bytes Our available flash inform
29. Self Energy Sensors Research and Development of a Prototype Self Powered Sensor Platform for Remote Roadside Monitoring Applications SUBMITTED IN FULFILLMENT OF THE Major Qualifying Project DEGREE REQUIREMENT FOR WORCESTER POLYTECHNIC INSTITUTE BY Mark A Zayac Ryan J Thornhill Ryan D Welch Lifeng Lai Advisor ON March 27 2015 Abstract The Wind Reaper is a prototype for a highway based sensor platform that powers itself by collecting energy from passing cars via a vertical axis wind turbine and the sun via a photovoltaic module The system utilizes an ultra low power microcontroller to collect and process data from user configurable sensors The collected data is transmitted wirelessly at specified intervals throughout its daily operation The system is also capable of detecting abnormal readings and signaling their occurrence with an emergency transmission The Wind Reaper is intended to be the base for future research and development into applications for self energy sensors ii Executive Summary The Wind Reaper is a prototype for a highway based sensor platform that powers itself by collecting energy from passing cars via a vertical axis wind turbine and the sun via a photovoltaic module It was developed as a result of extensive research in the areas of energy harvesting techniques and remote sensing applications examples found in research include solar powered deep sea tsunami detectors wildfire d
30. ack to zero Our clock source was set to ACLK which operates at 32768Hz With the clock source and TimerAO counter set at the same value an interrupt was triggered once per second With the ability of Timer AO interrupting once per second we were able to tailor the rest of the timing sequence to sample every 6 minutes We wanted to have control over the frequency of sampling so we implemented nested IF statements with two macros that we could adjust As demonstrated in the pseudocode below we used the one second interrupt speed of TimerAO to increment an integer Once the integer was above the first macro 39 level it would increment another integer Only once the second integer was above the second set macro value the ADC would be tripped TimerA0 Interrupt increment valuel if valuel gt Macrol reset valuel increment value2 LE valued gt Macro2 reset value2 trigger the ADC to sample When the ADC interrupts it samples and converts the current value and stores it into a global buffer The ADC is initialized to format the data to 16 bit unsigned integers Once the data 1s stored into the buffer the ADC interrupt flags are reset and the buffer index is incremented Note that the number of tripped interrupt flags directly depends on how many sensors are being sampled In our case only one flag is being tripped The ADC gathers samples for a defined set of iterations that can be set using a macro value The value we chos
31. aft collar this serves to reduce friction A second nail protrudes from the top of the turbine this slides into a hole drilled into a steel block The block is suspended from a cross piece at the top of the frame by a pair of bolts The bolts allow the block to be lifted so that the turbine can be put into or removed from the frame Figure 3 1 Top View of Turbine Showing Lock Block and Adjustable Mountings 3l The current frame can be adjusted in two different ways These adjustment mechanisms are very important for the stability of the wind turbine if the rotational axis is not exactly vertical the turbine will wobble and lose efficiency The frame sits on four plastic feet that can be extended and retracted via screwing them out of or into the base Each foot can be extended by 1 25 inches which allows the frame to tolerate inclinations up to 4 764 Additionally the metal block supporting the top of the turbine can be slid in two dimensions within a 4 inch slot in the cross piece The cross piece itself can slide in a 3 inch slot in directions perpendicular to the metal block All told the top of the turbine can be adjusted 2 inches north south and x1 5 inches east west in a horizontal plane this allows the turbine to tolerate inclines up to 5 389 The feet and crosspiece combined give the frame the ability to tolerate inclines up to 10 153 The PV module is mounted on a second cross piece directly above the wind turbine Th
32. ation memory is allocated into four 128 bytes sections The following figure shows all four information banks successfully erased in red Oxl6de 0x1800 0x122 Memory Rendering 43 gt 3 16 Bit Hex TI Style Y x 16DE ex001718 0x001752 0x00178C 0x0017C6 0x001500 x 153A 0x001874 Ox0013AE exeelsEs8 x 01922 x 195C 0x001996 0x0019D0 OXOOIAOA 0x001A44 x 1A7E x 1ABS OxBBlAF2 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF 3FFF d FFFFFFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF
33. atural wind This leads us to state that our data collected cannot be conclusive in one way or another but instead supports further investigation The collected data is displayed below in Table 2 1 Table 4 1 Data Regarding Road Simulation Testing NN 15 15 15 25 25 25 35 35 61 It is clear it was challenging to get wind speed given the conditions and given the equipment we had However most of the 25 45 mph drives consistently left a noticeable wake in person which was not necessarily reflected in the collected data Given this result it is safe to say that a car driving at slow speeds does not affect wind very much We can also state that a car driving at 25 45 mph leaves a wake We cannot support or deny that a car driving closer to highway speeds will leave a wake sufficient to turn our turbine nor can we state that multiple cars driving will increase or decrease this wind speed Lastly we were unable to collect data on whether wind coming from two cars driving in opposite directions added constructively or de constructively add to spin a wind turbine Further research should be made to make conclusions on these statements 4 3 XBee Transmission Range Stationary Our next Results revolved around the XBee and how far it can transmit its information We sat in a car and backed up in 50 feet increments until we were unable to receive data and then until we were not connected After 100 feet we were unable to collect data
34. axis spins an electrical generator which is typically positioned immediately behind the blades Unfortunately horizontal axis wind turbines HAWTs require a good deal of clearance to prevent foreign objects from colliding with the rotating blades Because of this a roadside system would not be able to use a HAWT on account of the risks it would pose to passing vehicles which violates SR 4 For many years people have been concocting alternative configurations for wind turbines ranging from niche to wildly impractical A particularly common theme among these designs in the idea that a turbine could be powered by the wind generated by passing cars Because of the matters of space and safety 12 previously mentioned these hypothetical designs frequently adopt a vertical axis configuration The most significant advantage that vertical axis wind turbines VAWTs hold over HAWTs is that they do not need to be aimed into the wind which 1s useful when the wind is inconsistent However VAWTS are less efficient than their horizontal counterparts because they generate more turbulence as can be seen by the models in Figure 2 3 below Figure 2 3 Turbulence Models of Modern Horizontal and Giromill Wind Turbines When placed in a single stream of wind VAWTs lose some efficiency because half of their structure is always rotating into the wind However we hypothesized that this particular disadvantage could be overcome by installing the turbine
35. ch a configuration is shown in Figure 2 5 14 Figure 2 5 Darrieus Wind Turbines With Savonius Type Start Up Rotors We expect that traffic in a typical highway environment will be variable throughout the day At some times traffic will be light and the resulting wind will be minimal so we decided to use a Savonius turbine in our system because of its low startup speed After further research we discovered that the efficiency of a Savonius turbine can be increased by offsetting the blades as seen in Figure 2 6 This allows the air to flow through the structure more easily which reduces drag In an interview with Dr Maria Chierichetti of Worcester Polytechnic Institute we learned that the ideal size of this offset labeled as e in the figure below is 20 of the overall diameter D of the turbine 15 Direction of rotation Figure 2 6 Effects of Offset Blades in a Savonius Rotor In general the power available in wind has a cubic relationship with the speed of the wind The following equation describes the available power in watts in the wind Power E A W Where p is the density of air typically 1 2Kg m A is the cross sectional area of the turbine in meters and W is the speed of the wind in m s Unfortunately it is theoretically impossible to fully utilize the available power in wind In 1919 a German physicist by the name of Albert Betz published findings and calculations regarding the maximum effici
