Home

Final Report - Calvin College

1. Motor Motor Controller 1 Controller 2 Voltage Voltage ibi Current Current UM Sensors Sensors Motor 1 cr TIED Motor 2 MDL BDC24 MDL BDC24 Figure 7 Block Diagram of Delivery Design Hardware Topology This topology 1s necessary because of the restrictions imposed by the motor controllers Since they can only run one motor each it was necessary to have two motor controllers and then one main controller 1 e the ECU to synchronize the actions of each motor controller The ECU is necessary to provide the I O interfaces to the user and to coordinate the responses of each of the motor controllers independently receiving status updates from each motor controller and sending commands to each motor controller Team 3 Achieving Mobility Final Report Page 23 of 160 May 11 2011 based on the sampled data Overall a third MCU is not needed to alleviate the computation limitations of the motor controllers but rather it is needed to provide the functionality that 1s required It should also be noted that the voltage current and temperature sensors as seen in Figure 7 are built into the motor controllers 4 1 4 Microcontrollers The microcontroller MCU is the central computational unit for the system which must perform all the necessary real time calculations as well as provide a platform on which to run software The communication between sensors and other peripherals 1s coordinated
2. esses 51 Algorithmic Flowchart of Speed Check Function ooccccccncncncncncnnnonononononnnnnnnnnnnnnnnnnnnnnonnnnnnnnos 22 Drawing of LED Screen Used Tor DeDUSPID9 a dd 54 Sample UART Output of Debugging Data essere 55 Example Graph ot Motor Status Dildo 56 Proposed Production Design Hardware Topology cccccccccccccncnnnononononancnnnnncnnnnnnnnnnnnnononnnnnnos 58 Schematic Diagram of Simple H Bridge Cilia iia 58 Mechanical D sien Breakdown acid 60 Preliminary design used for base case analySIS oooccccccnnononcncnnnonononnannnnnnnnncnnnnnnnnnnnonononnnnns 65 Second design used for overall height and length purposes essen 65 Final design of the stroller frame seien nnne eene nnne nnn nnns 66 CAD Drawing of Frame Weldt ett ipie ii 67 TIC DOW IS ons mare an ene O OD e Do Nue DUM MELDE CR OL DEN 68 Pinal Desin of the Storage Ate di d SE er trae eu de i oput 72 Main Beddin Cus ON souris A 73 ORE riter A ER 73 FRG SUE ATTN A A A A 74 Parts OF a PMDE mo dd 76 Comparison of PMDC and SWDC motors 8 tette reete teen 76 Parts of a BLDC motor uenenum EN RAM UNDA PME IDE MU TI Diagram of the motors engaged and disengaged ooooonnnnnnncccnnnnnnonnnnnononnonnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnns 78 Test setup for donated motor functionNallty oooooonnnncccccnnnnnnnnnnnnnononononannnnnnnnnnnnnnnnnnnnnnnnnnnnnnos 81 Invacare Nutron R51 motors used for
3. llLLe E ET DLL LIII u m du a M A cae Lee mT LI ANNA NIN NANA NN CN NN ATAN HM AK INL WIAA AAA ALK N MALO AA AA LEN 1 M AN 0 100 1000 CURRENT IN AMPERES Figure 15 Average Time for Breaking of Fuse 4 1 8 3 Heat Dissipation Making sure the components do not overheat 1s an issue in the power regulation circuit The LM317 voltage regulator is the primary component that allocates the appropriate voltages to each electrical Team 3 Achieving Mobility Final Report Page 32 of 160 May 11 2011 device The LM317 is a linear device and it burns off the extra power as heat The first voltage regulator that transforms 24 V to 12 V is at the highest risk of overheating due to the large change in voltage and also because there is about 500 mA going through it The Mathcad calculations show that better heat dissipation is required for the particular voltage regulator These calculations are shown in Figure 16 Heat Dissipation of Voltage Regulator Specifications The package is TO 220 T Jis junction temperature which is the temperature T J mar 125 C of the semiconductor in the voltage regulator T Ais ambient temperature which is the temperature of the surrounding atmosphere B8 JA is the maximum allowable value for the junction to ambient thermal resistance Meaured Values Vin 24V Vout 12V IL 526mA Land _G were determined using an ammet
4. occccccnncccnnccnnnnnnonnnnonnnnnnnnnnnnnnnononononononnnnos 100 Mounting or PCB eneasiti9 oui oie ene quet a od rotes ei quot dices tein r 101 vii Figure 79 Comparison of LCD Screen mounting alternatives nccicicinicicicicoccncncnnononononanininincncncncnoo 102 Figure 80 Prototype of LCD mount cccccccccccccssseesensssssssseeseeceeeeccccccecececsassseeeeeeeeeeeecceeeseecesacsaseseees 103 Figure ol Final LED quot Sy SECT senres naie Erana E EE n siai 103 Figure 82 Clamp style MOUSE AiE E EEEE E EEEN 104 Figure 83 Final motor mounting System oooonnnnnnnnnnnnnnnnnnnnnnnnnannnnnn aran 104 Figure 84 Design of strap in arm rest with touch button attached sess 105 Figure 85 Three strap in arm rest prototypes that were built and used to test on the customer 106 Figure 86 Magnetic strap in arm rest storage device cece ec eeeeeeeeeeeseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeees 106 Figure 67 Acceleration Testing RESTIS css erecci n 110 Figure 8 Velocity Control Test CSCS serais Eran ibas Alias PP aA E EEEN o REDI 111 Figure 89 LM317 Temperature as a Function of TIME oooonnnnnnnnnnnncncnnnonononnnnnnnnncnnnnnnnnnnnnnononnnnnnnnannnnnnnnss 112 Figure 90 Calculations for Speed TeStUllig iesu m retirer pete gave aola a aaa 115 Figure 91 Motor Deceleration from 2 5 mph as a Function of Time seen 116 Figure 92 Competitive models in the market
5. won hizez IbtAin z7 1194 06 1077 09 9550 13 2393 167 T 150 203 5059 24 qa 25 2595 313 2090 99 1201 56 4 422514 Page 8 of 8
6. 2 13 0 19 5 10 100 Project Requirements 3 11 15 14 0 19 8 10 100 funcionalRequremens 3 2 3 14 0 13 8 10 1005 _ Electrical Requirements 3 2 4 faao 11 8 10 100 Mechanical BULL 3 11 4 10 100 SafetyRequirements 3 2 3 awso 11 8 10 100 2 100 5 1 0 0 W W e THH 0 W JIU W W W N NM ii Uy N Tr UJ Ul MajorDesignDecisions 3 5 5 11 5 10 11 9 10 100 Userinterface 3 2 1 115 0 11 9 10 100 UserConroMechansm 3 1 1 11 5 10 11 9 10 100 Smeue 3 2 f 3 15 0 19 9 10 100 Systembesign 13 90 98 10 25 10 11 10 10 100 Software 4 12 9 115 0 11 10 10 100 Micropracessor 12 15 12 10 25 10 11 9 10 100 esos os 2 f S omeno 11 0 0 100 Batey 6 3 taps 11 8 10 100 4CDSren 1 2 2 11 8 10 11 8 10 100 mera 1 2 2 10 8 10 19 8 10 100 MetrContol 6 3 2 1025 0 19 1 10 100 Mame a f 10 16 11 550 11 9 10 100 W W W N je IN W IN IN IUU IN Team 3 Achieving Mobility Final Report Page 148 of 160 May 11 2011 4 3 11 8 10 11 8 10 100 Storage PmjctCots J 3 5 4 11 8 10 11 10 10 100 Market Feasibility 3 1 2 11 840 11 0 00 1006 ProjectMamagement 3 u 15 1110 0 11 12 10
7. 5 00 Nuts bolts washers etc Labor 50 00 Mounting connecting to motor and controller PCB 1 hrs 50 hr 51525 IN Emergency 13 00 Toggle Switch Chrome Servalite 34 212U Modern Hardware Switch Team 3 Achieving Mobility Final Report Page 90 of 160 May 11 2011 Table 24 Delivery Design Costs for the Braking Systems liem Kentwood Cycling and Fitness 19 99 per brake set Hand lever Electronic 0 00 Cost included in controller and motor costs Brakes Emergency 0 00 Cost included in motor costs Brakes Emergency 13 00 Toggle Switch Chrome Servalite 34 212U Modern Hardware Switch Mounting 0 00 Nuts bolts washers free at metal shop Labor 000 Contingency 5 30 Total 828 5 7 Encasings The encasings provide a protective enclosure to house the critical electronic components of the device motor controller PCB ECU PCB power regulation PCB and batteries The motors are being outsourced and therefore its electronic components already include a proper encasing and no further design 1s required 5 7 1 Requirements 5 7 1 1 Protection The encasings must be water resistant meaning the electronic components function properly when they are sprayed or splashed on with water The encasings must also provide protection from dirt dust and other foreign particles 5 7 1 2 Heat Dissipation The encasings must be designed to provide adequate heat dissipation from the electronic components
8. If the motor controller data is in the ECU s message queue 1 e if the data is available then the fault detection and speed correction can occur It takes less than 1 ms for a status response from both motor controllers so a periodic sampling of 1 ms is adequate Also servicing the velocity controller on a period of 1 ms allows for immediate response from the motors in case of diagnostic test failure such as a destructive current spike The response time of the motors is on the order of 10 50 ms so this rate 1s considerably fast enough to service the motor voltages in the Speed Check function A flowchart of the Speed Check function 1s shown below in Figure 40 This function takes the current samples from the motor controllers and makes the appropriate voltage adjustments based on these readings Start Speed Check No No y Set voltage back to full speed Yes Yes No Voltage lt 10 V Voltage gt 6 V Yes Yes No Change 20 iteration 20 iteration No 1 5 seconds Yes Yes Increase voltage Decrease voltage to motors by 0 2 V to motors by 0 2 V Figure 40 Algorithmic Flowchart of Speed Check Function The upper threshold limit of 20 A and lower threshold limit of 3 A were determined by testing results The team decided that if any motor draws over 15 A of current for over 1 5 seconds it must be Team 3 Achieving Mobility Final Report Page
9. Sen dd 19 a E 9 H ca Q Q 70 90 110 130 150 Forced Convection Coefficient W m 2 K Figure 68 Resulting temperature of PCB due to forced convection The limit temperature that each PCB will be designed for is 50 C Even though the limit temperature for each PCB depends on its components this temperature was chosen as an upper limit because it provides a safe operating temperature that will not melt components and ensure long life For the ECU PCB natural convection will be enough to keep it below the desired 50 C however the motor controller PCB and the power regulation PCB will both require forced convection The motor controller PCBs were both donated and already include a built in fan system so no further design is required A cooling system is required to be designed for the power regulation PCB Detailed testing and design for this PCB is described in 4 1 8 5 7 3 Design 5 7 3 1 Production Design Table 28 shows the final selection for material sizing and heat dissipation for each component encasing Table 28 Design Selection for Electronic Component Encasings Encasin Component Encasing nnd Heat Dissipation B Material rin Method Motor Controller ECU Flame 5x825x4 Forced convection Power Regulation Retardant ABS l Flame Team 3 Achieving Mobility Final Report Page 94 of 160 May 11 2011 Plastic encasings will be purchased rather than manufactured because it is more cost effective for a
10. The selection of a battery type was necessary for the design of the device The different pros and cons of both sealed lead acid batteries and lithium ion batteries are shown below in Table 7 Team 3 Achieving Mobility Final Report Page 28 of 160 May 11 2011 Table 7 Assessment of Battery Types Battery Specifications Manufacturer A ES Pros Cons Type Dollars Proven to work 804 inches cubed effectively weighs 77 6 Readily pounds available Cost Weighs 16 5 Piedad ae VOS PingBattery 495 00 p documentation Ion 30 ampere hours 288 inches l cubed SSH This table illustrates the overall evaluation of the two different battery types By clearly stating the 24 Volts Interstate 55 ampere hours Batteries benefits and drawbacks of each type a decision matrix was constructed in order to help determine the best fit battery for the power supply of the device The most important factor when considering the battery type was whether the battery would work properly Sealed lead acid batteries are almost exclusively used in wheelchairs so it was obvious that this type of battery worked well Lithium ion on the other hand is not highly used in wheelchairs because this type of battery 1s usually used for smaller devices that do not need as much power Size and weight are also very important because storage is needed for the medical equipment of the customer Also a lighter battery would cause less strain on an attendant pushing the strol
11. Voltage Regulators El 24012 I adj 501A R23 11 y V_out usi 3 I adj R2 LCD Screen voltage LCD 12V voltage LCD usvfa adj R2 Find R2 2 044 x 10 12 Camera voltage camera 9V Given voltage_camera usvfa 3 I adjRE2 Find R2 1474 x 10 Host Controller voltage hostController 5V voltage hostController usi I adj R2 Find R2 713 1542 Team 3 Achieving Mobility Final Report Page 2 of 8 May 11 2011 12 6 Appendix F PCB Heat Dissipation Calculation Parameters A 025 Average surface area of PCB Tir 298K Average temperatureof air flowing over PCB Natural Convection Coefficient Forced Convection Coefficient 1 92 35 17 1223 192 W W f i 1 gi1 mK nK W nat 9 2 Mor d 2 m K m K nat Anat 1 f nfor Pfor 4 tg lis s Average range of Average range of convection coefficient for convection coefficient for forced air free flowing air Engineering Toolbox com Engineering Toolbox com Team 3 Achieving Mobility Final Report May 11 2011 Motor Controller MC PCB Power Inc 10amp Vme 10V Wi Vine me W inc 100W Natural Convection T iT mc nat alr i BITE Forced Convection T LI 44 mc for air J Mor Electronic Control Unit ECU Power loo 77 3 4amp Veci 24V Wecu Vecu lecu Wecu 81 6W Natural Convection T u Wecu T ecu nat A tair nat 1 Forced Convection Wecu T
12. Weighted Mounting 8 ML I1 Performance all 42 weather a E m a ee RR For the attendant controlled braking system a hand lever with rim brakes was used in both the production and delivery design A hand lever is inexpensive easy to mount and most accessible by the attendant Team 3 Achieving Mobility Final Report Page 87 of 160 May 11 2011 The rim brake was chosen because it 1s inexpensive easy to mount and performs well enough for its low amount of usage Figure 64 shows a picture of the purchased rim brakes and hand levers for the final product Figure 64 Rim brakes and hand lever used in final product 5 6 3 3 Parking Brake The parking brake 1s designed to keep the wheelchair in place when the motors are either engaged or disengaged When the motors are engaged the automatic braking mechanism described above within the motors acts as a parking brake but when the motors are disengaged there needs to be a braking system that locks the wheels in place Two different types of parking brakes were considered wheel locks and axle locks The advantages and disadvantages of both systems are shown in Table 22 Table 22 Comparison of Parking Brake Alternatives Parking Brake Advantages Disadvantages Alternative e Simple e Possibility of Slip Wheel Lock Less expensive e Wears tires out e Mounts easily e Effective lock mechanism e Requires axle connecting wheels Axle Lock e Aesthetic e Expensive
13. 100 WerDwson 3 2 3 11 10 10 11 12 10 100 Schedule 3 5 7 11 10 10 11 12 10 100 References 3 2 2 11 810 1110 0 100 Appendices 3 f 10 6 11 810 11 10 10 1006 FimalRepore 15 60 100 4 3 11 4 20 11 100 Budget 3 3 3 48 12 9 10 720 12 10 10 1006 Electrical S 140 554 739 20 4 10 4 10 11 100 Microcontroller 45 66 111 10 8 10 12 29 10 100 Processorselecti n 6 1 3 11 8 10 11 15 10 100 dmtegrtionwih Motos 8 20 so 12 1 10 12 10 10 100 Measure Display Battery Charge 5 12 8 2z zi1i 2 11 11 100 AmalyzeOptions 8 8 10 10 25 10 11 3 10 100 W W UY UY W W UY W o Team 3 Achieving Mobility Final Report Page 149 of 160 May 11 2011 Obtain Charging System 5 1 1 11 29 10 12 3 10 100 Micracomtraller 6 3 30 is 1 10 11 100 Mor 6 3 30 12 610 12 13 10 100 LCDScreen Camera 68 10 11 19 22 10 2 23 11 100 integrate intosystem 16 6 5 z n 2 23 11 100 Motor Controller Circuit 92 66 105 10 18 10 2 22 11 100 DesignMotorControPCB 16 35 60 1 31 11 2 21 11 1006 Determine Power Requirements 6 4 5 10 25 10 11 1 10 1006 Integration withmcu 16 15 20 zA 1 2 22 11 1006 115 23 Use
14. 5 3 1 2 Strength The second main requirement is that the storage shelves are strong enough to support the weight of all components placed on them The weight of the lead acid batteries analyzed 1s approximately 76 pounds which accounts for the majority of the weight on the stroller The weight of the customer s suction device and feeding bag are 8 pounds and 4 pounds respectively Therefore the storage compartment will need be strong enough to support at least 88 pounds The final design however was made to hold at least double this amount in case different objects of larger weight are ever placed in the storage compartments 5 3 2 Alternatives Most of the alternatives for storage shelves are very similar to the ones listed above for the frame therefore the choices made for the frame will influence the choices made for the storage shelves 5 3 2 1 Material There are several different materials that can be used for the storage shelves The most common are 304 stainless steel 1020 HRS and 6063 Aluminum The advantages and disadvantages for these materials are listed above in Table 9 Since aluminum was already chosen for the frame aluminum will also be used for the storage shelves The storage shelves will be on the bottom of the stroller therefore they are the most susceptible to corrosion making aluminum the best choice The customer s current stroller has removable storage shelves that are clamped to the frame however after speaking
15. 59 60 Table 39 Comparison of Performance of competitive models Pride Mobilit Achievi Performance Invacare FDX E Permobil Corpus t Dn ne Quantum Mobility Speed Type Top Speed 5 0 mph 5 0 mph 6 5 29 A mp Incline Capability c9 39 User Weight 300 Ibs 300 Ibs 300 lbs 250 lbs Capacity nao 320 Ibs 224 Ibs 313 lbs 200 lbs Weight The emphasis in our production design will be on special features that do not exist in the market currently Table 40 compares the competition s main features with our product s main features Invacare FDX Pride Mobility Quantum and Permobil Corpus have about the same basic features and are combined here for comparative purposes Table 40 Comparison of features in competitive models Adjustable seating width depth height tilt Lay down bed fixed 170 with adjustable length Several drive control options 8 in Joystick Simple touch button control Single switch Fixed speed Wafer Board Manual drive capabilities Head control Easy access to batteries Chin control Attendant controlled brake Sip N Puff Camera LCD vision system drive while lying Variable speed options down Manual drive capabilities Independent wheel suspension Easy access to batteries As the table shows the competition has some features that our production design 1s lacking The main feature that sets us apart from the competition is the ability for the user to drive the w
16. 73 74 75 76 77 78 79 80 8l 82 How often does Spina Bifda occur Spina Bifda Association 2009 Web 4 Dec 2010 lt http www spinabifidaassociation org site c l KWL7PLLrF b 2700313 k 28B2 How_Often_Do es_Spina_Bifida_Occur htm The Wheeled Mobility Market WheelchairNet n d Web 4 Dec 2010 lt http www wheelchairnet org wcn_wcu Research StakeholderDocs PDFs wpf_wheelchairsum mary_doc pdf gt Warehouses for lease LoopNet N p 2010 Web 29 Nov 2010 lt http www loopnet com Warehouses For Lease gt Magazine and Tabloid Advertising Rates Nationwide Advertising N p n d Web 1 Dec 2010 lt http www nationwideadvertising com newmagadrat html gt Amyotrophic lateral sclerosis Google Health A D A M 2010 Web 30 Nov 2010 lt https health google com health ref Amyotrophic lateral sclerosis gt Cerebral Palsy March of Dimes N p Dec 2007 Web 30 Nov 2010 lt http www marchofdimes com Baby birthdefects_cerebralpalsy html types gt Spinal Cord Injury Facts and Figures at a Glance The National SCI Statistical Center N p Feb 2010 Web 30 Nov 2010 lt https www nscisc uab edu public_content pdf Facts 20and 20Figures 20at 20a 20Glanc e 202010 pdf gt Spinal Muscular Atrophy SMA Fact Sheet FightSMA N p 2010 Web 30 Nov 2010 http www fightsma org index php fact sheet Medema Bob Personal interview 1 Dec 2010 Rile
17. 8 2 10 18 10 10 25 10 100 Visitfamilytoseebed 3 6 6 1115 0 11 17 10 100 choose mounting material 4 8 2 11 18 10 11 23 10 100 Contact bed manufacturers for estimate 11 24 10 11 30 10 10096 4 3 40 20 64 32 4 5 4 5 5 N O 74 Asemby oa 32 f 1 m an 100 CircuitBoards 86 32 38 1 3 11 3 14 11 100 4 ke J 1 8 6 uzm 2 22 11 100 Momtng Tt 8 6 vzm 2 23 11 100 12 25 3 10 11 5 5 11 100 IP re I N 6 S 11 22 10 11 26 10 1005 6 15 j i129 0 12 3 10 100 52 16 a 6 110 11 1 10 100 _ 6 11 2 10 11 9 10 1006 2 Power Torque Calculations Size motor Team 3 Achieving Mobility Final Report Page 151 of 160 May 11 2011 Mounting 3 8 12 6 10 12 8 10 100 Lamm 0 2 4 3e xum lsm aw Broking 0 0 95 34 4 1 5 0 3 17 11 100 HandBrae z 20 28 1 50 3 17 11 100 Aemaivs 1 8 f 6 11 5 0 11 5 10 100 Momm 1 4 8 aasma asm 100 Parking Brake 26 14 14 1 5 00 3 16 11 100 Afemaivs 1 a f 2 11 5 00 19 5 10 100 Seetn J 1 2 1 11 810 19 8 10 100 Momm 1 24 3 asm 3SAS M 100 ooo Asemby a 4 8 Sanwa mn 100 TouchButton 8 n 17 Ms Aii 7 11 1006 Mechanism 5 6 5 Ma
18. After being loaded in the van and locking the stroller in place there were once again no issues 6 2 2 4 Motor Testing This test determined the performance of the motors as a part of the entire system The motors are required to move the maximum rated weight 200 pounds up the maximum rated incline 10 For this test the wheelchair was loaded with 230 pounds and driven up a 30 gravel incline using the touch button system Team 3 Achieving Mobility Final Report Page 119 of 160 May 11 2011 The wheelchair successfully travelled up the slope with no slippage or stalling but a slight reduction in speed therefore passing this test 6 2 2 5 Mechanical Brake Testing The mechanical brake test involved two different parts The first part tested the overall performance of the attendant controlled brakes as a part of the entire system These brakes are required to stop the wheelchair in a maximum of 2 feet when traveling at the top speed of 3 5 mph For this test the wheelchair was loaded with 230 pounds and manually pushed down a 30 gravel decline The hand brakes were then applied when the rear wheels reached a line marked on the gravel When the wheelchair came to a complete stop a measuring tape was used to determine the distance between the line and the rear wheels Three separate trials were conducted with an average stopping distance of 1 7 feet therefore passing this test The second part of mechanical brake test involved testing th
19. Alternating Current Brushed DC BLDC Brushless DC Controller Area Network Computer Aided Design Electrical Engineer C C E Electromotive Force FPGA Field Programmable Gate Array GPIO Integrated Development Environment ITAG MOSFET NTSC PMDC M P R RTOS SWDC UART V Video Graphics Array W Work Breakdown Structure AC BDC AD AN DC EE MF GUI HW I O rc IDE LCD LED CU ME OS PCB PID WM QEI OM MA SPI SW UI USB GA BS O 2011 Calvin College and Daniel Evans Matthew Last Matthew Rozema Robert VanderVennen Xl Team 3 Achieving Mobility Final Report Page 1 of 160 May 11 2011 1 Introduction 1 1 Problem Definition The main focus of our project was directed toward an individual named Isaac Postma He is a ten year old boy who was diagnosed with Spinal Muscular Atrophy SMA as an infant This is a rare genetic disorder that causes his muscles to deteriorate over time Currently he has lost virtually all voluntary muscle ability in his body except for the use of his left index finger and slight facial expressions Unfortunately he may lose the ability to use these muscles as well but at this time it 1s hard to say when that will happen according to his doctors He has two primary means of transportation a stroller in which he lies down while someone pushes him as well as an electric wheelchair in which he sits up and controls with his finger Unfortunately both means of
20. Main software components Task Peripheral Interface Self Test 8 1 3 Mechanical The mechanical components were divided into major components and were split up between Rob VanderVennen and Matt Last Each team member was responsible for designing purchasing assembling and testing each component Rob and Matt both have a different set of skills that will be utilized in the design of each component Table 56 shows the mechanical task breakdown and the person assigned to each task Table 56 Mechanical task breakdown Bed Rob Frame Rob Storage Wheels Rob 8 2 Team Organization and Management One of the main goals of the team is maintaining a structure and schedule that is highly organized Managing the time and resources that are available are a crucial part of this course as well as this project The team s advisor Professor VanderLeest did a fine job encouraging the team to stay organized throughout the semester Weekly status reports were submitted each week to show him where the team was currently and where the team expected to go the following week These proved to be very helpful and gave team members a schedule to follow each week In addition a detailed Work Breakdown Structure WBS was made to outline the tasks for the entire year Appendix A Work Breakdown Structure The WBS shows each task the start and end date estimated hours to completion actual hours to completion and the percent complete for the e
21. Seal Lead Acid Deep Cycle Battery eeesseeeeeeeeeessessseeeeeeeeee nnne nnne nennen nnns 27 Figure I2 Lithium Ion Battery nissan aii ew ep 21 Iagpute 13 24V 60A Battery Charro lus ratita linia 29 Figure 14 Voltase Resulator Tor Came d sc20naccecranseccsaneaetsasanataannnhaneacnenasscesaesvengedasteneaatancasnomsasasaanaeusanaeats 30 Figure 15 Average Time for Breaking of Fuse esee 31 Figure 16 Heat Dissipation Calculations 00 ccccccceeeeeeesseseeeeeeeeeeeeeeeeeeeeeeeeeeceeeeseeeeeeeeeeeeeeeeeeeeeeeeeeeeeees 32 Figure 17 Examples of Common Heatsinks cscscovsecseseiesssctersseatasavvavesivveradesevawsiisssaViesetesareedessessidvesdienesers 33 Figure Es Oyer Voltage Protection SCHCMANC cea aa a i a a 33 Figure 19 Power Switch used on final product sesseesesseseeeeeeeeeeeeeeereeeeeeereseeeeeeeeeeeeeeeeereeeeeeeeeeeeeeeeeeees 34 Figure 20 L Spice C I CUT e APRA O PEU o A OTE EN E on eaaa 35 Figure 21 Bode Floto 1000r T ADAC errores EREE 35 Piette 22 Bode PIOC O DTUE Bo MOT MAP o A S 36 Pitre 2os BAOLDB S C M MA e E on E TE E E T TE TE EE OE 37 Fioure os iar SIM Nude mI ES 37 retire 25 Revised EAGLE Schematie serial EENE 38 Figure 26 Revised EAGLE Board Design cccccccessesseseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeees 38 Figure 27 Version of Power Regulation Circuit esesseeseseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeerereeee
22. T ecu for air j bgp A J Page 3 of 8 Team 3 Achieving Mobility Final Report May 11 2011 Power Regulation PR PCB Power Ipr 0 12amp Vor 5V Wor Vor Lor Wor 0 6W Natural Convection pr T pr nat 7 l air i nat A T Forced Convection W pr T T E r for air Mor A J Design PCB Limit Liu 50 C Page 4 of 8 Team 3 Achieving Mobility Final Report Page 5 of 8 May 11 2011 Natural Convection 120 Tme nat 9 i 100 IK Tecu nat O 80 IK Tpr nat 9 C IK 60 Tiimit 9 IK 40 Component Temperature C 20 UA m m Un N hnat 1 Natural Convection Coefficient W m 2 K Forced Convection 70 T 0 C 60 Fr me for 1K T 0 C 50 ENT uen ecu for 7 1K T 0 C 40 pr for 7 IK Tiimit IK 30 20 50 100 150 200 Component Temperature C hor J Forced Convection Coefficient W m 2 K Team 3 Achieving Mobility May 11 2011 Final Report Page 6 of 8 12 7 Appendix G Brake Mounting Stress Calculations Stresses on Brake Support Arm Parameters Dist 2ft W total 450lbf tota WE 0 4 Vel ax 3 5mph Earm 2 25n Lprake 31n Den On Syield 40ksi T yield 26ks1 Distance to stop after brakes are applied Total weight wheelchair is required to handle user frame Fraction of weight the rear wheels support Maximum speed the wheelchair will be traveling D
23. The customer donated one of their old electric wheelchairs which included two electric motors These motors were tested for functionality by connecting a 100 watt power supply to the motor terminals and a 25 watt power source to the brake release terminals Figure 60 The team ran both motors successfully for at least 5 minutes at various speeds DS 2 3 i gt T p z f d P D f r E J u B y HE ae PE n t Figure 60 Test setup for donated motor functionality Team 3 Achieving Mobility Final Report Page 82 of 160 May 11 2011 The two motors are PMDC motors rated for 24 volts and 3 3 amperes This means the maximum power of these motors 1s approximately 0 1 horsepower 24V 3 34 79 2W 0 106 hp Based on Table 17 a power of 0 106 horsepower 1s not enough power for our design requires at least 0 295 horsepower Also the condition of these donated motors is unknown The customer informed us that the wheelchair had been in use for about 5 years which is about the service life of BDC motors The team does not want to give the customer a final product with a motor that they will need to get serviced after less than one year s use Therefore the donated motors were used for benchtop prototyping and early testing only Purchasing new motors was necessary in order to meet the 0 295 horsepower requirement The new motors were selected based on several factors including weight capacity power and wheel s
24. 160 May 11 2011 recommended this fabric and stated the majority of current wheelchairs use Rubatex Therefore 1t was decided that the final design would use dark gray Rubetex fabric The proposed design of the bed closely resembled the customer s current bed since the design was simple enough and the customer was very happy with it Therefore the main cushion on the bed was attached with Velcro to a metal board underneath 1t as shown in Figure 53 In addition cushions were attached to the side of the bed to prevent the user from rolling off of the bed These are removable using Velcro as shown in Figure 54 Figure 53 Main Bedding Cushion Figure 54 Side Cushion The metal base board underneath the bed ensures maximum strength of the bed and was made from aluminum to keep the materials consistent and to ensure that 1t can be welded to the frame Team 3 Achieving Mobility Final Report Page 74 of 160 May 11 2011 The restraint system was designed to ensure that the user is completely fastened in during vehicle transportation There is a main seatbelt attached to the stroller that fastens around the user s waist with two more straps going over the user s shoulders and attaching near the seatbelt This restraint system ensures that the user is completely fastened This is one area of the design that really shows the design norms of trust and caring This restraint system will ensure that the user feels safe and comfortable The r
25. A AAA c cT 72 5 4 2 e DSU 72 5 4 3 Propose Desi ai 72 5 4 4 E e 74 5 5 MOOL aa E LEM Me 74 5521 FES A em D CREPE t MI t a i aie 74 3593 2 Requtement NS E TTD 75 5 5 9 P NITE A sha te cease ate eines taeda ated tere asta dake cnet ota tate cic a toca a oad tale als eosec tained dase 75 5 5 4 SS LECH OM ILC ides etter santas A Lu EE 78 5 5 5 DES en a E M EIE a M HN LII MEM NEUE 80 5 5 6 A O A 83 5 6 A EEN E MU M 84 5 6 1 io y A A un EN RE ERETER 84 5 6 2 A ae Alacer Pe ewe uem ett te an E o EER PAI MU ELE S SLE Oeo sse DL Fone 84 5 6 3 A A A UE 85 5 6 4 ju BSE 6 6 2 bs eR rae tn er eka RI Eee p DUM OR RSS M IUMI MU I IM I 89 S d AA adeb tto Len oaa A A e oia 90 5 7 1 IRROqiire ments aeos texere afeso edet ia 90 5 7 2 ES AAA lacs ics un lod utut ad awsome cim Ul eaque EEE 9 5 7 9 IB pam CD rS 93 5 7 4 sl RR 95 VS Moun cud nosset e Ot 96 5 8 1 Di L MM M E EU 96 5 8 2 E A M P EI EP ME IEEE I DE DM ES ME 98 5 8 3 Printed moult Boada a or m dad 101 5 8 4 Een 101 5 8 5 A E O Mta A FOUN Monee 103 5 8 6 A O R 105 5 8 7 o AO 106 Tesino annaa a TT 108 6 1 Component Lescano 108 6 1 1 SOL E T EMEN ut MM M 109 6 1 2 LCD Cimera Sy SUC IY sd TT UD UNS 111 6 1 3 Power Repbul ti n Circuit DO siii 112 6 1 4 Temperature of Electrical Components outre a atado osa tr tata er eden eens 112 6 1 5 is A A betont dido epe deuten ane abe A 113 6 1 6 PP m 113 6 1 7 A E TE ee
