Home

EE, NMT - Electrical Engineering

Contents

1. sees 15 S Michael Belges ooa eee etn hs teanchy ce huh foa E nO qu d cR ufa I5 3 2 Toute der rp 16 23 2 Antlionv Montoya ES soo aate us ioi ertt qus tudetit im Mm tdeo tut a E ded ears 19 4 Summary and Recommendations eese nennen nennen enne 21 References secos Rn ee UL eh RS Lh EE 23 Appendics M RE H 24 Appendix A 100 inch mirror coating simulation Matlab esses 24 Appendix B 100 inch mirror coating simulation ZPL Matlab Display 31 Appendix C Parallel library and system configuration scripts sese 40 Python auto configuration script ssessssseseseeeeeeeeeeee ener enne nenne nennen 40 Bourne Shell wrapper for the Python configuration script eeeeeee 47 Appendix D Beowulf Cluster User s Manual eene 49 iv List of Figures Figure 1 Comparison of the NAS MG benchmark using the Class S data set on 10Base Tand LOOB ase T Networks es ceecpetbe roit br eode E dps dee Vea an 7 Figure 2 Comparison of the NAS MG benchmark using the Class W data set on 10Base T and 100Base T networks sequar pe tis vi tube o atq Hori ete bid trvaucatvefun debes uve 7 Figure 3 Time required to complete the mirror resurfacing simulation program on a IOBaSsec T HebWOrE eese tastes des alise Meta d NE UR tend am aM a LER M 8 Figure 4 Time required to complete the mirror re
2. water_routine Jump to routine leave 1 Increment leave PORTP PORTP 0x40 Turns water pump off while s5 gt V MOISTURE Looks at the value of A D ADRSH PWDTYO 127 5096 duty cycle PORTP PORTP amp 0x40 Pump on delay 15000 Wait 15 seconds PORTP PORTP 0x40 Pump off PORTP PORTP 0xCO Turns moisture circuit off leave 1 Exits water_plant routine The routine to turn pump on while measuring soil moisture void water_routine while s5 gt V_MOISTURE PWDTYO 127 Charge battery with 5096 duty cycle PORTP PORTP amp 0x40 Pump on delay 5000 5 second delay Creates a delay of num 1ms void delay unsigned int num ae int i while num gt 0 i 1000 while i gt 0 i i 1 num num 1 48 Interrupt to take voltage measurements of systems every 2ms interrupt void rti_isr void s7 ADR7H Looks at position 1 amp 2 of switch select s6 ADR6H Looks at position 3 of switch select s5 ADRSH Looks at moisture sensor output level s4 ADRAH Looks at battery voltage level s3 ADR3H Looks at solar panel voltage level RTIFLG 0x80 Interrupt that allows the stop mode to work interrupt void irq_isr void 49 REMOTE SHIPBOARD HEAT DETECTION PROTOTYPE A Thesis Presented to the Faculty of The Electrical Engineering Department
3. cancel threads 33 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems thread cancel tasks R POS FIFO thread join tasks R POS FIFO NULL thread cancel tasks L POS FIFO thread join tasks L POS FIFO NULL free the Prometheus IRQ tl free irq PROMETHEUS IRQ Hc o s CU CO make sure the motors have stopped and the fan is off tl outb char 0 short R PWM ADDR tl outb char 0 short L PWM ADDR tl outb 0 DIO PORTO RARAN Figure 18 cleanup module example code When the user space control algorithm writes to the Control FIFO the function my handler is executed This handler reads the structure depicted below in Figure 19 from the Control FIFO my handler then wakes up the appropriate thread struct my msg struct int command int task unsigned int period int rw read 1 write 0 only used with PWM fifos char data only used with PWM fifos writes PWM command Figure 19 my_msg_struct When the Prometheus generates an interrupt our ISR intr handler is executed Similar to my handler this routine wakes up the appropriate thread but also clears the interrupt Now we come to the workhorse of our module thread code This is the code that actually reads the data from the hardware as well as writing PWM values to the motors Once a thread is created it begins executing Thus the first thing our threa
4. Cook off the temperature at which the ordnance detonates because of the thermal energy USS Forrestal Association http users erols com routts cv59 htm 2001 lives while an additional sixty were injured This disaster remains the single worst loss of life on a navy vessel since the USS Franklin was bombed in WWII while operating in the Pacific The total damage was estimated to be 72 million dollars as calculated in 1968 Following the accident the Navy was assigned the task of developing a remote firefighting system for Navel vessels 1 2 Customer Requirements Eric Wilson Materials Engineer NAVAIR China Lake CA provided requirements for the final prototype The prototype must be smaller than a cubic foot less than 10 pounds be a single use device that has a shelf life of at least thirty years and survive in a temperature of 1600 F while maintaining a temperature of 150 F internally These requirements are shown in the Table 1 1 Specification Requirement Specification Requirement Size 1 foot cubic Activation 250 F temperature Weight less than 10lbs Battery Life 5V for 30 minutes Life cycle non fire 30 years Shelf life 30 years Life cycle fire single use Frequency 2 39 2 45GHz Max External 1600 F Transmit distance 100 meters temperature Max Internal 150 F Transmit data Live and preprogrammed temperature data Table 1 1 List of Customer Requirements These customer requirements were the origin
5. max min 6 8K slope 0 1K power vs time scale Oto 400K 110K i max min 6 8K freq gt UT date Oct 28 UT 2001 301 20 27 52 LST 18 2 hrs Source Time 90 a a 80 179 Center frequency EI CygEMN E 1 1420 40 MHz s G90 60 spacing 0 04 MHz L hi 50 number bins 40 i Cass tsys not yet calibrated 0 40 o ij MI7 ij alan Andie Pd i i GNpole limits Hm ad Sgr 20 i Sun E 10 m MEI LI ft CL TIL pee eye ee tt 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 azimuth deg Recvr freq 1421 16 pwr 222 counts temp 111K atten dB Status slewing ERRORS gnd 1 radio 0 74 7 km s Vpeak Fpeak 1420 76 MHz Figure 3 2 2 Original SRT Control Panel There were a number of changes that had to be made to the software First and foremost the display as it arrives is fairly confusing for people with a technical background let alone people with little or no knowledge about astronomy I had to read through the manual a few times to understand how to use the display to control the SRT It is unreasonable to expect visitors to read 15 pages of instructions before using the SRT so the display had to be changed Also NRAO specified a bulletproof system which would be safe to leave in the hands of unsupervised visitors This ruled out leaving the mouse and keyboard accessible to the public Input would be accomplished vi
6. Bill Willems 1 3 02 SUBDESIGN Pos Pulse Accumulator Ch_A INPUT clock for accumulators Ch B INPUT RESET INPUT ACC OUT 15 0 OUTPUT VARIABLE Accumulator 15 0 DFF we gotta put the count somewhere BEGIN Set clocks and clears Accumulator clk Ch A Accumulator clrn RESET ACC OUT Accumulator q So count already IF Ch B VCC THEN Accumulator d Accumulator l q 1 ELSE Accumulator d Accumulator q 1 END IF END pwmgen tdf PWMgen Generate PWM with a duty cycle based on input value Kyle Lamb 1 3 02 subdesign PWMgen clock INPUT PWM cmd 7 0 INPUT RESET INPUT DC out OUTPUT dir out OUTPUT VARIABLE count 6 0 DFF BEGIN count clk clock count clrn RESET dir out PWM cmd 7 count d count q 1 IF PWM cmd 6 0 lt count q THEN DC out GND ELSE DC out VCC END IF END 52 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Appendix C Breakout Board Driver Note bb mod c bb h prometheus h bb c image c and Makefile are not included in text here You should have received a copy of these files in electronic format with this text 53 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Appendix D Breakout Board Schematic Note The schematic file is not included in text here You should have received a copy of this file in electronic format with this text 54 R
7. 2 3 4 5 6 then to EEPROM Write the program as a text file and save itas tea in the USER directory of the Brainstem folder Open the console window and type the following to compile the program steep file name To save to program to the brainstem execute the following With the Brainstem connected and powered up type the follow load file name 2 X where X is the location you want it stored in You can choose a number from 0 3 To launch a program enter the following launch 2 X where X is the program location you want to launch To have a program launch from EEPROM execute the following script e 218 X0 where X is 15 18 for file location 0 3 i e 15 loads the program stored in slot 0 e 219 this stores the new values to the EEPROM To stop a program from running there are two options e 2 24 executes a power up resent and stops all programs from running e 222 X where X is the program number that is running terminates just that one program Appendix VIII Budget Table 5 1 compares the project development cost with the cost of reproduction This information is important to demonstrate the feasibility of this project in teams of the long team goals Item Distributor Development Cost Reproduction Cost Operational National 0 00 1 11 Amplifier Semiconductor LMC6442 Voltage Controlled Mini Circuits 0 00 15 95 Os
8. 2MHZ E clock 9us conversion ATDCTLS 0X70 8 bit convert continuous convert multichannel Setup for real time interrupt RTICTL 0x82 Set interrupt rate at 2ms RTIFLG 0x80 Clear source of interrupt To use STOP MODE of HC12 46 INTCR 0x40 _asm andcc 6f Enable interrupts enable Main program to control battery charging amp planter watering cycle while TRUE DBug12FNP gt printf Start n r leave 0 if s4 lt V BATT MED amp amp s3 gt V SOLAR charge_battery Charge battery with determined level else if s4 gt V BATT MED PWDTYO 127 50 duty cycle water plant Jump to water plant routine DBugl2FNP printf Stop n r _asm stop Puts the controller into stop mode to conserve power Charges the battery using 100 duty cycle void charge_battery while s4 lt V_BATT_MED amp amp s3 gt V_SOLAR PWDTYO 254 Determines switch position amp moisture level value then determines if plant needs watering ui void water plant PORTP PORTP amp 0x80 Turns moisture circuit on PWDTYO 127 Charges battery with 5096 duty cycle while leave 0 i switch s7 Looks at the value of A D ADR7H case SWITCH_POSITION_1 47 water_routine Jump to routine leave 1 Increment leave case SWITCH POSITION 2
9. AT Through roof Cables run to Lab Figure 3 1 1 Concept Design for Mounting SRT A literature search confirmed that since the parabolic dish is of a metal mesh construction wind loading would not be a problem The architect s only concern was for the dish to be secure He was assured that it would be and as a follow up he was contacted again when the dish was installed to confirm it was a secure installation Since the antenna was to be controlled from a lab on the second floor of the Workman building via a personal computer about a hundred feet of control and signal cables had to be run between the antenna and the lab It was necessary to find out which labs the cables would go through and for access to conduits on the third floor and obtain permission to run the cable in those existing conduits One professor on the third floor would not allow access to the conduit run in his lab due to sensitivity of his experiments so an alternative route was needed and permission for routing the cable there was obtained The Closed Circuit TV video cable was run through a different set of conduits and permission to run that cable was also obtained I devised a bracket assembly for temporary mounting of the camera to the roof stanchions and hooked up a power cord Based on how the project goals were to be divided up I went ahead and made a Gant chart to be approved by everyone From this further division of tasks I was to work on the watchdog whi
10. How the SRT Works cis ziBix Tracking Information Figure 3 1 1 Tutorial Slide shows In addition because of my background as an electrician please see resume in appendix I designed and submitted for approval the power and signal conduit runs for cables run inside of conduit when the entire project is moved for use at the VLA Visitor s building Figures 3 1 3 and 3 1 4 Final design approval from VLA personnel and the NRAO safety officer was needed before construction could begin N Elevation of VLA Visitors Center Outside quit door p ey SRT Placement Pat 3 12 isch sips at 2 1 inch PVC anchored ux pot vith cowciete ehetieal pube condit Det L 1 for SIGNAL Waathar proof 1 for CONTROL Exectnical Box Figure 3 1 2 Preliminary Installation Details of Outside Designs e 310 seh ii pipe To existing telephone comit oo g see detail dravine 4 in dapat prie bends PVC x 1 2 in electrical grade PVC V itor Center Ding Ground Level PVC buried 12 to 18in underground Bs 1 2 in 90 degree r bends PVC To Electrical Box see detail diving 2 Detail 4 Detail 2 ein EMT existing telephone conduit abandoned X in electrical grade PVC mounted on post or
11. R1 550 ohm L Printed Circuit Layout for IR Schmitt trigger APPENDIX J Spreadsheet of all parts purchased for the project including vendors and costs a e VENUE INDE Egal m LEESE MMC wl ee ee CL EU ee cl RE Gia VOR votvotagedetedoreset 4 oe 724 Giolab HotlekHT 680 encoder 8 aer 1722 10 8 Giolab Glolab Giolab Glolab 08F2612 Hallogenic hall effect OHO9OU 16N8553 tri state dip switches d l 47 Gsr2133 i4pin dip socket o o 6 14 2 4 472 472 5 5 4 2 11 02 4 2 6 7 o 2 2 1 3 1 2 1 1 1 3 0 0 l o En 3 1 3 62 3 84 9 12 1 0 2 5 3 3 94 59 8 2 A or EN 5 8 9 0 0 6 18 ED3120 ND 8 CC1205 ND K4Ai6 ND 5 0 8 9 AT490 ND N 51F1543 Header st post 1 C1 c Co C1 1 6 7 o 2 2 1 62 14 6 1 5 1 6 10 1 6 1 62 1 84 0 0 7 3 0 3 Oj 0j N 0 5 3 Glolab Goao Gioie Glle Newark Newark 0 0 0 3 zi Al OO TH HM aN ola Ni o Tono gu exo AA a lt lt 4 N ua Digi Key Digi Key A1448 ND_____ CONN PLUG IN LINE 2 CIRCUTT 6 7 102 DigiKey Dig Key WMa6a1 ND HEADERSPOS i56FRICLOCK 6 037 222 Dig Key MWMetizND CONN HOUS 3 POS 156 031 186 Dig Key WM2300 ND CRIMP TERMINAL 156 70 oo 42 2 36 Dig Key OW ND RES 10 OHM 2W 5 METAL RES 20 OHM 2W 5 METAL RES
12. Some other possibilities for applications of our work include using our simplified interface and program to benefit science hub schools In addition the idea of making CD s of the slide shows available to generate revenue has been raised 5 2 Recommendations For Future Work A nice addition to this project would be to have the java program stay open all the time to keep the SRT running and still have a choice of selecting other educational demonstrations This will cut down on the time it takes to track an object As it is the SRT must find its motor limits every time the java program starts up Furthermore the java program must shut down to change celestial objects because the objects are selected through the command line The program must be restarted every time the user wants to select a different object Further work with the touch screen interface and the program s command structure would eliminate this requirement Automatic features in software could be used to stow the dish pointed up at the zenith when inactive or under severe weather conditions If we had more time we would like to incorporate the VLA weather station program to flag on a high wind indicator and the 3l SRT program would cue make the antenna automatically stow and load the touchscreen layout which will inform the public that it is inoperable Another idea would be to open two programs at the same time with one dos prompt Currently when a celestial object
13. o a 9007 5 md 40 20 0 HHHH 8 RA q de oo Time min Figure 3 4 Internal temperature plot for insulation prototype 19 3 3 Temperature Sensor by Stephen Stange This project required the temperature sensor to be small inexpensive lightweight and low power The sensor also had to be able to reliably read temperatures in the full range of the customer requirements 250 F to 1600 F see Section 1 2 To meet these requirements several types of temperature sensors were considered The data that these temperature sensors output had to be transformed into a signal that could be read by the data processor inside the device 3 3 Evaluation of Temperature Sensors Several different types of temperature sensors were considered for this project The first option considered was a thermocouple Type K thermocouples described in Section 2 3 have a temperature range that can accommodate the requirements of this project 250 F to 1600 F Another option considered was a RTD resistance temperature device Also considered for this project were infrared radiators and thermistors The project requirements were thoroughly considered while trying to choose a viable option for the temperature sensor One option for the temperature sensor that was considered was to use type K thermocouples like the ones originally intended for the Peltier junction power supply A thermocouple consists of two dissimilar metals joined together at
14. C 2 350926 15 02 39 5 2414 5 2393 5 3174 5 3955 Average Temp change in Fahrenheit 15 02 40 5 2414 5 2393 5 3955 5 3955 4 231667 15 02 41 5 2414 5 2393 5 3955 5 3174 15 02 42 5 2414 5 3174 5 3955 5 3174 Temperature Change in Celcius 15 02 43 5 2414 5 2393 5 3955 5 3174 2 350926 15 02 44 5 2414 5 2393 5 3955 5 3174 15 02 45 5 2414 5 2393 5 3955 5 3174 Power Absorbed Mass Specific Heat Change in Temp C 15 02 46 5 2414 5 2393 5 4736 5 3955 2 334187 Watts time 15 02 47 5 2414 5 2393 5 3955 5 3955 15 02 48 5 2414 5 2393 5 3955 5 3955 Energy Absorbed Mass Specific Heat Change in Temp C 15 02 49 5 3195 5 2393 5 3955 5 3955 116 7094 Joules 15 02 50 5 2414 5 2393 5 3955 5 3955 15 02 51 5 3195 5 2393 5 4736 5 3955 15 02 52 5 2414 5 2393 5 5518 5 3955 15 02 53 5 8195 5 2393 5 4736 5 3955 15 02 54 5 2414 5 2393 5 4736 5 3955 15 02 55 5 8195 5 2393 5 4736 5 3955 15 02 56 5 2414 5 3174 5 4736 5 3955 15 02 57 5 8195 5 3174 5 4736 5 3955 15 02 58 5 2414 5 3174 5 5518 5 3955 15 02 59 5 3195 5 3174 5 6299 5 4736 15 03 00 5 2414 5 3174 5 5518 5 4736 15 03 01 5 3195 5 3174 5 5518 5 4736 15 03 02 5 3195 5 3174 5 5518 5 4736 15 03 03 5 3195 5 3174 5 5518 5 4736 15 03 04 5 3195 5 3174 5 5518 5 4736 48 15 03 05 5 3195 5 3174 5 5518 5 4736 15 03 06 5 3195 5 3174 5 5518 5 4736 15 03 07 5 3195 5 3174 5 5518 5 5518 15 03 08 5 3195 5 3174 5 5518 5 5518 15 03 09 5 3195 5 3174 5 5518 5 5518 15 03 10 5 3195 5 3174 5 5518 5 5518 15 03 11 5
15. Clear all prior data in Matlab write output file clear all n n write output file fprintf Loading data file please wait n n write output file S S n write output file ci ci n write output file co co n write output file N N n write output file n R P write output file P 20 t P n n R XY write output file XY 20fNt XY n n R mirror write output file Coating 20fNt 10 8 Coating n n Coating is scaled up by 10 8 in the data file scale it back down write output file Coating Coating 10 8 MnWMn R mirror write output file Mirror 20fNt Mirror n n When printing more than 2 dimensions ZPL does not put any kind of break between the higher dimensions so I have to print it like this write output file G zeros 2 S41 2 S 1 QU 3 n for n 1 to 3 do 1 2 S 1 1 2 S 1 1 n write output file G d n 20f t G n n end write output file fprintf done n n Close the output file close output file All the computational work is now done get the total file I O time timervalue CheckTimer if dotiming then write File I O Time in seconds t 12 4f n timervalue end end 37 Matlab Display Code You must run the app
16. Figure 1 3 D electric Held Vector oov imp oterim tae xe cer nme o ero erai anaes 6 Pistite 2 Electric field meter sio oe be e Rb edes ede eR a e i s 9 Figure 3 Waveforms produced by original data sseueusuus 10 Figure 4 FFT and filtering of HG ese ERR E Red Mta SEE OQ ev etat 14 Figure 5 Original filtered and interpolated Hg data 15 Pigute o ET OF B Class cls deeem e da ta uated Maud Co vay vente 16 Ligure 7 Hilbert transformi eese nepi eian Rer CHE rao aes 17 Figure 8 Lowpass filter for quadrature demodulation 18 Figure 9 Constant phase shift FIR filter eee eee e eee eens 19 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Small section of original data cc ccc cece e eee e eee e eens 23 Horizontal and vertical components of E field 25 E field directional components relative to B field 26 Angle of Hg switch to horizon esee 27 Orientation of instrument c esses 20 E field components relative to altitude us 30 Chapter 1 A Brief History of Thunderstorm Data Analysis as Relates to Electric Field Meter Data Collection Project Description Thunderstorm data analysis has been going for many years This project came to us as the continuation of an older project intended to broaden the knowledge base of electric fields in thunderclouds
17. exclusive OR expression AND expression exclusive OR expression AND expression AND expression equality expression AND expression amp equality expression equality expression relational expression equality expression relational expression equality expression relational expression relational expression shift expression relational expression shift expression relational expression shift expression relational expression lt shift expression relational expression gt shift expression shift expression cast expression shift expression lt lt cast expression shift expression gt gt cast expression additive expression multiplicative expression additive expression multiplicative expression additive expression multiplicative expression multiplicative expression cast expression multiplicative expression cast expression multiplicative expression cast expression multiplicative expression cast expression cast expression unary expression type specifier cast expression unary expression postfix expression unary expression unary expression unary operator cast expression unary operator postfix expression primary expression postfix expression postfix expression primary expression identifier constant expression loop statement while expression statement do statement while expression for expression expression expression statem
18. http www ee nmt edu wedeward EE382 SP01 ee382 html GP2D120 distance sensor http www fcc gov FCC regulations literature APPENDIX M Gantt charts for the first and second semesters September Project Selection Contact Customer Initial Brainstorming Complete SOVV and send to customer Develop Gant Chart Research RF transmitter encoder decoder alpha Research RF transmitter encoder decoder beta Research sensor possibilities for signals Give 5 min presentation Research and order alpha and beta transmitters Research and order brake sensor Research and order turn signal sensor Research and order LED unit Brainstorm alpha transmitter receiver pair interference Brainstorm beta transmitter receiver pair interference Review FCC regulations Test alpha transmitter Test beta transmitter Test brake sensor Test turn signal sensor Test LED unit Choose alpha or beta transmitter based on testing Block diagram of chosen system Initial setup of prototype Sensor selection Report Transmitter selection report Research PCB process Oral presentation of progress Task to be completed by Brandon Task to be completed by Dave Task to be completed by Ivan Task to be completed by the entire Group October November December GANTT chart Bi weekly progress reports PCB Research Improvement of LED array circuitr Finalize output logic Power supply filtering protection Connectors PCB layou
19. outfile write output file plots plot skymap spectra and power vs time displays procs coordinate procedures in other classes sport communicate with controller via serial port Srt main class initialize software and telescope time calculate time in all the various formats SRT uses Table 3 2 1 SRT Classes and Their Functions When the SRT program starts it begins in the class named srt This class initializes all variables and runs the telescope to its mechanical stops so that it has a base position referred to by the software as the stow point This is necessary because the azimuth and elevation drives do not have absolute encoders Every time the software is shut down the computer forgets where the dish is pointed After finding the stow position it starts an endless while loop which hands over control of the dish to other classes At this point the user is free to issue commands using a combination of control buttons at the top of the screen and a text input box at the bottom Sources are selected by clicking on them on the skymap Figure 3 2 1 shows the SRT control display in its original configuration 13 E13 clear atten source Stow track Azel npoint bmsw freq offset Drift record EE 0 x Reman Cal Vane comerr j 1 n Omm5000 antenna drive status comerr 1 iv spectrum integ 0 1 min 420 40 span 1 60 MHz Antenna coordinates emd 95 0 4 0 deg waiting on azimuth 3 trans move 0 5000
20. s600 9700 3800 00 10000 Raw B data 1000 3100 9200 9300 9400 samples 9600 9700 9800 3800 10000 MU THU 9300 9400 Re 600 9700 doo 9100 Figure 10 Small section of original data To start we took the fast Fourier transform of the mercury switch data to find the components of the frequency Because the mercury switch data is similar to a square wave the FFT has several harmonics of the fundamental frequency Also all of the frequencies are aliased by the FFT process An initial filter is applied to the output of the Hg switch FFT to isolate the fundamental frequency from the spurious data See Figure 4 To clean up the filtered Hg data even more we converted the filtered data which was a rough sinusoid back into a square wave by making any positive value a 1 and any 23 negative value a 0 This new cleaner square wave was then filtered again using the same filter from before This accounted for slight changes in the amplitude of the filtered Hg switch data We also interpolated the filtered Hg switch data by adding three more points for each existing point to make the sinusoid smoother See figure 5 A phase shift was added to this filtered data to put it in phase with the B field data Once those were matched the phase of this new sine wave was compared with the data from the charge amplifier and the Hall Effect sensor Often the electric field in a cloud is completely horizontal By investigating the phase shift b
21. 14 32 09 5 3195 5 2393 5 2393 5 2393 14 32 10 5 2414 5 2393 5 2393 5 2393 14 32 11 5 3195 5 2393 5 2393 5 2393 14 32 12 5 3195 5 2393 5 2393 5 2393 14 32 13 5 3195 5 2393 5 2393 5 2393 14 32 14 5 2414 5 2393 5 2393 5 2393 14 32 15 5 3195 5 2393 5 2393 5 2393 14 32 16 5 3195 5 2393 5 2393 5 2393 14 32 17 5 3195 5 2393 5 2393 5 2393 14 32 18 5 2414 5 2393 5 2393 5 2393 14 32 19 5 2414 5 3174 5 2393 5 2393 14 32 20 5 2414 5 2393 5 2393 5 2393 14 32 21 5 3195 5 2393 5 2393 5 2393 14 32 22 5 3195 5 2393 5 1611 5 2393 14 32 23 5 3195 5 2393 5 2393 5 2393 14 32 24 5 3195 5 2393 5 2393 5 2393 14 32 25 5 2414 5 2393 5 1611 5 2393 32 14 32 26 5 3195 5 2393 5 2393 5 2393 14 32 27 5 2414 5 2393 5 1611 5 2393 14 32 28 5 2414 5 2393 5 1611 5 2393 14 32 29 5 8195 5 2393 5 2393 5 2393 14 32 30 5 2414 5 2393 5 1611 5 2393 14 32 31 5 2414 5 2393 5 1611 5 2393 14 32 32 5 2414 5 2393 5 1611 5 2393 14 32 33 5 2414 5 3174 5 2393 5 2393 14 32 34 5 2414 5 2393 5 1611 5 2393 14 32 35 5 2414 5 2393 5 1611 5 2393 14 32 36 5 2414 5 2393 5 1611 5 2393 14 32 37 5 2414 5 2393 5 2393 5 2393 14 32 38 5 2414 5 2393 5 1611 5 2393 14 32 39 5 2414 5 2393 5 1611 5 2393 14 32 40 5 2414 5 2393 5 1611 5 2393 14 32 41 5 2414 5 2393 5 2393 5 2393 14 32 42 5 2414 5 2393 5 2393 5 2393 14 32 43 5 2414 5 2393 5 2393 5 2393 14 32 44 5 2414 5 2393 5 1611 5 2393 14 32 45 5 2414 5 2393 5 2393 5 2393 14 32 46 5 2414 5 2393 5 2393 5 2393 14 32 47 5 2414 5 2393 5 1611 5 2393 14 32 48
22. 82 1 6 14 5 19 4 5 3 05 2 2 26 9 IR Distance Sensor eo A1550 ND CONN PLUG IN LINE 3 CIRCUIT WM2300 ND CRIMP TERMINAL 156 e oos 036 175 175 Resistors 54k 16k HOKAdjPoi Adjustable pot qom 1 1805 JBrakeCotactPad O O4Fi089 Snap dome Push button switch 436 ND CONN PIN 24 18 AWG MALE E oO il 3B K 2 TurnSignalContactSwith qo l l 1L 6 ait fect Sensor WNM2300 ND CRIMP TERMINAL 156 s 906 5 ome r 0 37 3i 25 pl Phototransistor Sensor PN 168 PHOTOTTRANSISTOR 2 0 05 0 1 MAX 492 MAX 492 OP AMP 2 est 0 50 1 A 1494 ND 4 CT PRE TIN PCB SOCKET hdr 1 0 37 0 37 A1436 ND CONN PIN 24 18 AWG MALE 4 0 102 0 4 XX Board layout material 2 x 4 1 0 5 0 5 TOTAL 2 37 K 3 APPENDIX L Literature review of major web sites referenced for parts schematics and project information http www holtek com pdf consumer 3 18e PDF Encoder http www linxtechnologies com ldocs f prod html Transmitter Receiver pair http www glolab com modules module1 html Transmitter Receiver pair http www digikey com Parts supplier http www e switch com Parts supplier http www newark com Parts supplier http www analogdevices com Tilt sensor http www optekinc com book pdf oh090u pdf Hallogenic Hall effect sensor OHO90U
23. Acknowledgments scies sdestutia citate a ARUM uctus deveebaqsn coe ct E ii List of Figures and Table esses eene nnne nnns iv I Introunction ue ee Se ea teste ead Bae ates eee 1 LA NNDE Senibr Desten OUESE oo sua dace openPR Ure ee dada baa EN EE EU EEEL SRE 1 1 2 The National Radio Astronomy Observatory cese 1 13 The Small Badio Pelescope c vede ved ed ree RE e oa hes Re areas 2 2 Overview of Radio Astronomy Project sese 4 2 1 CUSTOMER regule melits s aces se er eode ut pec aede e Rai bes Mop ey RE a e a 4 2 2 Assignment Of WOtk 5 ceni estes ento EAR TERR T EEEE E eye seed aiii ia 3 Discussion of Individual and Group Accomplishments ssssss ST Gerald cuestan sm um blat E RR DARIN GERA Lastabt p d oa a ssepe uae ea Dui ue tases 6 SM pM cT P C RENE 12 Dish BEDS tee toes a tcd Qu Ra delia 20 3 4 MG nuo e iaie 25 4 Final State of Project oos HUE ARIES Mus duis tis iude UNE E EE 29 4 Binal Testfig 2i re D RN RERO OR ER QU a eins RU ewe Fora te Ure ie t pane 20 4 2 TS CO ICA AP 30 5 Summary and Recommendations sss eene eene 31 5 Summary or Current WOEk x ioo odio ptt pondo o tee f totes eoo latin 31 5 2 Recommendations for Future Work esees essere nennen 31 5 3 Recommendations for maintaining the system sesesse cere 33 ADDe
24. Co23 boolean true Co24 boolean true inner clips 31 ci integer 4 number of clips per inside filament 1 6 inner clip positions true if a clip is placed in that position false if not Cipl boolean false Cip2 boolean true Cip3 boolean true Cip4 boolean true Cip5 boolean true Cip6 boolean false outer clips co integer 5 number of clips per outside filament 1 6 outer clip positions true if a clip is placed in that position false if not Copl boolean true Cop2 boolean true Cop3 boolean true Cop4 boolean true Cop5 boolean true Cop6 boolean false yog N integer 12 ci 24 co total number of clips used region R XY Eque Sed vs qd lJs RP zu 1 N Fo suis s RG 1 2 S 1 1 2 S 1 Ko We 34 R_D 1 2 S 1 1 2 S 1 Rep len Arle R delta 1 2 S 1 1 2 841 1 N 1 3 R rsqgrd 1 2 S 1 1 2 841 1 N 1 1 R mirror 1 2 S 1 1 2 S 1 1 1 1 1 var output file file timervalue double Cip array 1 6 of boolean Cop array 1 6 of boolean Ci array 1 12 of boolean Co array 1 24 of boolean Cx array 1 6 of double Fi array 1 12 1 2 of double Fo array 1 24 1 2 of double XY R XY double P R P double temp2 R G double temp3 R delta double delta R del
25. Of New Mexico Tech By Scott Dearie Camden Mullen Summer Rhodes Stephen Stange In partial fulfillment of the requirements For the course EE 482 Senior Design Project II May 1 2002 2002 Stephen Stange Scott Dearie Summer Rhodes and Camden Mullen Acknowledgements We would like to recognize the following for their contributions to this project Dr R H Bond NMT EE Dept Dr D Hirschfeld NMT MATE Dept Dr Scott Teare NMT EE Dept Dr Stephen Bruder NMT EE Dept Dr Kevin Wedeward NMT EE Dept Dr Richard Scott NMT EE Dept Dr Ron Thomas NMT EE Dept Dr Aly El Osery NMT EE Dept Scott Savage Savage Industries Mark Whitney Acroname Inc Steven Hunyady Langmuir Labs Tucson Amateur Radio Association Jim Sanchez Solectek Industries Paul Vinsand Mini Circuits Rita Moore Mini Circuits John Hardy NMT MATE Dept Kevin Loveall TA Mfg Co Rand Poplar Ferrotel USA Maxim IC Applications Department and Online Services Steve Wasson NMT Thomee Wright NMT Executive Summary A prototype of a heat detector was designed for the Naval Air Warfare Center NAVAIR China Lake CA The purpose of the prototype was to provide early detection and warning about the presence of fires and their relative danger to ordnance This device was to have an indefinite shelf life before activation Once activated by heat the device was to transmit via a radio frequency RF signal an encrypted data set The data se
26. PEs arranged in parallel Figure 1 below is a block diagram of the current RDPP design i4 Lg d i Figure 1 Block Diagram of the Reconfigurable Data Path Processor Each Processing Element is identical and other than five input data registers contains purely combinational logic There are three 24 bit data inputs to the PE that are stored in its internal registers There are two 24 bit registers that are used for storing constants There are 67 configuration bits that are used to configure the data paths within the PE for different computations Finally there are two 24 bit outputs Figure 2 below is a block diagram of the Processing Element that highlights all of its I O and control signals In2 In1 In0 RIn2 en RIn1 en RIn en ALU CLIP sel mux3 N wUx3 7 Ni d muxo A mux set pad In2 Int Ind DR1 DRO M oe EON v sel oN MUX T set al shift sel xb 04 m AL SHIFT 475 sel xb1 XBAR5x3 r1 alu X sel xb g MUL CLIP m ALU OUT jset alu shift set mul shift set mul round sel muxS set mul clip CLIP MUL_CLIP ALU SW MUL Sw ALU CLIP Out sel mux5 E Configuration Bits B Data Constant Registers Blinput Data MOutput Data W internal Nodes Figure 2 Block Diagram of Processing Element Highlighting I O amp Control Signals There are three primary data paths within the PE a delay path a multiplication path and an arithmetic path Figure 3 on t
27. The filtered B field data was then interpolated to smooth it out and keep it on the same time scale as the filtered and interpolated Hg switch data The phase relationship between the Hg switch and the B field is found using quadrature demodulation This method takes a sinusoid and then shifts it 90 degrees See figure 7 16 Hg in blue shifted Hg from Hilbert Transform in red 304 304 1 304 2 304 3 304 4 304 5 304 6 304 7 304 8 304 9 305 Figure 7 Hilbert transform magnitude PE e gt 1 1 L 1 l L Each of these sinusoids is then multiplied by the sinusoid that contains the phase shift This results in two different sinusoids that are twice the frequency of the original sinusoids plus a component that is the sine or cosine of the phase angle Using a lowpass filter to get rid of the high frequency sinusoid the sine and the cosine of the phase angle remains By dividing the sine by the cosine and taking the arc tangent of the result the exact phase angle is extracted This is far more accurate than just using a cosine or a sine to determine the phase See figure 8 for the lowpass filter that isolates the phase 17 component of the quadrature demodulation Also see chapter 4 for more detail regarding this quadrature demodulation technique Fast Fourier Transform of B Hg in blue with filter in red c magnitude 500 3000 3500 i 4000 4500 5000 D 1000 1500 TEE requency in terms of samples Fast Fouri
28. Type javac filename java If the program compiled with no errors filename class will be created CAU ea eo dc um Run the program by typing filename java Change directories to the one with the Java program that you want to compile 36 Kit Specs Schematics located on CD Radiometer Characteristics L O Frequency range L O Tuning steps L O Settle time Rejection of LSB image 3 dB bandwidth I F Center 6 dB LF range Preamp frequency range Typical system temperature Typical L O leakage out of preamp Preamp input for dB compression from out of band signals Preamp input for intermodulation interference Square law detector max Control 1370 1800 MHz 40 kHz 5 ms gt 20 dB 40 kHz 40 kHz 10 70 kHz 1400 1440 MHz 150K 105dBm 24 dBm 30 dBm 4000 K a 0 dB attenuation 40 000 K at 10 dB attenuation RS 232 2400 baud 37 Antenna Specifications Manufacturer Kaul Tronics Inc Model Number S 7 5 Diameter 90 2 3m F D Ratio 0 375 Focal Length 33 75 85 7cm Gain 4 2 GHz 38 1 dBi Gain 1 4 GHz Weight with mount 160 lbs Beam Width 7 0 Degrees L band 38 Gant Chart Explanations In the following diagrams as can be seen the individual tasks of the overall project are placed on the left hand side while the time line and deadlines extends out to the right The assignment of these tasks is distinguished by color code the key is in the bottom rig
29. Utilizing the above attributes we have successfully designed a revolutionary planter that meet all the milestones and will automatically water a plant species without human intervention During the duration of the project we met weekly with our advisor to discuss progress and problems with the project We emailed our customer weekly to update him on our status Biweekly we provided our customer and advisor with a progress report After conferring with our customer and advisor we have all determined that we have successfully met all criteria for the project 32 Chapter 5 References Books The Art of Electronics Paul Horowitz and Winfield Hill Websites Plants usda gov all plants are unique Siliconsolar com People Dr Greg Donohoe UNM Electrical Engineering Department I appreciate all this research This is an important part of the Senior Design effort and should be a prominent part of your reports this is science Dr Bill Rison NMT Electrical Engineering Department Dr Ron Thomas NMT Electrical Engineering Department 33 Figures 1 2 3 Doll cempore coy i e ge us PRU aboga dge El L 3l QO9QOA COD KR coco n0 00606 nv ttti Wubi vaasbge sc 34 Figures 4 5 for ey 9 go iof Yr E 10k R2 10k R 108 2 POT Q2N3904 Light Sensinq electrode to be dipped Into dirt 0 5 Vani Vb VbyR amp b VaiveVbe Yb B D Vb RI VsetV
30. and it was not until the end that took it upon myself to figure out how to properly connect the sensor The sensor required greater than 9V in order to operate properly and in all the tests which were performed earlier the voltage supply was always connected to 5V Once the sensor was connected to the proper 9V supply it worked beautifully and was mounted and connected up to the Breakout Board post haste 24 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Bump Sensors Heather Bitsoi The basis for this project was to see if the Prometheus SBC was feasible for a mobile robot project and to test the capability of the processing power it has The customer decided to add additional sensors to the provided robot One of these sensors were the bump sensors The bump sensors are momentary switches that were placed around the base of the robot The four switches were provided by the client The wire use to connect these switch was a wire coat hanger The coat hanger was difficult to bend making it more difficult to get a perfect half circle The switch had small rollers on metal that made contact This roller had to be grinded down for the wire coat hanger to fit Figure 13 Bump Sensors in between the roller and lever The switches were mounted as it is depicted in Figure 13 The wire was not extended to the front of the robot because the switches would not work properly It was a redundant process to have the wire cover t
31. brainstem that decodes the encrypted signal and reads it properly The data stream can also be sent via a RF signal It has been demonstration that the transmitter is transmitting a signal of the correct frequency When a quadrature receiver is available the signal can be decoded to ensure that the signal can be read This project was successful in nearly all the revised required specifications The total developmental cost was 234 00 and cost of reproduction was 203 50 See Appendix VIII Using some of the suggestions discussed in the Future Considerations Section seen in Section 6 can bring the prototype even further along 36 5 0 Future Work This project has many potential improvements The first and most obvious improvement would be to place all electronic components into one integrated chip The device should be integrated into the ordinance design to prevent aerodynamic interference Insulation for the electronics must be minimized to aid in the aerodynamics of the device A receiving station that has the ability to decode the device signal must be developed or purchased in order to evaluate the transmitted signal This will also provide information about power characteristics that the signal needs to possess Research and testing can be done to examine how a transmitted signal propagates onboard ship Cost must further be reduced to make the device a feasible option 37 Works Cited ASTM Standard Test Method for Thermal C
32. can t complain 38 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Future Work Breakout Board and Driver Bill Willems With the reconfigurable Altera chip on the Breakout Board the possibilities for improvement are endless USB for instance is very CPU intensive if a USB controller could be placed on the Altera chip it would ease the overhead that now rests with the CPU If some image processing firmware could be developed imagine what we could do with that The Altera chip has a JTAG programming interface that allows it to be programmed through the parallel port The Prometheus has a parallel port so theoretically the Altera chip could be reprogrammed on the fly Before the driver for our Breakout Board is released into the public domain it needs to become more robust The reason it works so well for us is because we wrote it and therefore we know exactly how to interface to it am sure there is room to improve the efficiency of the code as well as some error checking Such improvements in efficiency could include breaking up the one version of thread code routine into several flavors one for the Prometheus A D interface the Digital interface and the Breakout Board One could also break them up into periodic and non periodic threads Future error checking could include code to ensure that the Prometheus Data Acquisition Circuitry as well as the Breakout Board is out there 39 RTLinux On A Small Mobil
33. tuned to half of the desired frequency One advantage of this oscillator was that it was able to function at temperatures up to 85 C Another advantage of this packaged oscillator was its power supply requirement of only 5V The third main component of the transmitter was the amplifier The amplifier used was also provided by Mini Circuits it is model number GALI 6F This amplifier was one of the few found that provided a 10X gain at 2 2GHz with only a 5V source The power gain graph for this amplifier can be seen in Figure 3 11 A main advantage of this amplifier was its cost effective price of 1 29 iBi xi Power Gain GALI BF Power Gain dB vof 1030 2080 3070 4060 o D o D 6040 7030 8020 3010 10000 FREQUENCY MHz Warning Applet Window Figure 3 11 Power gain graph for amplifier 34 4 0 Conclusions Research was conducted into each necessary aspect of the heat detection device Testing of the material was completed and electronic components were chosen integrated and assembled Feasibility was demonstrated for each of the original specifications except for the requirement of scavenging all power from the fire This one specification that was not met was researched for further physical development While a single fully integrated component was not created and destructively tested due to lack of funding and materials it is clear that this full integration could be easily completed It was n
34. 1 2 ActiveSheet Cells cellindex 2 2 ActiveSheet Cells cellindex 3 2 ActiveSheet Cells cellindex 4 2 ActiveSheet Cells cellindex 5 Next cellindex 1 End Sub Private Sub CommandButton2_Click IeTimerl Enabled ValFalse UltimaSeriall Stop End Sub Private Sub IeTimerl Timer i UltimaSeriall AvailableData ActiveSheet Cells cellindex 2 Format UltimaSeriall AnalogInput Ch1 3276 8 0 00 ActiveSheet Cells cellindex 3 Format UltimaSeriall AnalogInput Ch2 3276 8 0 00 ActiveSheet Cells cellindex 4 2 Format UltimaSeriall AnalogInput Ch3 3276 8 0 00 ActiveSheet Cells cellindex 5 Format UltimaSeriall AnalogInput Ch4 3276 8 0 00 ActiveSheet Cells cellindex 1 Time cellindex cellindex 1 If cellindex gt 120 Then cellindex 1 End Sub A 2 Gantt Chart Gantt Chart Sep 01 Oct 01 Nov 01 Dec 01 Jan 02 Feb 02 Mar 02 Apr 02 Responsible Person 1 2 3 41 2314 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 412 34 Statement of work all Research of Materials Ernest Jim Research of Thermocouples Daniel Rodriguez Research of A D boards Joe Hart Design Status Report all Build prototype wedge Ernest Joe Test Optics Ernest Joe Test Temperature Sensors D
35. 1 4 Project History 2 Technical Background 2 1 Subsystem Solutions 2 1 1 Laser Beam Dump Design 2 1 2 Heat Sensing Hardware Design 2 1 3 Data Collection System Design 3 Individual Contributions 3 1 Josef Hart 3 2 Ernest Jim 3 3 Daniel Rodriguez 4 Summary and Results 4 1 Discussion of Results 4 2 Laser Beam Dump Summary 4 3 Heat Sensing Hardware Summary 4 4 Data collection System Summary 4 5 Budget 5 Conclusion References Appendix A 1 Computer Code A 2 Gantt Chart A 3 Trial Charts A 4 Trial 1 Data A 5 Trial 2 Data A 6 Trial 3 Data A 7 Trial 4 Data A 8 Josef Hart s Resume A 9 Ernest Jim s Resume A 10 Daniel Rodriguez s Resume ro Ed ge e we 00 11 1110 Ui 4 WN NNN m LRL eB Be ee ee ee smasSeooveereemeun bh RWN SO SO 22 23 23 24 25 27 34 40 48 55 56 57 List of Tables Page 1 Project Specifications 9 2 Trial Results 17 3 Project Budget 20 List of Figures Page 1 Laser Beam Dump with Laser Beams 11 2 Final Beam Dump Dimensions 11 3 Schematic of Temperature Sensor Package 15 1 Laser Beam Dump and Power Meter Introduction 1 1 Project Description The project our group was assigned was to design a high energy laser beam dump beam dump and power meter with a given configuration to meet given specifications When operating the laser beam is directed into the V shaped segments and the plates absorb its energy Temperature sensors that are mounted to the plates are used to measure th
36. 14 42 45 5 1633 5 2393 5 4736 5 3955 14 42 46 5 1633 5 2393 5 4736 5 3955 14 42 47 5 1633 5 3174 5 4736 5 3955 14 42 48 5 1633 5 2393 5 4736 5 3955 14 42 49 5 1633 5 2393 5 4736 5 3955 14 42 50 5 1633 5 2393 5 4736 5 3955 14 42 51 5 2414 5 2393 5 5518 5 3955 14 42 52 5 1633 5 2393 5 4736 5 3955 14 42 58 5 1633 5 2393 5 4736 5 3955 14 42 54 5 2414 5 3174 5 4736 5 3955 14 42 55 5 8195 5 2393 5 4736 5 4736 14 42 56 5 3195 5 2393 5 4736 5 3955 14 42 57 5 8195 5 2393 5 4736 5 3955 14 42 58 5 3195 5 2393 5 4736 5 3955 14 42 59 5 3195 5 2393 5 4736 5 3955 14 43 00 5 3195 5 2393 5 4736 5 3955 14 43 01 5 3195 5 3174 5 4736 5 3955 14 43 02 5 2414 5 2393 5 4736 5 4736 14 43 08 5 2414 5 2393 5 4736 5 4736 14 43 04 5 3195 5 2393 5 3955 5 4736 14 43 05 5 3195 5 2393 5 3955 5 4736 14 43 06 5 3195 5 2393 5 3955 5 4736 14 43 07 5 3195 5 2393 5 3955 5 3955 14 43 08 5 3195 5 2393 5 3955 5 3955 14 43 09 5 2414 5 3174 5 3955 5 3955 14 43 10 5 3195 5 2393 5 3955 5 3955 14 43 11 5 2414 5 3174 5 3955 5 3955 14 43 12 5 2414 5 3174 5 3955 5 4736 14 43 18 5 2414 5 3174 5 3955 5 4736 14 43 14 5 2414 5 2393 5 3955 5 3955 14 43 15 5 3195 5 2393 5 3955 5 4736 14 43 16 5 2414 5 3174 5 3955 5 4736 14 43 17 5 2414 5 3174 5 3955 5 4736 14 43 18 5 2414 5 3174 5 3955 5 3955 14 43 19 5 8195 5 2393 5 3955 5 3955 14 43 20 5 3195 5 3174 5 3955 5 4736 14 43 21 5 8195 5 2393 5 3955 5 4736 14 43 22 5 8195 5 2393 5 3955 5 3955 14 43 23 5 2414 5 3174 5 3955 5 4736 14 43 24 5 2414 5
37. 14 44 43 5 2414 5 3174 5 2393 5 3955 14 44 44 5 2414 5 3174 5 2393 5 3174 14 44 45 5 3195 5 3174 5 2393 5 3955 14 44 46 5 2414 5 3955 5 2393 5 3955 14 44 47 5 2414 5 3174 5 2393 5 3174 14 44 48 5 2414 5 3955 5 2393 5 3955 14 44 49 5 2414 5 3174 5 2393 5 3955 14 44 50 5 2414 5 3174 5 2393 5 3955 14 44 51 5 2414 5 3174 5 2393 5 3174 14 44 52 5 2414 5 3174 5 2393 5 3955 14 44 53 5 2414 5 3174 5 2393 5 3174 14 44 54 5 3195 5 3174 5 2393 5 3174 14 44 55 5 2414 5 2393 5 2393 5 3174 14 44 56 5 2414 5 3174 5 2393 5 3955 14 44 57 5 2414 5 3174 5 2393 5 3955 14 44 58 5 2414 5 2393 5 2393 5 3174 14 44 59 5 2414 5 3174 5 3174 5 3174 14 45 00 5 2414 5 3174 5 2393 5 3174 14 45 01 5 2414 5 3174 5 2393 5 3955 37 14 45 02 5 2414 5 3174 5 2393 5 3174 14 45 03 5 2414 5 2393 5 2393 5 3174 14 45 04 5 3195 5 3174 5 2393 5 3174 14 45 05 5 2414 5 3174 5 3174 5 2393 14 45 06 5 2414 5 3174 5 2393 5 2393 14 45 07 5 2414 5 2393 5 2393 5 3955 14 45 08 5 2414 5 3174 5 2393 5 2393 14 45 09 5 3195 5 2393 5 2393 5 2393 14 45 10 5 2414 5 3174 5 2393 5 2393 14 45 11 5 2414 5 2393 5 2393 5 2393 14 45 12 5 2414 5 2393 5 2393 5 2393 14 45 18 5 2414 5 2393 5 2393 5 2393 14 45 14 5 2414 5 3174 5 2393 5 2393 14 45 15 5 2414 5 3174 5 2393 5 2393 14 45 16 5 2414 5 3174 5 2393 5 2393 14 45 17 5 8195 5 2393 5 2393 5 2393 14 45 18 5 2414 5 2393 5 2393 5 2393 14 45 19 5 2414 5 2393 5 2393 5 2393 14 45 20 5 3195 5 2393 5 2393 5 2393 14 45 21 5 2414 5 2393 5 2393 5 2393 14 45 22
38. 2393 14 31 49 5 3195 5 2393 5 2393 5 2393 14 31 50 5 2414 5 2393 5 2393 5 2393 14 31 51 5 3195 5 2393 5 2393 5 2393 14 31 52 5 3195 5 2393 5 2393 5 2393 14 31 53 5 3195 5 2393 5 2393 5 2393 14 31 54 5 3195 5 2393 5 2393 5 2393 14 31 55 5 2414 5 3174 5 2393 5 2393 14 31 56 5 3195 5 2393 5 2393 5 2393 14 31 57 5 2414 5 2393 5 2393 5 2393 14 31 58 5 3195 5 2393 5 2393 5 2393 14 31 59 5 3195 5 2393 5 2393 5 2393 14 32 00 5 2414 5 2393 5 2393 5 2393 14 32 01 5 3195 5 2393 5 2393 5 2393 14 32 02 5 3195 5 2393 5 2393 5 2393 14 32 03 5 2414 5 3174 5 2393 5 2393 14 32 04 5 2414 5 2393 5 2393 5 2393 14 32 05 5 3195 5 2393 5 2393 5 2393 14 32 06 5 2414 5 2393 5 2393 5 2393 14 32 07 5 3195 5 2393 5 2393 5 2393 14 32 08 5 2414 5 2393 5 2393 5 2393 14 32 09 5 3195 5 2393 5 2393 5 2393 14 32 10 5 2414 5 2393 5 2393 5 2393 14 32 11 5 3195 5 2393 5 2393 5 2393 14 32 12 5 3195 5 2393 5 2393 5 2393 14 32 13 5 3195 5 2393 5 2393 5 2393 14 32 14 5 2414 5 2393 5 2393 5 2393 14 32 15 5 3195 5 2393 5 2393 5 2393 14 32 16 5 3195 5 2393 5 2393 5 2393 14 32 17 5 3195 5 2393 5 2393 5 2393 14 32 18 5 2414 5 2393 5 2393 5 2393 14 32 19 5 2414 5 3174 5 2393 5 2393 14 32 20 5 2414 5 2393 5 2393 5 2393 14 32 21 5 3195 5 2393 5 2393 5 2393 14 32 22 5 3195 5 2393 5 1611 5 2393 14 32 23 5 3195 5 2393 5 2393 5 2393 14 32 24 5 3195 5 2393 5 2393 5 2393 14 32 25 5 2414 5 2393 5 1611 5 2393 33 14 32 26 5 3195 5 2393 5 2393 5 2393 14 32 27 5 2414 5 2393 5 1611 5 2393 14
39. 2393 5 2393 14 52 55 5 2414 5 2393 5 2393 5 2393 14 52 56 5 3195 5 2393 5 2393 5 2393 14 52 57 5 2414 5 2393 5 2393 5 2393 14 52 58 5 3195 5 2393 5 2393 5 2393 14 52 59 5 3195 5 2393 5 2393 5 2393 14 53 00 5 3195 5 2393 5 2393 5 2393 14 53 01 5 2414 5 2393 5 2393 5 2393 14 53 02 5 2414 5 2393 5 2393 5 2393 45 14 53 03 5 2414 5 2393 5 2393 5 2393 14 53 04 5 2414 5 2393 5 2393 5 2393 14 53 05 5 2414 5 2393 5 2393 5 2393 14 31 41 5 3195 5 3174 5 2393 5 2393 14 31 42 5 3195 5 2393 5 2393 5 2393 14 31 43 5 3195 5 2393 5 2393 5 2393 14 31 44 5 3195 5 2393 5 2393 5 2393 14 31 45 5 2414 5 2393 5 2393 5 2393 14 31 46 5 3195 5 2393 5 2393 5 2393 14 31 47 5 2414 5 2393 5 2393 5 2393 14 31 48 5 3195 5 2393 5 2393 5 2393 14 31 49 5 3195 5 2393 5 2393 5 2393 14 31 50 5 2414 5 2393 5 2393 5 2393 14 31 51 5 3195 5 2393 5 2393 5 2393 14 31 52 5 3195 5 2393 5 2393 5 2393 14 31 53 5 3195 5 2393 5 2393 5 2393 14 31 54 5 3195 5 2393 5 2393 5 2393 14 31 55 5 2414 5 3174 5 2393 5 2393 14 31 56 5 3195 5 2393 5 2393 5 2393 14 31 57 5 2414 5 2393 5 2393 5 2393 14 31 58 5 3195 5 2393 5 2393 5 2393 14 31 59 5 3195 5 2393 5 2393 5 2393 14 32 00 5 2414 5 2393 5 2393 5 2393 14 32 01 5 3195 5 2393 5 2393 5 2393 14 32 02 5 3195 5 2393 5 2393 5 2393 14 32 03 5 2414 5 3174 5 2393 5 2393 14 32 04 5 2414 5 2393 5 2393 5 2393 14 32 05 5 3195 5 2393 5 2393 5 2393 14 32 06 5 2414 5 2393 5 2393 5 2393 14 32 07 5 3195 5 2393 5 2393 5 2393 14 32 08 5 2414 5 2393 5 2393 5 23
40. 2393 5 3955 5 3955 14 43 25 5 2414 5 3174 5 3955 5 4736 35 14 43 26 5 2414 5 3174 5 3955 5 4736 14 43 27 5 2414 5 2393 5 3955 5 3955 14 43 28 5 2414 5 3174 5 3955 5 4736 14 43 29 5 2414 5 3174 5 3955 5 3955 14 43 30 5 2414 5 3174 5 3955 5 3955 14 43 31 5 2414 5 3174 5 3955 5 3955 14 43 32 5 2414 5 2393 5 3955 5 3955 14 43 33 5 2414 5 3174 5 3955 5 4736 14 43 34 5 2414 5 3174 5 3955 5 4736 14 43 35 5 3195 5 2393 5 3955 5 3955 14 43 36 5 2414 5 3174 5 3955 5 4736 14 43 37 5 2414 5 3174 5 3955 5 3955 14 43 38 5 2414 5 3174 5 3955 5 3955 14 43 39 5 2414 5 3174 5 3955 5 3955 14 43 40 5 2414 5 3174 5 3955 5 3955 14 43 41 5 2414 5 3174 5 3955 5 3955 14 43 42 5 2414 5 3174 5 3955 5 3955 14 43 43 5 3195 5 3955 5 3955 5 4736 14 43 44 5 2414 5 3174 5 3955 5 4736 14 43 45 5 8195 5 3955 5 3955 5 4736 14 43 46 5 2414 5 3174 5 3955 5 4736 14 43 47 5 3195 5 3955 5 3955 5 4736 14 43 48 5 2414 5 3174 5 3955 5 4736 14 43 49 5 3195 5 3174 5 3955 5 3955 14 43 50 5 3195 5 3955 5 3174 5 3955 14 43 51 5 2414 5 3955 5 3955 5 3955 14 43 52 5 2414 5 3174 5 3955 5 3955 14 43 53 5 8195 5 3174 5 3955 5 3955 14 43 54 5 8195 5 3174 5 3174 5 3955 14 43 55 5 2414 5 3174 5 3955 5 3955 14 43 56 5 2414 5 3174 5 3955 5 3955 14 43 57 5 8195 5 3955 5 3955 5 3955 14 43 58 5 3976 5 3174 5 3174 5 3955 14 43 59 5 2414 5 3174 5 3955 5 3955 14 44 00 5 2414 5 3955 5 3955 5 3955 14 44 01 5 3976 5 3174 5 3174 5 3174 14 44 02 5 3195 5 3174 5 3174 5 3955 14 44 03 5 3195 5 3955 5 31
41. 3195 5 31745 5518 5 5518 15 03 12 5 3195 5 3174 5 7080 5 5518 15 03 13 5 3195 5 3174 5 5518 5 5518 15 03 14 5 8195 5 3174 5 6299 5 5518 15 03 15 5 8195 5 3174 5 6299 5 5518 15 03 16 5 3195 5 3955 5 5518 5 5518 15 03 17 5 3195 5 3174 5 6299 5 5518 15 03 18 5 3195 5 3955 5 5518 5 6299 15 03 19 5 3195 5 3174 5 6299 5 5518 15 03 20 5 3195 5 3174 5 5518 5 5518 15 03 21 5 8195 5 3955 5 5518 5 6299 15 03 22 5 8195 5 3955 5 5518 5 5518 15 03 23 5 8195 5 3955 5 5518 5 6299 15 03 24 5 3195 5 3955 5 5518 5 6299 15 03 25 5 8195 5 3174 5 6299 5 5518 15 03 26 5 3195 5 3955 5 5518 5 6299 15 03 27 5 3195 5 3955 5 6299 5 5518 15 03 28 5 3976 5 3955 5 5518 5 5518 15 03 29 5 3195 5 3955 5 5518 5 5518 15 03 30 5 3195 5 3955 5 5518 5 6299 15 03 31 5 8195 5 3955 5 5518 5 5518 15 03 32 5 8195 5 3955 5 5518 5 6299 15 03 33 5 4758 5 4736 5 5518 5 5518 15 03 34 5 3976 5 4736 5 5518 5 5518 15 03 35 5 3976 5 3955 5 5518 5 6299 15 03 36 5 3976 5 3955 5 6299 5 5518 15 03 37 5 3976 5 3955 5 6299 5 5518 15 03 38 5 3976 5 3955 5 6299 5 5518 15 03 39 5 4758 5 4736 5 5518 5 5518 15 03 40 5 3976 5 3955 5 5518 5 5518 15 03 41 5 3976 5 3955 5 5518 5 5518 15 03 42 5 3976 5 3955 5 4736 5 6299 15 03 43 5 3976 5 3955 5 5518 5 6299 15 03 44 5 3976 5 3955 5 6299 5 5518 15 03 45 5 3976 5 3955 5 4736 5 6299 15 03 46 5 3976 5 3955 5 4736 5 6299 15 03 47 5 4758 5 4736 5 5518 5 5518 15 03 48 5 3976 5 3955 5 4736 5 6299 15 03 49 5 3976 5 3955 5 4736 5 6299 15 03 50 5 3976 5 3955 5 4736
42. 39 OHM 2W 5 METAL RES 82 OHM 2W 5 METAL 0 45 225 Linx Technologies HT 640 Hoekencodforsuf mtmans 3 3 9l Radio Shack 3 58 Radio Shack 4 00 Radio Shack 1 99 RadioShack Msc Mages J oo 198 E Switch E Switch 1005P3T6BTIMIQE B way toggle switch 6 sample xxx E Switch ACE Hardware 35 132 4 E co Walmart Misc mini fuse holder 3 94 Walmart Misc Velcro 2 rolls ojo O1 C5 AJ x lt x lt x lt x lt C1 x x Misc Color print charges Tech Library Analog Devices AdxI203 g0 N AR oO 24 94 640 23 EE 89 19 729 42 SUB_TOTAL Oo Shipping handling taxes GRAND TOTAL APPENDIX K Build sheet for transmitter and receiver subsystems along with sensor build sheets TranmiterPartsList ee Se i al WM2300 ND CRIMP TERMINAL 156 fol oo 08 HOOSPGTGBITMIQE Toggle switches 4 est200 8 O SC 544 Receiver Parts List ng pedi RM1V Receiver module 418 MHz E Holtek HT 694 77 a 4 A4027 ND CONN SMB PLUG NICK RG17 A4042 ND CONN SMB JACK VERTICAL WM3204 ND 6 ckt pin header reciever socket A1490 ND 2 CT PRE TIN PCB SOCKET hdr A1448 ND ONN PLUG IN LINE 2 CIRCUIT C 0 17 A1436 ND CONN PIN 24 18 AWG MALE C 077 WMZSOUND CRIMP TERWINAL 158 ta 0 06 Electrolitic cap NES ZJRoTA os 96 8489 16 5 34 4
43. 3955 5 4736 14 50 57 5 3195 5 3955 5 3955 5 5518 14 50 58 5 3195 5 3955 5 3955 5 4736 14 50 59 5 3976 5 3955 5 3955 5 4736 14 51 00 5 3976 5 3955 5 3955 5 4736 14 51 01 5 3195 5 3955 5 3955 5 4736 14 51 02 5 3195 5 3955 5 3955 5 4736 14 51 03 5 3976 5 3955 5 3955 5 4736 14 51 04 5 3976 5 3955 5 3955 5 4736 14 51 05 5 3976 5 3955 5 3955 5 4736 14 51 06 5 3976 5 3955 5 3955 5 4736 14 51 07 5 3976 5 3955 5 3955 5 4736 14 51 08 5 3976 5 3955 5 3955 5 4736 14 51 09 5 3976 5 3955 5 3955 5 4736 14 51 10 5 3976 5 3955 5 3955 5 4736 14 51 11 5 3976 5 3955 5 3955 5 4736 14 51 12 5 8195 5 3955 5 3955 5 4736 14 51 13 5 8976 5 3955 5 3955 5 4736 14 51 14 5 8195 5 3955 5 3955 5 4736 14 51 15 5 8976 5 3955 5 3955 5 4736 14 51 16 5 3195 5 3955 5 3955 5 3955 14 51 17 5 3976 5 3955 5 3955 5 3955 14 51 18 5 3195 5 3955 5 3955 5 4736 14 51 19 5 3195 5 3955 5 3955 5 4736 14 51 20 5 3976 5 3955 5 3955 5 4736 14 51 21 5 3976 5 3955 5 3955 5 4736 14 51 22 5 8976 5 3955 5 3955 5 3955 14 51 23 5 8976 5 3955 5 3955 5 4736 14 51 24 5 8195 5 3955 5 3955 5 4736 14 51 25 5 8195 5 3955 5 3174 5 3955 14 51 26 5 3976 5 3955 5 3955 5 3955 43 14 51 27 5 8195 5 3955 5 3174 5 3955 14 51 28 5 8195 5 3955 5 3955 5 4736 14 51 29 5 8195 5 3955 5 3174 5 3955 14 51 30 5 3195 5 3174 5 3174 5 3955 14 51 31 5 3195 5 3955 5 3955 5 3955 14 51 32 5 3195 5 3955 5 3955 5 3955 14 51 33 5 3195 5 3955 5 3955 5 3955 14 51 34 5 3195 5 3955 5 3955 5 4736 14 51 35 5 3195 5 3955 5 3174 5 39
44. 4 History The history of the project comes from the existence of a similar device The problem was that there was little documentation for the high energy laser power meter system It consists of two circles of trapezoidal plates whose specific heat and thermal conductivity are known It looks similar to the hexagonal form shown in the viewgraphs Figures here except instead of a cone to direct the energy azimuthally around the wedge plates they used a rotating mirror spun at high speed The two surfaces of the hexagonal beam dump form a V facing inward 2 Technical Background The problem to be solved consists of three major details The first was to consider the specifications given in table one and design a power meter prototype that could be used Specifications Plates scaled to accept a 0 25 to 2 0 inch laser beam Plate material edge treatment and thickness designed to accept a 10 watt laser beam for 50 seconds operating at a wavelength of 1 3 microns Plate material and thickness chosen so that the temperature rise is sufficient to measure the incident power to 2 yet low enough so that the material and treatment is not ablated or destroyed so that multiple shots can be measured N 1000 Plate angle and surface treatment to be determined with a design goal to absorb 99 9 of the incident energy Table 1 Project Specifications to measure the incident power propagated by a laser The next was to investi
45. 5 2393 14 32 29 5 8195 5 2393 5 2393 5 2393 14 32 30 5 2414 5 2393 5 1611 5 2393 14 32 31 5 2414 5 2393 5 1611 5 2393 14 32 32 5 2414 5 2393 5 1611 5 2393 14 32 33 5 2414 5 3174 5 2393 5 2393 14 32 34 5 2414 5 2393 5 1611 5 2393 14 32 35 5 2414 5 2393 5 1611 5 2393 14 32 36 5 2414 5 2393 5 1611 5 2393 14 32 37 5 2414 5 2393 5 2393 5 2393 14 32 38 5 2414 5 2393 5 1611 5 2393 14 32 39 5 2414 5 2393 5 1611 5 2393 14 32 40 5 2414 5 2393 5 1611 5 2393 14 32 41 5 2414 5 2393 5 2393 5 2393 14 32 42 5 2414 5 2393 5 2393 5 2393 14 32 43 5 2414 5 2393 5 2393 5 2393 14 32 44 5 2414 5 2393 5 1611 5 2393 14 32 45 5 2414 5 2393 5 2393 5 2393 14 32 46 5 2414 5 2393 5 2393 5 2393 14 32 47 5 2414 5 2393 5 1611 5 2393 14 32 48 5 2414 5 2393 5 1611 5 2393 14 32 49 5 2414 5 2393 5 2393 5 2393 14 32 50 5 2414 5 2393 5 1611 5 2393 14 32 51 5 2414 5 2393 5 1611 5 2393 14 32 52 5 2414 5 2393 5 1611 5 2393 14 32 58 5 2414 5 2393 5 1611 5 2393 14 32 54 5 2414 5 2393 5 1611 5 2393 14 32 55 5 2414 5 2393 5 1611 5 2393 14 32 56 5 1633 5 2393 5 2393 5 2393 14 32 57 5 2414 5 2393 5 2393 5 2393 14 32 58 5 2414 5 2393 5 2393 5 2393 14 32 59 5 2414 5 2393 5 1611 5 2393 40 A 6 Trial 3 Data 14 49 05 5 0851 5 23935 1611 5 1611 14 49 06 5 0851 5 2393 5 1611 5 2393 14 49 07 5 0851 5 1611 5 1611 5 1611 14 49 08 5 0851 5 1611 5 0830 5 2393 14 49 09 5 1633 5 2393 5 1611 5 2393 14 49 10 5 0851 5 2393 5 1611 5 2393 Max Values from rows 1 50 14 49 11 5 1633 5 2393 5 16
46. 5 2414 5 2393 5 1611 5 2393 14 32 49 5 2414 5 2393 5 2393 5 2393 14 32 50 5 2414 5 2393 5 1611 5 2393 14 32 51 5 2414 5 2393 5 1611 5 2393 14 32 52 5 2414 5 2393 5 1611 5 2393 14 32 58 5 2414 5 2393 5 1611 5 2393 14 32 54 5 2414 5 2393 5 1611 5 2393 14 32 55 5 2414 5 2393 5 1611 5 2393 14 32 56 5 1633 5 2393 5 2393 5 2393 14 32 57 5 2414 5 2393 5 2393 5 2393 14 32 58 5 2414 5 2393 5 2393 5 2393 14 32 59 5 2414 5 2393 5 1611 5 2393 33 A 5 Trial 2 Data 14 41 53 5 0070 5 0830 5 0830 5 0049 14 41 53 5 0851 5 0830 5 0830 5 1611 14 41 54 5 0851 5 0830 5 0830 5 0830 14 41 55 5 0851 5 0830 5 0830 5 0830 14 41 56 5 0851 5 16115 1611 5 0830 14 41 57 5 0851 5 1611 5 1611 5 0830 14 41 58 5 0851 5 1611 5 0830 5 0830 14 41 59 5 0851 5 0830 5 0830 5 0830 14 42 00 5 0851 5 1611 5 1611 5 0830 14 42 01 5 0851 5 08305 1611 5 1611 14 42 02 5 0851 5 0830 5 0830 5 0830 14 42 03 5 0851 5 0830 5 1611 5 0830 14 42 04 5 0851 5 0830 5 1611 5 0830 14 42 05 5 0851 5 16115 16115 1611 14 42 06 5 0851 5 0830 5 2393 5 1611 14 42 07 5 0851 5 0830 5 2393 5 1611 14 42 08 5 0851 5 0830 5 2393 5 2393 14 42 09 5 0851 5 0830 5 2393 5 2393 14 42 10 5 0851 5 0830 5 2393 5 2393 14 42 11 5 0851 5 16115 16115 1611 14 42 12 5 0851 5 0830 5 2393 5 2393 14 42 13 5 0851 5 1611 5 2393 5 2393 14 42 14 5 0851 5 0830 5 2393 5 2393 14 42 155 0851 5 16115 3174 5 1611 14 42 16 5 0851 5 0830 5 2393 5 1611 14 42 17 5 0851 5 0830 5 2393 5 2393 14 42 18 5 0851 5 0830 5 2393 5 2393 14 42
47. 5 3174 5 6299 5 5518 14 28 53 5 8195 5 3174 5 6299 5 5518 14 28 54 5 3195 5 3174 5 5518 5 4736 14 28 55 5 8195 5 3174 5 5518 5 4736 14 28 56 5 3976 5 3174 5 4736 5 4736 14 28 57 5 8195 5 3174 5 5518 5 4736 14 28 58 5 8195 5 3174 5 6299 5 5518 14 28 59 5 3195 5 3174 5 5518 5 4736 14 29 00 5 3195 5 3174 5 5518 5 4736 14 29 01 5 3195 5 3174 5 4736 5 4736 14 29 02 5 3195 5 3174 5 5518 5 4736 14 29 03 5 3195 5 3174 5 4736 5 4736 14 29 04 5 3195 5 3174 5 5518 5 4736 14 29 05 5 3195 5 3174 5 4736 5 5518 14 29 06 5 3195 5 3174 5 4736 5 4736 14 29 07 5 3195 5 3174 5 4736 5 4736 14 29 08 5 3195 5 3174 5 4736 5 4736 14 29 09 5 3195 5 3174 5 4736 5 4736 14 29 10 5 3195 5 3174 5 4736 5 5518 14 29 11 5 3195 5 3174 5 4736 5 4736 14 29 12 5 8195 5 3174 5 5518 5 4736 14 29 18 5 8195 5 3174 5 4736 5 5518 28 14 29 14 5 3195 5 3174 5 4736 5 4736 14 29 15 5 8195 5 3955 5 4736 5 4736 14 29 16 5 3195 5 3955 5 4736 5 4736 14 29 17 5 3195 5 3955 5 4736 5 4736 14 29 18 5 8195 5 3955 5 3955 5 4736 14 29 19 5 3195 5 3955 5 3955 5 4736 14 29 20 5 3195 5 3955 5 4736 5 4736 14 29 21 5 3195 5 3955 5 3955 5 4736 14 29 22 5 3195 5 3955 5 3955 5 4736 14 29 23 5 3195 5 3955 5 3955 5 4736 14 29 24 5 3195 5 3955 5 3955 5 4736 14 29 25 5 8195 5 3955 5 3955 5 4736 14 29 26 5 3195 5 3174 5 3955 5 4736 14 29 27 5 3195 5 3955 5 3955 5 4736 14 29 28 5 3195 5 3955 5 3955 5 4736 14 29 29 5 3195 5 3955 5 3955 5 3955 14 29 30 5 3195 5 3955 5 3955 5 4736 14 29 31 5 3195 5 3955 5
48. 5 5518 15 03 51 5 3976 5 3955 5 4736 5 4736 15 03 52 5 4758 5 4736 5 4736 5 5518 49 15 03 53 5 3976 5 3955 5 4736 5 4736 15 03 54 5 3976 5 3955 5 4736 5 5518 15 03 55 5 8976 5 3955 5 5518 5 5518 15 03 56 5 3976 5 3955 5 4736 5 4736 15 03 57 5 3976 5 3955 5 4736 5 4736 15 03 58 5 3976 5 3955 5 4736 5 4736 15 03 59 5 3976 5 3955 5 4736 5 4736 15 04 00 5 3976 5 3955 5 5518 5 4736 15 04 01 5 3976 5 3955 5 4736 5 4736 15 04 02 5 3976 5 3955 5 4736 5 4736 15 04 03 5 3976 5 3955 5 4736 5 4736 15 04 04 5 3976 5 3955 5 5518 5 5518 15 04 05 5 3976 5 3955 5 4736 5 5518 15 04 06 5 3976 5 3955 5 4736 5 5518 15 04 07 5 3976 5 3955 5 4736 5 4736 15 04 08 5 3976 5 3955 5 3955 5 4736 15 04 09 5 3976 5 3955 5 3955 5 4736 15 04 10 5 3976 5 4736 5 5518 5 5518 15 04 11 5 4758 5 4736 5 4736 5 4736 15 04 12 5 4758 5 4736 5 4736 5 4736 15 04 18 5 4758 5 4736 5 4736 5 5518 15 04 14 5 3976 5 3955 5 3955 5 4736 15 04 15 5 4758 5 4736 5 4736 5 4736 15 04 16 5 3976 5 3955 5 3955 5 4736 15 04 17 5 3976 5 3955 5 3955 5 4736 15 04 18 5 4758 5 4736 5 4736 5 5518 15 04 19 5 4758 5 4736 5 4736 5 5518 15 04 20 5 4758 5 4736 5 4736 5 5518 15 04 21 5 3976 5 4736 5 4736 5 5518 15 04 22 5 4758 5 4736 5 4736 5 4736 15 04 23 5 3976 5 3955 5 3955 5 4736 15 04 24 5 4758 5 4736 5 4736 5 4736 15 04 25 5 3976 5 3955 5 3955 5 4736 15 04 26 5 4758 5 4736 5 4736 5 5518 15 04 27 5 3976 5 4736 5 4736 5 5518 15 04 28 5 3976 5 3955 5 3955 5 3955 15 04 29 5 3976 5 3955 5 3955 5 3955 1
49. Bitsoi Lamb Willems the value from the data bus and store it in an internal register Once our address disappears from the address bus we let go of the data bus This process is depicted below in Figure 21 databus addrbus AEN BALE IOCS16 IOR IOW Figure 21 Timing Diagram for an 8 bit Read amp Write on the PC 104 Bus 46 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Appendix B Altera Code pci04inter tdf Interface to PC 104 Bus for Senior Design RT Linux On A Small Mobile Robot Written for Diamond Systems Prometheus SBC Bill Willems 1 3 02 Written for 8 bit interface Bill Willems 1 13 02 updated for 16 bit interface made PWM regs R W Constants CONSTANT OUR_ADDRESS B 110100 We can use 0x340 0x34F CONSTANT CTRL R POS B 0000 Position Accumulators need CONSTANT CTRL L POS B 0010 16 bit transfers CONSTANT CTRL R PWM B 0100 PWM Registers only require CONSTANT CTRL L PWM B 0101 only 8 bit transfers SUBDESIGN pcl04inter 9 E NOTE n indicates line is active LOW 9 PC 104 Bus specific input lines RESET E INPUT addr 9 0 INPUT address bus there are 20 lines but we can get away with 10 IOW n IOR n INPUT I O R W lines AEN INPUT Address Enable used with DMA BALE INPUT Address Latch Enable when is the address valid Robot specific input lines R_pos 15 0