36. cleared by setting it to zero Transmission of data over the Tx pin is surprisingly easy All that is required is to put 8 bits of data in UCAOTXBUF The data is automatically transmitted at the pre defined baud rate 9600 The busy bit UCBUSY is high while a transmission is in progress and needs to be polled until UCAOTXBUF is ready We used a FOR loop to read through all the data in our 512 bytes of memory For each loop iteration we transmit one integer value from the memory over the Tx pin Due to the chip architecture one integer is 16 bits so it must be split into two 8 bit buffers because UCAOTXBUF only takes 8 bits at a time We shift the integer to the right by 8 bits and store the value into UCAOTXBUF We poll the UCBUSY bit until all the data is transmitted out of the buffer Next we give UCAOTXBUF the remaining 8 bits of the integer value and poll the UCBUSY bit Once complete the FOR loop iterates again to send the next value from the memory When all values from the memory are sent all memory is cleared using the erase memory function The following figure shows the data being sent out on the Tx pin 44 Pos O 000s CH2 Coupling BW Limit BOMHz volts Div 4 it it a gg gt i o a LE s eoos us mu rr 10 o A Tea oo BEEN iim AL Figure 3 14 Oscilloscope Reading Showing Transmission of Data on the TX Pin Additionally the Tx function is able to repeatedly transmit the same data for a user defined amou
37. despite still being connected to the other XBee at 125 feet we had a 14 signal strength from the XBee and were unable to collect data At 150 feet the signal strength dropped to 696 and after that distance we were not within range of the other XBee These results suggest any stationary object on the side of the road would be able to collect the data being transmitted from the turbine in the middle of the median Even stopping on the side of the road to collect the data would be fine Additionally if there was a relay that was able to collect the data and then transmit it over a much larger distance no one would have to stop and collect data This data also acted as a baseline for our next test which focused on driving by and collecting data something we aspired to do at the beginning of this project 4 4 XBee Transmission Range Drive By Another experiment we conducted was to see if the XBees could connect to each other and transmit data while traveling in a car as we passed the device We passed by at increasing speed and measured the distance from the device when we connected and when we stopped receiving data In this scenario data 62 was constantly transmitting so we were able to verify when a connection that would transmit was established The information we collected shows that we can receive transmitted data at or just past the device and continue receiving data until roughly 250 feet past the device This was surprising given the
38. dges age maintenance costs increase However if maintenance is deferred as it often is according to the 2013 Report Card structural deficiencies worsen and costs are driven up Simply put many bridges in the United States were not regularly serviced and as a result small problems that could have been repaired quickly and inexpensively when they first arose have now developed into serious structural issues that will take a good deal of time and money to repair In the conclusion of the 2013 Report Card the ASCE proposes solutions that can be immediately implemented one such solution is prioritizing bridge repairs based on a risk prioritization model Additionally one long term solution proposed by the report is to develop a plan to build more resilient bridges and make repairs to existing bridges more effective and longer lasting In both of these solutions holistic bridge health monitoring will play an important role 9 Evaluating the structural health of a bridge in person is a large and risky undertaking especially when structural elements are suspended above or below the roadway Even if automatic sensors are used instead of humans such sensors must also be maintained if they are to provide accurate readings To this end we considered the possibility of designing self sufficient sensors that could be used to monitor bridge health A sensor array that harvests energy from its environment would significantly reduce maintenance r
39. dule to a static load and call it good Instead the optimal way to achieve maximum power is with a device called a Maximum Power Point Tracker MPPT MPPTs monitor the conditions around the PV module and adjust a dynamic load so that it is always putting out maximum power Unfortunately commercial MPPTs are expensive and their mechanics are so complicated that building one for ourselves would be highly impractical given our time and monetary constraints Instead we implemented a potentiometer between the battery terminal and the power socket receiver for the PV module this serves as an adjustable load that we can set to approximate an ideal maximum power load We determined this value by conducting a series of test on the PV module Our first test was performed indoors so that the module would be exposed to a constant amount of light The module was connected to a resistor and the voltage across the load was measured This test was repeated for a number of different resistor values The actual resistance of each component was measured during the test to maintain accuracy in our calculations 52 Figure 4 1 Indoor Test Environment for the PV Module After a suitable number of iterations we plotted our data to create characteristic curves that describe the physical traits of the ALT 10 12P 53 Characteristics Relative to Resistance S Power E Voltage Power mW 30 40 50 60 70 B x 100 Resistance k ohms Figu
40. e The table below compares the many characteristics of the most common types of secondary batteries 27 Table 2 2 Comparison of Battery Characteristics by Chemical Composition Specifications Lead Acid NiCd Li ion Cobalt Manganese Phosphate Specific energy 30 50 45 80 60 120 150 190 100 135 90 120 density Wh kg Internal resistance lt 100 100 200 200 300 150 300 25 757 25 502 ma 12V pack 6V pack 6V pack 7 2V per cell per cell Cycle life 200 300 1000 300 500 500 500 1 000 1 000 80 discharge 1 000 2 000 Fast charge time 8 16h 1h typical 1h or less Overcharge High Moderate Low Low Cannot tolerate trickle charge tolerance Self discharge 2096 30965 10955 month roomtemp Cell voltage 2V 1 2V7 1 2V7 3 6V 3 8V 3 3V nominal Charge cutoff 2 40 Full charge detection 4 20 3 60 voltage V cell Float 2 25 by voltage signature Discharge cutoff 1 75 1 00 2 50 3 00 2 90 voltage V cell 1C Peak load current 5C 20C 5C gt 3C gt 30C gt 30C Best result 0 2C 1C 0 5C 1C lt 10C lt 10C Charge temperature 20 to 50 C 0 to 45 C 0 to 45 C Discharge temperature 20 to 50 C 20 to 65 C 20 to 60 C Maintenance 3 6 months 30 60 days 60 90 days Not required requirement topping chg discharge discharge Safety requirements Thermally Thermally stable fuse Protection circuit mandatory stable protection common In use since Late 1800s 1950 1990 1991 1996 1999 Ultima
41. e for the system is ten values which over 24 hours between transmission times will fill our available allocated memory of 240 values Once ten samples are stored a completion flag is set the ADC stops sampling and the index is set back to zero The following pseudocode illustrates the process including the emergency transmission explained after the pseudocode ADC Interrupii ADCbuffer index current ADC sample increment index reset all interrupt flags 1f index gt Macrol ADC is complete 40 reset index value An added function of our system is to immediately transmit data that is of high importance as opposed to waiting to transmit the data at the end of the 24 hour period The emergency sequence starts in the ADC interrupt where all incoming samples are compared to a threshold value If the current sample is above the threshold value user defined macro an emergency flag is tripped Once the flag is tripped transmission of all data in the memory starts instead of waiting for the memory to be completely full An example of an emergency buffer stored in the information memory can be seen below in Figure 3 11 0x001300 x 183A x 01874 x 18AE x 15E8 x 01922 x 6195C x 01996 0x0019D0 1D 1D3 1D3 1D 1D1 1CB 1CE 1CE 1CE 1CF 1D 1CE 1CF 1CD 1CF 1CC 1CB 1CE 1CB 1CD 1CD 1CD 1C6 1CC 1CE 1CD 1CE 1CD 1CD 1CE 1CC 1CF 1CD 1CD 1CE 1CD 1CF 1CE 1CE 1CD 01