26. Cir 3 272 ft RPM eg RPM ES wheel Cir wheel mn Required Motor RPM Velyeg GearRatio 3 1 RPM motor RPM hoti 2 259 x 10 Cir Force per Rear Wheel due to Friction N WE W wheel tot Ff M N wheel Nneel 1 80 lbf Fp 117 1bf Force per Rear Wheel due to Friction on Incline Ff inc M Nwhee c0s 8 Team 3 Achieving Mobility Final Report Page 159 of 160 May 11 2011 Force per Rear Wheel due to weight on Incline Fw incl Nwhee 12 9 Fy inc 15 688 1bf Combined Force per Rear Wheel due to Friction and Weight on Incline Fcomb inel F indl Fw inel Fcomb inc 132 243 1b1 Required Torque on Wheel T Ef _ P wheel 2 Theel 82 62 N m Required Torque on Incline on Wheel D Tincl wheel gt comb incl Tincl wheel 7 73 384N m Required Torque on Motors 2 E UR motor GearRatio Tmotor 5 445 Nc m Required Torque on Incline on Motor D 2 bind Tincl motor TA Tincl motor 3 891 N m Required Power Preg ES T actor RFM intor Preg 0 261 hp Required Power on Incline Preq incl ES TincI motor RPM motor Preq incl 0 295 hp Team 3 Achieving Mobility Final Report Page 160 of 160 May 11 2011 12 4 Appendix D LTSpice Power Regulation Circuit stable Switching Regulator 24 Volt Node n 12 Volt Node Voltage Regulator Team 3 Achieving Mobility Final Report Page 1 of 8 May 11 2011 12 5 Appendix E Voltage Regulator Calculations
27. Ma sse tette tetto nd 120 Fieure 93 Product Plow deri 125 Figure 94 Breakdown of revenue source for wheelchair market in 1996 iecerei 126 Figure 95 Cumulative weekly hours for the team cccceeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeees 135 Pisure 96 Overall teami organizadon seacor ze esadeuo even E pelos tege Evene eur va dea bussude peau su usa ota s du o smdea Ru rud nus 136 Figure 97 WBS outlying major tasks and completion dates essem 136 Figure 98 Simplified model used for hand calculations sese 152 Figure 99 Bed stress analysis showing a max stress of approximately 1850psi based on 2001b load 155 Figure 100 Deflection analysis showing a max deflection located in the bed of approximately 1 64 156 Figure 101 Max stress in the frame of approximately 510 psi located in the front supports with a 2001b load on the bed and a 100Ib load on the storage shelves oooccccccccccccccnnnononnnnoonncnncncnonononononnnonononnnons 156 Figure 102 Max stress in the storage area is approximately 6000psi with 1251b load 157 Figure 103 Max deflection in the storage area of 0 053 with 1251b load sess 157 List of Tables Table la Lito AAC OM MS ETT T ERES X Table 2 Pros and Cons of PID Algorithm with Optical Encoder coooooooooooononccnncnnnnnnnon
28. Page 139 of 160 May 11 2011 8 3 2 11 Integration of Electrical Components The individual electrical components all need to work simultaneously in a system Therefore once all the components were working independently they were implemented into the final system to make sure that they all worked collectively This milestone was completed April 22 2011 8 3 2 12 Final Prototype Assembly The final prototype assembly was one of the last mechanical milestones to be completed All mechanical and electrical components had to be purchased and ready to be assembled in order to complete this milestone This milestone was completed May 5 2011 8 3 2 13 Power Regulation In order for each component to work sufficiently it needs to be supplied with the proper amount of power Testing was conducted in order to make sure the components have all the power that they need This milestone was completed May 5 2011 8 3 2 14 Final Testing Final testing was the last milestone that was completed for the final product Final testing cannot occur until all other milestones are completed Even though this was the final milestone there still needed to be time to debug the system after testing This milestone was completed May 11 2011 8 3 2 15 Final Report The final report summarizes all of the work from the entire year This milestone was completed May 11 2011 9 Acknowledgements There are several individuals and organizations that have helped the
29. and storage and was tested once the frame was welded together Since the stroller is rated for a 200 pound user and the frame should be the strongest part of the product it will be tested to 400 pounds The weight test was accomplished by placing approximately 400 pounds on the bed while the team observed all joints and supports for signs of weakness and fracture Since the stroller will be susceptible to vibration and fatigue over the course of its life it was very important to observe any noticeable stress concentration factors and surface cracks at an early point to avoid fatigue failure Most cracks will be too small to see and will be negligible but any large cracks must be fixed at this time This test was accomplished by having two people whose combined weight was approximately 400 pounds sit on the stroller The stroller was then pushed around manually and visually inspected by members of the team for any signs of failure During the test there were no noticeable cracks or problems in the structure Both individuals on the frame were asked if they felt any deflection or weakness in the frame during the test and both answered No 6 2 Integration Test Plans This type of testing confirms that all the components work with each other meaning that each component still works when the others are integrated into the system Several components are connected in one way or another For instance the ECU and motor controllers may work great as a co
30. applications The maximum weight the wheels are required to handle is 450 pounds which includes the wheelchair and the user combined Table 17 summarizes the required speed torque and power for the motor and wheel based on the following assumptions see Appendix C Motor Power Speed and Torque Calculations for more detailed calculations e Therear wheels carry 80 of total weight 40 each e Friction coefficient of 0 9 rubber tire of dry asphalt e Safety factor of 1 5 e Rear wheel diameter of 12 5 in e Maximum speed of 3 5 mph 2 5mph 1 0 mph e Gearbox ratio of 24 1 Team 3 Achieving Mobility Final Report Page 81 of 160 May 11 2011 Table 17 Required power speed and torque for motor and wheel Power hp 0 261 0 295 0 261 0 295 For production design the motors with the above specifications will be out sourced to a motor manufacturer The main wheelchair companies all out source their motors to manufacturers located in China which is most likely what we would do Using a manufacturer in China not only saves on money by cutting out mark up costs but it allows us the ability to obtain motors that fit our specifications If we purchased motors through the main wheelchair companies such as Invacare we would have to choose motors based on their specifications available not ours For example this application requires a 0 295 horsepower motor but Invacare only sells 0 5 and 0 9 horsepower motors 5 5 5 2 Delivery Design
31. batteries will be added to the stroller for each weight test The batteries weigh approximately 80 pounds as stated in the batteries section Different weight tests were performed to test different weight distributions The first weight test consisted of having the batteries underneath with a 200 pound person on the bed The second test consisted of putting an additional 50 pounds on the storage shelves with a 150 pound person on the bed The last test put an additional 50 pounds on the storage shelves with 100 pounds on the bed The tests were conducted for at least ten minutes to ensure there are no effects of creep or fracture growth occurring If there was indication of failure or lack of confidence in the design the time would have been increased for that test however this was not needed since lack of confidence did not occur Second for each test all welded joints were inspected all seams on the bed were inspected motor shafts were inspected and wheels were inspected There were not any signs of fracture cracking and ripping during this test In addition to testing component failure the stroller was also be driven to ensure that the motors could move the required weight without burning out the motors In addition to driving the stroller on flat ground the stroller was also driven up a 10 incline to ensure that the incline requirement was met while fully loaded Team 3 Achieving Mobility Final Report Page 118 of 160 May 11 2011 6 2
32. determined to not be feasible because it requires an axle across the bottom of the frame to be mounted to in which case the current storage shelf is obstructing it A hand brake however would be mounted to the wheelchair handle bars and a squeezing force would be required to operate them Figure 62 Comparison of foot lever and hand lever for attendant controlled braking system Team 3 Achieving Mobility Final Report Page 86 of 160 May 11 2011 A separate braking mechanism for the attendant controlled braking system needs to be implemented during manual drive because when the motors are disengaged the user controlled braking mechanism via reducing current to the motor is not available The different braking mechanisms considered were a rim brake a drum brake and a disk brake A rim brake presses two pads together against the rim of the rotating wheel A drum brake presses two pads outward against the inside surface of the wheel hub A disk brake consists of pressing two pads together against a metal disk attached to the wheel axial Figure 63 compares the different types of braking mechanisms Drum Brakes i Rim Brakes Figure 63 Comparison of different braking mechanisms for attendant controlled brakes Table 21 shows a decision matrix for the braking mechanism alternatives The scores were based on online research of the different brake types Table 21 Decision Matrix for Braking Mechanism c e mem Disk Brake
33. direction to turn the motor 4 3 3 MCU Selection The MCU used in the production design would probably be a Luminary Micro Cortex M3 This MCU worked great for the delivery design and some of the software written for the delivery design could be ported to the production design Also the team already has experience writing software and using the drivers for this family of MCU so the design time would be less for the production design The selected MCU would need at least two PWM generators 6 8 ADC at least 10 15 GPIO a minimum of two timers and UART There are several possible MCUS that fit these specifications and the design would not be constrained to only one or two possibilities 4 3 4 PCB Design As mentioned earlier the goal of the production design would be to create a more cost effective solution This would be done by minimizing the number of redundant components and creating a single PCB that contains all the functionality of the delivery design s PCBs The delivery design has four distinct PCBs two motor controllers the ECU and a power regulation PCB Therefore the production design would be created on a single board using one MCU and as few redundant analog components as possible The power regulation system would also be placed on this PCB Overall this would have been the ideal solution for the project but the design of this PCB would have been too much work and taken too much time to complete with just two electrical e
34. e Small physically A wheel lock parking brake was selected for the delivery and production design due to its simplicity and ease of mounting An axle lock is not feasible because it requires an axle which is not included in the frame structure for both the delivery and production design because the motors run independent of each other For the production design the parking brake will be outsourced due to the low production volume For the delivery design the parking brakes off the donated wheelchair were used because they were Team 3 Achieving Mobility Final Report Page 88 of 160 May 11 2011 functional and free of cost Figure 65 shows a picture of the parking brake that was used in the delivery design Figure 65 Donated parking brake being used in delivery design 5 6 3 4 Emergency Brake It is a real possibility that the user s finger gets tired and he or she does not have the strength to lift their finger off the touch pad causing the vehicle to move continuously in a particular direction Clearly this is extremely dangerous especially if in the forward or reverse mode Therefore an emergency brake will be implemented into the system so that a bystander can immediately stop the device whenever necessary The motors that will be out sourced for both the production and delivery design include an automatic braking device that utilizes a powerful magnet When enough voltage is applied to this device the current creates a magnetic field
35. is necessary for this real time system because any extensive timing latency could cause harm to the user Additionally the cyclic executive initializes the system on startup by performing self test and initializing global configurations such as maximum speed temperature peripheral initialization and interrupts Then after the initialization sequence the cyclic executive cycles between polling for the state of the touch button every 1 ms and responding to CAN interrupts from the motor controllers as well as servicing the directional state machine refer to section 4 2 2 The cyclic executive is discussed in depth in section 4 2 1 3 2 4 PC Interface The PC interface allows a technician to connect the device to a PC in order to install future software and firmware upgrades For the delivery model the firmware installation can be done using the LM Flash Programmer application provided in the evaluation kit This is a PC application provided by Luminary Team 3 Achieving Mobility Final Report Page 16 of 160 May 11 2011 Micro that allows software downloads to the device through a USB JTAG interface In the case of a software update the binary file would be provided by the software developer 1 e a member of Achieving Mobility A USB cable will be provided with the stroller which can be connected to the Electronic Control Unit and programmed with the upgraded binary Also it should be noted that the so called technician would be a member of
36. it was capable of handling the necessary functions This selection was done on November 3 2010 8 3 2 2 PPFS The PPFS contains a detailed overview of all components of the system Several aspects that were analyzed are the design alternatives feasibility and financials of each component This document has proven to be very beneficial in pointing the team in the right direction and figuring out how to manage the project This milestone was completed December 6 2010 8 3 2 3 Website Posted Posting the team s website was a large milestone since it tracked the progress of the team and is available to anyone to view This milestone was completed November 24 2010 Team 3 Achieving Mobility Final Report Page 138 of 160 May 11 2011 8 3 2 4 Directional Control Software Completed The directional software is the most fundamental part of the product The product cannot be called a success without a working directional control system This software determined which direction the stroller would move depending on which LED was illuminated and thus turn the motors in the correct direction This milestone was completed January 20 2011 8 3 2 5 Finite Element Analysis To make sure the structure would hold up under stresses induced by the weight it carries a finite element analysis was conducted using the computer program Autodesk Algor This milestone was completed February 28 2011 8 3 2 6 Final Structure Design The final design of the s
37. low volume product Although the encasing material and dimensions will ultimately be dependent on the manufacture s product line specifications close to those in Table 28 will be sought A gasket will also be purchased and fitted around the main PCB encasing to prevent water from seeping in at the edges The main PCB will require forced convection using a fan system because both the motor controller and power regulation components require forced convection to keep them below the design limit of 50 C The plastic encasing for the main PCB will include air way slits at each end with a fan at the entrance of one side 5 7 3 2 Delivery Design The motor controllers for the prototype were both donated by Texas Instruments and already include the proper encasing and cooling system design Battery encasings were donated by the Postma family and were confirmed to properly work with the batteries that were purchased for the prototype The only electronic component that required proper encasing design was the ECU The ECU was placed in an aluminum encasing that covers all three of the circuit boards 2 motor controls with plastic encasing and 1 ECU Aluminum was used for the delivery design encasing instead of plastic because aluminum can be obtained free from the Calvin College metal shop An open celled foam gasket was placed around the aluminum encasing to prevent any water from leaking in at the edges Figure 69 and Figure 70 show the delivery design of th
38. lower strength alloy would not have any adverse affects Safety was a very large consideration since one of our design norms was trust therefore extensive calculations were down to be certain that this lower strength alloy would still provide enough strength Second aluminum 6063 T6 machines easier than 6061 T6 making the frame easier to manufacture ultimately saving money on manufacturing costs 5 1 3 Material Size The material for the frame 1s circular structural tubing since the motors and brakes are designed to mount to circular tube Since most components are designed to mount to 1 OD tube that is the size that we decided to use We went with 1 8 wall to provide more than enough strength as stated above Smaller tube could have been used such as 3 4 OD but this is actually more expensive than 1 OD because it is not a standard size All quotes of sizes were obtained from Alro Steel Corporation 5 1 4 Final Design The material of the final design is aluminum 6063 T6 Some of the main factors that influenced this decision were corrosion resistance as well as weight reduction Once all of the electronic components mounting components motors brakes and storage racks were selected the final frame was designed to incorporate all of these components The design shown in Figure 47 was a preliminary model that shows a rough estimate of his current stroller with two electric DC motors and two 12V batteries added This design was mainly
39. mind Safety features on the final design include three independent braking systems one for the person pushing the vehicle one for the driver and one emergency stop built into the motor There is a control system implemented to trigger the brakes if any electric or system failure occurs In addition detailed testing has been performed on all mechanical and electrical components to ensure that no one will be injured from this product A safety factor of two has been used on the frame to ensure that the user feels safe using the product Team 3 Achieving Mobility Final Report Page 8 of 160 May 11 2011 2 5 4 Transparency The design norm of transparency should demonstrate a design that is easily understandable and predictable to the user There should be little confusion with how to operate the final product and each feature functions This design norm is also very important for this project since an individual needs to learn how to drive the vehicle easily and in a relatively short amount of time This product will typically be designed for people with physical and possible mental disabilities therefore it must function in a way that people can understand very easily This design norm may conflict with the design norm of trust because if more safety features are added to the product it will complicate the design and make it less transparent Therefore the team has assessed the individual in each circumstance to see what balance is best for the pa
40. need replacement after some time would be the bearings 5 5 4 4 Safety Safety is very important in this entire design All of these electric motors are safe and the only difference between the three is that BDC motors have a slightly higher risk for sparks due to the brushes rubbing against the commutator 5 5 4 5 Speed Control Speed control is an important selection criterion because the customer needs to be able to effectively control his her wheelchair safely All three motors have excellent speed control The BLDC motors have the most precise speed control out of the three due to the fact that the current is controlled by a complex controller The next most precise are the SWDC motors because the current in the rotor and the stator are independent from one another PMDC motors are also precise because the relationship between output speed and voltage is linear so it is extremely easy to control 5 54 6 Power Power is important because some motors perform more efficiently in higher power applications while others do better in lower power applications The application for this design requires roughly 0 295 horsepower as shown in the motor power calculations in Appendix C Motor Power Speed and Torque Calculations For this reason SWDC motors do not fare as well because they are best suited for applications involving 5 horsepower or greater PMDC motors on the other hand perform the most efficiently at fractional horsepower Team
41. off site at Ebling and Son Blacksmiths 5 3 3 Financials The current price of aluminum sheet metal 3 32 thick is about 0 04 per square inch based on rates from Alro Steel Corporation The exact number of square inches needed for the shelves is approximately 834 square inches as shown in Table 15 Therefore 834 square inches will cost 51 84 Table 15 Storage Shelf Dimensions Overall Dimensions Area sq in Lower Shelf 17 x 34 88 502 88 Upper Shelf 13 38 x 18 240 75 Total Area 833 63 5 3 4 Final Design The final design of the storage area 1s shown in Figure 52 The lower storage area will be used mainly for the batteries and any medical equipment that the user has The upper storage area will be used for any extra supplies that the customer may have with them There was an edge ranging from 2 3 tall made on each storage unit to prevent objects from falling off the side The ends were bent up using a sheet metal press and the sides were welded on In addition all corners were rounded to prevent anyone from being injured on any sharp edges Team 3 Achieving Mobility Final Report Page 72 of 160 May 11 2011 Figure 52 Final Design of the Storage Area 5 4 Bed 5 4 1 Requirements The bed of the stroller must provide the user with a comfortable location to lie for the majority of the day In addition the bed must have removable exterior for cleaning There shall be enough cushioning in the bed so the user cannot feel an
42. other hand are much more suitable for this type of application Electric motors can be divided into two categories alternating current AC motors and direct current DC motors however only DC motors will be considered due to the power requirement of DC voltage battery All DC motors consist of a stator and a rotor The stator includes everything in the motor that is stationary The rotor includes everything in the motor that rotates Both of these components produce Team 3 Achieving Mobility Final Report Page 76 of 160 May 11 2011 magnetic fields that interact and turn the rotor The configuration of the stator and rotor determine the different types of DC motors 5 5 3 1 Permanent Magnet Direct Current PMDC Motor Permanent Magnets Coil Windings Commutator Battery Terminals Figure 56 Parts of a PMDC motor In PMDC motors the stator consists of two or more permanent magnets and the rotor consists of several coil windings Figure 56 The permanent magnets produce a constant magnetic field between themselves because they have opposite polarities The coil windings also produce a magnetic field around themselves when current is applied In order to apply current to rotating coils brushes and a commutator are necessary Brushes press lightly against the rotating commutator providing current to the commutator The commutator is then connected to the coil windings and provides current to the coil windings The comm
43. our product is estimated to be about 14 058 7 5 3 which is fairly competitive in comparison to the models above It must be taken into consideration though that the additional features that these models offer our product lacks such as wheel suspension system adjustable seating etc 7 3 Market Research 7 3 1 Customer Base The overall customer base for this product is people with conditions that limit them to wheelchair use Table 42 shows the leading conditions associated with wheelchair usage Team 3 Achieving Mobility Final Report Page 123 of 160 May 11 2011 Table 42 Leading conditions associated with wheelchair use Persons mb Condition 10005 wheelchair users 5 Stroke Arthritis Multiple sclerosis Absence or loss of lower extremity Paraplegia Heart Disease Cerebral palsy Rheumatoid arthritis Diabetes 2 7 3 6 3 6 3 23 3 3 0 2 4 60 Orthopedic impairment of lower extremity Within this broad customer base of all wheelchair users a number of people would prefer to lie down in a wheelchair due to the certain condition they have The more specific customer base for our production design is aimed at children to small adults that are physically limited to lying down The most likely reason for lying down would be no control over one s upper torso muscles making sitting up painful or uncomfortable Loss of control of the upper torso can be caused by conditions such as e Spinal Muscu
44. pound weight requirement should be more than sufficient since the average weight of a 60 child is 85 100 pounds In addition the frame shall be able to withstand any direct frontal impact with a wall at the max speed of the motor of 2 5 Team 3 Achieving Mobility Final Report Page 61 of 160 May 11 2011 mph Finally all welds on the frame shall be continuous with a nominal radius equal to the thickness of the tube to ensure maximum strength of the joints 5 1 1 2 Weight Since most wheelchairs are often pushed up and down wheelchair ramps the stroller must maintain a welght of less than 250 pounds The main concern 1s to make sure most individuals will be able to push the stroller up a 10 ramp and have the brakes strong enough to easily stop it going down a 10 ramp In addition the wheelchair must be able to be lifted by a hydraulic lift in a transport van According to BraunAbility a leading dealer in wheelchair lifts for vans most wheelchair lifts can handle a weight a weight capacity of 600 pounds Therefore with the weight of the wheelchair equipment and person the overall weight is much less than 600 pounds and should work with most wheelchair lifts 5 1 1 3 Aesthetics The overall look of the wheelchair must be pleasing to the eye since it will be the customer s main means of transportation Avoiding sharp corners and any sort of boxy look will be very important throughout the design 5 1 2 Material Selectio
45. sealed in order to avoid the risk of acid spill Since the battery is charged with 120V 60 Hz standard outlet there is also the possibility of electrocution from the outlet A frequency of 60 Hz is more dangerous than a DC voltage because alternating currents flow bidirectional which may cause one s heart into fibrillation due to the twitching of the muscles Team 3 Achieving Mobility Final Report Page 12 of 160 May 11 2011 When dealing with 60 Hz AC a dangerous amount of current is approximately 15mA Therefore the resistance needed from one s body to prevent this dangerous current amount is calculated V iR 120V 15mA R R 8kQ This resistance is high enough that in order for it to be dangerous one needs to touch the outlet with a wet finger or a metal instrument Although this may be a possibility it is slim enough that the team is confident in the safety of the product One may also believe there is a significant risk of electrocution if the device 1s accidentally driven into a lake or pool Electrocution in a bathtub may be possible but lakes and pools are much larger Due to the high volume of water it has a high resistance and the electrical current disperses throughout the water Therefore there is not a significant risk of electrocution if the customer accidently drives into a lake or pool The team greatly discourages this though because the device will be ruined 3 2 Software The software provides a dynamic use
46. sections describe the physical components of the device including their selection and how they work together within the system The software sections describe the software architecture plans 4 1 Hardware Electronic hardware serves as the central nervous system of the device providing all the necessary physical electronic components and connections to allow the system to perform as desired The electronic hardware is primarily responsible for executing software connecting each component with necessary interfaces and providing stable power inputs to each component 4 1 1 Motor Controller A motor controller is an electronic control system that uses both digital and analog components to control the behavior of a motor For this project the motor controller needed to control two 24 Volt motors independently The circuitry of a motor controller typically includes at least an MCU PWM digital interface an H Bridge and other protection circuitry such as diodes fuses and capacitors Refer to Team 3 Achieving Mobility Final Report Page 18 of 160 May 11 2011 4 1 12 for more information on the motor controller analog hardware Motor controllers are very intricately and precisely designed for safe usage and therefore there was not enough time to design a motor controller from scratch in fact designing a motor controller could be a design project by itself Therefore a pre made development kit was used instead of designing our own A devel
47. team out with the project A special thanks goes out to the following e Calvin College for providing the team with the resources to design the project as well as funding a portion of it e Professor VanderLeest the team advisor for pushing the team and offering advice and criticism wherever necessary His organization and project management techniques proved to be very helpful in organizing the project e Tim Theriault from GE Aviation for being the team s industrial consultant and providing us with contacts at Texas Instruments Team 3 Achieving Mobility Final Report Page 140 of 160 May 11 2011 e Russel Ramsay from Texas Instruments for donating the development kits and offering his time to help the team through problems e Ivanrest CRC for providing the necessary funding for this project A special thanks to Carole Pettijohn head of outreach ministries at Ivanrest for setting up the donations and making this project possible e Invacare for providing the team with guidance with wiring up the current motors e Greg Bush electrical design engineer at Gentex for providing the team with LCD screens and cameras as well as mentoring the team with the camera and LCD screen design e Jim Nammensma manager at Ebling and Son Inc Blacksmiths for assisting the team in welding and machining the frame e Professor Ribeiro the professor of Engineering 315 for allowing the team to use class time and resources to design components of the
48. that the manufacturing of this product would produce profits after two years estimating sales of roughly 100 strollers per year The final product is a working prototype of this electronic stroller which was distributed to the customer This prototype was a representative of a final design that could be either created into an individual business or sold to an existing rehabilitative wheelchair manufacturer The device was completed and presented May 7 2010 Table of Contents US MH Rn 1 Ec Gsm E Po O UP O mI T 11 Ps AP m vi A E A Vill ME Ioue s CIO AAA E I Ze Poe CU Te CU Me RIO S corra placa oleo 3 2 1 bonctrond E R guireme dit blas 3 22 INVES ANIC TI AAA e E PuOi send Ide nii 3 2 2 1 bounce rei 3 22 2 Product A A A 3 22 5 A A 3 2 2 4 E RS 3 2 25 ETES A Mice eaten E E PE O o bm EIU Eum Im MEET UE 4 2 2 6 IA e PR EE E A E 4 2 22 A A 4 228 NOUO O E on E E E E AEE OE A 4 22 9 A E E E 4 PAP E N TT 4 PST NNR TNR 4 2 3 Plectrical Requirements arepa io spot 5 22 4 FO TO PREND u 4 X 5 2 92 Ie METTE E 5 2 3 3 IB S PEOR m 5 2 3 4 PPP o m o E E T 5 2 3 5 SOL Woo oi oi E 5 2 3 6 CO A e E E 6 2 4 SAC yi COU NG 11 ceso TO OT 6 2 4 1 STD o p E Pe Un O e o A T E ne eee eee 6 2 4 2 PP E A 6 2 4 3 ET A Oe eee eee ec A E ied 6 2 4 4 A e O IA 6 2 4 5 Rai