50. City UT were contacted regarding Eneco s thermal diodes which are solid state Thermoionic devices for converting heat to electricity These devices are still in the development stages and were not available A third consideration was the use of lead telluride TE generation modules These produce more power and continue to work at a higher temperature than the bismuth telluride modules Dan Allen at Hi Z Corp San Diego CA was contacted about the possibility of Hi Z lending or selling at a reduced cost one of their power generation modules HI Z declined but did send technical information on their products The requirement of using the heat from the fire to power the device was not met In place of the TE device a 5V Lithium battery was selected to supply the required power This is a commercially available power source that has a shelf life of 30 years A thermal mechanical switch was connected to a relay to activate the entire device Several small bimetallic switches were ordered from Allelectronics Corp Van Nyes CA for integration and testing These close at 212 F 100 C When 14 imbedded under a small amount of insulation the switches close and activate the unit at the activation temperature of 250 F 3 2 Insulation Material Performed by Summer Rhodes Research into standard thermal testing was completed so that thermal properties of materials could be determined This could be used to determine these properties for material
51. E field and mercury switch data The second plot shows the phase of the E field versus the mercury switch The relevance of this plot is that it shows if there is a switch in the direction of the E field If the phase shift changes the E field direction changes accordingly If the phase shift remains constant the direction of the E field remains constant This graph is produced by doing the quadrature demodulation of E field data with the Hg switch data 26 E field in blue Hg in red 2000 1950 1900 1850 1800 1750 1700 1650 1600 250 300 350 400 ES 450 500 550 600 time in seconds hase of E vs h 100 E d PSPSANN IW NV 100 L 250 300 350 ADU time in seconds 1 450 500 550 600 Figure 13 Angle of Hg switch to horizon Third plot of finished product We also use quadrature demodulation to determine the position of the instrument relative to the B field Using the Hilbert Transform the Hg switch data is converted into a sine cosine pair Hilbert Acos wt gt Acos ot iAsin t Where the real part of the Hilbert transform is the original sinusoid and the imaginary part is the sinusoid shifted by 90 degrees The two parts of this pair will be individually multiplied by the B field data Bcos ot Acos t Bcos oxt Asin t 27 This results in two separate sinusoids as follows YAB cos 2at cos V2AB sin 20t sin o The results will be adjusted by
52. Item Water Pump Battery Solar Panel Shipping Plexiglas Potting Soil Velcro Silicon Super Glue 1 8 Hose Custom Built HC 12 Analog Components Figure 14 Self Watering Planter Budget Cost 3 75 15 00 19 00 20 00 12 00 5 00 1 20 5 00 3 00 1 00 70 00 Donated Sampled Total 154 95 44 Ni Cad Ni MH SLA Li Ion Li Polymer Reuse Alkaline 40 60 60 80 30 100 150 200 80 Energy Density Wh L Cell Voltage 1 2V 1 2V 2V 3 7V 2 7V 1 5V V Typical 2 00 to 2 00 to 4 50 to 1 00 to 1 00 to 3 00 to Battery Cost 11 50 12 50 35 00 9 00 9 00 10 00 Cycle Life 1500 500 200 500 500 1000 150 200 80 charge amp discharge Table 1 The various types of batteries available and their performance cost characteristics device Volumetric energy density Wh L 0 250 Bo po kapacitors 0 05 5 105 10 105 10 Table 2 The difference between batteries and capacitors Power density W L number of charge discharge cycles Discharge time s gt 1000 45 Appendix A Program Code include hc12 h include dbug12 h Battery amp solar panel voltages are attenuated by 5 define TRUE 1 define V_BATT_HIGH 160 Battery volt conversion const of 3 15V out of 3 0V define V_BATT_MED 153 Battery volt conversion const of 3 0V out of 3 0V define V_BATT_LOW 140 Battery volt
53. Lm data_in3 data_in2 data_int 2 e TxRx_clock Figure 10 Pin Out of JP11 Header Figure 11 Interface I O on 7 Segment LED s The Altera chip can be programmed either through the JTAG interface or through the configuration controller on the board Since the chip is volatile we would have to program every time we powered up the chip if we were to use the JTAG interface The way that the configuration controller works however is that it loads a configuration from the flash memory on the board To place our configuration in the flash memory we had to do several things First we set up our Altera project so that it wrote the configuration in a hexadecimal format We then used a program called hexout2flash developed by Altera to convert the hexadecimal configuration file into a file that can be downloaded into the flash memory Finally with a jumper on JP2 we connected our development computer to the Altera board through the serial port and used a program called nios run to download our configuration into the flash memory For more information on downloading configurations into the flash memory refer to the NIOS Embedded Processor Software Development Reference Manual from Altera 17 CHAPTER 4 TEST SOFTWARE DESIGN Thomee Wright 4 1 INTRODUCTION This section details the development of the PC side of the hardware and software First the available pins on a PC parallel port and their usage in this project is given Second
54. a low pass filter to remove the high frequencies incurred during this process This will leave two values V2ABcos Q YABsin o The two outputs will be divided by each other so the amplitudes will cancel out and only tangent of the phase shift will remain YABsin o V2ABcos Q Taking the arctangent of this result will reveal the phase difference From this we know the degree of rotation between the instrument and the B field This information can be adjusted for minor deviations by shifting the Hg switch by a small phase amount Figure 14 shows the raw B field data and the raw Hg switch data and a graph of the phase of the B field versus the mercury switch This data is found using the above quadrature demodulation technique This shows the angle of the instrument from the horizontal of B This should be a sine wave of relatively constant frequency 28 B field in blue Hg in red T T phase of B vs hg 80 90 100 110 120 130 Mg m 1 1 l 50 300 350 400 ime SOON Q4 450 500 550 600 Figure 14 Orientation of instrument Fourth plot of finished product Figures 5 8 show the same things as graphs 1 4 respectively but show the interpolated smoothed data in the top plots instead of the raw data This is relevant because the filtering takes place on the interpolated data If the resulting vertical and horizontal components of the E field or the phase shifts between E B and hg are not what it a
55. a preferred developmental platform for testing and loading code With the OOPIC one must use Windows 98 as the developmental and launch platform The overall theme of our project is a low power design and this product draws 1 3 the power of the OOPIC This is crucial because any power that can be saved does not have to be created form the fire Finally the BrainStem has the backing of full customer support from the manufacturer The OOPIC is a side project of a very small company and does not provide any customer support 3 4 2 Data acquisition transmission and reception The overall data acquisition system is broken down into two parts the transmitter and the receiver The transmitter takes the data from the temperature 27 sensor and them transmits a raw A D value to the receiver In conjunction to transmitting the A D value the transmitter will send a value that indicates the type of ordnance that is on fire This is done by using the data space that corresponds to the temperatures below the activation temperature The transmitter also sends a 2 bit number to the receiver that indicates the next frequency of transmission The receiver obtains the value from the transmitter converts it to Celsius and compares it to the previous value these values are displayed to a console window In addition to displaying the temperatures the receiver will indicate what type of ordnance is on fire every 5th sample and indicate if the external tempe
56. a set of test vectors to execute that will convince him that the design is correct DESIGN OVERVIEW To verify the RDPP we decided to test only a single Processing Element There were two factors involved in making this decision First earlier work done here at Tech on the RDPP showed that a single Processing Element takes up 5796 of the APEX20K s resources thus the entire RDPP cannot be emulated and verified in this chip Second all data manipulation occurs in the PE Because of this it is critical to test the PE for functional testing of the RDPP while the other components are much less complex and easier to verify later We decided to develop an interface between our computer and the PE on the Altera board This interface was used to send and receive data between the PE and PC We also decided to develop software that would dynamically generate test inputs to execute on the PE as well as their expected results This software used the interface to the Altera amp board to send and execute the test vectors on the PE and receive the results The software also compared the actual results with the expected results and reported if any errors were detected 10 3 1 32 3 3 Chapter 3 Altera Hardware Design Steven Wasson INTRODUCTION This section covers all the Altera hardware design necessary for this project This includes the synthesis of the Processing Element as well as the design and implementation of the interface SYNTHES
57. and were familiar with it was assigned to work on the transmitter board layout Once we each had our section laid out in Protel we then incorporated our designs into one package which was then sent out for etching After receiving our board from Advanced Circuits mounted and soldered all necessary connectors resistors diodes and chips to the board Next Brandon and preformed testing on our individual boards and then we tested them together as a pair The fourth part that was assigned to was sensor development Each of us was required to choose several sensors for the turn signal or the brake pedal decided to use a push button for the brake pedal It was a simple design but it required that a housing be built for the sensor to prevent it from being damaged used a square piece of Plexiglas to house the sensor This allowed the sensor to be easily activated without the pressure of the foot damaging it or the wiring used Velcro to attach the sensor to the brake pedal For the turn signal decided to use a phototransistor to detect the presents of light from the dashboard The phototransistor does not give off enough voltage to activate our transmitter circuit so added an amplifier from Maxim to increase the voltage The mounting of this sensor only required a piece of double stick tape Chapter 5 Conclusion The following chapter is broken down into four sections that include the project highlights budget summary fin
58. and are able to modify an image to show only the vertical and horizontal lines We can also use the Prometheus SBC to capture images from the USB cameras which can then be used by our image processing algorithms As far as behavior control goes we have programmed the robot to left and right wall follow as well as respond to front sensor input Code has been written to interface to and check for updates on the rest of the sensors as well We were unable to implement Image Processing into the functionality of the robot due to time and work constraints 42 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems References PC 104 Embedded Consortium PC 104 Specification Version 2 4 August 2001 http www pc104 org Messmer Hans Peter The Indispensable PC Hardware Book 3rd Edition Addison Wesley 1997 Diamond Systems Corporation PROMETHEUS High Integration PC 104 CPU with Ethernet and Data Acquisition Models PR Z32 E ST PR Z32 EA ST User Manual V1 35 http www diamondsystems com Barabanov Michael Getting Started with RTLinux Finite State Machine Labs Inc September 26 2001 Linux Online Inc What Is Linux March 2002 http www linuxonline com Umbaugh Scott E Computer Vision and Image Processing Prentice Hall 1998 43 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Appendix A Our Interface to the PC 104 Bus The PC 104 Interface is electrically identical to the ISA bus found on the c
59. beginning to end It illustrates the steps taken when approaching the problem the choices made while seeking a solution and the results from testing the final design This document also provides contributions from the individual group members To expedite the progress of the project it was divided up into equal parts among the three group members The combination of their time and efforts results in the device that this document outlines and describes Acknowledgements First we would like to thank Dr Scott Teare of New Mexico Tech Electrical Engineering Department who was our class instructor and faculty advisor Dr Teare did a great job advising us Dr Daniel Eastman of Boeing SVS in Albuquerque New Mexico for presenting this project to the New Mexico Tech Electrical Engineering Department And also for supplying the laser and facilities used to test our project Dr Deidre Hirschfeld of the New Mexico Tech Materials Department for directing us in the proper path on the materials portion of this project Norton Euart of the use of the machine shop in Workman Center Without use of the equipment we would have not been able to construct the laser beam dump And also for supplying us with the components needed in this project Table of Contents Abstract Acknowledgements Table of Contents List of Tables List of Figures 1 Laser Beam Dump and Power Meter Introduction 1 1 Project Description 1 2 Definitions 1 3 Literature Review
60. board and just leave space so that we could cut them in half During the PCB process we assumed that once we had the PCB layout of each of systems complete we could just copy one into the other That was not the case The whole system has to be complete in the schematic to update the PCB so we wasted quite a bit of time not knowing this We put the whole system in the schematic and updated the PCB to find over 135 errors Refer to Appendix E for the PCB layout This obviously took some time to debug Most of the errors involved resistors and capacitors being named the same and the wrong footprints for the parts Those were relatively easy problems to solve The hardest problems were ones involving footprints that we had to make for the part that were not available in Protel For those parts we had to make our own footprint that matched the real part exactly After all this was done and all the errors were fixed we printed out the lay out to make sure the holes for all the parts matched Everything looked successful so we continued on with the PCB process and sent the files to advanced circuits for manufacturing The printed circuit board is shown in Figure 4 Figure 4 Printed circuit board Upon receiving the board we found errors immediately Some of the hole sizes for the parts were too small The reason for us not seeing the problem from the printed layout was that it was just too close to see with the eye The resistor holes were tigh
61. board from also provided methods for interfacing the A D board with commonly used software packages such as Microsoft Excel Excel By downloading and installing a program reference and code from the vendor that enabled Excel to receive data from the computer s serial port the collection 14 and calculation of the data was as easy as using Excel to perform mathematical calculations 32 Ernest Jim Ernest was in charge of constructing the laser beam dump This involved determining what type of material the beam dump would be made from and what shape it would have to best fit the project needs The material was determined by the amount of energy that the beam dump needed to absorb over time This gave the information needed to determine the mass length thickness and width of the beam dump wedges Once the dimensions were calculated the plates were cut bent drilled and assembled in the machine shop To increase absorption of the beam dump the inside of the wedge plates was coated with black paint The wedge plates were long so the angle between the plates was small This enabled the beam dump to keep less energy from escaping Calculations for the wedge were made for several different types of materials copper alumina and steel 3 3 Daniel Rodriguez Daniel was responsible for designing and developing the heat sensing hardware system of the project First Daniel performed a literature search on the commercially available temperature se
62. clusters configuration operation maintenance User s Manual as well as weekly and formal reports Within this area I also assisted in the creation of material used for the presentations throughout the year 20 4 Summary and Recommendations This project has produced a functional Beowulf cluster for Dr Teare s use A user s manual and series of scripts have also been produced The scripts produced automate the configuration of the systems in the cluster and to configure the parallel libraries and ZPL for the user These features lower the entry level to expanding the provided cluster or to building a new cluster with a similar configuration As a result of the benchmarks performed on the finished Beowulf cluster the following recommendations are provided for areas of future improvement in the cluster configuration Memory improvements The most beneficial upgrade that could be made to the cluster at this point is adding more RAM to the cluster nodes Each node is currently equipped with 64MB of RAM and a 256MB swap partition When running the 100 inch mirror coating simulation S 165 resulting in a 331x331 grid mapped to the mirror surface the total memory in the nodes was over 90 used according to the top program on the involved nodes Adding more RAM to the nodes in the cluster will increase the limit on the maximum sized simulations that can be run If the operating systems on the node computers are ever reinstalle
63. complex because it also included the output logic necessary to control our LED arrays and the relays The main components in the receiver board are the receiver decoder programmable logic chip PAL 556 timers Voltage regulator BJT power transistors and relay switches Refer to Appendix D for a schematic of the receiver system On the receiver board we were required to use a battery for our main power source The battery had to have a long life and it also had to be rechargeable The battery selection was done after we completed the board The receiver board used so little power that it allowed us to use two small Ni Cad batteries in series to get the 14 volts necessary to power our circuit The receiver board was done in several steps First we built and tested the receiver and decoder This was done simultaneously with the transmitter side so that testing on interference and distance could be checked The entire receiver circuit is shown in Figure 3 Figure 3 Entire receiver circuit 2 4 Output Logic After we were confident that the receiver was working correctly we started the design of the output logic The output logic was necessary so that we could control not only the intensity of the LED s but also the hierarchy of each of the sensor inputs Our first approach to the output logic was successful However the first approach used individual IC s to accomplish this task We felt the individual IC s took up too much room on our
64. customer s requirements for the project The rest of the year is spent working to meet customer demands In addition to work for the customer regular presentations and progress reports are required This work culminates in a presentation of group achievements to Electrical Engineering faculty and the department s corporate advisory board as well as written documentation of the project Grades are based on customer satisfaction regular written and oral communication and overall team performance 1 2 The National Radio Astronomy Observatory Among the customers offering projects was the National Radio Astronomy Observatory NRAO The NRAO is a publicly funded research facility operated by Associated Universities Incorporated AUT for the National Science Foundation NSF Based in Charlottesville Virginia NRAO operates a number of world class radio telescopes including the Very Large Array VLA located west of Socorro New Mexico Composed of 27 antennas in the San Augustin plains the VLA is a one of a kind instrument used by scientists from around the world to conduct cutting edge research The VLA has a visitors center which NRAO is currently in the process of upgrading One of the plans in the expansion is to put a small single dish radio telescope in the Visitor Center where the public can access it The radio telescope would be one of the first stops on the self guided tour and would educate people about radio astronomy so that th
65. dirt Vv vs R 140000 120000 100000 80000 60000 40000 20000 Fig 12 Sensor Test for wet dirt Nr Vi R R 100000 wet dirt Vv vs l 0 0002 0 00018 0 00016 0 00014 0 00012 ONDA ER WN 1 46 10 0 0001 14600 0 0001 1 5 11 0 00011 13636 36 aiis 1 54 12 0 00012 12833 33 1 6 13 0 00013 12307 69 0 00006 1 64 14 0 00014 11714 29 T 1 58 15 0 00015 11200 1 72 15 0 00016 10750 0 00002 1 75 17 0 00017 10294 12 5 9888 889 0 02 04 06 08 1 12 14 16 18 2w 9473 684 wet dirt I vs R R 0 00002 0 00004 0 00006 0 00008 0 0001 0 00012 0 00014 0 00016 0 00018 0 0002 wet dirt Vv vs R 45000 35000 30000 20000 15000 10000 o Fig 13 Final Test sample resistance ohms Volts dry 1mA 28000 0 0001 2 8 moist 1mA 14250 0 0001 1 425 water 1mA 7500 0 0001 0 75 dry 14mA 22000 0 00014 3 08 moist 14mA 11500 0 00014 1 61 water 14mA 7000 0 00014 0 98 dry 18mA 20000 0 00018 3 6 moist 18mA 10000 0 00018 1 6 water 18mA _ 7250 0 00018 1 305 resistance ohms 30000 28000 El resistance ohms 25000 20000 15000 10000 5000 0d T T T water 14mA dry 18mA moist 18mA water 18mA dry 1mA moist 1mA water ImA dry 14mA moist 14mA
66. e mail jhart 9 nmt edu Seeking permanent employment in the area of electrical engineering B S Electrical Engineering New Mexico Institute of Mining and Expected Date of Graduation May 2002 GPA 3 14 4 0 after 118 graded hours Summer Intern Utilities and Energy Management Team UEMT United States Department of Energy Albuquerque New Mexico 2001 Provided clerical support for the UEMT Assisted studying of legal briefs Reviewed and commented on proprietary solicitation proposals Attended and participated in meetings with the Public Service Company of New Mexico Public Regulation Committee of New Mexico and Los Alamos County Power Pool Quality checked and assured construction drawings for accuracy in estimated material needs L security clearance being processed Summer Intern Environmental Analysis Intelligent Systems and RoboticsCenter Sandia National Laboratories Albuquerque New Mexico 1999 2000 Compiled soil sample data Analyzed maps for metal concentrations Generated data for nuclear fall out absorption Marked sample locations using a GPS receiver Entered chemical data for website database Collected and analyzed data using MATLAB for a proprietary project Lab Assistant amp Grader New Mexico Tech Digital Lab Socorro New Mexico January 1998 present Maintained lab equipment Entered data Monitored lab activity Prepared materials for lab classes Graded papers and recorded grades for courses previou
67. final PC board design After completely testing the output logic on the receiver end we decided it would be more beneficial to look into using a programmable chip for that logic Since an Altera chip is easily accessible to us we decided to implement our design on the Altera chip Although the design worked flawlessly we felt that the Altera chip was a little overkill for our specific design This forced us to do further research on other possible programmable chips that would not only better suit our design but also be more cost effective We got a hold of an old style programmer from BP Micro and several programmable chips called PAL s from a professor and looked into getting it to work That involved downloading some software for the programmer from BP Micro and also compiling software called PALASM All of this was new to the group so we spent time learning the new software and how to implement our logic in the form needed to compile After weeks of struggling we were finally able to successfully program a PAL AMD348217 chip The use of this chip allows for future expansion as one just needs to reprogram the chip the way they want it We added a dip switch to the unused inputs so that they could be used for more combinations in the future Adding the AMD348217 simplified the entire system greatly as it reduced the total chip count by four chips To implement the flashing of the LED arrays for left and right turn purposes and for continuous li
68. for porting to the ZPL programming language and executing on the finished Beowulf cluster Finally we would like to thank everyone who has contributed to the Linux Documentation Project The background information HOWTOs and troubleshooting guided contained in that archive were the foundation for the work that was accomplished in this project iii Table of Contents POS UR AG D s atestosapectiolsesc dai otestiqas dii T oia citt d aes ii JxckroWIbde ments de dos UE istud ofu d qtcatude Aorta ated ade Med iii Table of Contents E aes aneia EA es eae eta R aa iv Eist of Figure S en anre e aa aa Eo E EESE S v 1 Project Backeroundcana Goals arrier E E E ER i 1 lL Project Histor S ot ca scncneceted tot cactadenmcaeesiunccan A seu stances uc M tes au ptu deris 1 1 2 Project Description ceren ces mie Ra RUN EUER ESI QA dann duced Ea EEEo ia oE AOE ATERS 1 1 3 De TOMS i ood tte seinien eaa RAA SEE torpe ctas LOSS eet 2 2 Technology Review and Design Decisions eese enne enne 3 2 Eocali vs Network BOOt r cadi E aiii TE E EEEE 3 2 2 Overview of the NAS MG benchmark used to evaluate the cluster 5 2 3 10Base T vs 100Base T INGUWOEKS n na ho re te ph Torte deti meet 6 2 4 Overview of the ZPL Parallel Programming Language eene 10 2 5 MPI vs PVM Parallel Libraries intet tert tete reat ot he e te i ee ease o desee 11 3 Team Member Contributions and Finished Product
69. if it is ready Clint 45 Real Time Linux On A Small Mobile Robot A Thesis Presented to the Faculty of the Electrical Engineering Department of New Mexico Tech By Heather Bitsoi Kyle Lamb Bill Willems In partial fulfillment of the requirements for the course EE 482 Senior Design Project Il May 2002 2002 Bitsoi Lamb Willems RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Table of Contents Section Page List of Figures 2 List of Tables 3 List of Abbreviations 4 Introduction 5 Customer Requirements 6 Background 7 Red Hat Linux 7 Real Time Linux 8 RT FIFOs 10 Prometheus SBC 10 Breakout Board 11 Image Processing 12 Discussion 15 Image Processing 15 Image Processing Algorithm 15 Breakout Board Wiring 20 Behavior Control 23 Ultrasonic Distance Sensor 24 Bump Sensors 25 Hardware Connection 26 Altera Programming for Breakout Board 27 Altera Code for PWM Generation 28 Altera Code for 8 bit PC 104 Interface 29 Altera Code for 16 bit PC 104 Interface 30 Installation of RTLinux 30 Breakout Board Driver 31 Hindsight is 20 20 38 Breakout Board 38 Future Work 39 Breakout Board and Driver 39 Image Processing 40 Conclusion 42 References 43 Appendix A Our Interface to the PC 104 Bus 44 Appendix B Altera Code 47 Appendix C Breakout Board Driver 53 Appendix D Breakout Board Schematic 72 Appendix E Features of the Prometheus 73 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems List of F
70. if the product ever reached full scale production With the sensors that the team built and the parts selected the final product is under the 200 initial prototype budget that the requirements specify The transmitter receiver and sensors run about 160 dollars and a set of LED arrays would run about 40 50 to buy or manufacture The following figure shows the prototype cost of the entire system Appendix K contains tables which state the parts purchased along with tables that lay out the costs associated in building the individual assemblies and sensors CATEGORY COST Transmitter Assembly 54 Receiver Assembly 85 IR distance sensor 15 Momentary contact lever switch 4 5 Hall effect sensor 3 5 Push Button Sensor 3 Phototransistor sensor 2 5 TOTAL System Cost 146 160 Fig 10 Total prototype system cost final cost is the sum of the transmitter and receiver along with one brake and one turn signal sensor 5 3 Final Sensor Functionally The five sensors that we took to the final prototype status have worked well in the limited time we have had to test them The input sensors also tested very well in a variety of vehicle styles years and makes The sensors were tested on two full size Ford trucks along with a full size Chevy and Dodge truck The only issue that may be a problem is adapting the sensors to vehicles that were not a part of the test cases and thus would require modifications to
71. in valid zpl settings zpl target zpl commlayer print ZPL can use the following libraries on s zpl target for commlayer in valid zpl settings zpl target print t s commlayer print print NOTE the default on most platforms is seq sequential print but you will almost certainly want to choose one of the print parallel libraries such as mpi or pvm in order to make print use of ZPL s parallel execution print i print Which parallel library should ZPL use print default s valid zpl settings zpl target 0 zpl commlayer sys stdin readline 1 print if len zpl commlayer 0 zpl commlayer valid zpl settings zpl target 0 return zpl commlayer 4 end of zpl setup zplcommlayer def zpl setup print sce 5e 2es 9xee5 25 2gc wewemccemIll ocmoce geces em print Configuring the ZPL parallel programming language print MsSsser esse sere eSe SH soG ree es Sees sore Se E S zpl home zpl setup zplhome zpl target zpl setup zpltarget zpl commlayer zpl setup zplcommlayer zpl target zpl bin os path join zpl home bin zpl target print ZPLHOME S zpl home print ZPLTARGET s zpl target print ZPLCOMMLAYER sS zpl commlayer zplcommlayer sh file open os path join user parallel setting dir zplcommlayer sh w zplcommlayer sh file write export ZPLCOMMLAYER s n zpl commlayer zplcommlayer sh file close 44 zpl
72. interval The next slide button on figure 3 3 4a will change slides and load the layout on figure 3 3 4b This layout will not have the automatic feature The user will only have access to the next slide button At any time when a slideshow is selected the user will have the ability to exit the slideshow The exit buttons on figure 3 3 4a and figure 3 3 4b will exit the slideshow and send the user back to the layout on figure 3 3 3 The exit button must be selected to go back to the layout on figure 3 3 3 if next slide is selected The automatic button will load the layout on figure 3 3 3 after the slideshow is completed 22 Figure 3 3 3 Slideshow Presentation Layout Figure 3 3 4a Figure 3 3 4b The VLA weather button will load a NRAO weather program to display on the monitor and the layout on figure 3 3 5 The weather program will tell people what the current weather conditions are at the VLA site The back to main button will exit the weather program and send the user back to the main layout The VLA weather program looks like the layout on the touchscreen Data is received from the VLA control center via the network 23 The VLA Weather Station Figure 3 3 5 The VLA Weather Station Layout When the web links icons are selected on Figure 3 3 3 the layout will change layouts to figure 3 3 6 The user will have access to three NRAO web pages The FAQ web page has answers to common questions that people have about the VLA T
73. is selected two dos prompts are needed because two programs are used One prompt will run the tracking software and the other will run the PowerPoint presentations At the present time mouse positioning is used to change dos prompts This is not the best way to do this because over time the mouse can get out of position A Visual C program could be created which will open both programs using one prompt This will eliminate the need to position the mouse creating a cleaner running program A copy of Visual C would need to be installed on the computer and we would need to learn how to use it to accomplish this task More radio sources could be made available for selection including terrestrial sources and geosynchronous satellites This would be a simple addition to the program requiring minimal changes to code An interesting Senior design project would be to upgrade the SRT controller Possible improvements would be to add a better microcontroller and a DSP chip to perform signal processing that the SRT program is not capable of More information about the VLA could be made available in the SRT display like incorporating VLA observation schedules The SRT project slideshow could also be made to run along with the educational slide shows Another possible addition is having a Web cam available to allow access from the web Its website could be configured to control the SRT from a remote location Our project has inspired the NMT Phys
74. max sunlight would hit it during lighted hours Fig T 3 3 4 Charge Circuit The solar panel could not be connected directly to the battery for charging due to over and undercharging We needed to implement some type of charge circuitry At first we looked into various charge circuits and methods from our EE 322 Advanced Electronics book The Art of Electronics We thought we could use an LM 317 however after testing it we found it to be inapplicable due to the fact that there needs to be at least a 3V drop across the solar panel to the battery We researched various designs and received many ideas for charge circuits We tested several designs but could never get any of them to properly work After weeks of research and still no answer to charge circuitry I went to consult with Dr Ron Thomas He gave several leads on charge circuitry one of which included the use of mosfets Finally we came up with a design that theoretically worked Fig 3 Once we had the design we tested it We found charging success within the circuit It consists of an N channel and a P channel FET The gate of the N channel receives a pulse width modulated signal PWM The source is grounded and the drain is tied to a 100k resistor and the solar panel input voltage The gate of the P channel FET is tied to the drain of the N channel FET and the source is tied to the solar panel input voltage The drain is then tied to the battery and a 5V low 30 dropout voltage
75. not feasible to identify and eliminate all aspects of kernel operation that lead to unpredictability These sources of unpredictability include the Linux scheduling algorithm which is optimized to maximize throughput device drivers uninterruptible system calls the use of interrupt disabling and virtual memory operations The best way to avoid these problems is to construct a small predictable kernel separate from the Linux kernel and to make it simple enough such that operations can be measured and shown to have predictable execution This has been the course taken by the developers of RTLinux This approach has the added benefit of maintainability before the development of RTLinux every time new device drivers or other enhancements to Linux were needed a study would have to be performed to determine that the change would not introduce unpredictability Figure 1 shows the basic Linux kernel without hard realtime support You will see that the Linux kernel separates the hardware from user level tasks The kernel has the ability to suspend any user level task once that task has outrun the slice of time allotted to it by the CPU Assume for example that a user task controls a robotic arm The standard Linux kernel could potentially preempt the task and give the CPU to one that is less critical e g one that boots up Netscape Consequently the arm will not meet strict timing requirements Thus RTLinux On A Small Mobile Robot Bitsoi La
76. plotted during the thermal test for the above insulation where a propane torch was used to simulate fuel fire conditions The plot is found in Figure 3 2 300 250 200 150 100 50 0 5 10 15 20 25 30 85 40 Figure 3 2 Thermal test for first insulation casing using Fastblock 301 17 The specified maximum internal temperature was 150 F that was reached 20 minutes into the test While the temperature should not have reached this temperature until 30 minutes into the test the results were promising and changes to the design could be made in order to meet the specifications Using the results of the first prototype tested a second test was designed This prototype was designed with consideration for the electrical components that had to fit within the casing The design was again a hemisphere as stated previously The setup used for this test can be seen in Figure 3 1 Due to materials limitations the spherical section of the design was 1 inch thick while the bottom of the casing was 0 8 inches thick Again a thermocouple was fed into the center of the prototype and Cerablanket insulation was placed inside the sphere The use of the Cerablanket was cost effective and drastically reduced the weight of the overall insulation see Table 3 1 The two Fastblock 301 components were sealed while wet so that it would form its own seal While Sauereisen had worked previously its brittle nature did not provide a durable devic
77. position information into an array of distance information it will consist of N plates of arrays that are 3 D vectors arranged in a 2 S 1 x 2 S 1 matrix templ zeros 2 S 1 3 temp2 zeros 2 S 1 2 S 1 3 delta zeros 2 S 1 2 S 1 N 3 for t 1 1 N for a 1 1 2 S 1 templ a P t end for b 1 1 2 S 1 temp2 b templ end delta t G temp2 end temp3 delta 2 rsqrd sum temp3 4 clear templ1 clear temp2 UBA Jo delta clear temp3 clear clear at this point we need to convert these distances over to a thickness contribution from each filament so we will have a new array of plates but this one will consist of N layers of the thickness contributed from each individual point source clip over the 2 S 1 x 2 S 1 grid covering the mirror layers zeros 2 S41 2 S41 N 27 layers 5 4 M 4 pi rho rsqrd clear rsqrd setting up the thickness on the mirror from this arrangement Coating zeros 2 S 1 2 S 1 Coatingtemp zeros 2 S 1 2 S 1 Mirror zeros 2 S 1 2 S 1 temp4 ones 2 S 1 2 S 1 temp5 zeros 2 S 1 2 S 1 Coatingtemp sum layers 3 setting a limiting value to compensate for the square grid and the round mirror beyond the mirror s edge Edge Coatingtemp S 1 2 S 1 surface zeros 2 S 1 2 S 1 surface S 1 1 surface S 1 1 for e 2 Teiles for u 1 1 S if e S 1
78. programmed to generate a signal once day that would wake the HC12 up and then once the HC12 executes the program it would go to sleep until the next day 25 Two RTCS that I found are the Dallas DS1306 SPI RTC and the Unitrode BQ3287E Parallel RTC I preferred the DS1306 RTC due to its SPI clocked communication So with just a little time left before the Senior Design Presentation I started trying to program the RTC Unfortunately I was unable to get the device to accept the program so I had to resort to the light sensor But with power conservation being the main focus of this project it would be beneficial to further explore an RTC as the source for an external interrupt 3 2 3 Microcontroller The choices for a device that can handle the complexity and the length of a program that can handle the various mechanical and electrical functions of this project have to be considered The various controllers that could handle these parameters are the Altera Programmable Logic Device the Motorola 16 bit 68HC12 the PIC and the STAMP Ieventually chose the Motorola 68HC12 I utilized the Motorola 68HC12912B32 EVB which is an evaluation board that houses the 68HC12912B32FU8 80 Microcontroller chip This board was provided to us for our EE 308 Microcontrollers class You can buy this particular evaluation board from the manufacturer for 90 or buy the chip for 30 and then create your own board This 16 bit controller can handle rather complex
79. programs with minimal power usage It also has the ability to go into a power saving mode This mode consumes minimal power when processing is not required during the night as in our case Though this type of controller is a bit overkill for our small application I happen to be familiar with the various functions of the device that can allow me to prototype the program that 26 will be used From this prototype I could use it to program controllers that are less costly and just as functional To make this planter work I had to utilize 1 Pulse Width Modulated signal port 5 Analog to Digital ports and 2 Output ports Fig 6 3 2 4 Program In order for this planter to successfully work it is necessary to measure if the battery needs to be or is able to be charged and to determine if the plant needs watering So to realize this it is necessary to create a program that can execute these two criteria and then after it has done something or not tell the controller to go into a sleep mode For the code I used C a high level programming language This language allowed me to create a behavioral type program that is relativity short in comparison to other types of programming languages like Assembly BASIC FORTRAN etc To write the code and download the necessary file into the controller I used the COSMIC Compiler IDEA CPU12 to create the program and a HC12 HyperTerminal window to handle the data transfer For the completed program refer t
80. requirements include building and testing the telescope kit and designing an interface which is both easy to use and educational The system must provide visitors with a choice of objects to observe and educate them in the basics of radio astronomy At the same time it must require minimal maintenance and supervision and must be resistant to abuse the public is likely to inflict To complete the project we divided up major tasks among team members The main areas of work involved developing educational material arranging for installation of the antenna developing a control program and implementing a touchscreen interface Our team has achieved all of the SRT project s goals except for completing final installation at the VLA site There has been a delay in obtaining a kiosk to house the control computer Arrangements have been made to complete installation within two weeks of the date of this thesis Acknowledgments First we would like to thank the National Radio Astronomy Observatory for giving us the opportunity to work on such a noteworthy project In addition Clint Janes with whom we worked in obtaining invaluable resources for this project earns special mention because he offered advice and guidance throughout the development of this work Additionally Robyn Harrison played a major role in shaping the final product because even though she initiated the idea our team was given free reign to develop the product as we could concept
81. root file system served over NFS and local booting and local file system with the home partition and some configuration files served over NFS were all considered Network booting In a network boot configuration the nodes in the clusters do not have hard drives Instead a boot rom is used on the network card to locate a bootp server and load the operating system image across the network from that server The primary advantage of this method is that a group of computers with similar hardware a single operating system image can be maintained This simplifies updates and configuration changes to the nodes in the cluster Another advantage of this method is that hard drives eventually fail In a large cluster the mean time between failures of the hard drives will lead to the failure of a hard drive somewhere in the cluster every few weeks If the nodes do not have hard drives this problem is eliminated as the only hard drives are in the bootp server The disadvantages are the need for network cards that support network booting and the associated boot roms Local booting with NFS root In a local boot with NFS root configuration each node boots the Linux kernel from a floppy hard drive or cdrom and then mounts the root file system from the server This method has the advantage of a single operating system image as in the network boot but has the disadvantages of boot media that can fail floppies and cdroms are especially prone to fai
82. sent The choice was made to calculate the actual temperature on the receiver side instead of computing it on the transmitter This was done to minimize the amount of data sent through the fire It was clear that the smaller the packet sent the better and there is a less of a chance that the fire would corrupt it Rather than sending the inventory data it was decided to store the inventory data on the receiver side This had two benefits it freed the transmitter to send more data and it kept all secure data from being transmitted If the transmission were intercepted the only thing received would be a number between 0 and 75 or the possibility of an abstract number This was a more secure method that allowed for the random generation ordnance code numbers The code for the receiver is in Appendix IX The following is the flowchart for the receiver In Figure 3 8 below is a sample of what is displayed to the screen when the receiver was ive i active fal BrainStem Console B xl temperature flux is missle is on fire current temp is 569 temperature flux is current temp is 312 temperature flux is current temp is 192 temperature flux is current temp is 119 temperature flux is Figure 3 8 Sample of receiver display 31 Temperature read in Stream received n 0 x stream bit Temperature displayed Warning displayed Figure 3 9 Flow chart for receiver Delta displayed 32 3 5 Transmitte
83. sh file open os path join user parallel setting dir Uzpl sh g ow zpl sh file write export ZPLHOME s n zpl home Zzpl sh file write export ZPLTARGET s n zpl target zpl sh file write parallelrc zplcommlayer sh n Zzpl sh file write export PATH PATH s n zpl bin zpl sh file close 4 end of zpl setup def generate cluster files default node count 2 default cpu per node 1 default domain name clusternet default node base name node default ip base addr 172 16 0 1 print print Generating system configuration files print print What domain name should this cluster use print default s default domain name domain name sys stdin readline 1 print if len domain name 0 domain name default domain name print What IP address should the cluster start at print default s default ip base addr ip base addr sys stdin readline 1 print if len ip base addr 0 ip base addr default ip base addr print How many nodes are in the cluster print default s default node count node count str sys stdin readline 1 print if len node_count_str 0 node count default node count else node count int node count str print How many CPUs processors are in each node print choose
84. so that was decided upon for our final design Another task worked on was the creation of the flashing times for the LED arrays The lights need to flash similar to the rate and intensity that normal trailer lights flash So the turn signals were set to flash about every one third a second The flash levels were set by adjusting the resistors on the 555 timers chip and with the lights being off for more time than on Brake sensor investigation and development were two key roles took on the second semester Some of the ideas that implemented were a tilt sensor a Hall effect sensor and a adjustable IR distance sensor created printed circuit boards for the Hall effect and distance sensor circuits and used a previously developed PC board for the tilt sensor Other sensors looked into that were feasible would be a different IR emitter and detector pair a current sensor or a pressure sensitive device that would mount directly onto the pedal itself After the prototype brake sensors were developed proceeded to test the initial designs refine them work on mounting issues and verify the correct operation of the sensors found that the best way to verify the designs was to actually mount the sensors in several different makes and years of trucks to get a feel for whether the sensor would work in a real world application A final task that was in charge of was taking care of all the purchasing and budget issues This task allowed me
85. software has to interface with the touchscreen There are a number of ways of doing this The best would be to implement a driver which would interface the program directly with the touchscreen Brian and I investigated this possibility by speaking to Pat Van Buskirk one of NRAO s programmers She referred us to another programmer who did not have time to help us We looked in a book and also checked a few internet references and decided against trying to write a driver for the touchscreen mainly due to time constraints Also the touchscreen that Brian had selected came equipped with some powerful software which lets it emulate sequences of keystrokes mouse movements and other functions He had set it up to run the software through a DOS prompt for testing so we decided to continue along that route 17 As it arrived the SRT software uses three command line arguments These serve as switches for az el drive simulation receiver simulation and maintenance mode To keep things as simple as possible we decided to simply add another command line argument which would specify the source This works with the SRT catalog file so that the number entered in the fourth argument corresponds to the order of the desired source in the catalog file This proved to be a quick way to select sources using the touchscreen The rest of the necessary commands are hard coded When the touchscreen starts the program it passes the desired source through the fourth c
86. stored and they may be called at a later time 2 5 Transmitter The navy requires that shipboard transmission frequency be between 2 39 and 2 45 GHz This is known as the UHF Band or more specifically the S Band Tests preformed by the NAVAIR prior to the beginning of this project showed that JP8 fuel fires have little or no effect on signals transmitted at the specified frequency This is stated in Appendix I This project required no receiver to be built This limited the amount knowledge that can be provided regarding the transmitter The power density required by a receiver remains unknown There is also information needed about signal propagation onboard ship Knowing this information and assuming line of sight one can use Friis Equation P P G G Q 4nd The required power density of the signal can be calculated with this extra information RISC reduced Instruction Set Computer 13 3 0 Original Developments 3 Power Supply Research Performed by Camden Mullen Extensive research into the use of TE s was performed to determine the best TE module configuration Experiments with a bismuth telluride module indicated that this type of module would not be suitable for a power supply The module used had solder that melted at too low of a temperature Research into the use of a Thermionic device suggested that this type of device would be more suitable Dr Harold Brown and Dr Yan Kucherov of Eneco Inc Salt Lake
87. test cases Following these tests the multiplier is tested using numeric rather than bit pattern test cases Three test groups one at 0 one around the square root of the 24 bit max and one around the 24 bit max are used Each of the two inputs are run through each of the three test ranges yielding roughly 600 thousand test cases The third group of tests focus on the ALU These tests must make the assumption that the data path is correctly routing the data and thus are not run until the data path s integrity is verified The tests also make the assumption that the multiplier works correctly as it is used as a shifter to achieve some of the Y test values The identity functions of the ALU pass X pass Y are tested in the data path tests and are assumed to work in the ALU tests The unary NOT operator is checked with test inputs the same as the data path Because the test cases must be tested over the full 48 bit inputs on the ALU this produces roughly 1000 tests for the NOT operator The unary negation operator focuses on numerical values rather than bit patterns and tests a group of inputs at the 48 bit negative maximum a group centered around the 24 bit negative maximum a group centered around 0 a group centered around the positive 24 bit maximum and a group at the positive 48 bit maximum Several other values scattered across the full input range are also tested This yields roughly 6000 test cases for the negation operator Like the dat
88. that feature would greatly increase ease of use The phase shift that must be applied to the mercury switch is chosen by the user and is found through trial and error An automated system to find the phase would be preferable These two additions would be large strides towards reaching our goal of furthering existing research 31 Appendix 32 Reconfigurable Data Path Processor Verification A Senior Thesis Presented to the Faculty of the Electrical Engineering Department of New Mexico Tech By Steven Wasson Thomee Wright In Partial Fulfillment of the requirements for the course EE 482 Senior Design II May 1 2002 ABSTRACT NASA s Institute for Advanced Microelectronics is in the process of developing a Reconfigurable Data Path Processor A full design for the RDPP exists at this time but needs to be tested This project verifies the design of the Processing Element the workhorse component of the RDPP This report first documents the implementation of the Processing Element in an Altera APEX20K200EFC chip and an interface to allow testing from a PC The report next documents how the PC software interfaces with the Altera board to run test vectors This report then documents how the test vectors were generated and why they provide sufficient test coverage Finally this document details how testing the device on the Altera board was unsuccessful ACKNOWLEDGEMENTS The authors would like to acknowledge the sponsor of
89. the sensors mounting to work properly in other applications One example of this would be the need to change the resistor values in the schmitt trigger to accommodate vehicles that have more distance to the firewall than the test cases Another example would be the creation of a mounting bracket to meet a specific application where the sensor could not be mounted easily using the methods that were successful in the test cases 5 4 Final Project Status The entire prototype system has been completed and is in a full working state The System was tested inside a vehicle and functioned flawlessly The system operated perfectly at 200 feet so the 100 feet requirement has been met with flying colors The LED arrays have a brighter intensity than current production vehicles so they also meet our requirements All the initial tests of the sensors seem to make the product look viable as the front signaling box performs well using both the sensors and the manual switches on the control box The receiver box containing the output logic and light driver circuitry also is in full functioning order Further testing of the product is recommended before it is put into unrestricted highway usage APPENDICIES Appendix A PALASM code that was used to program the PAL with the output logic LLELLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLL PALASM PARSER LISTING io LLELLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLL LINEA Jepe ep cec deed pousse cet LPALASM Design De
90. the driver to have to adjust his driving style in order for the system to fit his needs 3 2 3 Phototransistor Description A simple PN168 along with a driver circuit Application We would mount these on the dash next to the turn signal indicator lights to relay back when one of the lights is blinking We will need one of these for each the left and right turn signal light We will also be looking for a prepackaged light indicator that might be easier to mount Final Status The overall design of the phototransistor worked great The only problem that may occur in the future is excess light given off by the dashboard This may cause the phototransistor to be activated falsely have added a 10K ohm adjustable resistor to the circuit which will allow the user to make adjustments to the sensitivity of the phototransistor Refer to Appendix H for a detailed schematic of the circuitry involved 3 2 4 Lever Switch Description This is a very small simple lever switch that outputs 5 volts when activated Application Two lever switches are needed for the turn signal application one each above and below the turn signal shaft The lever switches are attached with Velcro just above and below the shaft with each lever just barely resting upon the turn signal shaft When the driver pushes the turn signal shaft either up or down the correct lever is activated and a high is sent to the control box indicating the correct turn Final S
91. thermocouple typically between 270 C and 2300 C would limit the resolution that could be provided to the internal electronics The customer for this project determined that the resolution of the temperature was not as important as being able to use the full range of the original project goals and design was done accordingly 3 2 2 Temperature Sensor Signal Conversion Data acquisition systems that convert the output of a thermocouple into a temperature are commercially available from companies such as Omega These units take the millivolt output of a thermocouple and use a conversion factor to change this into a temperature However these units would not meet the requirements of this project because they cost between 200 and 2000 This cost makes these products unreasonable for this project The conclusion drawn from this information meant that 22 a circuit had to be developed to translate the millivolt output of a K type thermocouple into a voltage range that could be measured by the electronic processor The first step of this process was to determine the output voltage range of the thermocouple when it is exposed to the full temperature range of this project 250 F to 1600 F A K type thermocouple was acquired for this project from the New Mexico Tech Instrument Room Thermal testing was preformed on the K type thermocouple to determine the voltage range it put out when exposed to the extremes of the required temperature range T
92. this project Dr Greg Donohoe for providing such a challenging project that contributed greatly to our learning experience Also Dr Donohoe provided all of the equipment and materials that were needed to conduct this project when we began it We would also like to acknowledge our faculty advisor for this project Dr Kevin Wedeward for the light natured humor filled project meetings we had throughout the year Finally we would like thank all of the faculty of the Electrical Engineering Department at New Mexico Tech to whom we are in great debt for the exceptional instruction they have provided us TABLE OF CONTENTS CHAPTER THIN TRODUC TION sss sisi vests a ite ods cls asp ERE Mande dish rabie aed ir dre 4 I LBACRGROUNDE iios er eta tes ede A isl sats onset Quee UENIRE ERR dR UE 4 12 PROJECT OBJECTIVES reiini tetur pitae ue ts tu nel Vd tides 4 1 3 PROVIDED MATERIALS amp EQUIPMENT eeeeeeeeee eene eterne een ntn nth tn einen etna 4 CHAPTER 2 TECHNICAL BACKGROUND amp DESIGN OVERVIEW etm 5 2 TANCERODEICOEEIORN S sctocistet ones foot tse hne beef bi st toruft iuste safete ft cup bert eua tHd 5 2 2 THE RECONFIGURABLE DATA PATH PROCESSOR eee 5 2 3 ALTERA EXCALIBUR BOARD AND QUARTUS II SOFTWARE 8 24 FUNCTION TESTING ui ds get eE e E E EKE EEE uide ae dee e ENE UN airs 9 2 5 DESIGN OVERVIEW istis nasi Up t tet pec a Ea A I CUT do FUE 10 CHAPTER 3 ALTERAG HARDWARE DES
93. to implement was the Sobel operator which successfully executed the conversion of images and produced satisfactory output The Roberts operator was used to find the difference in the color of the pixels Umbaugh 1998 13 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems The code for the Image Processing was written in C The program depended highly on structures They were easy to access and cut the time to process each image down The members of the project knew a significant amount of C programming which made it an easier way to implement with 14 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Discussion Image Processing Heather Bitsoi In the image processing part of the project our primary objective is to take an image from a basic web camera store this image and apply some edge finding algorithms There are different ways to accomplish this task The most appropriate and efficient processing is based upon the application needs The need for image processing is to be able to find a line in the surrounding environment and perform wall following using cameras instead of the IR distance sensors There are many more applications which the customer wishes to explore which are unfortunately beyond the scope of this project We have some limitations to the image processing procedure one of which is the limited processing power of the Prometheus SBC which is the source of the surrounding information Another con
94. to work with many different vendors and allow me to interact with many different people in industry The documentation was divided among the group members so also worked on several progress reports along with the development of the Gantt charts and work plans 4 3 Tasks Accomplished by Ivan was assigned to work on four major parts of our project The first part was to research the FCC regulations that may effect or limit the use of our transmitter receiver pairs This was a very important part of our project because we needed to make sure that we purchased a transmitter receiver pair that would be within FCC regulations Even though we did not find any transmitter receiver pairs that were already FCC accepted we did find several that laid within the frequency range that is acceptable to the FCC According to the FCC the transmitter receiver pairs that we picked could only transmit a signal within a range of 300 feet Anything over 300 feet would be considered illegal and must be FCC approved in order to be used was also chosen to work with Brandon on the alpha transmitter receiver pair The alpha transmitter and receiver are from Glolab and are part of a kit The kit was to be interfaced with an encoder decoder pair from either Holtek or Glolab Brandon and testing several encoder decoder pair so that we would be able to determine which pair will work best Performing these tests allowed us to pick an encoder decoder pair that woul
95. use Thermoionics TI which also use the heat of the fire However TI is still under development but can be reconsidered in the future The final decision was to use a long life lithium battery in conjunction with a thermo mechanical switch to power the device The thermal switch would be connected to a relay and the battery When the switch closed at the appropriate temperature the relay was activated keeping the circuit closed if the switch failed The major variables involved with a TE are heat flux through the TE temperature difference AT between the hot and cold sides operating temperature and electrical and heat resistance of the TE Under ideal conditions a TE can have a high efficiency at approximately 7096 at a large AT CRC Handbook 1995 relatively low temperatures less than 200 C and a good heat sink to draw the heat away from the module are also required CRC Handbook 1995 For this project several requirements were under consideration the temperature on the outside of the device would rise rapidly to 1000 C while the internal temperature must stay at or below 150 C The hot side of the TE would be on the outside edge of the insulation while the cold side would be imbedded into the insulation While this would create a very large AT in the beginning the AT would drop as the cold side temperature of the TE would increase due heat flux from the hot side to the cold side of the device This would cause two problems Fir
96. was to be able to protect the internal electronics for at least 30 minutes In order to protect these components the internal temperature of the prototype was to not exceed 150 F see Appendix I The ideal characteristics for the insulating material are that the material has a minimal thermal conductivity and minimal density in order to minimize weight and size Additionally corrosion UV and weather resistance as well as brittleness and strength were considered All of these factors must be considered to chose the most favorable housing for the device so that it will be durable The requirements for the insulation to meet were quite stringent when one considers that space shuttles upon reentry into the atmosphere must insulate against 2000 F for 9 minutes with an internal temperature not to exceed 350 F www spaceshuttletille com faq html 2001 For this reason when considering insulating materials both commercially available and non commercial available insulating materials were considered Insulation materials must protect against a flux of heat Heat transfer can occur by three basic means conduction convection and radiation Conduction is defined to occur when energized collide and transfer energy through a material or space Convection is found as the transfer of heat by mixing Finally radiation is the transfer of energy by electromagnetic waves Heat flux by convection could be neglected for the given conditio