42. e module is secured to a small wooden frame that screws onto the swivel mount of a recycled camera tripod this gives the PV module 360 of rotational freedom in the horizontal plane and 60 degrees of rotational freedom in the vertical plane Al ir ni im te A j 7 Aer 7777 MATH Hinia WT Mim AS iii EEEN HHM af AE IN F HIH H o o ESF E SESE ELIE T riisi IH EPEN f f ON shits TS nd gt 1 MM oH AE TOL HIRE 4 E E e MIHI as ew n warden de v mos T 079 9 ma dt Oe BP 4 etm Figure 3 2 Adjustable Feet of the Wind Turbine Frame 32 The charge controller connects to the terminals of the battery via metal tab slots that are screwed into the board It is through these connections that the charge controller draws power The circuit uses a 5V regulator to supply the appropriate voltage to the 555 chip and we originally thought that we would use the same regulator to provide 5V to the MSP430 however we decided to install a second regulator so as to not interfere with the signals within the charge controller Both regulators are connected to a common heat sink Additionally to conserve space the MSP430 and XBee are mounted on the board In the case of the MSP430 two 2x10 rows of header pins were installed on the board so that the microcontroller can be removed for programming and independent testing The XBee connects to a break out board that is soldered on the main board and ca
43. elected As it suggests DHCP will configure all these fields when we connect to the other XBee Before connection these fields will likely be 0 0 0 0 or blank 16 Whenever the XBee has filled its memory based upon our program every 24 hours it will transmit the collected information If the other XBee is connected during that time it will receive the transmitted data An example of the visualization of the transmitted data is shown in XCTU under the Consoles Tab 50 E a Ce Gel 00000040905 B5E Console log De DADADADADA DADA DADA DAD 01 BD O1 BD O1 BC O1 BB O1 BD 01 BF 01 BD O1 BC O1 BD O1 BD O1 BC O1 A DADADADADADADADADADADAD BD 61 BC O1 BD 61 BD 61 BC O1 BD O1 BC O1 BD 61 BD O1 BC O1 BD 61 BC Dr DABADADADADADADADADAD 01 BD 01 BC O1 BC O1 BC O1 BC 01 BC 01 BB O1 BD O1 BC 01 BC 01 BD O1 2400240924002 MIE BC 61 BC O1 BD O1 BC O1 BD O1 BC O1 BC O1 BD O1 BC 01 BC O1 BC O1 BC 0101 01740 7 014012401240124012 4000401 o1 BF O1 BC O1 BE O1 BC O1 BC O1 BD O1 BC O1 BD O1 BC O1 BD O1 BD O1 2400240024 00249240024 0020024002400 0040 BC 01 BD 01 BC 01 BC O1 BD O1 BC 01 BD O1 BC O1 BC O1 BC 01 BD 01 BD DADADADADADADADADADADAD 01 BD 01 BD O1 BC O1 BC O1 BD 01 BE O1 BD O1 BC O1 BD O1 BC 61 BD O1 DADADADADADADA DADA DADAS BC O1 BD O1 BC O1 BD 61 BC O1 BC O1 BD 61 BC 61 CO 61 BD O1 BD O1 BC DADADADADADADADADADADAD 01 BD 01 BE 01 BD 01 BD 01 BC 01 BD O1 BC O1 BD O1 BB 01 BE 01 BC O1 DADADADADADADADADADADAD v BC 01 BC 01 BC 01
44. electing the Added Device will load all previous parameters of the XBee There are approximately 95 different parameters for our S6B but most of these parameters will be set to their default settings Our XBees used firmware version 2021 but newer versions are now available Each firmware version may change functionality of the XBees slightly and further investigation into these versions are needed 16 48 ea x Beo J HF Radio Configuration 000000409D5EBSEE T SIA lee AF v MAC PHY Change MAC PHY Settings B Radio Modules Name Function XBEE WI FI Port COMS 9600 8 N 1 N AT MAC 000000409D5EB5EE X 9 XCTU Al Association Indication DI Device Cloud Indicator CH Channel LM Link Margin D PL Power Level v Network Change Network Settings 1 AH Network type CE Infrastructure Mode ID SSID EE Encryption Enable PK Passphrase O IP IP Protocol MA IP Addressing Mode 1 TM TCP Client Connection Timeout TS TCP Server Connection Timeout i DO Device Options 0 4 6 FF Highest 4 Infrastructure 2 STA mode 2 xbee 00409D5DB7CA No security 0 UDP 0 DHCP 0 64 100 ms 258 100 ms Figure 3 17 XCTU Interface Configurable Parameters a Parameter A 06606006006 e 60606006 Y Y Nas ee Jea DO IC eo ev 0 Us cl RARE a In Figure 3 17 the blue highlighted fields are the ones we
45. ency of an ideal wind turbine which has come to be known as the Betz Law The law is derived from the principles of conservation of mass and momentum of the air stream and states that no turbine can capture more than 16 27 59 396 of the kinetic energy in wind The factor 16 27 0 593 is known as Betz s coefficient Practical utility scale wind turbines achieve at peak 7596 to 8096 of the Betz limit This is largely due to the inherent drag and friction forces that exist in the turbine equipment 16 Once again the size of the turbine is limited by SR 4 Dr Chierichetti also revealed to us that VAWTs have an ideal diameter to height ratio of 1 2 With this in mind we constructed a prototype turbine measuring 1 in total diameter by 2 tall this is smaller than the maximum area in SR 4 to allow for a supporting frame The turbine follows the previously discussed Savonius design with two offset blades The blades were repurposed aluminium stovepipe pieces which we used because they were inexpensive and already curved the latter was particularly important during later attempts to build a larger turbine we quickly discovered that it is very difficult to apply a uniform curve to large sheets of metal without the proper equipment Aluminium is also a lightweight metal that is easy to manipulate The blades were mounted to a rectangular piece of acrylic at both their top and the bottom to maintain structural integrity When we tried to calcu
46. equirements by eliminating the need for battery replacement In the same 2013 Report Card American roads received a grade of D poor 9 After considering the possibility of bridge health sensors that harvest energy we realized that the same idea could easily be applied to all roadside sensing applications in general 1 24 Criteria After extensive research we recognized our project needs to satisfy three criteria e Based on our Tsunami research using sustainable energy will increase the lifespan of the battery and therefore decrease the frequency of maintenance routines e Based on our Wildfire Application research the product should be powered in an environmentally friendly fashion e Based on our Bridge Monitoring research the sensors should be self powered because their location is inaccessible or dangerous These three criteria helped determine our final project which focused on highway applications 1 3 Our Project All of these applications were great stepping stones paving the way for our initial design Our project is a wireless self sufficient sensor platform that is located on the side of the road Our product will be located on the median of well traveled highways where wind created by high speed traffic will provide self sufficient energy via a wind turbine Sensors monitoring road conditions and atmospheric gases will be powered by a combination of the turbine and a solar panel The sensors will collect readings from
47. etectors powered parasitically through the metabolism of trees and roadside sensors The results of this research lead to a set of project requirements that govern the development of the prototype system The system is intended to operate in the median of a busy highway Because of this the primary power sources of the system are a photovoltaic module and a Savonius type vertical axis wind turbine The particular model of turbine was chosen because it was hypothesized that a Savonius rotor would be able to efficiently utilize two opposing streams of wind that would result from traffic passing in both directions Both primary sources charge a secondary battery that will provide a constant power supply when one or both primary supplies are inactive The battery implements a charge controller with electromagnetic relay that prevents overcharge Sensor data is collected by an ultra low power microcontroller that is powered by the battery The system can interface with any sensor within the design specifications a simple photodiode was used for testing Readings from the sensor are taken periodically throughout the daily operation of the system Data is stored in a local buffer and transmitted via a wireless module to an off site receiver The system is also capable of detecting unusual measurements from the sensor this ability is intended to notice sudden changes in the roadside environment In this event the system immediately transmits the buffer as