49. that wheelchairs cover more than one third of the home medical equipment market Table 44 Home Medical Equipment Market Breakdown Equipment Type of market 3 2 Wheelchair INE Home Care Beds E EN 8 7 Bathroom Safety Supplies Ambulatory Aids Miscellaneous Patient Aids Wheelchair usage in terms of the entire population is very small but because 1t dominates the home medical equipment market a low volume high priced approach 1s feasible Table 45 shows the estimated annual births per year in the United States for conditions that that would have the highest possible demand for a horizontal powered wheelchair based on average 4 million annual births in the United States Table 45 Annual births per year in the United States for conditions most likely requiring horizontal powered wheelchair per year Spinal Muscular Atrophy 700 Spina Bifida 400 Quadriplegia 4 000 5 100 A total of 5 100 people each year in the United States could be demanding a horizontal wheelchair This number is optimistic because 1t assumes that all these people are capable of controlling a powered wheel and that all these people live long enough to control a wheelchair on their own Team 3 Achieving Mobility Final Report Page 125 of 160 May 11 2011 7 4 Target Market The target market for this device 1s the home medical equipment sector more specifically the manual and powered wheelchair market Figure 93 shows the flow from manufac
50. therefore rather high because salespersons must be paid to go out and convince insurance companies and doctors to insure and recommend the product This model will only be feasible if there 1s product diversification because one wheelchair alone is not enough to sustain a company in such a small target market Invacare for Team 3 Achieving Mobility Final Report Page 120 of 160 May 11 2011 example is so successful because they have such a varied line of wheelchair products to meet the different needs wants of the customer 7 1 2 Wheelchair Supplier This model involves manufacturing the product and selling it to a wheelchair supplier who then turns around and sells the product to the end user The main advantage of this model is the product is backed by a brand name such as Invacare Insurance companies will most likely back the product because of the supplier name on it Hospitals and doctors will also be more likely to recommend the product because of their familiarity of the supplier name A disadvantage to using this model is a smaller margin because the product must be cheaper for the supplier to buy than to manufacturer themselves The product must also be marketed to the supplier instead of the end user The business model that will be proposed for our project is direct to the user This means manufacturing the product in house and then selling it directly to the end user This is the best option for a low volume highly customized high co
51. to prevent overheating The amount of heat that is required to be removed from each component assuming a worst case scenario all input power is converted to heat is roughly estimated to be Motor Controller PCB 10V x 10A 100W 341 BTU hr Power Regulation PCB 24V x 3 4A 82W 280 BTU hr ECU PCB SV x 0 120A 0 6W 2 BTU hr Team 3 Achieving Mobility Final Report Page 91 of 160 May 11 2011 5 7 2 Alternatives The design choices that need to be determined for the encasings include material size and heat dissipation The alternatives for material are steel aluminum and plastic Table 23 provides a decision matrix comparing the important characteristics of each alternative Table 25 Decision Matrix for Encasing Material Selection i ee d Weighted s Weighted Score Weighted CTA EA ET EE E Weight 9 2 18 5 45 10 90 cot 8 17 56 5 4 10 80 EE E E E s The size of the encasings is mainly dictated by the size and shape of the electronic components it houses The frame layout and mounting design also put restraints on the size of the encasings The dimensions of the electronic components along with their proposed encasing dimensions are shown in Table 26 The motor controller and ECU will be placed in the same encasings Table 26 Electronic Component and Encasing Dimensions Component Dimension Encasing Dimension Component in in Motor dad ECU A5x 75x23 5x 825x4 Power Re
52. to move the wheelchair from rest for the maximum combined weight requirement of 450 pounds This includes enough torque to travel up a ramp with the maximum incline requirement of 10 The minimum required motor power based on the necessary torque 1s 0 293 horsepower Appendix C Motor Power Speed and Torque Calculations 5 5 2 2 Speed The motors shall have a fixed speed once the device has accelerated from rest This fixed speed will be set at 2 5 mph 1 0 mph The motors must therefore be able to provide a minimum angular velocity of 2 259 rpm Appendix C Motor Power Speed and Torque Calculations 5 5 2 3 Acceleration The motors shall be capable of accelerating the wheelchair from rest at a rate of 1 5 feet per second squared This acceleration means 1t will take the wheelchair roughly 2 seconds to reach the desired top speed of 2 5 mph from rest 5 5 2 4 Power The motor shall be capable of being powered by DC voltage battery 5 5 2 5 Engage Disengage The motors shall have the capability of being engaged and disengaged easily by an attendant The motors will need to be engaged in order for the user to control the wheelchair The motors will need to be disengaged in order for an attendant to manually push the wheelchair 5 5 3 Alternatives Combustion engines were not considered because they are not feasible for this application for several reasons including noise safety usage and emission concerns Electric motors on the
53. transportation are no longer suitable for Isaac He has outgrown his stroller and the manufacturer does not make one large enough to fit his growing body In addition he cannot use his electric wheelchair for more than twenty minutes due to the amount of strain that it puts on his back and breathing difficulties encountered while sitting up After hearing about Isaac the team knew that this was a project worth pursuing and that this was a great way to make an impact in the community 1 2 Course Overview Engineering 339 and 340 are a two course sequence offered at Calvin College to give senior engineering students experience with how the design process works These two courses often referred to as Senior Design are capstone courses offered at Calvin College where students tackle a design problem over the course of two semesters The first semester places an emphasis on team formation identifying a viable design problem and conducting an in depth feasibility study Researching and prototyping are conducted with an emphasis placed on accomplishing design norms to incorporate the reformed Christian worldview The second semester focuses on completing the work initiated in the first semester This typically includes designing the final product on paper and producing a working prototype to demonstrate the design with functionality in mind The final part of the class entails testing and performing engineering analyses which are completed with the design n
54. with the customer there does not seem to be any need for removable shelves Therefore the delivery design did not have removable shelves The aluminum sheet metal that is used to make the shelves was welded to the frame since the Team 3 Achieving Mobility Final Report Page 71 of 160 May 11 2011 material was the same and provided more strength The upper storage shelf on the other hand is removable with fasteners in case the user ever wanted to remove 1t 5 3 2 2 Size The size of the storage shelves was most dependent on the final design of the frame The entire bottom of the frame will be used for storage therefore the overall width and length of the frame determined the size of the storage area With the current design of the frame there is enough storage for all equipment underneath the stroller since the wheelbase is longer than the customer s current stroller and the width is about the same 5 3 2 3 Strength The load capacity of the storage area is largely dependent on the type and thickness of material used The material is aluminum 6063 T6 which means the thickness will determine the strength of the shelves The customer s current stroller uses 14 gauge sheet metal but Calvin College s metal shop can only bend up to 16 gauge materials Since 16 gauge material will cause too large of a deflection in the shelves as shown in Appendix B Stress Calculations 3 32 thick sheet metal was used Consequently the shelves were fabricated
55. work with brushed de motors PWM and or CAN based closed loop speed control built in QEI inputs temperature control options includes analog HW Simple design and layout certain 1t will work with brushed de motors No LCD or video drivers less extensive motor control libraries Insufficient power to run our DM163029 Altera DE2 built in QEI inputs example SW included plenty of peripheral expandability inexpensive includes analog HW Very flexible FPGA solution plenty of peripheral expandability relatively inexpensive experience with this kit motors 8 bit MCU less helpful SDK Does not include any necessary analog HW no motor control libraries no PWM drivers 4 1 2 Electronic Control Unit As mentioned in section 4 1 1 2 an Electronic Control Unit ECU was needed in our design to coordinate the behavior of each motor controller and also to provide a user interface Since the motor controllers were decided first the selection of ECU was limited to choices that would be compatible with the motor controllers Team 3 Achieving Mobility Final Report Page 21 of 160 May 11 2011 Since the motor controllers came with a full set of CAN drivers the selection of ECU was limited to a development kit that uses the same set of CAN drivers However this was an acceptable limitation to overcome because the use of CAN as the inter MCU communication protocol is a feasible solution and the hardware topology create
56. 00 Home Depot e Steel sheet 14 gauge 0 40 e Aluminum sheet included in material cost of frame e Fabric included in bedding cost e Velcro 2 00 Home Depot e Straps included in bedding cost e Magnet 2 32 Home Depot 5 00 Nuts bolts washers et Mounting and connections sias sss SSS Touch Button Mounting 39 32 Team 3 Achieving Mobility Final Report Page 108 of 160 May 11 2011 Table 36 Delivery design costs for the encasing systems Delivery Design Cost e 0 585 OD heavy stiffness 18 length flexible tube donated e 0 00 e Aluminum sheet included in material cost of frame Encasing Mounting 0 00 Only requires bolts which is covered under miscellaneous 0 00 Included in material cost of frame e PVC pipe 5 00 Home Depot e Steel sheet free from metal shop e Aluminum sheet free from metal shop Touch Button Mounting 9 32 Fabric included in bedding cost e Velcro 2 00 Home Depot e Straps included in bedding cost e Magnet 2 32 Home Depot 0 00 ioe bolts washers are free from metal 50 00 50 9 Mem S 6 Testing Detailed test plans were utilized to ensure that all components were working properly before being integrated into the system In addition integration testing was conducted to ensure that the overall product passes certain tests with all of the components working together Each test plan was designed to test the worst case scenario to ensure
57. 10 740 Fri 12 10 40 E Electrical 158 days Mon 10 440 Fri 5 6 41 i Create System Level Block Diagram 1 day Mon 10 4 10 Mon 10 4 10 Create System Topology Diagram 4 days Mon 11 1 10 Thu 11 4 10 j Hardware 130 days Mon 10 18 10 Fri41541 ENS MUI Software 133 days Fri 11 5 40 Fri 5 6 41 5 x 0 Mechanical 143days Thu 10 740 Fri 42211 MM M d Structure 143 days Thu10740 rri4 2211 m H n FEA analysis 5 days Mon 2 7 11 Fri 2 11 11 Encasings 96 days Mon 12 6 10 Mon 4 18 11 R Motor 127 days Mon 10 25 10 Tue 4 19 11 Dp Braking 118 days Fri 11 5 10 Tue 4 19 11 Touch Button 21 days Mon 3 7 11 Mon4 441 NEM LCD Screen 42 days Mon2 2141 Tue 4 19 11 M M A Camera 31 days Mon3 741 Mon 4 18 11 Dn Testing 22days Mon 441 11 Fri 5 6 41 eT Business Plan 7 days Mon11 15 10 Tue 11 23 10 Figure 97 WBS outlying major tasks and completion dates Team 3 Achieving Mobility Final Report Page 137 of 160 May 11 2011 8 3 2 Milestones The team identified the major milestones that had to be completed for the project by the end of the year Table 57 shows the completion dates and status of all major milestones Table 57 Project Milestones 6 3 2 1 Microcontroller Selected The microcontroller is the part of the system that does all the processing The design team chose the Micro Cortex M3 as the microcontroller because
58. 2 2 Velocity Control Algorithm Since the interface between the user and the motor control system is a simple touch button there is no means for variable speed control by the user In other words the user is not able to control the speed of the device directly Therefore 1t was necessary to implement a speed control algorithm to drive motors at a constant speed Initially the chosen method for velocity control was closed loop PID control using an optical encoder This method was well documented in the Project Proposal and Feasibility Study PPES However as the semester progressed we realized that using encoders may not be the best idea The main reason why encoders would not work well is because there is not enough room on the motor shaft for both an electronic brake and the optical encoder Therefore using an encoder would require removing the electronic brake However if this brake was removed the device would have no means of braking in case of an emergency such as software failure Keeping the brake on the motor allows the attendant to quickly stall the motors regardless of the state of the electronic system Also the team decided that such precise velocity control was not necessary for the customer s safety A PID algorithm could easily maintain a speed within 0 5 mph but the required speed window is 1 mph Finally after looking for optical encoders to buy it was discovered that it would cost over 100 per encoder which is quite expens
59. 2 2 Weather Testing The goal of the weather testing 1s to make sure that the product will still operate in the rain as well as resist corrosion The easiest way to test the resistance to rain was to drive the stroller around when it was raining outside It should be noted that the final product is designed to be rain proof not water proof This means that components must be designed to resist penetration from rain There are attachments on the side of the stroller to attach a custom umbrella provided by the customer which will keep most of the components dry however the motors are still very susceptible to rain being on the bottom of the stroller as water will spray on them from the wheels The goal of this test was to ensure that all electronic components remained functional when exposed to the rain Bare wires that become exposed to water can pose a serious safety risk therefore this test is very important for functional and safety reasons After performing this test in the rain there were a few areas of concern First the LCD screen is not sealed on the sides and can easily get water inside of it Therefore it is very important that the customer store the screen completely under the stroller when not in use There was also some concern with water getting in the electronics encasing as it traveled down the gooseneck and followed the wires into the encasing To prevent this the hole in the encasing was sealed to prevent water from getting in Besi
60. 2010 lt http www permobil com USA Products Rehab C350 Corpus gt Invacare eForms Invacare Price Sheet Invacare Corporation 2010 Web 4 Dec 2010 lt http www invacare com doc_files 09 068 20FDX CG 20Tilt 20Elevate pdtf gt C350 Corpus Permobil Price Sheet Permobil 2010 Web 4 Dec 2010 lt http www permobil com Global US A ORDER 20FORMS 08 09 10 200RDER 20FORM S 08_C350_Corpus_2010_Rev03 pdf gt Kaye Stephen Taewoon Kang and Mitchell LaPlante Wheelchair Use in the United States Disability Statistics Center University of California San Francisco May 2002 Web 3 Dec 2010 lt http dsc ucsf edu publication php pub_id 1 gt Spinal Muscular Atrophy SMA Frequently Asked Questions FAO Spinal Muscular Atrophy Foundation 2010 Web 4 Dec 2010 lt http www smafoundation org faq gt Spina Bifda Symptoms eMedicine Health n d Web 4 Dec 2010 lt http www emedicinehealth com spina_bifida page3_em htm gt Spinal Cord Injury Quadriplegic and Paraplegic Injuries Apparelyzed 2003 2010 Web 4 Dec 2010 lt http www apparelyzed com gt Schworm Kimberly 1998 The Industry s Facts amp Figures HomeCare Magazine 20 7 p 51 58 Thorogood Christine and Michael Alexander Cerebral Palsy eMedicine 2010 Web 4 Dec 2010 lt http emedicine medscape com article 3 10740 overview gt Team 3 Achieving Mobility Final Report Page 146 of 160 May 11 2011 69 70 71 T2
61. 3 Achieving Mobility Final Report Page 80 of 160 May 11 2011 5 5 4 7 Control Complexity The complexity of the control is fairly important because there are time constraints to build this product for the customer Because of the linear relationship between speed and voltage the controller for BDC motors is very simple and easy to integrate BLDC motors on the other hand require a complex controller because the position of the rotor and the timing of alternating the current must be just right for the motor to run smoothly 5 5 4 8 Other Heat generation and noise were considered but not included in the selection criteria because they are almost negligible in this size of electric motors Weight and torque were also considered but not included because all three motors score about the same 5 5 5 Design 5 5 5 1 Production Design PMDC motors were implemented into the final design product This type of motor was chosen for two main reasons The decision matrix in Table 16 shows that PMDC motors have the highest score based on the weighing of the selection criteria All the emphasized categories such as cost power speed control and safety scored very well under PMDC motors Also based on information from Jill Kolczynski Research and Development Lead at Pride MobilityO PMDC motors are the motor of choice for electric wheelchair applications because they are the least expensive to purchase and are very efficient with fractional horsepower
62. 5 6 2 Requirements 5 6 2 1 Brake Functions The wheelchair s braking system shall provide adequate braking for four different functions user controlled touch button braking attendant controlled hand braking parking brake and emergency brake switch User controlled brakes involve the braking needed while the motors are engaged and the user is driving the wheelchair Attendant controlled braking involves braking needed while the motors are disengaged and an attendant is pushing the wheelchair manually The parking brake involves locking the wheelchair in place to prevent any undesired movement The emergency brake involves allowing an attendant close by the ability to quickly stop the wheelchair in the event of an emergency as well a means to immediately stop the wheelchair if the batteries die 5 6 2 2 Braking Distance The stroller shall be able to stop in less than two feet on dry pavement during both user controlled and attendant controlled braking Team 3 Achieving Mobility Final Report Page 85 of 160 May 11 2011 5 6 3 Design 5 6 3 1 User controlled Braking System One braking system must be designed to give the user driving the wheelchair the ability to slow down or stop The two options involve a mechanical system and an electrical system A mechanical system was determined to not feasible due to the physical limitations of the users intended for this device An electrical system on the other hand was considered and would involve d
63. 53 of 160 May 11 2011 experiencing a sizable opposing force so the voltage should be increased to overcome this The voltage is increased in small increments every 1 5 seconds to ensure that the stroller does not accelerate too quickly There is however a limit to how high the voltage 1s allowed to climb this value is set to 10 V which is 2 V greater than the typical 8 V in the forward direction Conversely the team decided that 1f any motor draws less than 4 A of current for over 1 5 seconds 1t must be experiencing a sizable helping force so the voltage should be decreased to slow down the stroller The minimum voltage 1s set to 6 V which is 2 V less than the typical 8 V Finally if the current returns to a value within the upper and lower thresholds the ECU sets the voltage back to 8 V 4 2 5 Device Drivers Device drivers are modules of software that allow higher level software to interact with a hardware device They control the data transmission to and from the hardware and must handle any asynchronous interrupts that may occur The Stellaris Cortex M3 MCUs that were used in the device utilize a section of ROM that contains an extensive library of drivers for PWM UART CAN USB GPIO graphics and external timers This library contains all necessary functionality needed for the device These built in drivers allow a programmer to focus on high system level code development rather than low component level driver development Also
64. CALVIN COLLEGE ENGINEERING MOBILIT Final Report Team 3 Daniel Evans Matthew Last Matthew Rozema Robert Vander Vennen ENGR 340 Senior Design 12 May 2011 Abstract The following senior design project designed and distributed a customized electronic wheelchair hereon referred to as a stroller because of its layout for a particular customer with Spinal Muscular Atrophy a neuromuscular degenerative disorder affecting all muscles of the body This design is meant specifically for this customer but could be expanded to any number of individuals with similar needs Since the customer must remain horizontally positioned the stroller has a completely horizontal layout It has the capability of being manually pushed or controlled electronically The user interface for the electronic controls is a single touch button which the user can press whenever the stroller is in the desired directional state There are four directional states which change at a set interval signified by a series of four LEDs Also there is a camera feeding live video of the forward direction to an LCD for the user to safely navigate the device The following report summarizes the final design of this stroller including the major design decisions involved as well as the feasibility of the device from both technical and financial viewpoints Financial estimates reveal a wholesale selling price for the stroller of approximately 14 000 These estimates also show
65. Leading conditions associated with wheelchair use cccccssesssessssssecscssscscscscscsstsesteeseees 123 Table 43 Wheelchair Use by Age Group in 20029 ttt ettet teen 124 Table 44 Home Medical Equipment Market Breakdown eeeeeeeeeereeeeeeeeene 124 Table 45 Annual births per year in the United States for conditions most likely requiring horizontal powered wheelchair s coc cena tied A 124 Table 46 Estimated final prototype budget for Achieving Mobility coooooooooononcnnnncncnnnnnnnnnonononononons 127 Table 47 Total On AOS ie taa ld ln ttl 128 Table 48 Final product BOM for Achieving Mobility ooooocccccncnnnnncnnnnnnonnononononnnnnnnnnnnnnnnnnnonononnnnnnos 129 Table 49 Total assembly time and COS iii easidedi hala a nena 130 Table 50 Selling price Tor the Hole di deis 130 Table 51 Fixed costs associated with forming a new business eeeeeees sese 131 Table 52 Projected net income for first three years of business operation oooooooooonccccnnnnnnnnnnnnnnnnnnnononons 132 Table 55 Estimated cash balance at the nd or each E keine 132 Table 34 Hardware Ww ORK dV 1S1O Mk odis iocus oe oe eed e de dao 133 Table 55 Main software Components asalariado 134 Table 56 Mechanical task DreakdOWHI 2 aateceseccexscvesatessaences dev eun te to ie co ee ERE euo tus aepo ch Ev eate Eee d Eee tua ia peu ER 134 Table 5 7 PCO eC CIVIC TOMES tosco iaa ains 137 1X Table 1 List of Acronyms Acronym
66. O FEV SCA1E 12 w EIC UE SHEELI OHI 5 1 Figure 50 CAD Drawing of Frame Weldment On the sides of the stroller a railing system was designed to prevent the user from rolling off the bed To accomplish this the bed was placed at a slight incline with two aluminum tubes that ran along side of the bed that were horizontal On the sides of the railing brackets were welded on to accommodate a wheelchair accessible canopy Another factor that affected many parts of the frame was component mounting Near the front of the frame the front bearing assembly had to be mounted The bearings are 1 110 OD which meant the ID of the tube had to be slightly larger than this In addition the bearing covers were designed to go over 1 25 OD tube As a result 1 25 OD X 1 120 ID tube was used On the sides of the frame the rim brakes needed to be mounted over the rear wheels so two pieces of round bar were welded on the sides They were placed so that they would be directly over the wheels for proper braking There were also holes drilled in four different locations for the tie down brackets In order for the stroller to be transported tie down hooks were required in four locations where straps in a wheelchair accessible van could hook onto These straps needed to make as close to a 45 degree angle as possible with the ground Figure 51 shows the angles that each tie down strap makes with the ground Team 3 Achieving Mobility Final Report Page 68
67. This test was performed by connecting the touch button and LED lines to the proper external pads of the ECU thus connecting 1t to the proper GPIO pins This was a very simple test to perform but it was necessary for the user interface to correctly display the current directional state of the device The time interval between LED transitions was timed to be exactly 2 seconds as expected This was timed using a stopwatch Also the select button was illuminated for the duration of a button press Finally it was confirmed that the next state following a button press was forward The user interface was also tested with Isaac Postma himself This test was conducted on May 7 2011 at the Projects Night ceremony He was able to press the touch button within the given 2 second intervals His reaction time is fairly slow so this test was to make sure that he would be able to operate the device easily 6 1 1 1 2 Motor Controls The motor control software was tested using the touch button as an input This test confirmed that the motors turn in the proper direction when the button is pressed and also provide the desired acceleration Each of the four possible directions were invoked by the touch button and the response of the motors was observed Also when the button is released the motors were expected to stop The duration of each directional state was also varied between 0 5 seconds to 20 seconds in order to confirm the total functionality
68. Time The total time taken to stop the device was exactly 0 5 seconds therefore this test was a success 6 2 1 3 Emergency Stop Testing It is a real possibility that the user s finger gets tired and he or she does not have the strength to lift their finger off the touch pad causing the vehicle to move continuously in a particular direction Clearly this is extremely dangerous especially if it is stuck in forward or reverse Therefore an emergency stop button was implemented into the system so that a bystander can stop the device The emergency stop scenario was tested by having the stroller move forward at full speed and then turning off the switch The vehicle came to a complete stop in 10 inches When moving backward at full speed stroller came to a complete stop 5 inches after hitting the emergency stop switch When turning either left or right the stroller stopped after 6 inches 6 2 1 4 LCD Camera Testing The camera and LCD were tested by turning on the power to the system and checking to see if LCD and camera are powered on Then a multi meter was used to check if the proper voltage was supplying them Team 3 Achieving Mobility Final Report Page 117 of 160 May 11 2011 The LCD was tested to ensure that there was no distortion in the video while the device was in motion The screen was tested by driving over grass bark mulch and asphalt to confirm that there was no lose in video quality 6 2 1 5 Battery Testing The battery mo
69. Velocity Control Function Configure flags for synchronization Yes No Touch Button Flag Set as full voltage been se Yes Yes as zero voltage been already set already No J Decelerate motors to Y OV Set synchronization flags Clear synchronization flags gt lt v Wait for CAN status data Test for diagnostic errors iagnostic Tests No x Stop motors and notify Passed user Yes umeni reading No Adjust voltage level within thresholds Yes l Display motor status to LCD and send to UART Figure 39 Algorithmic Flowchart of Velocity Control Software Team 3 Achieving Mobility Final Report Page 52 of 160 May 11 2011 The velocity controller first checks the state of the button and then checks status flags to determine if the voltage has already been set to the proper value If the proper voltage has not already been set the velocity controller sets the voltage and sends an acceleration command to the motor controllers If the proper voltage commands have already been sent then the ECU merely sets the proper synchronization flags and moves on In order to determine if there are faults or if the speed needs correction the ECU waits for the motor controller status over CAN As shown in Figure 35 this data is queried every 1 ms
70. a ym 100 Mounting 3 5 f 12 asm ueni 100 ICDSereen 5 18 26 iMs Ai sheAi 100 MowtngDesgn 3 12 6 aaa a s 11 100 MowtngDesgn 2 Assembly 0000000000000 0 d 6 3 15 11 4 30 11 1006 Testing 5 0 0 0 0 0 y y 4 63 70 zii amm 100 HardwareTesting 6 12 13 3 28 41 fasta 100 Thermaltestingonmotorcontoler a 6 f 3 3 30 11 4a 100 Camera LCD 6 10 3 23 11 3 29 11 100 Software Testing 35 3 14 11 0 12 11 100 Speed Control 6 10 3 14 11 3 16 11 100 amiga anm sns aw integration Testing and Debugging 16 20 25 3 22 1 5 1 11 100 Structural Testing 9 8 4 2 981 42shi 100 PowerTorqueTesing a 8 n 21911 s 7Ai 100 A _ A LM DC NN Business Plan 8 1115 0 11 28 0 1006 8 19 15 10 11 18 10 100 Competitor Analysis 16 11 18 10 11 23 10 100 investments 1 6 4 11 18 10 11 23 10 100 A EA e je je je WwW U1 J U1 He IP RN N Uy N N B Marketing Strategy Business Strategy HE Team 3 Achieving Mobility Final Report Page 152 of 160 May 11 2011 12 2 Appendix B Stress Calculations Frame Stress Hand Calculation Figure 98 Simplified model used for hand calculations Givens L 34 75n Length of the longest beam n n E Outer
71. a n d Web 28 Nov 2010 lt http en wikipedia org wiki Brushless_motors gt 28 How Brushless Motors Work HPI Europe n d Web 29 Nov 2010 lt http www hpieurope com walk php lang en amp id 21 gt Team 3 Achieving Mobility Final Report Page 143 of 160 May 11 2011 29 30 31 32 23x 34 39 36 37 38 39 40 41 42 Boley Brian L Overview of Motor Types N p 1996 Web 12 Nov 2010 lt http www oddparts com acsi motortut htm gt Brushless vs Brushed Motors Dynetic Systems N p n d Web 12 Nov 2010 lt http www dynetic com brushless 20vs 20brushed htm gt Rolling Resistance The Engineering Toolbox n d Web 2 Dec 2010 lt http www engineeringtoolbox com rolling friction resistance d_1303 html gt Parts Catalog Invacare 2011 Web 27 Apr 2011 http www invacare com cgi bin imhgprd inv_catalog partsPHII_home jsp s O amp partsHome partsHome amp newSearch trued amp ar ea Main amp WT svl topNavLink3 gt The Every Part Of Your Bike Blog Ed Ryan Jay Tumbir n d Web 17 Apr 2011 lt http beachbikesonline tumblr com post 617679945 the every part of your bike blog gt Mia Moda Cielo Review Strollers and Prams n d Web 17 Apr 2011 lt http strollersandprams com strollers 349 Mia Moda Cielo review 1 html gt Sturmey Archer Dynamo Drum Brake Front Hub 36h Harris Cyclery n d Web 30 Nov 2010 lt http harriscyclery net product sturm