97. we had helped wire up all of the sensors on a breadboard to see which ones would be good to try out on a vehicle From there we chose which sensors to take out to the vehicle and test After testing all of the sensors six were chosen as ones that would work for the final product Those six needed to be wired up in a way that was convenient to mount and pleasing to eye so we divided the sensors equally was given the two lever switches to mount and wire up also worked on the interface logic to take the signals from the driver s inputs and make them do what was necessary i e brake blink developed a series of AND and OR gates to implement the necessary patterns on the tail light system After that was finished worked with Dave to simplify either his design or mine After many brutal hours we were able to simplify my design down to a reasonable number of chips It involved four logic chips so we decided to look at a programmable logic chip to put all of that logic onto one chip got the programmer from Dr Rison and downloaded all the necessary software to run it and compile the code wrote the code Appendix A for the logic using a truth table Appendix B and successfully programmed it onto a PAL tested the PAL into the entire circuit and it worked exactly like it was suppose to This saved both money and space on the board worked with Ivan to get the alpha transmitter receiver pair from Glolab to work We ran into a few pr
98. while the right does the opposite and you turn away from the wall Conversely when the distance is large the left motor slows down while the right motor speeds up and the robot turns in to the wall Determining the K can be a little tricky if it is to small the robot wont correct 23 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems well enough and if it is too big the robot will oscillate and never get the desired distance The K was determined by computing the constant that would make the Left PWM 10 which is very small when the robot was a maximum distance away from the wall Our maximum distance was 20cm due to nonlinearities with the distance sensors so was able to calculate a K that would work well In order to right wall follow just swapped which sensor were reading from and changed the plus and minus signs on the PWM calculations Front wall detection was also implemented which used the front distance sensor to check if the robot was going to hit anything in front of it If the front sensor detected a wall the robot would turn in place to the right and continue left wall following or turn in place to the left and continue right wall following depending on which wall was being followed There are separate functions written for both right and left wall following and are included in the code in Appendix C Ultra Sonic Distance Sensor Kyle Lamb The Ultra Sonic distance sensor was a bit of a mystery throughout the year
99. work Once the computer was set up I 8 3 a installed Microsoft Office with PowerPoint We used PowerPoint to make educational slideshows for the public Our customer wanted a touchscreen for the user interface I found a touchscreen supplier in Albuquerque called MassWorks They sell a small color LCD touchscreen that connects to the computer via a USB port connection The layout software that comes with the touchscreen can be integrated with any computer application At this time I have set up the touchscreen to run the VLA 20 weather software the VLA web pages the educational slideshows and the tracking software with the PowerPoint presentation I set up three main layouts because some objects are not visible at certain times of the year figure 3 3 1 The layout on figure 3 3 1a has the Sun Cygnus A and Cassiopeia available for tracking This layout will be used from November through May The layout on figure 3 3 1b has only the Sun and Cassiopeia available Cygnus A is not on this layout because it is only visible in the northern hemisphere from June through October The layout on figure 3 3 1c will not have any objects available to track This layout will be used if the SRT must be inactive due to weather conditions or technical difficulties People will only have access to the slideshows web links and VLA weather A technician will have to change layouts at the appropriate time Please Selec a loda d Select an ican Please
100. 0 mA on the OOPIC The OOPIC has 24 digital I O ports and 10 6 bit A D ports contrasted by the 5 digital I O ports and 5 10 bit A D ports on the Brainstem In addition to these shared characteristics each device has its own unique features The OOPIC has an onboard prototype area that allows for expansion and also has developed objects to simplify the programming process The Brainstem is able to run four programs concurrently a small package size and a very active customer support system These benefits and detriments are listed in Table 3 5 TEA Tiny Embedded application small programs that are used concurrently to minimize program size 7 C style see Appendix VII 26 Device Pro s Con s OOPIC Object Oriented Must use WIN 98 ANSI standard High Power Requirement Prototype Area No User Support BrainStem Universal Platform Non ANSIC 10 Bit A D No Prototype Area Active User Support Limited Digital Ports Low Power Requirement Able to Run Programs Concurrently Table 3 5 Comparison of Processors The Brainstem was selected for the project due to its universal developmental platform low power requirements and high level of customer support These factors made this device stand out from the OOPIC because they were determined to be vital to the overall success of the project In testing it was determined that the processor can be programmed with any machine ranging from Windows XP to Linux or Palm to Windows CE This provided
101. 11 5 2393 5 2414 5 3174 5 5518 14 49 12 5 0851 5 1611 5 2393 5 2393 14 49 13 5 0851 5 1611 5 2393 5 2393 Min Values from rows 1 50 14 49 14 5 0851 5 2393 5 2393 5 2393 5 0851 5 1611 5 0830 14 49 15 5 0851 5 2393 5 2393 5 2393 14 49 16 5 0851 5 1611 5 2393 5 2393 Avg Value of Maxes F 14 49 17 5 0851 5 2393 5 2393 5 2393 5 3961 89 93417 14 49 18 5 0851 5 1611 5 2393 5 2393 14 49 19 5 0851 5 2393 5 2393 5 2393 Avg Value of Mins F 14 49 20 5 0851 5 2393 5 2393 5 2393 5 1226 85 37625 14 49 21 5 0851 5 2393 5 2393 5 2393 14 49 22 5 0851 5 2393 5 2393 5 2393 Av Max Av Min F 14 49 23 5 0851 5 1611 5 2393 5 2393 0 2735 4 557917 14 49 24 5 0851 5 2393 5 2393 5 2393 5 4736 5 1611 C 32 18565 C 29 65347 C 2 532176 14 49 25 5 0851 5 2393 5 2393 5 2393 Average Temp change in Fahrenheit 14 49 26 5 0851 5 2393 5 2393 5 2393 4 557917 14 49 27 5 1633 5 1611 5 2393 5 2393 14 49 28 5 1633 5 1611 5 3174 5 3174 Temperature Change in Celcius 14 49 29 5 1633 5 2393 5 3174 5 2393 2 532176 14 49 30 5 1633 5 2393 5 3955 5 2393 14 49 31 5 1633 5 2393 5 3955 5 3174 Power Absorbed Mass Specific Heat Change in Temp C 14 49 32 5 0851 5 2393 5 3955 5 3955 2 514147 Watts time 14 49 33 5 1633 5 2393 5 3955 5 3955 14 49 34 5 0851 5 2393 5 3955 5 3955 Energy Absorbed Mass Specific Heat Change in Temp C 14 49 35 5 1633 5 2393 5 3955 5 3174 125 7073 Joules 14 49 36 5 1633 5 2393 5 3955 5 3955 14 49 37 5 1633 5 2393 5 3955 5 3955 14 4
102. 15 5 8195 5 3174 5 2393 5 3174 14 52 16 5 3195 5 3174 5 2393 5 3174 14 52 17 5 2414 5 3174 5 2393 5 3174 14 52 18 5 2414 5 2393 5 2393 5 3174 14 52 19 5 8195 5 3174 5 2393 5 3955 14 52 20 5 2414 5 3174 5 2393 5 3174 14 52 21 5 2414 5 3174 5 2393 5 3955 14 52 22 5 3195 5 3174 5 2393 5 3174 14 52 23 5 2414 5 2393 5 2393 5 3174 14 52 24 5 2414 5 2393 5 2393 5 3955 14 52 25 5 3195 5 3174 5 2393 5 3174 14 52 26 5 2414 5 2393 5 2393 5 3174 14 52 27 5 3195 5 3174 5 2393 5 3174 14 52 28 5 3195 5 3174 5 2393 5 3174 14 52 29 5 2414 5 2393 5 2393 5 3174 14 52 30 5 3195 5 2393 5 2393 5 2393 14 52 31 5 3195 5 3174 5 2393 5 3174 14 52 32 5 2414 5 2393 5 2393 5 2393 14 52 38 5 2414 5 3174 5 2393 5 2393 14 52 34 5 2414 5 2393 5 2393 5 2393 14 52 35 5 2414 5 2393 5 2393 5 2393 14 52 36 5 2414 5 2393 5 2393 5 2393 14 52 37 5 2414 5 2393 5 2393 5 2393 14 52 38 5 2414 5 2393 5 2393 5 2393 14 52 39 5 3195 5 2393 5 2393 5 2393 14 52 40 5 2414 5 2393 5 2393 5 2393 14 52 41 5 2414 5 2393 5 2393 5 2393 14 52 42 5 2414 5 3174 5 2393 5 2393 14 52 43 5 2414 5 3174 5 2393 5 2393 14 52 44 5 3195 5 2393 5 2393 5 2393 14 52 45 5 2414 5 2393 5 2393 5 2393 14 52 46 5 2414 5 3174 5 2393 5 2393 14 52 47 5 2414 5 2393 5 2393 5 2393 14 52 48 5 2414 5 2393 5 2393 5 2393 14 52 49 5 2414 5 2393 5 2393 5 2393 14 52 50 5 3195 5 2393 5 2393 5 2393 14 52 51 5 8195 5 2393 5 1611 5 2393 14 52 52 5 2414 5 2393 5 2393 5 2393 14 52 58 5 2414 5 2393 5 2393 5 2393 14 52 54 5 2414 5 2393 5
103. 16 repeated nibble words 111111 222222 FFFFFF This yields roughly 200 test cases per path tested Because test cases don t check the interaction between inputs the number of test cases can be kept relatively small for each data path There are 27 data paths tested here producing roughly 6000 test cases Each shifter is tested with the full set of data path values being shifted by one bit producing 4000 test cases Each shifter is then tested shifting a single bit through its full range of possibilities producing roughly 530 thousand test cases A few assumptions are made in the data path testing which may make isolation of a problem difficult It is assumed that multiplication by 1 on either input works and that the ALU s pass X and pass Y functions work Also it is assumed that several data path components perform as expected before they are actually tested These test ambiguities are unavoidable as any input must pass through several components between the input and output This is not a problem for our project as the goal is to determine whether problems exist not where they are Many of the tests will be obvious where the failure is aiding in solving the problem but this is not guaranteed 22 The second group of tests check the Processing Element multiplier The first group tests check using the multiplier as a shifter They use the same 8 test patterns as the data path tests shifted up to 47 bits left This yields roughly 800
104. 174 5 2393 14 31 30 5 2414 5 3174 5 2393 5 2393 14 31 31 5 3195 5 3174 5 2393 5 2393 14 31 32 5 8195 5 2393 5 2393 5 2393 14 31 33 5 8195 5 2393 5 2393 5 2393 14 31 34 5 2414 5 2393 5 2393 5 2393 14 31 35 5 8195 5 2393 5 2393 5 2393 14 31 36 5 3195 5 2393 5 2393 5 2393 14 31 37 5 3195 5 2393 5 2393 5 2393 3l 14 31 38 5 3195 5 2393 5 2393 5 2393 14 31 39 5 2414 5 3174 5 2393 5 2393 14 31 40 5 3195 5 2393 5 2393 5 2393 14 31 41 5 3195 5 3174 5 2393 5 2393 14 31 42 5 3195 5 2393 5 2393 5 2393 14 31 43 5 3195 5 2393 5 2393 5 2393 14 31 44 5 3195 5 2393 5 2393 5 2393 14 31 45 5 2414 5 2393 5 2393 5 2393 14 31 46 5 3195 5 2393 5 2393 5 2393 14 31 47 5 2414 5 2393 5 2393 5 2393 14 31 48 5 3195 5 2393 5 2393 5 2393 14 31 49 5 3195 5 2393 5 2393 5 2393 14 31 50 5 2414 5 2393 5 2393 5 2393 14 31 51 5 3195 5 2393 5 2393 5 2393 14 31 52 5 3195 5 2393 5 2393 5 2393 14 31 53 5 3195 5 2393 5 2393 5 2393 14 31 54 5 3195 5 2393 5 2393 5 2393 14 31 55 5 2414 5 3174 5 2393 5 2393 14 31 56 5 3195 5 2393 5 2393 5 2393 14 31 57 5 2414 5 2393 5 2393 5 2393 14 31 58 5 3195 5 2393 5 2393 5 2393 14 31 59 5 3195 5 2393 5 2393 5 2393 14 32 00 5 2414 5 2393 5 2393 5 2393 14 32 01 5 3195 5 2393 5 2393 5 2393 14 32 02 5 3195 5 2393 5 2393 5 2393 14 32 03 5 2414 5 3174 5 2393 5 2393 14 32 04 5 2414 5 2393 5 2393 5 2393 14 32 05 5 3195 5 2393 5 2393 5 2393 14 32 06 5 2414 5 2393 5 2393 5 2393 14 32 07 5 3195 5 2393 5 2393 5 2393 14 32 08 5 2414 5 2393 5 2393 5 2393
105. 19 5 0851 5 0830 5 3174 5 2393 14 42 20 5 0851 5 16115 3174 5 2393 14 42 21 5 0851 5 0830 5 3955 5 2393 14 42 22 5 1633 5 0830 5 3174 5 2393 14 42 23 5 0851 5 16115 31745 3174 14 42 24 5 0851 5 1611 5 3955 5 2393 14 42 25 5 1633 5 16115 3174 5 2393 14 42 26 5 0851 5 1611 5 3955 5 2393 14 42 27 5 0851 5 1611 5 3955 5 3955 14 42 28 5 1633 5 2393 5 3955 5 3955 14 42 29 5 0851 5 2393 5 3955 5 3174 14 42 30 5 1633 5 2393 5 3955 5 3955 14 42 31 5 0851 5 2393 5 3955 5 3955 14 42 32 5 0851 5 16115 4736 5 3955 14 42 33 5 0851 5 2393 5 3955 5 3955 14 42 34 5 0851 5 16115 4736 5 3955 14 42 35 5 1633 5 2393 5 4736 5 3955 14 42 36 5 1633 5 2393 5 3955 5 3955 14 42 37 5 1633 5 2393 5 4736 5 3955 Max Values from rows 1 50 5 1633 5 2393 5 4736 5 3955 Min Values from rows 1 50 5 0070 5 0830 5 0830 5 0049 Avg Value of Maxes F C 5 3179 88 63208 31 46227 Avg Value of Mins F C 5 0445 84 07458 28 93032 Av Max Av Min F C 0 2735 4 5575 2 531944 Average Temp change in Fahrenheit 4 5575 Temperature Change in Celcius 2 531944 Power Absorbed Mass Specific Heat Change in Temp C 2 513917 Watts time Energy Absorbed Mass Specific Heat Change in Temp C 125 6959 Joules 34 14 42 38 5 1633 5 2393 5 4736 5 3955 14 42 39 5 1633 5 1611 5 4736 5 3955 14 42 40 5 1633 5 1611 5 4736 5 3955 14 42 41 5 1633 5 2393 5 4736 5 3955 14 42 42 5 1633 5 2393 5 4736 5 3955 14 42 43 5 1633 5 1611 5 4736 5 3955 14 42 44 5 1633 5 1611 5 4736 5 3955
106. 195 5 2393 5 6299 5 4736 Min Values from rows 1 50 5 0851 5 0830 5 0049 5 0830 Avg Value of Maxes F C 5 4156 90 25958 32 36644 Avg Value of Mins F C 5 0640 84 4 29 11111 Av Max Av Min F C 0 3516 5 859583 3 255324 Average Temp change in Fahrenheit 5 859583 Temperature Change in Celcius 3 255324 Power Absorbed Mass Specific Heat Change in Temp C 3 232146 Watts time Energy Absorbed X Mass Specific Heat Change in Temp C 161 6073 Joules 27 14 28 26 5 1633 5 2393 5 5518 5 3955 14 28 27 5 1633 5 2393 5 6299 5 3955 14 28 28 5 2414 5 2393 5 5518 5 4736 14 28 29 5 3195 5 2393 5 6299 5 3955 14 28 30 5 2414 5 2393 5 5518 5 4736 14 28 31 5 2414 5 3174 5 5518 5 4736 14 28 32 5 3195 5 2393 5 5518 5 4736 14 28 33 5 2414 5 2393 5 6299 5 4736 14 28 34 5 2414 5 2393 5 6299 5 4736 14 28 35 5 2414 5 3174 5 6299 5 4736 14 28 36 5 3195 5 3174 5 6299 5 4736 14 28 37 5 3195 5 2393 5 6299 5 4736 14 28 38 5 3195 5 2393 5 6299 5 4736 14 28 39 5 3195 5 2393 5 6299 5 5518 14 28 40 5 3195 5 3174 5 6299 5 5518 14 28 41 5 3195 5 2393 5 6299 5 5518 14 28 42 5 3195 5 3174 5 6299 5 5518 14 28 43 5 3195 5 3174 5 5518 5 4736 14 28 44 5 3195 5 2393 5 6299 5 5518 14 28 45 5 8195 5 3174 5 6299 5 5518 14 28 46 5 3195 5 3174 5 6299 5 5518 14 28 47 5 3195 5 3174 5 6299 5 4736 14 28 48 5 3195 5 3174 5 5518 5 4736 14 28 49 5 3195 5 3174 5 6299 5 4736 14 28 50 5 3195 5 3174 5 5518 5 4736 14 28 51 5 8195 5 3174 5 6299 5 5518 14 28 52 5 8195
107. 2 u S 1 2 gt S 2 u surface e surface 2 S 2 e u 0 surface e 2 S 2 u 0 surface 2 S 2 e 2 S 2 u 0 else u surface e surface 2 S 2 e u 1 surface e 2 S 2 u 1 surface 2 S 2 e 2 S 2 u 1 end end end 28 temp5 Edge temp4 surface Coating temp5 Coatingtemp surface Mirror Coatingtemp G 3 surface G S 1 2 S 1 3 temp4 surface Plotting the resulting coating figure 2 mesh XY XY 10 8 Coating XLABEL Mirror Position cm YLABEL Mirror Position cm ZLABEL Coating Thickness angstroms TITLE Perfect run no miss fires determining the range of thicknesses and other values as output A B min Coating S 1 C D max Coating S 1 Gi co Minimum A 10 8 minradius abs 26 B 5 1308 Maximum C 10 8 maxradius abs 26 D 5 1308 Var Maximum Minimum Ave 10 8 mean mean Coating writting soem of the data to files wklwrite e glass xls G S 1 3 2 2 wklwrite e mirror xls Mirror S 1 2 2 wklwrite position xls XY 2 2 curve fitting for use in zemax optical analysis gc polyfit XY G S 1 3 7 wklwrite glass cross section curve fit xls gc mc polyfit XY Mirror S 1 7 wklwrite mirror cross section curve fit xls mc 29 coatfit polyfit XY Coating S 1 7 wklwrite coating cross section curve fit xls coatfit 30 Appendix B 100 inch mirror co
108. 32 28 5 2414 5 2393 5 1611 5 2393 14 32 29 5 8195 5 2393 5 2393 5 2393 14 32 30 5 2414 5 2393 5 1611 5 2393 14 32 31 5 2414 5 2393 5 1611 5 2393 14 32 32 5 2414 5 2393 5 1611 5 2393 14 32 33 5 2414 5 3174 5 2393 5 2393 14 32 34 5 2414 5 2393 5 1611 5 2393 14 32 35 5 2414 5 2393 5 1611 5 2393 14 32 36 5 2414 5 2393 5 1611 5 2393 14 32 37 5 2414 5 2393 5 2393 5 2393 14 32 38 5 2414 5 2393 5 1611 5 2393 14 32 39 5 2414 5 2393 5 1611 5 2393 14 32 40 5 2414 5 2393 5 1611 5 2393 14 32 41 5 2414 5 2393 5 2393 5 2393 14 32 42 5 2414 5 2393 5 2393 5 2393 14 32 43 5 2414 5 2393 5 2393 5 2393 14 32 44 5 2414 5 2393 5 1611 5 2393 14 32 45 5 2414 5 2393 5 2393 5 2393 14 32 46 5 2414 5 2393 5 2393 5 2393 14 32 47 5 2414 5 2393 5 1611 5 2393 14 32 48 5 2414 5 2393 5 1611 5 2393 14 32 49 5 2414 5 2393 5 2393 5 2393 14 32 50 5 2414 5 2393 5 1611 5 2393 14 32 51 5 2414 5 2393 5 1611 5 2393 14 32 52 5 2414 5 2393 5 1611 5 2393 14 32 58 5 2414 5 2393 5 1611 5 2393 14 32 54 5 2414 5 2393 5 1611 5 2393 14 32 55 5 2414 5 2393 5 1611 5 2393 14 32 56 5 1633 5 2393 5 2393 5 2393 14 32 57 5 2414 5 2393 5 2393 5 2393 14 32 58 5 2414 5 2393 5 2393 5 2393 14 32 59 5 2414 5 2393 5 1611 5 2393 54 A 8 Josef Hart s Resume HOME ADDRESS SCHOOL ADDRESS OBJECTIVE EDUCATION Technology WORK EXPERIENCE ACTIVITIES 12112 Glen Canyon Rd NE Albuquerque New Mexico 87111 505 293 3098 Post Office Box 2083 Socorro New Mexico 87801 505 835 6853
109. 34 PROGRAMMING THE ALTERA CHIP With the Processing Element successfully compiled and our interface developed we were ready to program the Altera chip Below is a diagram of the Altera board with all the components that we used labeled JP11 40 pin header for 5 0 v U4 Prototype Connector amp PEX JP2 Configuration Jumper Header Controller for Configuration Controller x u3 D1 Two Digit Flash 7 Segment Display Memory evice Push Buttons he EP20K200E Device Figure 9 Altera amp Board Diagram In our program we designated JP11 for the I O to the computer Figure 10 on the next page shows a pin out for the JP11 header We set up the Two Digit 7 Segment Display to display all of the I O between the computer and the Processing Element for debugging purposes Figure 11 on the next page shows which LED corresponds with each I O signal Finally we used one of the Push Buttons for a reset NIOS Embedded Processor Development Board 3 18 JP11 1 2 3 e 4 50 6 data out 7 a data_outt data out3 9 e e 10 data out2 PE clock 12 TxRx clock TxRx state1 3 14 TxRx state data in3 5 16 data in2 TxRx_statet data_ini17 18 data in aw GND 19 e 2 data cdi data out data _out0 j N 21 2 7 X 23 e 4 e m m ul xd aged fos 01 TxRx_state0 27 e 28 gt W ad s we aoe 31 v e amp ale 3 p vas fatica ocx Y SPE clock coll teal
110. 393 5 2393 5 2393 14 31 50 5 2414 5 2393 5 2393 5 2393 14 31 51 5 3195 5 2393 5 2393 5 2393 14 31 52 5 3195 5 2393 5 2393 5 2393 14 31 53 5 3195 5 2393 5 2393 5 2393 14 31 54 5 3195 5 2393 5 2393 5 2393 14 31 55 5 2414 5 3174 5 2393 5 2393 14 31 56 5 3195 5 2393 5 2393 5 2393 14 31 57 5 2414 5 2393 5 2393 5 2393 14 31 58 5 3195 5 2393 5 2393 5 2393 14 31 59 5 3195 5 2393 5 2393 5 2393 14 32 00 5 2414 5 2393 5 2393 5 2393 14 32 01 5 3195 5 2393 5 2393 5 2393 14 32 02 5 3195 5 2393 5 2393 5 2393 14 32 03 5 2414 5 3174 5 2393 5 2393 14 32 04 5 2414 5 2393 5 2393 5 2393 14 32 05 5 3195 5 2393 5 2393 5 2393 14 32 06 5 2414 5 2393 5 2393 5 2393 14 32 07 5 3195 5 2393 5 2393 5 2393 14 32 08 5 2414 5 2393 5 2393 5 2393 14 32 09 5 3195 5 2393 5 2393 5 2393 14 32 10 5 2414 5 2393 5 2393 5 2393 14 32 11 5 3195 5 2393 5 2393 5 2393 14 32 12 5 3195 5 2393 5 2393 5 2393 14 32 13 5 3195 5 2393 5 2393 5 2393 14 32 14 5 2414 5 2393 5 2393 5 2393 14 32 15 5 3195 5 2393 5 2393 5 2393 14 32 16 5 3195 5 2393 5 2393 5 2393 14 32 17 5 3195 5 2393 5 2393 5 2393 14 32 18 5 2414 5 2393 5 2393 5 2393 14 32 19 5 2414 5 3174 5 2393 5 2393 14 32 20 5 2414 5 2393 5 2393 5 2393 14 32 21 5 3195 5 2393 5 2393 5 2393 14 32 22 5 3195 5 2393 5 1611 5 2393 14 32 23 5 3195 5 2393 5 2393 5 2393 14 32 24 5 3195 5 2393 5 2393 5 2393 14 32 25 5 2414 5 2393 5 1611 5 2393 39 14 32 26 5 3195 5 2393 5 2393 5 2393 14 32 27 5 2414 5 2393 5 1611 5 2393 14 32 28 5 2414 5 2393 5 1611
111. 3955 5 4736 14 29 32 5 8195 5 3955 5 3955 5 4736 14 29 33 5 8195 5 3955 5 3955 5 4736 14 29 34 5 3195 5 3955 5 3955 5 4736 14 29 35 5 3195 5 3955 5 3955 5 4736 14 29 36 5 3195 5 3955 5 3955 5 4736 14 29 37 5 3195 5 3955 5 3955 5 4736 14 29 38 5 8195 5 3955 5 3955 5 3955 14 29 39 5 3195 5 3955 5 3955 5 3955 14 29 40 5 3976 5 3955 5 3955 5 4736 14 29 41 5 3195 5 3955 5 3955 5 4736 14 29 42 5 3195 5 3955 5 3955 5 4736 14 29 43 5 3195 5 3955 5 3955 5 4736 14 29 44 5 3195 5 3955 5 3955 5 4736 14 29 45 5 3195 5 3955 5 3955 5 4736 14 29 46 5 3195 5 3955 5 3955 5 4736 14 29 47 5 3195 5 3955 5 3955 5 4736 14 29 48 5 3195 5 3955 5 3955 5 4736 14 29 49 5 3195 5 3955 5 3955 5 4736 14 29 50 5 3976 5 3955 5 3955 5 3955 14 29 51 5 3976 5 3955 5 3955 5 4736 14 29 52 5 3195 5 3955 5 3955 5 4736 14 29 53 5 3195 5 3955 5 3955 5 4736 14 29 54 5 3195 5 3955 5 3955 5 4736 14 29 55 5 8195 5 3955 5 3955 5 3955 14 29 56 5 3976 5 3955 5 3955 5 3955 14 29 57 5 3976 5 3955 5 3955 5 3955 14 29 58 5 3976 5 3955 5 3955 5 3955 14 29 59 5 3195 5 3955 5 3955 5 4736 14 30 00 5 3976 5 3955 5 3955 5 3955 14 30 01 5 3195 5 3955 5 3955 5 4736 29 14 30 02 5 3976 5 3955 5 3955 5 3955 14 30 03 5 3195 5 3955 5 3955 5 3955 14 30 04 5 3195 5 3955 5 3955 5 3955 14 30 05 5 3195 5 3955 5 3955 5 3955 14 30 06 5 3976 5 3955 5 3955 5 3955 14 30 07 5 3976 5 3955 5 3955 5 3955 14 30 08 5 3976 5 3955 5 3955 5 3955 14 30 09 5 3195 5 3955 5 3955 5 3955 14 30 10 5 3976 5 3955 5 3955 5 395
112. 3976 5 3955 5 3955 5 3955 15 05 10 5 3195 5 3955 5 3955 5 3955 15 05 11 5 3976 5 3955 5 3955 5 3955 15 05 12 5 8195 5 3174 5 3174 5 3955 15 05 13 5 3976 5 3955 5 3955 5 3955 15 05 14 5 3976 5 3955 5 3955 5 3955 15 05 15 5 3976 5 3955 5 3955 5 4736 15 05 16 5 3195 5 3955 5 3955 5 4736 15 05 17 5 3976 5 3955 5 3955 5 4736 15 05 18 5 3195 5 3955 5 3955 5 3955 15 05 19 5 3976 5 3955 5 3955 5 3955 15 05 20 5 3195 5 3174 5 3174 5 3955 15 05 21 5 3976 5 3955 5 3955 5 4736 15 05 22 5 8195 5 3955 5 3955 5 4736 15 05 23 5 3976 5 3955 5 3955 5 3955 15 05 24 5 3976 5 3955 5 3174 5 3955 15 05 25 5 3976 5 3955 5 3174 5 3955 15 05 26 5 3195 5 3955 5 3955 5 3955 15 05 27 5 3976 5 3955 5 3955 5 3955 15 05 28 5 8195 5 3955 5 3174 5 3955 51 15 05 29 5 3195 5 3955 5 3174 5 3174 15 05 30 5 3195 5 3174 5 3174 5 3955 15 05 31 5 3195 5 3174 5 3174 5 3955 15 05 32 5 8195 5 3955 5 3174 5 4736 15 05 33 5 3976 5 3955 5 3955 5 3955 15 05 34 5 3195 5 3955 5 3955 5 3955 15 05 35 5 3195 5 3174 5 3174 5 3955 15 05 36 5 3195 5 3174 5 3174 5 3955 15 05 37 5 3976 5 3955 5 3174 5 3174 15 05 38 5 2414 5 3955 5 3174 5 3174 15 05 39 5 3976 5 3955 5 3955 5 3955 15 05 40 5 3195 5 3174 5 3174 5 3955 15 05 41 5 3976 5 3955 5 3955 5 3955 15 05 42 5 3195 5 3174 5 3174 5 3174 15 05 43 5 3976 5 3955 5 3174 5 3174 15 05 44 5 3976 5 3955 5 3174 5 3174 15 05 45 5 8195 5 3955 5 3955 5 3955 15 05 46 5 3976 5 3955 5 3174 5 3174 15 05 47 5 3195 5 3174 5 2393 5 3174 15 05 48 5 3195 5 317
113. 4 5 2393 5 3174 15 05 49 5 3195 5 3174 5 3174 5 3174 15 05 50 5 3195 5 3174 5 2393 5 3174 15 05 51 5 2414 5 3955 5 2393 5 3955 15 05 52 5 3195 5 3174 5 3174 5 3174 15 05 53 5 3195 5 3955 5 3174 5 3174 15 05 54 5 3976 5 3955 5 3174 5 3174 15 05 55 5 2414 5 3174 5 2393 5 3955 15 05 56 5 8195 5 3174 5 2393 5 3955 15 05 57 5 2414 5 3174 5 3174 5 3174 15 05 58 5 2414 5 3955 5 2393 5 3955 15 05 59 5 2414 5 3955 5 2393 5 3955 15 06 00 5 3976 5 3955 5 3174 5 3174 15 06 01 5 2414 5 3174 5 2393 5 3955 15 06 02 5 3195 5 3174 5 2393 5 3955 15 06 03 5 2414 5 3174 5 2393 5 3955 15 06 04 5 2414 5 3174 5 2393 5 3955 15 06 05 5 2414 5 3955 5 2393 5 3955 15 06 06 5 2414 5 3174 5 2393 5 3955 15 06 07 5 2414 5 3955 5 2393 5 3955 15 06 08 5 3195 5 3174 5 2393 5 3955 15 06 09 5 2414 5 3955 5 2393 5 3955 15 06 10 5 2414 5 3955 5 2393 5 3955 15 06 11 5 2414 5 3955 5 2393 5 3955 15 06 12 5 2414 5 3174 5 2393 5 3955 15 06 13 5 2414 5 3955 5 2393 5 3955 15 06 14 5 2414 5 3955 5 2393 5 3955 15 06 15 5 2414 5 3955 5 2393 5 3955 15 06 16 5 3195 5 3174 5 2393 5 3955 52 15 06 17 5 2414 5 3955 5 2393 5 3955 15 06 18 5 2414 5 3955 5 2393 5 3955 15 06 19 5 2414 5 3174 5 2393 5 3955 15 06 20 5 2414 5 3174 5 2393 5 3955 15 06 21 5 2414 5 3955 5 2393 5 3955 15 06 22 5 3195 5 3174 5 2393 5 3955 15 06 23 5 2414 5 3955 5 2393 5 3955 14 31 45 5 2414 5 2393 5 2393 5 2393 14 31 46 5 3195 5 2393 5 2393 5 2393 14 31 47 5 2414 5 2393 5 2393 5 2393 14 31 48 5 3195 5 2393 5 2393 5
114. 48 5 2414 5 2393 5 1611 5 2393 14 32 49 5 2414 5 2393 5 2393 5 2393 14 32 50 5 2414 5 2393 5 1611 5 2393 14 32 51 5 2414 5 2393 5 1611 5 2393 14 32 52 5 2414 5 2393 5 1611 5 2393 14 32 58 5 2414 5 2393 5 1611 5 2393 14 32 54 5 2414 5 2393 5 1611 5 2393 14 32 55 5 2414 5 2393 5 1611 5 2393 14 32 56 5 1633 5 2393 5 2393 5 2393 14 32 57 5 2414 5 2393 5 2393 5 2393 14 32 58 5 2414 5 2393 5 2393 5 2393 14 32 59 5 2414 5 2393 5 1611 5 2393 47 A 7 Trial 4 Data 15 02 19 5 1633 5 1611 5 2393 5 2393 15 02 20 5 1633 5 2393 5 2393 5 2393 15 02 21 5 2414 5 2393 5 2393 5 2393 15 02 22 5 2414 5 2393 5 2393 5 2393 15 02 23 5 1633 5 2393 5 2393 5 2393 15 02 24 5 2414 5 2393 5 2393 5 2393 Max Values from rows 1 50 15 02 25 5 2414 5 2393 5 2393 5 2393 5 3195 5 3174 5 6299 15 02 26 5 1633 5 2393 5 2393 5 2393 15 02 27 5 2414 5 2393 5 2393 5 2393 Min Values from rows 1 50 15 02 28 5 1633 5 2393 5 2393 5 2393 5 16833 5 1611 5 2393 15 02 29 5 2414 5 2393 5 2393 5 2393 15 02 30 5 2414 5 2393 5 3174 5 2393 Avg Value of Maxes F 15 02 31 5 2414 5 2393 5 2393 5 2393 5 4547 90 91083 15 02 32 5 1633 5 2393 5 2393 5 2393 15 02 33 5 1633 5 2393 5 3174 5 2393 Avg Value of Mins F 15 02 34 5 1633 5 2393 5 2393 5 2393 5 2008 86 67917 15 02 35 5 2414 5 2393 5 3174 5 3174 15 02 36 5 2414 5 2393 5 3174 5 3955 Av Max Av Min F 15 02 37 5 2414 5 2393 5 3955 5 2393 0 2539 4 231667 15 02 38 5 2414 5 2393 5 3174 5 3955 5 5518 5 2393 C 32 72824 C 30 37731
115. 5 14 30 11 5 3195 5 3955 5 3955 5 3955 14 30 12 5 3976 5 3955 5 3955 5 3955 14 30 13 5 3976 5 3955 5 3955 5 3955 14 30 14 5 83195 5 3955 5 3955 5 3955 14 30 15 5 3976 5 3955 5 3955 5 3955 14 30 16 5 3195 5 3955 5 3955 5 3955 14 30 17 5 3195 5 3955 5 3955 5 3955 14 30 18 5 8195 5 3955 5 3955 5 3955 14 30 19 5 3195 5 3955 5 3955 5 3955 14 30 20 5 3195 5 3955 5 3955 5 3955 14 30 21 5 3195 5 3955 5 3955 5 3955 14 30 22 5 3976 5 3955 5 3955 5 3955 14 30 23 5 3195 5 3955 5 3955 5 3955 14 30 24 5 3976 5 3955 5 3955 5 3955 14 30 25 5 3976 5 3955 5 3955 5 3955 14 30 26 5 3195 5 3955 5 3955 5 3955 14 30 27 5 3976 5 3955 5 3955 5 3955 14 30 28 5 3976 5 3955 5 3955 5 3955 14 30 29 5 3976 5 3955 5 3955 5 3955 14 30 30 5 3195 5 3955 5 3955 5 3955 14 30 31 5 3976 5 3955 5 3955 5 3955 14 30 32 5 3976 5 3955 5 3174 5 3955 14 30 33 5 3195 5 3955 5 3174 5 3174 14 30 34 5 3195 5 2393 5 3174 5 3955 14 30 35 5 8195 5 3955 5 3955 5 3955 14 30 36 5 3976 5 3955 5 3955 5 3955 14 30 37 5 3976 5 3955 5 3955 5 3955 14 30 38 5 3976 5 3955 5 3174 5 3174 14 30 39 5 3976 5 3955 5 3955 5 3955 14 30 40 5 3195 5 3955 5 3955 5 3174 14 30 41 5 3976 5 3955 5 2393 5 3955 14 30 42 5 3976 5 3955 5 3955 5 3955 14 30 43 5 3195 5 3955 5 3955 5 3955 14 30 44 5 3195 5 3174 5 2393 5 3174 14 30 45 5 3195 5 3955 5 3174 5 3174 14 30 46 5 3195 5 3955 5 3174 5 3174 14 30 47 5 3195 5 3174 5 2393 5 3174 14 30 48 5 3195 5 3955 5 3174 5 3174 14 30 49 5 3195 5 3955 5 3174 5 3174 30 14 30
116. 5 04 30 5 3976 5 3955 5 3955 5 3955 15 04 31 5 3976 5 3955 5 3955 5 3955 15 04 32 5 3976 5 3955 5 3955 5 4736 15 04 33 5 3976 5 3955 5 3955 5 3955 15 04 34 5 3976 5 4736 5 4736 5 5518 15 04 35 5 3976 5 3955 5 3955 5 3955 15 04 36 5 3976 5 3955 5 3955 5 4736 15 04 37 5 3976 5 3955 5 3955 5 3955 15 04 38 5 3976 5 3955 5 3955 5 3955 15 04 39 5 3976 5 3955 5 3955 5 3955 15 04 40 5 3976 5 4736 5 3955 5 4736 50 15 04 41 5 3976 5 3955 5 3955 5 3955 15 04 42 5 3976 5 3955 5 3955 5 4736 15 04 43 5 3976 5 3955 5 3955 5 3955 15 04 44 5 3976 5 4736 5 3955 5 4736 15 04 45 5 3976 5 3955 5 3955 5 3955 15 04 46 5 3195 5 3955 5 3955 5 3955 15 04 47 5 3976 5 3955 5 3955 5 3955 15 04 48 5 3976 5 3955 5 3955 5 4736 15 04 49 5 3195 5 3955 5 3955 5 4736 15 04 50 5 3195 5 3955 5 3955 5 3955 15 04 51 5 3976 5 3955 5 3955 5 4736 15 04 52 5 3195 5 3955 5 3955 5 3955 15 04 53 5 3976 5 3955 5 3955 5 3955 15 04 54 5 3976 5 3955 5 3955 5 4736 15 04 55 5 3976 5 3955 5 3955 5 4736 15 04 56 5 3976 5 3955 5 3955 5 4736 15 04 57 5 3976 5 3955 5 3955 5 4736 15 04 58 5 3976 5 3955 5 3955 5 4736 15 04 59 5 3976 5 3955 5 3955 5 3955 15 05 00 5 3195 5 3955 5 3955 5 4736 15 05 01 5 3976 5 3955 5 3955 5 3955 15 05 02 5 3195 5 3955 5 3174 5 4736 15 05 03 5 3976 5 3955 5 3955 5 4736 15 05 04 5 3976 5 4736 5 3174 5 3955 15 05 05 5 3195 5 3955 5 3955 5 4736 15 05 06 5 3195 5 3955 5 3955 5 3955 15 05 07 5 3976 5 3955 5 3955 5 4736 15 05 08 5 3195 5 3955 5 3174 5 3955 15 05 09 5
117. 5 2414 5 2393 5 2393 5 2393 14 45 23 5 2414 5 2393 5 2393 5 2393 14 45 24 5 3195 5 2393 5 2393 5 2393 14 45 25 5 3195 5 2393 5 2393 5 2393 14 45 26 5 2414 5 3174 5 2393 5 2393 14 45 27 5 2414 5 2393 5 2393 5 2393 14 45 28 5 3195 5 2393 5 2393 5 2393 14 45 29 5 2414 5 2393 5 2393 5 2393 14 45 30 5 2414 5 2393 5 2393 5 2393 14 45 81 5 2414 5 2393 5 2393 5 2393 14 45 32 5 2414 5 2393 5 2393 5 2393 14 45 33 5 2414 5 2393 5 2393 5 2393 14 45 34 5 2414 5 3174 5 2393 5 2393 14 45 35 5 2414 5 3174 5 2393 5 2393 14 45 36 5 2414 5 2393 5 2393 5 2393 14 45 37 5 2414 5 2393 5 2393 5 2393 14 45 38 5 2414 5 2393 5 2393 5 2393 14 45 39 5 2414 5 2393 5 2393 5 2393 14 45 40 5 2414 5 2393 5 2393 5 2393 14 45 41 5 3195 5 2393 5 2393 5 2393 14 45 42 5 2414 5 2393 5 2393 5 2393 14 45 43 5 2414 5 2393 5 1611 5 2393 14 45 44 5 3195 5 2393 5 2393 5 2393 14 45 45 5 8195 5 2393 5 2393 5 2393 14 45 46 5 2414 5 2393 5 2393 5 2393 14 45 47 5 2414 5 3174 5 2393 5 2393 14 45 48 5 2414 5 2393 5 2393 5 2393 14 45 49 5 3195 5 2393 5 2393 5 2393 38 14 45 50 5 2414 5 2393 5 2393 5 2393 14 45 51 5 2414 5 2393 5 2393 5 2393 14 31 40 5 3195 5 2393 5 2393 5 2393 14 31 41 5 3195 5 3174 5 2393 5 2393 14 31 42 5 3195 5 2393 5 2393 5 2393 14 31 43 5 3195 5 2393 5 2393 5 2393 14 31 44 5 3195 5 2393 5 2393 5 2393 14 31 45 5 2414 5 2393 5 2393 5 2393 14 31 46 5 3195 5 2393 5 2393 5 2393 14 31 47 5 2414 5 2393 5 2393 5 2393 14 31 48 5 3195 5 2393 5 2393 5 2393 14 31 49 5 3195 5 2
118. 5 5 3174 5 5518 5 4736 14 50 18 5 3195 5 3174 5 5518 5 4736 14 50 19 5 8195 5 3174 5 5518 5 4736 14 50 20 5 3195 5 3174 5 5518 5 4736 14 50 21 5 3195 5 3174 5 5518 5 4736 14 50 22 5 8195 5 3174 5 5518 5 4736 14 50 23 5 8195 5 3174 5 5518 5 4736 14 50 24 5 3195 5 3174 5 5518 5 4736 14 50 25 5 8195 5 3174 5 5518 5 4736 14 50 26 5 3195 5 3174 5 5518 5 4736 14 50 27 5 3195 5 3174 5 4736 5 5518 14 50 28 5 3195 5 3174 5 4736 5 5518 14 50 29 5 3195 5 3174 5 4736 5 5518 14 50 30 5 3195 5 3174 5 4736 5 5518 14 50 31 5 3195 5 3174 5 4736 5 5518 14 50 32 5 8195 5 3955 5 4736 5 5518 14 50 33 5 3195 5 3174 5 4736 5 5518 14 50 34 5 3195 5 3955 5 4736 5 5518 14 50 35 5 8195 5 3174 5 4736 5 5518 14 50 36 5 3195 5 3955 5 4736 5 5518 14 50 37 5 3195 5 3955 5 4736 5 5518 14 50 38 5 3195 5 3174 5 4736 5 5518 42 14 50 39 5 3195 5 3955 5 3955 5 4736 14 50 40 5 3195 5 3955 5 3955 5 4736 14 50 41 5 3195 5 3955 5 4736 5 5518 14 50 42 5 8195 5 3955 5 3955 5 5518 14 50 43 5 3195 5 3955 5 3955 5 5518 14 50 44 5 3195 5 3955 5 4736 5 5518 14 50 45 5 3195 5 3955 5 3955 5 4736 14 50 46 5 3195 5 3955 5 4736 5 4736 14 50 47 5 3195 5 3955 5 3955 5 5518 14 50 48 5 3195 5 3955 5 3955 5 4736 14 50 49 5 3195 5 3955 5 3955 5 5518 14 50 50 5 3195 5 3955 5 3955 5 4736 14 50 51 5 3195 5 3955 5 3955 5 4736 14 50 52 5 8195 5 3955 5 3955 5 4736 14 50 53 5 3195 5 3955 5 3955 5 4736 14 50 54 5 3195 5 3955 5 3955 5 5518 14 50 55 5 8195 5 3955 5 3955 5 4736 14 50 56 5 3195 5 3955 5
119. 50 5 3195 5 3955 5 2393 5 3955 14 30 51 5 3195 5 3955 5 3174 5 3174 14 30 52 5 3195 5 3174 5 2393 5 3174 14 30 53 5 2414 5 3174 5 2393 5 3955 14 30 54 5 3195 5 3955 5 3174 5 3174 14 30 55 5 3195 5 3174 5 3174 5 3174 14 30 56 5 2414 5 3174 5 2393 5 3955 14 30 57 5 3195 5 3174 5 3174 5 3174 14 30 58 5 2414 5 3174 5 2393 5 3955 14 30 59 5 3195 5 3174 5 2393 5 3174 14 31 00 5 2414 5 3174 5 2393 5 3955 14 31 01 5 2414 5 3174 5 3174 5 3174 14 31 02 5 2414 5 3174 5 2393 5 3955 14 31 03 5 2414 5 3174 5 3174 5 3174 14 31 04 5 3195 5 3174 5 2393 5 3174 14 31 05 5 2414 5 3955 5 2393 5 3955 14 31 06 5 2414 5 3174 5 2393 5 3174 14 31 07 5 2414 5 3174 5 2393 5 3955 14 31 08 5 3195 5 3174 5 2393 5 3174 14 31 09 5 2414 5 3174 5 2393 5 3955 14 31 105 3195 5 3174 5 2393 5 3174 14 31 11 5 2414 5 3174 5 2393 5 3955 14 31 12 5 2414 5 3174 5 3174 5 3174 14 31 13 5 2414 5 3174 5 3174 5 3174 14 31 14 5 3195 5 3174 5 2393 5 3174 14 31 15 5 2414 5 3174 5 2393 5 3955 14 31 16 5 3195 5 3174 5 2393 5 3955 14 31 17 5 2414 5 3174 5 2393 5 3174 14 31 185 2414 5 3174 5 2393 5 3174 14 31 19 5 2414 5 2393 5 2393 5 2393 14 31 20 5 2414 5 2393 5 2393 5 3174 14 31 21 5 2414 5 3174 5 2393 5 3174 14 31 22 5 2414 5 2393 5 2393 5 2393 14 31 23 5 2414 5 2393 5 2393 5 2393 14 31 24 5 2414 5 3174 5 2393 5 3174 14 31 25 5 2414 5 3174 5 2393 5 2393 14 31 26 5 2414 5 2393 5 2393 5 2393 14 31 27 5 3195 5 2393 5 2393 5 2393 14 31 28 5 3195 5 2393 5 2393 5 2393 14 31 29 5 3195 5 3174 5 3
120. 55 14 51 36 5 3195 5 3955 5 3955 5 3955 14 51 37 5 3195 5 3955 5 3174 5 3174 14 51 38 5 8195 5 3955 5 3955 5 3955 14 51 39 5 3195 5 3174 5 3174 5 3955 14 51 40 5 3976 5 3955 5 3174 5 3174 14 51 41 5 3195 5 3955 5 3174 5 3174 14 51 42 5 3195 5 3174 5 3174 5 3174 14 51 43 5 3195 5 3174 5 3174 5 3955 14 51 44 5 8195 5 3955 5 3955 5 3955 14 51 45 5 8195 5 3174 5 2393 5 3955 14 51 46 5 2414 5 3174 5 2393 5 3955 14 51 47 5 2414 5 3174 5 3174 5 3955 14 51 48 5 2414 5 3174 5 3174 5 3174 14 51 49 5 2414 5 3955 5 2393 5 3955 14 51 50 5 3195 5 3174 5 3174 5 3955 14 51 51 5 3195 5 3174 5 2393 5 3955 14 51 52 5 8195 5 3955 5 2393 5 3955 14 51 53 5 2414 5 3174 5 2393 5 3955 14 51 54 5 8195 5 3174 5 2393 5 3955 14 51 55 5 2414 5 3174 5 3174 5 3174 14 51 56 5 2414 5 3174 5 3174 5 3174 14 51 57 5 2414 5 3955 5 2393 5 3955 14 51 58 5 8195 5 3174 5 2393 5 3955 14 51 59 5 2414 5 3174 5 2393 5 3955 14 52 00 5 2414 5 3174 5 3174 5 3174 14 52 01 5 2414 5 3174 5 2393 5 3174 14 52 02 5 2414 5 3174 5 2393 5 3955 14 52 08 5 2414 5 3174 5 3174 5 3174 14 52 04 5 2414 5 3174 5 2393 5 3955 14 52 05 5 2414 5 3174 5 2393 5 3955 14 52 06 5 2414 5 3174 5 3174 5 3174 14 52 07 5 3195 5 3174 5 2393 5 3955 14 52 08 5 3195 5 3174 5 2393 5 3174 14 52 09 5 2414 5 3174 5 2393 5 3955 14 52 10 5 2414 5 3174 5 3174 5 3174 14 52 11 5 3195 5 3174 5 2393 5 3955 14 52 12 5 2414 5 3174 5 2393 5 3955 14 52 18 5 2414 5 3955 5 2393 5 3955 14 52 14 5 2414 5 3174 5 2393 5 3955 44 14 52
121. 7 5 1633 5 1611 5 0830 5 0830 14 27 48 5 0851 5 16115 16115 1611 14 27 49 5 0851 5 1611 5 2393 5 1611 14 27 50 5 0851 5 1611 5 2393 5 2393 14 27 51 5 0851 5 16115 2393 5 2393 14 27 52 5 0851 5 16115 2393 5 2393 14 27 53 5 0851 5 1611 5 2393 5 2393 14 27 54 5 0851 5 0830 5 2393 5 2393 14 27 55 5 0851 5 0830 5 2393 5 2393 14 27 56 5 0851 5 0830 5 2393 5 2393 14 27 57 5 0851 5 0830 5 2393 5 2393 14 27 58 5 0851 5 0830 5 2393 5 2393 14 27 59 5 0851 5 1611 5 2393 5 2393 14 28 00 5 0851 5 1611 5 2393 5 2393 14 28 01 5 1633 5 1611 5 2393 5 1611 14 28 02 5 1633 5 1611 5 2393 5 2393 14 28 03 5 0851 5 1611 5 2393 5 2393 14 28 04 5 0851 5 1611 5 3955 5 2393 14 28 05 5 0851 5 1611 5 3955 5 2393 14 28 06 5 0851 5 1611 5 3955 5 2393 14 28 07 5 0851 5 2393 5 3955 5 2393 14 28 08 5 0851 5 1611 5 3955 5 2393 14 28 09 5 0851 5 1611 5 3955 5 2393 14 28 10 5 1633 5 1611 5 3955 5 3174 14 28 11 5 1633 5 2393 5 3955 5 3955 14 28 12 5 1633 5 2393 5 3955 5 3955 14 28 13 5 1633 5 2393 5 3955 5 3174 14 28 14 5 1633 5 2393 5 3955 5 3174 14 28 15 5 1633 5 1611 5 3955 5 3955 14 28 16 5 1633 5 16115 3955 5 3955 14 28 17 5 1633 5 2393 5 3955 5 3955 14 28 18 5 1633 5 2393 5 4736 5 3955 14 28 19 5 1633 5 1611 5 4736 5 3955 14 28 20 5 1633 5 1611 5 4736 5 3955 14 28 21 5 1633 5 1611 5 4736 5 3955 14 28 22 5 1633 5 1611 5 4736 5 3955 14 28 23 5 1633 5 16115 4736 5 3955 14 28 24 5 1633 5 2393 5 4736 5 3955 14 28 25 5 1633 5 16115 4736 5 3955 Max Values from rows 1 50 5 3
122. 74 5 3955 14 44 04 5 2414 5 3174 5 3955 5 3955 14 44 05 5 3195 5 3174 5 3174 5 3955 14 44 06 5 3195 5 3174 5 3174 5 3174 14 44 07 5 3195 5 3955 5 3174 5 3174 14 44 08 5 3195 5 3174 5 3174 5 3955 14 44 09 5 3195 5 3174 5 2393 5 3955 14 44 10 5 2414 5 3174 5 2393 5 3955 14 44 11 5 3195 5 3174 5 3174 5 3174 14 44 12 5 8195 5 3174 5 3955 5 3955 14 44 18 5 2414 5 3955 5 3955 5 3955 36 14 44 14 5 2414 5 3955 5 3955 5 3955 14 44 15 5 2414 5 3174 5 2393 5 3955 14 44 16 5 2414 5 3174 5 3174 5 3174 14 44 17 5 2414 5 3955 5 2393 5 3955 14 44 18 5 2414 5 3174 5 2393 5 3955 14 44 19 5 2414 5 3174 5 2393 5 3955 14 44 20 5 2414 5 3174 5 3174 5 3955 14 44 21 5 3195 5 3174 5 2393 5 3174 14 44 22 5 2414 5 3174 5 2393 5 3955 14 44 23 5 2414 5 3174 5 2393 5 3955 14 44 24 5 2414 5 3174 5 3174 5 3174 14 44 25 5 2414 5 3174 5 2393 5 3955 14 44 26 5 2414 5 3174 5 2393 5 3955 14 44 27 5 2414 5 3174 5 2393 5 3955 14 44 28 5 2414 5 3174 5 2393 5 3955 14 44 29 5 3195 5 3174 5 2393 5 3174 14 44 30 5 2414 5 3174 5 2393 5 3955 14 44 31 5 2414 5 3174 5 2393 5 3955 14 44 32 5 2414 5 3174 5 2393 5 3955 14 44 33 5 2414 5 3174 5 2393 5 3955 14 44 34 5 2414 5 3174 5 2393 5 3955 14 44 35 5 2414 5 3174 5 2393 5 3955 14 44 36 5 2414 5 3174 5 2393 5 3955 14 44 37 5 2414 5 3174 5 3174 5 3174 14 44 38 5 2414 5 3174 5 2393 5 3955 14 44 39 5 2414 5 3174 5 3174 5 3174 14 44 40 5 2414 5 3174 5 2393 5 3955 14 44 41 5 2414 5 3174 5 2393 5 3955 14 44 42 5 2414 5 3174 5 2393 5 3955
123. 8 data bus 7 data bus 6 data bus 5 data bus 4 data bus 3 data bus 2 data bus 1 data bus 0 Tri state the data bus TRI B O enable16 TRI data 15 TRI data 14 TRI data 13 TRI data 12 TRI data 11 shitbrick TRI data 10 shitbrick TRI data 9 shitbrick TRI data 8 shitbrick TRI data 7 thats a read TRI data 6 thats a read TRI data 5 thats a read TRI data 4 thats a read TRI data 3 thats a read TRI data 2 thats a read TRI data 1 thats a read TRI data 0 thats a read shitbrick shitbrick shitbrick shitbrick 0 0 0 0 0 Bitsoi Lamb Willems q CTRL R PWM q CTRL L PWM q gt CTRL L PWM q CTRL R PWM q CTRL L PWM 49 RTLinux On A Small Mobile Robot dig filter tdf Digital Filter dP ge oe Bill Willems 1 3 02 oe o oe SUBDESIGN Dig Filter clk In Out VARIABLE INPUT INPUT OUTPUT Digital Filter Machine MACHINE BEGIN WITH STATES s0 Digital Filter Machine clk clk CASE Digital Filter Machine IS WHEN WHEN WHEN WHEN WHEN WHEN s0 gt Out VCC IF In GND THEN Digital Filter Machine ELSE Digital Filter Machine END IF sl gt Out VCC IF In GND THEN Digital Filter Machine ELSE Digital Filter Machine END IF s2 gt Out VCC IF In GND THEN Digital Filter Machine ELSE Digital Filter Machine END IF s3
124. 9 38 5 2414 5 2393 5 3955 5 3955 14 49 39 5 1633 5 2393 5 3955 5 3955 14 49 40 5 1633 5 2393 5 3955 5 3955 14 49 41 5 1633 5 2393 5 4736 5 3955 14 49 42 5 1633 5 2393 5 4736 5 3955 14 49 43 5 1633 5 2393 5 4736 5 3955 14 49 44 5 1633 5 2393 5 4736 5 3955 14 49 45 5 1633 5 2393 5 4736 5 3955 14 49 46 5 1633 5 2393 5 4736 5 3955 14 49 47 5 1633 5 2393 5 4736 5 3955 14 49 48 5 1633 5 2393 5 4736 5 3955 14 49 49 5 1633 5 2393 5 4736 5 3955 14 49 50 5 1633 5 2393 5 4736 5 3955 41 14 49 51 5 1633 5 2393 5 4736 5 3955 14 49 52 5 1633 5 2393 5 5518 5 3955 14 49 53 5 2414 5 3174 5 4736 5 3955 14 49 54 5 2414 5 3174 5 5518 5 4736 14 49 55 5 2414 5 2393 5 5518 5 4736 14 49 56 5 2414 5 3174 5 5518 5 4736 14 49 57 5 2414 5 2393 5 6299 5 4736 14 49 58 5 2414 5 2393 5 6299 5 4736 14 49 59 5 2414 5 2393 5 6299 5 4736 14 50 00 5 2414 5 2393 5 6299 5 4736 14 50 01 5 3195 5 2393 5 6299 5 5518 14 50 02 5 3195 5 3174 5 5518 5 4736 14 50 03 5 3195 5 2393 5 6299 5 5518 14 50 04 5 3195 5 3174 5 6299 5 5518 14 50 05 5 3195 5 2393 5 6299 5 5518 14 50 06 5 3195 5 3174 5 6299 5 5518 14 50 07 5 3195 5 3174 5 5518 5 4736 14 50 08 5 3195 5 3174 5 5518 5 4736 14 50 09 5 3195 5 3174 5 5518 5 4736 14 50 10 5 3195 5 3174 5 5518 5 4736 14 50 11 5 3195 5 3174 5 5518 5 4736 14 50 12 5 3195 5 3174 5 5518 5 4736 14 50 13 5 3195 5 3174 5 5518 5 4736 14 50 14 5 3195 5 3174 5 5518 5 4736 14 50 15 5 3195 5 3174 5 5518 5 4736 14 50 16 5 3195 5 3174 5 5518 5 4736 14 50 17 5 819
125. 90 909 0 00008 0 82 12 0 00012 6833 333 0 88 13 0 00013 6769 231 ru 0 96 14 0 00014 6857 143 0 00004 1 18 15 0 00015 7866 667 0 00002 1 22 16 0 00016 7525 1 25 17 0 00017 7352 941 7111 111 6894 737 water I vs R 20000 15000 10000 5000 E 0 00002 0 00004 0 00006 0 00008 0 0001 0 00012 0 00014 0 00016 0 00018 0 0002 water Vv vs R Fig 11 Sensor Test for drv dirt vi R R 100000 0 62 0 65 0 0000065 95384 62 0 78 0 65 0 000065 120000 0 82 0 7 0 000007 117142 9 0 95 0 95 0 000095 100000 13 1 4 0 000014 85714 29 14 1 8 0 000018 77777 78 1 5 2 2 0 000022 68181 82 12 3 0 00003 56666 67 1 8 3 5 0 000035 51428 57 19 4 0 00004 47500 2 4 7 0 000047 42553 19 24 5 5 0 000055 38181 82 2 2 6 5 0 000065 33846 15 23 7 1 0 000071 32394 37 25 8 8 0 000088 28409 09 2 6 9 8 0 000098 26530 61 28 11 4 0 000114 24561 4 29 12 2 0 000122 23770 49 3 13 0 00013 23076 92 32 144 0 000144 2222222 3 33 15 5 0 000155 21483 87 3 45 16 41 0 0001641 21023 77 3 6 17 3 0 000173 20809 25 dry dirt Vv vs I 0 0002 0 00018 0 00016 0 00014 0 00012 0 0001 0 00008 0 00006 0 00004 05 1 15 2 25 3 35 E dry dirt vs R 120000 100000 80000 60000 40000 20000 0 0 00002 0 00004 0 00006 0 00008 0 0001 0 00012 0 00014 0 00016 0 00018 0 0002 dry
126. 93 14 32 09 5 3195 5 2393 5 2393 5 2393 14 32 10 5 2414 5 2393 5 2393 5 2393 14 32 11 5 3195 5 2393 5 2393 5 2393 14 32 12 5 3195 5 2393 5 2393 5 2393 14 32 13 5 3195 5 2393 5 2393 5 2393 14 32 14 5 2414 5 2393 5 2393 5 2393 14 32 15 5 3195 5 2393 5 2393 5 2393 14 32 16 5 3195 5 2393 5 2393 5 2393 14 32 17 5 3195 5 2393 5 2393 5 2393 14 32 18 5 2414 5 2393 5 2393 5 2393 14 32 19 5 2414 5 3174 5 2393 5 2393 14 32 20 5 2414 5 2393 5 2393 5 2393 14 32 21 5 3195 5 2393 5 2393 5 2393 14 32 22 5 3195 5 2393 5 1611 5 2393 14 32 23 5 3195 5 2393 5 2393 5 2393 14 32 24 5 3195 5 2393 5 2393 5 2393 14 32 25 5 2414 5 2393 5 1611 5 2393 46 14 32 26 5 3195 5 2393 5 2393 5 2393 14 32 27 5 2414 5 2393 5 1611 5 2393 14 32 28 5 2414 5 2393 5 1611 5 2393 14 32 29 5 8195 5 2393 5 2393 5 2393 14 32 30 5 2414 5 2393 5 1611 5 2393 14 32 31 5 2414 5 2393 5 1611 5 2393 14 32 32 5 2414 5 2393 5 1611 5 2393 14 32 33 5 2414 5 3174 5 2393 5 2393 14 32 34 5 2414 5 2393 5 1611 5 2393 14 32 35 5 2414 5 2393 5 1611 5 2393 14 32 36 5 2414 5 2393 5 1611 5 2393 14 32 37 5 2414 5 2393 5 2393 5 2393 14 32 38 5 2414 5 2393 5 1611 5 2393 14 32 39 5 2414 5 2393 5 1611 5 2393 14 32 40 5 2414 5 2393 5 1611 5 2393 14 32 41 5 2414 5 2393 5 2393 5 2393 14 32 42 5 2414 5 2393 5 2393 5 2393 14 32 43 5 2414 5 2393 5 2393 5 2393 14 32 44 5 2414 5 2393 5 1611 5 2393 14 32 45 5 2414 5 2393 5 2393 5 2393 14 32 46 5 2414 5 2393 5 2393 5 2393 14 32 47 5 2414 5 2393 5 1611 5 2393 14 32
127. A NE KEW cb e euo TU Ve ED ERO DN 27 Ded R ndy Send oido RT Do EDD DUI ENIM Na ded EATER 27 3 3 1 Planter Design iescost eese nrbe ty PEN E E TE bare EET R ERR 28 Sd Pufhpaseccett o brote eee tu Ei Vt Ex bre UA 28 3 3 3 Solar Panel eere eI er Rer E vals Rue cp eoe ers 29 3 34 Charre C ICH S Gs cerca rE EA nag Y RR ER CE NUR ER UAR ude ENARRARE E REN RR 30 3 3 5 Adjustable Program S Wiel 2355s edr eta o ea PR roO ero Nux Ta SR TU DES 31 4 0 Suminiary COnCIUSIONS lt 2 sceu astarvusyexvessuousaue RUE ETET TENURE R CLEVER NA CQ A FEM ES 32 SO TA edic DRM EE pa See nate EE E EEEE EEEN E naa 33 UN LAPpendi x Aig pcvsieccusces ii iese TE EEE EEEE E E E E EEE E 46 62 Appendix Borer bte E AE A d Ed IUE 50 GF Appendix OUR R 51 List of Tables Table 1 Performance Characteristics of Batteries Table 2 Comparison of Batteries and Capacitors List of Figures Figure 1 5V Low Dropout Voltage Regulator sese 34 Figure 2 Adjustable Voltage Switch cesses mene 34 Figure 3 Charge Circuits sier eor heri x PE tra VOR DE te ee PUR mast URP ee ed PRU dE 34 Figure 4 Moistite Sensob ooo P EE IRR TOES UN d PUR deca TECH 35 Ligure 5 lieht CIECUIL cei xsesese ove RESET vee Reo ae EMEN an Sete VER RGB ee MD Qn 35 Figure 6 HCI2 Layouten oerda ea a eoe Ea ERS E 36 Figure 7 Planter Design i cet dst IR EEEE EEE E EE EA Ye EN EUER S 37 Figure Program Flow Cro eso aS EYES ERR NOR EER DONA Ree ERE UIS 38 igure 9 Plan
128. C angle sensor capable of seeing tilt relative to the X and Y planes Application The sensor is mounted to the brake shaft and when the user applies pressure to the brakes the shaft tilts and changes the sensor output a few tenths of a volt when a 5 10 degree tilt is seen on the brake lever This change is fed into a Schmitt Trigger and the signal would be conditioned and sent to the transmitter box to be relayed out to the receiver box Final Status The sensor initially looked good during initial prototype testing but when a vehicle accelerates or there is weight in the back of the tow vehicle the tilt seen by the sensor is great enough to emit false signals This sensor had to be discarded and the other brake sensors were further developed 3 2 Turn Signal Sensor Development The turn signal sensor would need to sense whether a turn is being initiated and relay back in which direction the turn is intended This sensor can in no way hinder the drivers usual driving habits or give the driver any added responsibility 3 2 1 Toggle Switch Description This is a basic three way toggle switch with a very long shaft Application The base of the switch is attached to the steering column while the shaft of the toggle switch is attached to the turn signal shaft As the turn signal shaft is moved up and down the toggle switch moves with it The correct signal is then relayed to the control box indicating which direction the driver is t
129. Electrical Engineering Department New Mexico Tech Presents the Theses From the Course Senior Design Project EE 481 amp EE 482 for 2001 2002 Instructor Scott W Teare Assistant Professor of Electrical Engineering Preface The Senior Design Project course in the Electrical Engineering Department at New Mexico Tech is a two semester course that provides students with the opportunity to develop a technical project in Electrical Engineering or closely aligned discipline while building their project management skills This course introduces the students to working in teams developing a project s statement of work reporting milestones presentations design reviews status reports and of course bringing in a project on time and under budget In this year s class the students were presented with 20 projects from industry Department of Defense and faculty The students rated each project by their interests and then were assigned to the projects nominally in groups of 3 based on their choices combined with a lottery to break ties As a result 9 projects were undertaken The final tasks in this course are for the students to present to the faculty their sponsors and their peers a talk and a thesis on the work the accomplished over the past 8 months This document is a collection of the thesis prepared by the students describing their projects The page numbers in this document are set for each thesis As such the easiest way
130. Excalibur Board Data Constants Input Data TxRx clock Computer Output Data Figure 5 Block Diagram of PC Interface to the Processing Element As shown data is transferred to and from the Altera board in 4 bit packets When data is sent to the Altera board it is stored in temporary 4 bit registers Once all the input data to the Altera board has been sent it is transferred to the PE and clocked into it s internal registers This immediately produces a new result This result is sent back to the computer There are four different transmission states send configuration bits send data constants send input data and receive output data Two bits TxRx_statel and TxRx stateO are used to represent the transmission state Table 1 on the next page shows the values of TxRx statel and TxRx_stateO that correspond to each transmission state 12 TxRx statel TxRx state0 Transmission State 0 0 Send Configuration Bits PC PE 0 1 Send Data Constants PC PE 1 0 Send Input Data PC PE 1 1 Receive Output Data PE PC Table 1 Interface Data Transmission States Finally there are two clock signals that run from the computer to the Altera board The first clock signal TxRx clock is used to control the 4 bit transmission and clock in the data that is being transferred to the Altera board to the temporary 4 bit registers The second clock signal PE clock is used once all the data has been s
131. IGN eere eene ene ttntn nitet tnt tnt nont tnan 11 SA INTRODUCTION 2 ridus E E EEE e E E RE E ences pus educa ub edere 11 3 2 SYNTHESIZING THE PROCESSING ELEMENT eene 11 3 3 INTERFACING TO THE PROCESSING ELEMENT eee 11 3 4 PROGRAMMING THE ALTERA CHIP eese enne entente 16 CHAPTER 4 TEST SOFTWARE DESIGN it iet Etro edita ported nip iiaiai 18 TI UINTRODUCTIONS C o ostenta nitide etta tu tip tte teta Vu ees iode 18 4 2PC HARDWARE INTERFACE cenna o otadpte eani dete referee do doses kei adole 18 4 3 S0FTIWAREDESIGN2a idend ue o na rr dged aote p nde i ia qud geo n UNE RB ead 19 44 TEST CASE GENERA TION u a cote toe egestate opes bic aea Ei peu aod ee 2l CHAPTER 5 RESULTS amp CONCLUSIONS 5 etaescicce cutus aorta terc Yvon oben Vea eene eR R N 25 REFERENCES ona a a e a tende Ea E A a EREE 26 LIST OF FIGURES Figure 1 Block Diagram of the Reconfigurable Data Path Processor eee 5 Figure 2 Block Diagram of Processing Element Highlighting all I O amp Control Signals 6 Figure 3 Block Diagram of Processing Element Highlighting the Data Paths 7 Figure 4 The Altera Excalibur Board d oa a vei et es 9 Figure 5 Block Diagram of PC Interface to the Processing Element eee 12 Figure 6 Initial Design to Route the Incoming Data eene 13 Figure 7 Initial Design f
132. INPUT L pos 15 0 INPUT PC 104 specific output lines IOCS16 BIDIR Asserted by us to allow a 16bit transfer instead of 2 8bit transfers Robot specific output lines R pwm 7 0 OUTPUT L pwm 7 0 OUTPUT Bidirectional Lines data_bus 15 0 BIDIR PC 104 Data Bus VARIABLE R_pwm_latch 7 0 DFF PWM counters compare to these L pwm latch 7 0 DFF addr latch 9 0 DFF latch what s on the addr Bus data 15 0 NODE Internal Data Bus thats_a_read NODE Is the SBC reading from us hey_thats_us NODE Are you talkin ta me latch R pwm NODE use these to clock the PWM 47 RTLinux On A Small Mobile Robot BEGIN Bitsoi Lamb Willems latch L pwm NODE latches enablel6 NODE tri state enable for IOCS16 shitbrick NODE tri state enable for the upper half of the data bus during 16 bit transfers DEFAULTS latch R pwm VCC latch L _ pwm VCC thats a read GND hey thats us GND enablel6 GND shitbrick GND END DEFAULTS Setup clocks and clears R pwm latch l clk latch R pwm L pwm latch l clk latch L pwm addr latch clk BALE R pwm latch clrn RESET L pwm latch clrn RESET addr latch clrn RESET wire latch inputs R pwm latch d data bus 7 0 L pwm latch d data bus 7 0 addr latch l d addr wire pwm latch outputs R pwm R pwm latch q L
133. IZING THE PROCESSING ELEMENT The first step we undertook to synthesize the Processing Element was to convert Dr Donohoe s VHDL design files to Quartus II Block Symbol Files BSF These BSF files were then placed in a Quartus II Block Diagram File BDF and interconnected as specified by the block diagram of the Processing Element showed in Figures 1 and 2 Due to the magnitude of this file it cannot be included in this report It is however available on our project CD We then compiled this BDF file with the APEX20K200E chip specified as the target device This took about fifteen minutes on average and compiled correctly once various bugs were eliminated The Processing Element used 57 of the chip s total resources INTERFACING TO THE PROCESSING ELEMENT We decided to develop the interface between a computer and the Processing Element using the parallel port We briefly considered using a Data Acquisition card but they are expensive as well as complicated to set up and use The parallel port on the other hand exists on every computer thus requiring no additional equipment costs and making our design portable In addition to this we are both intimately familiar with how to setup and 11 use the parallel port due to coursework in EE 352 Microcomputer Interfacing offered at New Mexico Tech Figure 5 below shows a block diagram of the interface we designed between the computer and the Processing Element on the Altera board Altera
134. Interrupt Request Industry Standard Architecture Interrupt Service Routine kilobyte 1 000 bytes Logical Block Addressing Light Emitting Diode milliamps 0 1 Amps MegaByte 1 000 000 bytes Mega bits per second Megahertz 1 000 000 hertz Operating System Personal Computer Personal Computer Interface 104 pin connector Peripheral Component Interconnect Serial Port 2 Generation Pulse Width Modulation Random Access Memory Red Green Blue Read Only Memory Real Time Single Board Computer Synchronous Dynamic Random Access Memory Universal Serial Bus Volts RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Introduction This project is the basis of our capstone course at New Mexico Tech We selected this project based on our interests and our skills at the beginning of this year The purpose of this project is to design and implement a Robot which would use RT Linux on a Single Board Computer to bring in sensor data and respond to the environment based on the sensor data Our client has also asked that we do some image processing using two USB cameras such as vertical and horizontal line detection On the robot a wide variety of sensors would be used to get a good sense of the environment and the robot could respond to these variables accordingly We were also to use the SBC along with wireless networking in order to have feedback from the robot as it ran its different tasks A possible future use of the work we com
135. K K K K K K K K K K K K K K K K K K K K Date Wed 17 Oct 2001 16 05 10 0600 MDT From Clint Janes lt cjanes aoc nrao edu gt To rharriso zia aoc NRAO EDU Cc brajala nmt edu gbivens nmt edu sfield 9 nmt edu Subject Project Hi Robyn I am so pleased at the progress so far of the student project I suggest having the Visitor Center Committee meet at Tech next time so that the students can demo and get additional suggestion Let me know if you agree and I can set up Tuesday afternoon is a good time for them Clint 43 LLELLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLDLLLLLLLEL Date Wed 10 Apr 2002 17 18 29 0600 From Robyn Harrison rharriso 9 cv3 cv nrao edu To Clint Janes cjanes 9 zia aoc NRAO EDU Sam Field lt sfield nmt edu gt Gerald Bivens gbivens nmt edu gt Brian lt brajala nmt edu gt Dr Shanechi lt shanechi ee nmt edu gt Subject SRT installation at site Hi The fence went out to the site today I think Charley plans to put it up either tomorrow Friday or Monday at the latest so it should be ready for you to move the dish itself out next week We can t install the internet cable until we have an actual cabinet to put the I forget what Fred told me goes between the end of the cable and the computer in but anyway it would be great if we could go ahead and test it w o the fiber link though I know we need to update the clock etc Let me know and we can plan a time we can s
136. Managed night shift Related Course Work Microcontrollers Electromagnetic Wave Transmission and Radiation Intermediate Control Theory Analog Electronics Adobe Photoshop Computer Skills LabView Matlab 6 Altera Max plus II C C Adobe Photoshop MSWord MSExcel MSAccess Protel 99 Honors Placed 7 in the Trinity College Fire Fighting Home Robot Contest 57 Wireless Vehicle Signaling System A Thesis Presented to the Faculty of the Electrical Engineering Department of New Mexico Tech By David Byrd Brandon Lattimore Ivan Olguin In partial fulfillment of the requirements for the course EE 482 Senior Design II May 2002 2002 David Byrd Brandon Lattimore Ivan Olguin Abstract A major disadvantage to conventional trailer hookups is that they rely on hardwiring the tow vehicle wiring to trailer wiring The Wireless Signaling System will eliminate the need for consumers to struggle with faulty wiring and bad grounds common on trailers today The Wireless Signaling System Team is responsible for designing a universal system to bypass the normal trailer wire assemblies on current trailers and replace them with an innovative RF wireless control and LED array system The system is portable and can be installed with simple hand tools allowing the average consumer to simplify the amount of wiring necessary to setup a trailer signaling system After completing an extensive literature search on the necessary components the protot
137. P This component clips the 48 bit operand to a 24 bit result It also produces a four bit condition word that represents if the result is negative or zero or if an overflow or underflow occurred CMUX This component is a conditional multiplexer It can either unconditionally select one of the two inputs or select one based on the condition word that is provided to it MUX This component is a simple multiplexer that selects one of the two inputs and places it on the output depending on the value of the select line The PE contains both 24 bit and 48 bit multiplexers and one 4 bit multiplexer for selecting the condition word from the multiplication or arithmetic path PAD This component pads a 24 bit number with Os or 1s to make it into a 48 bit number REGISTER There are five 24 bit registers two for the data constants DRO amp DR1 and three for the data inputs RINO RIN1 amp RIN2 ROUND This component rounds its input or simply passes it through depending on the value of set round SHIFT This component is just like the AL SHIFT except that it also has the capability of performing a circular shift XBARS5x3 This component has five 24 bit inputs three 24 bit outputs and three 3 bit select lines Each of the outputs can select one of the five inputs For a more thorough description of the RDPP and its operation refer to RDPP Overview by Dr Greg Donohoe ALTERA EXCALIBUR BOARD AND QUARTUS II SOFTWARE Th
138. RC ICHECK RF LF ITRACE OFF A 2 Appendix B Truth table of the inputs and outputs used for PALASM code Design 1 with flash and brake 5 chips and 12 gates INPUTS OUTPUT S Czcontinous intensity F flash B Brake intensity X Off Here is the design with Brake over riding the flash This design adds a few gates and chips to the previous design Appendix C Transmitter circuit implemented in Protel 88485555 nn NES griibiikil Appendix D Heceiver circuit implemented in Protel APPENDIX E PCB layout of entire system This layout comprises of the transmitter and receiver boards Appendix F Output Logic that was later implemented on programmed chip Use J pue eyxeJg S9je5 e pue sdiu v APPENDIX G Schematics of Hall effect sensor circuit and board layout Hallogenic Hall effect sensor Printed Circuit Board Connector 5 gnd sigout Printed circuit board for hall effect sensor APPENDIX H Phototransistor circuit schematic that includes a Max 492 Op Amp and a phototransistor Phototransistor Circuit 10K ohm u PN168 10K POT Output APPENDIX I Distance sensor layout along with the schmitt trigger circuit layout and printed circuit board design from GPD 120 To transmitter box LM 319 comparator schematic
139. Results which at first was discouraging since our power meter was only returning values in the 2 3 to 3 3 watt range After the trials the laser operator placed a power meter in front of the beam to take an actual power reading The result from this measurement was 2 42 watts When compared to our data we had a 9 7 percent error Not only were we pleased with the results but the customer was also A copy of the data can be found in appendices A 4 through A 7 4 1 Discussion of Results The results were pretty much what we were expecting The plates heated up and cooled down as the charts in appendix A 3 show The heat up time lasts from zero to fifty seconds After fifty seconds the laser was turned off and the cool down phase began The series one and two lines represent the sensors that were on the back third of the beam dump They appeared to heat up the most The other two sensors were on the 18 out third of the beam dump and did not heat up quite as much as the first two All of the sensors appeared to cool down to the same value between trials The actual power value was only computed from the first fifty cells which were when the laser was on From those cells we found the overall change in temperature multiplied it by the mass and specific heat of aluminum and then divided it by the time or fifty seconds The final values can be seen in either Table 2 or in appendices A 4 through A 7 4 2 Laser Beam Dump Summary The body of th
140. Select an loom f V u Figure 3 3 1a Figure 3 3 1b Figure 3 3 1c Figure 3 3 1 The Main Layouts There are three slideshow presentations One slideshow will run when the Sun Cygnus A or Cassiopeia is selected on the touchscreen The SRT takes about five minutes to track a celestial object This slideshow is set up to keep people interested during this tracking down time To run two programs at the same time two dos prompts must be open The mouse pointer is first positioned on a DOS prompts to run the java program Then mouse pointer is then positioned on the other DOS prompt to run the PowerPoint presentation The mouse pointer is positioned using the touchscreen software Once the PowerPoint presentation is finished the java software will run until a user selects the exit 21 button on figure 3 3 2 The exit button will exit the java program and load the main layout Figure 3 3 2 The Exit Layout The other two slideshows are accessed when a user selects slide show on one of the main layouts Both How the SRT works and The Science of Radio Astronomy buttons on figure 3 3 3 will run the PowerPoint presentation and load the layout on figure 3 3 4a The back to main button will sent the user back to the main layout The layout on figure 3 3 4a will allow the user to automatically go through the presentation or go to the next slide The automatic button on figure 3 3 4a goes through the selected slideshow at a timed
141. T will be tracking for example the Sun Cygnus A and Cass A which are strong radio sources Figure 1 3 1 Small Radio Telescope on top of Workman Center NRAO ordered an SRT kit with the intention of installing it at the VLA Visitors Center for use by the public as an interactive tutorial on radio astronomy As furnished by MIT the kit is not suitable for this application for a number of reasons First the control software is not intuitive and requires a significant amount of time to learn Second a mouse and keyboard are needed to run the telescope both of which are subject to theft and damage and allow too much access to the computer running the telescope Third the software is intended for users with significant knowledge of radio astronomy and is therefore of little value in educating the general public Therefore our project was to modify the kit to make it suitable for NRAO s needs This involved making it easy for the public to use The intent of the project is to demonstrate the functional operation of a radio telescope and provide a tutorial on radio astronomy for the public 2 Overview of Radio Astronomy Project 2 1 Customer requirements In order to find out what the project was all about we made contact with Clint Janes Electronics Division Head and Robyn Harrison Educational Liaison for NRAO We first met with Robyn Harrison and discussed NRAO s needs We found that we would be furnished with a kit to build a radio t
142. TLinux On A Small Mobile Robot Bitsoi Lamb Willems Appendix E Features of the Prometheus SBC System Features Processor 486 DX2 processor running at 100MHz with co processor Pentium class platform including burst mode SDRAM and PCl based IDE controller and USB 32MB SDRAM system memory 50MHz memory bus for improved performance 2MB 16 bit wide integrated flash memory for BIOS and user programs 8KB unified level 1 cache VO 4 serial ports 115 2kbaud max 2 ports 16550 compatible 2 ports 16850 compatible with 128 byte FIFOs 2 full featured powered USB ports 1 ECP compatible parallel port Floppy drive connector IDE drive connector 44 pin version for notebook drives Accepts solid state flashdisk modules directly on board 10BaseT full duplex PCI bus mastering Ethernet 10Mbps IrDA port requires external transceiver PS 2 keyboard and mouse ports Speaker LEDs Other System Features Plug and play BIOS with IDE autodetection 32 bit IDE access and LBA support Built in fail safe boot ROM for system recovery in case of BIOS corruption User selectable COM2 terminal mode On board lithium backup battery for real time clock and CMOS RAM ATX power switching capability Programmable watchdog timer Power surge monitor for fail safe operation Zero wait state capability for flash memory and PC 104 bus 5V only operation Exten
143. This sensor is in working condition and has been tested on a vehicle It is a little awkward at first but after a while it feels a lot like the original brake pedal It is a very reliable sensor This sensor is depicted in Figure 6 Figure 6 Push button switch 3 1 3 Lever Switch Description This is a basic lever switch with a decent size lever on it Five volts is outputted when the lever is pressed Application The idea behind this one was to mount the base on the shaft of the brake pedal and it would be pushed up against the back firewall as the brake pedal was pushed This is a good application because it is totally out of the way of the driver s foot Final Status We had problems locating a good place on the brake pedal shaft where the switch would make contact when the pedal was pushed due to the severe differences in vehicles brake mounting among different models Because of this issue this switch didn t make it past the initial selection 3 1 4 Hall Effect Sensor Hallogic OHO90U Description This is a Hallogic hall effect sensor that detects the presence or absence of a magnetic field Application The best way to incorporate this into the design is to mount a magnet onto the brake lever arm while mounting a hall effect sensor board in a fixed location near the magnet When the magnet is within 3 8 inches to 5 8 inches depending on the intensity of the magnetic field the transistor switches and logic low is seen
144. a a touchscreen so the software had to be modified to allow this Before I could modify the software I had to learn how it worked The code had minimal documentation something on the order of one useful comment every few hundred lines and many of the variables did not have names reflecting their use I spent over a week reading through the program and marking the parts I could recognize before modifying any of the code 14 My first modifications were to the display The display is drawn by the class procs which calls methods defined in plots Plots contains methods for drawing the skymap the spectrum plots and the power vs time graph I started by temporarily removing all of the display elements This proved to be a time consuming task as numerous classes contribute to the display Tracking down all of the drawn elements and commenting them out of the code took many hours Starting with a clean slate I began work on the spectrum plots specifically working to make them movable and scalable to fit our needs This work began in October 2001 and continued into March 2002 The SRT program generates two spectrum plots as shown in Figure 3 3 1 The one on the right is redrawn with every sweep of the receiver across the observing band The one on the left is the average over time of the individual sweeps We chose to include the integrating plot in the final display for the reason that it is less susceptible to noise and provides a cleane