48. eveloped and sold by Texas Instruments for their embedded microcontrollers It enables editing of source code automatic building of the code debugging of the code and an interpreter making it an integrated development environment IDE Using CCS gives us complete control to design and build our code for the MSP430F5529 and any other TI products All of our code for the MSP430F5529LM was designed and written in CCS The following sections will detail and explain our code and how it works 3 2 1 1 Software Overview The overall function of our software was to determine when to sample gather ADC samples store those samples locally and then transmit the data over UART to the XBee By using real time embedded techniques we created functions that satisfied the requirements above The following state diagram Figure 3 10 shows the overall function of our code 38 Put Sample ADC12SC in Buffer Figure 3 10 State Diagram for Microcontroller Logic 3 2 1 2 Analog to Digital Converter The operation of our ADC was split into two main components ADC timing and ADC sampling In our system the ADC needed to sample at regular intervals throughout the day and utilize as much memory space as possible available for the samples We initialized our ADC to operate with TimerAO Timer AO was set in count up mode with a max value of 32768 Every time Timer A0 reaches the max value an interrupt is triggered and the timer count is set b
49. find Voltree a company that has patented a way to harvest energy from trees Voltree creates power from trees by using the PH difference from the tree s trunk and the soil shown in Figure 1 3 7 Figure 1 3 Voltree Bioenergy Harvester The power generated is very small but provides enough power to run their fire detection product named the Javelin The unit can sense many different data points including humidity temperature wind speed direction precipitation 360 visual fire alert and more In addition the Javelin sends the data wirelessly to an offsite location 8 We learned from Voltree that having an environmentally friendly and sustainable energy source was something that we wanted to include into our project However the Javelin would have been major competition if we decided to enter this market For this reason we decided not to pursue a product in this area 1 2 3 Bridge Monitoring In their 2013 Report Card for America s Infrastructure the American Society of Civil Engineers ASCE issued a grade of C mediocre to the category of Bridges This is largely due to the fact that over 67 000 of the nation s 607 380 bridges are considered structurally deficient Furthermore the report states that the Federal Highway Administration FHWA estimates that spending on bridges must increase by 8 billion annually to eliminate the nation s bridge deficient backlog by 2028 9 Clearly bridges are expensive installations As bri
50. fter defining these requirements we look at each individual component of the system and use these requirements to rationalize our decision for each component We also elaborate on the specifications of these chosen components After these components are explained we describe how all the components are combined and applied in the final prototype We split this section into hardware and software to separately illustrate the physical prototype and the data processing coding within 2 1 System Requirements As discussed in Section 1 3 Our Project we designed and constructed an energy harvesting sensor platform that collects and transmits data from a roadside environment The methods by which the data is collected can vary by application Before developing our prototype we established a set of requirements and constraints for the design based on intended operating conditions The system requirements what the product needs to do and achieve set forth in the following sections are also influenced by the three criteria developed in Section 1 Motivation The System Requirements SR are as follows SR 1 The system will be located on the highway next to high speed traffic as to allow the sensors easy access to the environment SR 2 The system itself will be a platform which will interface with third party sensors that perform tasks as required by the user SR 3 The system will regularly process data from its sensors and periodically transmit the
51. hese cells are made of materials like crystalline silicon that produce electricity when exposed to photons There are two kinds of crystalline silicon that are used in the production of solar cells polycrystalline p Si which results in bright blue rectangular cells and monocrystalline m S1 which results in dark blue octagonal cells 10 4 o 4 4 4L el BUE el IRE A De e gt Figure 2 1 Monocrystalline left and Polycrystalline right PV Modules PV modules with p Si cells are less efficient in terms of energy produced per unit area than modules with m Si cells but are also less expensive and more tolerant of low temperatures Furthermore PV modules with m Si solar cells can face drastic drops in power output if they are partially covered It was for this reason in particular that we elected to use a polycrystalline solar panel in our design A solar panel in a roadside environment could possibly be subjected to snow which would render an m Si panel inoperable until its surface was fully cleared whereas a p Si panel would continue to provide some amount of power so long as part of its surface was exposed PV modules are available in almost any size and their power output is related to their surface area so bigger modules produce more power than smaller modules Note that this is not an exact linear relationship but instead a general trend because there are many other characteristics
52. if a better design exists In addition to a better design more research should be done on constructive and destructive wind flow When the turbine is in the middle of a highway on the Jersey barrier traffic flow passes the turbine on both sides A vertical wind turbine might be able to benefit from two different opposing wind sources 64 Another suggestion would be to create a larger turbine of the current design We built a 1 5 scale model of the turbine but ran out of time to build a housing structure for it We also believe that it might be too heavy so reconstructing it with lighter materials could be a viable option Our power calculations indicate a bigger turbine would create more power but keeping the startup speed low and the weight of the blades small might pose a challenge The current generator for the turbine is a DC motor that was found in a cardboard box in a storeroom with the year 1977 stamped on it In addition we never found a data sheet or any documentation on the motor Needless to say a generator that is tailored to the system could greatly improve power generation Note that using gears to gear up or gear down the generator does not improve the performance of the design We tried adding gears to the turbine to turn the generator at a faster speed but the result was making the startup speed of the turbine higher Keeping a low start up turbine speed is essential to the project so leave gears out of the equa
53. ill repeatedly fill and convert the ADC buffer until a stop bit is enabled We decided not to use this mode because we felt we had more control over the ADC in Single Channel mode by telling it exactly when we wanted an ADC buffer of new data 13 A flow diagram of Single Channel mode is shown below CONSEQx 00 ADC120N 1 Ty ADCI2ENC 4 x CSTARTADDx Wait for Enable SHSx 0 and ADC12ENC 1 ord and A ADC12SC 4 i ADC12ENC 4 Wait for Trigger x SAMPCON 4 ADC12ENC 0 E SAMPCON 1 S Sample Input Channel Defined in ADC12ENC 0 ADC12MCTLx see Note A __SAMPCON Y N NR x ADC12CLK ADC12ENC 0 Note A m NT x ADC12CLK Conversion l Completed Result Stored Into ADC12MEMx ADC12IFG x is Set x pointer to ADC12MCTLx ne gt Figure 2 7 Flow Diagram Showing Single Channel Mode In both the Repeat Single Channel and Repeat Sequence of Channels modes the other two modes mentioned above multiple inputs into the ADC can be sampled and converted It should be noted that this 20 functionality of the ADC could be used for future improvements to the system design to accommodate for multiple sensor inputs 13 2 2 4 2 Memory The MSP430F5529LP flash memory is non volatile and partitioned into different size segments The code memory segments are 32kB while the information memory segments are 128B Initially we