72. also requested that the width of the bed be the same as the customer s current stroller therefore the frame had to incorporate a 13 wide bed Strength was also an issue that affected the overall design of the frame After performing some preliminary FEA tests the team determined that the bed would deflect too much without support bars under it As a result two support beams were added under the bed to increase the strength of the bed Support bars were also placed on the front and rear of the bed to support the weight of the user These Supports were placed at angles that were in 10 increments with the ground for ease of manufacturing as shown in Figure 50 A full size CAD drawing is shown in Figure 50 Team 3 Achieving Mobility Final Report Page 67 of 160 May 11 2011 TEM NO PART NUMBER Default ary Bhort Side Support b otto d C cross Beam 2 hrer eria n A 25 E 8 tearing spacer 1 OD X 1 5 WALL X 2 LG TUBE E A AE E A Pe A Rear slide in Bar lo 2 ij ES 0 42 3 Er gt Foot cross Beam Rear Support 20 J MOD b 2 Y c Bed Support Umbrella Mount 16 brake Mount DETAILS 4 4 12 MUST NOT BE WELDEDTANSENTTO EACH OTHER THERE MIST A 1 16 GAP BETWEEN THEM FILLED WITH WELD THERE MUST NOT BE ANY WELD FOR THETOP 5 16 OF THE GAP SEE VIEW A A FILL GAP WITH WELD UP TO LOCATION i SHOWN SECTION AA SCALE 1 2 FRAME WELDIVIENT SEE DWC N
73. ands Free Bluetooth car audio systems He plans to attend graduate school for computer engineering with an emphasis in software systems after working a couple years in the industry Dan Evans is from Chelmsford MA and will be graduating with a degree in engineering with an electrical and computer concentration Dan has had three internships two at Avid Technology and one at Raytheon Company He spent last summer conducting research on the electrophysiology of the brain Team 3 Achieving Mobility Final Report Page 3 of 160 May 11 2011 lacrimal gland and duct cells at Calvin College He will be attending Worcester Polytechnic Institute in Worcester MA next year as he pursues his PhD in biomedical engineering 2 Project Requirements This section describes the top level requirements for the device It has been divided into functional mechanical and electrical requirements Refer to Figure 2 for a conceptual block diagram of the system 2 1 Functional Requirements The device provides users with the ability to drive themselves around in a stroller while lying down The user will be able to drive the device utilizing a touch button control system with LED directional display as well as an LCD showing video output from a camera in the front of the stroller The stroller will move in the desired direction of the user as long as the touch button is engaged Whenever the touch button is not engaged the device will be stationary The stroller s
74. appropriate PWM duty cycle and outputs the correct voltage to the motors This CAN message is a 32 bit 16 16 signed fixed point number The desired output voltage to the motors is a configuration variable that will be set at a level the produces 2 5 mph but can be changed by a technician if necessary see Figure 3 For the scope of this project this technician would need to recompile and load the software onto the ECU thus the technician would need to be a member of the team Alternative methods for speed configuration would be sought if this design were mass produced Since the voltage should not be set instantaneously to its final value a voltage ramping function was used to increase the motor voltage at a constant acceleration This acceleration is also fully configurable by a technician with a similar method as the maximum speed 4 2 4 3 Algorithm The algorithm for controlling the motors using voltage control has been fully developed This algorithm has the structure of an infinite loop continuing until the power is turned off The device must be disabled by cutting the power to the device via manual shut off The following flowchart Figure 39 shows the velocity control function which resides in the ECU Team 3 Achieving Mobility May 11 2011 No Accelerate to full voltage based on current directional state Clear synchronization flags Final Report Page 51 of 160 Enter
75. ate timely data Causes of bandwidth reduction and increased latency would be mainly due to poor timing of CAN signals being sent If two messages are sent at the same time the one with a higher priority will be sent and the other will be ignored Therefore efficient timing of these messages will be important in order to reduce the latency in the CAN bus The CAN interface provides a large number of possible commands grouped together by the type of command The types of commands include broadcast messages system level commands motor control commands configuration commands and status information The message fields for a CAN message on the RDK BDC24 board is shown below in Figure 10 Byte 3 Byte 2 Bvte 1 Byte 0 aS OES SS A AAA Device Type Figure 10 CAN Message Identifier Fields Bits 6 through 10 represent the API class and the second 4 bit field is the API index which determines the particular API class to use Refer to 5 for more information on the CAN protocol with this device 4 1 7 Power Supply The power source of the stroller is a fundamental design decision The power supply provides electrical energy to the device Most of this energy is converted into mechanical energy through the motors in order to move the stroller but some is needed to power the MCU LCD and camera The battery is rechargeable so that the customer does not need to continuously buy new batteries for the device The charger can be used in the hom
76. atteries with the electronics turned on this could approach 30 V although testing showed that it only reached 28 V Regardless if the batteries were to have a voltage of 30 V or above the ECU would beep loudly until the voltage dropped back to a normal level 1 e until the user disconnects the battery charger from the batteries The ECU cannot control the amount of voltage being delivered to the motor controllers so this was the best solution Again this error will most likely never be encountered because the batteries should never be charged about 28 V even being connected to the charger 4 2 4 Velocity Control A mechanism for controlling the speed of the device 1s necessary for the safety of the user The speed requirement states that the stroller must not exceed 3 5 mph for an incline ranging from 10 to 10 This velocity control was accomplished through an closed loop voltage control system which will be described in the following section It should be noted that this approach is substantially different than that which was described in the team s PPFS This change was made as a result of factors discussed in section 3 2 2 4 2 4 1 Theory A fundamental theory of DC motor control is that voltage 1s directly proportional to motor speed and current is directly proportional to motor torque Therefore since the goal is to maintain a steady speed setting the input voltage of the motors to a constant value makes the motors run at a constant sp
77. ble for past designs For the delivery design the wheels from Isaac s old Power Tiger wheelchair will be used Re using these not only saved money but also guaranteed that the wheels were compatible with the motors since the new motors are nearly identical to the old motors The rear tires are Power Express tires made by PRIMO model number 62 203 and they are 12 1 2 in diameter and 2 1 2 wide The tread is still good Team 3 Achieving Mobility Final Report Page 70 of 160 May 11 2011 on them and they should easily be able to last the life of the wheelchair The front tires are also Power Express and are 6 in diameter and 2 wide The front tires are mounted to a front bearing assembly that is attached to the frame therefore this front bearing assembly was designed to accommodate the front wheels 5 3 Storage The storage compartments are used to store the batteries and all medical equipment The storage shelves provide enough space and strength to support all equipment 5 3 1 Requirements 5 3 1 1 Storage Space The main storage requirement is that there is sufficient space for storing batteries medical equipment and electronic components The batteries are enclosed in plastic encasings to keep them separate from all other equipment The final design uses two lead acid batteries which are stored underneath the stroller In addition there needs to be sufficient storage for the customer s suction device and feeding bag
78. built and used to test on the customer Unlike the multi pivot system the strap in arm rest requires a separate storage design A shelf a hook and a magnetic system were all considered for storing the strap in arm rest underneath the bed The magnetic system was the chosen design because it took up the least amount of space underneath the bed limited space with all the other components underneath the bed and it provided the most stability while the wheelchair was in motion do not want the strap in arm rest vibrating or falling out during operation Figure 86 shows the design of the strap in arm rest storage device mounted underneath the bedding Figure 86 Magnetic strap in arm rest storage device 5 8 7 Financials Table 29 shows the projected costs of the production design Table 30 shows the costs of the delivery design Team 3 Achieving Mobility Final Report Page 107 of 160 May 11 2011 Table 35 Production design costs for the mounting systems Production Design Cost Rim Brake Mounting 0 00 Included in material cost of frame Parking Brake Mounting 0 00 Included in cost of parking brake 0 00 Included in material cost of frame e 0 585 OD heavy stiffness 18 length l flexible tube 20 00 PEDD eei Mounung 20 00 e Aluminum sheet included in material cost of frame Encasing Mounting 0 00 Only requires bolts which is covered under miscellaneous 0 00 Included in material cost of frame e PVC pipe 5
79. button only a limited number of directions can feasibly be achieved If the user s means of control were something more versatile like a joystick there would be many more possible directions Allowing any more possible directions would significantly Team 3 Achieving Mobility Final Report Page 46 of 160 May 11 2011 increase the complexity of the control system from the user s perspective and would increase the potential latency between direction changes The directional control system consists of an event driven state machine Since there are four possible directions available to the user there are four directional states The events that drive the current directional state are a timer and the current state of the touch button Figure 36 shows the state diagram for the directional controller As shown by the state diagram as long as the touch button is not being pressed i e Button from Figure 36 the timer controls the current state of the system such that each state 1s visited for two seconds before transitioning The transition time of 2 seconds is adjustable by a technician but this option is not available to the average user for safety reasons Whenever the touch button is pressed the current state of direction controller software is held until the button is released Upon release the device comes to a complete stop before a new command issued by the user can be registered by the software O H t Timers 0 Fi
80. by the MCU making it necessary for the MCU to be able to interface with the peripherals The MCU also serves as a bridge between the UI and the desired output of the device As mentioned in section 4 1 1 the selection of the motor controllers dictated the selection of MCU for the delivery design The motor controller kits include the Luminary Micro Cortex M3 microcontroller specifically the LM3S2616 so this particular MCU was be used in the delivery design This MCU operates at 50 MHz with 256 kB Flash memory and 96 kB SRAM It includes all the necessary features such as PWM outputs and extensive serial interfaces A block diagram summarizing the features of this MCU is shown below in Figure 8 retrieved from Luminary Micro s product brief 256 KB Flash ARM Cortex M3 96 KB SRAM 80 MHz ROM Clocks Reset 3 LAR Te System Control I ET E Timer 4 Timer waa Coe Pach 3 0 21316 bE A Timers cn GPIOs E Erga ri EPI ium z z E z l s or o Precision Oscillator E Battery Bnekced Hibernate 2 Quadrature Encoder Inpubs I LOO Voltage a PWM utputa Regulator Tuer 3 Anale i Camper alters Camps FAO FR 2x 10 bit ADC Each B channel 1 Mapa DO TIVNV PM PWM A jE aT MOTION CONTROL Dead Band j mirador Temp Sensor Figure 8 Block Diagram of LM3S2616 MCU Team 3 Achieving Mobility Final Report Page 24 of 160 May 11 2011 Also the selection of developmen
81. cation Protocols ni a 25 4 1 7 PARTE dt ti 25 4 1 8 Power Re claudia dia 29 4 1 9 Connectors did Cables riera Me CA DER CEE 40 LE A E E D e E e i EEEE A nena tact E ees tam sd ec atau bores no nea auras EEA 42 E ENT AAA II 43 ELI Motor Control Hard Ware a eee AG leo 43 4 2 A A Re rene Lic uu I E LR Seen T ene 44 4 2 1 Cyclic cBxecutrye Architect idos 44 4 2 2 Dire cional AO a a 45 4 2 3 Didanostes ia pcc E 48 4 2 4 Velocity COMTO sodes e e M M E ML DE E ILI 49 4 2 5 DEVICE CES oe diccionarios Dens EIE 53 4 2 6 DEDUS SINE iS iS PER 53 4 2 7 IRE AE o a E is od 56 4 3 Future Work Production DeSISTI ranei A A da ERR PPuCEDULe I DiDdS 23 4 3 1 la iren cs e II 57 4 3 2 Motor Control Circa diia 58 4 3 3 IVICU SC CO oct shits cence nade edie RN 59 iii 5 4 3 4 PCB DONT 59 Mechan Al BD MR ET E Um 60 5 1 Jui PPS RETE 60 5 1 1 nda Vin HIEDES AA A A OE Eo em M D MUS NU I LEA 60 5 1 2 Karera Se CCU Ol ddr aui eru RE 61 5 1 3 IVT CSA SZ see tetas t cocenses Jeqcsccnc eters Rc Eus dde tac Lene ses anes cea Nnm tee onus este c eA UL E 64 5 1 4 Po DO A O O bad aee tectessctoteecis 64 5 1 5 A A CODI AS Lee 68 5 1 6 OG CU RE 69 3 2 VTS AAA A T 69 5 3 O ana oce Ua Mx Moles MMC mE EM IER CM casu RUE 70 5 3 1 REQUIEM ENS ad D UT NU EIN 70 213 2 SS A E tel lea ra racist cm a a le OT 70 5 3 3 PUVA CLAS A Mcd EE 71 5 3 4 A 71 5 4 li T 72 5 4 1
82. compatible canopy Team 3 Achieving Mobility Final Report Page 5 of 160 May 11 2011 2 3 Electrical Requirements 2 3 1 Motor The motors shall be powerful enough to provide the necessary torque to move the wheelchair from rest for the maximum combined weight requirement of 450 pounds This includes enough torque to travel up a ramp with the maximum incline requirement of 10 The minimum required motor power based on the necessary torque 1s 0 295 horsepower Appendix C Motor Power Speed and Torque Calculations The motors also require an emergency braking mechanism in case the electronic brakes fail The motors must also be able to provide a minimum angular velocity of 2 259 rpm Appendix C Motor Power Speed and Torque Calculations Finally the motors must have the capability of being disengaged in order to provide the possibility of manual operation 2 3 2 Power The battery provides the power to the system The customer has communicated to the team that the product will be primarily used during recess at Isaac s school and after school activities and should not be needed for more than two hours The battery needs to provide enough energy to the device so that it is operable for over two hours when the device is traveling at full speed for a quarter of the time with maximum occupant weight The battery is recharged by use of a standard American outlet in less than 8 hours so that the batteries will have completed recharging while the c
83. control system e The Business 396 team which consists of Ji Won Choe Audrey Petrini Will Richert Theo Van Hoek and Peter Wierenga for providing the team with help on the business plan and financials e Josh Schroyer for assisting the team in making videos for the RESNA competition e Mary Free Bed Rehabilitation Hospital for designing and fabricating the bedding and restraint system e Johnson Controls Inc for fabricating the power regulation PCB 10 Conclusions The final product designed for this project was a success The team successfully designed and built a working prototype for Isaac Postma that he can operate electronically while lying down Several difficulties were encountered along the way such as finances but through the generosity of several organizations this project has become a reality The delivery design was specifically designed for Isaac Postma however if the product were to hit the market it would be modified to meet the needs of more individuals The group made substantial progress throughout the year and has completed the final project by the end of the school year Throughout the course of the year the team has logged in approximately 1800 hours on this project This product has proven to be a great example of how we can use God s gifts to make a difference in someone s life Knowing that this product will touch someone s life and give them the opportunity to see life in a new perspective is what makes this pr
84. ction criteria The first notable difference between the production design and delivery design would be consolidating all the motor control hardware onto a single board This way a single MCU could be used to control two motors independently instead of two Most of the analog components could not be consolidated for instance there would still need to be two H bridges one per motor and there would still need to be two voltage and current sensors Also a single MCU could most likely handle the directional state machine and the user interface Testing would be conducted to make sure that controlling the user interface along with both motors did not cause the MCU to become over utilized 1 e over 90 utilization A simple block diagram of the proposed production design topology is shown below in Figure 44 As the figure shows there is only one board including a single MCU and all the analog hardware needed to run the motors Team 3 Achieving Mobility Final Report Page 58 of 160 May 11 2011 Touch Button Input LED Output MCU H Bridge Power Regulation m Voltage Voltage Current Current Sensors Sensors A A Motor 1 Figure 44 Proposed Production Design Hardware Topology 4 3 2 Motor Control Circuitry The H bridges are one of the main pieces of analog hardware that would need to be designed in the production model A schematic of a typical H bridge circuit
85. ctronic Component Encasings eese 93 Table 29 Production design costs for the encasing systems occcccccnnnnccnnnonnnnonnnnnnnnnnonnnnnnnnnononnonnnnnnnnnnnnnnss 95 Table 30 Delivery design costs for the encasing systems oooccccccnnnnnncnnnnnnnnnnnnnnnnnnnnnnnnnnnnnonnnononnnnnanannnnnoss 96 Table 31 Component Mounting Type ii ii ii e dais 96 Table 32 Calculated bending and torsional stress in the brake support arM ooocccccncnnnnnnncnnnnnononanannnnnnnoss 97 Table 33 Comparison of LCD screen mounting SYSteM oooooooononncncnnnnnnnnnnnnnnonononnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnos 102 Table 34 Comparison of the two touch button mounting design alternatives oocccccccccnnnnnnnnnnnnnnnnnno 105 Table 35 Production design costs for the mounting systems oooccccccccnnncncnonnnononnnnnnnonnnnnccnnnnnnnnnnnnnnnnnnnnns 107 Table 36 Delivery design costs for the encasing systems ooooooonnnnccnnnnnnnnnnnnnnnnnonononannnnnnnnnnnnnnnnnnnnnnnnnnnnnnns 108 Table 37 Directional Soltware Test DII id 110 Table 38 Time Trials to Calculate Speed of Stroller oooooonnnnnnccnnnnnnnccnnnonoonononnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnos 115 Table 39 Comparison of Performance of competitive models gt 00 77st ettet tenete nt 121 Table 40 Comparison of features in competitive models 90 acciciciciciciconinononononononononononononononos 121 Table 41 Cost comparison of competitive models 9 eere eere 122 Table 42
86. d by this design is also feasible see O for more detail on the hardware topololgy This limited the selection of kits to those using a Stellaris Cortex M3 MCU which was also not a severe limitation because this is a very widely used and well documented microcontroller Also the development kit used for the ECU needed to have all the necessary CAN hardware such as CAN bus connections for the protocol s Physical Layer compatibility and a CAN transceiver and controller for protocol s Data Link Layer compatibility Finally the development kit needed to have expansion GPIO ports for interfacing with the touch button LEDs and character LCD After researching all potential options for the ECU there was only one option that fit all of the above requirements The selected development kit was the LM3S2965 CAN Board by Luminary Micro This board uses a Stellaris Cortex M3 MCU and uses the same driver library as the motor controllers This board has all the necessary CAN components for complete compatibility with both the CAN protocol and the motor controllers Finally this board has a full set of expansion headers which allow for easy external peripheral interfacing A labeled picture of the LM3S2965 CAN Board is shown below in Figure 6 As Luminary Micro continues to improve their lines of MCUs a concern could be that the board would not be replacable in case it ever broke However newer lines of Stellaris MCUs such as the upcoming Stellaris Cor
87. d into two categories fixed and adjustable as shown in Table 31 Table 31 Component Mounting Type Bakes X Camera X Ps X p X Mer X Touch Button x 5 8 1 Brakes In order for rim brakes to function properly they must be centered over each wheel Based on this requirement a support arm was built out from the frame structure to mount the rim brakes over the rear wheels The support arm will extend out 2 3 inches from the frame and consist of a 0 75 inch diameter solid bar for better support during braking To determine 1f this design would be able to safely handle the braking load both the maximum bending stress and the maximum torsional stress were calculated see Appendix G Brake Mounting Stress Calculations The stresses in the welded joint were also calculated using the computer program Autodesk Algor see Appendix G Brake Mounting Stress Calculations The stresses calculated were far below the yield strength of the material making the design suitable as shown in Table 32 Team 3 Achieving Mobility Final Report Page 97 of 160 May 11 2011 Table 32 Calculated bending and torsional stress in the brake support arm E Bending Stress Torsional Stress Bending Stress in Weld Calculated 12 0 ksi A PO The designed support bar along with rim brake can be seen in Figure 71 Figure 71 Mounting of rim brake for delivery design The parking brakes are mounted dir
88. d was connected to the chassis of the device so that if for some reason the user touches a live wire and ground the current will travel through the metal chassis of the device instead because it has a lower resistance than the human body 4 1 6 8 EAGLE Schematic By creating the schematic in EAGLE it becomes much easier to make the PCB layout since the software shows the creator where all the connections need to be The schematic in EAGLE is shown below in Figure 23 Team 3 Achieving Mobility Final Report Page 37 of 160 May 11 2011 Figure 23 EAGLE Schematic 4 1 8 9 EAGLE Board Design The final design of the printed circuit board created in EAGLE is shown in Figure 24 Figure 24 EAGLE Board Layout 4 1 6 10 Revised EAGLE Schematic The revision was made to the original and several changes were made Diodes were added in order to protect components from someone accidentally mixing up the hot and ground A switching diode replaced the first LM350 and large heatsink in order to increase efficiency Over voltage protection was added at the 12 V 9 V and 5 V nodes The mounting holes were increased to an eighth of an inch The spacing was improved between components such as capacitors and voltage regulators The inner diatmer of the pads where the wires connected were increased so that the preferred wires could fit through the pads The holes to the LED were increased because the LED s terminals were larger than expec
89. delivery design see 83 Comparison of foot lever and hand lever for attendant controlled braking system 85 Comparison of different braking mechanisms for attendant controlled brakes 86 Rim brakes and hand lever used in final product ooccccccccncccccncnnncnnnnnnonnnnnnnnnonnnnnnnnnononnnnnnnons 87 Donated parking brake being used in delivery design esses 88 On off toggle switch for emergency braking ooooooonnnnnncnnnnnnnnnnnnononnonnnoonnncnnnnnnnnnnnnnnnnonononnnos 88 Resulting temperature of PCB due to natural CONVECTION seen 92 Resulting temperature of PCB due to forced CONVECTION sese 93 Motor controller ECU and power regulation PCB encasing delivery design 94 Donated battery encasing for final product oooooonnnnnncnncnnnnonnnoncnnnnnnnonononnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnos 95 Mounting of tim brake Tor delivery desist iu A A 97 Parkins brakes Mounted onto tame ca ask 98 First desien Of camera mounting SM A it acia 98 Second design of camera mounting SYStOIM occcccccccnnnnnnnnnonnnonnnnnnnnnonnnnnnnnnnnnnnnnnnnnnnrnnnnnnnnnnnnnnnnss 99 Prototype Of second Camera destbn uie coder trie nenna eade pe Co a nue ee tat ainiin es 90 Second prototype of third camera mounting system sseessseseeeeeeeeeeeeeeeeeeeeeeennnnnnnnns 100 Final camera mounting system for final produCt
90. des the possible problems with the LCD there were no other issues with components not functioning correctly after being exposed to the rain 6 2 2 3 Portability Testing The portability test was conducted to make sure that the stroller was compatible with a wheelchair accessible van The customer s current stroller is loaded into a van using a standard wheelchair lift and then locked down with four tie down straps that attach to the stroller This test verified that the stroller could be loaded into a wheelchair accessible van using a standard wheelchair lift Even though the wheel base for this stroller is much longer than most wheelchairs it is still short enough to fit on a hydraulic wheelchair lift The longest wheelbase that the customer s wheelchair accessible van can handle is 40 which is much longer than the wheelbase of the final design This test was performed at Calvin College with the family s wheelchair accessible van The first test loaded the stroller into the van using the hydraulic lift There were not any problems during this test due to excess weight or length of the stroller Once the stroller was in the van it was locked in using the four tie down hooks and the parking brakes were engaged The stroller was rocked back and forth without any movement and the family was very happy with how solid the stroller was secured Another test was then performed with Isaac in the stroller completely fastened using the restraint system
91. direction The schematic of the motor controllers was analyzed and the motor controllers contain the flyback diodes therefore flyback diodes were not added into the team s power regulation circuit 4 1 9 Connectors and Cables There are several different types of cables and connectors needed throughout the device 4 1 9 1 USB Updating the ECU with the newest software is done using a Universal Serial Bus USB cord The USB cord also allows for the use of UART through a virtual COM port UART was used for real time status updates to a terminal while testing the stroller This is primarily used to monitor the voltage and current there are applied to each motor It also allowed the team to measure the temperature of the motor controllers These measurements allowed the team to adjust the software appropriately Refer to 4 2 6 for further detail on UART debugging 4 1 9 2 CAN 6 Pin Connector Cable CAN is a bus standard that allows the MCUs to communicate with each other on a single bus The product has one motor for each rear wheel and each motor is controlled separately The two motor controllers need the ability to communicate with the ECU in order for the device to maintain accurate speeds The CAN vehicle bus standard provides this capability with a CAN 6 pin connector A diagram of the CAN connector is shown below in Figure 29 CANH CANL GND RS232 CL AN Socket Viewed from Top Tab down Figure 29 CAN Socket Team 3 Achieving Mob
92. e Table 19 and Table 20 shows the projected costs of the production and the cost of the delivery design respectively Table 19 Production Design Costs for the Motor Production Design Cost Wem ost Sum 2 PMDC 1 600 Based on Invacare s Nutron R51 Powered Wheelchair motors 0 5 hp 2 pole PMDC motor and gearbox assembly 800 00 each Mounting 500 00 Nuts bolts washers etc 00 Mounting motors connecting to batteries and Labor controller attaching wheel 1 hr 50 hr 1650 ses OSS Team 3 Achieving Mobility Final Report Page 84 of 160 May 11 2011 Table 20 Delivery Design Costs for the Motor Delivery Design Cost liem 2 PMDC 1 600 Based on Invacare s Nutron R51 Powered Wheelchair motors 0 5 hp 2 pole PMDC motor and gearbox assembly 800 00 each Labor Contingency 5 6 Brakes 5 6 1 Research Based on research of power wheelchairs produced by Invacare Pride Mobility and Permobil the trend in the market is using an electronic braking system that slows the motor by decreasing the current to the motor The majority of powered wheelchairs do not have a hand brake system for manual drive To find information on hand brakes manual strollers had to be researched Manual stroller applications that employ a hand brake generally utilize a disc type braking mechanism The customer s current manual stroller produced by Thomashilfen has a hand brake that operates using drum brakes
93. e ARM Cortex M M3 based 32 bit LM3S2616 MCU with CAN UART and advanced motion control capabilities It also includes the analog components necessary to run DC motors such as an H bridge voltage regulators and a current shunt monitor The kit includes an extensive driver SDK hardware design files and software design files The RDK BDC24 uses CAN to communicate between individual modules 4 1 1 1 3 DM163029 This is a Picdem Mechatronics development kit from Microchip which serves as a learning tool for motor control systems It makes use of the PIC16F917 MCU as well as LEDs an LCD display switches and a built in DC motor It also includes all the necessary analog amplification components necessary and is easily customizable for a vast number of applications 4 1 1 1 4 DE2 This is a general purpose development kit from Altera that is based on the Cyclone II FPGA The soft core processor used on the FPGA is the Altera Nios 2 processor This FPGA based MCU provides extra flexibility and expandability to any design It also includes a vast array of peripherals such as LEDs an LCD screen buttons switches UART USB and VGA This development kit is familiar to students of Calvin College Electrical Engineering but it 1s not specifically designed for digital motor control Team 3 Achieving Mobility May 11 2011 Final Report Page 20 of 160 4 1 1 2 Selection The final decision was to use RDK BDC24 as a development kit for mot
94. e PCB encasing and the battery encasings respectively Figure 69 Motor controller ECU and power regulation PCB encasing delivery design Team 3 Achieving Mobility Final Report Page 95 of 160 May 11 2011 Figure 70 Donated battery encasing ing for final product 5 7 4 Financials Table 29 shows the projected costs of the production design Table 30 shows the costs of the delivery design Table 29 Production design costs for the encasing systems A Design Cost Based on pricing of Invacare battery M Encasings 41 84 encasing 20 92 each DC 96F DC Series Heavy Duty eS 12 64 Electronics Enclosure 10 x 6 x 3 5 00 Nuts bolts washers etc Mounting and connections JE MNA O Team 3 Achieving Mobility Final Report Page 96 of 160 May 11 2011 Table 30 Delivery design costs for the encasing systems Delivery Design Cost liem Battery Encasings 0 00 Donated Motor Controller Encasings 0 00 Donated ECU Motor Controller Power Regulation Encasing Mounting 0 00 Nuts bolts washers are free from metal shop Labor 0 00 Donated Contingency 0 00 10 0 00 0 00 Free aluminum sheet from metal shop 5 8 Mounting Mounting consists of determining how each component will be attached to the frame of the wheelchair The main components that require mounting are the camera PCBs brakes LCD screen touch button and the motors The mounts for these components can be divide