145. a path tests the unary operator tests benefit from a lack of interaction between inputs keeping the number of test cases small 23 The bitwise binary operators of the ALU are tested next Each 4 bit nibble is run against each 4 bit nibble at all 44 positions plus all the full word patterns of repeated nibbles are tested These test cases are run for each of 6 bitwise operators for a total of about 70000 test cases The arithmetic binary operators are the last part of the ALU to be tested Each of the two inputs are run through the ranges of inputs used in the unary negation test for each of three operators X Y X Y Y X producing roughly 1 2 million test cases Once the multiplier and ALU have been successfully tested the connection from the ALU to the multiplier can be tested The connection from the multiplier to the ALU gets sufficiently exercised generating test inputs to the ALU that it doesn t need independent testing The test cases passed from the ALU path to the multiplier go over the same range as the data path test cases producing roughly 200 test cases The final stage of testing is the Conditional MUX Its abilities to pass a constant value are tested in the data path section leaving just the conditional switching to be tested There are 8 conditions to test on each of two clippers for 16 conditions to test The inputs via Mux4 and Mux5 then via the crossbar are configured to use the two data constants DRO and DRI Thi
146. ab to ZPL so that it could be run on the cluster The completed conversion from Matlab to ZPL is in Appendix B This provided a real world test case to use in our performance benchmarks as well as providing a valuable example for the professors and graduate students at New Mexico Tech for converting existing Matlab simulation and data processing programs to ZPL My final contribution was the python and shell scripts I worked on to simplify the cluster configuration process for the end user scripts contained in Appendix C These scripts automate the generation of several system and parallel library configuration files and allow the user to easily switch ZPL between the available parallel libraries on the system The scripts are also part of the installation recovery procedure 3 2 Kevin Fisher At the onset of this project most of the machines to be used as cluster nodes had a limited amount of RAM others had absolutely no RAM Dr Scott Teare purchased many RAM DIMMS and I installed this memory in the cluster nodes Care was taken to 16 blow out any dust present in the memory sockets before installing the DIMMS To ensure that all of the new DIMMS functioned properly the program memtest86 was run on all machines for at least a twenty four hour period I assisted in this memory testing process Midway through the project we concluded that the upgrade to RedHat Linux 7 2 was required for further advancement in the development of t
147. al goals for the final project However these goals were modified as restraints involving time and money were encountered The modified goals are discussed further in Section 1 3 Project Description USS Forrestal Museum Inc http forrestal org fidfacts page13 htm 2001 1 3 Project Description This project was a step towards an intelligent firefighting system for ordnance carrying aircraft This design team was assigned the task of developing one part of this firefighting system The device being designed will be placed within the ordnance on the aircraft and will send a signal to a central firefighting receiving station when a fire is present It will also send the type of ordnance inventory information and the expected cook off temperature A heat detection unit was developed to identify temperatures that would be dangerous to shipboard ordnance This device was to be capable of transmitting an RF signal through the hostile conditions of a JP8 fuel fire A JP8 fuel fire produces temperatures as high as 1600 F in which the device must be able to survive for at least thirty minutes The transmission signal must contain the following information an ordnance authentication code the ordnance type the inventory data the current temperature and the heat flux These requirements are shown below in Table 1 2 below Ordnance authentication code Ordnance type Inventory data Current temperature Heat flux Table 1 2 Transmission req
148. al sensor functionality and overall project functionality 5 1 Project Highlights The wireless signaling system incorporates many features that make this product a success Robust keyed connectors are used in the design to insure that the user is able to easily hook up various sensors and does not have to worry about a difficult installation The system incorporates an encoding and decoding system that allows several of the same devices to operate in a close proximity without crosstalk Another key feature of the system is that the output logic is developed using a programmable logic device so additional features can be added in the future if necessary The sensors developed can be used on many different vehicle applications A final highlight is that the transmission system and sensors have worked well throughout the initial testing process 5 2 Budget Summary The Wireless Signaling System team was given 740 to build up the prototype system The team developed two wireless transmission systems with one being chosen to take into the full prototype stage Other items that the team developed include LED arrays and sensors The entire budget was spent with a good portion going to the two transmission systems Spare parts and shipping and handling fees also added to the final budget Appendix J lists how the project funds were dispersed The team was given constraints on how much the project should cost in the developmental stage and also
149. alyzing the data we had than collecting ideal data Before this element of the project was completely done away with I helped Seth prepare the instrument for refurbishing We met with Dr Rison and gave suggestions on how the new sensors ought to be mounted in the EFM Quite a bit of hunting and collecting had to be done to relate true north and elevation for each flight as they were mostly done in different cities I compiled a list of this information for use in determining the actual direction of the horizontal component of the B field I worked with Seth to develop a menuing system and helped Chris debug the final Matlab program Once we had an acceptable tool I did some data sampling for customer to locate relevant data ranges within each flight Communication was my core strength in the group I organized weekly meetings with our customer and advisor and kept everyone abreast of changes Most of the year s writing and presentation preparations were also my responsibility 22 Chapter 4 Final Product We began with several sets of flight data each of which contains digitized readings from a charge amplifier Hall effect sensor and a mercury switch To understand the three dimensional E field the direction of the EFM must be extracted The position of the conductive spheres must be known at any given moment so the Hg switch data has to be compared to the B field data See figure 10 Raw E data anna 9100 9200 9300 9400 R an
150. ams AutoCad C Microsoft products Altera e One Year work experience from summer internships College Course and Lab experience with digital analog electronics linear systems digital signal processing communication systems micro controllers PC interfacing electro magnetics 3 phase power systems and transmissions lines Available upon request 56 A 10 Daniel Rodriguez s Resume Campus Address Home Address P O Box 3154 1106 David Ct S W Socorro NM 87801 Albuquerque NM 87105 505 835 6302 505 877 251 dmr78 G hotmail com dmr78 Q hotmail com Objective To obtain a position for full time employment in the Electrical Engineering industry that will allow me to apply the educational training in a commercial environment Education Bachelor of Science in Electrical Engineering New Mexico Institute of Mining and Technology May 2002 2 95 GPA Major Experience Energetic Materials Research and Testing Center Socorro NM Office Assistant Fall and Spring Quarters 1997 to 2002 Assisted in data entry for the Purchasing Department Filed and maintained purchasing accounts The City of Albuquerque Family and Community Services Dept Albuquerque NM Special Events Assistant Summer Quarters 1999 to 2000 Developed and maintained badge system of community centers Supervised special events at community centers Bob s Burgers Albuquerque NM Cashier Cook Assistant Manager 1993 1997 Worked the cash register
151. aniel Rodriguez Connect and test TS with A D boardjall Integration of temperature sensors all System test all Final Documentation all 24 A 3 Trial Data Charts 5 7000 5 6000 5 5000 5 4000 5 3000 5 2000 Voltage 6 5 1000 5 0000 4 9000 4 8000 4 7000 4 6000 5 6000 5 5000 5 4000 5 3000 5 2000 Voltage 6 5 1000 5 0000 4 9000 4 8000 4 7000 11 oo oc r d r NO x 21 31 41 51 s ire 61 71 73 82 91 100 109 81 91 101 111 121 131 141 Trial 1 151 161 171 181 Time s Trial 2 191 201 211 221 231 241 251 261 271 281 291 301 311 262 271 280 289 298 307 316 Series Series2 Series3 Series4 Series Series2 Series3 Series4 25 Trial 3 5 7000 5 6000 5 5000 5 4000 5 3000 Series Series2 ET Series3 5 2000 44 Series4 5 1000 LH 5 0000 4 9000 4 8000 Time s Trial 4 5 8000 5 7000 5 6000 5 5000 5 4000 5 3000 5 2000 5 1000 5 0000 4 9000 4 8000 Time s 26 A 4 Trial 1 Data 14 27 40 5 0851 5 1611 5 0049 5 0830 14 27 41 5 0851 5 1611 5 1611 5 1611 14 27 42 5 0851 5 0830 5 0830 5 1611 14 27 43 5 1633 5 0830 5 0830 5 0830 14 27 44 5 0851 5 0830 5 0830 5 0830 14 27 45 5 1633 5 1611 5 0830 5 0830 14 27 46 5 0851 5 1611 5 1611 5 1611 14 27 4
152. ansform FFT command in Matlab I was able to see all of the major frequencies that were present in each of the signals As Seth mentioned above the Hg switch contained several frequencies the fundamental frequency and its harmonics The harmonics are what make the signal a square wave instead of a sinusoid so we decided to filter them out and isolate the fundamental frequency See Figure 4 I wrote a few lines of Matlab code that 13 would determine the peak value of the FFT and use that as the center for a FIR bandpass filter Frequency response of Hg switch in blue and Filter in Red T T o Magnitude 0 1000 2000 3000 4000 5000 6000 Frequency response of filtered Hg switch data T T E c Magnitude wm 0 1000 2000 3000 4000 5000 6000 Figure 4 FFT and filtering of Hg This effectively isolated the fundamental frequency Although the fundamental frequency was isolated the result was a sinusoid that was not very clean and had shifts in the amplitude where the original Hg switch data stuck Because these amplitude shifts can affect some of the data processing I wrote a for loop that re filtered the data and got rid of many of those shifts The loop took the sinusoid and any value that was greater than zero was turned into a 1 and any value less than zero became a 0 This made a much nicer square wave without the data anomalies This new square wave was filtered using the same technique as the or
153. apacity therefore erasing the memory The unused capacity of a cell cannot be utilized if the cell is not fully discharged Also at high temperatures the battery has possible performance problems Nickel Metal Hydride Ni MH It has many benefits It has low maintenance a long cycle life a high energy and power density In addition it is non toxic and it does not suffer from the memory effect However it does have its drawbacks It has a very high self discharge rate and it is very expensive In addition it requires specialized charging techniques and it is more sensitive to overcharging and complete discharging than Ni cads Lithium Ion Polymer etc It has many benefits It has a high energy density a long cycle life very low self discharge rate and it has a very good power to weight ratio In addition it is compact and it has more power than alkaline However it does have its drawbacks It is very 22 flammable when exposed to water very expensive Li ion is difficult to charge and Li polymer is cheaper and it is a higher energy density version of Li ion Reusable Alkaline It has many benefits It has a long storage life it is very powerful for a limited time it has high energy densities and it is available everywhere However it does have its drawbacks Itis very expensive over time it is not very dependable its power drops off with repeated charging it has low power densities and it is environmentally unfri
154. arges The vertical component was extracted by filtering out the high frequency component of the E field and multiplying it by the mercury switch data The horizontal component was derived by multiplying this same high frequency component of the E field but by the mercury switch data shifted by 90 degrees E field in blue hg in red 50 300 350 400 450 500 550 600 ime in seconds zi Vertical blue amp horizontal red component versus time T T ee ee g IN MAL Yu hen MA AA Pos phat BAA f A A 44M 1 f i i aha MEM NEM NENNEN MM ML LT ES NAMEN i f i fi i i 250 300 350 400 450 500 550 600 time in seconds Figure 11 Horizontal and vertical components of E field First plot of finished product 25 Figure 12 shows 2 subplots also the second is the exact same as in Figure 11 The first however is of the raw B and mercury switch data The relevance of the graph is to monitor the B field as the components of the E field change and also to show that the filtering is not affecting the vertical and horizontal components B field in blue Hg in red E T 1 doo 300 350 400 a 450 500 550 500 time in seconds Vertical blue amp horizontal red component versus time NANT pom DUE fu L QAM UL Ra i m ME RR 450 a 550 500 400 ne in seconds Figure 12 E field directional components relative to B field Second plot of finished product Figure 13 shows raw
155. as been routine and predictable This project still required the skills learned in those labs but the final answer was not known The successful completion of this project illustrates our ability to come together as a team and solve a problem that is anything but routine 21 References Contact Temperature Sensors Temperature Sensors 1996 2002 lt http www temperatures com csensors html gt 2 September 2001 LM34 Precision Fahrenheit Temperature Sensors Datasheet 30 November 2000 lt http www national com pf LM LM34 html gt 2 September 2001 Temperature Products By Handbook 5 January 2001 www omega com 2 September 2001 DI 194 Starter Kit Dataq Instruments 28 March 2002 lt http www dataq com products startkit dil94rs htm 29 April 2002 Huang Manna How to Use ActiveX in Excel Ulima Serial 17 May 2001 http WwWw geocities com ultimaserial exceltutor html 29 May 2002 Serway R A and Faughn J S College Physics 4 edition Harcourt Brace College Publishers 1995 22 A 1 Computer Code Dim cellindex As Integer Private Sub CommandButton1_Click UltimaSeriall CommPort 1 UltimaSeriall Device 194 UltimaSeriall SampleRate 100 UltimaSeriall EventLevel 0 UltimaSeriall ChannelCount 4 UltimaSeriall Key variable variable 35FFDOBD UltimaSeriall Start IeTimerl Interval 1000 IeTimerl Enabled 1 For cellindex 1 To 120 ActiveSheet Cells cellindex
156. ate performance increase of the NAS MG benchmark using a Class A data set This performance increase is similar to that observed with the mirror resurfacing simulation program 3000 4 p 2500 2000 seconds 1500 E 100Base T E 10Base T 1000 500 Processors 8 Figure 5 Comparison of the NAS MG benchmark using the Class A data set on 10Base T and 100Base T networks 2 4 Overview of the ZPL Parallel Programming Language ZPL is described as follows on the ZPL main page ZPL is an array programming language designed from first principles for fast execution on both sequential and parallel computers Because ZPL benefits from recent research in parallel compilation it provides a convenient high level programming medium for supercomputers with efficiency comparable to hand coded message passing Users with scientific computing experience can generally learn ZPL in a few hours Those who have used MATLAB or Fortran 90 may already be acquainted with array programming style ZPL main page 2002 10 This Beowulf cluster supports ZPL as one of the design requirements and is able to utilize ZPL with the user s choice of the PVM MPICH or sequential libraries 2 5 MPI vs PVM Parallel Libraries The MPI Message Passing Interface and PVM Parallel Virtual Machine libraries are the most common message passing parallel libraries in use today PVM is the older of the tw
157. ating simulation ZPL Matlab Display ZPL Computation Code program mirror config var setting all parameters output filename string data m dotiming boolean false Ri double 63 5 radius of inner track of filaments centimeters Ro double 127 0 radius of outer track of filaments centimeters M double 0 065 mass per clip grams rho double 2 7 density of the material grams per cubic centiemter H double 70 51675 height of filament plane above the mirror s center centimeters S integer 25 radius of the grid to cover the mirror integer which clips fire and which fail CiO01 boolean true Ci02 boolean true Ci03 boolean true Ci04 boolean true Ci05 boolean true Ci06 boolean true Ci07 boolean true Ci08 boolean true Ci09 boolean true Ci10 boolean true Cill boolean true Cil2 boolean true Co01 boolean true Co02 boolean true Co03 boolean true Co04 boolean true Co05 boolean true Co06 boolean true Co07 boolean true Co08 boolean true Co09 boolean true Co10 boolean true Coll boolean true Co12 boolean true Co13 boolean true Co14 boolean true Co15 boolean true Col16 boolean true Co17 boolean true Co18 boolean true Co19 boolean true Co20 boolean true Co21 boolean true Co22 boolean true
158. ay en T s Je Figure 3 4 1 Reassembling the SRT at the VLA Site 27 Once the installation was complete we connected the development computer and controller to verify that the telescope still worked We were able to control the SRT with no difficulties but since the kiosk wasn t ready we brought the computer back with us Appropriate VLA building and maintenance personnel were asked to move the cabinets and displays that might be in the way of the SRT s kiosk The Coke machine authorities would also have to be notified We have since arranged for a couple more needed parts like a multi strip outlet and grounding lugs When the kiosk arrives the space needed for it will be clear and we can schedule final installation and testing We have also made arrangements to pass on documentation of the project to Phil Dooley a VLA engineer 28 4 Final State of Project Figure 4 1 SRT at Its Final Location at the VLA 4 1 Final Testing Initial development and testing was an on going process With the touch screen we tested operation and functionality every time we made a change or added a new feature Proceeding in this manner we could ensure the system worked both in simulation runs and actual tracking of selected objects With the final version of the software we have tested the system and it has worked every time The system was tested on the site using the original software This test worked and verified that the wiring was re insta
159. be Vb Vsn Vbe RI A RbAZRb 10K as we want 10k since Rb 10k 2 56V 10 ASMA Vbe theoretically 7V but actually 64V for Figure 4 Mositure Sensor 35 E n PADO AHO PADI AML XTAL PAD SAN PADS SANG PADHANE PADS amp N5 PADS SANE PAD SANT P8 IxD ADDES DATAS PAl P IxD ADDELO DATALO PA2 P K ADDELS DATALS SPAT PST Cs aE PRO PWO PPl PWl PP2 PWI PPS PWS PR PPS Prt EXIAL ADDED DATAN SPE ADDES DATAS PES ADDES DATAt PEt ADDES DATAS PES ADDES DATAS SPBC ADDES DAYA7 PB7 ADDES DAIAS PAO ADDE DATAll SPAS ADDE DATAL2 P 4 ADDELS DATAL3 PAS ADDEL DATAL SPAS ED PH TD VPP SPL RSW PEL TTS EGO ECLER PE MODA IPIPED PES MDDE IPIPFI PE TEREPE Toto PIO IOCl Pfl 1002 PT2 1003 PI3 Tote PT OCs PIS Toe Spre PRAI IOC7 P17 MESAC AESC FONS FIGURE 6 39 Apr 2002 A 2002 E WINDOW SDa kh p Ha w F Hor 2 HC 2 Figure 6 Motorola 68HC12912 80 Microcontroller Dion 36 Figure 7 Self Watering Planter ry E a lt 3 G Z HeD Filter HeO storage Power Storage PLD Circuitry 37 Interrupt Signal Generated Light Sensor or RTC Controller Wakes Up Charge Battery 1 Measure Battery Voltage 2 Measure Solar Panel Voltage Is Solar Does Ba
160. be capable of retrieving and transmitting the data Design this device so it could transmit through the thermal insulation The transmission signal must travel a distance of 100 meters The transmission signal must be between 2 39 2 45 GHz Table 1 3 Design requirements This project was broken up into two parts an electrical section and a materials section Both parts were completed simultaneously with the final goal of integrating the parts together to build one working prototype Chapter 2 0 Technical Data 2 Power Supply Basic type K thermocouples wired together in series were originally considered as a possible power supply This power supply would obtain power from the fire and have an indefinite shelf life Research indicated that this would not be feasible as these and other types of thermocouples produce a very small voltage on the millivolt scale Omega Corp 2000 Another consideration for the power supply was to run a Peltier Junction or Thermoelectric Module TE in reverse A TE is a bimetallic or semiconductor module that performs as a heat pump when a voltage is applied across it When operated in reverse with a temperature a voltage is produced Most commercial TE modules are composed of a Bismuth Telluride semiconductor Bismuth Telluride operates from below 0 C to about 300 C www ferrotec usa com 2000 The second most common TE composition is of Lead Telluride which operate at 500 C A second option was to
161. board or build our own Each choice has advantages as well as disadvantages A third party board means we do not have to spend the time to design our own However designing our own board allows us to tailor it to suit our specific needs as well as the ability 11 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems to easily change such features should those needs change After much deliberation it was decided that we would design and build our own timer pulse accumulator board The details of this design will be discussed later The Breakout Board makes use of the PC 104 interface bus to communicate with the Prometheus This interface is electrically identical to the ISA interface found on the common PC The key differences between these two architectures are some additional ground pins in the PC 104 interface for the added ruggedness inherent to embedded systems design and a different form factor A brief overview of the PC 104 bus and our interface to that bus are outlined in Appendix A Image Processing Image Processing is the convergence of image data and digital signal processing The concept and procedures do not come first hand to everyone Research was conducted using the Internet books simple code examples and finally peer communication There were several resources for Image Processing encountered a book provided by the client was the primary resource The different types of methods used to manipulate images are summarized
162. bracket 1 2 in compression 7 12 in electrical grade PVC Data and Control buried 12 to 18in ae g To Buried PVC bep pulineebow 7 g Cables to Antenna see detail drawing 3 Ji teal PVC connector Visitor Center Building A 1 2 in 90 degree bends PVC Ground Level y e U 3 12 inrigid pipe anchored in ground see detail drawing 1 Figure 3 1 3 Details of Installation 10 Coordination efforts with the key personnel for the final SRT installation had a few issues that had to be resolved While the concept designs were generally accepted as well as the preliminary parts lists by the time February rolled around several e mails were received regarding some confusion in these installation plans for the SRT Personal contact with the lead people overseeing the different construction phases the SRT Project installation was needed to get everyone on the same page I had to address specification issues like location materials parts installation design grounding concerns lightning protection and a fence for protection Talking personally to the Grounds Projects Coordinator the Head Electrical Engineer and the Lead Electrician seemed to clear up the idea of the entire project This saved a lot of time and confusion in emails phone calls and inter department memos Next my conceptual design for the housing cabinet for touchscreen and monitor and computer and controller was submitted for approval and revision
163. cessor Verification Project Sponsor Gregory Donohoe UNM Authors Michael Carpenter Rocky Ginnani Randy Sena Scott Dearie Camden Mullen Summer Rhodes Stephen Stange Josef Hart Ernest Jim Daniel Rodriguez David Byrd Brandon Lattimore Ivan Olguin Gerald Bivens Sam Field Brian Rajala Heather Bitsoi Kyle Lamb Bill Willems Micheal Berg Kevin Fisher Anthony Montoya Jr G Martin C Reiten S Schuyler Steven Wasson Thomee Wright Design of Self Watering Planter with Emphasis on Power Scavenging A Thesis Presented to the Faculty of the Electrical Engineering Department of New Mexico Tech By Michael Carpenter Rocky Ginnani Randy Sena In partial fulfillment of the requirements for the course EE 482 Senior Design Project II May 2002 2002 Michael Carpenter Rocky Ginnani Randy Sena ABSTRACT A lot of people in this day and age go on sabbaticals or vacations and they have a tendency to neglect the watering of their household plants Without human assistance plants will not gain the necessary watering for survival Obviously plants cannot water themselves so we have developed a revolutionary planter that will automatically water any plant species based on its type water intake level utilizing a microcontroller moisture sensor pump battery charging through solar cells and other vital components This whole project utilizes the idea of power scavenging This is a basic system which in
164. ch was required to restart the system in the event of a power failure I had no idea how to find a watchdog and after extensive literature searches and interviewing the network system administrator I found out how we could restart our system without human interaction and without additional costs For the responsibility of education I started to educate myself in the basics of radio astronomy from websites and other reference material recommended by Robyn Harrison to develop the instructional displays Preliminary ideas like Physics and Radio Astronomy for the tutorials were initiated and in the second semester I started compiling literature search information I finally decided on three slide shows for tutorials and one slideshow for the project itself The largest slide show is The Science of Radio Astronomy which tells about some of the complex topics of radio astronomy and breaks them down to a basic high school grade level How the SRT Works is the next largest slide show which describes the major parts and pieces of this small radio telescope The Tracking slide show which runs automatically and gives time for the antenna to start tracking a source before the viewer gets bored tells of what the SRT is doing and what will be seen The final slide show The SRT Project was developed for presentations Educational Slideshows The Science of Radio Astronomy laid
165. chmark This particular benchmark solves Poisson s equation calculates how a charge distribution will attempt to reach equilibrium over a three dimensional array The NAS MG benchmark was packaged with ZPL and is what was used to locate some bottlenecks of the Beowulf Cluster 2 3 10Base T vs 100Base T Networks It was theorized that moving from a 10Base T network to a 100Base T network would bring great improvement in the Beowulf Cluster s performance While great improvement was observed in the NAS MG benchmark using Class S and Class W data sets the improvement observed was moderate for the larger more computation intensive tasks such as the NAS MG benchmark using Class A data set and the mirror resurfacing simulation program Readily observable from Figures 1 amp 2 below is the substantial reduction in the time required to complete the NAS MG benchmark using the Class S data set Figure 1 and using the Class W data set Figure 2 on a 100Base T network relative to the time required on a 10Base T network The Class S and Class W data sets are substantially smaller than the Class A data set or the set of data manipulated in the mirror resurfacing program Hence a greater percentage of the time required to complete the NAS MG benchmark program is devoted to communication between the cluster nodes This translates to an increase in network traffic with the addition of each new cluster node Moving from a 10Base T network to a 100Bas
166. chmitt trigger to implement this The sensor could now sense when an object was within range and then output 5 volts Refer to Appendix for a printed circuit board layout and schmitt trigger circuit details Final Status This is proving to be a reliable sensor when a good mounting place is found This sensor is depicted in figure 5 below It works in a good variety of vehicles but it will be up to the customer to mount the sensor 3 5 to 6 inches from the firewall and then adjust the switching level by adjusting the 10k multi turn pot until the schmitt trigger is giving a logic low We found that the R2 value on the schmitt trigger needed to be increased in order to work at distances above 6 inches The current generation of full size Ford pickups has the firewall set back further than the other tested vehicles Figure 5 Distance Sensor 3 1 2 Push Button Switch Description This is a small discrete general purpose push button switch This sensor is a button that outputs 5 volts when pushed Application The hardest thing with the sensor is how to mount it without it interfering with the driver We created housing for the push button out of thick plexi glass to protect it from damage from the foot This block also enabled us to put sandpaper on the face of the block so that the foot wouldn t slip off the block The plexi glass block is placed over the entire brake pedal with the button being in the center of the brake Final Status
167. cillator ROS 1121V Amplifier Mini Circuits 0 00 1 29 GALI 6F Transmitter Maxim 0 00 3 23 MAX2721 K type NMT Instrument 0 00 0 10 foot Thermocouple Room Fastblock 301 Kirkhill TA 84 00 90 Cerablanket Thermal Ceramics 0 00 0 32 Brainstem Acromane Inc 64 00 79 00 DC Relay Radio Shack 5 00 5 00 Thermal Switch Allelectronics 8 00 0 50 Lithium Battery Radio Shack 66 00 7 00 Total 227 00 203 50 Table 5 1 Total project budget Laser Beam Dump and Power Meter A Thesis Presented to the Faculty of the Electrical Engineering Department of New Mexico Tech By Josef Hart Ernest Jim Daniel Rodriguez In partial fulfillment of the requirements for the course EE 482 Senior Design Project II May 2002 2002 Josef Hart Ernest Jim Daniel Rodriguez Abstract High energy lasers are used for a variety of experimental purposes In a laboratory the laser beam produced must be stopped and absorbed by means of a laser beam dump Applying such a beam dump can also be useful in monitoring the energy that is left after them beam passes through its target The given problem is to design and build a beam dump that can withstand and measure the energy directly from a laser as well as a beam that has passed through a target This project was sponsored by Dr Daniel Eastman from Boeing SVS and the faculty advisor for the project was Dr Scott Teare from New Mexico Tech This thesis provides a description of the project from
168. cludes the main sub systems of recyclable watering solar rechargeable battery power and user selectable program The moisture sensor supplies the analog measurement of moisture within the soil sending the value to the microcontroller This will allow the watering system to supply water for the plant based on one of the chosen sub programs The chosen sub program depending on the plant s water intake level will decide the amount of water to release as it activates the pump for a predetermined length of time The water will be released directly into the soil since the tube is submerged in it and wrapped around the plant s roots For some plants too much water can be devastating therefore if there has been too much water induced into the plant the water will fall through the soil enough to be recycled back into the holding tank for future use Once the water drops below the predetermined level more water will be added The power will be delivered from a battery and recharged from solar cells that will be attached to the planter The choice of battery and solar panel was based on conceived power consumption as low as possible The user selectable program is sort of the brains of the project The HC12 microcontroller will have several sub programs that are based on the user s plant choice A switch will also determine the program used as well Until another plant selection has been made the microcontroller will continue executing the cho
169. consists of several chips such as transmitter encoder voltage regulator and several other components that were necessary to get these chips to work together Due to the simplicity of the board there was only the need for one input logic gate to transmit our signals We first decided to use an OR gate as a way to control when the encoder would transmit however after further evaluation we decided to make an OR gate out of diodes This not only cut down on cost but it also used less space on the PC board Refer to Appendix C for the entire transmitter schematic The main requirements for the transmitter board were that it had to be powered by a cigarette lighter We also added a hazard switch and override switches for both the turn signals and brake signal Another main task was to add a tri state dipswitch to our encoder which would allow us to exceed the requirement of at least 256 addressing The 256 addressing would also prevent any possibility of cross talk or interference from another signal The four channels on the encoder were used for the brake signal left and right turn signal and the hazard signal There were several stages that we took in building the transmitter board The first stage was to connect up the transmitter and encoder and verify that a signal could be transmitted properly The entire transmitter system is shown in Figure 2 Figure 2 Entire transmitter circuit 2 3 Receiver Circuit The receiver board was a lot more
170. conversion const of 2 75V out of 3 0V define V SOLAR 179 Solar panel volt conversion const of 3 5V out of 3 5V define SWITCH_POSITION_1 20 Switch select 2 0V water dry define SWITCH_POSITION_2 0 Switch select 0 0V water medium define SWITCH_POSITION_3 204 Switch select 4 0V water wet define V_MOISTURE 128 Voltage level of fully dry soil void charge_battery Charge battery function void water plant Water plant function void water routine Function that holds the amount of water to be pumped void delay unsigned int num Function that implements a delay for num Setup the appropriate variables that will be needed for the program volatile unsigned char s7 0 s6 0 s5 0 s4 0 s3 0 volatile signed int leave 0 main Enable bit 6 7 output PP7 moisture sensor MOSFET PP6 pump MOSFET DDRP DDRP 0xCO PORTP PORTP 0xCO Setup for PWM on Channel 0 PWCLK PWCLK amp OxF8 Choose 8 bit mode CO Select N 0 for Channel 0 38 PWCTL PWCTL amp 0x08 Choose left aligned PWPOL PWPOL Ox1f Choose positive polarity Select clock mode for Channel 0 PWSCALO 0x02 Select M 2 for Channel 0 PWPERO 255 Select period of 256 for Channel 0 PWEN PWEN 0x01 Enable PWM on Channel 0 PWDTYO 254 100 duty cycle on Channel 0 Setup for A D conversion on all channels ATDCTL2 ATDCTL210x80 Power up A D ATDCTLA 0X01
171. cook the sensors All the sensor data lines were fed to the appropriate pins on the 50 pin connector and in cases where a 5V supply was needed the power and ground connections were made to the sensor connectors This made for a clean and easy way in which to connect all of the sensors up and to interface the sensor data to the Prometheus When began decided to bring our power in through the PC 104 bus and use this to power the sensors looked at the specifications for the PC 104 bus and saw a place were it stated that the bus was capable of supplying up to 500mA decided to put a 250mA fuse on the board between the bus power and all of our connections as a safety precaution When had everything connected up discovered that weren t getting enough power to the sensors so decided that the PC 104 bus wasn t supplying enough current despite what the specifications said decided to bring in our own power supply to the Breakout Board also using a 250mA fuse and connect all of the sensors up to it while leaving the Altera power connections separate on the PC 104 bus This worked well and made sure that was not trying to draw too much current from the PC 104 bus After all connections were made and tested our customer asked us to draw up a schematic to show all the connections Our customer is planning to have a board made that will use the functionality which implemented and which can be changed around to meet the customer needs It wil
172. cott W Teare has given us both in and out of class Dr Teare has given us knowledge on how to succeed in the real world and how to develop our engineering skills Table of Contents Chapter 1 l1 Oyeryiew Of Projeti morira h 1 2 Introduction E EE Beg urea Snc M LA UP nuoc M S 2 1 5 Design Specifications oco nee E ea pe sa Ye rea lavas aea E Rea ENES 4 Chapter 2 Technical Work 2 DL Work OVervie WALedessdeseceta neck eee rte ue vu ERR E WU e Ege UU EE Chapter 3 Final Sensor Development 3 1 Brake Sensor System Development coire eei dero eR a ea SALTS Hamp GP DI 20 4 aue ci eyevks eo EE E ERE REEE Del 2 PushiButtoli SWIQh rinnas iir pN EE a bsp HEUS 3 1 3 Lever SWiteh s iscrisse irinari ee exper Teste utra ie eae 3 1 4 Hall Effect Sensor Hallogic OH090U esee S Current Sensorer ai iere i vto haeo RE vd E etae PE RETTEG v EM 3 T 6 Fouch Pad n coser cde e EE A SENI UII ST T 3 2 Turn Signal Sensor System Development ccceeeeeeeeeene eee eens 2 2 UO SSIES WAL Luo 120i Eee ipei Ires o ea DU de ERR ensue eS 3 2 2 Push Button Switeh eiie eret eerit Ret R eS Pee nba RU VERE TUE 3 2 3 PhotofransistUr iiie e Loo eroe D OV HELGA ERE IU HEU 32A bever mWIICH ersero naire o ude i E ee ES oe Desi vi rer ES EITEN 3 29 Sbhatp GP2DT20 ra sence tee A E A RM RU EE ET REACTS Y Chapter 4 Individual Contributions 4 1 Tasks Accomplished by Brandon Latti
173. ctive time cost to be amortized over many test cases While this would be a graceful and reasonably efficient solution the language we chose for the project C does not have the ability to generate functions on the fly in a clean manner so we can t use this solution A less graceful but feasible alternative to this method is to write the expected value 20 4 4 function when writing the test configuration word This adds to the software size but keeps the execution speed of test cases down This is the solution we used in our test software For the full source code to the test software refer to our project CD TEST CASE GENERATION Because it is not possible to test every possible combination of inputs to the Processing Element a subset of inputs which sufficiently test the device must be used This can be accomplished by testing the interconnections between components then testing each individual component However even using this approach most components cannot be exhaustively tested A subset of input cases for each component which should insure full functionality are used In order to fully test the Processing element several types of tests must be run The data path control mechanisms through the processor are tested first The selection inputs on muxes are easy to test exhaustively but only a subset of all possible data inputs are tested Because each output bit depends on only a single input bit 4 bit wide patterns are swept ac
174. ctrical requirements it was determined that the device should be pre mounted or non surface mounted and be able to execute programs on power up The requirements are summarized in Table 3 4 below Nominal current 100 mA Start up current 300 mA Voltage requirement 5V Ports 2A D 5 IO Memory 128KB of EEPROM Characteristics non surface mount or pre mounted Executes on power up Table 3 4 Minimum requirements for central processor 25 Three processors considered met these requirements the Brainstem form Acroname Inc the OOPIC from Savage Innovations and the 1800 series PIC from Microchip Samples of the Brainstem and the OOPIC were received and tested The 1800 series PIC was not tested due to the fact it is not pre mounted and would require extensive modification to become a viable solution It was determined that if the OOPIC or the Brainstem worked more time could be devoted to learning how to operate the PIC rather than trying to modify it Tests on the Brainstem and the OOPIC revealed that the two processors were substantially different even though they both met the design requirements The Brainstem is able to run on a variety of platforms that include Windows Linux Unix Windows CE and Palm operating systems The OOPIC must be run on Windows 98 The OOPIC is programmed in Object Oriented ANSI C whereas the Brainstem is programmed using TEA packets in a C style language The Brainstem runs at 75mA compared to 30
175. d increasing the swap space to 512MB or higher would also raise this limit although swap space is slower than RAM Network improvements As was seen in Figures 1 and 2 the switch from a 10Base T hub to a 100Base T hub 21 significantly reduced the performance penalty of the network overhead However even with a 100Base T hub when results were being gathered to the server node or during periods of high inter node communication when processing the collision light on the hub would indicate a frequent network packet collisions Packet collisions require the packets to be resent wasting bandwidth and decreasing network performance Moving to a 100Base T switched network would eliminate the network inefficiencies due to packet collisions and allow simultaneous communication between different sets of nodes in the cluster increasing total network throughput 22 References The Beowulf Project http www beowulf org Parallel Processing HOWTO http www tldp org HOWTO Parallel Processing HOWTO html Beowulf HOWTO http www tldp org HOWTO Beowulf HOWTO html ZPL main page 2002 http www cs washington edu research zpl NASA Advanced Supercomputing Software Descriptions NAS Parallel Benchmarks http www nas nasa gov Research Software swdescription html NPB PVM resource list http www csm ornl gov pvm pvm home html PVM FAQ http www netlib org pvm3 faq html nodel html 23 Appendices Appendix A 100 inch mirror coating s
176. d needs to do is go to sleep until woken up by the user space control algorithm When it does wake up it goes into an infinite while loop where 34 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems the first thing it does is check the Command FIFO for start stop instructions The start instruction sets up how the thread will function That is is this thread periodic or non periodic interrupt driven threads are a special case of non periodic threads The stop command simply puts the thread to sleep in the case of a PWM thread we send a PWM command of 0 to stop the motors before it goes to sleep Next the thread must gather the data To read the Position Accumulators we simply read a word two bytes of data from the appropriate address using the Linux function inw short address To write to the PWM command registers we use the RTLinux function rtl_outb char data short address The variable data is a signed char data type This is an 8 bit value where the most significant bit the sign bit determines the direction relative to the robot forward or reverse we wish the motors to spin The remaining seven bits represent the speed that the motors should spin A value of 0 represents a speed of 0 stopped and a value of 127 represents a speed of 100 full speed Any Analog I O data we need to gather IR Distance Sensors Ultrasonic Distance Sensor amp IR Flame Sensors makes use of the Prometheus A D module which was set
177. d not encounter any future interference Testing was done to determine which encoder decoder pair would work best with our transmitter receiver from Glolab We decided to use an 18 pin encoder decoder pair from Holtek HT694 This encoder decoder pair turned out to work best for our application The Holtek was not only simple to interface but it also gave us an acceptable amount of encoding possibilities The range of our transmitter receiver pair is about 250 feet My second assignment was to design the taillight LED arrays as depicted in figure 9 that would be mounted to the trailer have done research on both the pre manufactured LED arrays and also on a design of my own Due to the high expense of the pre manufactured LED arrays decided to design my own have done some simple calculations that will give us maximum light intensity with minimal power consumption At the beginning of the spring semester we were asked to add on the possibility for the use of regular incandescent light bulbs This had no major effect on the LED array output Instead we used a jumper to access either the LED s or the light bulbs Figure 9 LED taillight arrays The third part was assigned to design was a PC board that would consist of all the necessary hardware to have a complete prototype by the end of the spring semester Brandon and were in charge of different parts the PC board design We decided to use Protel to layout our design since both Brandon
178. d with 5 working Cat5 cables 6 additional cables were created to complete the network portion of the hardware evaluation Hardware Installation I handled the purchase and assisted in the installation of a CD RW to assist in completion of the Emergency System Recovery Plan and add to the system s backup and archiving capability Research Other Beowulf Clusters Current designs of Beowulf clusters were researched with the goal of finding the best techniques to apply to our Beowulf cluster Also in this capacity historical information and general background knowledge was sought for presentation and question aspects 19 Person of Contact I had the responsibility of communicating messages or changes to the project between the team customer advisor and instructor for the duration of the project The lines of communication between our group and the customer were the most important to keep strong and clear in order to ensure that we were progressing through the project in the right direction I also was responsible for ensuring communications to our advisor Dr Wedeward We were required to report to our advisor on a weekly basis to report progress and seek advice if needed for the project I also received and shared information that was passed to me from the instructor pertaining to changes in assignments or due dates as well as any other modification in the class Documentation I assisted in the documentation pertaining to the Beowulf
179. data coming from the receiver into a display e g spectrum that demonstrates VLA operation Time and Location Update Implement software or hardware to update computer clock for accurate tracking Survey site to find exact dish location Implement Tracking Software Look into using LabVIEW TheSky STS control system or existing SRT software to track astronomical objects Calibration of SRT Calibrate receiver to obtain accurate data Electronic Noise Method of calibration more accurate than motorized vane Motorized Vane Alternative method of calibrating receiver Watchdog Reset Set up computer so that it and the display program restart without human intervention in the event of power failure SECOND SEMESTER Tasks Control Software Implementation further development and use of existing SRT software for tracking astronomical objects and deadlines to reject other control software Research order parts Deadline for other controls software such as LabVIEW TheSky or STS control system Develop Public Use System further development of touchscreen interface that the public can use to control SRT and learn about radio astronomy As part of implementing the touchscreen for public use with intuitive commands and that has an attractive appearance Research order parts this might be to get together with someone to design a counter to house the computer components that denies public access to the main computer except for the monitor and touch
180. ded temperature range operation 40 to 85 C Data Acquisition Subsystem Analog Input 16 single ended 8 differential inputs 16 bit resolution 100KHz maximum aggregate A D sampling rate Programmable input ranges gains with maximum range of 10V 0 10V Both bipolar and unipolar input ranges 5 ppm C drift accuracy Internal and external A D triggering 48 sample FIFO for reliable high speed sampling and scan operation Analog Output 4 analog outputs 12 bit resolution 10V and 0 10V output ranges Simultaneous update Adjustable output range optional Digital I O 24 programmable digital I O 3 3V and 5V logic compatible Enhanced output current capability 8 12mA max Counter Timers 1 24 bit counter timer for A D sampling rate control 1 16 bit counter timer for user counting and timing functions Programmable gate and count enable Internal and external clocking capability 55 Beowulf Cluster Research and Implementation A Thesis Presented to the Faculty of the Electrical Engineering Department of New Mexico Tech By Michael Berg Kevin Fisher Anthony Montoya Jr In partial fulfillment of the requirements for the course EE 482 Senior Design Project II May 2002 2002 Michael Berg Kevin Fisher Anthony Montoya Jr Beowulf Cluster Research and Implementation Michael Berg Kevin Fisher Anthony Montoya Jr El
181. device is valid It is in the computation of the expected result that timing is critical There were several options for computing the expected result Our customer has a C object which simulates the Processing Element of the RDPP It takes a configuration string and a set of inputs and will generate the expected output The biggest advantage to this option is that it is a complete ready to go solution Unfortunately though its versatility and flexibility which are its strongest points on its own cause it to require significantly more computational overhead This made it a poor solution for our project The next option was to develop our own Processing Element simulator focussed on efficiency rather than flexibility This option would require significant effort to develop and probably would have a high computational overhead as well Because the test cases are broken down into groups of many tests with the same configuration other methods of computing the expected results are possible A function which computes the expected value for a test configuration given the inputs could be extremely efficient since it only has to perform the necessary calculations and doesn t have to spend time checking the configuration for what to do One solution would be to generate these expected value functions at run time based on a configuration word This process would not have to be as efficient since it would only be done once per set of tests allowing its effe
182. directly by Excel and we were able to perform the necessary calculations in another cell Excel proved to be a straight forward easy to use method of collecting and calculating the data 4 5 Budget The following was our expenditure for this project Additional items that were donated are marked with zero in the price column Total 1 Resistors 1K ohm 0 10 0 40 CawcioriuF 4 3900 0 00 5 00 SmalCometos s 5040 5450 5132 Large Ring Comestos 2 3008 80 16 Tp 4 Sus S94 S164 240 30 16 Table 3 Table of total expenditures in the building of the Laser Beam Dump 3040 0 00 NN 1 Resistors ISK ohm 4 3010 80 40 Lec cte ESO ty 4 4 4 4 4 1 2 4 20 5 Conclusions The goal of this project was design and develop a high power laser beam dump and power meter This was not the type of project we had expected to see due to the interdisciplinary aspects Having a strong foundation in electrical components and data collection we were left to learn material properties as well After learning about and choosing a material that could be set in front of a laser the rest of the project was a matter of applying our electrical engineering skills with our newly acquired material knowledge The result was a successful project that returned values very close values shown by a commercial laser power meter This project was a very valuable experience Up until now our practical lab work h
183. djustable level threshold circuit or an A D analog to digital converter which is conveniently in the HC12 microcontroller OV for resistance below certain level 5 for above The whole setup requires an OP AMP a resistor a potentiometer seems like a Schmitt trigger displaying hysteresis characteristics We were also thinking about creating a circuit with LP low pass frequency characteristics to eliminate potential noise I found a JFET that can do this however this plan also fluttered for one simple reason it draws way too much power in the W range From this point on I tested several circuit designs One was somewhat of a discrete circuit It is a comparator or once again a single supply OP AMP LM319 LF411 is considered a dual supply Vref and Vin input voltage displays hysteresis Once the probe and some resistors are connected to the positive input and Vref is connected to the negative input the resistance connecting Vout displays the hysteresis width I plan on hooking the output to the input of the HC12 To realize the output current one must subtract the base voltage by 7 Then divide that by the R connected to GND ground perhaps 100k Another circuit was a generic transistor The base would be connected to the HC12 5V on OV off Vcc would be sending 5V through to the emitter It would go to the A D converter where it s 0 ohms Basically the comparator will switch the sensor on and off with a stable curren
184. do ones tek ees vev pati ORE TER FERA NAA Exe QE 22 Figure 3 3 3 Slide Show Presentation Layout eese 23 Figure 33 saand D cocco ve AO wav R NEED A ERER ORE EN 23 Figure 3 3 5 The VLA Weather Station Layout ccc ccc eeceeeee eee eees 24 Figure 3 3 6 The VLA Web Links Layout sesseseII 24 Figure 3 4 1 Reassembling the SRT at the VLA Site ssusuueususs 27 Figure 4 1 SRT at Its Final Location at the VLA eeseeeeeeeeeeee 29 iv 1 Introduction This paper is in partial fulfillment of course requirements for the New Mexico Institute of Mining and Technology Department of Electrical Engineering s Senior Design course This is a detailed description of the work accomplished by Gerald Bivens Sam Field and Brian Rajala on the National Radio Astronomy Observatory s Small Radio Telescope project between August 2001 and May 2002 1 1 NMT Senior Design Course As part of its requirements for graduation the Electrical Engineering Department at New Mexico Tech offers a two semester design course At the beginning of the course seniors in the department are offered a number of projects by outside entities representing a wide variety of public and private enterprises Students are placed in groups of two to four members based on their project selections Each team is responsible for meeting with their customer to draft a statement of work outlining the
185. ds for use with the PVM library the MPICH library the LAM MPI library and for single computer sequential use The documentation that comes with ZPL only describes how to install one of these variants and we were interested in at least installing all of the PVM and MPI variants So I compared all of the files in each download variant and found that there is a common ZPL core set of binaries and libraries that then compile ZPL code against separate directories for the parallel or sequential library being used We were then able to safely install all of the ZPL variants that we were interested in in order to facilitate easily switching between parallel libraries when compiling for testing purposes Once this was completed I had to configure the PVM and MPICH parallel libraries so that ZPL could make use of them This involved editing several configuration files for the libraries and setting shell environment variables in the user s login configuration files 15 This step went smoothly except for some problems with the RSH configuration in the cluster that prevented PVM and MPI from successfully starting programs on the other nodes in the cluster Once the team had solved the RSH problem the parallel libraries were functioning properly and we could use ZPL in conjunction with all of the parallel libraries that we had installed My next contribution was porting Denis Oesch s 100 inch mirror coating simulation see Appendix A from Matl
186. duced by the hazardous environment of a jet fuel JP8 fire to an intelligent fire fighting receiver station The final prototype was created using the following components The insulation used a ceramic resin material Fastblock 301 Kirkhill TA Co which had a low thermal conductivity of 0 07btu ft hr F at 400 F The inner layer of insulation was composed of Cerablanket Thermal Ceramics Co which had a thermal conductivity of 2 05btu ft hr F at 1800 F This insulating casing was capable of maintaining an internal temperature near 150 F for nearly thirty minutes when exposed to a continuous flame Electronic components where found that are rated to operate at temperatures above the 150 F originally required The electronics employed a K type thermocouple to read the external temperature a Brainstem processor Acroname Co an op amp circuit to amplify the signal from the thermocouple and a quadrature transmitter Maxim Dallas A lithium battery in conjunction with a thermal switch served as the power supply Table of Contents List of Figures and Tables eet eee seien io ern e o eete eee pr enero een eev ers seen sepe eas eceueroes L 1 0 IntrodUuctHtOn iiis visites essa ae esa cu toan Ps kv ou cha k3d ev iege dota Centuria asiro LET HISUOEY S oed eco er Conn NER onte pite do rec ieri ec een eek enin edes sa era ies dg 1 2 Customer Requirements e ecce ee eee ee ee eere eese ense eoo sees cose eo ssee 1 3 P
187. e We chose to use Excel because we were already familiar with the software package and this allowed us to stay on schedule 13 3 Individual Contributions 3 1 Josef Hart Josef was the group s point of contact group leader and responsible for interfacing the laser beam dump hardware with a data collection system As the point of contact Josef kept in contact with both Dr Scott Teare class instructor and the group s faculty advisor and Dr Daniel Eastman the customer from Boeing SVS Josef then passed on the information to the rest of his group so that they were aware of all communications and project developments The responsibilities of the group leader coincided with the duties of point of contact In addition to keeping everyone s work on track Josef also arranged meetings and scheduled tests at the customer s facilities The task of interfacing the laser beam dump hardware with a data collection system consisted of two main areas The first was to connect the sensors to a computer through an analog to digital A D board and the second was to process the data with a software package that would calculate and return values that were useful in our experiment The A D board was a four channel board and connected to a computer through the serial port It took its power from the computer and therefore required no external power source The board also did not require any programming which made set up simple The vendor that we bought the
188. e Altera Excalibur board shown in Figure 4 on the next page is designed to implement embedded systems such as NIOS processor design provided with the board The APEX20K200EFC chip is a high speed SRAM programmable chip The board allows you to program this chip through a JTAG interface or by installing your configuration in on board flash memory The Altera Quartus II software like MAX plus allows the user to design digital systems textually with a Hardware Description Language or graphically It also serves as the means to program the PLD Mi mieten Windil Figure 4 The Altera amp Excalibur Board 2 4 FUNCTIONAL TESTING A functional test is one of the many tests executed on a device as it is being developed It answers the question Is this device actually doing what we are designing it to do To answer this question test inputs or vectors are executed on the device and the output produced is compared with what the output should be according to the device specifications If the actual and expected outputs do not equal each other then there is an error in the design Ideally one would like to test the device with every possible input combination In reality however the device has so many input combinations that it is 2 5 impossible to test for every case Consider the Processing Element of the RDPP To test it would require the execution of 2 test cases Since it is impossible to test every case the tester must derive
189. e Robot Bitsoi Lamb Willems Image Processing Heather Bitsoi For the continuation of the image processing here are a few details of where the project ends and how to continue To compile To run the program Outputs gcc o root lt filename gt The altered lt filename gt lt image filename w o images from lt filename gt c extension gt program The commands above will compile the program The current program finds vertical and horizontal lines There is additional code provided which was used for testing purposes In the testing algorithm the test was conducted to determine the convolution matrix that was most efficient and effective for finding edges There are many more applications that could be added to the image processing part of the project Applications like determining how far the line is relative to the robot the magnitude and orientation of the line determining whether a line has moved or not using the subtraction of images This method would be an approach for multi frame line tracking Other applications like object recognition would be possible For instance the furniture in the maze for the firefighting robot is yellow The color value of yellow can be compared to the pixels in the image to determine whether a furniture object is in the image or not The image can be re configured to a 2 bit 4 bit or 8 bit image Making it possible to determine the difference in colors easier being that the
190. e T network drastically increases the speed of such network traffic thus substantially decreasing the time required to complete the program E 100Base T B 10Base T Processors 7 Figure 1 Comparison of the NAS MG benchmark using the Class S data set on 10Base T and 100Base T networks 160 140 120 100 seconds 80 60 E 100Base T E 10Base T 40 20 3 4 Processors 5 Figure 2 Comparison of the NAS MG benchmark using the Class W data set on 10Base T and 100Base T networks A comparison of Figures 3 amp 4 the performance of the mirror resurfacing simulation program on a 10Base T network and a 100Base T network respectively will show a moderate performance increase with the 100Base T network The mirror resurfacing simulation program unlike the NAS MG benchmark program outputs the time required for I O and the actual computation time Also observable in Figures 3 amp 4 is the increase in time spent on I O activities with the addition of each new cluster node 140 120 100 80 E O Time E CPU Time seconds 60 40 20 2 3 4 5 6 7 8 Processors Figure 3 Time required to complete the mirror resurfacing simulation program on a 10Base T network 140 120 100 80 E O Time E CPU Time seconds 60 40 20 0 2 3 4 5 6 7 8 Processors Figure 4 Time required to complete the mirror resurfacing simulation program on a 10Base T network Figure 5 shows a moder
191. e beam dump worked like we thought it would We did not have the means or time to test it individually but when the sensors were attached and laser energy absorbed the beam dump performed according to calculations Given more time for design and with the laser we would have worked harder to ensure that all light and energy was absorbed Due to our results we can assume that all of the energy and power were absorbed but we could not see if any light was reflecting out of the beam dump 4 3 Heat Sensing Hardware Summary The use of the temperature sensor package helped tie in the actual beam dump with the data collection The simple reliable package made it possible to measure the heat exchange in the aluminum This measurement was then feed into the A to D board for data collection and manipulation Without the sensor package one of the main goals of this project could not have been possible The actual LM34 temperature sensor in the sensor package worked as specified in the factory specifications for these sensors 19 4 4 Data Collection System Summary The data collection system worked exactly like we thought it would The A D board was very easy to connect and interface with a computer Interfacing the A D board with other software packages was also simple since such resources were provided by the manufacturer Huang Starting and stopping the program was a simple push button that was inserted into an Excel spreadsheet The data was collected