54. information wirelessly to a user specific receiver The system will also immediately transmit if it detects values beyond a given threshold emergency transmission SR 4 The system will not create an unsafe driving environment This limits the size of the entire system to an area approximately 2 feet wide by 2 feet long by 5 feet tall SR 5 The system will in all respects by environmentally friendly SR 6 The budget for prototype development and testing is 375 SR 7 The system will collect enough energy from its environment to power a data processor a wireless transmitter and an array of sensors SR 8 The system will be powered by a sustainable source of energy in order to minimize the costs associated with regular maintenance SR 9 The system will store excess energy which will serve as a secondary source of power 2 2 Components The following sections will detail all the components of the system We will describe the components and give reasons to why we chose them 2 2 1 Solar Panel Typical highways are uncovered and are exposed to the open sky therefor solar power is a readily available source of energy Powering our system with a solar panel satisfies SR 5 and SR 8 While the term Solar Panel is a general term that describes a number of panel like objects that can be used to capture energy from the sun most people specifically think of photovoltaic PV modules which is a package of several connected solar cells T
55. is useless to develop a long term self sufficient power source because the battery can be replaced during biennial maintenance As a result we decided to forgo the idea of designing a tsunami warning system 1 2 2 Wildfire Detection Wildfire detection is currently on the rise in the United States due to the increase of intensity and frequency of fires in recent years In 2007 more than 2 5 billion of damage was done by wildfires Modern wildfire technology is quickly advancing to implement ground sensors that can remotely detect smoke and heat signatures Surprisingly the current most viable method is a watch tower where a trained worker keeps 4 an eye out for smoke and flame 5 An application of remote sensing can cover a vast area of forest without increasing the cost and maintenance of the sensors 6 pp 7 8 Therefore using sustainable energy would be an excellent idea for this type of application Unfortunately there is not a consistent power source that would be able to run a low power device Sunlight and wind would not be reliable energy sources due to the sensor being on the ground floor of the forest A piezoelectric energy harvester uses a flexible membrane which converts mechanical vibrations into electrical power In this case the vibrations would come from the movement of the tree which would only move during very windy times and thus were unreliable energy producers Trying to find a reliable power source led us to
56. l it is set to O by the CPU not user controlled Once the BUSY bit is 0 the flash controller needs to be configured properly and set into erase mode For the actual erase step we are really writing all 1 s to the memory location As mentioned before all 1 s in a memory segment is a clean slate that can be written to So erasing is actually writing over the old data Finally the LOCK bit must be re enabled 13 The steps can be seen below Disable watchdog Setup flash controller and set WRT 1 Write byte or word Set WRT 0 LOCK 1 reenable watchdog Page 344 from the MSP430x5xx Family Users Guide Data can only be written to a bank four times before the memory segment must be erased Failing to do so causes the memory segment to display all 0 s because the flash memory can only write 1 to O individually and requires an erase function to set them back to 1 The smallest size memory that can be erased is one segment which is 128B Due to this we opted to erase all four information memory segments at once once all the information from the memory is sent over the UART 13 2 2 4 3 Universal Serial Communication Interface UART Mode The Universal Serial Communication Interface USCT can be set in two different modes Universal Asynchronous Receiver and Transmitter mode UART and Serial Peripheral Interface mode SPI The main difference between UART and SPI is that UART is asynchronous and SPI is synchronous
57. lag Le secondFlag 0 j end if statement end while loop end for loop erase mem int erase the banks of memory erase mem int 0x1880 uu ue erase mem int erase mem int 0x1980 g start addr int 0x1800 reset the start address since we are going to erase P3OUT 0x80 Output to high to put the XBee to sleep P3DIR 0x00 set P3 7 to Input for the time being 74 Appendix B ALT10 12P Indoor Test Data IEIEIEIETE en en JE i 8 8 8 8 8815 8 0 162 0 222 Power mW enken al A son gt 2 o al ma em e P oi re Ti 8189 9332 SAS 8 olala eielesdGlaidicd i co cw co wn nf A eal on SERIE Current mA stage 75
58. late the power output of the wind turbine we ran into a problem Due to the fact that HAWTs outperform VAWTS in most applications Savonius turbines are not often the subject of scientific research and analysis The findings of studies that have been conducted are inconsistent with each other As a result we were unable to generate a comprehensive calculation to determine the efficiency of our turbine Instead estimated the efficiency based on published measurements of other similar machines Overall we found that Savonius turbines have an efficiency between 20 and 40 of the Betz Limit We assumed that our turbine being a rough prototype designed and assembled by a team with no experience in the field of aerodynamic machines would likely fall in the low end of the range We also decided that it would be better for us in the long run to perform calculations with pessimistic assumptions and therefore design our system around a smaller power budget in case we found that the actual output was higher thusly providing excess power to our design which would be handled in accordance with SR 9 2 2 3 Generator In order to actually utilize the power produced by the wind turbine we needed a generator to convert the mechanical energy of the turbine into electrical energy for our system We elected to use a DC generator because the PV module also produces DC power Even small DC generators are expensive starting at 100 so in accordance with SR 6 we
59. logging into a web application and viewing all transmitted data We suggest building a complete UI from the data transmitted to the receiving XBee module It was always the dream of our group to expand the current sensing module into a complete network of sensors that span along the highway Each module would stand alone how it currently is but would communicate to adjacent modules placed at regular distances down the highway An ultimate goal would be to have each module produce Wi Fi signals to give drivers the ability to connect to the internet while driving Constant Wi Fi will require a larger power consumption thus optimizing the turbine generator and solar panel are heavily suggested 66 Bibliography 1 2 3 4 5 6 7 Precision Drone Drones for Agricultural Crop Sureillance Black Frod Designs Online Available http www precisiondrone com agriculture Accessed November 2014 U S Energy Information Administration Primary Energy Consumption by Source and Sector 2014 Online Available http www ela gov totalenergy data monthly pdf flow css_2013_energy pdf Accessed January 2015 Commonwealth of Australia Bureau of Metherology Deep Ocean Tsunami Detection Buoys Commonweatlh of Australia 2015 Online Available http www bom gov au tsunami about detection_buoys shtml Accessed January 2015 Mid Atlantic Regional Association Costal Ocean Observing System Fe
60. n also be removed The microcontroller is powered by the regulated 5V supply and also outputs its own regulated 3 3V supply to power the XBee The battery and circuit board are installed in a trough several inches below the generator The walls of the trough are low enough that they do not block the heat sink Figure 3 3 Battery and Circuitry Mounted in Turbine Frame 33 Figure 3 4 Shaft Collar Connecting the Turbine and the Generator nt AR A x 4 Eri Aat t RS Le AME LN Tas M Ms r ba 4 Figure 3 5 Battery Terminal Connectors The PV module and wind turbine are connected to the charge controller through a pair of sockets The female sockets are fixed to the board and the power sources terminate in a male socket The connecting wire is a shielded 3 line 20 gauge cable this is particularly useful because it encloses both the positive and 34 negative wires and significantly reduces the chances of the wires getting tangled We chose to use this connection method instead of permanently attaching the supply lines to the board so that we could easily disconnect the board for troubleshooting The generator is close enough to the battery that the connector cable is only a few inches long but the connector coming from the PV module runs through a plastic tube that is secured to one of metal legs of the frame This tube assures that the connector will not interfere with the wind turbine during normal operation