95. e capital investment a bank line of credit of 150 000 will be taken out at an annual interest rate of 6 and yearly invested capital will also be 150 000 The bank line of credit will be paid off by year 3 based on projected sales Based on this initial capital investment and net income at the end of each year the net cash balance can be calculated and is shown in Table 53 It is also assumed that as the company grows it will purchase more equipment to increase the rate of production and decrease renting costs Table 53 Estimated cash balance at the end of each year Year 2 Beginning Cash Balance 16 200 99 571 Net Income After Tax 279 515 55 289 229 172 Depreciation expense 715 4 083 12 918 Invested Capital Equity 150 000 150 000 150 000 Increase decrease in borrowed funds 150 000 106 000 56 180 Equipment Purchases 5 000 20 000 50 000 Ending Cash Balance 16 200 99 571 385 480 Team 3 Achieving Mobility Final Report Page 133 of 160 May 11 2011 8 Project Management Project management is an essential part of this project that must be maintained throughout the year Clear goals and a structured schedule were very important in completing the project on time as well as staying within the budget Project management has been divided into three main categories work division team organization and overall schedule 8 1 Work Division Since this team 1s split between two concentrations dividing up the work betw
96. e of the customer and it is not designed to travel with the device while it 1s in use because it would add extra weight as well as occupy necessary space Team 3 Achieving Mobility Final Report Page 26 of 160 May 11 2011 4 1 7 1 Requirements The stroller shall be operable for at least two hours of continuous use The power supply must be able to provide the adequate amount of energy to all of these components simultaneously A chart of the amount of power necessary for each component is shown below in Table 6 Table 6 Typical Power Values of a emm votar cre reset tum Tower 4 1 7 2 System Current Draw The system requires a battery to supply a current that standard batteries e g NiMH NiCad are not capable of The motors alone require an average of about 5 A each Therefore a battery with large ampere hours is needed so the device can be operable for at least two hours of continuous use A power supply with 50 Ah was used in the Power Tiger wheelchair The stroller however is heavier and less efficient and operable for a shorter time than the Power Tiger Therefore 55 Ah is desired for the stroller in order to err on the cautious side 4 1 7 3 Battery Type There are several options for the type of battery Although reaching the minimum specifications of 24 V and 50 Ah is a necessity the team is also seeking to reduce the physical size and weight of the battery as much as possible A standard car battery cannot be u
97. e overall functionality of the parking brakes as a part of the entire system These brakes are required to hold the wheelchair in place on the maximum rated incline 10 with the maximum rated weight 200 pounds with no slippage or sliding For this test the wheelchair was loaded with 230 pounds and manually held on a 30 gravel decline Each parking brake was then engaged and the wheelchair was let go of The wheelchair stayed in place with no slippage or sliding therefore passing this test da Business Plan The following chapter discusses the different aspects involved in a business plan for this project and the end product 7 1 Business Models This section describes the possible business models that could be developed around the product Two different models will be examined each with their own advantages and disadvantages 7 1 1 User This model involves manufacturing the product and selling it directly to the end user The advantages of this model include the possibility of a higher margin and more customized products better satisfying the needs of the customer One disadvantage is that it would be difficult to get insurance companies to back the product in this model Without the help of the insurance companies usually cover between 50 and 80 most users will not be able to afford the product Also hospitals and doctors will be unaware of the product and will not be able to recommend to a patient to buy from us Setup costs are
98. e wheelchair s market revenue comes from the actual end user This further shows that the main market target should be insurance companies and doctors hospitals Team 3 Achieving Mobility Final Report Page 126 of 160 May 11 2011 Figure 94 Breakdown of revenue source for wheelchair market in 1996 7 5 Project Financials The final design of Achieving Mobility is innovative and offers a product to the market that could help many people As with any new business or product a financial analysis must be analyzed to see 1f the product has any potential to yield a profit for a company The following sections go through a detailed financial analysis of what 1t would take to make this product profitable 7 5 1 Delivery Design Budget For this class the team designed a delivery design for a specific individual Isaac Postma The costs for this design differ greatly from the production design due to donations from companies and organizations In addition several components were salvaged from old wheelchairs to keep the final cost within our budget Table 46 shows the delivery design budget of all components that have been discussed in depth in previous sections of this report All raw materials already have a volume discount factored in because they were purchased through Ebling and Son Inc Blacksmiths which already gets a volume discount Team 3 Achieving Mobility Final Report Page 127 of 160 May 11 2011 Table 46 Estimated final prot
99. ecreasing the current into the motor therefore slowing the wheelchair down to a stop includes required 10 decline This method can be implemented through software and would provide smooth efficient braking This system is feasible because the user interface works around the user s physical limitations An electronic user controlled braking system was selected for the both delivery and production design because it is regulated by a controller limiting the user s work it provides smooth effective braking and it can be implemented with various user interfaces depending on the physical limitation of the user The speed of the PMDC motors being used for the delivery design is directly proportional to voltage as voltage decreases the speed of the motor decreases and vice versa This relationship will be used to slow down and stop the wheelchair When the user de presses the touch button the voltage 1s slowly cut from the motor and the wheelchair slows down and stops 5 6 3 2 Attendant controlled Braking System When the motors are disengaged and an attendant is manually pushing the wheelchair a brake system controlled by the attendant is required to slow stop the wheelchair This system involves two main components the attendant interface and the braking mechanism A hand lever and a foot lever were both initially considered for the attendant interface part of the attendant controlled braking system Figure 62 A foot lever however was
100. ectly to the aluminum tube of the frame just in front of each wheel The parking brakes are built to fit snug around the tube and bolt into the tube The clearance hole for the bolt 1s large enough so the horizontal position of the parking brake can be adjusted 1f 1t 1s too close or too far from the wheel Figure 72 shows the parking brakes mounted to the frame Team 3 Achieving Mobility Final Report Page 98 of 160 May 11 2011 Figure 72 Parking brakes mounted onto frame 5 8 2 Camera A camera needs to be mounted to the front of the frame to provide the user with a view of where he or she is going The camera also needs to be adjustable to provide the user with the option of altering their view if necessary The camera will therefore be able to be adjusted in the vertical position a maximum of 30 The camera will be mounted to the bottom of the bed which has a 7 incline for this reason the total 30 will be split 10 upward and 20 downward net angles from horizon are 17 upward and 13 downward The initial design of the camera mount is shown in Figure 73 A mounting plate with an arc hole will allow an arm holding the camera to rotate about a fixed pin 30 When the desired location is found the pin through the arc will be tightened by the user fixing the camera arm in place Figure 73 First design of camera mounting system Team 3 Achieving Mobility Final Report Page 99 of 160 May 11 2011 After talking with the custo
101. eed The issue arises when an external torque is applied to the motors such as traversing an incline or decline When this happens the voltage stays constant but the speed of the motors decreases as a function of the amount and direction of external force applied Also when an external force is applied the current flow through the motor windings changes because the force causes torque on the motor shaft Therefore voltage control cannot maintain a constant speed making the current sampling necessary Since the motor controllers have built in current sampling capabilities the ECU can periodically query the current of each motor and make decisions based on these values For instance if the motor current is above a certain threshold it can be assumed that the motor is experiencing an opposing force and therefore the voltage and thus the speed should be increased Conversely if the motor current is below a certain threshold it can be assumed that the motor is experiencing a helping force such as driving down a hill and therefore the voltage should be decreased to slow down the motors A simple block diagram of the closed loop voltage control system is shown below in Figure 38 Note that the correction signal C s is actually a voltage that is delivered to the motors producing a current Y s Then the motor current is sampled Q s and read by the ECU Team 3 Achieving Mobility Final Report Page 50 of 160 May 11 2011 Desired Curre
102. eeeeeeeeesesees 39 Figure 28 Final Version of Power Regulation Circuit sessi 39 Figure 29 CAN Socket AN 40 Poore SUM ately ODD LORS iei tune Diod Iba A a aaa 4 Figure 31 Battery Charger CONOCIO estrias aia que Rue AIE nc Dua oso RO oia Ro EDU E 4 151066 22 Motor Power Connect ita 42 Figure 33 Liquid Crystal Display Monitor ccccceceeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeees 43 Figure 34 oia lea diro ac OT OO LISSE 44 Figure 35 Figure 36 Figure 37 Figure 38 Figure 39 Figure 40 Figure 41 Figure 42 Figure 43 Figure 44 Figure 45 Figure 46 Figure 47 Figure 48 Figure 49 Figure 50 Figure 51 Figure 52 Figure 53 Figure 54 Figure 55 Figure 56 Figure 57 Figure 58 Figure 59 Figure 60 Figure 61 Figure 62 Figure 63 Figure 64 Figure 65 Figure 66 Figure 67 Figure 68 Figure 69 Figure 70 Figure 71 Figure 72 Figure 73 Figure 74 Figure 75 Figure 76 Figure 77 Figure 78 System Boot and Cyclic Executive Software Flowchart oooccccccccncccccnnnononnnooonnncnnnncnnnnnonnnoss 45 Event Driven State Machine Software Flow Diagram cccccsseeeecceceeeeeeeeeeessssssseeeeeeees 46 Algorithmic Flowchart of Button Debouncing Process esses 47 Block Diagram of Closed Loop Voltage Control System esee 50 Algorithmic Flowchart of Velocity Control Software
103. een all four members of the team was important in order to stay organized and on schedule At the beginning of each week certain tasks are laid out for each team member to complete over the course of the week Then at the end of the week status reports were utilized to analyze the work completed as well as what needs to done the following week The three main areas that the work has been divided into are hardware software and mechanical Even though certain people are assigned to specific tasks there will be some overlap between tasks and tasks that require more than one person to complete 8 1 1 Hardware The hardware components were split between Dan Evans and Matt Rozema since they were the two electrical engineers on the team The components were split up based on different skills between the two electrical engineers and their varying abilities Table 54 shows how the electrical hardware components will be divided between Matt R and Dan Table 54 Hardware work division consta na LCD Selection Motor Control Hardware Power Regulation PCB Power Supply Selection 8 1 2 Software The majority of the software components were handled by Matt Rozema Matt has experience working with software design from an internship and feels that he will be able to handle the design of the software components The main software components are shown in Table 55 Team 3 Achieving Mobility Final Report Page 134 of 160 May 11 2011 Table 55
104. egulation circuit is used to ensure that each component is receiving a sufficient amount of power The circuit also protects against damaging levels of current and voltage as well as unwanted noise The circuit has been fabricated on a printed circuit board A schematic of the power regulation circuit in LTSpice is shown in Appendix D LTSpice Power Regulation Circuit 4 1 8 1 Voltage Regulation The motor controllers motor brake releases LCD screen camera and ECU all need their own voltage level The motor controller is powered directly from the battery because they use approximately 24V Team 3 Achieving Mobility Final Report Page 30 of 160 May 11 2011 They should not have more than 30 V and may become damaged at 35 V When the batteries were first purchased and fully charged they put out 26 V There 1s no reason to believe that the total voltage should be higher than 30 V let alone 35 V The motor brake releases need more than 18 V of initial voltage in order to unlock the brake therefore the motor brake releases are also connected directly to the batteries The LCD needs 12 V the camera needs 9 V and the ECU needs 5 V The battery provides between approximately between 19 and 26 V depending on the state of the charge In order to provide all the components with the correct voltage several voltage regulators were utilized and they are set up in cascade for efficiency reasons The LM317 was used as the voltage regulator device A diag
105. ely important if Isaac s finger gets tired and he is not able to lift it up from the touch pad which has happened in his previous wheelchair Also 1f the software malfunctions causing the stroller does not stop an attendant who notices this can flip the switch so that the stroller stops The switch can be seen in Figure 19 Figure 19 Power Switch used on final product 4 1 8 6 Filter Capacitors Filter capacitors are at each node in the circuit in order to reduce noise from the surroundings The wires which carry DC voltages will have ripples or spikes in it from the surroundings A capacitor can even out the voltage by absorbing peaks and filling in the valleys An electrolytic large value filter capacitor of 100uF is added in order to smooth the DC and remove unwanted lower frequency noise Small value non polarized ceramic capacitor of O 01uF is more effective in shorting very high frequency noise spikes to ground In order to prove that each capacitor is affecting the node differently a basic circuit was implemented in LTSpice The circuit is shown below in Figure 20 Team 3 Achieving Mobility Final Report Page 35 of 160 May 11 2011 AC 10 0 ac dec 10 1 40 Figure 20 LTSpice Circuit The impedance of the capacitor is assumed to be negligble therefore the value of the resistance of the entire circuit was estimated as 430 using Ohm s law V TIx R 24V 560mA R R 430 The voltage was set at 24 Volts because that
106. er IG 523mA Calculations PD Vin Vout I L VinI G 6438 W T R max 95 A C We want 8 JA actual high since 8 JA actual is lower than 6 JA rating we therefore need better heat dissipation Figure 16 Heat Dissipation Calculations Since the integrated circuit is not able to dissipate the generated heat fast enough better heat dissipation is needed In order to fix this problem a heatsink and a fan were added in order to keep the_LM317at an acceptable temperature Examples of heatsinks that were used for testing and the final design are shown in Figure 17 Team 3 Achieving Mobility Final Report Page 33 of 160 May 11 2011 Ez f TARTA 2 ABLA 5 gt E A EL aa f A PL T 2X eZ EL Zu er PT ER gt AES Ea CALI AA 22727 P ea Figure 17 Examples of Common Heatsinks 4 1 5 4 Over Voltage Protection The motor controllers should not have more than 30 V applied to them and at 35 V the motor controllers may get damaged The team 1s confident that this voltage from the batteries will never reach 30 V because they are 26 V when fully charged and this maximum voltage should only decrease as the batteries age The batteries were also hooked up to an oscilloscope in order to check 1f there was a voltage spike During testing there was a spike of approximately 1 5 V This spike paired with a fully charged battery would be 27 5 V Yet there are nodes of 12 V 9 V and 5 V within the power regulation circuit that could pote
107. er individuals in the area Team 3 Achieving Mobility Final Report Page 69 of 160 May 11 2011 Table 14 Total tube length needed for the frame Material Total Qty Needed in LENGTH 1 OD X 1 8 WALL 6063 T6 374 5 ae 15 75 24 35 37 56 ee eee NN 38 88 3 4 OD X 1 8 WALL 6063 T6 191 0 i AAA AR e a N 5 1 6 Financials As stated above the material cost of the frame was around 100 The only other cost that had to be factored in was the cost of labor to machine the ends of the tubes bend the tubes and weld the beams together Labor time for this project was 40 hours Ebling and Son Inc Blacksmiths agreed to do our project at a reduced rate of 55 hour Since they did not have the dies to do the tube bending they had to outsource some parts to another company to make The final cost for the labor provided by Ebling as well as the outsourced parts and material costs was 3 000 5 2 Wheels The wheels and tires for the wheelchair could easily be purchased from a wheelchair manufacturer Since they are inexpensive and readily available designing our own wheels would be impractical Some considerations are the size of the wheels as well as the durability The size of the wheels and tires on the delivery design will be the same to those on Isaac s old wheelchair the front tire is 6 3 4 in diameter and the rear tire is 12 1 2 in diameter These sizes have proven to be adequate and have been very relia
108. er to the system Since the device relies solely on the battery for power the battery provides sufficient energy for the entire use of the device 3 1 5 Power Regulation The allocation of power through the system has been carefully designed so that the device can be operable for an extended amount of time If power is improperly regulated the device will not function properly because power will be converted to heat Thus if a significant amount of power dissipation occurs the device may only be operable for a short amount of time 3 1 6 LCD The LCD provides the user with the ability to see what is in front of the device 3 1 7 Video Camera The video camera sends the live video feed directly to the LCD The camera is mounted to the front of the wheelchair so the user knows what is ahead when driving forward Team 3 Achieving Mobility Final Report Page 11 of 160 May 11 2011 3 1 8 Motor Control Circuitry The motor control circuitry contains all the electronic components that connect the digital outputs of the MCU to the motor It 1s designed for high current motor control and utilizes safety components in order to protect the components 3 1 9 Electrical Safety For electrocution to occur one needs to be touching two different conductive pieces that have different voltages between the two parts Current cannot travel through a person s body without a voltage difference so for that reason voltage levels have been examined in order t
109. estraint system is shown in Figure 55 5 4 4 Feasibility The fabrication of the bed was very feasible with the help of Mary Free Bed Rehabilitation Hospital more specifically their OrthoSEAT division They design and fabricate bedding like this on a regular basis and have agreed to do our project They have also agreed to do our project for a reduced rate The final cost of the bedding was 286 which accounts for 174 for material and 112 for labor 5 5 Motors 5 5 1 Research The three main manufacturers of powered rehabilitation wheelchairs are Invacare Pride Mobility and Permobil Based on communication with each company brushed DC BDC motors are the motor of choice for powered wheelchair applications involving rehabilitation These companies use BDC motors in all of their products except a few models which use brushless DC BLDC motors The major reason these companies choose to go with BDC over BLDC is cost Jill Kolezynski Research and Development Lead at Pride Mobility states that BLDC motors cost 15 to 20 percent more than BDC motors Though this research indicates that BDC motors are the most popular for this type of application both BDC and BLDC motors were analyzed equally to determine which motor for this application was the most feasible Team 3 Achieving Mobility Final Report Page 75 of 160 May 11 2011 5 5 2 Requirements 5 5 2 1 Torque The motors shall be powerful enough to provide the necessary torque
110. ey archer dynamo drum brake front hub 36h 2177 htm gt SDM 2 0 Bicycle Disc Brake Products Alibaba n d Web 30 Nov 2010 http www alibaba com product tp 112287745 SDM2 0 Bicycle Disc Brake html Power Wheelchairs Invacare 2011 Web 9 May 2011 http www invacare com cgi bin imhaprd inv_catalog prod_cat jsp s O0 amp catOID 536885238 gt Quantum Rehab Pride Mobility 2010 Web 9 May 2011 lt http www pridemobility com quantum powerbases q6series q6edge asp gt Rehab Series Permobil n d Web 9 May 2011 lt http www permobil com USA Products Rehab gt Single Strollers JoggingStroller com n d Web 9 May 2011 lt http www joggingstroller com Strollers Single Strollers index cat gt Easys Rehabilitation Stroller Thomashilfen 2010 Web 12 Nov 2010 lt http www thomashilfen com th index php en gt Shimano Ultegra Brake Caliper Carrollwood Bicycle Empoirum n d Web 30 Nov 2010 lt http cbebikes com product shimano ultegra brake caliper 43 120 1 htm gt Team 3 Achieving Mobility Final Report Page 144 of 160 May 11 2011 43 44 45 46 4T 48 49 50 Sl 52 33 54 Bicycle brake systems Wikipedia 2010 Web 30 Nov 2010 http en wikipedia org wiki Bicycle brake systems twhittbs Wilson David G Bicycling Science 2nd ed Cambridge The MIT Press 1982 237 41 Wikipedia Web 30 Nov 2010 Brandt Jobst Brakes from Skid Pads to V b
111. g Pa cos Opar F Find Fop Fep Fap Fap 50 33 6 63 F pa lbf Finar 32 831b 52 83 hd 3 ibe Page 154 of 160 Team 3 Achieving Mobility Final Report Page 155 of 160 May 11 2011 Axial Stress in beam AD r 05in a 0 375in a A w t 5 A 034 in PAP 1 Tai C wal 153 73 p 1 A axial P Bending Stress in beam AD Map Pasin Pg ay Lan Fagcos Pg Ax 8cgg LAp Map 109 19 in Ibf s Jj ip 085 4 a gt Manto bending I ALGOR Stress Calculations Stress von Mises E do Cum Ibf in 2 s 1857 421 Figure 99 Bed stress analysis showing a max stress of approximately 1850psi based on 2001b load Team 3 Achieving Mobility Final Report Page 156 of 160 May 11 2011 Displacement Magnitude in 0 01587775 0 01428998 0 0127022 0 01111443 0 009526652 0 007938877 0 006351101 0 004763326 0 003175551 0 001587775 0 Load Case 1 of 1 Maximum Value 0 0158778 in x Minimum Value O in 0 000 12 592 in 25 184 37 776 lt Design Scenario 1 AAA Figure 100 Deflection analysis showing a max deflection located in the bed of approximately 1 64 Worst Stress Absolute Value Ibf in 2 560 504 448 392 336 280 224 168 112 56 0 Load Case 1of1 Maximum Value 566 473 Ibf in 2 Minimum Value 9 66338e 013 Ibf in 2 0 000 13 466 in 26 911 40 367 1 lt Design Scenario 1 gt Figure 101 Max stress in the f
112. gulation The design alternatives for heat dissipation include natural convection and forced convection Natural convection will be utilized as much as possible because it is the easiest and least expensive way to dissipate heat because it requires no special fan system To determine whether or not a component can be sufficiently cooled by natural or forced convection Newton s law of cooling is utilized as shown below Qconv AconvArce Tpcp 5E Tair The estimated value for each variable is shown in Table 27 Team 3 Achieving Mobility Final Report Page 92 of 160 May 11 2011 Table 27 Values used to roughly estimate cooling of PCBs by convection Natural convection coefficient Forced convection coefficient 10 200 Surface ars Air temperature The convection coefficient is very difficult to calculate and requires extensive testing therefore for a rough estimate a range was used for each Appendix F PCB Heat Dissipation Calculation Figure 67 and Figure 68 show the resulting temperatures of each PCB for both natural convection and forced convection respectively Motor Controllers ECU Power Regulation ee Design Limit PCB Temperature C 10 12 5 15 Natural Convection Coefficient W m 2 K Figure 67 Resulting temperature of PCB due to natural convection Team 3 Achieving Mobility Final Report Page 93 of 160 May 11 2011 Motor Controller ECU Power Regulation ee Design Limit N Q e9
113. gure 34 Also note that each component described in this block diagram was implemented in the ECU in fact the software residing on the motor controllers was provided with the development kits Therefore the software for the ECU was designed to coordinate with the software already residing on the motor Achieving Mobility Software Block Diagram Cyclic Executive controllers Direction Controller High High Ambient Temperature Diagnostics Current Voltage Current Sampler Velocity Controller Voltage Sampler Timer CAN Interrupt Handlers SysTick Fault Driver Library SysTick ARM Core Figure 34 Software Block Diagram 4 2 1 Cyclic Executive Architecture A cyclic executive is a software architecture that is comprised of an infinite loop with all the tasks of the system contained in that loop Since an RTOS was ruled out this architecture is necessary for this device because the software must always be running and responding to inputs performing calculations and producing outputs This architecture was implemented on the ECU however a similar cyclic executive was implemented on the motor controllers however this will not be discussed because it was not designed by the team Figure 35 shows a flowchart of the cyclic executi
114. gure 36 Event Driven State Machine Software Flow Diagram The directional controller also resides on the ECU because it must respond to user inputs Also the method to determine the state of the touch button is polling every 1ms via SysTick interface as shown in Figure 35 Every Ims when the SysTick timer expires the state of the button is polled and debounced with software Interrupts could be used for the button state changes but the SysTick is used for CAN heartbeating as well so it was easier to just poll the button state in the same interrupt handler A general flowchart of the debouncing process is shown below in Figure 37 Team 3 Achieving Mobility Final Report Page 47 of 160 May 11 2011 Read GPIO Pin State Changed om last reading Yes No Y v Increment Reset Debounce Debounce Counter Counter Debounce Counter 4 No az End Polling 2 A Yes Yes utton Status No Pressed Set Button Press Clear Button Press Flag Flag v Save New Stable State for Comparison Figure 37 Algorithmic Flowchart of Button Debouncing Process The directional controller also controls the LED user interface called the directional display The main function of the directional display is to provide the user with information regarding the current direction of both the directional state machi
115. hall also be able to be disengaged for attendants to manually push it 2 2 Mechanical Requirements 2 2 1 Weight Capacity The device shall be designed to support up to a 200 pound user 2 2 2 Product Weight The device itself shall not exceed 250 pounds Most electric wheelchairs are around 300 pounds therefore this will ensure that the stroller is light enough to be lifted by a hydraulic wheelchair lift 2 2 3 Size The size of the device shall be compatible with the customer s current environment This includes being able to drive through a standard doorway that is no more than 32 wide In addition the overall length must adjust in size so that the smallest length is at least 56 and the longest length 1s at least 60 The height of the center of the stroller shall be at least 25 tall so that the customer can be at an appropriate level to children his age per customer request 2 2 4 Storage The device shall be capable of securely storing all customer medical equipment The customer s suction motor weighs approximately 8 pounds and is 9 wide x 14 long x 10 tall The customer s feeding bag is approximately 4 pounds when full and is 9 long x 5 wide x 13 tall Team 3 Achieving Mobility Final Report Page 4 of 160 May 11 2011 2 2 5 Material The device shall consist of material that is rust free and strong enough to support the desired weight requirement 2 2 6 Transportation The device must be ab
116. hat the motor controllers can safely deliver is 40 A however a current of 30 A would only be attained if the stroller was being driven straight into a wall or if the internal brake failed to release If either motor receives 30 A both motors will gradually slow to a halt the LEDs will blink four times and the on chip speaker will chirp loudly four times After the user releases the touch button the stroller will be fully functional again The second diagnostic error is high temperature The maximum operating temperature of the motor controllers is 50 C Therefore in order to protect the motor controllers the maximum temperature was Team 3 Achieving Mobility Final Report Page 49 of 160 May 11 2011 set to 45 C In the case that this temperature is reached the motors will gradually slow to a halt and the speaker will beep three times Also the LEDs will flash constantly until the temperature reaches 40 C Additionally the fans inside the motor controllers will continue to run even though the motors will not be running The third diagnostic error is high bus voltage on the motor controllers 1 e the battery voltage being delivered to the motor controllers The maximum safe voltage that the motor controllers can handle is 35V so a maximum value of 30 V was set as the threshold The maximum voltage from the batteries at full charge is about 26 V which is well under the safe voltage limit however if the user starts charging the b
117. he flexible gooseneck was chosen The gooseneck was also a good choice because it was donated and therefore free and we were able to come up with a way of storing it effectively underneath the bed Figure 80 shows a prototype of the gooseneck system using wood to simulate the dimensions of the frame Page 103 of 160 Final Report Team 3 Achieving Mobility May 11 2011 Figure 80 Prototype of LCD mount This prototype helped determine the correct length needed for the final gooseneck If the length is too short the user will have no viewing angle and if it is too long the LCD will vibrate more and it will be more difficult to store away Once the gooseneck was designed the attachment from the gooseneck to the LCD needed to be designed A coupling was welded to both the LCD plate and the underside of the bed to screw each end of the gooseneck into Also the customer requested that the LCD plate include some sort of sun shield that would help prevent glare on the screen Figure 81 shows the final LCD mounting system ve d 44 MATT LLL i E i Figure 81 Final LCD mounting system 5 8 5 Motor The motors need to be mounted to the frame in a certain orientation because the wheels are directly connected to the motors They need to be both parallel to each other and perpendicular to the ground to Team 3 Achieving Mobility Final Report Page 104 of 160 May 11 2011 provide the smoothest ride and least amount of wear The moto