192. e casing Sauereisen was used to seal the hole that the thermocouple was inserted into Two thermocouples were attached one to measure the internal temperature and one the external temperature The internal temperature reading was fed to a computer where the data was plotted using Omega temperature acquisition software The plots of the external and internal temperature versus time and of internal temperature versus time are found in Figures 3 3 and 3 4 respectively Temperature readings were taken every twenty seconds as seen in Figure 3 3 As can be expected from a fire the external temperature had a large variability however 18 it remained close to 800 C or 1472 F The internal temperature gradually increased The internal temperature can be seen in Figure 3 4 There was an increase in temperature as expected during the test The temperature at thirty minutes was 85 C or 184 6 F close to the specified temperature of 150 F It was determined that this prototype was feasible and it could be modified to fully meet the specifications 1000 800 700 600 500 400 4 Temperature C 300 4 Rom do Qo o o Qm oc KR io 5 t Sf wo r o c c oO Q O O e 53 t F K bo 9 ory Time t 20 seconds 5 9t 13 17 1 25 t J 29 33 Figure 3 3 External temperature is plotted above and internal temperature below 140 129 4 100 2 5 80
193. e consists of consecutive bytes representing the pixels in the scan line in left to right order All BMP files contain RGB data In our case we have e 24 bit 16777216 colors mixes 256 tints of Red with 256 tints of Green and Blue The set of the first 55 positions in the image is the image header First we extract the image width followed by the height the depth the compression and finally the compressed size After reading all this information into a buffer the buffer is a character array that is allocated to store the bitmap information and retrieve it for later use The next section of the bitmap is the image data The image data is put into another buffer This information would be altered and 16 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems manipulated The original data is untouched and remains the same for comparison Now that the information in the BMP is stored the next procedure is to use the convolution process to highlight the horizontal and vertical lines in the image The convolution process constitutes of two processes one is defining the horizontal lines and the other is defining the vertical lines Each process uses a different mask the mask is a 3x3 matrix with the values that make the lines more prominent The mask for horizontal lines is called the row mask and for vertical lines it is called the column mask The masks used for the convolution process are depicted in Figures 2 and 3 These matrices a
194. e it down in to workable pieces and added comment lines so that my other group members could use the tools I was uncovering The most useful tool I got from this diary was spike filtering Though I had to do each point by hand this made filtering go so much smoother and gave a foundation for the C code that Seth wrote to automate flaw filtering There is a Matlab function called DISP that does partial derivatives between data points meaning x2 x1 x3 x2 etc I thought that perhaps I could add or multiply these functions to take each of these differences and somehow divide that value in half and place it between the existing points to create a smoother curve The answer turned out to be no but I spent quite a lot of time experimenting with this hypothesis I also tried to fit a sine wave to our B field data Our customer suggested the method of least squares I searched the web and looked in the library and could only find 20 equations derived for linear fit and polynomial fit for least squares no sine fit equations I finally found the IEEE standard IEEE STD 1057 for sine fitting which involves matrix multiplication This works wonderfully for small data sets but our set contains 49000 points Even a small section of our data would result in the multiplication of a 100x100 matrix This would be cumbersome to compute and would possibly crash Matlab Finally I derived the equations based on the polynomial fit equations that I found I
195. e lower byte of Port C to output To finish the setup of the Prometheus I O module we enable the interrupt the I O module uses IRQ 5 and assign an ISR intr_handler Finally to complete the initialization of the module we assign a device file handler my hanaler for our Control FIFO This code is demonstrated below in Figure 17 int init module void pthread attr t attr struct sched param sched param char temp make sure the RT FIFOs we want aren t already in use rtf destroy CONTROL FIFO rtf destroy R POS FIFO create our FIFOs rtf create CONTROL FIFO CMD FIFO BUFF SIZE tf create R POS FIFO FIFO BUFF SIZE H create threads thread attr init amp attr ched param sched priority POS PRIORITY thread attr setschedparam amp attr amp sched param thread create amp tasks R POS FIFO amp attr thread code void R POS FIFO thread attr init amp attr ched param sched priority POS PRIORITY thread attr setschedparam amp attr amp sched param thread create amp tasks L POS FIFO amp attr thread code void L POS FIFO O O0 0 O O 0 0n o setup Prometheus I O Reset FIFO depth to 0 clear all the int flags CLR INTS disable Analog Input Interrupts temp rtl inb INT DMA COUNT CTL temp amp ENABLE AIO INTS 32 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems rtl outb temp INT DMA COUNT CTL Se
196. e pedal is being pushed but without any interference to the normal driving habits and safety of driver in any way The sensor would need to be kept away from the driver s foot and not hinder the foot reaching the brake pedal in any way 3 1 1 Sharp GP2D120 Description This is a general purpose IR distance measuring sensor The output voltage on the sensor increases as an object gets closer Application The first application involved placing the sensor directly above the brake pedal so that it would see the driver s foot on the brake pedal This wasn t very feasible for a couple of reasons It needed to be placed within 3 inches of the path of the foot for an accurate reading which meant it could easily be kicked off Also since it was looking directly for the foot it might see the foot just resting by the pedal or under it and assume that braking is intended A more feasible way of using this sensor was to set it up to look for the movement of the brake pedal shaft We had to add some extra circuitry to make this sensor work for this application The output voltage of the sensor ranges from about 1 3 volts to 1 5 volts when an object is within 3 5 to 6 inches This sensor needed external circuitry since the encoder needs 3 7 volts to switch levels To fix this we added an adjustable schmitt trigger to take inputs of 1 3 volts to 1 5 volts and output 5 volts With a few calculations we were able to figure the exact resistor values for the s
197. e temperature rise of the plates and the total incident power The total incident power is computed as a function of this temperature rise 1 2 Definitions HEL High Energy Laser Specific Heat The quantity of heat required to raise the temperature of a given mass of a substance Serway 323 342 Protel 99 A Windows based printed circuit board PCB design tool 1 3 Literature Review To begin this project we started by searching the internet and library for sources that concern our project Since our project was based on a device that already existed we tried to find previously written documentation about it The problem was that there was none and even our customer had trouble finding anything regarding our topic While we found many sources for commercially available laser beam dumps there was little about beam dumps that were integrated with power meters The available literature was useful for the design of the beam dump It gave us an idea of what materials to use and the geometry of portion that the laser beam strikes but the power meter design was left to us to develop In addition to design ideas other literature searches helped us find what hardware to use for power measurement These sites were primarily manufacturer s websites that gave us ideas of how to take commercially available parts and implement them to solve our given problem Our reference section outlines the sources we reviewed while solving our design problem 1
198. e the SRT for safety considerations and to protect it from abuse This information 26 was passed on to appropriate personnel Then tentative final delivery and installation dates were scheduled After development of the antenna control programs and educational tutorials the team got together to deliver the kit to the VLA site in April Together the team got the antenna and mounts taken apart the cables pulled out and everything loaded into the NRAO s panel truck We had a backup plan to store everything in the warehouse in case the installation went awry At the site we reassembled the kit and mounted it on the provided post in the enclosure built by VLA site personnel Figure 3 4 1 Contact was made with the electricians to borrow a wire pulling fish ladder and some other tools We managed to pull the signal wire into the installed conduit between the enclosure and the Visitor Center building and assemble the antenna without any problems When we were trying to pull in the control cable it got stuck because of too many 90 degree bends that were put into the conduit run The VLA electricians were called to expedite cutting the conduit in the middle and to install an extra pulling elbow The control cable was finally pulled through to the junction box at the base of the dish The cables were pulled through the outer wall of the Visitor Center but the junction box there was not completed In addition there was a Coke vending machine in the w
199. e the benchmark process to a small degree These scripts would run through each benchmark program incrementing the number of processors utilized in the benchmark programs and writing the output data from each program to separate and organized files These scripts allowed for the running of many benchmarks in sequence without the need of a human operator to monitor the process and input commands thus allowing the running of benchmarks overnight The scripts also greatly reduced the probability of user error in setting up the benchmark parameters on the command line Near the end of the project the Beowulf Cluster was moved from a 10Base T network to a 100Base T network for performance comparison purposes I was in charge of this move replacing the Beowulf Cluster s 20 port 10Base T hub with one 8 port 100Base T hub 18 3 3 Anthony Montoya Jr Hardware Evaluation The initial phase of the project consisted of an evaluation of the hardware we received from our customer Dr Teare I aided in the testing of the individual machines and the cluster s internal network This included adding memory to the machines and testing the existing and upgraded memory in all of the machines We also verified the working condition of the hub and its ports as well as the existing and newly manufactured Cat5 cables Hardware Construction I made and tested several Cat5 Cables that were used in the networking of the cluster Because we were only provide
200. e transmission setup must be encoded to safeguard against cross talk The worst case transmission distances is 100ft and the entire system should be designed with industry standard parts The receiver side of the setup will be battery powered and be designed to consume low power The system will run on a user supplied power source of between 10 14 volts There is a decoder on the receiver side to ensure that the received signals are coming from the correct tow vehicle The system will include light driver circuitry that sets the LED arrays to run at a low level for continuous level lighting along with a full intensity level when the brakes are applied Logic circuitry is needed to control the flashing braking of LED arrays The brake input state must override the hazard state The light emitting diode arrays must consist of red high intensity LED s These LED s must meet or exceed the brightness of current production trailers The bulbs must have a long life and consume low power The prototype board built will be able to give the customer a general idea of how bright the LED arrays are along with a sample of how the braking and turning functions are output from the rear signaling box Chapter 2 Technical Work This Chapter describes in detail all the work that was needed to put together a successful end product It is broken up into sections detailing the transmitter and receiver subsystems along with Protel layout and design work inv
201. e wrong results xy rsqrd R rsqrd layers 5 0 4 0 M 4 0 M PI rho setting up the thickness on the mirror from this arrangement 1 2 841 1 2 S 1 1 1 1 Coatingtemp lt lt R rsqgrd layers setting a limiting value to compensate for the square grid beyond the round mirror s edge Flood Coatingtemp S 1 2 S 1 1 1 into Edge R mirror Edge gt gt S 1 2 S 1 1 1 1 1 Coatingtemp This makes surface a circle filled with ones and zeros outside the circle R mirror surface Indexl S 1 2 Index2 S 1 2 gt S 2 R mirror temp5 Edge 1 0 surface R mirror Coating temp5 Coatingtemp surface Flood G S 1 2 S 1 1 3 into temp6 R mirror temp6 gt gt 1 2 S 1 1 2 S 1 1 3 G Flood G S 1 2 S 1 1 3 into temp7 R mirror temp7 gt gt S 1 2 S 1 1 3 G R mirror Mirror Coatingtemp temp6 surface temp7 1 0 surface All the computational work is now done get the total runtime Do it here since the file I O below shouldn t be in the computational runtime benchmarks 36 timervalue CheckTimer if dotiming then write Computation Time in seconds t 12 4f n timervalue end Write out all the data needed in Matlab for post analysys and graphing the results ResetTimer reset the timer for I O timing
202. ectrical Engineering Department New Mexico Tech ABSTRACT A Beowulf cluster is a method of building a super computer by connecting numerous machines together and getting them to work in parallel on a single task The design and construction of a Beowulf cluster is still a highly technical activity requiring a proficiency in operating systems networking and system design The primary goal of this project is to provide the customer with a functional and easy to use Beowulf cluster constructed from surplus computers A secondary goal of this project it to allow existing parallel computing techniques to be more easily applied to future Beowulf cluster work at New Mexico Tech and elsewhere This project promotes the use of surplus computers readily available to university environments in Beowulf clusters and will simplifies the process of constructing a Beowulf cluster This makes Beowulf clusters more accessable to researchers in other disciplines such as physics and biology ii Acknowledgments First we would like to thank Dr Scott W Teare for sponsoring this project Dr Kevin Wedeward was our faculty advisor and we would like to thank him for advising our team throughout the year Dr David Westpfahl Jr was a great help by providing the surplus computers used in the construction of this Beowulf cluster Denis Oesch provided the project with a copy of his Matlab 100 inch mirror coating simulation as a real world application
203. ed lab area for EFM experiments and acted as advisor for software questions Paul Krehbiel NMT Physics Department Acted as technical advisor and director of project William Rison NMT EE Department Contributed his time and knowledge to assist in using Matlab tools W David Rust NSSL Furnished EFM and Labview program for experiments Invaluable resource of knowledge concerning project s history Bill Winn NMT Physics Department Provided documentation of past work on project Table of Contents Chapter 1 A Brief History of Thunderstorm Data Analysis as Relates to Electric Field Meter Data Collection 6 Proj t DOSCrIpHOlls oes usos etui nua beers Das demi eS ui bn o oed ato oe Sone 6 Definitions and AGFOHY MS 1 rie eouo etx tex da Che Nene Cen ono AVR EERRE 7 Literature Review nee N cae Pale oer ia ee ds edd 8 lario Em 9 Chapter 2 Past Findings on Electric Field Meter Data 10 Chapter 3 Steps Toward Development of Data Analysis Anoma 14 25 0554 Lie iss rte e aa e esM Or bap 12 Seth Schuyler Menuing and User Interface Llsuusuuus 12 Chris Reiten Matlab Coding and Digital Signal Processing 13 Ginny Martin Experimentation and Communication 20 Chapter 4 Final Product yess oedoci m Dp eI ER eiosaaloan Exod Bus od 23 Chapter 5 Future Work x 2r eee REEE 3 Appendix Einal ode s ra Eos fetetu netosb tonem VS ERI uU 32 List of Figures
204. elescope a development computer and access to NRAO facilities We would have to manipulate the software to track an assortment of astronomical objects and display telescope data in a manner that is both educational and easy to understand for the general public We would have to implement an interface that would be foolproof and easy to use The interface had to control the telescope without a keyboard or mouse but would still allow public interaction It was suggested that this be done with a touchscreen The top priority of the entire project was to educate the public about radio astronomy The Project budget was briefly discussed and we were told that a distinct ceiling was not set but all expenses must be approved by the NRAO This would also include installation of things like support piers conduits and electrical see e mail Appendix A page 63 2 2 Assignment of Work A clear idea for this open ended project was obtained when we presented our initial ideas attending a meeting with Visitor Center Committee From this meeting we were also able to finalize the Statement Of Work contract After outlining the customer s requirements for the SRT Project we had to decide how to break up the project among the team The logical divisions of work seemed to be in the areas of control software touchscreen interface educational tutorials and final installation design In addition some tasks were determined to be immediate needs that the entire
205. ems board followed by the many sensor connections and the 50 pin ribbon cable connector which connected to the Prometheus I O The Altera chip was programmed to interface to the PC 104 bus as well as generate a PWM signal for each motor to control the speed and two pulse accumulators which were used to tell motor position and derive motor speed The Altera chip was connected to the PC 104 bus and two four pin connectors for motor encoder input and two two pin connectors for PWM and ecru soul TONO GStUME Figure 12 This is our Breakout Board isn t it beautiful the connections were wire wrapped when available and soldered otherwise The Altera connections included connections to the clock from the PC 104 bus which was divided down for use on the Altera Chip as well as connections to the Address and Data Bus lines on the PC 104 bus Since we were using 16 bit addressing connections to the 16 bit addresses on the PC 104 bus had to be added which meant connecting up the C and D rows of the PC 104 connector as shown in the schematic in Appendix D All sensor connections were ran to the Breakout Board including distance sensors flame sensors the white line sensor the ultra sonic distance sensor and the connection to the relay for the fan For the sensor connections used 21 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems audio connectors wherever possible so that there could be no bad connections that would
206. en was the LM34CAZ The LM34CAZ is a precision integrated circuit temperature sensor The output of the sensor is linearly proportional to the Fahrenheit temperature The sensor gives out a 10mV change per degree Fahrenheit An example of this is at 83 2 degrees the sensor gives out 832 volts The accuracy of the sensor is plus or minus one and one half degrees Fahrenheit The price for these sensors is 6 64 each These factors make this temperature sensor a good alternative to higher price temperature Sensors 12 2 1 3 Data Collection System Design The data collection subsystem collected the data from the temperature sensors passed the values into a computer and then performed calculations on the data to return a useful value The temperature sensors returned a temperature value in the form of a voltage so that voltage number had to be calculated with other known values to produce a power value To solve this problem we chose the DI 194 A D board from Dataq Instruments This board did not provide the level of precision our customer desired but due to budget constraints it served its purpose very well The DI 194 is a four channel eight bit A D board It connected very simply to the external serial port of a computer and it received its power from this port The A D board was then interfaced with Microsoft Excel through the use of a macro written in visual basic Huang A copy of this program can be found in the appendix on page twenty thre
207. endly Note The various types of batteries and their performance characteristics are shown in Table 1 Capacitors Capacitors are devices that store energy by maintaining two charged interfaces of opposite polarity separated by a dielectric electrolyte The energy stored by capacitors is related to the charge at each interface q and potential difference del V between the electrodes Ecapacitor q V Since capacitors store charge at the interface only rather than within the entire electrode they tend to have lower energy densities The charge discharge reaction is not limited by ionic conduction into the electrode bulk so capacitors can be run at high rates and provide high specific powers Typical numbers for capacitors and batteries are given in Table 2 Since there is no bulk change in the electrodes the charge discharge reactions can typically be cycled many more times than batteries 10 cycles device have been achieved 23 Supercapacitors The operating principle of the Supercapacitor is based on an electric double layer appearing at the interface between activated carbon particles and sulfuric acid solution as electrolyte An ionically conducting but electrically insulating porous membrane separates the two electrodes Conductive rubber membranes contain the electrode and electrolyte material and make contact to the cell Several cells are stacked in series to achieve the 5 5V and 11V rated voltages Since Supercapacitors e
208. enever any subset of the 24 available digital input lines goes low Unfortunately this is not the case the DIO Interrupt has nothing to do with the 24 digital I O lines This interrupt is only generated when a separate pin the External Trigger line goes low If we want an interrupt to be generated when any of the DIO lines go high we would have to tie them to the Digital I O lines as well as the External Trigger line through an N channel NOR Gate Instead of doing this we only tied the tone decoder digital I O line to the External Trigger When we catch this interrupt we simply place a default value into the appropriate RT FIFO 36 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Now that all of the data has been gathered we need to place it into the appropriate RT FIFO for the user space control algorithm to read We do this with the RT Linux function rtf_put int fifo char data int size Finally the thread must go to sleep If the thread is periodic the thread must sleep until the next period using pthread wait np This function tells the thread to wait until the next period or until the user space control algorithm wakes it up If the thread is non periodic the thread is put to sleep by executing pthread suspend np pthread self which tells the thread to sleep until it is explicitly woken up by the control algorithm 37 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Hindsight is 20 20 Breakout B
209. ent Appendix IV Code for transmission and data acquisition this program will send the value on A2D 4 to Digital port 2 ones digit first with a buffer of 10 1 s folled by a zero It also sends the quad signal to the transmitter include lt aCore tea gt include lt aDigIO tea gt void main int a 0 x 0 t 0 m 0 f 1 mis 0 int zz 500 aDigital Config 2 DIG OUTPUT aDigital Config 3 DIG OUTPUT aDigital Config 4 DIG OUTPUT aDigital Config 0 DIG INPUT while 1 print value a aAnalog Read Int 0 display_int_dec a display_char n sleep zz cue the trasmitter freq if f 1 aDigital_Write 3 0 aDigital_Write 4 0 if f 2 aDigital_Write 3 1 aDigital_Write 4 0 if f 3 aDigital_Write 3 0 aDigital_Write 4 1 if f 4 aDigital_Write 3 1 aDigital_Write 4 1 f 0 cue the reciever while m 10 aDigital Write 2 1 sleep zz m m l m 0 aDigital_Write 2 0 sleep zz end cue start xmit if mis 0 a 0x0001 mis 5 else mis mis 1 for x 0 x lt 10 x x 1 if a amp 1 aDigital_Write 2 1 else aDigital_Write 2 0 a a gt gt 1 sleep zz end xmit while m lt 10 aDigital_Write 2 0 sleep zz m m m 0 return to idle f f 1 change freq Appendix V Code for the rec
210. ent to the Altera board to clock it in to the PE s internal registers In our initial design of the interface we used simple counters and demultiplexers to route the incoming data to the appropriate register Figure 6 below illustrates how this was done Register 0 Register 1 data in 0 3 Register 2 Register n TxRx clock Figure 6 Initial Design to Route the Incoming Data 13 As shown data is input to the demultiplexer The value from the counter determines which output of the demultiplexer the input is placed on The demultiplexer also provides an enable signal for each output The outputs and enable signals go to different 4 bit registers When the enable line for one of the registers is high and a rising edge of TxRx clock occurs the data is clocked into the register On the falling edge of TxRx clock the counters are incremented and the input data is routed to the next register Three of these demultiplexer and counter blocks for routing the incoming data were used for receiving the configuration bits the data constants and the input data Only one of these three block were enabled at a time based on the value of the transmission state bits For sending the output from the Processing Element back to the computer we used a mulitplexer and counter as shown in Figure 7 From PE data out 0 3 TxRx clock Figure 7 Initial Design for Sending Data to Computer As shown the output data is broken
211. er Transform of B Hg after filtering c magnitude 0 500 1000 1500 7000 3 3500 shoo 3500 4000 4500 5000 requency in terms of samples Figure 8 Lowpass filter for quadrature demodulation In the program the quadrature demodulation technique was implemented using the Hilbert transform function in Matlab The Hilbert function shifts a sinusoid by 90 degrees The filtered and interpolated Hg switch data was shifted using the Hilbert function giving the two separate cosine and sine waves This was then used to demodulate both the B field data and the E field data When demodulating the E field I had to keep the two phase components separate This told us the vertical component and the horizontal component of the E field The cosine component of the quadrature demodulation told us the horizontal E field and the sine component of the quadrature demodulation told us the vertical 18 component of the E field Because the Hg switch changes before it is straight up and down it is not an exact description of when the spheres were vertical To account for this we had to design a FIR filter that would pass all frequencies and have a constant phase shift This would shift the Hg switch data by a certain phase angle to match up with the Magnitude response of phase shift filter 1 05 Magnitude e e a 0 3 0 85 i 1 4 L 1 L L 150 100 50 50 100 150 Input frequency in degrees Phase response of phase shift filter Phase shi
212. er is a divider which has perforations with a coffee filter The coffee filter allows added protection against soil entering into the water holding area The idea behind the filtering is to implement water recycling The total dimensions of the planter are 9 x 13 Under the planter is a separate water tight compartment to house the controller circuitry and battery It has dimensions of 2 5 x 9 x 9 A 6 pin wiring harness was designed to allow communication between the pump moisture sensor and solar panel Inside the lower circuitry containment area Velcro was used to hold the components in place 3 3 2 Pump A pump was necessary in effort to get water from the holding area to the soil We looked at two methods to perform this task a pump and a bladder pressure method 28 The bladder method involves steady watering which does not quite fit our needs since we must provide variable watering for different plant categories The pump turned out to be the best fit for our needs We found numerous pumps that would accommodate our needs After comparing cost voltage and current consumption we decided to go with a Techtronics 6V 500mA water pump The pump will pump 12 oz min at a height of 42 cm We ordered the pump and tested it as soon as we received it The pump did live up to its minimal expectations and thereafter we decided upon it as our final choice for water pumping The cost on this pump was also very minimal in comparison with it
213. er to make a decision upon which transmission pair to use compared spec sheets from several vendors to narrow the selection down to just a few choices In this phase also spoke with the Linx Technical Department to verify compatibility of their prefab receiver with the Holtek encoding was planning on using for the design Another task worked on was the interface circuitry for the Linx Module In this stage looked at different ways to interface the Linx transmitter and receiver with the input sensors and taillights Two separate designs with the linx module were developed They incorporated either two arrays per side or additional logic after the pre fab receiver and thus were thrown out after the preliminary design phase Along with the interface research looked into a manner in which to send a PWM signal through the transmitter for the turn signal flashing After the group decided upon using the Glolab transmission system and going with one LED array on each side the Linx prefab receiver was set aside and investigation began more on the way our output logic would be attacked Brandon Lattimore and spent several days working on finding the most efficient way to simplify the logic the group initially created Appendix F shows the simplest design that we came up with The four chip design was prototyped on a breadboard and was viable but an inexpensive programmable logic device would allow for more options and expansion in the future
214. eral ways to utilize this bus DMA IRQ driven and simple memory mapping We decided to go with memory mapping the registers we needed as it is the simplest to implement The Base Address of the Breakout Board is 0x340 This address was decided because it is below the 1MB boundary making driver programming easier and this address was not in use on neither the development machine nor the Prometheus The Position Accumulators are at addresses 0x340 and 0x342 16 bit addresses and the PWM command registers are at 0x344 and 0x345 8 bit addresses 29 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Altera Code for 16 bit PC 104 Interface Kyle Lamb When we were working on the PC 104 bus connections on the Breakout Board we ran into troubles with Altera timing issues using the 16 bit reads was able to write a nested if statement in Altera which solved the timing issue by bringing the IOCS16 line low before IOR or IOW went low With the 16 bit reads fixed we were able to continue with the development of the PC 104 bus and enable read in 16 bit data from the motor pulse accumulators Installation of RT Linux Bill Willems We are using RedHat Linux 7 2 with the 2 4 19 pre6 kernel and RTLinux V3 1 We chose to use such a bleeding edge kernel because of several USB controller fixes available only in this kernel Due to the limited size of our Compact Flash drive we had to find a way to get a minimal set of instructions To do thi
215. eramic resin material Fastblock 301 Kirkhill TA Co Valencia CA which had a low thermal conductivity of 0 07btu ft hr F at 400 F The inner layer of insulation was composed of Cerablanket Thermal Ceramics Co which had a thermal conductivity of 2 05btu ft hr F at 1800 F This insulating casing was capable of maintaining an internal temperature near 150 F for thirty minutes when exposed to a continuous flame Electronic components where found that are rated to operate at temperatures above the 150 F originally required The electronics employed a K type thermocouple to read the external temperature a Brainstem processor Acroname Co Boulder CO an op amp circuit to amplify the signal from the thermocouple and a quadrature transmitter Maxim Dallas TX A lithium battery in conjunction with a thermal switch served as the power supply Abstract A prototype of a heat detector was designed for the Naval Air Warfare Center NAVAIR China Lake CA The purpose of the prototype was to provide early detection and warning about the presence of fires and their relative danger to ordnance This device was to have an indefinite shelf life before activation Once activated by heat the device transmitted via a radio frequency RF signal an encrypted data set The data set provided stored inventory data related to the ordnance and real time temperature readings This signal would be transmitted through the material casing and the noise pro
216. es The coordinates are specified in the catalog file in right ascension declination form and to transform these into SRT s native azimuth elevation coordinates accurate knowledge of current time is absolutely necessary In addition it is necessary to specify the coordinates of the telescope s location on earth With knowledge of its position both physically and temporally the telescope can figure out where to point Physical location would be easy to determine A GPS reading at the dish location will be more than accurate enough to point the dish Deciding how to update the clock required much more work One option was a radio receiver which would plug into the computer s serial port and update the clock using the signal broadcast by the National Institute of Standards and Technology NIST in Fort Collins Colorado I ruled this out due to the fact that we could not test it The receiver was specified for use in wood frame buildings and it seemed highly unlikely that it would work in our lab in Workman Another option was software which would log in to one of NIST s time servers over the internet and update the clock This was attractive because it would be cheap and would require no additional hardware NIST distributes a 19 very nice program for free which seemed to suit our needs perfectly It could be set to update the clock as often as once per hour more than enough to keep SRT pointed accurately This did not work however due to t
217. es of thermocouples are effective over different temperature ranges A K type thermocouple was found to be best for the requirements of this project which required a temperature range of 250 F to 1600 F A K type thermocouple is rated up to 1800 F It is composed of Chromega and Alomega For the purpose of this project the end wires were welded together and left as an exposed junction A K type thermocouple is a general purpose thermocouple which is very dependable in the necessary range 2 4 Central Processor The project required a central processor to compile the data obtained from the temperature sensor and transform it into a useable form for transmission This processor needed to have the ability to change a voltage into a binary stream transmit and receive digital numbers and compute basic math functions These are the minimum requirements necessary for this project to meet specification It was decided to use a Programmable Integrated Circuit PIC from Acroname Incorporated known as The Brainstem to implement the needed functionality The 4 See section 3 4 1 for selection criteria used in selection of the PIC 12 Brainstem is an integrated controller running of a 40 Mhz RISC processor that has five ten bit Analog to Digital Converters A D five general purpose digital Input Output I O ports and 10KB of EEPROM EEPROM is on board memory that can only be erased electronically This ensures that programs can be
218. etween the charge amplifier and the idealized mercury switch the angle that the mercury switch needed to be shifted by could be determined This is important to the quadrature demodulation technique we planned to use to separate the horizontal and vertical components of the electric field This process finds the component of the charge amplifier data that is in phase with the mercury switch and the component that is 90 degrees out of phase Since we wanted these two components to represent horizontal and vertical we had to correct the mercury switch data With the vertical and horizontal components of the E field at the instrument known just the direction of the instrument remained to be determined Since the earth s magnetic field is basically constant across the instrument throughout the flight the Hall effect sensor s data was investigated for this information There were two primary frequencies in the B field data the rotation and precession of the EFM Only the low frequency is needed to determine direction so the data is passed through a filter to extract it This yields a sine wave that describes the direction of the EFM relative to the Earth s magnetic field Figure 11 shows 2 actual plots One of the raw E field and mercury switch data and the other of the horizontal and vertical components of the E field The raw data is 24 plotted so one can monitor the E field to see if any odd patterns in the components are due to lightning disch
219. ey pin cut Fiaure 20 PC 104 Connector 44 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems AEN Address Enable line This line indicates whether or now the address that is now on the address bus pertains to a read or write from memory When this line is high the SBC is reading from memory When it is low the SBC is reading from the ISA or PCI bus The Position Accumulators are simple 16 bit memory mapped up down counters at address 0x340 and 0x342 When the SBC requests this information it places the address on the address bus and brings AEN low We need to recognize this action and bring IOCS16 low because this is a 16 bit read or write before the SBC brings the IOR or IOW line low Once the IOR line is low we can place the data onto the data bus If the IOW line goes low we do not do anything because the Position Accumulators are Read Only registers Once our address disappears from the address bus we let go of the data bus This process is depicted below in Figure 20 databus addrbus AEN BALE 10CS16 Figure 20 Timing Diagram for a 16 bit Read amp Write on the PC 104 Bus The PWM command registers are 8 bit memory mapped registers at addresses 0x344 and 0x345 When the SBC requests this information it places the address on the address bus and brings AEN and IOR or IOW low If IOR goes low we place the data onto the data bus If the IOW line goes low we read 45 RTLinux On A Small Mobile Robot
220. ey could get the most out of their visit to the VLA 1 3 The Small Radio Telescope The project offered by NRAO involved the adaptation of an existing radio telescope kit intended for students and amateur astronomers for use by the public at the VLA Visitors Center The Small Radio Telescope SRT kit was developed by the Massachusetts Institute of Technology MIT as a fully functional radio telescope and is sold for use primarily by universities and amateur astronomers The kit consists of a commercially available satellite dish originally made for satellite television systems along with a custom azimuth elevation drive L band receiver and a controller which allows the telescope to be interfaced with a PC The control software is furnished as open source Java code and provides full control of the SRT through a simple graphical user interface GUI The SRT Figure 1 3 1 collects data from the universe in the radio part of the electromagnetic spectrum The radio spectrum ranges from 1mm to 30 meters Data can be collected in both the day and night The emissions that are measured by the SRT come from the astronomical objects that are in the universe The radio telescope uses a large parabolic metal dish to reflect radio waves to the sub reflector The sub reflector then directs the radiation to the feed at the center of the reflector Then the signal is amplified through the receiver A computer then processes the amplified radio signal The SR
221. f the SRT project followed by a section describing work that performed as a group 3 1 Gerald From the start of this Senior Design project my tasks fell under two main categories communication and education It seemed that more effort went into communication with people to pool information and to organize the many tasks that would bring the SRT Project to completion Many people were involved in completing this project and gave it a professional polished look The education aspect was certainly a challenge as I had no prior experience with radio or any other astronomy The idea for educational tutorials from PowerPoint slides had its basis in many courses I have taken in my undergraduate studies at NMT Initial communication and coordinating efforts started with some personal contacts I had at Tech s Physical Plant I took the task of finding out about putting the antenna on a roof at Tech We thought our hardware was due to arrive soon and getting an area laid out before the equipment arrived would be a good idea At Tech s Physical Plant I was put in touch with the architect for Tech Together we surveyed the roof of the Workman building for installation of the SRT and discussed possible alternatives We found we could run the cables down through the existing conduit We checked the blueprints for access points for the cables Since the Workman building was designed for experiments to be mounted on top of it along with penetrations
222. fault zpl home none possible zpl home dirs for directory in common zpl home dirs if os path isdir directory possible zpl home dirs append directory print The following possible ZPL HOME directories were found for directory in possible zpl home dirs print t s directory print ni if len possible zpl home _ dirs gt 0 default zpl home possible zpl home dirs 0 The character sequences at the end of the following two lines suppress a newline from print and remove the newline from readline print Which ZPL HOME directory do you wish to use print default s default zpl home zpl home sys stdin readline 1 print if len zpl home 0 and len possible zpl home dirs gt 0 zpl home possible zpl home dirs 0 return zpl home end of zpl setup zplhome def zpl setup zpltarget zpl target default zpl target x86 linux while not zpl target in valid zpl settings keys zpl target print The choices for ZPLTARGET are for target in valid zpl settings keys print t s target print print Which ZPLTARGET is appropriate for this configuration print default s default zpl target zpl target sys stdin readline 1 print if len zpl target 0 43 zpl target default zpl target return zpl target 4 end of zpl setup zpltarget def zpl setup zplcommlayer zpl target zpl commlayer while not zpl commlayer
223. file and open Follow steps 8 and 9 Once the main layout has been set all other corresponding layouts will load automatically cass sun amp cygnus The main layout is cass2 lay Open this by choosing file and open Follow steps 8 and 9 Once the main layout has been set all other corresponding layouts will load automatically cass amp sun The main layout is cass2noCygnus lay Open this by choosing file and open Follow steps 8 and 9 Once the main layout has been set all other corresponding layouts will load automatically 4 Oncea layout is open a buttons is created by right mouse clicking on an empty space on the layout and selecting insert and choosing Macro A macro can be a list of keyboard or mouse actions or events to perform when the button is pressed or released The event is received from the controller by the button in the layout Each button can have many macros of actions to carry out 34 10 11 12 A button is modified by right mouse clicking on it and selecting properties All of the macros can be modified from this window For a more detailed method of creating or editing buttons select the help menu Text can be added to the layout by selecting insert and choosing Text This will not create more than one line of text at a time A layout is sent to the device by left mouse clicking on the yellow lightning bolt icon on the toolbar The message The layout has been sent to the device will appear clic
224. for power control and data cables to various points in the building the Workman roof seemed to be perfect Roof mounting would avoid obstacles from the SRT looking at the sky Another reason for roof mounting was to reduce the possibilities of tampering with the dish since locked doors control the roof access We went through some more brainstorming of ideas with our project advisor and supervisor on topics such as permissions for NRAO to have the antenna on Workman or another place on Tech property liabilities other design planning other politics and even other possible sites like Etscorn Observatory and my house We needed to find out if we would need administrative permission Brown Hall or if we would be infringing on any positions of Union workers from the P Plant or other government workers and whether we would need work orders to complete our installation Further tasks were assigned and I got to find out about mechanical cabling and power requirements When we finally obtained permission for the roof mounting we had to decide whether or not to order the base for it I also investigated the other technical aspects from data sheets for the SRT From my literature search I decided we did not need a base as the antenna would fit over the existing stanchion as in Figure 3 1 1 Existing Weatherproof Electrical Junction Box Existing Supply Power _ 2 fits over Existing 3 5in pipe ee and Control Cable
225. ft in degrees 150 100 50 50 100 150 Input frequency in degrees B field data See figure 9 Figure 9 Constant phase shift FIR filter With the help of Dr Rison and Seth I was able to create the filter in Matlab The filter is set up so that the user can input a certain phase in degrees and the Hg switch data will be shifted by that amount When the unshifted Hg data was used to demodulate the E field horizontal and vertical components the ripple from the horizontal component could be seen in the vertical component When the Hg data was shifted the ripple was gone 19 effectively isolating the vertical and horizontal components See figure 11 for the individual components My main role in the project was to write the majority of the code in Matlab that would do the data analysis With the help of Ginny and Seth our final product successfully met the requirements of our customer Ginny Martin Experimentation and Communication Initially all three team members had the same task understand the data As it turns out there an infinite number of solutions and we all took different paths I haven t touched Matlab since my freshman year so I had a considerable amount of catch up to do My team members graciously gave me a crash course on what they had learned in DSP Our customer gave us a diary file that one of his graduate students worked on years ago I stepped through this file and learned much about the data I brok
226. g off the coast of North Vietnam The Forrestal had been conducting combat operations for 4 2 days including a strike early that morning At 10 52 am the crew was beginning the second launch cycle of the day when a Zuni rocket from an F 4 Phantom was accidentally launched by stray voltage during the ignition of the Phantom It soared across the deck hitting a parked and armed A 4 Skyhawk The impact caused a rupture in the belly of the fuel tank on the Skyhawk spilling fuel and causing a chain reaction fire of planes parked on the deck The impact also caused a 1 000 pound bomb to fall off into the spreading fire Within a minute and a half the first bomb reached cook off and detonated killing the flight deck chief and the first wave of firefighters This initial detonation caused a massive chain reaction of explosions that engulfed half the airwing s aircraft and blew huge holes in the steel flight deck Fed by jet fuel and ordnance from other aircraft that were armed and ready for the coming strike the fire spread quickly Many pilots and support personnel were trapped and killed Meanwhile fuel and bombs began spilling into the holes created by the exploding ordinance spreading the fire further into the ship The fire burned for thirteen hours before being extinguished Once the fires were extinguished the extent of the devastation was apparent The most tragic event was the loss of the crew One hundred thirty four men had lost their
227. gate commercially available beam dumps to determine the optimal material and angle needed to meet the given specifications yet use commercially available materials and surface treatments Then we were to build a prototype capable of measuring the power output from a given laser and with a laser furnished by the customer demonstrate its capability 2 1 Subsystem Solutions The device consists of two rectangular plates subtending an angle determined by the designer The designer also determined the material thickness and surface treatment Several temperature sensors are attached to the outer surfaces and the temperature sensors are linked to an analog to digital board A D board 2 1 1 Laser Beam Dump Design Designing and constructing the laser beam dump This involved determining what material to use what dimensions the beam dump would have and the availability of each material First several materials were investigated Materials with high melting points and good thermal conductivity were selected The properties melting point thermal conductivity specific heat mass reflectivity of each material were documented for later calculations Since most of these materials were highly reflective it was decided that we should apply a dark coating to the inner surface layer of the material This would increase the absorption of the laser beam Two selections were black anodized and black paint Black anodized was very durable but it could only be u
228. ghts we used a LM 556 which has dual 555 timers 2 5 Relays As an added feature we gave the user an option of using the power transistors or the relays This gives the user an opportunity to use the LED arrays or normal incandescent light bulbs The user can accomplish this by positioning the jumper that is located at the inside of the box to the correct setting The LED arrays were designed in house and made specifically for our setup Each of the LED arrays is powered by a power transistor and requires only one input line and a ground wire The incandescent light bulbs are from a normal utility trailer and a relay switch powers each of the bulbs This setup requires the use of additional incandescent bulb on each side to run as continuous lights 2 6 Protel Once both the transmitter and receiver boards were tested we decided to use the PCB making software Protel to lay out the circuit design The reason why we decided to use Protel was because the group was vaguely familiar with it and it was easily accessible to us in the lab To lay the board out quicker we simultaneously laid the transmitter and receiver out on different computers The layout took a little longer than expected because we only knew the basics of the software We laid out the entire transmitter and receiver systems and then began to design the PCB part of it In order to save money on the PCB manufacturing we decided to put both the transmitter and receiver on the same
229. grams usr bin rsh usr bin ssh usr local bin rsh usr local bin ssh pvm root pvmgetarch path print print Configuring the PVM parallel library print i while not os path isdir pvm root and os path isfile pvmgetarch program pvm root pvm arch pvmgetarch program default pvm root none possible pvm root dirs for directory in common pvm root dirs if os path isdir directory possible pvm root dirs append directory print The following possible PVM root directories were found for directory in possible pvm root dirs print Nt s directory print if len possible pvm root dirs O0 default pvm root possible pvm root dirs 0 The character sequences at the end of the following two lines suppress a newline from print and remove the newline from readline print Which PVM root directory do you wish to use 41 o print default s default pvm root pvm root sys stdin readline 1 print if len pvm_root 0 and len possible pvm root dirs gt 0 pvm root possible pvm root dirs 0 pvmgetarch program os path join pvm root lib pvmgetarch if os path isfile pvmgetarch program pvm arch sS pvmgetarch program else print Could not find file s let s try this again n o pvmgetarch p
230. group would address These included e Set up development computer with network capabilities e Decide on control software options e Download NRAO web links and VLA Weather Station program in design e Acquire the Java compiler the open source Java code for SRT tracking software e Acquire Microsoft Office for educational slideshow presentations in PowerPoint e Obtain a touchscreen for input and control e Communicate vehemently with the vendors to remind them to send the SRT kit e Obtain permission for testing location Viable locations discussed and debated Once we had a breakdown of the individual tasks we could assign them according to our levels of expertise in each area Sam would work on the Java software Brian would work on the touchscreen interface and Gerald would be responsible for developing educational material and coordinating installation at the VLA 3 Discussion of Individual and Group Accomplishments After initial contact with our client we devised methods of putting together the tasks that the project entailed The first task was to break up the work among the group members This was accomplished by evaluating our strengths and weaknesses In doing this we tried to consider each group member s strengths Sometimes our group was lacking any knowledge of a subject but we plunged into research of the topic anyway and found some valuable information The following sections are our individual personal accountings o
231. gt Out GND IF In VCC THEN Digital Filter Machine ELSE Digital Filter Machine END IF s4 gt Out GND IF In VCC THEN Digital Filter Machine ELSE Digital Filter Machine END IF s5 gt Out GND s1 sl s0 s2 s0 s3 s0 s4 s3 s5 s3 Bitsoi Lamb Willems s3 s4 s5 50 RTLinux On A Small Mobile Robot IF In VCC THEN Digital Filter Machine s0 ELSE Digital Filter Machine s3 END IF END CASE END div_clk tdf Divide the Clock Down The PC 104 Bus has a 14 31818 MHz clock Divide it down to get 1 Mhz and 128 kHz clocks Bill Willems 1 3 02 SUBDESIGN div clk clk in INPUT pwm cilk OUTPUT filter clk OUTPUT VARIABLE div 2 DFF div 4 DFF div_ 8 DFF div_16 DFF div 32 DFF div 64 DFF div 128 DFF BEGIN div 2 clk clk in div 4 clk div 2 q div 8 clk div 4 q div 16 clk div 8 q div 32 clk div 16 q div 64 clk div 32 q div 128 clk div 64 q div 2 d div 2 q div 4 d div 4 q div 8 d div 8 q div 16 d div 16 q div 32 d div 32 q div 64 d div 64 q div 128 d div 128 q pwm clk div 128 q 111 86 kHz filter clk div 8 q 1 7857 MHz END o ge oe o oe Bitsoi Lamb Willems 51 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems pos pulse accumulator tdf Pulse Accumulator for Position Registers Accumulator value is latched in pcl04inter
232. gures 8 and 9 As more processors are involved in the execution of the program the execution time reduces but the I O time to reassemble the data and output the result file on the server node increases The end user is advised to only output data to the final result file that is critical to the analysis of the results If unnecessary data is collected and output the network and disk overhead in gathering and writing this data will eventually cancel the performance benefits of the parallel execution 13 140 120 100 80 I O Time E CPU Time seconds 60 40 20 0 2 3 4 5 6 7 8 Processors Figure 8 Performance of mirror resurfacing simulation program using the MPI library 140 120 100 80 E l O Time El CPU Time 60 seconds 40 20 0 2 3 4 5 6 7 8 Processors Figure 9 Performance of mirror resurfacing simulation program using the PVM library 3 Team Member Contributions and Finished Product 3 1 Michael Berg My first contribution was a basic set of instructions to follow while installing Linux on the nodes This allowed all of the nodes to have almost identical configurations for development and testing and it made sure that the libraries and tools that we would need were available for use Kevin then took over modifying this list as we made changes to the cluster My next contribution was obtaining and installing ZPL ZPL comes in several variants For example under Linux there are ZPL downloa
233. h lower density than ceramics 10 Information on insulation materials that were researched was listed in Table 2 1 The materials are arranged from highest thermal conductivity to lowest It should be noted that there are general trends found in the density and cost columns With decreasing thermal conductivity there was generally an increase in cost i e the better the insulation the more expensive it was Additionally density generally decreased with decreasing thermal conductivity This may be attributed to the fact that the large volume of air in porous materials both reduces the thermal conductivity and the density Cotronics Liquid 110 4 5 50 Lo nnd MEE NENNEN MEE Thermal Ceramics 2 05 at 1800 F Tom ECC e Yttria Stabilized Cotronics Rescor 1 at 2300 F a RR Fastblock 301 39 9 44 9 0 07 at 400 F 40 0 04 at 78 F MicroTherm 21 2 25 0 0 22 at 800 F Moulded Form Space Shuttle Tiles p 14 028 BTU ft hr F at 70 F 2200 Table 2 1 Materials Considered bulk price The information obtained from research into these materials was used in choosing materials to test and the tests to be completed 11 2 3 Temperature Sensor A thermocouple is a device that can be used to measure temperature It is composed of two metals that when attached at one end produce a small voltage on the order of millivolts There are many types of thermocouples composed of different metal and alloy combinations These different typ
234. he front of the robot in addition to the Sharp GP2D12 IR distance sensor The removing of the wire allowed for better positioning of the wire and more consistent results The wire did extend to the back of the robot since there is no way of knowing if we made contact When the wire made contact with a surface surrounding the robot the output of logic high was sent to the Prometheus s digital input The difference in the logic depends on the area of contact The contact areas are depicted in color in Figure 14 25 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Outputs from Contact Switches Switch 1 Logic High 0 2 Logic Low 1 1 1 0 0 0 0 2 3 0 1 1 0 0 0 4 Figure 14 Color Code for Contact 0 0 0 0 1 1 Table 2 Output Logic for Contact Areas in Fig 14 The outputs from the switches are shown in Table 2 The bump sensor design worked consistently and there was no significant difference in the output The positioning of the wire played the major role in maintaining consistence outputs The wire used was the main problem in this design That was the only difficulty encountered The overall design was successful Hardware Connection Kyle Lamb The Robot came pre assembled so there were not a lot of hardware connections that needed to be made but a new power supply needed to be added to supply power to the Prometheus and the variety of sensors that would be using In orde
235. he Beowulf Cluster To accomplish this task I downloaded ISO image files of the RedHat 7 2 OS burned these images to compact disc installed the OS on every machine in the cluster and set up an initial OS configuration that would facilitate the overall goals of the project To ensure that working cluster configuration files under the old OS were not lost during the upgrade I upgraded half of the machines to RedHat 7 2 stored backup copies of all relevant configuration files to these machines then upgraded the other half of the machines Realizing that the final Beowulf Cluster product would include a recovery method that would instruct the user in repairing nodes on the cluster I recorded all steps taken in the RedHat 7 2 installation process This measure has ensured that no information vital to the OS installation process was forgotten when the actual creation of a recovery method took place To identify possible performance improvements in the Beowulf Cluster design extensive 17 benchmark data had to be compiled I was in charge of running all benchmark programs compiling all benchmark data and coming up with easy to read graphs of this data for inclusion in the final Beowulf Cluster user manual Each benchmark program was run three times with the same parameters data set number of processors etc and these three results were averaged to arrive at a final benchmark value I wrote many miscellaneous shell scripts to automat
236. he Careers at NRAO web page has the current job openings at the NRAO The Radio Astronomy Images web page displays celestial images that the VLA produced The back to main button will exit the opened web page and send the user back to the main layout The page up and page down buttons will allow the user to scroll up and down the web pages Figure 3 3 6 The VLA Web Links Layout 24 This feature was necessary because the web pages could not fit on one screen The user will only have access to these three web pages The web pages are accessed on the computer as a htm extension file This will increase the reliability because it does not depend on the network Each main layout had its own identical set of corresponding sub layouts This was necessary to link the layouts together because the sharing of layouts would create errors when trying to control the programs One problem with the touchscreen was when a button was pushed which had a command to change layouts If the button on the second layout were in the same place as the button on the first layout the button on the second layout would also be selected This led to the problem of choosing an option that was unwanted I tried to fix this problem by adding a delay after each layout I thought a delay would give the touchscreen time to change before the selection becomes available but this did not work so I was forced to have buttons in different locations when a layout changed I thin
237. he EE Department s firewall I explored many options downloading and testing numerous freeware and shareware internet time update programs None of them worked through the firewall I checked with Chris Pauli and found that changing the firewall settings was not an option One thing I noticed about the NIST software was that it used TCP port 13 with no provision for changing this setting I downloaded the source code for the program and changed it to use port 80 but found that this would not work The time server the program accesses only listens to port 13 As a last resort I talked to James Robnett one of NRAO s computer experts He told me that NRAO maintains two time servers on its network for the specific purpose of updating PC and workstation clocks He said that connecting our PC which runs Windows 2000 would be no trouble at all once it was on the NRAO network The problem was solved but I spent about two weeks on it when all it took was a five minute conversation This was quite a learning experience 3 3 Brian My first task was to set up the NRAO computer The computer did not have a java compiler so I downloaded one on the internet A d java compiler is required for the SRT software In order to compile w a program the instruction set CLASSPATH must be typed in V 1 the directory with the file before the program can be compiled I tried to add the directory with the java code to the CLASSPATH variable but this did not
238. he next page highlights these various data paths Li In2 B Int E ind Lu d In2 Ini Ind DR1 DRO v sel_mux1 set_al_shift sel_xb L3 4 alu v AL_ SET 3 C3 purs X sel xb1 LE XBAR5x3 sel xb g MUL CLIP a op C ep SHIFT Ka don in ALU OUT set mul shift i r1 set alu shift set mul round FY et alu round set mul clip Set alu e cue REED P set alu clip CLIP tos MUL_Sv E ALU CLIP E Delay Path W Multiplication Path W Arithmetic Path Figure 3 Block Diagram of Processing Element Highlighting the Data Paths As shown in the above figure there is a data path that goes from the multiplication path to the arithmetic path where the MUL_OUT node is an input to MUX1 This can be used to implement Multiply and Accumulate routines which are useful in signal processing applications There is also a path that goes from arithmetic to multiplication where the ALU_CLIP node is an input to MUX2 There are a total of 10 different components in the Processing Element Below are functional descriptions of the components ALU This is the arithmetic logic unit Based on the alu op bits it performs the following operations addition subtraction and nand or nor xor xnor not negate pass X operand pass Y operand AL SHIFT This component shifts 48 bits It can shift both left and right It can shift logically and arithmetically The set al shift bits determine what kind of shift to do and how much the operand is shifted CLI