61. nit Timer void TAOCCTLO CCIE enable interrupt TAOCCRO ind Blade set timer counter to 1 sec TAOCTL TACLR clear the counter reset the clock divider reset count direction TAOCTL TASSEL 1 MC 1 ID 0 select ACLK use upmode and go from 1 to 50 i void init ADC 12 void DISCLAIMER not sure how this works so will need to look in the user guide Page 742 748 ADC12CTLO amp ADC12ENC turn off to modify ADCI2CTLO ADC120N ADC12MSC On and Single sample convert ADC12CTL1 ADC12SHP ADCI2SSEL 1 ADCI2CONSEO 0 set the ADC clock source to be ACLK and divide clock by 8 ADC12CTL2 ADCI2RES 2 ADCI2PDIV set resolution to 12 bit resolution we can change this easily Also pre divide the clock source by 4 ADC12CTL2 amp ADC12DF Data format in memory register currently stored at Unsigned Int can be 2s Complement ADC12IE 0x0001 enable interrupt bit ADC12IFG amp 0x0000 flags that get tripped when there is a conversion result ADC12IV amp 0x0000 when the memory is used err is like has a result or whatever and then it sets the flag ADC12CTLO ADCI2ENC turn ADC12ENC back on so we can do conversions P6DIR 0x01 Enable as input P6SEL 0x01 Enable P6 0 for use by ADC void init Ports void PASEL Ox00 apparently initializing these reduces current consump ron PBSEL 0x00 PCSEL 0x00 PDSEL 0x003 pragma vector
62. nt of transmissions Occasionally the beginning of a transmission is lost because the XBee modules are still trying to connect with each other Other times it is possible that an entire transmission is not received as a result of interference In either case multiple transmissions increase the likelihood that the receiver gets a complete set of data This practice is called best effort The amount of iterations are easily set using a defined Macro that is used in a FOR loop The following pseudocode shows the Tx function Toc TIENDE LOL Configure USCI registers Clear UCSWRST bit Set P3 3 for the amount of times you send the same data Check TEBUSY Dams 45 Shift integer by 8 bits to the right Give shifted data to UCAOTXBUF Check UCBUSY bit Give remaining 8 bits of the integer to UCAOTXBUF Erase all memory Reset memory address pointer 3 2 2 XBee Wireless Module The other software aspect of this project is the Implementation of the XBees As mentioned in Section 2 2 5 XBee we used XCTU to visualize the data received Before receiving this data we first had to program the XBees slightly to connect correctly The XCTU interface is fairly straight forward When first opening the application you see the something similar to Figure 3 15 An open interface with no Radio Modules XBees connected 16 46 a g XCTU HR wear LEME de Radio Modules t Radio Configuration Click on f
63. o further realize our goal of designing a self sufficient sensor system we aimed to use components that will operate effectively for as long as possible and will be recyclable at the end of their life Overall we decided to pursue a design that would fill a gap in remote sensing technology and ultimately prove the feasibility of implementing renewable and environmentally friendly energy sources in daily applications 1 2 Beginning Ideas The following sections explain the first ideas we had for the project in chronological order Each idea helped lead to our final project idea and gave us insight on possible parameters for our project 1 2 1 Tsunami Detection One idea was to use wave energy to power sensors in tsunami detection systems Tsunami detectors are water pressure sensors that are sunk to the bottom of the sea or suspended at a certain depth If there is a certain pressure change in the water it may be a sign of a possible tsunami The pressure sensor sends a signal to an onshore location via an attached buoy platform As we began to research the idea it became clear that the sensors were not powered by alternative energy but instead with a battery However we soon discovered that tsunami detectors were serviced every 2 4 years to replace sensitive parts 3 Figure 1 2 Individual servicing Tsunami sensor buoys 4 Servicing the sensors is important to keep the accuracy and integrity of the sensor in working order It
64. on the UART pin using an oscilloscope and see the received data on XCTU providing complete diagnostic opportunities Additionally we were able to connect a laptop to the XBee transmitter and use a packet sniffer application called Wireshark to see the individual packets being received 2 2 6 Sensor The sensors in our product follow the System Requirements set forth from SR 2 and SR 3 Namely the sensors will be from the third party and simply need to meet the power requirements from the ADC which states the voltage needs to be no higher than 2 5V and no lower than OV with its analog readings The sensor data will be collected and sent to via the XBee to a computer off site Because of this what the 26 sensor data looks like and how a program will have to handle the data will not be available until the data has gone off site There can be many sensors in our network with minimal modification to the software For these reasons and the simplicity of our prototype we took the liberty of having just one sensor a photodiode with a voltage range from roughly 150mV to 425mV This sensor has a threshold emergency that trips at 305mV this threshold is modifiable Beyond these specifications there are no other requirements for the sensors 2 2 7 Battery and Charge Controller As has been made clear by this report our system will not be able to generate a constant and stable amount of power We have implemented many features in our design
65. on the median of the highway In such a position the turbine would have two sources of wind one from each direction of travel that would be opposing and theoretically complementary 13 savonius VAWT Modern HAWT Giromill Darrieus VAWT Figure 2 4 Visual Comparison of Vertical and Horizontal Axis Wind Turbines VAWTs follow one of two general designs which are named for their respective inventors Savonius turbines and Darrieus turbines Savonius turbines are characterized by having a number of scoop like blades and most modern turbines utilize two or three semi cylindrical blades Some turbines use a twisted Savonius design which is a variation of the original where the blades take on a helical twist Savonius turbines have a low startup speed meaning that they do not require high wind speed to start rotation and are very simple to construct The Darrieus model of VAWTs has curved vertical airfoils mounted around a central axis this look has earned Darrieus VAWTs the nickname of eggbeaters A giromill is a variation of the Darrieus turbine with straight instead of curved blades the second turbine in Figure 2 3 is a giromill type turbine Darrieus turbines are more efficient than Savonius turbines but are far more complex in their construction because they require airfoil blades Darrieus turbines also have a high startup speed and require either an external motor or smaller startup turbine to begin rotation an example of su
66. p pin The current is reduced to 2mA in associated sleep mode and 6uA in deep sleep In this way we can have the XBee off charge the battery and when enough charge is built up the XBee will turn on and transmit 14 The XBee is very small about 1 inch by 1 inch in size and can send data of 7 or 8 bits at a time with the Least Significant Bit first If our data is of greater size 7 or 8 bits we can modify the microcontroller to put different segments of the data one after another avoiding loss of information The XBee can operate in the main wireless network standards 802 11b g and n It can operate using a TCP or UDP protocol and has a data rate of up to 72MBps The complete data rates standards and their relative current consumptions are shown below in further detail in Table 2 1 14 24 Table 2 1 Electrical Specification of the XBee Wireless Module Electrical Specifications Specification XBee WI FI 3 14 3 46 VOC Supply Voltage 802 11b Operating Current transmit max output power MCS 0 6 5 7 22 Mbps MCS 1 13 14 44 Mbps MCS 2 19 5 2L67 Mbps MCS 3 26 28 89 Mbps MCS 4 33 43 33 Mbps MCS 5 52 57 78 Mbps MCS 6 58 5 65 Mbps mcs 7 65 72 22 Mbps iMm 100mA Operating Current Receive Deep Sleep Current 6 pA 25 C Associated Sleep current 2 m asleep 100 mA awake See AP Associated Sleep section for details The XBee is able to communicate
67. re 4 2 Power and Voltage Plotted Against Resistance Characteristics Relative to Resistance Power Voltage 16 Resistance k ohms Figure 4 3 Power and Voltage Plotted Against Resistance Scaled Axis 54 Microsoft Excel provides a function to plot a trend line for a set of data We made use of this ability to generate a fifth order polynomial curve as seen in Figure 4 4 We used Wolfram Alpha to evaluate the local maximum of this equation the results told us that 6 124895 kQ is the ideal load for maximum power in the given light environment Characteristics Relative to Voltage Power Trendline Power mW mi d Voltage V Figure 4 4 Power Plotted with Trend Line Against Voltage 55 Wolfram Al ha computational knowledge engine E E E Examples 22 Random Input interpretation maximize 0 0003 x 0 0064 xt 0 0576 x 0 1915 x 1 8257 x 0 0497 Global maxima no global maxima found Local maximum max 0 0003 x 0 0064 x 0 0576 x 0 1915 x 1 8257 x 0 0497 14 283 at x 9 10714 Plot 5 5 10 15 ix from 5 4 to 14 8 10 20 4 Download page POWERED BY THE WOLFRAM LANGUAGE Figure 4 5 Evaluating the Local Maximum of the Power Curve The curves revealed by this experiment are physical characteristics of the ALT10 12P which means that it can be scaled to rep