118. he graphic on top of the video being displayed on the screen This option would require LCD drivers and software written specifically for the LCD screen Additionally interfacing to an external LCD would mean that the LCD drivers included in the software SDK would most likely not work so custom drivers would need to be built Team 3 Achieving Mobility Final Report Page 13 of 160 May 11 2011 LCD t Figure 4 Display Option 1 Overlaid Arror System The alternative to the overlaid arrow system is an LED system mounted beneath the LCD screen as shown in Figure 5 This is less visually appealing to the user but has the same functionality and provides the same information Also the software for this option is much simpler than the overlaid arrows This option would not require LCD drivers or software written for the LCD rather it would require LED drivers which are very simple to implement in software t e poe 8 Figure 5 Display Option 2 LED System 3 2 1 2 Display Selection The selected form of directional display was the LED system shown in Figure 5 Due to the time constraints inherent to this project the time that was saved by using LEDs instead of an overlaid arrow Team 3 Achieving Mobility Final Report Page 14 of 160 May 11 2011 system on an LCD was very desirable Also the aesthetic appeal gained by the overlaid arrow system from Figure 4 was determined to be minimal 3
119. he monitor A 5 6 LCD screen was donated by the Gentex Corporation in Zeeland MI The LCD screen can be seen in Figure 33 Greg Bush an employee of Gentex worked closely with the team to make sure the product s visual system functions properly The LCD screen is NTSC compatible because that is the format that the camera outputs and the two devices need to be compatible The LCD provides the user with vision of the front view of the stroller Each LCD screen needs approximately 12 V therefore a voltage regulator was implemented so that the LCD receives the appropriate voltage Mirrors were not used because a camera and LCD system was preferred by the customer The customer believed that 1t would be more convenient and easier to use 1f a camera and LCD system was used Team 3 Achieving Mobility Final Report Page 43 of 160 May 11 2011 Figure 33 Liquid Crystal Display Monitor 4 1 11 Camera An NTSC compatible camera was used for the final design Gentex donated the camera and LCD system and each device needs to use the same format and since the LCD is NTSC compatible so is the camera Since Gentex works with rear view cameras mounted on the back of automobiles the company has many experts who have had a lot of experience with rear view camera systems The cameras that Gentex deals with all have flipped images because the cameras are used for rear view video Therefore since the camera shows the front view to the user the camera should
120. heelchair while lying down using the camera LCD vision system In order to sustain a business this feature needs to be emphasized and other competitive features need to be designed for such as an adjustable seat various drive controllers variable speed and wheel suspension system Team 3 Achieving Mobility Final Report Page 122 of 160 May 11 2011 The cost of the different competitive models needs to be analyzed in order to determine if our production design cost is competitive or if we need to revise something to reduce the cost To compare the wheelchairs on the same level options will be selected that are most similar to our production design Table 41 shows a cost comparison of the various similar wheelchair models with the pricing of the different options Pride Mobility was not included because detailed price sheets were not readily available Table 41 Cost comparison of competitive models Invacare FDX Permobil C350 Corpus Base Model 5 790 00 Base Model 6 645 00 Transport Tie Down 250 00 Proximity Switch 524 00 Brackets Proximity Switch and 3 852 00 Input Module 1 012 00 Digital Scanner mE Digital Interface 1 500 00 Locking Base 1 806 00 OX Emergency Stop Switch 1 185 00 R net Controller 545 00 ili Controller Mount 475 00 MK6 Auxiliary Power 200 00 Controller Mount Source Attendant Stop Control 1 103 00 Wheel Locks 45 00 Attendant Push Handles 345 00 The price of
121. here is a diagnostic problem and the particular sequence of LEDs and beeping would be mapped to a particular diagnostic message The team would provide the user with all possible diagnostic errors and the corresponding responses allowing them to quickly diagnose the problem The second alternative for displaying diagnostic information is to interface the ECU with a small character LCD allowing for real time diagnostic information available to the user This option would be much more convenient and helpful for the user because there would be no need to decipher the message 4 2 3 2 Selected Design The LED and speaker option was selected for diagnostic information because the user will rarely see any diagnostic errors so the inconvenience of having to decipher the messages is minimal This method also saved a lot of time approximately 15 hours of design work because an external LCD did not need to be selected or interfaced to the ECU There are three diagnostic errors that are checked for periodically approximately every 1 2 ms The user will only be notified when the motors are operating at a dangerous level therefore 1f all components are behaving properly the user will see no diagnostic information The first diagnostic error that 1s periodically tested is too much current When the motors are running the ECU constantly checks to make sure that neither motor is providing more than 30 A of current The maximum amount of continuous current t
122. icients of Friction The Engineering Toolbox n d Web 18 Apr 2011 lt http www engineeringtoolbox com friction coefficients d_778 html gt Dual Swinging Arm Desk Mount Parity Medical 2008 Web 1 Dec 2010 lt http www paritymedical com mounts desk fixed height single monitor dual swing arm htm gt Team 3 Achieving Mobility Final Report Page 145 of 160 May 11 2011 55 56 JJ 58 59 60 61 62 63 64 65 66 67 68 Arkon IPM525 S 15 Inch Tall Flexible Steel Gooseneck Seat Bolt or Floor Mount for iPhone 4 with Slim Grip Phone Holder Bizrate 2010 Web 1 Dec 2010 lt http www bizrate com gps accessories o1d2200436801 html gt Central Iron Steel Inc Phone interview Apr 2011 Electric Wheelchair Qualifying for Medicare and Insurance Coverage Mobility Advisor 2005 2010 Web 3 Dec 2010 lt http www mobility advisor com electric wheelchair medicare html gt Invacare Corporation Product Catalog FDX with Formula CG Powered Seating Invacare Corporation 2010 Web 3 Dec 2010 lt http www invacare com cgl bin imhgprd inv_catalog prod_cat_detail jsp s O amp prodID FDX CG amp catOID 536891033 gt Quantum Rehab Power Bases Rear Wheel Drive Series R 4000 Pride Mobility 1995 2010 Web 4 Dec 2010 lt http www pridemobility com quantum powerbases rearwheel r4000 asp gt USA Products Rehab Series C350 Corpus Permobil Permobil 2010 Web 4 Dec
123. ility Final Report Page 41 of 160 May 11 2011 4 1 9 3 Battery The positive and negative terminals of the batteries are connected in series to the rest of the electrical system Standard powerpole connections were used to provide safe yet removable connections The connection to the batteries 1s shown below in Figure 30 Figure 30 Battery Connector In order to recharge the batteries the main power switch should be turned off and the charger connected to the designated charging port The designated charging port is shown in Figure 31 Figure 31 Battery Charger Connection 4 1 9 4 LCD Monitors and Cameras The LCD and camera are connected via shielded copper wires that transfer the video signal The video input to the LCD is a shielded signal wire with a standard monaural jack connection The yellow wire Team 3 Achieving Mobility Final Report Page 42 of 160 May 11 2011 from the camera connects to the signal line of the mono plug and the white wire from the camera connects to ground 4 1 9 5 Motor Each motor is connected to the battery by large copper wires that are approximately 10 gauge A picture of how the wires from the motor are connected to the rest of the circuit is shown in Figure 32 Note that the red ports supply power to the motors and the black ports supply power to the internal brake releases Figure 32 Motor Power Connection 4 1 10 LCD Screens The camera provides the live video feed to t
124. is shown below in Figure 45 Figure 45 Schematic Diagram of Simple H Bridge Circuit Team 3 Achieving Mobility Final Report Page 59 of 160 May 11 2011 Along with the H bridge itself some extra protective equipment is necessary for safe operation A slow blow fuse would be inserted between the power supply Vcc and the top node of the circuit to prevent voltage spikes from destroying the motors Also a small capacitor would be used before the fuse to filter out high frequency noise disturbances greater than about 20 kHz and a larger capacitor to absorb excess voltage spikes before reaching the fuse 20 kHz would be the approximate cutoff frequency because the PWM frequency will be a maximum of about 15 kHz Finally because of the large amount of current that can flow through the MOSFETs protection would be needed against excess heat If the transistors get too hot they could melt and start a fire Therefore heat sinks would be used on all of the transistors to channel the heat from the transistor to the ambient air for convective cooling The FWD and REV ports in Figure 45 represent outputs from the MCU PWM signals sent to the FWD port cause the motor to turn in the forward direction and PWM signals to the REV port cause the motor to turn in the reverse direction The magic behind the H Bridge is that when voltage is applied to the gate of one of the input ports only that path is opened Thus all current is delivered in the desired
125. is test was conducted on the bench for functionality using just one motor and observing the response via UART output similar to the acceleration test The threshold limit was set to 4 5 A because that amount of current is easily attainable by adding torque with the tester s bare hand The results of this speed control are shown below in Figure 88 This graph shows that once the current was Team 3 Achieving Mobility Final Report Page 111 of 160 May 11 2011 above the minimum threshold in this case 4 5 A for about 1 5 seconds the voltage would increase by 0 5 V Then if the current remained above 4 5 A for another 1 5 seconds it would increase again until it reached 10 V Also if the current ever dropped below 4 5 V throughout the test the voltage would be reset to 8 V All of this functionality was carefully designed to keep the stroller within the required speed window and the results prove that the velocity control software 1s fully functional Velocity Control Test Results Vout Aa Current Em 25 s gt d E a E ai A 3 ee rm ai Lower Current Threshold 10 15 20 25 Time seconds Figure 88 Velocity Control Test Results 6 1 2 LCD Camera System The LCD was tested by powering it the power regulation circuit board Since the optimum voltage of the monitor is approximately 12 V a multi meter was used to confirm that a constant voltage of approximately 12 V is being outputted form the power regu
126. is the approximate value of the two batteries in series The current through the entire power regulation circuit was measured to be approximately 560mA Therefore a 430 resistance was added to the circuit above The value of the capacitor in the simplified circuit was changed in order to see the separate affect of the 0 01uF ceramic capacitor and the 100uF capacitor The Bode plot of the circuit with a 100uF capacitor is shown in Figure 21 Figure 21 Bode Plot of 100uF Capacitor This Bode plot shows that the capacitor acts like a low pass filter The corner frequency is at 34Hz meaning the capacitor 1s allows frequencies of below 34Hz and filters out frequencies of higher than Team 3 Achieving Mobility Final Report Page 36 of 160 May 11 2011 34Hz The plot shows that this capacitor is indeed a good choice of a filter capacitor at low frequencies The Bode plot of the circuit with a 0 01uF capacitor is shown in Figure 22 100KHz Figure 22 Bode Plot of 0 01uF Capacitor This plot looks very similar to the previous one but the frequencies are very different The corner frequency is at 340kHz meaning the capacitor is allowing frequencies of below 340kHz and filters out frequencies of higher than 340kHz Although it may seem that the 100uF is unnecessary the small capacitor is more focused on filtering out high frequencies and the large capacitor is better at filtering out low frequencies 4 1 8 7 Ground As a precaution the groun
127. istance of arm support from frame that is supporting rim brakes Distance of brake caliper from brake pad to arm support Outside diameter of brake arm support Tension yield strength of Alumimun 6061 T 6 Mechanics of Materials Shear yield strength of Alumimun 6061 T 6 Mechanics of Materials Weight amp Mass on each rear wheel W WF W wheel total W wheel 8 In NW W whee 180 1bf whee m 180 lbm Time to stop once brakes applied Dist Velmax time MWWA Wheel de acceleration 7 Velnax time a time 0 39s ft a 13 176 2 S Team 3 Achieving Mobility Final Report May 11 2011 Bending Force amp Moment on brake arm F ma F 73 712 Ibf NA M apm F Larm Mari 18 739J Second moment of area and distance from neutral plane T 4 od Don c4 Darm Lug 0 016 1n c 0 3751n MM Worst case bending stress on brake support arm Sworst 4004 ksi Torsion Polar area moment of inertia 4 T Darm 4 Jam 5 Jarm 0 031 in Torque on brake arm Lorake Tm F Tarm 12492 Worst case torsional stress on brake support arm _ Tarm ur oor 1 335 ksi arm Safety Factors u yield SFbendine pr SF bending 9 989 worst B T yield SFrorsion SFrorsion 19 479 worst Page 7 of 8 Team 3 Achieving Mobility May 11 2011 Figure 104 Bending stress in the weld of the brake arm simulated in Autodesk Algor Final Report Stress
128. it margin for wholesaler MSRP BEEN 21 087 Assuming 50 profit margin for distributor Team 3 Achieving Mobility Final Report Page 131 of 160 May 11 2011 7 5 4 Fixed and Variable Costs Since this production operation will be a small business the warehouse and machinery will be rented and there will be an estimated four salaried employees A certain percent of the company s time and money will be spent on research and development and advertising each year to further sell and develop the product The main variable costs are due to labor and material A full breakdown of the fixed costs is shown in Table 51 Table 51 Fixed costs associated with forming a new business Fixed Costs component Cost Quantty Cut note Component Cost Quantity Cost Note I 9 000 5 of engineering time 1 6000 00 00 AA First Year Total 675 500 Subsequent Year Total I 489 500 Excludes initial design cost and prototype cost 7 5 5 Three year financial outlook A three year financial forecast was used to determine whether or not starting a new business for this product would be feasible The cost analysis was based on concepts learned in Business 357 and the BizPlan financial template was used to analyze a three year product outlook Based on information from various sources there is a market for this product of about 15 000 people The number of units sold is based on assuming a small percent of the market
129. ity Final BOM Unit Cost Quantity Total Cost Aluminum Shelf per sq in 3008 900 100 00 Aluminum Tube perin 3029 60 102 08 PCB Components 2900 520 00 550 00 Tie down hooks Team 3 Achieving Mobility Final Report Page 130 of 160 May 11 2011 Table 48 continued Achieving Mobility Final BOM Upholstery Seat 600 00 600 00 30 00 30 00 M 4 431 46 7 5 3 Sell Price The final cost for the end product is based on material and assembly labor costs Labor hours were estimated and broken down into major categories The average cost for assembly labor was priced out at 50 hour to include benefits and overhead Table 49 shows the total labor cost for assembling the stroller Table 49 Total assembly time and cost Note Installing all drivers and software 20 1 000 Time for all component and product testing Contingency EM 510 1096 of total assembly cost A profit margin of 4096 was used for our product since it is a low volume high price product Based on material cost assembly cost and a profit margin of 40 the final product should have a wholesale price of approximately 14 000 as shown in Table 50 The suggested retail price for the product should be set at 21 000 based on a 50 profit margin for the distributor Table 50 Selling price for the stroller Assembly Costs 502 Material Costs J sei o Product Total A Wholesale Price 14 058 Assuming 40 prof
130. ive Table 2 shows a list of the pros and cons of using the PID algorithm with an optical encoder as feedback Table 2 Pros and Cons of PID Algorithm with Optical Encoder Provides precise speed Would require removing control within 0 5 mph electronic brake Selected motor controllers Not necessary for safety have built in encoder interface QED Encoders are expensive about 100 Increased design time After carefully evaluating the pros and cons of each velocity control method the chosen technique was a voltage control algorithm The algorithm utilizes a closed loop control system that maintains a constant voltage to each motor unless the current is above or below specific thresholds discussed in 4 2 4 The current samplings of the motors are used as feedback to determine if there is positive or negative torque on the motors and therefore determine if the speed of the motors needs adjusting While this is not as precise as a PID algorithm it does maintain the required speed window on a 10 slope See section 4 2 4 for a detailed description of the voltage control algorithm Team 3 Achieving Mobility Final Report Page 15 of 160 May 11 2011 3 2 3 Operating System The software to control the execution of functions and applications is called the operating system For this design there were two main options for executive software a commercial OS more specifically an RTOS or a custom designed cyclic executive Examp
131. ize The motors needed to have a weight capacity of at least 450 pounds at least 0 295 horsepower and be compatible with 12 5 inch wheels Table 18 compares the various motors offered by Invacare Several models are missing values due to unavailable information from Invacare Table 18 Comparison of Invacare motors Combined MOS TA ent Power Speed Size a 2 pole Gearbox P 4 pole Ratio m Lm mim omm A 4m i 1 68 4 Taman oper os ee eel NuronRS 455 350 05 2poe 40 125 PonxbpT 392 350 05 40 125 Pronom51 49 350 05 425 105 PnoM9 57 4ple 14 3GTorqe SPRWD 566 4poe 1 amp 1 65 14 3GRangeXRWD 666 4poe 231 50 14 Motors off the Nutron R51 model wheelchair were selected for our design because they fit our specifications the best Even though 0 5 horsepower is much more than what we need 0 295 hp it is the smallest motor that Invacare uses on their wheelchairs Figure 61 shows one of the new Nutron R51 motors that was in the final product Team 3 Achieving Mobility Final Report Page 83 of 160 May 11 2011 Figure 61 Invacare Nutron R51 motors used for delivery design 5 5 6 Financials For the production design the PMDC motors will be purchased from a motor manufacturer in China but because their prices were not available the price of the Nutron R51 motors from Invacare was used as a rough estimat
132. lar Atrophy e Spina Bifida e Quadriplegia The delivery design has a specific customer Isaac Potsma The delivery design is therefore being customized to fit the particular needs of Isaac The touch button for example is utilized due to the fact that Isaac has some movement in one of his fingers Other customers might not have this capability so the production design needs to be expanded to a larger customer base The production design will therefore have an independent control system that works with several different user interfaces including a joystick a sip and puff device or a tongue device 7 3 2 Market Size There are an estimated 1 6 million Americans using wheelchairs outside of institutions according to data from the National Health Survey on Disability Of these 1 5 million use manual devices while only 155 000 use powered wheelchairs Wheelchair use increases rapidly with age as shown in Table 43 Team 3 Achieving Mobility Final Report Page 124 of 160 May 11 2011 Table 43 Wheelchair Use by Age Group in 2002 All Wheelchairs Manual Wheelchairs Powered Wheelchairs 1000s population 1000s population 1000s population gr so oss 90 006 s 897 564 0 1 IS oo Elderly people are the group with the highest rates of both manual and electric wheelchair use However more than two thirds 69 7 of powered wheelchair users are not elderly under 65 years of age Table 44 shows
133. lation circuit board The voltage from a power source was varied from 18 V to 26 V and the input voltage to the monitor remained constant at approximately 12 V The camera had similar tests to the LCD monitor because the camera needs a constant voltage of 9 V This means that the camera will also use a voltage regulator In order to make sure the camera worked in the final design the camera was be hooked up to the power regulation and transmitted a signal to the LCD Team 3 Achieving Mobility Final Report Page 112 of 160 May 11 2011 Once the LCD and camera are working together the brightness and clarity of the picture was tested The LCD screen was taken outside during the early afternoon and the picture was still easily seen with direct sunlight hitting the screen The LCD screen had excellent quality when used in the dark although the camera had difficulty capturing the surrounding objects 6 1 3 Power Regulation Circuit Board The power regulation circuit was tested by inputting a 24 V battery source to the input terminals Then a multi meter was used to measure the voltage at the 12 V 9 V and 5 V nodes They read 11 9 V 8 7 and 4 9 V which were all acceptable 6 1 4 Temperature of Electrical Components The electrical components of the circuit board may overheat due to the potential of high current traveling through them The electrical components were closely monitored to make sure they did not overheating during operation The fi
134. le AGM batteries This monitor was mounted to the battery so the customer could check the capacity of the battery at the present time The monitor provides the voltage of a battery as well as the capacity of the battery as a percentage These values give the customer a better indication of when the batteries need to be recharged The selected battery monitor is shown below in Figure 9 Figure 9 Battery Monitor Team 3 Achieving Mobility Final Report Page 25 of 160 May 11 2011 4 1 6 Communication Protocols The Stellaris MCU has several options for peripheral communication namely these are UART SPI C and CAN For the purpose of communicating between the three MCUs any of these protocols would theoretically work However since CAN is the primary built in method of communication with the MDL BDC24 motor controllers it was decided to use a CAN bus for bus communication CAN is considerably more reliable than the alternative protocols such as UART SPI and C because it has excellent error detection and confinement capabilities It has a much more robust error correction field than any other alternative including three error detection mechanisms at the message level and two at the bit level which is important for safety critical systems operating in harsh environments The tradeoff of using CAN is that it is slower than SPI and FC however a high speed CAN bus with data transfer speeds up to 1 Mbit s are fast enough to maintain accur
135. le to fit on a wheelchair hydraulic lift that is approximately 40 long and 36 wide In addition there must be tie down hooks to lock the stroller in place in a wheelchair accessible van The hooks must be approximately 24 apart on the stroller to ensure they are compatible with the locking system currently in the customer s van 2 2 7 Bedding The bedding material shall provide enough cushioning to keep the user comfortable and shall be made from a material which provides enough friction to prevent the user from sliding around The bedding shall also include a restraint system for when the customer is being transported in a vehicle 2 2 8 Mounting Component mounting for the LCD touch button and camera shall be in a convenient location for the user and be flexible enough to adjust to the user s needs 2 2 9 Encasings The encasings shall provide adequate protection for all electronic components from precipitation as well as enough heat dissipation to avoid overheating 2 2 10 Maintenance The stroller shall be essentially maintenance free The only maintenance that may be required is replacing the motor brushes after five years based on usage greasing the bearings if the front wheels start to stick and inflating the rear tires if they become soft 2 2 11 Weather The stroller shall be rain proof and be completely operable in all forms of weather precipitation The stroller shall also have attachments to attach a wheelchair
136. ler during manual operation The final decision matrix is shown in Table 8 Table 8 Battery Type Decision Matrix Weighted Score o Sm 6 The decision matrix above shows that sealed lead acid batteries are the clear choice This battery type has been very successful in past wheelchairs Although the size and weight of a lithium ion battery would be beneficial this advantage does not outweigh the negative aspects thus the use of a lithium ion battery is unacceptable Team 3 Achieving Mobility Final Report Page 29 of 160 May 11 2011 When talking to an Interstate Batteries representative who specializes in batteries he confirmed the decision to choose deep cycle lead acid batteries He described lithium ion batteries as too delicate meaning the technology for lithium ion batteries has not reached the point where they would be reliable enough to use in an electronic vehicle 4 1 7 5 Battery Charging A charger for the lead acid battery will be provided to the customer The charger is necessary so that the battery can be recharged when the product is not in use The charger connects to a standard American outlet and hook up to the batteries through a simple connection The present battery charger was donated to the team and is being used in the final design due to the cost or purchasing a new one A picture of the battery charger is shown in Figure 13 Figure 13 24V 6A Battery Charger 4 1 8 Power Regulation The power r
137. les of possible RTOSs include QNX VxWorks or eCos These are commercial operating systems with a relatively small overhead and could be used to control the flow of the software Cyclic executive software would be designed by the team and customized to meet the needs of the device A table evaluating the pros and cons of an RTOS is shown below in Table 3 Table 3 Basic Evaluation of Pros and Cons of an RTOS Come with device drivers Increased code space Flash ROM and necessary RAM space Priority based preemptive High learning curve to write multitasking code for RTOS Built in memory management Can be expensive to purchase if not open source up to several thousand dollars for licenses For this project the cons of an RTOS severely outweigh the pros In fact none of the advantages to using an RTOS would be highly beneficial for the project No TCP IP USB video or file system drivers were needed for the device and there was no need for extensive memory management Also priority based preemptive multitasking was not necessary because the software is not complex enough to need it A round robin scheduling approach was used which was implemented in a cyclic executive A cyclic executive is a control structure or program for explicitly scheduling the execution of several periodic processes on a single CPU This scheduling is done in a deterministic fashion such that the execution of any program is predictable This determinism
138. llaris LM352963 Evaluation Board User Manual LM3S2965 CAN Evaluation Board Texas Instruments 6 Jan 2010 Web 19 Jan 2011 http www luminarymicro com products Im3s2965 can evaluation kit html 9 LM382616 Microcontroller Luminary Micro 2010 Web 15 Oct 2010 lt http www luminarymicro com products LM352616 html gt 10 Argus Analyzers Battery Bug Deep Cycle Battery Monitor N p 2011 Web 15 Apr 2011 lt http www argusanalyzers com battery monitors products bb dcm12 agm battery bug battery monitor argus analyzers copy 1 html gt 11 Chaurasia Neelima Serial Buses in Industrial and Automotive Applications Communication Protocols N p n d Web 12 Nov 2010 lt http www scribd com doc 33221949 Communication Protocols gt 12 Controller Area Network CAN Overview National Instruments 7 Nov 2009 Web 7 Nov 2010 http zone ni com devzone cda tut p 1d 2732 13 24V 30AH V2 5 LiFePO4 Battery Pack ProStores 2010 Web 23 Oct 2010 Team 3 Achieving Mobility Final Report Page 142 of 160 May 11 2011 14 Interstate Batteries Personal interview 4 Nov 2010 15 Collinson Andy Over Voltage Protection Circuit Exchange International N p n d Web 9 Mar 2011 lt http www zen22142 zen co uk Design overvoltage htm gt 16 Slo Blow Fuse 313 315 Series Jameco Electronics n d Web 2 Apr 2011 17 Alternating and Charging Systems N p 2010 Web 6 Dec 2010 lt http www glass
139. ls ds O CO saasaboecaatanepeccnstareactamdess 6 2 5 Desea INGOTS ee o nn o m 7 2 5 cuu Pu e o ER 7 2 5 2 lr ugln T 7 2 5 3 boc 7 2 5 4 TAS A toos 8 A A PEE o yu A A 8 3 1 Electrical a AY t E E A A E TT 9 3 1 1 Cea TS OM EO E NORTE EET 10 3 1 2 TOUCO DUO sisas E A E aos 10 3 1 3 E O O ee ee DT 10 11 3 1 4 DAME site 10 3 1 5 Power Reunidas 10 3 1 6 A UEM E MM UM 10 3 1 7 As A II A O O M ee e ERE 10 3 1 8 Motor Control CiU ono 11 BEY Eco aks Oy e Se an S 11 3 2 SN E E E EEE AE E E A EEEE EE 12 idol AAA TO E 12 322 Velocity Control JATSPOEIEDITI nenea a eee eae 14 3 2 3 Operatie SyS A sack er ahs a oe dee a 15 3 2 4 LE eal al ic AA A S 15 3 2 0 Vo NR A 16 3 3 Mechanical Hard Wat a dde 16 3 3 1 A ec d ILI M E pM M EM E T 16 3 3 2 MS Wes Be eT RN UM 16 3 3 3 Bede ceci eM is M LE MA MU 16 3 3 4 SOL NR OT O EN 16 3 3 5 A e TERIS REUR 16 3 3 6 MOORS NR iaa 17 3 3 7 A O MCI ASCENT E LM UU NEE IRON 17 3 3 8 AMC la MALE edo e di e ad o e 17 is E NENNT 17 4 1 PALO O 17 4 1 1 Motor C OnO ET creia a dos 17 4 1 2 Electone Contool Uiit anestesia nne lt aanedoe Sano nsGuaaneducsedetoshunanwdaaSomerosteeaeednasenen 20 4 1 3 Hardware TODOlOBS de ite used a b mt ies te E Er A EE eea OEE 27 4 1 4 MICEFOCOBEEOLIBESE os create ricco e uci cA ILI M AIL due iL d d EL M Sekai 23 4 1 5 E A E A II EI INR 24 4 1 6 Communi
140. mance 6 1 7 2 Attendant controlled Braking System For the attendant controlled brakes the rim brakes must be able to stop the wheels when the hand lever is compressed by the attendant The motor was setup as described above and a rim caliper was held over the wheel rim connected to a hand lever Once the wheel on the motor had been running for approximately 2 minutes the hand lever was compressed The wheel on the motor then came to a stop while the lever was compressed therefore passing this test This test determines whether the brakes function and not their performance Team 3 Achieving Mobility Final Report Page 114 of 160 May 11 2011 6 1 7 3 Emergency Braking System The emergency brakes provide the attendant with a convenient way to quickly stop the wheelchair for any given situation To test this functionality the motors were setup as described above The power to the brake release terminals was then cut to 0 V by turning off the power supply this simulates an attendant switching the emergency brake switch to the off position The wheel on the motor came to an immediate stop almost simultaneously with cutting power to the brake release terminals therefore passing this test This test determines whether the brakes function and not their performance 6 1 8 Structure The goal of testing the structure is to ensure that it is strong enough for the maximum user weight requirement of 200 pounds The structure consists of the frame bed