239. he sensory circuitry that needs to be added to bring the outputs to appropriate levels would have to be extremely precise This is because many turn signal shafts have an extremely small range of motion thus the GP2D120 would have issues picking up those small angle changes Chapter 4 Individual Contributions This chapter is broken up into sections describing each team member s individual contributions to the overall success of the project 4 1 Tasks Accomplished by Brandon Lattimore The first task was given was to do the sensor research and selection was to research several possibilities for sensors to be used for both the brake and turn signals They would need to be affordable reliable and compact They also could not interfere with the driver s normal habits The customer also requested that we try out several different ways to do each task with the possibility of giving the customer the option to pick and choose the sensor that would work best for his vehicle For the brake sensor looked for sensors that would not get in the way of the driver s foot Along the same lines chose sensors for the turn signal that would not interfere with the driver s usual turn signal routine searched through Digi key online web sites and through previous work to determine which sensors to order ordered a total of about nine sensors that were wired up and tested for feasibility David Byrd found a few more sensors to add onto a later order
240. hese tests involved using a torch to heat one end of the thermocouple while the temperature and the voltage at the output end where measured It was determined to yield a maximum output of 60mV 0 06 Volts The next step was to use the output range of the K type thermocouple and convert it to a voltage signal that the internal electronic processor could use Calculations to determine the gain can be seen in Appendix VI This gain was determined to be 67 LMC6442 Thermocouple Figure 3 5 Schematic of Temperature Sensor Circuit 23 The next step was to implement the circuit The circuit was designed using a two stage op amp circuit The first stage was a difference circuit which took the difference between two signals The two signals were provided by the two different metals in the thermocouple as shown in the schematic in Figure 3 5 The second stage was a non inverting amplifier with a gain of 67 The requirements of this project limited the choice of operational amplifiers or op amps that could be implemented The source voltages in the device were limited to ground and 5 volts This meant that a single source op amp had to be used There were also limitations due to the nature of the signal that was being amplified The thermocouple output was on the order of millivolts which a Rail to Rail op amp had to be used to accommodate this A Rail to Rail op amp is an op amp that can have input signal voltages that are eq
241. hey all had a downward trend i e as current increases resistance decreases which makes sense as the voltage is increasing This is the comparator I was looking for as it compares one voltage to another I set it at 15k since that is the highest resistance about 14 25k at the lowest current If it is above the threshold then the soil is dry If below consider it wet In addition I can use the lower current 1mA due to the widest marginal behavior very high resistance for dry dirt near 30k to moderately low resistance for wet dirt 14 25k So 1mA is good and I thought maybe even lower The lower the better since we need to consume as little power as possible The maximum power allowed running the sensor now would be about 3mW power 1mA 2 28k The minimum voltage necessary is about 3V voltage 1mA 28k The next step was to connect all this to one of the A D 17 converters available on the HC12 microcontroller and realize the HC12 s input impedance 3 1 2 Plant Species Investigation In addition to developing the moisture sensor I investigated various plant species mainly their water intake level Fig 9 Each plant species display various characteristics and are inconsistent in comparing with others The moisture intake level is the amount of water a plant can should receive which are typically categorized as low medium or high This as well as plants humidity tolerance is the typical charac
242. his device is required since all of our analog components require a 5V input 2 1 5 Mosfet Switches To limit our power usage we implemented some switches which will provide power to components such as the adjustable program switch and the moisture sensor only when power is needed The use of switches will then consume power by not supplying power to devices that don t need to be continuously on To do this we used analog Mosfet IRF9510 switches From the microcontroller we can then send a signal to 11 activate the switches when power is desired at various components 12 Chapter 3 Member Contributions 3 1 Michael Carpenter My contribution to this project includes research and development of the moisture sensor distinguishing plant species and helping Rocky incorporate the sub programs into the HC12 and power distribution Our customer advised that I spent time in the fall semester researching on what type of moisture sensor to use or develop try to categorize plant species if need be in some fashion for watering and scavenge ways to distribute power accordingly to utilize as little as possible Although I got off to a slow start in the spring semester due to many mishaps I still was able to pull through and make my decisions on all three aspects 3 1 1 Moisture Sensor My contribution of the project includes developing a moisture sensor as well as researching the various plant species moisture intake At first I decided to ap
243. ht hand corner along with the coding for the tasks that everyone black will participate in The longest tasks paths have been divided up according to experience and preference The shorter paths have been assigned intuitively and thought of as challenges Some tasks have already been accomplished while some tasks which were scheduled for later were moved up and started on already Still there are some tasks that had to be bumped later because these time constraints have been relying on outside sources that have bogged down Following the two full versions Updated and Previous of the time line Gant charts are listed short explanations of the individual tasks FIRST SEMESTER Tasks Build and Mount SRT Assemble the SRT kit and mount on Workman for testing Run Cables Pull cables through existing conduit to second floor lab Setup Development PC Install necessary software on PC provided by NRAO Network Address Connect development PC to Tech network Java Compiler Install Java Compiler on development PC Simulate Dish Operation Use SRT software to simulate tracking and data collection Develop Public Use System Set up interface that the public can use to control SRT and learn about radio astronomy Research Order Parts Find out about necessary components particularly a touchscreen Touchscreen Implementation Set up touchscreen for public use with intuitive commands and attractive appearance 39 Interface Receiver Data Process raw
244. ia in determining energy storage requirements was that the storage device had to be cheap easily rechargeable has a high energy density has a long life cycle and has low to no maintenance So after a considerable amount of time looking for ways to fulfill the requirement I narrowed our choices down to chemical batteries and or supercapacitors Below is a list of commonly available batteries and capacitors as well as their benefits and drawbacks Batteries Sealed Lead Acid It has many benefits It is inexpensive reliable and relatively forgiving to overcharging and deep charging In addition it has a fairly long cycle life has high energy densities and its cells have a nominal discharge potential of 2V However it does have its drawbacks It has a relatively low power density and relatively rapid self discharge 21 Nickel Cadmium Ni Cd It has many benefits It has a longer lifetime than Lead Acid it has a greater density than Lead Acid and it has a high constant discharge rate However it does have its drawbacks It has lower cell voltage 1 1V to 1 3V it is more expensive than Lead Acid it is sensitive to overcharging or undercharging and it requires special charging techniques for full usage In addition at shallow discharge cycles the batteries can develop voltage depression or memory effect A memory effect occurs when the battery repeats a few full cycles of complete discharging and charging restoring the full c
245. ialize the Cip and Cop arrays from the config vars Cipla cCEXpl Cip 2 Cip2 Cip 3 Cip3 Cip 4 Cip4 Cip 5 Cip5 Cip 6 Cip6 Cop 1 Cop1 Cop 2 Cop2 Cop 3 Cop3 Cop 4 Cop4 Cop 5 Cop5 Cop 6 Cop6 vectors to center of inner filaments coordinates first for n 1 to 12 do Fi n 1 Ri cos M PI 6 M PI 24 Fi n 2 Ri sin M PI 6 M PI 24 end vectors to center of outer filaments coordinates first for m 1 to 24 do Fo m 1 Ro cos M PI 12 M PI 16 Fo m 2 Ro sin M PI 12 M PI 16 end cartesian cartesian m 1 m 1 now we want to convert those centers of filaments into a line segment that will place the four or five point sources along the filament R P P 0 inner filaments for n 1 to 12 do if Ci n true then ni xm qe 34 for wi 1 to 6 do if Cip wi true then ci n ci ni 1 P Fi n 1 Cx wi Fi n 2 Ri ci n ci ni 2 P Fi n 2 Cx wi Fil n 1 Ri ni ni 1 end end end end outer filaments for m 1 to 24 do if Co m true then Hoc for wo 1 to 6 do if Cop wo true then 12 ci co m co no 1 P Fo m 1 Cx wo Fo m 2 Ro Fo m 2 Cx wo Fo m 1 Ro no no 1 12 ci co m co no 2 P ll end end end end setting u
246. ical Research June 20 1978 William P Winn C B Moore C R Holmes and L G Byerley III Electric Field Structure in an Active Part of a Small Isolated Thundercloud Journal of Geophysical Research February 20 1981 W P Winn C B Moore C R Holmes History This project originated in the seventies An instrument called a balloon borne electric field meter shown below was attached to a helium filled balloon and released into a thunderstorm Two conductive spheres rotated about a shaft as the device traveled through the clouds As they rotated the electric field component E field perpendicular to this axis produced equal and opposite charges on the two spheres The rotation caused the charges to vary sinusoidally which caused a small current This current was detected by a charge amplifier Inside one of the spheres was a mercury switch that indicated up down orientation Data was also gathered from a Hall Effect sensor which detected the B field These devices have been used by several universities as well as the NSSL to collect and compile data To parachute and radiosconde Bearing Spin 0 08 Hz Monofilament nylon 2m Rhomboid Conductive vane A d A p RT ra Vane battery Bearing q 2m andmotor housing Figure 2 Electric field meter and rotational rates Chapter 2 Past Findings on Electric Field Meter Data For some years our customer Dr Paul Krehbiel and a
247. ics Department to begin work on an array of multiple SRT antennas This project had been planned previously but they needed incentive to initiate production 32 A printer could be made available for screen shots of the control display easy to implement to generate revenue for the VLA Visitor Center from it 5 3 Recommendations For Maintaining the System In providing for some minimal maintenance a technician should grease it at the provided fittings check bolts for tightness and verify alignment of vertical axis chassi A technician should run the SRT program once upon startup to verify operation A technician should Change layouts in November and June when objects to be selected are not visible please see Touchscreen Operator s Manual in appendix 33 Appendix Touchscreen Operators Manual 1 Left click on the start icon 2 Select programs Massworks ID 75 Layout Software to start the layout program 3 To open a layout click on file and open gt gt Program Files MassWorks ID 75 Layouts In the Layouts folder there are three folders that the SRT uses stow Used when the SRT is inactive cass sun amp cygnus Used when all three sources are visible Cass Sun and Cygnus Used from June through October cass amp sun Used when only two sources are visible Cass and Sun Used from November through May Each of these folders has three main layouts stow The main layout is stow lay Open this by choosing
248. iever this prgram will decode a serial string of 10 bits sent with a buffer of 10 1 s with a trailing zero ones digit first It also changes the transmitter number into a temperature or references it to a data value include lt aCore tea gt include lt aDigIO tea gt void main int a 0 x 0 t 0 s 0 m 0 i 0 delta 0 pos int zz 500 display string Reciever is on n display char n sleep 200 while 1 aDigital Config 2 DIG OUTPUT aDigital Config 0 DIG INPUT while m 9 a aDigital_Read_Int 0 a a amp 0x0001 if a amp 0x0001 m m l else m 0 sleep zz if m 9 a aDigital_Read_Int 0 a a amp 0x0001 while a 1 a aDigital_Read_Int 0 a a amp 0x0001 j sleep zz gt gt 1 sleep zz if a 0 start decode x 0 pos l for i 0 1 lt 10 i i 1 a aDigital_Read_Int 0 a a amp 0x00001 if a 1 x x I pos pos pos lt lt l sleep zz if x 1 display_string missle is on fire n display_char n sleep zz else t x 17 t t 10 display_string current temp is display_int_dec t display_char n sleep zz delta t delta display_string temperature flux is display_int_dec delta display_char n sleep zz delta t Appendix VI How to compile and load programs onto the Brainstem and save
249. ifiers and one way connectors made the temperature sensor package not only durable but also highly precise The Schematic for the sensor is given in Figure 1 I T 4u LM34CAZ ful 75k Figure 3 Schematic of the IC temperature sensor and op amp 16 When the temperature sensor circuit was complete Daniel constructed a wiring harness to connect the sensors to the A D board The wiring harness consisted of one way connectors for the sensor PCB in order to avoid confusion between ground and power leads on the sensor package The wires were also clearly labeled so the ground power and sensor output would be connected correctly With the sensor hardware fabricated Daniel was able to attach the sensors to the body of the beam dump He attached the sensors to the aluminum plates with a non conductive epoxy Besides not conducting electricity the epoxy also could withstand high temperatures so the sensors were held firmly to the beam dump during the entire test 17 4 Summary and Results Overall this project was successful We performed four trials with an HEL up at Boeing SVS in Albuquerque New Mexico on April 11 2002 The trial results are shown in table X The front panel of the laser read that it was propagating an eight watt beam Trial Results Trial 1 3 23 Watts Trial 2 2 53 Watts Trial 3 2 53 Watts Trial 4 2 33 Watts Table 2 Trial
250. iginal Hg switch data and a cleaner sinusoid was extracted I then interpolated the filtered data using the interpft command to smooth the 14 sinusoid For every existing point in the sinusoid three new points were inserted Figure5 shows the original square wave the filtered Hg wave and the interpolated Hg Wave Raw Hg in blue Filtered Hg in red Filtered and interpolated Hg in black 0 8 0 6 magnitude 0 2 0 4 0 6 219 219 5 220 eno 221 221 5 222 Figure 5 Original filtered and interpolated Hg data The phase difference between the B field data and the Hg switch data gives us the direction that the instrument is facing This is very important for the final step of determining the direction of the E field Taking the FFT of the B field data revealed two primary frequencies See Figure 6 15 Fast Fourier Transform of B Field T T T T Magnitude w n lll Buieehivesutu man eget histo 0 2000 4000 40000 12000 8000 GOD Frequency in terms of samples Figure 6 FFT of B field The higher frequency was due to the rotation of the spheres about the horizontal axis and the lower frequency was due to the precession about the vertical axis For determining the phase between the B field and the Hg switch the higher frequency of the B field had to be isolated This was easy to implement since the filter for the Hg switch was already created
251. igures Figure Number and Title Page 1 Normal Linux vs RT Linux 9 2 Row Mask 17 3 Column Mask 17 4 Image Data 18 5 Buffer 18 6 Image Processing 18 7 Image Data 19 8 Result Placement 19 9 Edge Equation on Color Image 19 10 Vertical Lines 19 11 Horizontal Lines 19 12 This is our Breakout Board Isn t It Beautiful 21 13 Bump Sensors 25 14 Color Code for Contact 26 15 Quadrature Signal From The Motors 27 16 Digital Filter Stage Machine Diagram 28 17 init module Example Code 32 18 cleanup module Example Code 33 19 my msg struct 34 20 PC 104 Connector 44 21 Timing Diagram for a 16 bit Read amp Write On The PC 104 Bus 46 22 Timing Diagram for an 8 bit Read amp Write On The PC 104 Bus 47 RTLinux On A Small Mobile Robot List of Tables Table Number and Title 1 Useful Image Processing Operations 2 Output Logic for Contact Areas in Figure 14 Bitsoi Lamb Willems Page 14 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems List of Abbreviations A A D AC BIOS BMP C CCW CMOS COM CPU D A FIFO Amps Analog to Digital Alternating Current Basic Input Output System Bitmap degrees Celsius counterclockwise Complementary Metal Oxide Semiconductor Communications Port Central Processing Unit Clockwise Digital to Analog Direct Current Digital Input Output Direct Memory Access First In First Out Input and Output Integrated Drive Electronics Infrared Data Association
252. imulation Matlab o by Denis Oesch clear all setting all parameters Ri 63 5 radius of inner track of filaments centimeters Ro 127 radius of outer track of filaments centimeters M 0 065 mass per clip grams rho 2 7 density of the material grams per cubic centiemter H 70 51675 height of filament plane above the mirror s center centimeters Sm 25 radius of the grid to cover the mirror integer inner clips Cu SrA number of clips per inside filament 0 6 inner clip positions 1 if a clip is placed in that position 0 if not Cip 0 12 1 1 T7 0 Souter clips Co m5 number of clips per outside filament 0 6 Souter clip positions 1 if a clip is placed in that position 0 if not Cop 1 1 1 1 1 0 N 12 ci 24 co total number of clips used setting up filament tracks positions of the clips on each filament for later shifting Cx 2 0 1 2 0 4 0 4 1 2 2 0 setting up the filaments positions F zeros 36 2 Sinner vectors to center of filaments in cartesian coordinates for i 1 1 12 F i 1 Ri cos pi 6 i 1 pi 24 24 F i 2 Ri sin pi 6 i 1 pi 24 end vectors to center of outer filaments in cartesian coordinates for j 1 1 24 F j 12 1 Ro cos pi 12 j 1 pi 16 F j412 2 Ro sin pi 12 j 1 4pi 16 end now we want to convert those centers of filaments into a line segment that will place the point sources along
253. in Table 1 below 12 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems How it can be applied to f Operation What it could be used for this project TO Used to brighten an image If the image from camera is Mutiplication too dark ez Used to darken an image If the image from camera Division has too much light Used to determine whether Multi frame line tracking Subtraction there is a change in two image with some additional manipulation Addition Image morphing or adding Not applicable at this time noise Bit Operations Crop and image and get the These operations were not AND part you are needing incorporated in the final Get rid of the additional image ee i OR attached after AND and getting 8lgorithm It could be o the cropped image alone later use in Advanced Image Processing Change the image to a NOT 3 5 negative image Table 1 Useful Image Processing Operations The Image Processing is highly dependent upon the convolution process The convolution process functions with the use of 3x3 matrices These matrices vary significantly and depend primarily on the application needs In the Image Processing the matrix used was the Sobel operator There are many different operators i e Prewitt operator Kirsch Robinson Laplacian operators Frei Chen etc used in image processing and the more complicated it gets to implement in C programming The easiest
254. ing for dryness or wetness The HC12 microcontroller is the brain that sends the water All components utilize as little power as possible In order to gain conclusive measures one must realize 19 the problems There are three main problems objectives that I had to solve First I came to the realization that dirt can easily become dry which in turn will easily and quickly kill the plant So a moisture sensor had to be developed in order to alleviate this problem Realizing that the majority of the commercial sensors are inapplicable to our situation I developed my own circuit after looking at an immense number of circuits a saturated transistor switch with a current source This utilizes a simple copper electrode which is submerged into the dirt This electrode tries to realize if the dirt has no ion flow or current i e no water Upon a few experiments I set it to a certain threshold and when it approaches that voltage the electrode sends a signal and triggers the HC12 to execute one of the desired programs Once executed water flows through the tube which is embedded in the dirt and very close to the root as possible Second since I mentioned one of the desired programs I came to the realization that there are many species of plants just like animals Since watering is our main objective here I classified each plant according to their water intake level The three main ones are lots of water medium water little water he
255. k ok Slideshow timing can be modified The time between slides can be changed by opening up the corresponding layouts that have an automatic button There are three layouts for each slideshow with an automatic button that will need to be changed if any one is changed Program Files MassWorks ID 75 Layouts stow stowsrt science lay Program Files MassWorks ID 75 Layouts stow stowsrt works lay Program Files MassWorks ID 75 Layouts cass amp sun stowsrt science lay Program Files MassWorks ID 75 Layouts cass amp sun stowsrt works lay Program Files MassWorks ID 75 Layouts cass sun amp cygnus stowsrt science lay Program Files MassWorks ID 75 Layouts cass sun amp cygnus stowsrt works lay To do this count which slide needs to be modified Right click on automatic button and select properties The delay for each slideshow change is before the corresponding space bar command Double click on the chosen delay to change the timing example 22 seconds 22000 Change this same line in each of the other two automatic button layouts for the slideshow Close the program so the window will not hamper the operation Do a test run by selecting the VLA weather button on the layout and then press the back to main button once the weather station is up For more information go to the website at http www massworks com This website will contain help on using the touchscreen 35 Compiling and Running a Java Program Type set CLASSPATH on the dos prompt
256. k this is a flaw with the touchscreen software 3 4 Group Efforts In addition to weekly meetings in which we met with our clients and faculty advisor for updates there were some tasks that we performed as a group The first task we accomplished as a group was finding a place to mount the SRT antenna for testing and development We went through some brainstorming of ideas with our project advisor and supervisor on topics such as permissions for NRAO to have the antenna on Workman or another place on Tech property liabilities other design planning other politics and even other possible sites like Etscorn Observatory and Gerald s house We needed to find out if we would need administrative permission Brown Hall or if we would be infringing on any positions of Union workers from the 25 P Plant or other government workers and whether we would need work orders to complete our installation Because there are stanchions on the roof which are designed to securely mount rooftop experiments such as weather sensors and antenna hardware the idea of roof mounting the SRT on top of the Workman Building was appealing Other reasons it was an appealing location was its accessibility and ability to be secured Another factor that made Workman attractive was that our faculty advisor Dr Shanechi graciously agreed to let us set up the development computer in his lab on the second floor This saved us time as we did not have to run the cables all
257. l also be used as the basis of an embedded systems platform 22 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Behavior Control Kyle Lamb The behavior control algorithms that were utilized were designed to use the RT threads developed by Bill First off the sensors were characterized and a math function was developed that would give back a distance in centimeters based on the readings from the analog to digital conversions The calculations work very well for distances between 3cm and 20cm Since was planning to have the robot follow the wall from about 9cm away these values worked very well The wall following function first uses the distance equation to determine distance errors which are used by a proportional controller that will speed up or slow down the motors based on the distance The basic layout of the proportional controller for left wall following is like this Left PWM Speed Desired Distance Distance Desired K Right PWM Speed Desired Distance Distance Desired K The Speed Desired is the desired speed in terms of the PWM take the distance which is an error measurement and multiply it times a K to get the desired change in speed to the motor Notice that the left side subtracts the error while the right side adds This is so that when you are getting close to a wall and your distance is larger than your distance desired then the PWM will be a larger number and the left motor will speed up
258. l specifications As seen in Table 3 1 the Cotronics Liquid Silica would not be feasible because of the thickness required Fastblock 301 is a more viable option because of its lower thermal conductivity Thermal and flame tests were conducted on the materials available for testing These were conducted to test for failures such as melting cracking or ignition and the resultsof flame tests are found in Table 3 2 Cerablanket Sauereisen adhesive and Cotronics fused silica are ceramic materials As expected there was no noted change in the materials when subjected to heat or flame Pictures of these tests can be seen in Appendix VII Material Ignition time Notes Cerablanket N A Became less pliable following test Sauereisen N A Thermal shocking on cooling caused fracture Adhesive brittle after heating Fastblock 301 25 sec Cracked ignited material expanded Cotronics N A No noted changes in material Silica Table 3 2 Observations from flame testing While Fastblock will not hold its shape during exposure to flame and high temperature its low thermal conductivity still made it the most viable material for design Sauerisen was also chosen to seal Fastblock 301 parts together It was used despite its thermal shocking because testing showed that the adhesive did not thermally shock and rupture a junction during a thermal test as seen in Appendix V 16 Temperature F Figure 3 1 Setup for flame test The internal temperature was
259. lled properly From this we have every reason to believe that it will work with our developed software We plan to meet with Phil Dooley an engineer at the VLA May 9 after this paper is due The purpose of this meeting will be to pass on responsibility for the project and give Mr Dooley the knowledge he will need to set up the system when the last of the needed parts are assembled and put in place 29 4 2 Functionality Using a touchscreen a person can select to operate the SRT to track celestial objects It will take approximately five minutes to initialize the system to start tracking During this time an educational slide show will run explaining what the viewer is about to watch There are also tutorials on The Science of Radio Astronomy and How the SRT Works Additional features include NRAO Web pages and The VLA Weather Station a LabVIEW program used by VLA astronomers 30 5 Summary and Recommendations This section describes effect of our work on this project In addition we have outlined some possibilities for further improvement of the SRT and some recommendations for maintaining it 5 1 Summary of Current Work Our work has enable NRAO to provide a sophisticated instrument that the general public can access and use to learn about radio astronomy The SRT kit designed by MIT is suitable for astronomy students or scientists We have adapted it to be suitable for people who do not necessarily have any technical knowledge
260. lure over extended periods of use and needing a custom kernel and a complicated setup for the exported root file system Local booting and root with NFS home In the local boot and root with NFS home configuration each node has its own hard drive containing a complete Linux system The node boots and loads programs from this drive The server node exports the home directory and several cluster wide configuration files which each of the client nodes then mount over NFS The primary advantage of this method is the simplicity of setting up the cluster as a nearly default Linux install can be used In a configuration with varied surplus hardware this method is also simpler than construction of a tagged network image for each computer The disadvantages are the difficulties in upgrades and the potential for hard drive failures in nodes of the cluster Decisions made for this cluster After considering the options the decision was made to use the local boot and local file system with an NFS mounted home Since this project aims at simplifying the use of surplus computers in a cluster the network boot and local boot with NFS root methods both have complications Many surplus computers are already equipped with networking hardware but most of this network hardware does not currently support boot rom chips and booting from the network Even if it does support a boot rom locating or making a functional boot rom will likely be beyond the skill le
261. map I wrote an explanation which is to the right of the image A final touch was to replace the crosshairs SRT uses to denote the active source with a small transparent background GIF image When the sun is being tracked an image of the sun appears on the skymap For other sources an image of a star marks the source in the sky I also wrote an explanation of the spectrum and placed information on parameters such as source coordinates telescope coordinates receiver frequency and time on the display In addition there are two pictures of the active source in the upper left corner of the display One is produced using SRT s mapping function and the other is produced by the VLA The map function on SRT is very impressive and we would have liked to have made it available to the public but it takes a very long time so the image shown on the display is from a previously produced file The main purpose of these two images is to show visitors the quality of the work taking place at the VLA Figure 3 2 2 is a screenshot of the new display panel as it would look tracking the sun It shows all of the elements mentioned above 16 iBi x The graph to the leftis a spectral plot of the signal produced by the Sun It plots frequency on the horizontal axis versus power on the vertical axis The spike in the middle is due to Source Sun hydrogen emissions at 1420 Source Coordinates Azimuth 96 8 degrees megahertz Elevation 29 4 degrees Anten
262. mb Willems in trying to be fair to all tasks the kernel can prevent critical events from occurring Figure 1 also shows a Linux kernel modified to support hard realtime An additional layer of abstraction termed a virtual machine in the literature has been added between the standard Linux kernel and the computer hardware As far as the standard Linux kernel is concerned this new layer appears to be actual hardware More importantly this new layer introduces its own fixed priority scheduler This scheduler assigns the lowest priority to the standard Linux kernel which then runs as an independent task Then it allows the user to both introduce and set priorities for any number of realtime tasks User Processes User Processes RT Linux Processes if Real Time Kernel Linux Kernel Hardware Normal Linux Hardware Figure 1 Normal Linux Vs RT Linux Real Time Linux The abstraction layer introduced by RTLinux works by intercepting all hardware interrupts Hardware interrupts not related to realtime activities are held and then passed to the Linux kernel as software interrupts when the RTLinux kernel is idle and the standard Linux kernel runs Otherwise the appropriate realtime ISR is run The RTLinux executive is itself nonpreemptible Unpredictable delays within the RTLinux executive are eliminated by its small RTLinux On A Small Mobile Robot Bitsoi Lamb Willems size and limited operati
263. more eese 4 2 Tasks Accomplished by David Bytrd eese 4 3 Tasks Accomplished by Ivan Olguin esses Chapter 5 Conclusion ast Project HIG SINS aj oor eov e pvp ta x CEPR iet pp del 5 2 B dget SUMIMALY ooieoe eoe a Su et hebes RU EIS SPERM RR tance RUPES S pP SuU MEER ER 5 3 Final Sensor Functionality vasco pete OE NE PRETEREA MUS 5 4 Final Project Status iiie eorr ry ere ou e eoa EAUX XY CX X aao a ea V veers Appendix Chapter 1 This chapter gives a background and overview to the Wireless Vehicle Signaling System 1 1 Overview of Project This document presents an outline to our approach reviews the requirements for the Wireless Vehicle Signaling System and states the design processes taken to develop a successful end product It shows all the steps taken to develop an innovative remote signaling system for automobiles 1 2 Introduction A major disadvantage to conventional trailer hookups is that they rely on hardwiring the tow vehicle wiring to trailer wiring The Wireless Signaling System will eliminate the need for consumers to struggle with faulty wiring and bad grounds common on trailers today The Wireless Signaling System Team is responsible for designing a universal system to bypass the normal trailer wire assemblies on current trailers and replace them with an innovative RF wireless control and LED array system The system is portable and can be installed wi
264. mphasis in Digital Signal Processing A Thesis Presented to the Faculty of the Electrical Engineering Department of New Mexico Tech by G Martin C Reiten S Schuyler In Partial Fulfillment of the requirements for the course EE 482 Senior Design Project II May 1 2002 2002 Ginny Martin Chris Reiten Seth Schuyler Abstract The storm activity in clouds strips charges from air molecules and eventually generates a powerful electric field This field is detected by an instrument called an electric field meter EFM which ascends into the storm and telemeters the data to ground where it is recorded The scope of the senior design project is to extract the direction and magnitude of the electric field vector from the existing data Over the past 9 months we have worked to achieve just this Different types of filtering and transforms were used to separate out relevant components We have written an interactive Matlab program that analyzes flight data allowing the user to select time ranges and phase angles The output of this program provides 9 comprehensive plots that show the user the raw data the horizontal and vertical components of both the electric and magnetic field relative to the instrument and the strength of the magnetic field as the instrument ascends Acknowledgements Ken Eack NMT Physics Department Set up receiver and decoder for EFM experiments Tim Hamlin NMT Physics Department Maintained computers provid
265. n of the software as painless as possible and to make sure our customer would be able to use and modify it it is implemented in Linux This provided several advantages It provides a stable multitasking environment capable of running long computational tasks without interruption Access to the PC s parallel port is very easy in Linux Most importantly the operating system and a full set of development tools are available freely ensuring that our customer will be able to use and modify the testing system without having to purchase any additional tools or worry about licensing restrictions The other primary consideration in designing the testing software was speed In order to effectively test the device it is necessary to run a vast number of test cases In order to run enough test cases in a reasonable amount of time the speed at which tests can be run is critical The maximum speed at which data can be written to the parallel port is fixed and the PC interface in the Altera chip can easily receive at this speed The time necessary to compute a result in the Processing Element is also fixed from the point of view of the PC software This means that the time to run a test case on the Processing Element is a fixed constant The remaining part of each test case is computing the expected result In order to 19 test the device it is necessary for the PC to independently compute the correct result so it can determine whether the response from the
266. nDdIX 5 5 nro etat susto toot t tiep bie Sr fuh cant fucus fuf 34 Touchscreen User s manual essere enne eene nennen 34 Compiling and Running a Java Program see ea eeseeeesseeeeses 36 Kit SDeCS ocio otia UMP a a Had toad ric ee DCN EL EIU 37 Gant Chart and explanations sse ener enne 39 Chent Enialls nee deu bb RT E Es 43 R siies 5 I ot ae een oe ethene Boge E nile tede d etiaai een dia Sieb a 46 iii List of Figures Figure 1 3 1 Small Radio Telescope on Top of Workman Center 3 Figure 3 1 1 Concept Design for Mounting SRT ssseueesesseeeess 7 Figure 3 1 2 Tutorial Slide Shows ccc cece ence eee e ence ene ence eae tenet nena 9 Figure 3 1 3 Preliminary Installion ccc cece cece ence cence eee eee neeen ease 10 Figure 3 1 4 Details of Installation 0 0 ccc cece cece ence cece HII 10 Figure 3 1 5 Visitor Center Kiosk Concept sss 11 Table 3 2 1 SRT Classes and their Functions ccc eceece ence ee eeeeeeee es 13 Figure 3 2 2 Original SRT Control Panel 00 00 c cece ence ee ence ceea eens 14 Figure 3 2 3 Modified SRT Display Panel 0 ccc ccc ecc ence ee eee eee eens 17 Figure 3 3 1 The Main Layouts ig rte t p ke e eran eH Eee x VE NS coed YR E gute 21 Figure 3 3 2 The Exit Layout o
267. na Coordinates Azimuth 91 5 degrees Elevation 23 9 degrees Receiver Frequency 1420 96 Integration Time 2 3 Minutes max min 2 5K slope 0 3K The two images shown above are images of the Sun The one on the left was produced by the VLA and the one on the right was made using SRT Both images were made at 1420MHz which is the spectral line produced by Hydrogen As you can see from the image quality the VLA s complexity pays off The image to the left is called a skymap It helps astronomers visualize the sky in terms of antenna coordinates The antenna uses coordinates called azimuth and elevation to pinpoint objects in the sky Here s how it works The numbers along the bottom of the skymap are azimuth coordinates Imagine the skymap wrapping around you in a huge circle 180 degrees corresponds to South 90 degrees is due East 270 degrees is West and 0 and 360 meet at North The numbers on the side of of the skymap are elevation coordinates The horizon is at 0 degrees elevation and the Bonema oa m 180 JU 220 240 zenith is at 90 degrees The dotted lines ioien ac Azmuth deg eS TIE ST represent the antenna s mechanical limits dffstat te GB readysetao Bil oldstartup java srt 1 101 jslideshows gt WV dg sssam Figure 3 2 2 Modified SRT Display Panel My other primary modification to the software involved controlling it Without a mouse and keyboard SRT would be useless so the
268. nals depicted above by Ch A and Ch Bcw As the motor spins counterclockwise we get the signals depicted above by Ch A and Ch Bccw As indicated the direction in which the motor is spinning is easily determined by looking at the state of Ch B on the rising edge of Ch A We decided not to take full advantage of the quadrature encoding for two reasons 1 The Prometheus is fast enough to read the position 27 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems accumulators several times before they overflow and 2 our Altera design barely fit on the chip as it is Thus the Position Accumulators are simple up down 16bit counters These counters assume that the signal from the motors is a clean square wave As this is a physical system the quadrature signal is prone to noise To combat this interference we designed a digital filter with a 6 state state machine depicted below in Figure 16 This filter was applied to both channels of both motors 0 Figure 16 Digital Filter State Machine diagram Due to the layout of the motors on the robot chassis forward rotation relative to the robot is clockwise on one motor but counterclockwise on the other This could be solved in software by negating the value read from one of the accumulators but we decided to solve this in hardware by simply inverting Ch B of one motor as it comes out of the digital filter Altera Code for PWM Generation Kyle Lamb In order to control the motor s
269. nce I utilized three different subprograms Fortunately for us if we cannot decide what kind of plant species it is or some other circumstance arises the golden rule of botany is that if the dirt is dry water it It may not grow to its maximum capabilities but as long as it is watered it is still alive As long it is not watered frequently and excessively it will be okay Frequent watering will cause root rot among other extremities Third just like in real life applications I came to the realization that all components use power In addition just like a typical consumer they want to utilize as little power as possible Hence Randy and I specifically distributed the power to make sure the overall planter is sufficiently satisfied with power The water pump 20 utilizes 3W the moisture sensor utilizes 1W the HC12 and circuitry utilizes 01W Once the calculations were realized there is very little power being dissipated 3 2 Rocky Ginanni My contribution to this project lies in the area of energy storage microcontroller external interrupt device and programming At the beginning of the fall semester we were approached by our client to construct a self watering planter that employs many features These features include storing energy that is derived from a means other than the power from a wall outlet and then some way of controlling the watering of a plant based on some form of sensor input 3 2 1 Energy Storage So my criter
270. ng systems networking and system design The goal of this project is to simplify this process so that the existing techniques can more easily be applied to future Beowulf cluster work at New Mexico Tech and elsewhere This project targets using surplus computers that are readily available to university environments simplifies the construction of a Beowulf cluster for researchers in other disciplines such as physics and biology This completed project delivers a Linux based Beowulf cluster consisting of eight functional nodes At the time of this document several nodes were not functioning properly and have been excluded from the final cluster This issue will be addresses as well as possible with the next week This cluster runs the ZPL parallel computing software in conjunction with the user s choice of PVM or MPI parallel libraries Software implementing a parallel solution to a real world computation problem of interst to the client is also provided A final collection of documentation is included The documentation describes the results of out investigation into other Beowulf cluster implementations and why we made certain design decisions on this Beowulf cluster The documentation includes setup and usage instructions scalability analysis of the hardware and number of nodes and supporting performance benchmarks results for the cluster 1 3 Definitions MPI Message Passing Interface a specification for implementing message passing pa
271. ngs mips irix seq pvm valid zpl settings powerpc aix seq valid zpl settings sparc solaris seq mpi mpich pvm valid zpl settings x86 freebsd seq valid zpl settings x86 linux seq mpi lam mpi mpich pvm Parallel System Settings valid zpl settings enterprise seq mpi valid zpl settings origin seq mpi pvm Platform Settings valid zpl settings sp2 seq mpi valid zpl settings t3e seq mpi pvm shmem Convert 32 bit integer form of an IPv4 address to dotted quad form def int to ip4 ip4 addr num octet ip4 addr num amp Oxff000000 gt gt 24 ip4 addr num amp 0x00ff0000 gt gt 16 ip4 addr num amp 0x0000ff00 gt gt 8 ip4 addr num amp 0x000000ff f ip4 addr str s s s s octet 40 return ip4 addr str end of int to ip4 Convert dotted quad form of an IPv4 address to 32 bit integer form def ip4 to int ip4 addr str octet re split ip4 addr str ip4 addr num long octet 0 lt lt 24 long octet 1 lt lt 16 long octet 2 lt lt 8 long octet 3 return ip4 addr num end of ip4 to int def pvm setup i Common PVM install locations extend this list as needed common pvm root dirs usr share pvm3 usr lib pvm3 opt pvm3 usr local pvm3 usr local lib pvm3 common pvm root dirs valid pvm rsh pro
272. nohoe thought of the self watering planter as a solution to proper plant care while away from the plants for weeks at a time Chapter 2 Background Information 2 1 Technical Background 2 1 1 Moisture Sensor In the Fall Semester we researched common commercial sensors but realized the majority of them are not applicable to our situation More specifically most of them are either too big too expensive or cannot be submerged into the dirt We went on to use a saturated transistor with a current source That is the sensor itself as it is connected to a copper probe The saturated transistor allows more current larger to flow into the current source as it supplies little current since we are allowing little power as possible The current is what gives it its power The copper probe is like the arm of the sensing device as it senses for ion flow in the dirt i e water 2 1 2 Light Sensor Real Time Clock The light sensor idea began at the beginning of the spring semester After choosing the components during the fall semester we decided that we needed a way for the planter to wake up Since the light sensor is an IR phototransistor this requires an output circuit that takes the analog voltage from the emitter of the phototransistor and then sends that to the positive input of the comparator Then based on a 100mV change that is determined from a selection of resistors the output of the comparator will swing from OV to 5V for darkness and s
273. ns While radiation in a fire can be appreciable and is treated by the NIST article on heat flux in fire Womeldorf International Conference 2001 this also was neglected for the purposes of this project By considering conduction to be the method of heat transfer Fourier s law Q k A dT dx Geankoplis 1993 can be applied where Q is the heat flow A is the cross sectional area dT dx the change in temperature change in thickness and k is the thermal conductivity This general equation was applied to non steady state conditions as it exists for the conditions of this project Calculations using heat transportation equations are shown in Appendix II These were used to determine size requirements for insulation casing Another class of material considered was silica ceramics Silica has a low thermal conductivity Very high purity silica is used to make the space shuttle tiles www spaceshuttletile com faq html 2001 While shuttle tiles are far too expensive for the price range requirements of this project silica materials were considered for their good thermal insulation see Table 2 1 Additionally zirconia ceramics have been considered as they too have very good thermal insulation Foam and blanket materials made from very porous or fibrous ceramic were also considered These take advantage not only of the good thermal insulation of the ceramics they were composed of but of air which has an even better insulation and muc
274. nsors The result was the available thermocouples thermistors and IC temperature sensors Thermocouples were the first considered elements to be used The high sensitivity and fast output reading were the main characteristics that led to the choice of these sensors The weak aspect of the thermocouples was that they were expensive and due to budget constraints Daniel had to consider another device So 15 following a suggestion from our costumer Daniel Eastman Daniel Rodriguez decided to use an IC temperature sensor The final sensor chosen was the LM34CAZ The sensitivity was lower then the thermocouples and the response time was slower but the cost was much cheaper After selecting and purchasing the temperature sensors Daniel could begin testing the temperature sensors The initial testing was done on the breadboard A heated knife placed against the sensor was used as a stimulus to the sensors so the response from the sensors could be measured Later testing showed that a gain would be required so that reliable data could be sent to the A D board Using the information provided by Ernest Jim on the degree change that would be seen and the limitations of the A D board the sensor output was amplified by six Once the temperature sensor circuit was designed Daniel etched a printed circuit board PCB for the temperature sensors He used Protel 99 to create a simple easy to use design The use of one percent resistors LM 741 operational ampl
275. nuclear Third we needed to realize the type of moisture sensor to employ We used a basic saturated transistor with a current source This sensor needs to utilize as little power as possible to accommodate the power distribution 1 3 Literature Review The Art of Electronics book by Horowitz and Hill covers a wide range of subjects in electronics This includes transistors FETs field effect transistors OP AMPS operational amplifiers active filters and oscillators voltage regulators low noise circuits digital electronics microprocessors high frequency techniques low power designs and signal processing It helped us develop the majority of our components for the self watering planter The plants usda gov website provided vital information in plant water intake level among other characteristics Siliconsolar com website provided answers for solar cell needs among other vital parts Human assistance included Dr Greg Donohoe who was our customer and offered assistance in the moisture sensor development Dr Bill Rison who greatly assisted us in the HC12 microcontroller programming and Dr Ron Thomas who greatly assisted us in solving most of the circuitry problems 1 4 History The history behind the self watering planter revolves mainly around the inconvenience of constantly taking good care of the plant s Our customer Dr Greg Donohoe is constantly away from his office on trips hence the plants die easily under improper care Do
276. number of colors is less than the 24 bit image obtained from the web camera Reducing the number of colors in the image could make determining the objects in the field of 40 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems view easier to recognize The walls could be one color and the background or door could be another All in all the future of image processing is endless The procedure on how to get there is the hard part which the same attribute faced in this project 41 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Summary amp Conclusion The final product of our labors is a robot that is using the Prometheus SBC along with RT Linux to read and respond to sensor data in real time We loaded RT Linux onto a small 128MB Flash disk and were able to boot from it as well as interface to our breakout board and the Prometheus I O The robot is also able to network and use remote login capabilities which allow the user to connect with the robot over great distances A wide variety of sensors have been connected to the robot all of the existing sensors as well as bump sensors and the ultrasonic distance sensor The Breakout Board we built is fully functional and is highly adaptable so that it can be changed to meet customer desires Along with the Breakout Board hardware we have supplied a driver for the breakout board that can be easily modified to add more functionality We have also accomplished some Image Processing
277. nvaluable for this along with the many other aspects of working with this program that I did not understand at the start of the project Also I did my best to add comments to the code hopefully clearing it up a bit for future work Commenting and coding in general became much easier one night in February when Dr Stephen Bruder walked into the lab and asked why I was using Wordpad He suggested downloading a Java Integrated Development Environment 18 IDE I downloaded and installed the RealJ IDE that night and life immediately became much easier I did not use many of its features but the color coding in RealJ helped immensely My version of the code can be found on the accompanying CD In summary I would describe my work as reshaping the visible parts of the program for NRAO rather than rebuilding it in any way The things I would certainly not claim in the program are all of the major program functions such as calculating source coordinates calculating dish coordinates low level dish control receiver control data processing and time calculations I also did not do any work on the serial interface the program uses While the program consumed the bulk of my time on this project I had other tasks One of these was figuring out how to update the computer clock automatically and accurately as well as determining the dish s physical location This is necessary because of the manner in which the software calculates source coordinat
278. o and is the de facto standard PVM has a user shell that is used to configure the cluster interactively MPI is the newer of the libraries and is backed by a formal industry standard MPI is configured by a default configuration file but the user can specify a different configuration file to use on the command line There are several implementations of the MPI standard available under the Linux operating system The two most common are LAM Local Area Multicomputer and MPICH MPI CHameleon The MPI and PVM libraries both make use of the rsh program to connect to the nodes in the cluster and begin the execution of the parallel program on that node The parallel libraries then manage the network communication between the processes on each node Figures 6 and 7 illustrate the results of the PVM and MPI benchmarks performed on the Beowulf cluster constructed Figure 6 illustrates the PVM and MPI message passing overhead Since the S data class is a small data set the predominant time factor is the communication between nodes From these results MPI has a lower communication overhead on the test cluster Figure 7 involve more computations relative to network 11 communication and thus shows little difference between the performance of PVM and MPI Though in the majority of the test results MPI held a slight advantage Figures 6 also shows increasing execution time as the number of processors involved increases for the same amount of c
279. o Appendix A I created a flowchart that shows the structure of the program that controls the function of the planter Fig 8 3 3 Randy Sena My contribution to the project includes research and development of the planter design pump solar panel charge circuit and adjustable program input switch For the majority of the first semester research was done on all of these areas in an effort to investigate the methods we used along with alternative methods Through research I was 27 able to narrow down my decision by comparison of cost and effectiveness in order to make my final choice on specific components 3 3 1 Planter Design The planter design was first thought out to be a modification of an existing planter The intention was to modify a planter in effort to meet our needs Shortly after trying to accommodate various planters it was finally decided to build our own planter The planter is constructed entirely of clear Plexiglas Fig 7 The use of Plexiglas offers a unique style which allows visibility of all circuitry With the clear Plexiglas you can also visibly monitor the water level inside the water storage area The lower two thirds of the planter offers a 285 cubic inch water holding area which has a refill spout and a drain plug Also on the lower two thirds is a containment area for the pump The upper third of the planter houses the soil which holds the plant Between the upper third and the lower two thirds of the plant
280. oard Bill Willems If given the chance to restart this project knowing what know now there are a few things would do differently For one would have gone with buying a third party timer pulse accumulator board instead of designing and building our own This aspect of our project occupied all of my time during the first semester and became a project within a project If we used a third party board we would only have to convert or rewrite the manufacturer supplied driver to a Realtime driver This would have allowed more time to be devoted to other aspects of the project There are also other smaller changes would have made Had known that our client was waiting for the go ahead from us to order our Prometheus would have ordered it sooner would have given the task of interfacing to the Prometheus I O to Kyle because he was waiting for me to finish it before he could develop his Behavior Control algorithms wouldn t have nuked the WindowsNT boot partition on our development system would also have setup our weekly team meetings outside of our weekly meetings with our advisor during the first semester Despite these grievances am satisfied with the end result of this project Had we gained the time lost to development of our Breakout Board this project could have achieved so much more However our client is very happy that we did design and build our own expansion board and he has big plans for it guess
281. oblems at the beginning trying to get a complicated encoder working but ended up using a simpler Holtek encoder decoder pair to get the Glolab transmitter receiver to receive signals The major task completed for this project was the board layout in Protel We decided to implement the transmitter board and receiver board onto one to save on cost of manufacturing the board developed the receiver board in Protel while Ivan did the transmitter board Once we laid out the circuits in Protel combined the two into one board and placed all the parts in the PCB part of Protel had to make certain that no traces from the transmitter side were creeping over to the receiver side After the board was laid out the way we wanted it Dave sent it off to the manufacturer was in charge of soldering all the parts onto the receiver board and getting it to work once it came in then rigged up a plastic box so that it would hold the circuit switches wires and antenna Once all of this was done in correlation with the other members work we were able to test the entire project on a vehicle Along with the work on the project also played a big part in the writing of the papers was given every third progress report to write and helped write every other paper that was due 4 2 Tasks Accomplished by David Byrd One of my main initial tasks consisted of investigating several receiver transmitter pairs along with encoding decoding options In ord
282. oisture it needs It will inform the HC12 whether if it s dry or wet by the predetermined threshold This device will send a signal to the HC12 and execute the desired program 3 1 3 Power distribution The self watering planter will require power to various subsystems that include moisture sensor the PLD and pump The theme of the project is power scavenging hence to consume power components were chosen based on their performance and power draw The water pump pulls 6V 500mA the moisture sensor utilizes 5V 20mA and the PLD circuitry utilizes 5V 2mA In order to determine the total power draw calculations were computed and the planter draws 533A Hr day With this in mind I was then able to predict what type of solar panel and battery we would need For the other electronic components we installed a MAX667 LDO 5V regulator This low dropout voltage regulator only uses 20uA of current This 5V regulator provides power to the moisture sensor HC12 and adjustable switches Mosfet IRF9510 switches were placed on the circuit board to provide power to the moisture sensor circuitry and the pump when needed These switches are activated by a low signal from HC12 ports A low dropout voltage regulator was used to regulate a steady 5V source the moisture sensor HC12 and switches used 5V Fig 1 My contributions were an overall success with all milestones achieved Appendices B C The moisture sensor is efficient as it acts like a hand sens
283. olved 2 1 Work Overview The project s first technical research was to decide on a transmitter and receiver pair We decided to choose two different pairs alpha and beta Each pair was tested and evaluated for simplicity transmission distance and cost The alpha transmitter and receiver pair was not pre wired to any other components This allowed us to have full flexibility on both the transmitter and receiver This system allowed the group to add all external logic and a light driver circuit to meet exact specifications This system will work with a variety of encoders where the beta system will only work with a specific encoder because of the prefabricated receiver The beta transmitter and receiver pair came in a package that only needed an encoder to make a completely working system The receiver was enclosed in a box that included the decoder and relays to power an output maximum of 5 amps The problem that we ran into with this system was that it would not allow us to have the flexibility to add our own specific requirements The beta pair would also use up excess power to run the relay switches Another important decision was to find an encoder and decoder which would work with our transmitter and receiver and also allow us to meet the specific requirements of our customer After making our decision we then separated the project into two parts the transmitter and receiver boards 2 2 Transmitter Circuit The transmitter board
284. ommand line argument The software takes note of this and goes about its routine of moving the dish to the stow position I had to add a mechanism to wait while the dish moves due to the fact that the software moves much faster than the dish The original software depends on the user waiting until the dish stows before making commands Adding this waiting mechanism turned out to be a lengthy process but it is in place Once the dish is stowed commands are issued within the program to select the previously noted source and begin tracking There are a few other settings that occur but this is the basic routine While I would by no means claim ownership of the new version of the SRT software there are some new functions that I implemented that did not exist in the old version The routine that waits until the dish stows before issuing commands did not exist It was also fairly difficult to implement due to the lack of information on dish status available within the program This is because the two Basic Stamp microcontrollers in the SRT control module handle the most basic drive tasks Interfacing the program with the touchscreen is another accomplishment I made While very simple this ability was also not present in the original software A third contribution I made to the program is a method for drawing images While it sounds simple I had never used images in a program before and I had to figure out how to do it I found Google s usenet archives i
285. ommon PC with a few minor differences Figure 20 below shows the PC 104 connector and the pins we are interested in IOCHCHK Ground P SD7 RESET4 SD6 5V All of the Altera code to handle the interface SD5 IRAQ SD4 5V SD3 DRO2 to the PC 104 bus can be found in Appendix SD2 12V SD1 ows B SD0 12V i IOCHRDY Key pin cut gt AEN SMEMW SD0 SD15 Data Bus SA19 SMEMR SA18 IOW SA17 IOR SA0 SA9 Address Bus this bus is actually SA16 DACK3 SA15 A16 B16 DRO3 aad BAe 24 bits wide but due to the backwards SA13 DRO1 SA12 Refresh iSi irement of the ISA w nulo bibi compatibility requirement of the ISA bus we SA10 IRO7 SA9 IRQ6 can get away with using only ten SA8 IRQS SAT IRO4 SA6 ROS 4 RESET System Reset SA5 DACK2 SA4 TC l ae A BALE IOR IOW Read Write signal lines when SA2 5V SA1 osc this line goes low the system is reading or SA0 Ground Ground Ground Ground Ground writing to the ISA device MEMCS16 SBHE gt iocsie yes OSC System Clock IRQ10 eg LA22 IRO11 C4 LA 1 IOCS16 Signal line to indicate a 16 bit read IRQ12 LA20 IRQ15 LA19 IRO14 C7 LA18 or write DACKO LA17 DRAG MEMR DACKE HENW BALE Address Latch Enable when this DROS SD8 DACK6 D12 C12 SD9 line goes high the data on the address line DROG SD10 DACKT SD11 DRO7 SD12 has had enough time to settle and is valid 5V SD13 MASTER D17 SD14 Ground SD15 Ground K
286. omponents of the waves such as high frequencies and low frequencies We also want to isolate the fundamental frequencies of the data sets to compare phase shifts Our goal is to detect patterns or correlations among the different data types There is erroneous data that must be removed such as data spikes We will have to find where the actual data begins since the meter records on its way up to the storm as well as in it The physicists involved have analyzed and begun to filter the data and from this work have gained significant understanding of the electric fields in a cloud The vertical component as well as the magnitude of the horizontal component of the E field has been extracted from the data by hand This has been a slow process and thus much of the analysis is still not done Ultimately we hoped to automate the analysis so the knowledge buried in the data would be revealed more readily This would greatly aid the NOAA and the NSSL in their life protecting efforts by increasing their understanding of electric fields in thunderclouds 11 Chapter 3 Steps Toward Development of Data Analysis Automation Seth Schuyler Menuing and User Interface The initial idea was to all do the same work on the data independently None of us were proficient with Matlab and we thought this division of labor would allow all of us the opportunity to gain the skills necessary for the completion of this project As it turned out however each of us con