68. re in many ways dangerous to humans For this reason we sought to design a remote monitoring system that would be entirely self sufficient in order to significantly reduce its need for human interaction after being installed Remote sensors in many systems are powered by either long life batteries or the public electrical grid While batteries are an appropriate power source for short term systems permanent installations mostly rely on commercial power 88 of all electric power generated in the United States comes from non renewable sources oil coal natural gas and nuclear fuels 2 Primary Energy Consumption by Source and Sector 2013 Quadrillion Btu Percent of Sources Percent of Sectors Total 97 5 Petroleum 35 1 36 Natural Gas 26 6 Residential amp 27 Commercial 10 7 11 Coal 18 1 Electric Power 1 99 5 38 4 3996 Renewable Energy 9 3 1096 Nuclear Electric Power 8 3 8 Source Sector Does not include biofuels that have been blended with petroleum biofuels are included in Electricity only and combined heat and power CHP plants whose primary business is to Renewable Energy sell electricity or electricity and heat to the public Includes 0 2 quadrillion Btu of electricity net Excludes supplemental gaseous fuels imports not shown under Source Includes less than 0 1 quadrillion Btu of coal coke net imports Notes Primary energy in the form that it is first accounted
69. resent any light environment This is demonstrated by Figure 4 6 which shows curves for power versus load for three different light environments These conditions were achieved by placing a piece of cardboard over the PV module in different positions as to partially cover the cells see 56 Figure 4 7 Each position is approximately 2 inches up the long side of the module Position 1 is the least covered position Power Characteristics in Different Light Levels mjn Uncovered S Position 1 de Position 2 Position 3 40 50 60 Load k Ohms Figure 4 6 Comparison of Power Vs Load Curves in Different Light Environments 57 Figure 4 7 Cardboard Cover in Position 3 In a second test we used a multi turn potentiometer in place of individual resistors see Figure 4 8 We calculated the power at various loads in an outdoor environment and plotted the measurements on a similar curve Figure 4 9 compares the results of the outdoor test to the curve from the indoor test The two curves are plotted on different scales in order to show their similarities The outdoor curve orange has a much sharper peak than the indoor curve blue and also has a greater maximum value 58 Figure 4 8 Outdoor Environment for Testing the PV Module 59 Indoor and Outdoor Power Characteristics ndoor S Outdoor 16 2000 14 7000 6000 5000 3 og E 5 m LO i cg B 4000 e AL 23 P 22 E 3000 4 2000 2 j 1000
70. sought out used electric drills from which we could remove and repurpose DC motors DC motors and generators are physically identical and only differ in the 17 fact that motors consume electrical energy to produce mechanical energy and generators do the exact opposite Unfortunately two of the drills that were donated to us ran off of AC power and as a result did not have motors that could be used as generators A third drill did contain a DC motor but after preliminary testing we realized that the force required to turn the shaft was impractically high drill motors are designed to apply a lot of torque to the drill bit We eventually borrowed a DC motor from the Electrical Engineering Workshop at WPI for free The motor was primarily chosen because it was free and readily available It has a rated maximum power of 24 V but we do not know much more about it because it was manufactured in 19777 and no data sheet could be found When calculating the power that we can ultimately extract from the wind we need to consider the efficiency of both the wind turbine and the generator Given the lack of significant data available on both components we assume an overall efficiency of 10 2 2 4 Microcontroller The function of our system is to collect data from an array of sensors and wirelessly transmit it to another location to eliminate the need for human interaction with the unit The data collection is handled by an MSP430F5529LP microcontroller on a
71. tOr annae nennen nnen e eenen ennnnnen 16 Figure 2 7 Flow Diagram Showing Single Channel Mode nanne eenen eneen en eennen 20 Freire 2 5 AC TU User Ita eones 26 Figure 2 9 Schematic for a Battery Charge Controller aaan venen een nennnerereeensennnn 30 Figure 3 1 Top View of Turbine Showing Lock Block and Adjustable Mountings 31 Figure 3 2 Adjustable Feet of the Wind Turbine Frame nnn nnnn nennen nennen eenen ennennen 32 Figure 3 3 Battery and Circuitry Mounted in Turbine Frame nennen nennen nennen nnenennen 33 Figure 3 4 Shaft Collar Connecting the Turbine and the Generator noen 34 Fronte 9 5 Battery Terminal CONE COn ic 34 Figure 3 6 Safety Tubing for Power Cable from PV Module eeen 35 Figure 3 7 Cable Connector to be Attached to DC Generator oooocccnccnnnnncnnnnnnnonnnnnonononcnnnnnnnnnnanoss 36 Figure 3 8 Break Out View of Charge Controller and Circuitry aaan 37 Figure 3 9 Top View of Assembled Circuitry add 37 Figure 3 10 State Diagram for Microcontroller LOZIC nennen eneenenennnnnnensnsrsersnerevevevens 39 Figure 3 11 Example of an Emergency Buffer Stored in Memory aaneen eenen 4 Figure 3 12 One Buffer of ADC Samples Stored into the Flash Memory 43 vii Figure 3 13 Evidence that All Four Memory Banks Have Been Erased ccc eessesesesseeeeeeeees 43 Figure 3 14 Oscilloscope Reading Showing Transmission of Data on the TX Pin
72. tch Tsunami Node amp Wave Glider Ocean Observing Demonstation Project University of Delaware 2015 Online Available http maracoos org content fetchtsunami node wave glider ocean observing demonstration project Accessed February 2015 Firefighter Nation Using Surveillance Systems for Wildfire Detection Firefighter Nation 5 June 2013 Online Available http www firefighternation com article wildland urban interface using surveillance systems wildfire detection Accessed December 2014 K D e al Clean Technologies Energy Harversting European Union 2014 Voltree Power Bioenergy Harvester Voltree Power 2014 Online Available http voltreepower com bioHarvester html Accessed December 2014 67 8 Voltree Power Javelin Lifetime Voltree Power 2014 Online Available http voltreepower com javelinLifetime html Accessed December 2014 9 American Society of Civil Engineers 2013 Report Card for America s Infrastructure ASCE 2014 Online Available http w ww infrastructurereportcard org a p home Accessed December 2014 10 Aeroqual AQM 60 Air Quality Monitoring Station Aeroqual 2014 Online Available http www aeroqual com product aqm60 air quality monitoring station Accessed November 2014 11 T Chen Power Generation System Utilizing Wind Draft from Vehicular Traffic United States Patent US 7427173 B2 9 May 2006 12 Texas Instrumen
73. tely we decided to use a 12 V lead acid battery because it would be easy to charge and would not require complicated protection circuitry Admittedly this is a violation of SR 5 however we decided 2 oo that for the purposes of developing a proof of concept prototype that particular requirement could be walved To further satisfy PR 3 we used a charge controller in conjunction with the battery The charge controller is an analog circuit that is the connection between the primary power sources and the battery It takes the voltage difference across the terminals of the battery compares it to a regulated threshold and determines if the power from the primary sources should be used to charge the battery 1f the battery voltage is below a set threshold or dumped into a dummy load if the battery is fully charged The circuit itself was designed by DIY enthusiast Mike Davis and is published on his website There he described how he realized that he could use the popular 555 chip as a set of comparators instead of a clock The circuit also uses an electromagnetic relay to switch the destination of the input power and is intentionally composed of common components so that it can be assembled by just about anyone The entire circuit is shown below in Figure 2 9 17 20 555 Based Solar Wind Charge Controller O 1 12 V Battery Pos C 2 Wind Solar Pos Input O 3 Dummy Load RLY 1 Dumping O 4 Ground IC1 780