141. maximum reliability of the stroller 6 1 Component Test Plans This type of testing served to test the individual components of the system It was important to determine whether each component worked before integrating it into the system By testing each component separately from the others it simplified the complexity of the project and increased the chances of the integration tests being successful Instead of one complex and intricate product it was broken down into smaller steps for preliminary testing Team 3 Achieving Mobility Final Report Page 109 of 160 May 11 2011 6 1 1 Software Testing of software was conducted on each piece of software to make sure that it was functioning properly in all circumstances The following section describes the component testing that was performed on the ECU and the motor controllers The ECU software was tested and debugged regularly throughout the development process The three main aspects of the software design that were tested rigorously are the user interface velocity control and direction control The integration of these three systems was then tested to make sure that they work properly together 6 1 1 1 1 User Interface The LEDs for steering were tested for the time delay between directional states and to confirm that the select LED remained illuminated for the duration of the button press Also this test made sure that the state transitioned to forward after any button press
142. mer the customer requested that the camera be removable so they could move the camera around by hand and show the surroundings to the user while he she is lying down This made the camera design a little more complicated and a new design had to be created To incorporate an adjustable camera that can still be fixed during operation a square tube system was designed as shown in Figure 74 Figure 74 Second design of camera mounting system The smaller tube was dimensioned so the camera would fit snug inside 1t with epoxy applied The smaller tube is the one that will be capable of being removed by the user attendant The smaller square tube will then fit into the larger square tube and snap in place utilizing a spring button The same mounting plate design will be used from the initial design to allow for the 30 vertical adjustment Instead of the making the entire device aluminum stainless steel will be used for the large tube because steel on aluminum only has a coefficient of friction of 0 6 as opposed to aluminum on aluminum which has a coefficient of friction of 1 35 A lower coefficient of friction will allow for lower clearances which in turn will decrease vibration and it will also be easier for the user attendant to remove and insert the smaller square tube as well as slide the arm up and down A prototype of the second camera mount design Figure 75 was built to test the viewing angles and size of the mounting plate Figure 75 Pr
143. mponent but unless they are correctly integrated together with the entire system then they are not beneficial to the product 6 2 1 Electrical Testing 6 2 1 1 Speed Testing The vehicle s speed was tested to ensure that it is operating at a velocity that is safe for the user as well as those located in the same vicinity A simple speed test was performed by having a user direct the vehicle in the forward direction for 30 feet Then the duration it takes to move 30 feet will be measured using a stopwatch If it takes approximately eight seconds then the speed of the vehicle is acceptable because this is approximately 2 5 mph which is the average walking pace These calculations are shown below in Team 3 Achieving Mobility Final Report Page 115 of 160 May 11 2011 Figure 90 The stroller was also tested on a 30 degree hill in order to test the speed difference on an incline and decline Since the team could not find a 10 degree hill readily available a 30 degree hill was used instead If the stroller stays within the required speed window at 30 degrees 1t will work even better with a 10 degree slope Distance Rate Time Distance Time Rate Distance 30ft rate hour 3600seconds second B l seconds Figure 90 Calculations for Speed Testing The results are summarized below in Table 38 These test results show that the top speed on flat ground is roughly 2 4 mph which is very close to the desired 2 5 mph Also
144. n The first decision made regarding the frame was what material to use The most readily available and common materials are low carbon steel stainless steel and aluminum In order to decide what material to use a decision matrix was set up to analyze what material would be best for the design Table 9 shows the decision matrix for the wheelchair material Table 9 Frame material decision matrix 304 Stainless 1020 HRS 6063 Aluminum Selection Criteria 3 Score Weighted Score Weighted Score Weighted 0 10 Score 0 10 Score 0 10 Score Cot d5 X jJ 5 4 2 9 4 Corrosion Resistance 8 5 40 1 8 9 2 Weldabiiy 5 4 20 9 45 3 15 Strength 7 8 56 8 56 5 35 Weight Deflection Team 3 Achieving Mobility Final Report Page 62 of 160 May 11 2011 5 1 2 1 Cost The cost of each material was calculated based on information from Alro Steel Corporation located in Grand Rapids MI Table 10 shows the cost calculations for each material based on 374 5 inches of 1 OD x 1 8 wall and 191 inches of 3 4 OD x 1 8 wall The length of tubing needed was based on the current design and is calculated in Table 14 Table 10 Cost analysis for each material Material M 1 OD length M ve 3 4 OD length Total Cost cost in cost in 304 Stainless 374 5 243 26 1020 HRS 374 5 143 17 6063 Aluminum 0 09 3745 98 61 5 1 2 2 Corrosion Resistance The corro
145. ne as well as the physical direction of movement of the device This information is generated through LEDs positioned around the LCD screen refer to Figure 5 The LED that is illuminated is a direct output of the current state of the state machine refer to Figure 36 such that the LEDs represent the current state of the directional state machine Also when the touch button 1s Team 3 Achieving Mobility Final Report Page 48 of 160 May 11 2011 pressed the illuminated LED remains illuminated until the user releases the button After the user releases the button the LEDs continue circumnavigating the LCD screen 4 2 3 Diagnostics The diagnostics system tests the electronics for faults and displays the appropriate information to the user regarding the state of the device This section describes the diagnostics system including the user interface and the available diagnostic information 4 2 3 1 Alternatives The other function of the user interface 1s to provide the user with diagnostic information regarding the status of the device There are two main options for displaying the diagnostic information The first option is to use an LED indicator system separate from the directional display system described previously Since the ECU 1s not interfaced with the LCD this is one of few options to notify the user that there is a problem With this configuration a sequence of LEDs would be illuminated and the on chip speaker would beep only 1f t
146. ngineers If the team had had another electrical engineer this production design PCB would have been a possibility Team 3 Achieving Mobility Final Report Page 60 of 160 May 11 2011 5 Mechanical Design The mechanical design of the system consists of all major structural components dynamic parts mounting equipment protection and surface material Figure 46 shows the breakup of the mechanical design of the project The larger tasks near the top must be completed before moving on to tasks further down the tree which means following the schedule will be essential to complete the project on time The following section analyzes each task in detail explaining the decision criteria alternatives financials final design and final product Mechanical Design mo GS x 1 oae S Brakes Figure 46 Mechanical Design Breakdown 5 1 Frame The frame consists of the main structure that supports the bed storage motors and mounting components The frame supports most of the weight of the wheelchair and must be designed with strength in mind Several different materials and designs were analyzed for the final design of the frame 5 1 1 Requirements 5 1 1 1 Strength The frame for the delivery design shall be strong enough to support at least 200 pounds on the bed This specific stroller is designed for individuals less than 60 tall therefore a 200
147. nitor was installed on the battery so that the customer would be able to check the charge of the battery By continually operating the vehicle for several hours 1t was confirmed that the battery had the appropriate capacity to power the stroller and that battery monitor was working properly 6 2 1 6 Temperature Testing The vehicle s temperature was tested at room temperature A thermocouple was used to make sure the main electronics compartment was not overheating The ambient temperature was measured at 23 C After the stroller was used for several hours a thermocouple was inserted into the main electronics compartment and it measured 23 4 C which is acceptable 6 2 2 Mechanical Testing 6 2 2 1 Weight Testing The main goal of the weight test is ensure that the final product can hold a 200 pound user as well as all equipment weight Before the final product was built a Finite Element Analysis was performed on the stroller however the FEA was not able to analyze the strength of several components such as the bed material Therefore the weight test is very important and will show the team where any weak areas are on the final product The customer for the delivery design only weighs 54 pounds and the weight requirement includes all equipment and batteries underneath the stroller therefore several different weight tests will be performed to test different distributions of weight Since the stroller will always have two batteries on it the
148. nnnnnnnnnonononoss 158 12 4 Appendix D LTSpice Power Regulation Circuit ooooonnnnnnncnnnoncncnonononononononnnnnnnnnnnnnnnnnnnnnnnnnnnnnos 160 12 5 Appendix E Voltage Regulator Calculati0DS ooooooonnnnncnnccnnnnnnnnnnnnonononononnnncnnnnnnnnnnnnnnnnnnnnnnnnnns I 12 6 Appendix F PCB Heat Dissipation CalculatiOl ooa rone topic coo oer paa eoo wok inae Suv o esee eua a taies A 12 7 Appendix G Brake Mounting Stress Calculations oooooonnnnnnnccncnnnnnnnnnononononnnonnnnnnnnnnnnnnnonnnnnnnnnnnos 6 List of Figures Figure 1 Picture of Team 3 L to R Matt Last Rob VanderVennen Matt Rozema Dan Evans 2 Fig re 2 Toplevel System Areces A E EDA E Road du dae 8 Figure 3 Use Case Diagram of Achieving Mobility eese eene 9 Figure 4 Display Option 1 Overlaid Arror System esses eene 13 Fig re 5 Display Option 2 LEDS iii dax Se uia luca toas ta ia aaaea 13 Figure 6 Labeled Picture of the LM3S2965 CAN Board o c cccccscscsesesescscscesesececscscscscscssssscseeesseseen 21 Figure 7 Block Diagram of Delivery Design Hardware Topology oooooooooooncccnnnononnnnnnnnnnnonononanannnnnnnoss 22 Figure 8 Block Diagram of LM3S2616 MCUJ isssssseeeeetteteetetetetntetetetntnttee ettet teet teni 23 SU AAA Ein II Dee Uos EIE EUR dDCUS 24 Figure 10 CAN Message Identifier Fields ccccccesessssssssssscececscsesescsesesescscsessesecscacavavavenstsesesesssseeees 25 Figure 11
149. nnnnss 69 Table 15 Storage Shelf Dimensi ns essien areik cni Ei e EEEE E e RETN rrr nnn nnne nennen 71 Table 16 Motor Design Decision Matrix 7 7 0 sse tetetetne tette tete entree teet 78 Table 17 Required power speed and torque for motor and wheel cccccccccceeeeceeeeeeesseeeeeeeeeeeeeeees 81 Table 18 Comparison of Invacare motors c ceesesssssssesesescecscscscscsescsvsvscscscssescsvevssacscesesesvevsvacsceseseseees 82 Table 19 Production Design Costs for the Motor sess 83 Table 20 Delivery Desten Costs tor the Mot scada is 84 Table 21 Decision Matrix for Braking Mechanism Alternatives 86 Table 22 Comparison of Parking Brake Alternatives ccccccccccccccccccccaeeeesssssseeeccceeeeseeseeaaeeeesssseesees 87 Table 23 Production Design Costs for the Braking Systems ooooonnnccccnncnncncnnnnnononnnonnnnnnnnnnnnnnnnnnnnnonnnnnnnnnns 89 Table 24 Delivery Design Costs for the Braking SyStemWS ooooooonnnnnnccconnnononcncnononononononncnnnnnnnnnnnnnnnonnnnnnnnnns 90 Table 25 Decision Matrix for Encasing Materal ccccccccccccccccccccseseseeseseseeecccceceeeeseeeaaaaessseseseeeeeeess 91 Table 26 Electronic Component and Encasing Dimensions ccccccnnnnncnnnnnnnononononnnnnnnnnnnnnnnnnnnnnononnnanannnnnnnss 91 Table 27 Values used to roughly estimate cooling of PCBs by convection oooooooooocccccncncnnnnnnnnnonononononons 92 Table 28 Design Selection for Ele
150. not have a flipped image Gentex has graciously reprogrammed the camera so that the image is not flipped The camera needs 9V supplied to it 4 1 12 Motor Control Hardware The motor control hardware is the link between the digital circuitry and the mechanical action of the motors This hardware is composed of analog circuitry including transistors heat sinks capacitors diodes and a fuse The main component of this circuitry is the H Bridge a device that allows the rotational direction of an actuator to change The H Bridge is composed of four MOSFET transistors two PNP on top and two NPN on bottom along with protection diodes to protect the digital circuitry from back EMF current spikes It should be noted that all of this motor control hardware is included in the RDK BDC24 development kit therefore the team did not design any of the motor control hardware See section 4 3 for a discussion of the production design or the motor control hardware Team 3 Achieving Mobility Final Report Page 44 of 160 May 11 2011 4 2 Software The software must control the speed and direction of the motors as well as provide a means of displaying information to the user It is absolutely necessary for the software design to be robust and precise providing safe and reliable transportation for the user This section 1s a description of the software design that was used in the device A general block diagram of the software architecture is shown below in Fi
151. nt Error Correction Actual Output Voltage C s Y s E o gt Controller gt ES gt T s Sampled Output Q s Current Sensors H S Figure 38 Block Diagram of Closed Loop Voltage Control System Also in order to slowly accelerate and decelerate the motors Pulse Width Modulation PWM was needed for the motor controllers Control of DC motors using PWM is a very common method Controlling the pulse frequency 1 e the duty cycle of the PWM voltage signal from the MCU changes the average power to the motors A higher frequency produces a higher average voltage producing faster motor rotations Conversely lower frequencies produce slower motor rotations This PWM signal is sent out by the motor controllers at varying frequencies depending on the voltage level commanded by the ECU 4 2 4 2 Implementation The voltage controller module was implemented within the ECU Since the ECU is directly interfaced with the external touch button and runs the directional state machine the ECU 1s better suited to send voltage commands to the motor controllers depending on the directional state selected by the user If the voltage control methods were implemented on the motor controllers the directional state would have to be queried before any speed adjustments could be made Thus the ECU sends the desired voltage signal as a CAN message to the motor controllers which then computes the
152. ntially rise above the desired voltage if a voltage regulator would stop working Therefore instead of destroying the LCD camera or ECU the over voltage protection would break a fuse instead Therefore over voltage protection was added to the 12 V 9 V and 5 V nodes on the power regulation circuit After reading Andy Collinson s article and discussing it with him the crowbar method was used to protect the motor controllers from a voltage exceeding the desired value The figure below in Figure 18 shows over voltage protection part for the 5 V node of the circuit Figure 18 Over Voltage Protection Schematic Team 3 Achieving Mobility Final Report Page 34 of 160 May 11 2011 If the voltage gets above 5 6 V the zener diode will close and current will to the following node This will then open the 2N 1595 thyristor which is rated for 1 6 A This will create a short circuit causing the 1 6 A fast blow fust to break This produce an open circuit and the LCD camera and ECU will lose power The 4 7 kQ resistor was recommended by Collinson so that the majority of the current travels through the thyristor 4 1 8 5 Switch A mechanical switch 1s in the circuit as another precautionary measure If the switch is flipped the circuit will become an open circuit and thus cut off power to everything The most important thing that cutting the power does is reapplying the brakes and the stroller will come to a stop immediately This would be extrem
153. ntire project Management within the team was done through intra team status reports At the end of the every week each team member went over what they accomplished as well as what would be completed the following week Issues such as team members spending too much time on a task or being behind schedule were brought up and addressed at this time No consequences were given to any team member for poor quality Team 3 Achieving Mobility Final Report Page 135 of 160 May 11 2011 or lack of work Instead issues were addressed by talking them over and figuring what can be done to overcome them A daily time card was filled out each day and totaled at the end of the week to make sure all team members were putting in enough time and to avoid one person putting in significantly more time than others Figure 95 shows the monthly hours the team has logged over the course of this school year Team 3 Cumulative Monthly Hours September October November December January February March April Month Figure 95 Cumulative weekly hours for the team In addition to having the guidance of Professor VanderLeest several outside mentors and sponsors have volunteered their time and resources to help the team Ivanrest CRC has graciously donated the funds to complete the project In addition Texas Instruments Gentex Mary Free Bed Rehabilitation Center and Ebling and Son Inc Blacksmiths have donated their products time and resources Without these sponso
154. o find out what is dangerous for the user The average person can only perceive electricity flowing through them that is about 4mA Using Ohm s law of Voltage Current Resistance it can be determined how much voltage is needed to make a person notice electricity running through them Since the product is being powered by 24V DC battery this is the voltage that the calculations use in order to err on the cautious side Although the electrical resistance of the human body varies the resistance between two dry hands is approximately 1 MQ and the resistance between two wet hands 1s 17kQ Therefore the voltage needed for one with wet hands to perceive the electrical current is V 4mA 17kQ 68V Therefore one cannot perceive electrocution from the product with dry hands or even wet hands If a person has wet hands and touches a terminal with a metal object such as a ring the resistance between their hands greatly decreases to about 1k 2 Therefore the voltage needed is V 4mA 1kQ AV The product does provide this much voltage so there is a small risk The customer should not be worried about electrical shock from the product because the voltage is insignificant According to most industry standards 30 Volts is considered to be a conservative threshold for a dangerous voltage In order to provide the safest product possible the terminals are covered with insulation including the terminals of the battery Also the batteries are
155. of 160 May 11 2011 Figure 51 Tie Down Angles The frame was also made with an extendable leg portion to adjust as the user grows in height This was accomplished by making a foot weldment that could slide into the main weldment and be secured using four bolts The overall length of the stroller could then be lengthened by up to 4 Another design factor that influenced the final design of the stroller was aesthetics Wherever possible rounded corners were used as opposed to sharp corners to avoid a boxy look to the stroller Even though bending the tube is more expensive and harder to make it was designed this way to be appealing to the user 5 1 5 Feasibility Even though the final design was difficult to manufacture due to the welding of the aluminum and the inability of Calvin to bend large tubing and the manufacturing of the frame alone cost more than Calvin s budget it was still feasible with the help of outside organizations and companies Ivanrest CRC made a large donation towards the project which helped pay for the added cost of aluminum The current design uses approximately 566 inches of aluminum as shown in Table 14 which cost approximately 100 Since welding the aluminum will be difficult and there is no equipment to bend 1 tube at Calvin College the team partnered with Ebling and Son Inc Blacksmiths to weld the frame This company is located in Grand Rapids MI and has worked on projects similar to this one for oth
156. of the system In order to achieve a constant acceleration of 1 5 ft s the desired time until full speed was approximately 2 5 seconds using the following equation 2 5mph _ ft 1 47 2 5s lt 2 The inputs and outputs that were tested are summarized in Table 37 Team 3 Achieving Mobility Final Report Page 110 of 160 May 11 2011 Table 37 Directional Software Test Plan Directional Counter Clockwise Counter Clockwise Counter Clockwise Counter Clockwise The results of the direction test were a success The observed response of each motor matched the expectations The results of the acceleration test are shown below This test was conducted on flat ground with a 200 lb load on the stroller The results are shown in Figure 87 Motor Voltage as a Function of Time Output Voltage V Time s Figure 87 Acceleration Testing Results As seen in this graph the maximum speed 8 V was reached after about 2 6 seconds This is extremely close to the desired 2 5 seconds such that the difference 1s negligible Therefore this acceleration was used for the design 6 1 1 1 3 Velocity Control Software The velocity control software was tested to ensure that the speed control functions worked properly such that if the stroller experienced opposing force such as an incline the voltage would increase to compensate and if the stroller experienced a helping force such as a decline the voltage would decrease to compensate Th
157. oject so special It is our goal that through this project Isaac Postma will be able to put life into drive and achieve mobility Team 3 Achieving Mobility Final Report Page 141 of 160 May 11 2011 11 References l Electrical Safety lt http www allaboutcircuits com vol_1 chpt_3 4 html gt Ze Ermer Gayle and Steve VanderLeest Using Design Norms to Teach Engineering Ethics Calvin College N p June 2002 Web 22 Nov 2010 lt http www calvin edu academic engineering about faculty svleest abstracts asee02 htm gt 3 Measuring speed and position with a quadrature encoder Pager motors and robot gearmotors Altera Corporation n d Web 17 Oct 2010 lt http www didel com mot RomEnco pdf gt 4 Motor Control and PFC Developer s Kit Texas Instruments n d Web 10 Oct 2010 http focus ti com docs toolsw folders print tmds mtrpfckit html gt 5 RDK BDC24 Firmware Development Package User s Guide Stellaris Brushed DC Motor Control Reference Design with CAN 24V Texas Instruments 3 Aug 2010 Web 27 Nov 2010 lt http www luminarymicro com products rdk bdc24 html gt 6 PICDEM Mechatronics Demonstration Kit Microchip Technology Inc July 2005 Web 13 Oct 2010 lt http ww 1 microchip com downloads en DeviceDoc 5 1568a pdf gt DE2 Development and Education Board Altera Corporation n d Web 17 Oct 2010 lt http www altera com education univ materials boards de2 unv de2 board html gt 8 Ste
158. on Table 16 Motor AT Decision Matrix 3 PMDCMotor Motor SWDC Motor Brushless Motor Selection Criteria s 5 ai Control 54 48 2 12 mE CEN es ae a a 5 5 4 1 Cost The cost of the motors was the most important selection criteria because the motors make up 50 percent of the project budget BLDC motors cost 15 to 20 percent more than BDC motors BLDC motors require a complex controller to run and that increases the cost substantially Team 3 Achieving Mobility Final Report Page 79 of 160 May 11 2011 5 5 4 2 Size The size of the motors is not as crucial because there is room to work with in regards to the structure of the wheelchair Also there is not much variation in size for motors for applications with fractional horsepower However for the same mechanical work output BLDC motors will usually be smaller 5 5 4 3 Service Life The service life of a motor is fairly important because the customer does not want a product that requires servicing often The actual lifetimes of all three options are negligible but the service life is quite different The service life of BDC motors is short 2 5 years because the brushes and commutator need replacement due to constant mechanical wear SWDC fair a little better than PMDC because there is no loss of magnetism in the stator BLDC motors on the other hand have a long service life due to no parts rubbing or wearing during use The only part that might possibly
159. onnnonnnnnnnnannnnnnnnns 14 Table 3 Basic Evaluation of Pros and Cons of an RTOS coooooonnnnnncnnncnononononononnonononnnnnnnnnnnnnnnnnnnnnonnnnanannnnnnnos 15 Table 4 List of Development Kit Alternatives cccccccccccnnnononnnnonnnccnncnnnonononononnnononnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnns 18 Table 5 Qualitative Assessment of Development Kit Alternatives 97 20 Table 6 Typical Power Values of Components ccccccceccccccceccceceeeesssesseseseeecccceeeeeeeeeeaaasasssseeseeeeeeeess 26 Table 7 Assessment of Battery Ty pe seirer ert rep tet buuec e orta ee eorr e a Ie hg repetidas iii 28 Table 8 Battery Type Decision Matrix 2225 1 hide e tasto sao di eio exu di 28 Table 9 Frame material dec SiO malik 10652otivett 29 9209089922612 05892698090 9201 0895620701090 16 302959 26 289 292550205000 200 0920 6dfs 61 Table 10 C ospanalysis Tor each Material recocido cocecib said a nudes Ed euros ANENA 62 Table 11 Strength properties for each material ccccccccccccccccceecceeaeesesseseeeeccccceeseeeseeeaausaesssseeeeeeeeeees 63 Table 12 Maximum deflection for each matertal ooooonnnnnncnnnnnnoncnnnnnononnnnnnnnnnnnnnnnnnnnnononononnnananannnnnnss 63 Table 13 Material Weight Comparison eeeeeesssssessseeeeeeeee nennen nennen enne ene ssnn nnn a rrr r nn nnne n enne es 64 Table 14 Total tube length needed for the frame ooooooonnnnncccnnnonononnnnnononnnnnonnnnnnnnnnnnnnnnnnnnnnnonnnnnanannn
160. opment kit although a very general term is a product that utilizes a certain processor and contains connections to a vast array of peripherals This is a very helpful tool when learning how to use a particular processor and interface 1t with a set of peripherals In general they are used primarily as a learning tool and not often used in a final design because they are often too expensive to include in a mass produced product and they often contain much more peripherals and extra features than needed in the final design They are especially useful when attempting to quickly evaluate the effectiveness of a given processor For the design of this project the selection of a development kit was pivotal for the design of several other components specifically the MCU hardware topology and communication protocols While adapting the design to available development kits is often not very cost effective the benefit of using a pre constructed development kit far surpassed the increased cost of the device Moreover the production design of the device will differ from delivery design in ways that make the device less expensive with similar functionality see section 4 3 for more information However as mentioned earlier the time saved by designing the system around the development kit will be crucial to finishing the project on time 4 1 1 1 Alternatives Due to the time constraints resulting from a team of only two electrical engineers it was necessary
161. or control This decision was based on the fact that it was powerful enough to drive our motors which is critical for early testing capabilities The temperature control capabilities were also an advantage because it helped us create a safer device Also as stated earlier Texas Instruments offered to supply us with these kits free of charge However since these only control one motor and do not have an expandable set of GPIO pins for peripheral interfacing another development kit was needed to provide the user interface and to control the operation of each motor controller independently This development kit will be referred to as the Electronic Control Unit ECU from now on While the term ECU is primarily used in the automotive electronics industry the purpose of this unit is the same as in the automotive industry to coordinate and synchronize the actions of several independent MCUs over a single CAN bus Refer to section 4 1 2 for the selection of the ECU Table 5 Qualitative Assessment of Development Kit Alternatives gt Unsure if it works with brushed dc motors limited peripheral connectivity several unnecessary components most expensive solution does not fit optimal HW topology Enough power to run both motors high resolution PWM signals built in QEI inputs impressive motor control SDK GUI PC control application includes analog HW TMDSIMTRPFCKIT RDK BDC24 Powerful enough to run one motor certain 1t will
162. or the production design the ECU power regulation and motor controllers will be on the same PCB and then enclosed by a single plastic encasing This encasing will consist of threaded flanges that can be bolted to the bottom of the bed to provide the most protection For the delivery design two motor controller PCB encasings were bolted to the bottom of the bed using the bolt holes already present in the encasing that was included with the donation of the motor controllers The ECU PCB and power regulation PCB were directly bolted underneath the bed using spacers to provide some space between the bed and PCB for air movement The large aluminum encasing covering all the above PCBs include threaded flanges that were used to bolt 1t to the bottom of the bed for the most protection Figure 78 shows the large aluminum encasings covering all the PCBs mounted to the bottom of the bed Figure 78 Mounting of PCB encasing 5 8 4 LCD The LCD mount requires the most adjustability for two reasons One reason 1s that the user 1s not going to be in the same position each time so the screen needs to be capable of adjusting to a position that is most suitable for the user s line of sight The second reason is that the LCD needs to be capable of being moved out of the way while the system is not in use For example when the user is being placed into or being removed from the wheelchair the LCD needs to be stored away to prevent any damage to the LCD or incon
163. orms in mind Finally teams present their projects to faculty family and friends at the Senior Design Banquet in May 1 3 Project Statement The main objective of this project was to design prototype and deliver a motorized stroller that would provide Isaac with adequate mobility and comfort The focus of the design was in the areas of safety reliability and ease of use Isaac controls the stroller with the touch of his finger and he sees where he is going via an LCD that is wired to a camera in the front of the stroller With two electrical and two mechanical engineering concentration students this project presented several electrical and mechanical design challenges that were appropriate for this team Even though this product has been specifically designed for Isaac Postma it has the potential to benefit individuals with similar disorders Providing the gift of mobility to otherwise immobile people was the overall goal of this project Team 3 Achieving Mobility Final Report Page 2 of 160 May 11 2011 14 Meet the Team The team 1s comprised of four senior engineering students two with a mechanical engineering concentration and two with an electrical and computer concentration Pictured from left to right in Figure are Matt Last ME Rob VanderVennen ME Matt Rozema EE and Dan Evans EE Figure 1 Picture of Team 3 L to R Matt Last Rob VanderVennen Matt Rozema Dan Evans Matt Last 1s from Syracuse NY and will be grad