287. omputations This is due to the large overhead of network message passing in relation to the small data set that calculations are performed on Figure 7 involves 512 times the number of calculations which overwhelms the network overhead and the benefits of the parallel computing come into play The Class S data in Figure 6 is a 32x32x32 element data set and the Class A data in Figure 7 is 256x256x256 element data set From the results in Figures 6 and 7 the following guidelines for this cluster can be made data sets of fewer than approximately 125 000 total elements will in general perform better on a single processor than on the cluster Larger data set will reap the benefits of the parallel execution EI MPI Bl PVM Processors 7 8 Figure 6 Performance comparison of NAS MG benchmark with the Class S data set using the MPI and PVM libraries 12 2500 2000 seconds 1500 EI MPI 1000 Bl PVM 500 Processors Figure 7 Performance comparison of NAS MG benchmark with the Class A data set using the MPI and PVM libraries Figures 8 and 9 illustrate the performance of the MPI and PVM libraries in conjunction with the 100 inch mirror simulation program written in ZPL see Appendix B Since this simulation is a computation intensive task the differences between MPI and PVM were not readily visible However another valuable suggestion is visible in Fi
288. on the output of the sensor As the brake pedal is depressed 1 8 of an inch or more the magnetic field is not strong enough to switch the internal transistor of the hall effect sensor and logic high is seen on the hall effect output To mount this sensor zip ties or Velcro can be used and sensor adjustment consists of changing the distance between the magnet mounted on the brake pedal and the sensor body The Hallogic sensor can sink up to 30mA so this sensor is more than capable of driving the encoder chip Refer to Appendix G for a printed circuit board layout and the circuitry involved Final Testing The sensor is designed to work in harsh environments and there is an internal voltage regulator along with built in histeresis internal to the chip In vehicle testing proved the sensor worked flawlessly with no false signals being seen For testing in the current generation of full size Fords a bracket was made to mount into factor holes underneath the dash to extend the sensor over the brake lever assembly This sensor is shown in Figure 7 Figure 7 Hall effect sensor 3 1 5 Current Sensor Description This is a relatively simple sensor that senses the presence of current in a wire Application Our first approach to using this sensor was to search for one we could wrap around the brake wire that would sense the presence of current while the brake pedal is compressed This would be easy to implement if we could find the c
289. onductivity of Refractories Annual Book of ASTM Standards American Society for Testing and Materials Philadelphia PA Bryant Rodney Erik Johnsson Thomas Ohlemille Carole Womeldorf International Fire Science and Engineering Conference 9th Proceedings September 17 19 2001 Edinburgh Scotland 2001 Estimates of the Uncertainty of Radiative Heat Flux Calculated from Total Heat Flux Measurements Cotronics Corp High Temperature Materials and Adhesives for use to 3000F Volume 01 Number 41 3379 Shore Parkway Brooklyn N Y 11235 CRC Handbook of Thermoelectrics D M Rowe Ed CRC Press New York 1995 Ferrotec USA Inc www ferrotec usa com Geankoplis Christie J Transport Processes and Unit Operations 3rd Edition Prentice Hall New Jersey 1993 Loveall Kevin Manager Business Development and Planning Kirkhill TA Letter to the author 20 November 2001 Omega Temperture Handbook and Encyclopedia Omega Corp 2000 Space Shuttle Tile Company http www spaceshuttletile com faq html Speyer Robert F Thermal Analysis of Materials Marcel Dekker Inc New York 1994 Radial Heat flow Calorimiter methods Thermal Ceramics P O Box 923 Augusta Georgia 30903 0923 Kirkhill TA 28065 Franklin Parkway Valencia CA 91355 3M http www 3m com market industrial ceramics pdfs outerspace_apps_brochure pdf 38 Appendix I Customer Proposal Appendix II Power Supply Calculations The calculations in this appendix illustrate a
290. one end known as a junction that produces a small thermoelectric voltage when the junction is heated Omega Engineering Inc 2000 A thermocouple thermometer measures the electromotive force emf between two different materials and translates it into a temperature 20 Resistance temperature devices or RTD s work because as temperature changes the resistance of the RTD material changes Resistance of an RTD rises relatively linearly with temperature RTD s have a moderately wide temperature range from 270 C to 850 C RTD s were the most stable kind of sensors considered They have specifications to be off by only 0 2 C after 10 000 hours at the maximum temperature The main problem with using an RTD for this project is the tendency of RTD s to self heat This not only leads to inaccuracies but also the internal circuit for this project needs to be kept at a temperature of 150 F Therefore any internal heat due to self heating would be damaging Infrared radiators work by measuring the thermal radiation of an element These were strongly considered because they do not require physical contact and are good at measuring large surfaces The long range goal of this project is to have a disposable product that will only be used in a fire once The price of such sensors that would meet the temperature requirements of this project was excessive Thermistors are a specific type of RTD Thermistors work because a change in temperat
291. ons Realtime tasks have two special attributes they are privileged that is they have direct access to hardware and they do not use virtual memory Realtime tasks are written as special Linux modules that can be dynamically loaded into memory They are not expected to execute Linux system calls The initialization code for a realtime tasks initializes the realtime task structure and informs RTLinux of its deadline period and release time constraints Non periodic tasks are supported via the use of interrupts In contrast with some other approaches to realtime RTLinux leaves the Linux kernel essentially untouched Via a set of relatively simple modifications it manages to convert the existing Linux kernel into a hard realtime environment without hindering future Linux development Barabanov 2001 RT FIFOs Realtime FIFOs are Linux character devices with the major number of 150 RT FIFOs are First In First Out queues that can be read from and written to by Linux processes and RTLinux threads FIFOs are uni directional you can use a pair of FIFOs for bi directional data exchange Prometheus SBC The Prometheus Single Board Computer SBC is an embedded PC 104 CPU that integrates three separate circuits onto a single compact board CPU Ethernet and Analog I O The Prometheus conforms to the PC 104 standard an embedded standard that is based on the ISA and PCI buses and provides a compact rugged mechanical design for embedded system
292. or Sending Data to Computer esee 14 Figure 8 Final Design to Route the Incoming Data eese 15 Figure 9 Altera Board Diagram euin etos ien ted epu o t tbc elie cele ibe 16 Ligure 10 Pur Our or JPII BeS3detusis aan erit r a ona us aq med on 17 Figure 11 Interface I O on 7 Segment LEI Scoot act e ette ieee e oa dest tat anda 17 Figure 12 PC Parallel Port Dia grannies sce iiie eer e bees tenete ttp Meebo 18 CHAPTER 1 INTRODUCTION 1 1 1 2 1 3 BACKGROUND The Reconfigurable Data Path Processor RDPP is a custom processor being developed by NASA s Institute for Advanced Microelectronics It is being developed for high throughput processing of streaming data aboard spacecraft It s primary application will be image and signal processing Donohoe 2002 PROJECT OBJECTIVES The objective of this senior design project was to verify the functionality of the RDPP Verification or functional testing means to execute test inputs on the device under test DUT and verify that the output is functionally correct for each test input Our customer Dr Greg Donohoe specified that the RDPP be verified using an Altera Programmable Logic Device PLD that he supplied Using a PLD for device verification has two advantages 1 it s closer to the final chip layout than a software implementation but doesn t require that the device be fabricated and 2 design revisions can be easily implemented The end prod
293. orrect sensor and then locate the correct wire Final Status We finally found a clamp on type sensor but for no less than a few hundred dollars which was much more than we could allot for one sensor Many vendors did carry a through hole device that would work but with the stipulation of not being able to tamper with vehicle wiring we had to dismiss the idea That would be hard to do without cutting the wire Another problem we were running into was that we would have to instruct future customers of this product on how to find the exact wire That would be hard since the location of the wire is different on all vehicles For these reasons this sensor didn t work out 3 1 6 Touch Pad Description This is a simple momentary contact touch sensor that outputs five volts when touched and stays high until untouched Application The idea behind this sensor was to find a very thin touch sensor that could be placed over the brake pedal and sense the touch of the foot We wanted something very small and discrete that would not change the feel of the brake pedal drastically This sensor in theory would be the best way to go Final Status The problem we had with this was that we could not find a vendor that carries such a sensor to meet our needs We later searched for the touch sensors similar to those used for the touch pad of a microwave We were unsuccessful in finding this sensor 3 1 7 Angle Sensor Description Analog devices ADXL202JQ
294. ost name per line with either n file write hostname n file write or n file write hostname n n file write where n is the number of processors in an SMP The hostname should n file write be the same as the result from the command hostname n for node in node list if cpu per node 1 file write s d n node 1 cpu per node else file write s n node 11 file close end of generate cluster files if sys argv 1 pvm 46 pvm setup if sys argv 1 zpl setup zpl setup if sys argv 1 zpl lib zpl target os environ ZPLTARGET if zpl target in valid zpl settings keys zpl commlayer zpl setup zplcommlayer zpl target zplcommlayer sh file open os path join user parallel setting dir zplcommlayer sh zplcommlayer sh file write export ZPLCOMMLAYER s n zpl commlayer zplcommlayer sh file close else print ZPL environment is not correctly configured n print You must run zpl setup before using ZPL n QW if sys argv 1 gen cluster files generate cluster files Bourne Shell wrapper for the Python configuration script This script is a wrapper around the Python script above It provides shell function that handle calling the Python script with the necessary arguments and then sourcing the resulting files to activate the changes This file should be placed in the parallel directory of the user s directory and then
295. ot demonstrated that the insulation maintains an internal temperature of 150 F while subjected to an intense fire for thirty minutes However components are available that have operating temperatures of up to 85 C or 182 F The insulation tested did demonstrate that it could keep the internal temperature below this temperature when subjected to the extreme temperatures for thirty minutes The insulation prototype tested was of the appropriate size to house all of the current electronic components While the power is not being supplied in a manner consistent with the original specifications the power supply system chosen meets all other specifications in terms of voltage current and shelf life The device is activated when a switch is subjected to heat of 212 F which is an improvement over the original specified activation temperature of 250 F A latching relay ensures that regardless of temperature excursions following the activation of the device the power supply remains constant The electronic components have been fully integrated and perform required functions The power supply was integrated with each component There is a K type thermocouple temperature sensor integrated with an amplifier circuit which the signal 35 can be accurately detected and read by the processor The processor has been shown to encrypt stored data live temperature and temperature flux This has been demonstrated by being hard wired to a second receiving
296. ot share a clock or a method of synchronizing off each other To accommodate for this the receiver synchronized with the transmitter The receiver remains in an idle state until it received a cue from the transmitter The cue was seen as a series of 9 ones instead of a series of 10 ones witch was the standard protocol This was done to ensure that the receiver had to wait for the zero bit that was transmitted at the end of the cue After the receiver had received 9 ones it waited for a zero to be transmitted When the zero was received it waits for one and a half clock cycles and then received the 10 bits from the transmitter The timing of the receiver was shown in Figure 3 7 below After the receiver had received the data stream it performed several Cue Data Count 9 1 s A Receive data Delay 1 Wait for clock cycles Zero Figure 3 7 Timing diagram for the receiver checks on the number If the number is between 0 and 75 it was recognized as the inventory data for the ordnance and the ordnance data is displayed on the console If the data is greater than 75 it was interpreted as a temperature and was multiplied by a scaling factor to convert it to Celsius and it was compared with the previous data value The current temperature and change in temperature was 30 displayed on the console The current temperature is also compared to the cook off temperature of the ordnance if it was within a specified margin a warning message was
297. p the position vectors to describe the mirror as coordinates Since it will be a grid we need only assign points a single row the x and y terms matching setting up an array to hold the position vectors for each point on the mirror grid 1 2 8S41 1 2 S 1 1 G 128 27 256 54 2 S Index2 1 1 2 841 1 2 S 1 2 G 128 27 256 54 2 S Index1 1 1 2 S 1 1 2 S 1 1 D G 2 5130 8 0 5 H 1 2 S 1 1 2 S 1 2 D G 2 5130 8 0 5 H 1 2 841 1 2 S841 3 G R_D D This line if for simplifying output of Matlab compatable results R XY XY 128 27 256 54 2 S Indexl 1 at this point we need to convert these distances over to a thickness contribution from each filament so we will have a new array of plates but this one will consist of N layers of the thickness contributed from each individual point source clip over the 2 S 1 x 35 2 S 1 grid covering the mirror for t 1 to N do R G temp2 1 t 1 3 P 1 2 S41 1 2 S 1 t 1 3 delta gt gt 1 2 S 1 1 2 S 1 1 1 3 G temp2 end R delta temp3 delta 2 1 2 841 1 2 S 1 1 N 1 rsqrd lt lt R delta temp3 You must use 5 0 and 4 0 instead of 5 and 4 or else some integer math is done which is incorrect and produces th
298. peed we needed to have a Pulse Width Modulation signal incorporated into our Breakout Board since the Prometheus did not have any PWM generators The PWM signal was easily incorporated into the Altera code as a counter which would count up to a certain value sent over 28 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems the PC 104 Bus and remain high as long until the value was reached and then send out a digital low Bill and decided to send over a signed 8 bit number that would use the most significant bit to determine the direction desired out of the motor Any number over 64 would be considered to be negative and hence send a signal to the direction on the PWM and reverse the motor direction Another problem that we encountered was setting the frequency of the PWM signal that we wanted From previous experience in Junior Design knew that a frequency of about one kilohertz would work great So in order to get the PWM function to work we needed a clock signal to come in that was about 128 kilohertz because we were counting between 0 and 128 This required that we have a divided down clock based on the PC 104 bus clock signal With the divided down clock it was easy to implement the counter and have the PWM and direction bits connected to an output from the Altera Altera Code for 8 bit PC 104 Interface Bill Willems The Prometheus utilizes the PC 104 interface bus and so does our Breakout Board convenient huh There are sev
299. pleted is to design robots that can collaborate efforts over a network and possibly control the robot over the Internet Along with the work collaborating robots the Breakout Board that we constructed can be used in many applications possibly used as a reconfigurable data processor RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Customer Requirements The requirements set forth by our customer were to develop a package using RT Linux that would run on the Prometheus Single Board Computer to access and respond to multiple sensors We were also required to do some Image Processing to find vertical and horizontal lines We were provided with an assembled Robot Kit that the customer asked us to add some sensors to as well as add a better battery We were also provided with a Prometheus SBC which due to its lack of Timer Pulse Accumulators needed an external board to interface with the motors on the robot As far as the external board went we were told we could either design our own or order a prefabricated board We decided it would be better for us to design our own board but this unfortunately ended up taking most of our time RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Background RedHat Linux Linux is an operating system that was initially created as a hobby by a young student Linus Torvalds at the University of Helsinki in Finland Linus had an interest in Minix a small UNIX system and decided to develop a system
300. ppears they should be one can look back at the raw data and see if odd data anomalies are due to instrument failure or are real data points Figure 15 shows the individual components of the E field vs time with time on the vertical axis and the components on the horizontal axis As time passes the instrument is rising If it is assumed that the instrument is rising at a fairly constant rate it can also be assumed that time and altitude are linearly related Therefore the graph can be interpreted as the E field vs altitude 29 Vertical component in blue and Horizontal component in red 1000 T T T y T T T 950 900 850 800 750 700 650 600 550 un 150 100 50 0 50 100 150 200 Figure 15 E field components relative to altitude Ninth plot of finished product Using this one can see how the fields change with altitude and when lightning discharges occur For implementation questions please refer to the text version of the finished product in the appendix 30 5 0 Future Work The three dimensions have been extracted for a cylindrical coordinate system but it would be preferable to compute them in the Cartesian coordinate system This can be done by taking the phase shift of the low frequency component of the B field with the horizontal component of the E field Once the electric field has been extracted in this form the three dimensional E field can be displayed a function of time Adding
301. proach the conventional method of researching various commercial moisture sensors I realized there were two types capacitive and resistive Capacitive sensors are typically more efficient as they are more tolerant to high humidity levels The most important feature to focus on is one with internal logic These sensors typically require a 5V source to provide input frequency With the internal circuitry capabilities it will convert the frequency to voltage then convert that analog signal to a digital one There is no need for time and power consuming calibration Other traits that were investigated upon were accuracy only 2 error long term stability ability to recover from condensation chemical physical contaminant resistance size cost and low current consumption uA 13 or mA little power used a cover of some sort to protect it from dust dirt oil etc linear voltage output stability with low relative humidity error operation in wide humidity ranges fast response time once fed with power needs to operate quickly However everything came to a complete halt when 2 events occurred First a Honeywell engineer bluntly informed me that I was investigating the wrong type of sensors and none of them would be able to be submerged into any forms of media I realized that was 2 5 months of wasted research Second I spoke with an environmental engineering professor and his descriptions of his homemade moisture sensors match pe
302. pwm L pwm latch q you talkin ta me Robert DeNiro kicks a IF addr_latch 9 4 q OUR_ADDRESS AND AEN GND hey_thats_us VCC END IF Put Get stuff on from the pc 104 bus Q reads I F hey thats_us VCC AND addr_latch 3 0 q CTRL_R_POS THEN enablel6 VCC IF IOR n GND THEN thats a read VCC shitbrick VCC data R pos END IF END IF IF hey thats us VCC AND addr latch 3 0 q CTRL L POS THEN enablel6 VCC IF IOR n GND THEN thats a read VCC shitbrick VCC data L pos END IF THEN 48 RTLinux On A Small Mobile Robot END END IF IF hey thats us VCC AND addr latch 3 AND IOR n GND THEN thats a read VCC data 7 0 R pwm latch 7 0 q data 15 8 H 55 END IF IF hey thats us VCC AND addr latch 3 AND IOR n GND THEN thats a read VCC data 7 0 L pwm latch 7 0 q data 15 8 H 55 END IF IF hey thats us VCC AND addr latch 3 AND IOR n GND THEN thats a read VCC data H 55 END IF writes IF hey thats us VCC AND addr latch 3 AND IOW n GND THEN latch R pwm GND data H 55 END IF IF hey thats us VCC AND addr latch 3 AND IOW n GND THEN latch L pwm GND data H 55 END IF Q IOCS16 data bus 15 data bus 14 data bus 13 data bus 12 data bus 11 data bus 10 data bus 9 data bus
303. quation on the output from the convolution process using the row mask These mono images show exactly where the line is and whether or not the line is vertical or horizontal After this edge finding equation is applied the data in the result buffer is then saved as a bitmap file The file header will be attached followed by the data in the buffer Although the newly saved images are only viewable in certain image viewing applications some of which are Microsoft Paint for Windows and XV for Linux The reason for this problem was not solved or researched any further This would be on enhancement for this algorithm Another programming constraint was each process the row mask and column mask process had to be executed independently If some linear algebra operations were conducted on them to obtain one matrix to convert the image The resulting image after the convolution process with this new matrix would not be same as the images shown in this paper However the time for the program to execute completely and write the data to files and save them is on average 1 5 seconds Which is fairly good for this application The processing is faster than the amount of time it takes to grab an image with the web camera Breakout Board Wiring Kyle Lamb The Breakout Board was constructed using a PC 104 board kit which was ordered from Diamond Systems An Altera chip was selected and placed on the 20 RTLinux On A Small Mobile Robot Bitsoi Lamb Will
304. r Development by Stephen Stange The transmitter for this project was required to be low power small inexpensive and lightweight It was also needed to transmit the signal at a frequency of 2 39 to 2 45GHz A block diagram demonstrating the manor in which the transmitter components were configured is shown in Figure 3 10 Data Signal Amplifier Oscillator Figure 3 10 Block Diagram for Transmitter Circuit 3 5 Transmitter Design Maxim Dallas Dallas TX provided a quadrature transmitter for this project The transmitter was a prepackaged chip the MAX2721 This chip included a LO Doubler local oscillator doubler this means that the oscillator used to tune the frequency of the transmitter needed to oscillate at half the desired frequency and an integrated I Q modulator see Section 3 4 2 1 for transmission protocol One of the main advantages of this chip is that a low input voltage is required where typical operation requires 3V Another advantage of using this transmitter was 33 that has been tested to work at temperatures up to 85 C with only a 1 5dB loss To ensure that the required distance of 100 yards line of site was met an amplifier was added into the final transmitter package The oscillator used for this project was provided by Mini Circuits Brooklyn NY The oscillator was a surface mount voltage controlled oscillator model number ROS 1121V Since this oscillator was voltage controlled it could be
305. r spectrum To move the plot around on the screen the programmer has only to specify the upper left and lower right corners of the plot Either of the plots can be turned on or off using a switch in procs which calls the method that draws the spectra During this time I also worked on the skymap changing it in much the same way as the spectra plots The skymap is now easy to move and resize by setting the upper left and lower right corners in three classes The outlines and mechanical limit lines are drawn in plots the sources are drawn in geom and sport draws the crosshairs which indicate the antenna position I encountered several scaling problems in altering the spectra and skymap mostly due to the fact that SRT uses methods in many classes to draw these objects The drawing commands pass through a number of scaling routines before actually being drawn and I had to keep close track of where scaling happened to get the items to display properly The second phase of modifying the display involved deciding where to put all of the display elements as well as making them clearer and more attractive than in the original 15 With artistic direction from Gerald and Brian I placed the skymap in the lower left corner of the display and the spectrum in the upper right In addition we superimposed the skymap on a picture of the VLA landscape so that zero degrees elevation lies along the horizon To further clarify the function of the sky
306. r to meet the power requirements of the robot decided to go with a 12V lead acid battery which gave us a surplus of power and made charging simple The battery was strapped to the robot using Velcro which made it easy to take off the battery if so needed Along with the 12V battery added a 3A DC to DC converter in order to supply 5V to the Prometheus and the sensors also added a 3A fuse and a connector that is used to recharge the battery using an AC adapter Connections to the various sensors had to be 26 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems added as well as wiring connections to the H Bridges for motor control and power connections to the Prometheus SBC and the Breakout Board Altera Programming for Breakout Board Bill Willems The Altera code within the Breakout Board has three functions Position Accumulators to gather encoder data from the motors PWM generation to control the speed of the motors and a PC 104 interface to get accumulator data to the Prometheus as well as PWM commands from the Prometheus PWM Generation was discussed previously and the complete schematic and Altera code can be found in Appendices B amp D Ch A Ch Bcw Ch Bccw Lu LI Figure 15 Quadrature Signal from the motor We receive data from the motors in the form of a quadrature signal as depicted above in Figure 15 A quadrature signal consists of two channels A amp B as the motor spins clockwise we get the sig
307. ral types of input sensors and RF transmitter receivers Choose several different sensor options and 2 transmitter options Design sensors that seem feasible past the elementary stage and test them to see if they work well for this application Build prototype LED array for system testing Report stating how the sensor selections and working end of October Report of how the transmitter and decoder work is going end of November Bi weekly status reports Prototype RF circuitry and encoding decoding system Design PCB s for system and combine into an integrated system Documentation of project including a users guide Testing of our final product Final document stating the functionality of system Figure 1 Chart listing of the main deliverable requirements 1 5 Design Specifications The following section states the necessary specifications of the transmitter receiver input sensor and LED array subsystems The transmitter needs to interface with the vehicle by connecting through the cigarette lighter plug This is the only direct connection to the vehicle for the entire system There needs to be built in noise immunity and surge protection on the control box The system must contain override switches on control box to manually signal driver intentions along with status LED s to display when the sensors are tripped or the manual switches are set Th
308. rallel libraries for use in parallel programming There are multiple MPI implementations the two most common on Linux being LAM MPI Local Area Multicomputer and MPICH MPI CHameleon PVM Parallel Virtual Machine the de facto standard message passing parallel library It predates the MPI specification and is used in many older parallel applications ZPL An array based parallel programming language intended to support engineering and scientific applications NFS Network File System The de facto standard for exporting file systems for use on a network consisting of UNIX or Linux computers NIS Network Information Services NIS handles user authentication in a network environment and is commonly used in conjunction with NFS to decide when a user can access a shared file system RSH Remote Shell This program allows the execution of commands on a remote computer and is used to start parts of parallel programs by the PVM and MPI parallel libraries 2 Technology Review and Design Decisions Affordability simplicity of construction and maintenance and reasonable performance were the major factors considered in the design and construction of this Beowulf cluster The design decisions that were available to this project and the final decisions made will now be discussed 2 1 Local vs Network Boot Several different file system configurations were considered in the design of this Beowulf cluster Network booting local booting with the
309. rature below 18 Figure 3 4 Internal temperature plot for insulation prototype eeeeee 18 Figure 3 5 Schematic of Temperature Sensor Circuit see 22 Figure 3 6 Basic flow chart for data acquisition and transmission 28 Figure 3 7 Timing diagram for the receiver us ones etr ooa eie eoe 29 Figure 3 8 Sample of receiver display ues cesses ces tsi eaaet de eub Gus Ex ERAN TE RUE Ed Eo etd 30 Figure 3 9 Plow chart for receiver eed ode ot dtes atq loan Me 3l Figure 3 10 Block Diagram for Transmitter Circuit eee 32 Figure 3 11 Power gain graph for amplifier 1 eei reticere 33 Tables Table 1 1 List of Customer Requirements eeeceesceceenceceeeeeceeneeceeeeeceeeeeceteeeesaes 3 Table 1 2 Transmission requirements uocare atrae crie et o ete eee re raa t 4 Table 1 3 Design requirements 21 225 eo v eae e pas 5 Table 2 1 Materials Considered 5 bulk DrIee Y eese edestets eee sansa tate nated agi uoa 10 Table 3 1 Material thickness needed to meet thermal specifications 15 Table 3 2 Observations from flame testing eeseeeeeeneeeeneee nennen 15 Table 3 3 Comparison of Temperature Sensors seen 21 Table 3 4 Minimum requirements for central processor 24 Table 3 5 Comparison of Processors 26 1 0 Background Information 1 1 History On July 29 1967 the USS Forrestal was operatin
310. rature exceeds the cook off temperature for that ordnance The method of loading programs onto the Brainstem is found in Appendix VII 3 4 2 Transmission protocol The onboard A D of the Brainstem is used for data acquisition After executing the necessary setup commands the device reads a value from the A D port as a 10 bit number at a set sampling rate After the value has been acquired the processor cues the receiver by sending out a series of 10 1 s followed by a zero then a value every clock cycle The cue of 10 ones is used because this pattern will never occur in the transmitted data After the cue has been sent the processor sends the value from the A D bit by bit least significant bit first This continues every clock cycle until the all ten bits have been sent After the tenth bit is sent the processors send zeros every clock cycle until the new sample is data space Activation temperature approximately the numbers from 0 to 75 allowing for 75 different ordnances to be used 28 taken The simple program structure is shown in the flow chart in Figure 3 6 below The code for the transmitter can be found in Appendix VIII Voltage from Temp Sensor Analog to Digital conversion Figure 3 6 Basic flow chart for data acquisition and transmission 29 3 4 2 2 Receiver protocol The receiver side of the system was more complicated This was attributed to the fact that the receiver and the transmitter did n
311. re a number of options available including software such as TheSky by Software Bisque TheSky would have provided a very nice graphical user interface GUI but would have taken work to interface with the SRT We decided early on to use the software that MIT ships with SRT This software is shipped in the form of Java code which must be compiled by the user to run the telescope We had a number of good reasons to use this software First it would save us the work of figuring out how to calculate azimuth and elevation coordinates of sources how to point the telescope and how to process information coming in from the receiver In addition the fact that the software was open source made it very easy to modify to suit NRAO s needs Thus the decision was made to use SRT s existing control software I volunteered to do the programming because I had the most programming experience in the group having written much of the code for my junior design robot Still I had not done any Java programming or worked with a program of this magnitude before The SRT control software consists of about 5400 lines of code divided into 12 classes Their names and functions are given in Table 3 2 1 12 Class Name Cat Primary Function parse srt catalog file checkey watch mouse and keyboard disp define functions for graphics and text geom compute az el pointing coordinates global store and provide access to global variables map contour 25 point map
312. re also known as the Sobel convolution mask 3 2 3 3 0 3 0 0 0 2 0 2 3 2 3 3 0 3 Figure 2 Row Mask Figure 3 Column Mask The convolution process takes each of the mask and is virtually placed over the shaded area depicted in Figure 4 actual image data The data in the image is the pixel information The mask values 3 2 3 0 0 0 3 2 3 are multiplied by the pixel values of the corresponding value when the mask is in placed after the multiplication the results are summed together and the results of this operation is stored in X Figure 5 The mask is moved to the right by one pixel The same operation is performed on the corresponding pixel values The buffer holds the result of the convolution process This process continues until the mask covers the entire image and the mask is at the lower right corner 17 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Figure 4 Image Data Figure 5 Buffer The following pictures in Figure 6 are the actual results from the convolution process Result from convolution process using Column Mask pe In this transition the colored image is converted to a grayscale image Result from convolution process using Row Mask Original Image Figure 6 Image Processing In each of the resulting images the information it contains is mo
313. re then we need for our application To extract the unnecessary information and retain the useful information we apply an edge point equation The equation we used is X D AHB CI A B C and D are the pixel values in the image The shaded area is the area the operation is applied to Figure 7 The result is placed in a buffer at the location that corresponds to the D location in the original image as seen in Figure 8 18 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems A B C X Figure 7 Image Data Figure 8 Result Placement This procedure continues until the entire image is covered by shifting the shaded area over one pixel to the right The equation checks to see if the pixel at position D is an edge If the color or the difference in the values are not equal to zero it is possible that there is an edge If this equation were directly applied to the colored image then you would get an output pictured in Figure 9 Figure 9 Edge Equation on Color Image This resulting image is not a clean and precise output as to the following mono images in Figures 10 and 11 Figure 10 Vertical Lines Figure 11 Horizontal Lines 19 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Figure 10 is the result of the edge equation on the output image from the convolution process using the column mask Figure 11 is the output from the edge e
314. regulator LDO The large value of the resistor was used to limit current By varying the PWM signal you can limit the amount of voltage passing through the charge circuitry to the battery You can pass anywhere from OV to Vs Vs is the solar panel voltage which is around 7V The charge circuit is placed between the solar panel and battery It is located under the planter in the circuitry housing area 3 3 5 Adjustable Program Switch The adjustable programming switch consists of a three pole toggle switch that outputs OV 2V and 4V Fig 2 High values of resistors were used in order to consume power The switch outputs are tied with the HC12 A D converters From this point the program monitors which value corresponds to which of the three watering programs that is to be implemented 31 Chapter 4 Summary Conclusions The Self Watering Planter was a project that required research and development Our customer Dr Greg Donohoe emphasized that we heavily researched various areas that would contribute to the project for the majority of the Fall semester Through research we were able to look at various methods of problem solving along with ideas on how to achieve the common goal of designing a self watering planter The Self Watering Planter provides power scavenging water recycling water holding tank adjustable programs for various plant species microcontroller moisture sensor battery charger battery pump and a unique planter design
315. rfectly to what the specs desire however he did not allow me to use it for the following semester At this time I came to the simple realization that I had to construct one Dr Donohoe suggested we spent the first semester researching the project and the second semester actually developing it Already 2 5 months went under way so I had about half a month to figure out a way to develop an efficient moisture sensor I spoke with Dr Donohoe and we came up with a schematic When I returned for the next semester I had problems right off the bat other problems that plagued me throughout the semester was bad parts overall mainly multimeters and resistors and I decided to flirt with other schematic designs The original schematic was in the form of a resistance to voltage converter The resistive probes would be in the soil and it would feed into the converter Vref reference voltage of preferably 5V is used from a voltage divider That connects to a series resistor and connects to a negative input of the inverter The positive input is grounded Then a potentiometer of 100k variable feedback resistance moisture probe connects itself to the output from the negative input Vout output voltage would be the negative R 14 resistance of the pot divided by the series R As long as Vref is fixed Vout depends only on the variable feedback R This is like a single supply OP AMP Furthermore the output connects to some level detector some a
316. rogram pvm rsh while not os path isfile pvm rsh pvm rsh default pvm rsh none possible pvm rsh programs for program in valid pvm rsh programs if os path isfile program possible pvm rsh programs append program print The following possible PVM RSH programs were found for program in possible pvm rsh programs print t s program print if len possible pvm rsh programs gt 0 default pvm rsh possible pvm rsh programs 0 The character sequences at the end of the following two lines suppress a newline from print and remove the newline from readline print Which PVM RSH program do you wish to use print default s default pvm rsh pvm rsh sys stdin readline 1 print if len pvm_rsh 0 pvm rsh possible pvm_rsh_programs 0 print PVM ROOT s pvm root print PVM ARCH s pvm arch pvm sh file open os path join user parallel setting dir pvm sh w o export PATH PATH s n os path join pvm_root pvm sh file write export PVM_ROOT s n pvm root pvm sh file write export PVM_ARCH s n pvm arch pvm sh file write export PVM RSH sWn pvm rsh oe pvm sh file write bin pvm arch 4 end of pvm setup 42 def zpl setup zplhome common zpl home dirs opt zpl zplhome opt zplhome usr local zpl zplhome usr local zplhome zpl home while not os path isdir zpl home zpl home de
317. roject Descriptions eese ein tre tpa epe e eere ee top een ein sese Dos edee e eEpS ne pere D 2 0 Technical Data e 1 PAN M ursa EE taeseaasoeeeaneeesantatee T Se CASTING oa ck sabre MEN RD PROB etd O 2 3 Temperature Sensor seccsssvccesssvecssosseresesessvcocsteesesosvecssesunsevccesesvcsssasavsen L2 2 4 Central Processor inet DIO mensium etuoenlZ 2 5 Transmitter i atiioi pni oi ici SO Ar ESI E ER Eel ir esiispaito LO 3 0 Original Developments e cere ee eee eee eee oe ee eese ee sese eesoseeesoseeeseses L4 3 1 Power Supply Research A e eere e eee eee eee ee eeeeo sees seseess 14 3 2 Insulation Mat rlIal iid oiioet sees veto yr gun oes PER ert paru et res epeu iere L5 3 3 Temperature Sensor ssvccseceosscsnceovscsscsuncssvceeyscveceovecoversecssncoesssoncvovessvers 20 3 4 Central Processor sseesseessooesooesoocesoccssocesocesoosesoessccesocesoosesoeessesesocssss 24 Kx Miriam iesirea aei OO a RUR DS TH LEITI i EE EET E TEE 5 0 Future Considerations esssesssesssecssocesooessossssecssocsooeesosesoesssocesocesoosssossssesssesese J 7 Works Cited Mec I etae 1 Appendices List of Figures and Tables Figures Figure S T Set ptor Tate TCSE tesconao n me oret b estet oeil cane s 16 Figure 3 2 Thermal test for first insulation casing using Fastblock 301 17 Figure 3 3 External temperature is plotted above and internal tempe
318. ropriate data output file from the ZPL simulation before running this file oe oe oe The data file contains results from the ZPL code s run on the Beowulf Cluster S ci ci N P XY G Mirror and Coating are all dumped to this file Creating a plot of the clip positions above the mirror figure 1 for v 1 1 N plot P v 1 P v 2 hold on end axis square xlabel Mirror Position cm ylabel Mirror Position cm title Clip Firing Pattern Plotting the resulting coating figure 2 mesh XY XY 10 8 Coating view 2 axis square colorbar vert xlabel Mirror Position cm ylabel Mirror Position cm zlabel Coating Thickness Angstroms title Coating Thickness Angstroms figure 3 mesh XY XY 10 8 Coating axis auto xlabel Mirror Position cm ylabel Mirror Position cm zlabel Coating Thickness Angstroms Stitle Perfect run no miss fires determining the range of thicknesses and other values as output A B min Coating S 1 C D max Coating S 1 ci co Minimum A 10 8 minradius abs 26 B 5 1308 Maximum C 10 8 maxradius abs 26 D 5 1308 Var Maximum Minimum ll ll Ave 10 8 mean mean Coating o writing some of the data to files 38 wklwrite e glass xls G S 1 3 2 2 wklwrite e mirror xls Mirror S 1 2 2 wklwrite position xls XY 2 2 curve fitting for u
319. ross the input Next the multiplier and ALU can be tested Again while all possible configuration inputs to the ALU can be easily tested all possible data inputs are prohibitively large Some operations of the ALU are tested the same as the data path while others are tested with numbers in interesting areas After a few general tests the numerical boundary conditions are more thoroughly explored The multiplier is tested in a similar manner Following that the link between the ALU and multiplier can be tested Because this is simply a data connection performing no computation the test cases which are used for the data path are sufficient Finally the conditional parts of the CMUX can be tested This is a simple test because the data passing mechanisms of the CMUX are tested in the 21 data path tests so all that remains is testing the conditional selection This can be tested exhaustively The data path phase of the testing ensures that the components which guide data through the ALU function appropriately These components include the muxes shifters and crossbar as well as the interconnections between all the components These tests check for paths which are stuck high stuck low stuck to an adjacent line and for reversed busses The test inputs for these configurations are generated by using each of 8 test patterns 1 11 101 111 1001 1011 1101 1111 and shifting it through all possible positions in the input word plus using the
320. s Figure 3 1 5 The Safety officer for the VLA Jon Spargo was contacted to address his concerns on the SRT project He was assured that the National Electric Code regulations for power and signal voltages were observed For warehouse parts that might be needed at the site we had to have authorization for dispersion so I contacted appropriate personnel for those authorizations In the beginning of March I came up with a physical design for implementing the lightning protection devices I consulted with Communications installation technician Pete Martinez of the Information Services Department of NMT about lightning protection For the end of this project I will oversee the final installation of this lightning protection equipment that was not a part of our original statement of work Figure 3 1 4 Visitor Center Kiosk Concept 11 32 Sam My work on the SRT project focused on developing software which 3 would control the telescope and provide a display of useful information for users This is a detailed description of the work I accomplished between August 2001 and May 2002 While we were in the initial stages of determining what needed to be done and how to split up the work one of the tasks that came up was the implementation of some sort of control data processing data display software that would run the dish and display the information it gathered in a manner suitable for educating the public about radio astronomy There we
321. s PC 104 modules feature a pin and socket connection system in place of card edge connectors as 10 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems well as mounting holes in each corner The result is an extremely rugged computer system fit for mobile and miniature applications PC 104 modules stack together with 0 6 spacing between boards 0 662 pitch including the thickness of the PCB Prometheus uses the PCI bus internally to connect the ethernet circuit to the processor It uses the ISA bus internally to connect serial ports 3 and 4 as well as the data acquisition circuit to the processor Only the ISA bus is brought out to expansion connectors for the connection of add on boards Diamond Systems 2002 The features found on the Prometheus SBC are outlined in Appendix E Breakout Board The Prometheus SBC has only two timer pulse accumulators One of these timers is already dedicated to the onboard I O module the other due to a flaw in the programming of the controller chip is also unavailable In order to control a single motor in a closed loop we need one timer to generate a PWM signal to control speed and one accumulator to read the encoder data generated by the motor We are using two motors creating the need for a minimum of four timer pulse accumulators Due to the lack of timer pulse accumulators provided by the Prometheus it was clear that we would need to either buy a third party timer pulse accumulator
322. s competitors 3 3 3 Solar Panel Power scavenging is one of the themes of the project so we had to look at various sources for obtaining power The choices we had to choose from include harmonic vibrations wind generated power and solar power For our application we could not find a steady source of wind in a window pain area so the method of wind generated power was eliminated Harmonic vibrations were recommended by our customer however in a planter and its stand we could not find sufficient vibrations to produce the desired power As a result we turned toward solar power since plants require sun and they are usually placed near a window When we turned toward solar power we investigated many solar panels From resources we found that we would have to allow for around 20 loss of power due to imperfect sun conditions Through calculations we estimated that we would require at most 533 A H day With the desired voltage and needs calculated we now knew what voltage and current we would be needing We needed an 8V 400mA solar panel We were allowing for 2V 200mA loss since our 29 desired need was 6V 200mA When we received the panel we ordered we immediately tested it Amazingly we were proved wrong by the panel as we only had 2V 20mA loss The solar panel was nearly perfect With this in mind we immediately chose the Silicon Solar Panel as our final choice The panel was placed at the top portion of the planter side so
323. s we installed the Swiss Army Knife of Embedded Linux BusyBox version 0 60 2 We started out with the 2 4 4 kernel because RTLinux was designed around this kernel We upgraded to a newer kernel to support devfs and upgraded again to the 2 4 19 pre6 kernel to solve our USB controller issues The first thing we needed to do was to compile the kernel in such a way that it would fit in the boot sector of the Compact Flash drive Once that was complete we designed the root filesystem using BusyBox We then wrote a script to reformat the Flash disk create the ext3 filesystem copy our filesystem source tree and then ran LILO from the Flash disk to correctly write the boot sector This does not sound like much but it took months to perfect the technique Once RTLinux was 30 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems up and running on the Prometheus we were then able to test our Breakout Board the corresponding driver and control algorithms Breakout Board Driver Bill Willems In order to read and write data to and from the Breakout Board and the Prometheus I O module we needed to write a driver for it Our module sets up threads to gather data from the hardware and pass that to the Behavior Control algorithm This bi directional communication is established using RT FIFOs There are three types of FIFOs a Command FIFO used by the user space control algorithm to start and stop threads as well as to pass PWM commands to the hard
324. s has the advantage that controlling input registers to create certain conditions only affects which output is selected and doesn t change the possible output values Hand written test cases are used to produce each of the desired conditions Running the full test suite runs roughly 2 5 million test cases although when testing changes to the design it will often require only a subset of the full test suite to be run to test that component Preliminary timing results indicate that each test case can be run in roughly 0 2 milliseconds requiring approximately 10 minutes to run the full test suite 24 CHAPTER 5 RESULTS amp CONCLUSIONS The most important aspect of our project the test cases are a complete success We have developed software which will generate a comprehensive set of tests and can be easily adapted to different interfaces In addition new sets of tests can easily be added and blocks of tests can easily be removed for testing a single component of the Processing Element Our second task was to develop an interface between a PC and an Altera board running the Processing Element This also was a success with both a hardware design for interfacing via a PC parallel port and software modules to communicate with it Our third task was to produce a report on the current state of the Processing Element design Our initial set of data path tests fail on about half the test cases When we tried to reproduce this error in simulation
325. s where they were not available from the manufacturer Several tests have been devised that adapt standard thermal conduction insulation tests to the available equipment in the department Speyer 1994 ASTM standards test E1225 was to be the principle thermal conductivity test This test method describes steady state techniques for thermal conductivity where the values are between 0 2 200 W mK up to 1300K or 1880 F ASTM Standards 2000 Such testing was not needed however because thermal conductivity data was found for all materials under consideration The tests may be used to verify thermal conductivity values given in the literature Using data for the insulating materials modeling was conducted using principles of unsteady state heat transfer Cotronics Corp 2000 The insulation was modeled to be a hollow hemisphere The dimensions of the electronic components are 6 5cmX6 5cmX7cm which for simplicity was assumed to be a hemisphere with radius of 4cm It was determined what was the minimum thickness required to insulate in the worst case scenario as described in Section 2 2 was determined Table 3 1 lists the thickness required for each available material which can serve as the main thermal insulation 15 Material Thickness m Appendix X Cotronics Liquid Silica 2700 F 2 0 Fastblock 301 castable ceramic 0 07 Fastblock 301 castable ceramic 0 21 Table 3 1 Material thickness needed to meet therma
326. screen Data Processing source incoming signals to appropriate displays Spectrum Analysis source incoming signals to appropriate displays Educational Graphics completion of educational displays Other Displays integration of other data files like weather or observation schedules and explanations of their relevance for public knowledge 40 Complete Installation physically install the SRT at the VLA Visitors Center Research order parts figure out what parts are needed for the installation Planning Engineering Contacts design the final installation layout and make contacts with appropriate personnel to get the jobs completed Reports Papers Presentation Material ongoing production of paperwork Final Documentation write this SRT operation manual complete with diagrams Compile Ongoing Diary Putt together all email and correspondence as a chronological history of the SRT project 41 42 Client E mails LLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLDLLLLLLLLLLLLLLLLLLLLLI Date Wed 17 Oct 2001 14 25 37 0600 MDT From Clint Janes cjanes aoc nrao edu gt To brajala nmt edu gbivens nmt edu sfield nmt edu Subject project My apologies for not making it to the meeting yesterday Sam showed me what has been going on and I m really impressed thanks Clint PS I ll try harder to make it next Tuesday Clint KEKE eoe K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K
327. scription 2 1 3 Declaration Segment 4 TITLE Trial 1 5 IPATTERN 6 IREVISION 7 AUTHOR us 8 ICOMPANY 9 IDATE 02 17 02 10 1 ICHIP triall PALCE16V8 Iz 13 l PIN Declarations 14 IPIN2 L Leftturn 15 IPIN 3 B Brake 16 IPINA R Rightturn 17 IPIN 5 C Continuous 18 IPIN6 A CLOCK2 19 IPIN 14 RF RIGHTFET 20 IPIN 15 LF LEFTFET 2 22 Boolean Equation Segment 23 EQUATIONS 24 ILF B L A L C L R B 25 IRF B R A R C L R B 26 _ Simulation Segment 27 ISIMULATION 28 ITRACE ONLBRCARFLF 29 30 JISETF A L B R C 31 ICHECK RF LF 32 33 ISETFA L B RC 34 ICHECK RF LF Oo Nn 36 37 38 39 40 4 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 ISETF A L B R C ICHECK RF LF ISETF A L BRC ICHECK RF LF ISETF A L B R C ICHECK RF LF ISETF A L B R C ICHECK RF LF ISETF A L B R C ICHECK RF LF ISETF A LBRC ICHECK RF LF ISETF AL B R C ICHECK RF LF ISETFAL B R C ICHECK RF LF ISETF AL BR C ICHECK RF LF ISETFAL BRC ICHECK RF LF ISETF AL B R C ICHECK RF LF ISETF AL B RC ICHECK RF LF ISETFAL B R C ICHECK RF LF ISETFALB
328. se in zemax optical analysis gc polyfit XY G S 1 3 7 wklwrite glass cross section curve fit xls gc mc polyfit XY Mirror S 1 7 wklwrite mirror cross section curve fit xls mc coatfit polyfit XY Coating S 1 7 wklwrite coating cross section curve fit xls coatfit 39 Appendix C Parallel library and system configuration scripts Python auto configuration script The following script parallel config py is a python program that should be placed in a parallelrc directory in the user s account This program handles writing shell files that contain the necessary environment variables to use ZPL and the parallel libraries The shell wrapper that follows in the next section allows the functions in they Python program to be called from the users shell In order to support other shells such as csh C shell follow the examples in this script where the Bourne shell data is exported to files and do the same for csh equivalents usr bin python import sys import os import re user home dir os environ HOME user parallel setting dir os path join user home dir parallelrc The following are the available ZPLCOMMLAYERs for the supported ZPLTARGET platforms listed on the zpl website This dictionary will require periodic updates valid zpl settings Workstation Settings valid zpl settings alpha osf seq valid zpl settings hppa hpux seq valid zpl setti
329. sed on aluminum and copper It also had to be applied to the surface of the material through a chemical process and could only be applied after the beam dump parts were machined to the correct size Black paint could be used on any material it was easy to apply and it was available the locally We decided to use the black paint because applying it did not involve a chemical process like black anodized involved Next the dimensions of the wedge plates had to be determined The basic shape of the wedge was a V as seen in figure XX The wedge consisted of two identical 10 Figure 1 Beam Dump with Laser Beams entering plates that were bolted together The opening of the V would be one inch and each plate would be one and one eighth of an inch wide The length and thickness of each plate were calculated using specific heat mass laser power and expected change in temperature of each material Which basically comes from the equation of specific heat c Q m AT Now that the dimensions of the beam dump for each material were determined this further narrowed the selection of material type Finally when all calculations were complete the availability of each material was checked into Aluminum was the only material that was available locally It could be purchased at the local hardware store in the thickness and width determined in our calculations This meant that machining would be made easier Therefore we decided to use aluminum because it was ea
330. sen program to water the plant All selections battery charger solar panel water pump programmable logic device etc have been made in the fall semester and the actual assembly began this semester which includes moisture sensor development The programming of the logic device began as well We found that the majority of our tasks went according to the gant chart and had very little errors Acknowledgments First we would like to thank our customer Dr Greg Donohoe for offering us the opportunity to endeavor on an exciting project His assistance and advice on energy scavenging as well as moisture sensor development were greatly appreciated In addition to Dr Bill Rison and our advisor Dr Stephen Bruder as they were always willing to offer us helpful advice mainly pertaining to the HC12 microcontroller Dr Ron Thomas was instrumental in helping with the actual circuitry and analog design mainly pertaining to the battery charger and we thank him gratefully for that We would also like to thank the engineer at Honeywell who informed us that we were investigating the wrong type of sensors for detecting moisture level as well as Environmental Engineering Professor Dr Mark Cal for presenting us helpful advice pertaining to a moisture sensor Without the donated parts and assistance of the Instrument Room Manager Mr Norton Euart we would not have been able to actually construct our planter Finally we would like thank our ins
331. sideration we had to acknowledge was the placement of the cameras There were two cameras provided by the customer The positions of the cameras were going to allow us to determine the distance to objects via depth perception The results from the image processing would be used to enhance and create a more accurate perception of the robot s surroundings Image Processing Algorithm Heather Bitsoi The Image Processing algorithm consists of five different parts The first part is reading a file that is saved to memory on the SBC This file is a 24 bit bitmap BMP image The image is determined to be a bitmap image if the file extension is bmp The image is retrieved from the IBM web camera which is mounted to 15 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems the robot The retrieval of this image takes on average 4 5 seconds per image The format of the image is as follows e abitmap file header this contains information about the type size and layout of a device independent bitmap file e abitmap information header which specifies the dimensions compression type and color format for the bitmap e The color table is not present for bitmaps with 24 color bits because each pixel is represented by 24 bit red green blue RGB values in the actual bitmap data area e an array of bytes that defines the bitmap bits These are the actual image data represented by consecutive rows or scan lines of the bitmap Each scan lin
332. sier and faster to get a hold of and since it came in some of the dimensions we wanted The final size and dimensions of the beam dump for aluminum are shown in figure 2 0 5000 0m i 10 1 0000 oq d o aD i 6 0000 Figure 2 Final dimensions of the beam dump 11 2 1 2 Heat Sensing Hardware Design Measuring the temperature rise in the plates was an important part in the laser beam dump design At first glance there was a wide range of surface temperature sensing hardware to choose from The options that were explored were thermocouples thermistors and IC temperature sensors Thermocouples are interchangeable have standard connectors and can measure a wide range of temperatures Their response time is in the milliseconds One of their weak aspects is that they require special connectors and test equipment which would significantly increase the cost per unit The prices for these sensors are about 13 00 each Thermistors have some of the same advantages as thermocouples The draw back of using thermistors is their slow response time Their response time can be in the range of seconds The price for this type of sensor is about 15 00 each IC temperature sensors are inexpensive fairly precise and can easily integrated into other circuits The solution to the heat sensing hardware was to use an IC temperature sensor The sensor chos
333. sly completed New Mexico Tech Rugby Team References available upon request 55 A 9 Ernest Jim s Resume Objective Education Employment Qualifications References Phone 505 838 0145 PO Box 128 Msg 505 368 5908 Shiprock NM 87420 E mail ejim800 hotmail com Seeking employment in the electrical engineering field that will allow me to utilize previous employment experience and skills obtained from my educational background Areas such as hardware analog digital design controls and testing 1998 2002 New Mexico Tech Socorro NM Electrical Engineering major to receive BS in May 2002 e GPA3 01 e Honor Roll Fall 2001 5 01 8 01 Arizona Public Service 4 Corners Power Plant Fruitland NM Electrical Engineering Intern e Assisted in designing UPS back up power system for server network e Responsible for cost estimates of several projects e Designed circuitry for several systems 5 00 8 00 BHP Minerals Fruitland NM 5 99 8 99 Electrical Engineering Intern e Revised and updated Drag Line electrical schematics e Assisted with electrical tribulations and solution development 5 98 8 98 Lawrence Livermore National Laboratory Livermore CA 5 97 8 97 Engineering Intern Assisted in design and fabrication of targets for two stage gas gun e Designed database for data collected from experiments e Prepared weekly reports based on performance and development e Knowledge of computer aided progr
334. sourced from their bashrc file Similar scripts can be written for other shells such as csh C shell if f parallelrc zpl sh then parallelrc zpl sh else echo ZPL had not been configured yet echo You must run zpl setup before using ZPL echo nn fi if f parallelrc pvm sh then parallelrc pvm sh else echo PVM had not been configured yet echo You must run pvm_setup before using the PVM libraries echo nmn fi if f parallelrc mpi mpich sh then parallelrc mpi mpich sh fi if f parallelrc mpi lam sh then 47 parallelrc mpi lam sh fi parallel script dir parallelrc function zpl case 1 in setup parallel_ script dir parallel config py zpl setup source parallelrc zpl sh change lib parallel script dir parallel config py zpl lib source parallelrc zplcommlayer sh a echo Usage zpl setup change lib esac function pvm setup parallel script dir parallel config py pvm source parallelrc pvm sh function beowulf cluster case 1 in setup zpl setup pvm setup beowulf cluster gen files gen files parallel script dir parallel config py gen cluster files x EF echo Usage beowulf cluster setup gen files esac 48 Appendix D Beowulf Cluster User s Manual Following this page is the User s Manual for the produced cluster 49 Thunderstorm Data Analysis Progress Exercises in Matlab with an E