74. that affect the power output of PV modules such as their cell configuration and composition Large PV modules also tend to come with an equally sizable price tag Unfortunately SR 4 limits the size of our system and SR 6 limits our budget Because of this we needed a cost effective moderately sized polycrystalline PV module that could produce a decent amount of power even in less than ideal situations like cloudy days We bought a 11 suitable module for 40 from altE a company that specializes in manufacturing alternative energy products The ALT10 12P pictured below in Figure 2 2 is 10 8 efficient and has a rated power output of 10 watts 12 volts It measures 14 6 x 9 8 x 0 71 and can operate in temperatures between 49 and 185 F Furthermore altE specifically promotes the ALT10 12P for use in remote monitoring systems Figure 2 2 ALT10 12P Polycrystalline Photovoltaic Module 2 2 2 Wind Turbine Highways are also excellent sources of wind energy Their long corridor like shape funnels wind like a wind tunnel and high speed traffic can generate gusts of wind up to 15 MPH Placing a wind turbine next to the road to harness this energy would satisfy SR 1 SR 5 and SR 8 Like solar panels wind turbines come in many different sizes and configurations In residential and commercial applications the most common configuration has three blades rotating on a horizontal axis mounted on a pole or tower The rotation of the
75. the environment at predetermined intervals and data will be stored on site between daily transmissions The data will be wirelessly transmitted over Wi Fi to a data logging device where it can be analyzed and published Anomalous data will trigger an immediate emergency transmission 1 4 Existing Products and Patents Currently there are no commercial products that provide roadside monitoring in the manner that we are proposing However there are technologies on the market that have simila goals One product the AQM 60 Air Quality Monitoring Station by Aeroqual is a mobile air quality station that can be placed in many different locations The product boasts the ability to measure up to 10 different pollutants and environmental parameters The AQM 00 uploads all measured data wirelessly in real time giving an operator the most up to date information 10 However the product does not use renewable energies such as wind or solar power The AQM 60 is the only commercially available monitoring station available on the market at this time However many U S patents cover similar aspects of our product One such patent US 7427173 B2 is a roadside wind turbine that utilizes wind energy from cars to power a vertical turbine 11 The patent states that the turbine would be located alongside the roadway or in the middle of a highway barrier 2 METHODOLOGY In this section we go through the basic requirements our project will need A
76. tion The solar panel can have two major additions a maximum power point tracker MPPT and a mechanism that always points it towards the sun A MPPT has a dynamic load that changes to give the maximum power output from the solar panel An MPPT can simply be bought but usually cost upwards to around 100 which was over our budget Being able to point the panel towards the sun would definitely help in generating more power for the system We recommend looking into the previous MQP Azimuth Altitude Dual Axis Solar Tracker by Adrian Catarius and Mario Christiner advised by Alexander Emanuel In regards to the data network of the system there are several improvements and suggestions to make it better Current existing problems with the XBee modules include not using a TCP connection and their inability to use pin sleep mode When we configured the XBees to use TCP they refused to connect to one another only UDP connection was successful Fixing this problem would make the system more secure and the data transfer more reliable Another strange bug we encountered was the XBee pin sleep To sleep the XBee one must put a high signal on the sleep pin Although we implemented the signal into our code the XBee never enters sleep mode Putting the XBee to sleep would save a considerable amount of power 65 The data collected over the Wi Fi signal could be organized into a web based User Interface UI A customer might want to check on the module by
77. ts MSP430F551x MSP430F552x Mixed Signal Microcontroller Rev L 2009 2013 13 Texas Instruments MSP430x5xx and MSP430x6xx Family Users Guide Texas Instruments 2008 14 D International XBEE WiFi RF Module S6B User Manual Minnetonka Digi International 2015 15 D International XCTU software download DIgi International 2015 Online Available http www dig1 com products wireless wired embedded solutions zigbee rf modules xctu Accessed January 2015 16 D International XCTU User Guide 20 08 2008 Online Available http ftp1 dig1 com support documentation 90001003_A pdf Accessed 01 2015 68 17 M Davis A New amp Improved Charge Controller Based on the 555 Chip 2014 Online Available http mdpub com 555Controller Accessed November 2014 69 Appendix A main c include lt msp430 h gt include lt string h gt include lt stdio h gt include lt stdlib h gt include lt math h gt include lt stdint h gt include lt in430 h gt define MAX 10 max number of ADC samples to store into the buffer define THRESHOLD 500 max value to trigger the emergency bit define SPEED 1 how fast we get one sample for the ADC buffer Currently should take one minute for each of these loops define SPEED2 1 The higher the number the longer the time it takes should be 6 minutes define TXMAX 10 how many times the Tx sends out the same information
78. us locations while using eco friendly energy We researched ideas involving tsunamis wildfires and bridges until we finally settled on a roadside application After choosing a roadside application we developed system requirements that outlined size power generation and functionality of the system Each component was chosen based upon these system requirements We describe our development of the product in our Implementation section specifying how each component functions both by itself and as part of the overall product We present preliminary testing results in Section 4 5 2 Project Recommendations Project recommendations are suggestions to improve the system as it currently stands Although our project functions in the manner we wanted it to many improvements throughout the system can be implemented Improvements can be divided into two general topics power efficiency and data network Power efficiency refers to the mechanical aspects of the system including the turbine generator and solar panel The data network mainly involves creating a stronger and more reliable data network for the collected information A continuous theme throughout our project was the management of power in the system The three major contributors to power generation turbine generator solar panel can all be improved upon The design of the turbine should be researched more extensively Although we do think the design fits the application we are unable to say
79. voltage change the ADC can detect The MSP430 has the ability to set the reference voltage to 1 5V 2V or 2 5V We chose the 2 5V range because it allows us greater flexibility to add a diverse range of sensors into the network 13 In our system the ADC samples at a rate of once every six minutes At this speed 240 samples are collected daily We chose this sample rate due the available size we had in the flash memory addressed in the next section Every sample from the ADC is a 16 bit unsigned integer and we have 512B of memory available to store the data in Although the memory size is our main constraint sampling more than 10 times an hour would most likely not give us any new information thus making it an appropriate sampling rate The ADC has four different conversion modes Single channel Single Conversion Repeat Single Channel Sequence of Channels Autoscan and Repeat Sequence of Channels Repeated Autoscan Each conversion mode offers a different method for getting the conversion samples We decided to use Single channel mode due its simplicity In Single Channel mode the ADC waits to be enabled by the user 19 Once enabled it will get one ADC sample buffer and convert it into an integer value When complete the enable bit is automatically set back to zero and waits to be re enabled by the user As a comparison Repeat Single Channel mode operates in a similar fashion but with one key difference Once enabled the ADC w

Download Pdf Manuals

Related Search

Related Contents

Manual do Produto  MB11 Classic - Operating Instructions  Blueball First Class Readme Manual  1 informazioni generali  Manual de Instruções  Kohler K-1418-H2 User's Manual  Porter-Cable 4216 User's Manual  

Copyright © All rights reserved. Failed to retrieve file