335. specifier declaration specifier type specifier void char int string declarator identifier declarator parameter list parameter list parameter declaration parameter list parameter declaration parameter declaration declaration specifier declarator compound statement statement list statement list statement list statement statement asm statement selection statement jump statement labeled statement loop statement expression statement labeled statement identifier statement case constant statement default statement selection statement if expression statement if expression statement else statement switch expression statement asm statement asm asm list asm list opcode labeled statement jump statement return expression continue break goto identifier expression assignment expression expression assigment expression assignment expression conditional expression unary expression assignment operator assigment expression assignment operator conditional expression logical OR expression logical OR expression expression conditional expression logical OR expression logical AND expression logical OR expression ll logical AND expression logical AND expression inclusive OR expression logical AND expression amp amp inclusive OR expression inclusive OR expression exclusive OR expression inclusive OR expression exclusive OR expression
336. st the TE module would lose efficiency from the low AT Secondly the overall high temperatures would cause an even greater loss of efficiency and at a higher rate than the first Essentially the TE would only be able to produce the required power for a short time before failure Connecting the TE to a bank of capacitors was considered a solution to this problem However a TE only produces a small amount of current making this solution unreasonable 2 2 Casing The casing for the prototype protected the internal electrical components from the worst conditions that the device might encounter onboard a navy aircraft due to a JP8 jet fuel fire These conditions were assumed to be a maximum temperature of 1600 F see Appendix I and a heat flux of 6 8 BTU ft s ares ame arizona edu materials ysz html 2000 The true worst case heat flux value was not available and could not be determined by our design team In general the heat flux from a fire has been found to be variable Estimates of the variability of heat flux in a fire was quantified by NIST and with more information on the conditions of a fire the variability in heat flux may be determined using empirical equations found by NIST research ares ame arizona edu materials ysz html 2000 Because the harshest conditions the prototype was designed to meet are thermal the prototype casing must be composed of a thermally insulating material Under the conditions outlined above the insulation
337. surfacing simulation program on a IOBase T network iieiea erae S PUpHAs atout A Sas UR S NIARDIE Qva Doce qu ERE ARV 9 Figure 5 Comparison of the NAS MG benchmark using the Class A data set on IOBase T and TOOBase T networks eei eoe terrre rete eonun to aea do ono 10 Figure 6 Performance comparison of NAS MG benchmark with the Class S data set using the MPI and P VM HBEAEIS Loto erts tpe pure deben 12 Figure 7 Performance comparison of NAS MG benchmark with the Class A data set using the MPI and PVM libraries eicit re tette enitn ERR EXPE O ran UL ee Hgn eei 13 Figure 8 Performance of mirror resurfacing simulation program using the MPI library deditio sie ie Dind sane San hala don eas E T E Saul anon Bes eddy Deda p vo E 14 Figure 9 Performance of mirror resurfacing simulation program using the PVM library toco c iet Eo am euet deo cae cos Leto aeos 14 1 Project Background and Goals 1 1 Project History The first Beowulf cluster was built by Donald Becker in 1994 at NASA s Goddard Space Flight Center Since then further research has been done by other national labs universities researchers businesses and hobbyists This Beowulf cluster project is based on this existing body of work with the goal of simplifying the process for application to small scale university projects 1 2 Project Description The design and construction of a Beowulf cluster is still a highly technical activity requiring a proficiency in operati
338. t SDOCIGS suerte rues tma tcess se a ie teem eae T Poele TR 39 Figure 107 Sensor Test Water sess eevtiasSe PUER PeI Ee EORRRA SA RV M NV EPI E Ve HERRERA NT ED 40 Figure 11 Sensor Test dry OUt cies eo S P E eet S eet ees 41 Figure 12 Sensor Test wet dirt esses em emen 42 y E AER Final Test NA RT 43 Fieure 14 BUE oestox eee vta won DORIA REA unto avc aes aw SEN EEEE eri 44 Chapter 1 Introduction 1 1 Project Description On September 10 2001 Dr Greg Donohoe presented a project for us to undertake His idea was for us to design a self contained system that would automatically water a plant when the plant needed it all without human intervention We decided to accept the challenge He also wanted us to scavenge energy from an external source to maintain its power storage requirements 1 2 Definitions There are several definitions to this project First we needed to determine if it is feasible to design a self contained system that will water a plant when necessary The type of plant to be used will be pre programmed into the HC12 microcontroller AII the user has to do is input his her plant species The HC12 will control the water schedule of that plant from that point on or until the input has changed Second we needed to realize what type of power source to use We used solar cells and to store the power we used a lead acid battery Other ideas thrown around include harmonic vibrations wind chemical mechanical or
339. t and design Further sensor development Finalize sensor design intensity research Final assembl Thesis Formal presentation Final testing User s manual Task to be com Task to be com leted by the entire Grou Milestones c I S 3 9 gt April Ma M 2 Small Radio Telescope Project A Thesis Presented to the Faculty of the Electrical Engineering Department of New Mexico Tech By Gerald Bivens Sam Field and Brian Rajala In partial fulfillment of the requirements for the course EE 482 Senior Design Project II May 2002 O 2002 Gerald Bivens Sam Field and Brian Rajala Abstract The National Radio Astronomy Observatory NRAO is a publicly run institution under the direction of the National Science Foundation NRAO runs a number of world class radio astronomy research facilities One of these is the Very Large Array VLA located in the San Augustin Plains west of Socorro NM The VLA has been receiving record numbers of visitors and because of this public education about radio astronomy has risen in importance for NRAO For this reason Smal Radio Telscope SRT is being built for NRAO for use at the VLA Visitors Center where it will play a major role as an interactive display and educational tool to explain radio astronomy to the public The goal of the SRT project is to build a small radio telescope which will be installed for use by visitors to the VLA Project
340. t appropriate Gain rtl outb UNIPOLAR GAIN 0 5 V AD GAIN SCAN SET FIFO depth to 1 rtl outb 0x01 FIFO THRESH Configure the DIO temp rtl inb DIO CTL get current config temp DIR PORTA INPUT DIR PORTB INPUT PORTs A amp B are input temp amp DIR PORTC 3 0 INPUT DIOCTR PORT Cl to output and DIOCTR to 0 rtl outb temp DIO CTL Enable DIO interrupts temp rtl inb INT DMA COUNT CTL temp ENABLE DIO INTS rtl outb temp INT DMA COUNT CTL enable the Prometheus IRQ and assign a handler rtl request irq PROMETHEUS IRQ intr handler Disable timer interrupts and DMA operation temp rtl inb INT DMA COUNT CTL temp amp ENABLE DMA ENABLE TIMER INTS rtl outb temp INT DMA COUNT CTL assign a device file handler for the Control FIFO rtf create handler CONTROL FIFO amp my handler return 0 Figure 17 init module example code Just as a module can be loaded it can also be unloaded For this we need a routine to be executed that frees any memory we allocated shuts down the hardware etc In our case the routine cleanup module is executed to destroy the FIFOs we are using free our threads release the Prometheus IRQ and send a PWM of zero to stop the motors This code is demonstrated below in Figure 18 void cleanup module void let go of fifos rtf destroy CONTROL FIFO rtf destroy R POS FIFO
341. t but they fit The 7805 and transistor holes were too small but after we re drilled the holes they fit perfectly The problem was with the connectors We had ordered 156 mil spacing for all our connectors so we had to design the footprints for all the connectors We got the spacing correct but failed to make the holes larger not noticing that there was a difference between our connectors and the standardized Protel connector pads This affected about 15 connectors To fix this we had to grind down the pins on all the connectors so they would fit in the board Resizing the current connectors was the only option as for the vendors we found only used pins of at least 45 mils wide for connectors with 156 spacing The PC board was a success and worked perfectly even with these few setbacks Chapter 3 Final Sensors Status For the wireless transmitter project the driver s brakes and turning intentions needed to be sensed without tapping into any of the wires To do this sensors were needed to sense the driver s intentions of two different functions turn signal and brake We researched and tested several possibilities for each task using some of the sensors to try and accomplish both functions by applying it in a different way to accomplish the needed task This chapter is broken up into two sections a brake sensor system and a turn signal system 3 1 Brake Sensor System Development The brake pedal sensor would need to sense when the brak
342. t source acting like a buffer The problem I realized is that this particular OP AMP requires a lot of power 15 Another circuit was a comparator type of device determining dirt is dry or not at a specific resistance but it can only go up to 2V with the power supply for efficiency to be discussed later the minimum necessary voltage for dry dirt in order for a sensor to operate is 3V Finally the working circuit consisted of simple current sources Current sources supply 18mA when Vb base voltage is properly tuned using the variable resistor at the base of the transistor 18mA is provided so Vo output voltage from each of the sensor should match the graphs I provided Figs 12 15 I could not set up a special voltage OV dry 5V wet without adding extra active circuitry I did not really do a quantitative analysis little wet really wet etc but I figure the HC12 microcontroller should switch on the waterer when the sensor goes out of a certain range of voltages between the dry and wet bars on 18mA bar graph There is also a saturated transistor switch incorporated into it Fig 4 This allows a small control current to enable a much larger current to flow in another circuit for this case the current source The result is once it is fed 18mA it should say 3 6V near necessary minimum according to the graph for the dry dirt The actual value resulted in 3 61V Fig 13 I decided to measure resistance of dirt as
343. t would provide stored inventory data related to the ordnance as well as real time temperature readings This signal would be transmitted through the electronic material casing and the noise produced by the hazardous environment of a jet fuel JP8 fire to an intelligent fire fighting receiver station The materials chosen for the casing of the prototype had to insulate against the heat of a JP8 fuel fire withstand flames and as stated allow the signal generated by the internal electronics to pass through this casing Secondly this casing had to be lightweight cost effective and durable The electronics must be reliably powered during operation Use of the heat produced by a fire to generate this power was considered in addition to alternatives Several power supply sources were considered These included a thermopile or array of thermopiles in a Peltier junction configuration This could scavenge the needed power from heat A battery in conjunction with a thermal switch was also considered The insulation materials considered were commercial ceramic based materials These included dense refractory ceramics ceramic foams fibrous materials adhesives and coating A processor was needed that could accommodate stored data and interpret incoming thermal readings provided by a temperature sensor A transmitter was needed to send this signal to a receiving station The final prototype was created using the following parts The insulation used a c
344. ta double rsqrd R_rsqrd double layers R_rsqrd double G R G double D R D double Edge R mirror double 32 surface R mirror integer temp5 temp6 temp7 R_mirror double R_mirror double R_mirror doubl e Coating R mirror double Coatingtemp R mirror double Mirror R mirror double n m ni no wi wo t b temp integer integer integer integer integer integer integer integer double procedure mirror begin write S d S dx 2 S 1 d mirror grid n 2 S 1 ResetTimer Clear the timer for benchmarking or runtime info Open the output file for writing output file open output filename w Cx 1 Cx 2 Cx 3 Cx 4 Cx 5 Cx 6 l I NEHO Specify whether each set of clips fires or not Ci 1 2767 01 Ci 2 C102 Ci 3 Ci03 Ci 4 Ci04 Ci 5 C105 Cile Ci06 Ci 7 C107 Cil 8 Ci08 Cil9 Ci09 Ci 10 C110 Ci 11 CP Ti Ci 12 Ci12 Co 1 Co01 Co 2 2 00025 Co 3 Co03 Co 4 Co04 Co 5 Co05 Co 6 Co06 Co 7 Co07 Co 8 Co08 Co 9 Co09 Co 10 Co10 Co 11 Co11 23 Co 12 Co12 Co 13 Co13 Co 14 Co14 Co 15 Co15 Co 16 Co16 Co 17 Co17 Co 18 Co18 Co 19 Co19 Co 20 Co20 Co 21 Co21 Co 22 Co22 Co 23 Co23 Co 24 Co24 Init
345. ta sets This made starting a project a snap but working within it was still cumbersome To this 12 end I programmed the subplots of any given panel to be interlaced Now when one graph is re scaled the other graphs are automatically re scaled to match Our customer found this to be a very useful improvement I was reassigned to modify the data acquisition software to work with the new generation of EFM however it was later discovered that the new hardware would not be completed in time I then assisted Ginny with making preparations to the old EFM for a forthcoming upgrade but then that too was taken out of our domain Finally I worked with Chris on the FIR to phase shift the mercury switch See figure 9 Chris Reiten Matlab Coding and Digital Signal Processing From the beginning of the project we all knew that there was going to be a lot of digital signal processing and electricity and magnetism involved Fortunately I took Digital Signal Processing in the fall semester This helped me to better understand some of the tools that would be very important in determining the magnitude and direction of the E field vector An important primary step in extracting the E field vector from the data is determining the relationship between the Hg switch data the E field data and the B field data We all figured that this relationship is primarily a matter of the phase difference between the individual data components Using the Fast Fourier Tr
346. tatus This sensor has proven to be very reliable and works well for this application in a large variety of vehicles due to the simplistic design and ease of mounting since turn signal levers and steering columns are reasonably uniform This lever switch is shown in Figure 8 Figure 8 Lever switches 3 2 5 Sharp GP2D120 Description This is a general purpose IR distance measuring sensor The output voltage on the sensor increases as an object gets closer Application To apply this sensor to work for the turn signal there would need to be a sensor attached to both the bottom and top of the turn signal shaft Each sensor would be angled toward the steering wheel column so that there will be no voltage reading when the turn signal shaft is at the idle position When a turn is indicated the sensor would then be at such an angle that it reflects off of the steering column indicating whether it is a left or right turn Final Status This was a little too complicated of a design to implement in such a space constrained place Using two distance sensors made the cost of this sensor fairly considerable We also ran into mounting issues with this design because the sensors themselves were too bulky to place anywhere so they would not interfere with the driver and normal operation of the turn signal lever This sensor worked for the brake but wasn t a good approach for the turn signal application Another issue that this sensor brings up is that t
347. team of physicists have been collecting data from thunderstorms with an electric field meter Over the course of a thunderstorm they recorded orientation of the instrument the earth s magnetic field relative to the instrument and the electric field relative to the instrument Our job is to analyze the data collected from the thunderstorms Dr Krehbiel wants a more accurate interpretation of the electric field While the magnitude of the electric field has been extracted from the data the direction remains a mystery Using the magnetic field data as a reference we hope to determine the three dimensional electric field vector This project makes extensive use of digital signal processing Our raw data are just numbers At time t the other data has values of x y and z If we plot this data over a certain time interval and connect our points we get waveforms Raw E data T T T T T T 1 1 1 1 1 0 5 1 4 5 2 25 amp samples m Raw B data x10 4000 000 2000 i 1000 1 1 1 1 1 0 5 1 tS 2 2 5 3 samples 4 Raw Hg data x 10 T T T T T T 1 48 1 6 5 4 0 2 0 0 5 1 tS 2 25 3 samples 4 x 10 Figure 3 Waveforms produced by original data 10 There are data sets from several storms and over 40 000 individual data points per storm It isn t practical to plot this by hand so we are automating the process using Matlab These waveforms show patterns of data behavior We want to filter out different c
348. technique FET fast Fourier transform Common digital signal spectrum analysis Hall Effect The Hall effect occurs when the charge carriers moving through a material experience a deflection because of an applied magnetic field This deflection results in a measurable potential difference across the side of the material that is transverse to the magnetic field and the current direction Hg switch mercury switch In one position mercury flows to the end of a small tube where it touches two contacts closing a switch and indicating a digital 1 In the opposite position the switch is open indicating a digital 0 This is used to determine up down orientation of the EFM NOAA National Oceanic and Atmospheric Administration for more information visit http www nws noaa gov NSSL National Severe Storms Laboratories for more information visit http www nssl noaa gov Literature Review Books Applications of Linear Algebra 2 Edition Chris Rorres and Howard Anton Data Reduction and Error Analysis for the Physical Sciences Philip R Bevington Digital Signal Processing A Computer Based Approach Sanjit K Mitra Digital Signal Processing Using Matlab V 4 A Bookware Companion Problems Book Vinay K Ingle and John G Proakis Papers Evaluation of a Standardized Sine Wave Fit Algorithm Peter Handel Thunderstorm on July 16 1975 Over Langmuir Laboratory A Case Study Journal of Geophys
349. teristic of determining proper water usage for plants Plants that allow fewer levels of moisture than normal tend to tolerate humidity more and vice versa No exact recommendations were exactly given for watering plants including a lady that insisted that it had to be watched by a human but there are some general guidelines The soil at each watering should be thoroughly wetted or moist water should drain out the bottom of the pot or wherever after watering hence we should use containers with drainage holes Plastic pots what we intended to use but went with Plexiglas which is still alright is typically better than clay ones Clay pots dry out faster and are more difficult to clean than plastic ones due to the fact that clay absorbs chemicals and salts Frequency of watering depends on many factors but mainly plant species moisture intake level If all cases fail we should not water more frequently than required wet soils can lead to root rot problems In addition excessively wet soils lack the oxygen required for root growth Watering plants at room temperature water is also a must as long as the water is not cold Since plants do not conform to just one moisture intake level and as explained each plant displays various characteristics we will be implementing three 18 programs deduced from the original five ten into the HC12 microcontroller The sensor will act like a well sensing device to measure how much m
350. th simple hand tools allowing the average consumer to simplify the amount of wiring necessary to setup a trailer signaling system 1 3 Project Goals The objective for this project is to build a wireless vehicle signaling system to eliminate having to run a cable from the vehicle to plug into the trailer We were given several requirements from the project sponsor to include in the project Two separate transmission systems needed to be researched tested and built capable of meeting the Wireless Vehicle Signaling System s requirements A huge requirement was that the whole product must be universal It had to have the capability of working in most makes of vehicles The two systems needed to be compared and evaluated in order to select the best one to take into the final design A main requirement was that it be user friendly in all aspects It needed to be set up by the average consumer in minutes with no needed training To ensure this there could be no direct wiring to the vehicle electrical system We had to design and develop several sensors capable of detecting the intentions of the driver since we were unable to tie into the vehicle s wiring The sensors needed to be reliable and sturdy The whole system would need to be laid out and then designed in PCB design software The board would then be sent out to a manufacturing house to be professionally etched The system also required a set of prototyped LED arrays and light driver circuitry
351. that exceeded the Minix standards He began his work in 1991 when he released version 0 02 and worked steadily until 1994 when version 1 0 of the Linux Kernel was released The current full featured version is 2 4 released January 2001 and development continues Linux is developed under the GNU General Public License and its source code is freely available to everyone This however does not mean that Linux and its assorted distributions are free companies and developers may charge money for it as long as the source code remains available Linux may be used for a wide variety of purposes including networking software development and as an end user platform Linux is often considered an excellent low cost alternative to other more expensive operating systems Due to the very nature of Linux s functionality and availability it has become quite popular worldwide and a vast number of software programmers have taken Linux s source code and adapted it to meet their individual needs At this time there are dozens of ongoing projects for porting Linux to various hardware configurations and purposes Linux Online 2002 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Real Time Linux RTLinux is a hard realtime operating system that coexists with the Linux OS With RTLinux it is possible to create realtime POSIX 1b threads that will run at precisely specified moments of time The basic premise underlying the design of RTLinux is that it is
352. the filaments correctly P zeros N 3 Sinner filaments for mn e dxzlilr2 no 1 for w 1 1 6 if Cip w 1 P ci n ci no 1 F n 1 Cx w F n 2 Ri P ci n ci no 2 F n 2 Cx w F n 1 Ri no no l end end end Souter filaments for m 1 1 24 no 1 for w 1 1 6 if Cop w 1 P 12 ci co m co no 1 F m 12 1 Cx w F m 12 2 Ro P 12 c1i co m co no 2 F m 12 2 Cx w F m 12 1 Ro no no 1 end end 25 end clear F creating a plot of the clip positions above the mirror figure l for vz1 1 N plot P v 1 P v 2 hold on end setting up the position vectors to describe the mirror as coordinates since it will be a grid we need only assign points to a single row the x and y terms matching XY zeros 2 S41 1 XY2 zeros 2 S 1 1 for k 1 1 2 S 1 XY k 1 2 128 274 256 54 2 8 k 1 XY2 k 1 2 XY k 1 2 5130 8 0 5 H end setting up an array to hold the position vectors for each point on the mirror grid G zeros 2 S 1 2 S 1 3 D zeros 2 S4 1 2 S 1 2 for q 1 1 2 S 1 G q 1 XY G q 2 XY D q 1 XY2 D q 2 XY2 end G 3 8um D 3 clear XY2 clear D 26 now we have an array P that lists the positions of the N point sources in vector form and another array G that lists the positions Sof the points on the grid placed across the mirror s surface now to convert all this
353. the lowest number that applies for all nodes print default s default cpu per node cpu per node str sys stdin readline 1 print if len cpu per node str 0 cpu per node default cpu per node else cpu per node int cpu per node str print What base name should be used for the node names print default s default node base name 45 node base name sys stdin readline 1 print if len node base name 0 node base name default node base name node list ip base addr num ip4 to int ip base addr for i in range node count v node name s 03d node base name i 1 node ip int to ip4 ip base addr num i node node ip node name node list append node output etc hosts file open hosts w file write Do not remove the following line or various programs Mn file write that require network functionality will fail n file write 127 0 0 1Ntlocalhost localdomainNtlocalhost Wn for node in node list file write s t s s t s n node 0 node 1 domain name node 1 file close output etc hosts equiv file open hosts equiv w for node in node list file write s s n node 1 domain name file close output machines LINUX file open machines LINUX w file write Change this file to contain the machines that you want to use n file write to run MPI jobs on The format is one h
354. the requirements and end solution for the design of the PC test software are given Finally a detailed explanation of the generation of test vectors is provided 4 2 PC HARDWARE INTERFACE The standard PC parallel port provides twelve output and five input pins The output pins correspond to two eight bit hardware registers an output byte and a control byte Only the four low bits of the control byte are available as output pins The five input pins are available as an eight bit status register Only the high five bits correspond to actual input pins D7 D6 JD5 DA D3 DZ D1 DO J bL IL LA 1 12011 in LEE a 7 amp 5 4 3 g wie ssieeee D c EEEE TT T LT TES Figure 12 PC Parallel Port Diagram Our project required only eight output bits so we used only the data register Bits 3 0 were used for data in 3 0 bits 5 4 for TxRx state bit 6 for TxRx clock and bit 7 for PE clock The output from the Altera interface data out 3 0 was read via the status 18 4 3 register bits 7 4 For a full wiring diagram of the Altera to PC cable refer to the project CD SOFTWARE DESIGN The purpose of the PC test software is to run enough tests on the hardware to either find any problems with the design or to provide a reasonable assurance that there are no problems Several factors influenced the design of the test software The design needed to be easy to implement portable and fast In order to keep the implementatio
355. the resistance of the generator equals the resistance of the load Rgen NS Rm Np NS is the number of modules in series 2 and Np is the number of modules in parallel 1 Rgen 6 2ohms not far off of 5ohms The total heat input to the generator is QhzNS Sm Th D Np 0 5 2 2 Rm4 Km DT Qh for a our module set up is 167 2watts Generator efficiency Eg is PO Qh PO is the power output from the module configuration Eg 2 99 The heat transferred to the cold side is Qc Qh PO 167 2watts Swatts 162 5 watts The equations and figures above only represent one DT For our device the TE modules would be embedded into the insulation If we apply the above equations to our device the 162 5watts of heat that is transferred to the cold side will have no place to dissipate to as the insulation stops the heat from moving in that direction Essentially over time the heat builds up on the cold side and as a result the DT drops This continues until DT is zero At large temperatures more heat is produced regardless of DT The smaller the DT and the higher the temperature the more modules are needed to produce the same power therefore this design is not feasible at this time Appendix III TEA Grammar Index translation unit external declaration translation unit external declaration external declaration function definition function definition declaration specifier declarator compound statement declaration specifier type
356. the same test cases work flawlessly All the synthesis settings in the Altera software had no effect on our results This leads to one of two conclusions The first possibility is that a piece of the Processing Element design simulates well but cannot be properly synthesized If this is the case then we have succeeded and saved our customer a significant amount of time and money by identifying the problem at this stage of the design rather than it being discovered when the chip is manufactured The second possibility is that there is a flaw in the Altera synthesis software If this is the case it is out of our hands but our testing system will still be a valuable tool for our customer once Altera has corrected the problem Our customer will further investigate the problem to determine whether it is a flaw in his design or a flaw in the Altera software The bottom line is that our project is a success and our customer is pleased with the final product 25 References Donohoe Greg Reconfigurable Data Path Processor Overview Institute for Advanced Microelectronics 2002 Altera Corporation NIOS Embedded Processor Development Board January 2002 http www altera com literature ds ds nios devboard pdf February 2002 Altera Corporation NIOS Embedded Processor Software Development Reference Manual January 2002 lt http www altera com literature manual mnl_niossft pdf gt February 2002 26
357. the way down to the senior design lab on the first floor We finally obtained permission to mount the SRT on the roof of Workman from the Physics Department chair Dr Minschwaner after Clint Janes spoke with him We met with the university architect Dan Jones to verify that the mounts on Workman were suitable for this project In early November when the SRT kit finally arrived our team got together to assemble and mount it on the roof of Workman When we received the electronics and cable assemblies a short time later we again got together to run them from the roof to the second floor lab We all became familiarized with the antenna control program and the educational slide show demonstration We also obtained a closed circuit television camera and got together to run the cable again from the roof to the lab The camera served two purposes First it provided us with visual feedback as to the dish s motion Also it provided a measure of safety by letting us know when people were on the roof or near the antenna In March we took a trip to the VLA and saw the completed antenna mount and the installation of the conduit runs for grounding control and signal wires along with a spare conduit that might be used for power Communication with the carpenters about the design of the Monitor cabinet led to the conclusion that the cabinet would have to come from somewhere else A decision was agreed upon as to the size of the fence that would enclos
358. then applied these equations to a perfect sine wave to see if it would produce a perfect sine wave as it should The answer was yes However when these equations were applied to a pseudo sinusoidal wave which our real data turned out to be the result was not regular enough to use as a sine wave I tried many approaches and had several people check my logic and math My conclusion was that with only a third degree polynomial one degree being DC offset we couldn t possibly have enough information to produce a sine wave Originally we were going to collect data in an ideal environment using an EFM I located the meters talked to the appropriate powers and coordinated acquiring one This was no small task I had to search the web for contact information and convince a stranger to let me borrow his expensive scientific equipment I sent him special EFM boxes which Federal Express promptly destroyed so we were back near square one Once we received the instrument we still needed an area to set it up in equipment to receive the transmitted data and software to organize the data I developed a mounting system to simulate the release of the instrument into a cloud I also kept in close contact with the owner of the receiver equipment and learned how to operate it I then devised a series of experiments that I wanted to perform Unfortunately I only got 21 to run one experiment before our customer decided that he would rather we focus on an
359. ties to sustain a high continuos load draw over time 3 2 2 Light Sensing Real Time Clock The key critical component of the project is actually the device that generates the external interrupt that is fed into the IRQ pin of the controller The device that the planter is presently running is an IR Phototransistor circuit Fig 5 The light sensor when used in conjunction with a comparator circuit will generate a 0 or 5V signal for darkness and sunlight respectively This signal is CMOS compatible with the HC12 s IRQ input When I tested this circuit it worked perfectly It was able to output 5V when light hit the sensor and OV when the light was taken away The only drawback to this sensor was that it responds to any kind of light sunlight or artificial light So for this planter as soon as the sunlight hits it in the morning it will constantly trigger until sunset thus causing the controller and the rest of the circuitry in the planter to use power needlessly So I decided to use this type of sensor as a last resort if I could not find some other kind of device that was more efficient So after searching extensively for an alternative replacement I found a suitable test subject called a Real Time Clock RTC This device functions almost like your bedside alarm clock You program time and date as well as the alarm time and date When the alarm time matches the current time it generates an interrupt that can be fed into the HC12 It can be
360. tinued to find and explore new aspects of Matlab I bumped into the Filter and Design and Analysis Tool and shortly there after found the Signal Processing Tool Through a great deal of experimentation and study I managed to isolate the fundamental frequency of the mercury switch data Chris s work with the Z transform was a tremendous help At this point I hit a limit The data being digital does not yield sine waves Instead the filtered signal was an obfuscated triangle wave When this same filter was applied to the Hall effect sensor the output was interesting but non intuitive We later solved this problem by interpolating the data I invested a great deal of effort into a now abandoned task Initially I created a program written in the C programming language which removed data spikes from both the B field and E field data sets The E field spikes which were formally believed to be data anomalies were closely inspected and determined to be lightening Because of this a decision was made to allow them to remain We do not know the cause of the B field data spikes but they do not correspond with spikes in the E field Thus it is impossible to use the B field as a trigger for filtering E field data anomalies These discoveries were side effects of code rewrites and experiments As the data analysis portion of our project began to take form the product became difficult to interact with I worked with Ginny to build a menuing system for the da
361. to get to a particular thesis is through the bookmarks of this PDF file I would like to express my sincere gratitude to the faculty and staff of the Electrical Engineering Department for their support of the Senior Design program in particular Robert Bond Stephen Bruder William Rison Hassan Shanechi Ronald Thomas amp Kevin Wedeward for their participation as faculty sponsors on the various projects and Carrol Teel for her patience in dealing with all of us I would also like to thank Dr Deidre Hirschfeld of the Materials Engineering Department for her support and help on a multidisciplinary project between our departments Scott W Teare New Mexico May 2002 Table of Contents Thesis Title Design of Self Watering Planter with Emphasis on Power Scavenging Project Sponsor Gregory Donohoe UNM Remote Ship Board Heat Detection Prototype Project Sponsor Eric Wilson NAVAIR Laser Beam Dump and Power Meter Project Sponsor Dan Eastman Boeing SVS Wireless Vehicle Signaling System Project Sponsor Jason Vandehey Small Radio Telescope Project Project Sponsor Clint Janes NRAO Real Time Linux On A Small Mobile Robot Project Sponsor Stephen Bruder EE NMT Beowulf Cluster Research and Implementation Project Sponsor Scott Teare EE NMT Thunderstorm Data Analysis Progress Exercises in Matlab with an Emphasis in Digital Signal Processing Project Sponsor Paul Khrebiel Physics NMT Reconfigurable Data Path Pro
362. to help gain an understanding of lightning An instrument called an electric field meter was attached to a helium filled balloon and released into a cloud where it collected and transmitted important data regarding a mercury switch the earth s magnetic field and the electric field of the thunderstorm This data was given to us as dimensionless lists of numbers indicating different parameters in each storm The goal of our project was to assess this data and from it extract the 3 dimensional electric field vector at the instrument E Vector Figure 1 3 D electric field vector This was accomplished through several forms of data analysis including filtering interpolating demodulating and a vast array of digital signal processing tools Our final product is an interactive Matlab program that allows the user to select the flight and section of data they wish to study It not only returns the 3 dimensional vector of the electric field in the cloud but also displays other useful data in graphical format that is indicative of lightning strikes Definitions and Acronyms B field Earth s magnetic field as seen by the Hall Effect sensor in the EFM E field storm cloud s electric field as seen by the current induced across the conductive spheres of the EFM EFM Hlectric Field Meter the instrument used to collect the data for our project Illustrated on page 9 FIR finite impulse response filter Common digital signal filtering
363. to power the arrays The whole system would need to cost less than 150 while in full in production and a 200 goal for early production 1 4 Deliverables for the Wireless Vehicle Signaling System The project started on September 10 2001 and was completed May 1 2002 The group created a Gantt chart for each of the first and second semesters of the project APPENDIX M The tasks were divided up and distributed to the three group members depending on interest and the skills that each possessed One of the most important tasks that was handled was the documentation of project design Bi weekly reports were written to give the faculty and project sponsor a heads up on the progress and a summary of the successes and issues that were arising in the project Also more formal work plans and project summary reports were required giving the customer a good look at what work was being done and that the project was going as scheduled By reporting on any issues the group had come across the team was able to see all the issues that arose early and was able to tackle them in a timely manner The following figure lays out the items that had to be presented to our customer during the project On May 3 the project was fully completed and the complete prototype system was ready to be handed over to the customer Refer to Appendix L for a list of the web sites containing information we referenced for the project Deliverables Investigate seve
364. tore the computer in the projector room if we need to The cabinet builder gets back April 24 I expect we ll have the cabinet within a few days thereafter And Fred will probably connect the cable shortly thereafter or the same day I ll be sure he gets it ordered before hand I m sure I ve forgotten something If you think of it let me know Robyn Date Tue 30 Apr 2002 11 24 40 0600 From Robyn Harrison rharriso 9 cv3 cv nrao edu To Clint Janes cjanes 9 zia aoc NRAO EDU Sam Field lt sfield nmt edu gt Gerald Bivens gbivens nmt edu gt Brian lt brajala nmt edu gt Dr Shanechi lt shanechi ee nmt edu gt Subject kiosk 44 Okay This one was my screw up I evidently miscommunicated what I needed to the builder of the kiosk at least as far as when we needed it I just talked to him and he thought I was still waiting to have it approved At any rate he ll work on it this week and try to have it done first of next week What do you want to do Robyn KKK KKK K Robyn Harrison Education and Public Outreach National Radio Astronomy Observatory Socorro New Mexico 87801 505 835 7243 Date Tue 30 Apr 2002 13 16 28 0600 MDT From Clint Janes lt cjanes aoc nrao edu gt To brajala nmt edu cjanes 9 zia aoc NRAO EDU gbivens nmt edu rharriso 9 cv3 cv nrao edu sfield 9 nmt edu shanechi 9 ee nmt edu Subject Re kiosk The students plan to meet with Phil middle of next week We can take it out
365. tructor Dr Scott Teare for his witty and beneficial comments regarding to the Senior Design course itself We feel he has prepared us tremendously on what to expect once we graduate from college Thank you Table of Contents 1 1 Project Description EE 8 IPAE DG NP 8 1 5 Literature Revie Woecoaue cse tO aeu Yer EPI N SEE AFER EUH Yuseb oa we cu PESERN BEY yw 8 iE Moi mI neenaedd es 9 2 1 Technical Background i ee cote erret ttt rb E vea P ER YU YN seuss 10 2 1 1 Moisture Sensor orco hU ERE CLIE UE E EE DRE U p Edea 10 2 1 2 Light Sensor Real Time Clock ccceeceenccencceenceenseeeeeeenees 10 23 Microcontroller sedeo rer EX D E ora endi 11 2 1 4 5V Regulator eoe IR E rRNA REX VR FEDERN E EN ECPAR es 11 212d MOSFET SWIC hes ico ooa OE NR ne bade ER ERR E EE aces 11 3 1 Michiae FC Arpentets c ureseeenqet ire iba cage viva ee eileda persone VERS FRE STER CRSRPO PN M NETS 13 Sols PMISBUTG SEDISOES cd eee cute DOR vas ute Fa VER D ARR vd cesi 13 3 1 2 Plant Species Investigation ooi eI E Sear 18 3 1 3 Power Dist butones dies eee ole rye eere Lv re E PRRKR EE vex EP bad 19 32 Rocky GIndnni essee vx oor reb E PLE REOR EUR EePER NU EE R P EE RUM RES aed 21 32 1 Ener y Storage ise sete e p ta er E ries 21 3 2 2 Light Sensing Real Time Clock 0 0 cee cce eee e ence ence eee eeaeeeenens 25 3 2 0 Mitrocontrollet 5 5 2 oder AEE esta E Sea esL Peu degna eeu e 26 9 2 4 Prosta e core E UNS YeXR o FRE SANE ties ie
366. ttery Panel Need Voltage Sufficient Charging 1 Measure Moisture Level 2 Measure Switch Position Does Plant Need Watering Yes Go To Sleep Based on Available Data Water Plant Figure 8 Flowchart of controller program 38 Figure 9 Plant Species Moisture level Common name Thoroughly wet Dracaena African Violet Begonia Bromeliads Chinese Evergreen Chrysanthemum Ivy Evenly Moist Fern Fiddleleaf fig Gardenia Hibiscus Pine Palm Lily Philodendron Spineless Yucca Velvet Plant Drench then di Cactus Citrus Orchid Poinsettia T Umbrella Tree Wax Plant Description of terms for a few examples of common plant species above 1 Thoroughly wet low typically tolerant of humidity Daily watering generally required May stand in water for brief periods 2 Evenly moist medium Frequent watering required but must never stand in water Soil surface should always feel moist 3 Drench then dry high typically intolerant of humidity Soak root ball thoroughly then allow the soil to become fairly dry before watering again Do not allow the plant to wilt however 39 Fig 10 Sensor Test for water water Vv vs I Vy Vi l R R 100000 0 0002 0 00018 0 00016 0 00014 0 00012 oon C C ER WN 0 0001 0 73 10 00001 7300 0 78 11 0 00011 70
367. ual to the source voltages The model LMC6442 op amps were donated by National Semiconductor All of the components used in this circuit are rated to work at temperatures up to 85 C This temperature is equal to182 F which exceeds the internal temperature of 150 F required by the original proposal see Appendix I 24 3 4 Central Processor by Scott Dearie The central processor was implemented to transform the voltages sent for the temperature sensor and change it into a useful form The data was transformed by the processor into a binary stream that was sent to the transmitter The transmission protocol was an asynchronous stream that had to automatically synchronize the transmitter and the receiver station 3 4 Selection of the Microprocessor It was determined than any processor used in this project must meet a specific set of requirements These requirements included that the processor have a nominal current of 100 mA a start up current of less than 300 mA and an overall voltage of less than 5 volts These values were chosen such that the device would be low power It was also necessary that it have minimal noise impact on the transmission form the transmitter It was also determined that the PIC must have at least 2 A D ports 5 digital I O ports and 128 KB of EEPROM These specifications were selected to ensure that all the necessary functions of receiving and transmitting data could be achieved In addition to the power and ele
368. ualize it The following is a list of persons who all deserve our thanks and we ask at this point to forgive us if we have mistakenly forgotten anyone Thank you With NRAO Clint Janes Electronics Head Senior design projects implementation Robyn Harrison NRAO Education Liaison Marie Glendenning computer Sheila Reasner shipments Purchasing agent Pat Van Buskirk Java and network consultation Phil Dooley Intranet Weather Station Lothar Dahlmeyer Password control Richard Rupp delivery of kit to site helping with kit installation James Robnett consultation about timeserver With VLA Jon Spargo NRAO VLA Safety Officer Bob Broilo VLA Electrical Division Head Guy Stanzione Grounds Projects Jaime Montero Lead Electrician Other Key Personnel Jim Shepherd Outsourced Carpenter for kiosk Steve Wasson programming consultation Physics Department chair Dr Minschwaner permission for SRT antenna Dr Hasan M Shanechi Faculty Project Advisor and laboratory space Dr Scott W Teare Senior Project Coordinator Dan Jones NMT Architect Physical Plant Consultation for installation Chris Kaiser contact at MassWorks 8401 Washington Pl N E Albuquerque New Mexico 87113 Chris Pauli EE Dept Network Administrator Carrol Teel EE Dept Secretary Pete Martinez communications consultation Dr Dan Klinglesmith NMT Physics Astronomy Basic Astronomy Science Information consultation ii Table of Contents Fels m D E i
369. ucts of this project are twofold first this thesis documenting our design and reporting on the current state of the RDPP and second a software and hardware package containing test procedures and tools for verification of future design revisions PROVIDED MATERIALS amp EQUIPMENT When we began this project Dr Donohoe provided us with the hardware design files for the RDPP written in VHDL We were also provided with an Altera Excalibur board containing the APEX20K200EFC Programmable Chip and a computer with the Altera Quartus II software to program the Altera chip 2 1 2 2 CHAPTER 2 TECHNICAL BACKGROUND amp DESIGN OVERVIEW INTRODUCTION This chapter will provide a brief technical background to this project First a basic description of the Reconfigurable Data Path Processor will be given This will include schematics and definitions of components Next an overview of the Altera Excalibur board and Quartus II software will be given We ll then provide a brief introduction to the theory and practice of functional testing of devices Finally we will give an overview of our design for this project THE RECONFIGURABLE DATA PATH PROCESSOR The Reconfigurable Data Path Processor RDPP is a custom processor being developed by NASA s Institute for Advanced Microelectronics It is being developed for high throughput processing of streaming data aboard spacecraft The current RDPP design contains 16 reconfigurable Processing Elements
370. uirements In order to provide power to a mechanism with a thirty year shelf life the suggested power supply was a high temperature Peltier junction device To ensure the survivability of the electronic components of the device a highly insulating material must be used as protection The long term goals are to develop a cost effective device that will be small durable and lightweight It will conform to the ordinance so that it does not affect the aerodynamics of the ordinance The time and funding constraints imposed a change on this facet of the long term project The redefined goals were as follows Study the feasibility of using a thermopile to provide power for the device Develop a way to insulate the electronics of this device from the heat of the fire while maintaining a maximum temperature of 150 F Select a processor that is able to compile the necessary data and store it in a format that cannot be destroyed over time Design a transmission system that can communicate at 2 39 2 45 GHz These requirements are shown in Table 1 3 below Study the feasibility of using a thermopile to provide power for the device Find a way to insulate the electronics of this device from the heat of the fire This insulation had to be designed to keep the temperature of the electronics at or below 150 F Select a device capable of processing the necessary data This device must also be capable of storing inventory data in memory The device was to
371. unlight respectively If a Real Time Clock RTC is to be implemented we should use the DS1306 SPI RTC This requires the SPI communication to be set up on the microcontroller After attaching a 32KHz clock to the X1 input of the RTC as well as the various power and ground lines the RTC can then be programmed It is programmed Address first and Data 10 second The device comes with the appropriate data sheet though it is a little vague as to how to program the device 2 1 3 Microcontroller The controller we chose to go with was the 68HC12912EVB which already contains all the necessary hardware for the entire controller to be programmed If we chose to buy the controller chip separate then a board will be required to house the necessary components as well as the controller chip Then utilizing a compiler program like the COSMIC Compiler IDEA CPUI2 this will allow us to generate and then compile our programs Then all is needed is a file transfer application An HC12 HyperTerminal or even the WIN 3 11 Terminal will do in order to transfer the s19 file that programs the controller 2 1 4 5V Regulator The 5V regulator is a chip which receives an input voltage greater than 5V With an input greater than 5V it then outputs a steady 5V The chip we used was a MAX667 which is a low dropout positive linear voltage regulator that supplies 250mA of output current The regulator was placed between the battery and the circuitry The use of t
372. up in the init module routine The first thing we need to do is to clear the A D interrupt and reset the A D FIFO depth to zero by writing a 0x11 to the Command Register 0x280 for the Prometheus Data Acquisition Circuitry We then reset the FIFO Threshold to 1 by writing a 0x01 to the FIFO Threshold Register 0x285 and set which channel to scan by writing to the A D Channel Register 0x282 Next we trigger the A D Scan by writing a 0x80 to the 35 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Command Register 0x280 and wait for the scan to finish by waiting for bit 7 of the Analog Input Status Register 0x283 to go low Finally we must read the value of the A D conversion a 16 bit value by reading the least significant byte from address 0x280 and the most significant byte from address 0x281 To make the code easier to read we wrote macros for each of these steps these macros are defined in prometheus h found in Appendix C To read Digital I O values bump sensors and the white line sensor we simply read from the appropriate DIO register 0x288 for Port A and 0x289 for Port B To write to the Digital I O the fan we write a 1 to the appropriate bit of Port C 0x284 The Prometheus has only one interruptible Digital I O line The documentation of this DIO Interrupts implies that an interrupt could be generated on the falling edge of the input line We took this to mean that we could setup this interrupt to be generated wh
373. up into 4 bit packets and passed into the multiplexer The counter determines which 4 bit piece is placed on the output and sent to the computer The TxRx clock increments the counter on the falling edge to select all of the 4 bit pieces sequentially This whole block is enabled when the transmission state bits are both high 14 corresponding to the transmission state in which data is being sent from the PE to the computer After implementing and testing this design we upgraded it to a more efficient design We took out the counters and demultiplexers that were used for routing incoming data and connected the input data to all of the temporary 4 bit registers We replaced the counters with a state machine that would drive a single enable line at a time for all of these registers On the falling edge of TxRx clock the next enable line was driven high for the next register Figure 8 below shows a block diagram of this design data in 0 3 Register Enable Control Machine TxRx state 1 0 Figure 8 Final Design to Route the Incoming Data We changed our design to send data back to the computer in a similar manner The counter was removed and replaced with a state machine that drove singe select lines We changed the multiplexer so that it took in individual lines to select which input to select rather than a string of bits encoding which input to select For the Altera design files of our interface refer to our project CD 15
374. ure which causes a resistance change in a ceramic semiconductor The resistance drops nonlinearly with temperature rise These devices are very accurate The limited temperature range typically between 40 C and 150 C was not enough for the required temperature range of this project It was necessary to choose between these of temperature sensors to find the one that best suits the needs of this project Table 3 3 lists the pros and cons of the different types of temperature sensors considered 21 Device PROS CONS Thermocouples Wide temperature range Not as accurate as others Inexpensive Rugged Not as stable as others RTD s Stable Accurate Not as rugged as other sensors Fairly wide temperature range Subject to self heating Infrared Radiators Good at measuring large surfaces Not as accurate as other sensors Do not require physical contact Sensitive to surface radiation Range includes high temperatures Expensive Thermistors Very accurate Limited temperature range Subject to self heating Highly nonlinear Table 3 3 Comparison of Temperature Sensors Using the information contained in Table 3 3 the final choice was to use a K type thermocouple The thermocouple had a temperature range that best met the needs of this project When this consideration was combined with cost effectiveness K type thermocouples were the only choice Originally there was concern that the large temperature range of a
375. urning This in a sense would work exactly like the normal mechanical turn signal does Final Status The shaft on the switch needs to be long enough so that it can move easily with the turn signal This is what created the problem because we couldn t find a toggle switch that had a long enough shaft or with a sensitive enough switching mechanism that moves smoothly and keeps the base of the switch immobile We tried using a toggle switch that we had that didn t have a very long shaft but it failed to move smoothly with the turn signal shaft 3 2 2 Push Button Switch Description This is a small discrete general purpose push button switch This sensor is a button that outputs 5 volts when pushed Application This application would involve two general purpose push button switches with one each being placed on the top and bottom of the shaft When the driver operated the turn signal he would have to make sure his hand was placed on the appropriate button The hope was that this sequence would become routine to the driver and not a hassle although anything extra the driver has to do could become a distraction so this was thought to pose a problem from the beginning Final Status It was later discovered through testing that in order for this sensor to work the driver would have to hold the button throughout the entire turning or process or hit the button again when the turn is complete This sensor wouldn t work since we could not expect
376. vel of the targeted end user and contribute an extra cost to building the cluster The NFS root method is also probably beyond the target skill level as it involves the creation of root file system images and building custom Linux kernels that support NFS root file systems The local boot and file system method has several other attractive features Surplus or off the shelf computers usually come equipped with a functional hard drive so this resource should not just be discarded This allows an almost default Linux install to the hard drive with only slight modifications made for mounting the home partition across NFS and user authentication with NIS Installing Linux on each node has the added bonus of the install having the opportunity to detect and configure the different types of hardware that are present in the different computers that will be used since not all the nodes are guaranteed to be identical Surplus computers also tend to have limited memory RAM and in such cases the hard drive provides valuable swap space disk based virtual memory to compensate for this limitation 2 2 Overview of the NAS MG benchmark used to evaluate the cluster The NASA Advanced Supercomputing NAS Parallel Benchmarks are a collection of eight benchmarks used to evaluate the performance of highly parallel supercomputers http www nas nasa gov Research Software swdescription html NPB Among these eight benchmarks is the Multigrid Array MG ben
377. very simplified case These calculations show how many modules are needed to meet the power requirements for a dT of 100C This dT is used if we assume a 100 degree difference between the Devices internal and external temperatures This would most likely be the dT at the power on temperature A module consists of 127 junctions The calculations show that the 3 modules are required to meet the power requirements Other things in consideration are the internal resistance of the modules versus the load resistance They should match or be fairly close for maximum power output There is also a calculation for overall efficiency at dT100 I think the short paragraph explains that 8 modules are needed for a dT of 50 and that these dTs are for relatively low temperatures A dt of 50C from 800C to 750C is still dT of 50C but there is much more heat flow than the same dT from 100C to 150C The equations below are simplified to allow for an ideal case Average Temperature Tav Th Tc 2 Th hot side temp Tc cold side temp Th 403K Tc 303K so Tav 353k Load resistance is R V I 5V 1A 5Ohms Max Power Pm Sm DT 2 4 Rm Where Sm is the Seebeck coefficient 0 05544V K DT is the temperature difference between Th and Tc 100K Rm is the resistance of the module 3 00994Ohm The result is that the module will produce 2 48 watts of power The minimum number of modules needed is Po Pm which for our case equals 2 modules The maximum generating efficiency is when
378. ware Several Data FIFOs from which the user reads the data the threads have received from the hardware and a Command FIFO that is used by the module to pass user space commands to the appropriate thread The complete code for this driver can be found in Appendix C and uses five subroutines init module cleanup module my handler intr handler and thread code This driver is designed to be a loadable kernel module When a module is loaded into kernel memory the code that is executed to initialize the module is init module This subroutine sets up any hardware the module needs to use memory spaces pointers to functions etc Our module accesses our Breakout Board and the Prometheus I O module using threads and RT FIFOs First we need to destroy the FIFOs we will be using in case they are already in use by another process We then create the FIFOs to get the data to and from the user space Behavior Control algorithm as well as the threads to gather all of the data 31 RTLinux On A Small Mobile Robot Bitsoi Lamb Willems Next we setup the Prometheus I O we do not need to setup the Breakout Board as it is simple memory mapped registers We enable Digital interrupts and disable Analog I O and Timer interrupts as well as DMA operations We then configure the Analog I O module by setting the gain to a unipolar 0 5V as well as setting the FIFO depth to one Next we configure the Digital I O module by setting Ports A amp B to be input and th
379. well as it may vary with different voltages applied to dirt Voltage varies with a set resistance range The dirt s resistance changes with applied voltage because there were significant changes in the resistance readings when switching ranges Our advisor Dr Bruder threw in a few ideas One was to use two sensors and taking the average of the two readings to receive a better resolution I decided to go with one because as long as it s placed near the actual roots that is precise enough to receive 16 accurate readings He also thinks I should derive measurements at specific water levels hysteresis analysis even utilizing a measuring cup I decided to just measure wet dirt and dry dirt and the readings produced constant results I decided to provide graphs to search for consistencies Figs 10 13 I dipped the electrodes copper probes into dry soil wet soil and water at a specific depth and distance apart I got several voltage and current readings and noticed specific patterns Of course I had to fiddle with several current sources since several were malfunctioned Variables in this test include length of submerged portions of electrodes at certain depth independent voltage dependent current measured value resistance voltage current dependent Furthermore I took three separate points current on a current vs resistance graph on these graphs There are nine parameters dry dirt moist dirt water at 1mA I4mA and 18mA T
380. xhibit relatively high ESR they are not recommended for ripple absorption in DC power supply applications Supercapacitors or electrochemical double layer capacitors EDLC take advantage of the charge stored in the electrochemical double layer and provide extremely high capacities of more than 1000 farads These devices have applications in computer power backup power electronics electric vehicles and space flight technology However power and energy demands of these applications vary significantly The Supercapacitors do have its drawbacks It has a low energy density energy to weight ratio its varying voltage supply is proportional to the stored energy and it has high material costs However it does have its benefits Little maintenance is required it has a long life it has a high cycle efficiency it has high power and power density it does not use toxic or explosive fluids and it has a high recharge and discharge rates After evaluating all batteries and capacitors I decided to go with a Sealed Lead Acid Battery type The battery that I got has a 6V 2 8 A Hr capacity Besides being readily available they are more forgiving to overcharging and undercharging which means that our charge controller need not have to be complex in function They also meet all of the criteria stated previously 24 From investigation the use of supercapacitors would not be a cost effective alternative because they do not have high enough energy densi
381. ype circuit was developed and preliminary testing was done on all the subsystems These designs were then transferred onto a professionally made printed circuit board which was also designed by the group The sensors which are used to determine if the brake or turn signals are being activated were then integrated into the complete system The overall results of the project were a success and the components have been tested inside vehicles during real world road testing Acknowledgements We would like to thank our sponsors Jason Vandehey and Jeff Duyck for the opportunity that they have given us This opportunity is one that has not only benefited us this semester but will also be a great help in our success as electrical engineers We would also like to thank Jason for his continued support and knowledge that he shared with us We would like to give a special thanks to the faculty of the Electrical Engineering Department at New Mexico Tech for the continued support and knowledge Thanks to Dr William Rison for his support and contribution of the programmable chips and equipment Special thanks to our faculty advisor Dr Ronald Thomas for his guidance and support along with all the extra time he set aside for us out of his busy schedule We would also like to thank Dr Richard Scott and Dr Robert Bond for the support and knowledge about RF transmission We would also like to offer our appreciation for all the support and guidance that Dr S

Download Pdf Manuals

Related Search

Related Contents

Multicontroller User Manual  SBS Disney  Fujitsu ESPRIMO Q920  KOHLER K-2057-96 Installation Guide  Software Modelling and Architecture: Exercises  Région - Journal de la Région Alsace  給湯暖房用熱源機取扱説明書 保証書付  新機能および補足説明について（ソフトウェア Ver.1.40）  USER`S MANUAL - American Panel  MODE D`EMPLOI  

Copyright © All rights reserved. Failed to retrieve file