164. otors The motors convert electrical energy from the battery into mechanical energy to the shaft giving the wheelchair the ability to move The motors have enough torque to provide the required speed for the maximum weight capacity They are controlled by the user and have a fixed operating speed Finally the motors have the ability to be disengaged by hand for manual drive 3 3 7 Brakes The braking system consists of a mechanical hand brake a mechanical parking brake an electrical brake and an electrical emergency brake The mechanical hand brake provides an attendant pushing the wheelchair in manual drive the ability to slow stop the wheelchair The mechanical parking brake consists of an independent lever system on both wheels that provide the user the ability to lock the wheels in place The electrical brake provides the ability to stop when the user releases the touch button Finally the emergency electrical brake provides an attendant with the ability to shut off the motors using a button The emergency brake also automatically shuts off the motors if the batteries die 3 3 8 Enclosures The enclosures protect the critical electronic components including the battery motor controller PCB ECU PCB and the power regulation PCB from harsh environments 4 Electrical Design This section describes the design of the electrical system for the device The electrical design is split up into two main subsections hardware and software Hardware
165. ototype of second camera design Team 3 Achieving Mobility Final Report Page 100 of 160 May 11 2011 After hooking up the camera to the prototype and viewing the image on the LCD screen the team determined that the downward angle was too small and the upward angle was too large The design was revised to make the upward angle 0 and the downward angle 30 net angles from horizon are 7 upward and 23 downward It was also determined that the mounting plate was too large and could be shrunk to better fit underneath the bed and protect the camera from hitting objects upon accidental impact 1 inch was removed from both the length and width of the plate The third camera mount design with the above mentioned revisions is shown in Figure 76 Figure 76 Second prototype of third camera mounting system After testing the third prototype the team determined that using a touch button did not allow for a sturdy enough camera mount As the stroller moved the camera would vibrate and skew the picture on the LCD screen For this reason a set screw was chosen to replace the touch button A set screw still allowed for the camera to be removed but also stabilized the camera in place preventing any vibration Figure 77 shows the final camera mount system for the final product i Figure 77 Final camera mounting system for final product Team 3 Achieving Mobility Final Report Page 101 of 160 May 11 2011 5 8 3 Printed Circuit Boards F
166. otype budget for Achieving Mobility Aluminum Shelf persq in 3008 900 10000 Aluminum Tube perin 3029 600 10208 PCB Components 245 Team 3 Achieving Mobility Final Report Page 128 of 160 May 11 2011 Table 46 continued Website Template 67 00 67 00 40 00 40 00 AAA 5 952 53 In order to pay for the delivery design substantial donations were acquired over the course of the year Table 47 shows the funding that was acquired over the course of the year Table 47 Total Donations Ivanrest CRC Donation 5 000 00 50 00 Calvin College 1 000 00 The original budget for the delivery design was approximately 2 000 Some of the main reasons that the final delivery budget was nearly triple this were buying new motors higher than expected welding cost and under budgeting most components 7 5 2 Production Design Budget The budget for the production design 1s the one that will be used for the cost analysis of starting a new business Some of the components on the final design are more expensive since discounts were given on several components and there are no donated components on the final design There are also some components that will be cheaper due to large volume orders The final product budget is shown in Table 48 Team 3 Achieving Mobility Final Report Page 129 of 160 May 11 2011 Table 48 Final product BOM for Achieving Mobility Achieving Mobil
167. ph for an incline range of 10 to 10 which will handle most normal situations Moreover the software shall be able to carry out any task associated with other design requirements 2 3 6 Reliability The electrical system should be able to withstand jolts due to pot holes and speed bumps It should also be able to withstand the vibration of traveling on surfaces of gravel bricks and asphalt The electrical components shall be able to perform for at least 5 years The electrical system shall be able to operate between the temperatures of 0 C and 50 C 2 4 Safety Requirements 2 4 1 Speed The device shall not exceed 3 5 mph 2 5 1 0 mph This includes traveling up down an incline decline with a 10 slope 2 4 2 Brakes The device shall be able to stop in less than 2 feet from full speed Applying the brakes can be accomplished by three different methods user controlled touch button system attendant controlled hand brake and attendant controlled emergency brake switch 2 4 5 Harnessing The device shall include a seatbelt to secure the user in place This seatbelt will be an adjustable strap that will run comfortably across the user s waist and shoulders It will be mounted to the frame and be secured and unsecured by use of a buckle mechanism 2 4 4 Transportation The device shall be able to be secured in the family s van The van consists of four adjustable snap hooks mounted to the floor The snap hooks need four loca
168. ports of 36 and a 5 Ib in weight distribution Table 12 Maximum deflection for each material 304 Stainless 0 00073 1020 HRS 0 00068 6063 Aluminum 0 00204 As expected the deflection for each material is very low which is why deflection was weighted the lowest in the decision matrix Any of the three materials would be sufficient based on maximum deflection 5 1 2 6 Weight The overall weight of the frame is very important because one of the major design requirements is keep the overall weight of the final product under 250 pounds Based on the total weight of the frame shown below in Table 13 aluminum weighs approximately 30 pounds less than steel Therefore using aluminum as opposed to steel would reduce the overall weight by approximately 15 since the final product weighs around 200 pounds Team 3 Achieving Mobility Final Report Page 64 of 160 May 11 2011 Table 13 Material Weight Comparison Density Ib in 3 Total Weight Ibs 304 Stainless 0 285 1020 HRS 0 284 6063 Aluminum 0 098 5 1 2 7 Final Material Choice After analyzing the possible options and making a decision matrix aluminum was the obvious choice for the frame material Originally aluminum 6061 T6 was selected due to 1ts high strength availability and cost However there were two reasons that 6063 T6 was selected instead First the team determined that strength was not going to be an issue after performing extensive FEA analyses so going with a
169. r interface and utilizes directional and velocity control algorithms In order to properly understand the directional software a distinction must be made between the directional state of the software and the directional state of the device The directional state of the software is a result of the directional state machine refer to section 4 2 2 for more detail and the directional state of the machine is the direction in which the device is physically moving These terms must be distinguished in order to properly understand the functionality of the software 3 2 1 Information Display The information display component of the user interface provides information regarding the directional state of both the software and the device as well as takes directional commands from the user This component displays the current directional state of the software to the user which allows him or her to navigate in their desired direction The information display also provides diagnostic information to the user in case of a system warning or failure such as high temperature current and voltage 3 2 1 1 Display Alternatives There were two main mechanisms considered for the directional display The first mechanism is overlaying a system of four arrows onto the corner of the LCD screen as shown below in Figure 4 This would be the most aesthetic option for the user but it would also be rather difficult to implement with software because it would require overlaying t
170. radius of the structural tube T 3m a Inner radius of the structural tube r 35m E 9993100psi Modulus of Elasticity P 150 Assuming max loading of 4001b with 300lb on the main bed and 100 i s on the rear bed Half of 300lb will be felt on each side wat Weight distribution hh L W 216 Team 3 Achieving Mobility Final Report May 11 2011 Driving Equation 3 4 d y dy P VzEI W zE IL Shear Farce and distributed load dc d d EIC m1 cy Integrating from above de d p EC Cy Integrating Again dx 3 2 d Jp pp ELO teyx 0 Integrating again dx 6 ay at both ends therefore 2 _ wL L 227 13 3 4 2 ya a cy bx E I y 0 24 6 12 4 y 0 at both ends therefore L ty 01 3 Deflection of Bed 4 4 Tf f 4 W L WL Cy L n 2 3 4 2 Wa OE wita La 24 6 12 4 Y Page 153 of 160 Team 3 Achieving Mobility Final Report May 11 2011 Stress in the front support beam Lengths Angles cpp 23deg Epc 2deg Foca deg On Ay deg Driving Equations Pa 375lbf Pg 37 51bf Guess Fog 30lb Fpp bf Fan 50lbf Fan 301bf Given Ma Log Fog sin Scpe Bas Lag Ppcos Pp A Leg Fep sin Opp Pax Fy Fap cos 9p a Fap cos pc Fep cos ppe Fop cos cpg Fy Fapsin p ay Fapysin pca Feg sin dggc Fopsin cpe Pa Pp Mg Lap Fapsin pca Bas LA
171. rakes Jobst Brandt Frequently Asked Questions about Bicycles and Bicycling Harris Cyclery 2005 Wikipedia Web 30 Nov 2010 lt http www sheldonbrown com brandt brakes html gt Aivd Code 185mm Dics Brake Galvanized Tree Fort Bikes n d Web 30 Nov 2010 lt http www treefortbikes com product 333222337854 454 A vid Code 185mm Disc Brake html gt Invacare Corporation Parts Catalog Invacare n d Web 30 Nov 2010 lt http www invacare com cgl bin imhgprd inv_catalog partsPHII_home jsp s O amp pagePostCount 3 amp bvModelNbrName null amp bvFormNumber null amp partsHome searchCategoriesToParts amp formNbr 10076 amp categoryDropDo wn 24898 gt Tektro R556 Long Reach Brake Set Bicycle Warehouse n d Web 30 Nov 2010 lt http bicyclewarehouse com product tektro r556 long reach brake set 12028 htm site g00gle_base gt Nexus Right Hand Brake Lever Bike Parts n d Web 30 Nov 2010 http www bikeparts com search results asp idZBPC345244 Convection Heat Transfer The Engineering Toolbox 2011 Web 27 Apr 2011 lt http www engineeringtoolbox com convective heat transfer d_430 html gt DC 96F DC Series Heavy Duty Electronics Enclosure Polycase 2011 Web 18 Apr 2011 lt http www polycase com dc 96f seriesqty 1 amp x 13 amp y 12 gt Riley William F Leroy D Sturges and Don H Morris Mechanics of Materials 6th ed Hoboken John Wiley amp Sons Inc 2007 700 01 Print Friction and Coeff
172. ram of the voltage regulator for the camera is Shown below in Figure 14 LM317 Voltage Regulator 9 Volt Node Figure 14 Voltage Regulator for Camera Although R1 could be any value R1 was chosen to be 240Q because that was a recommended value The value for R2 was determined from the equation R2 Vout 1 25V 1 Po This equation was provided by the LM317 datasheet Ig was measured to be 50u A using a ammeter Using the equation R2 was determined to be approximately 1 5kQ The voltage regulator calculations laa R2 for each component can be found in Appendix E Voltage Regulator Calculations 4 1 5 2 Current In this particular circuit the current is a concern because the batteries are able to output 825A if the circuit is short circuited If this much current travels through the electrical components they would burn up Therefore protection is added into the circuit design in order to provide safety The most significant amount of current goes to the motors Through extensive testing of the motor control system by the electrical engineers on the team it was determined how much current the motors can draw from the batteries The current when the stroller initially starts moving requires the most current Team 3 Achieving Mobility Final Report Page 31 of 160 May 11 2011 due to the lack of momentum This initial current can reach up to about 18 A The current used when the stroller is moving forward at a constant veloci
173. rame of approximately 510 psi located in the front supports with a 2001b load on the bed and a 1001b load on the storage shelves Team 3 Achieving Mobility Final Report Page 157 of 160 May 11 2011 Stress won Mises Ibf in 2 6000 5400 4800 4200 3600 3000 2400 1800 1200 600 D Load Case 1 of 1 X Maximum Value 21082 Ibf in 2 Minimum Value 3 84211 Ibf in 2 0 000 7 463 in 14 925 22 388 1 lt Design Scenario 1 gt Figure 102 Max stress in the storage area is approximately 6000psi with 125lb load Displacement Magnitude in 0 055 0 0495 0 044 0 0385 0 033 0 0275 0 022 0 0165 0 011 0 0055 0 Load Case 1 of 1 Maximum Value 0 053761 in Minimum value D in 0 000 7 463 in 14 825 22 388 1 Design Scenario 1 gt Figure 103 Max deflection in the storage area of 0 053 with 125lb load Team 3 Achieving Mobility May 11 2011 Final Report Page 158 of 160 12 3 Appendix C Motor Power Speed and Torque Calculations Motor Parameters W tot 450lbf WF 0 40 u 0 65 0 5deg FS 1 5 D 12 5in Veleg 3 5mph GearRatio 24 Circumference Cir TD Required Wheel RPM Maximum weight requirement user wheelchair Weight Factor fraction of how much each rear wheel carries of total weight Rubber on Dry Asphalt Engineering Toolbox Maximum incline requirement Factor of Safety Diameter of rear wheel Maximum speed requirement Motor gearbox ratio
174. rmterfue n 58 60 2 1 11 2 16 11 1006 DetailedSoftware Requirements 5 8 10 11 810 11 12 10 100 Howcars 6 20 25 115 0 11 2 00 1006 Temperature Regulation 6 15 10 2 28 11 3 7 11 100 CANDrvrs 8 35 12 14 10 12 25 00 100 rection Control Software 30 50 60 1 10 11 2 18 11 100 SpeedControlSoftware 60 60 40 12 24 0 3 17 11 100 Mechanical O 126 Structure 126 291 224 10 710 3 30 11 100 SelectBetween Aluminim or Steel 4 8 6 _ 10 11 10 10 14 10 100 ModelCurrentWheelhairin SW 6 16 10 10 7 10 10 14 10 100 12 8 1o 14 10 10 21 10 1005 1 1 Analyze frame support 16 4 10 22 10 10 29 10 10096 angles storage IP Ul IP N OY tO Team 3 Achieving Mobility Final Report Page 150 of 160 May 11 2011 Adjust frame to meet angle reqts 5 12 8 11 15 10 11 19 10 100 Redesign Frame based on Ebling 20 11 29 10 12 3 10 100 Imput 2 Mounting components attachments 12 6 10 12 10 10 10096 12 6 10 12 10 10 10095 Storageshelfattachment 5 12 6 Build Frame Prototype 21 so so 2 4 11 3 14 11 100 10 Wheel 37 24 10 10 710 12 9 10 100 8 8 11 22 10 11 26 10 10096 Design bearing subassy for front 5 1 wheels FU 12 6 10 12 9 10 100 ModelBedinSolidWorks 6
175. rs and mentors this project would not have been possible therefore the team has utilized these resources to manage and organize the project more efficiently Figure 96 shows the overall structure and organization of the team Team 3 Achieving Mobility May 11 2011 Bed Brakes Encasings Mo unting Final Report Page 136 of 160 Mentors Tim Theriault GE aviation Jim Nammensma Ebling 4 Son Sponsors lvanrest CRC Calvin College Texas Instruments Communication 4 Protocol Development kit Microcontroller Motor Control Hardware Software Team Advisor Professor VanderLeest Russel Ramsey TI Greg Bush Gentex Rob VanderVennen Dan Evans Cables Camera LCD Screen PCB Power regulation Power supply Sensors Motor Figure 96 Overall team organization 8 3 Schedule and Milestones 8 3 1 Figure 97 shows a portion of the WBS showing the start and finish dates and duration of major tasks A more detailed and complete WBS is shown in Appendix A and outlines all tasks for the entire year Task Name Duration Start Finish EE May Project Management 165days Thu 10 740 Fri 5 2041 Sree ee AE ES Presentations 155 days Wed 10 20 10 Fri 5 20 11 PPFS Draft 15 days Mon 10 2540 Fri 11 12 40 Finalizing PPFS 6 days Mon 11 29 10 Mon 12 6 10 Final Report 141 days Thu 10 740 Wed 4 20 11 Create Website 13 days Mon 11 8 10 Wed 11 24 10 Budget 48 days Thu
176. rs come standard with several threaded holes in its top housing that can be utilized for bolting to another plate Initially the motors were going to be mounted using the clamp style mounts from the old donated motors but these were found later on to not align correctly with the new purchased motor s threaded holes The two alternative designs for a mounting plate involve one that is permanently welded to the frame and another that is fitted around the frame pipe using bolts Figure 82 shows the clamp style mounting system for the motors Figure 82 Clamp style mount system The weld style mount was chosen as a better design because it is much easier to make sure the motors are directly perpendicular to the ground and parallel to each other Also the clamp style mounting system must be re adjusted every time the motors are removed while the weld style system 1s always in the correct orientation Figure 83 shows the weld style mounting system with the motors attached Figure 83 Final motor mounting system Team 3 Achieving Mobility Final Report Page 105 of 160 May 11 2011 5 8 6 Touch Button The touch button should be adjustable to meet the needs of the user Not every user is going to want the touch button in the same location so 1t should not be fixed The touch button should however be configured in such a way that prevents the button from moving rotating or sliding away from the user during operation The difficulty in design i
177. rst voltage regulator that attenuated the voltage from 24 V to 12 V had the highest chance of overheating because 1t also had the most current traveling through 1t A thermocouple was used to measure the temperature of the LM317 voltage regulator The excel graph below in Figure 89 shows the temperature under various conditions LM317 Temperature as a Fuction of Time no heatsink g AAAAA AA om 444 O0000000000000000000000000 BA 90900000 B with heatsink Q o cD Sen ded o 19 oa E 19 gt A heatsink with 12V fan O heatsink with 24V fan 100 150 Time seconds Figure 89 LM317 Temperature as a Function of Time It was determined that a heat sink and fan should be used for the voltage regulator if the LM317 was used Team 3 Achieving Mobility Final Report Page 113 of 160 May 11 2011 6 1 5 Battery Charger The battery s voltage was tested using a multi meter When the batteries are being fully charged and are directly connected to the charger still the voltage 1s at 29 4 V When the batteries are unhooked from charger the batteries were at 27 V Then the batteries were run for a minute in order to remove the surface charge and the voltage was 25 9 V The battery charger was used to recharge the final product s batteries 6 1 6 Motor The functionality of the motors was tested to determine if they run when a voltage is applied to them The brake release terminals were connected to a 25
178. rticular situation 3 System Design The following section examines the system design at a component level Achieving Mobility is divided into three systems based on functionality electronic hardware software and mechanical systems These systems are further broken down in several subsystems or components A diagram of the system architecture is shown in Figure 2 Blue blocks represent electronic components yellow blocks represent mechanical components and green blocks represent user interface components A use case diagram of the device is shown in Figure 3 This diagram shows all major operations available to any type of potential user Green User Interface ie Electrical Component Yellow Mechanical Component Figure 2 Top level System Architecture Team 3 Achieving Mobility Final Report Page 9 of 160 May 11 2011 Achieving Mobility extends Change direction K Move Full Speed IN extends View Diagnostic Information View current direction Manual Operation Emergency Stop chnician Stop Device Change Top Speed Calibration Attendant Figure 3 Use Case Diagram of Achieving Mobility 3 1 Electrical Hardware The electrical hardware includes all physical electronic components that are needed in order for the stroller to function according to the requirements These can be either digital hardware components such as the microcontrollers or analog such as sensors and power reg
179. s making it adjustable at certain times and fixed at others Also the touch button along with its mounting should be capable of being removed and stored away when it is not in use by the user The two design alternatives for the touch button included a multi pivot system that folds out from underneath the wheelchair The other design involves attaching the button to a mount that the user s arm is strapped into A comparison of both alternative designs is shown in Table 34 Table 34 Comparison of the two touch button mounting design alternatives Touch Button Advantages Disadvantages More adjustable e More difficult to manufacturer Multi pivot More aesthetic professional looking e More expensive Easier to store More comfortable e Only certain positions Strap in arm Less expensive e Requires more room to store rest More stable for finger during operation Easier to build in house Based on the difference in advantageous and disadvantageous of both options a strap in arm rest was chosen as the means of mounting the touch button Figure 84 shows the design of the strap in arm rest with the touch button attached Figure 84 Design of strap in arm rest with touch button attached Figure 85 shows three prototypes that were built to test on the customer to determine the most comfortable design and dimensions Team 3 Achieving Mobility Final Report Page 106 of 160 May 11 2011 Figure 85 Three strap in arm rest prototypes that were
180. sed because it 1s designed to provide large amounts of power for only a few seconds whereas the battery needed for the device needs to be used for at least two continuous hours An analysis of the two primary battery options is required in order to choose the best battery type 4 1 7 3 1 Deep Cycle Sealed Lead Acid Battery These types of batteries are used primarily in wheelchairs scooters and industrial equipment An example of a sealed lead acid battery is shown below in Figure 11 Team 3 Achieving Mobility Final Report Page 27 of 160 May 11 2011 3 POISON Pg SEALED AGM DEEP CYCLE F F Tu s Figure 11 Seal Lead Acid Deep Cycle Battery The Power Tiger wheelchair which was donated to the team used this type of battery This battery is 402 cubic inches weighs 38 8 lbs and is rated for 12 V 55 Ah The strong track record of using this type of battery in most wheelchairs suggests that this type of battery is capable of providing sufficient power to the product 4 1 7 3 2 Lithium Ion Battery Lithium ion batteries shown in Figure 12 are known for their high energy density which makes them lighter than other leading battery types A lithium ion power source would make the stroller easier to manually push provide more storage and require less torque from the motors Therefore lithium ion batteries were heavily researched as a possible alternative E Figure 12 Lithium Ion Battery 4 1 7 4 Battery Selection
181. since these libraries are stored in ROM they do not need to be allocated in the internal flash memory allowing for more system level code to be stored in the flash Moreover these libraries have been extremely valuable to the software development of the device Throughout the design of the software these built in drivers were used to configure GPIO ports and pins for reading the state of the touch button and for driving the LEDs configuring the CAN interface and controller initializing the timer and SysTick and configuring their interrupts and configuring the UART interface for run time debugging and testing 4 2 6 Debugging Several functions were written to allow for easy debugging of the software These functions mainly use the on chip LCD and UART to display run time information The on chip LCD can be used to display real time voltage current temperature and fault statuses for each of the motor controllers This was helpful for testing the motor controllers to make sure that they were behaving as expected A drawing of the LCD screen in debug mode is shown below in Figure 41 The top and bottom status section monitor each motor controller individually Team 3 Achieving Mobility Final Report Page 54 of 160 May 11 2011 Achieving Mobility Current Direction Forward Vbus 24 9 V Vout 12 2 V Curr 5 43 A Temp 25 9 C Vbus 25 1V Vout 12 2 V Curr 5 28 A Temp 24 8 C Figure 41 Drawing of LCD Screen Used for Debugging Al
182. sion resistance of each material is very important because this product will be exposed to rain snow and salt Having a rusty product is not acceptable therefore this decision alone makes it very difficult to choose any kind of low carbon steel The level of corrosion resistance is based on the oxidation of elements as well as corrosion resistance tables found in The Engineering Toolbox 5 1 2 3 Weldability Having the ability to weld each material is very important because it determines if the frame can be made at Calvin College or if it must be welded off site Hot rolled steel HRS can be metal inert gas MIG welded using the equipment at Calvin College quite easily On the other hand stainless steel must be welded using MIG welding with stainless wire or tungsten inert gas TIG welded This could be done at Calvin College but not very easily since the welding stations are set up for low carbon steel Similarly aluminum would be very difficult to weld at Calvin College because it involves only TIG welding Phil Jasperse the head of the Calvin College machine shop stated that learning how to weld aluminum well enough would be extremely difficult by the end of the year Aluminum welding would therefore have to be done off site with the help of another individual or company 5 1 2 4 Strength The strength of each material is important since the wheelchair is required to support 200 pounds in addition to the weight of all components and ma
183. so the UART module was used to print out these same real time measurements to a terminal Using the terminal these numbers can be logged saved and evaluated after the test This data was used to correctly design the power regulation system to determine the proper gauging of wires and to set the proper current thresholds for the velocity controller A sample printout of UART measurements is shown below in Figure 42 This is a real sample of the motors accelerating from idle to 2 5 mph Each column of data represents the status of one of the motor controllers The number on the far right is a timestamp giving the exact time of the sampled data This debugging technique allows for constant real time monitoring of bus voltage output voltage current and temperature Also an Excel macro was created as a supplement to this data allowing the team to quickly produce graphs and analyze the data received in the terminal Team 3 Achieving Mobility Final Report Page 55 of 160 May 11 2011 Figure 42 Sample UART Output of Debugging Data Team 3 Achieving Mobility Final Report Page 56 of 160 May 11 2011 A typical graph showing motor voltage and current for each motor controller is shown in Figure 43 This graph shows a visual depiction of the status of the motors throughout the test and the relationship between voltage increases and the corresponding current spike These graphs were very useful throughout the testing process to make sure that the mo
184. st product The production design is currently not differentiated enough to sustain a company and have a large enough market To solve this problem several different options can be designed into our product in the future to expand the target market This includes various user interfaces touch button joystick sip and puff device tongue device variable speed increased top speed weight capacity incline capability adjustable recline and height positions 7 2 Competition The wheelchair market consists of more than 100 different suppliers of wheelchairs Of the vast amount of suppliers only the manufacturers of powered wheelchairs with the largest market share will be analyzed and used in a comparison to our product These suppliers include Invacare Pride Mobility and Permobil For each company it was determined which model was the most similar to our design as shown in Figure 92 Invacare FDX Model Pride Mobility Quantum Model Permobil Corpus Model 58 59 60 Figure 92 Competitive models in the market Team 3 Achieving Mobility Final Report Page 121 of 160 May 11 2011 These main competing models have different performance specifications as shown in Table 39 Please note that these competing models are being compared to the base production design To sustain a company a new production design will need to be developed with differentiated options that would make the product more competitive with the similar models in the market 58
185. t eaten E tema 113 6 1 8 imde aida 114 02 Inmtestation Test Plans sita eons apa tes ua tear aec alan Ia us ta d o cR hac pui onu ecu 114 6 2 1 Elecmical Tesine mr CER 114 6 2 2 Mechanical De SILIO aono Pe E 117 T SA dl A A O AS 119 7 1 Busmess Mod A 119 7 1 1 NSC a TS MS EAA S EEA SE AE SEE AEA S ASE 119 7 1 2 ls E E 120 P OD a E E ta Moon eo E UE 120 7 3 Muke t ReSe IET a nin rasta lt e obse Copied entem lao E e e 122 7 3 1 COH SEOTHE T D SEL A o 122 1 922 Maket SIZE ENTRE m UU MT 123 7 A LN eH A hada tieamuaeanseidealtaseuad aaeatentereaeenaaaes 125 7 5 a O ter ep erent A an Rae A 126 dl Dever Desom BUC uc de E 126 TEL Production Desten Bud reto N a 128 7 5 3 A O Mi ate 130 7 5 4 Iuxed and VAS e ta 131 T99 Thiee year Hilancrab olt lOOK ninia ii a 131 o Project Management odo dtes iet stp A 133 8 1 AN OPK DINOS A ll ii 133 8 1 1 Hard Wal ONT dial iaa 133 8 1 2 A ci otto uncos dient tain c iot aded odeur shuts eed cedes 133 8 1 3 Mecanica OE N 134 5 2 tMTeamOreanizati n and Mana eMe aaa 134 8 3 Schedule and Milestone da o e is ad be de uo el o 136 8 3 1 vieille E N DM ETE 136 9 3 2 AU EUIS ATI S ERNEUT HET 137 OS AACKMOW AU T TEE 139 10 ose DIOE ETE Ser 140 11 SA O 141 12 JNDDeldiCe Sidi A A tet le oed oque deka eon 147 12 1 Appendix A Work Breakdown Structure eaei e E E 147 12 2 Appendix D Stress Calculations iii 152 12 3 Appendix C Motor Power Speed and Torque Calculati0DS oooooooonnnnncnnnonnnn
186. t kit for the ECU dictated the particular selection of MCU The ECU also includes a Luminary Micro Cortex M3 microcontroller specifically the LM352965 This MCU operates at 50 MHz with 256 kB Flash memory and 64 kB SRAM It includes all necessary features such as fully configurable GPIO pins timers and 2 CAN modules with 32 message objects with individual identifier masks Overall this MCU is very similar to the LM3S2616 used in the motor controllers and they can essentially be viewed as the same unit for the purposes of this project 4 1 5 Sensors There are sensors for three independent values current temperature and battery life These sensors are discussed in the following sections 4 1 5 1 Current Sensor The velocity control system cannot be utilized unless there is a feedback sensor to provide data to the ECU regarding the current draw of the motors The current sensor will actually measure the current that 1s being applied to each motor Refer to section 4 2 4 for more detail about the velocity control system 4 1 5 2 Temperature Sensor The ambient temperature of the motor controllers is monitored by the ECU on a periodic basis The selected motor controllers come with a temperature sensor that monitors the temperature This sensor 1s vital to the temperature control system discussed in section 4 2 7 4 1 5 3 Battery Monitor The battery monitor was purchased from Argus Analyzers It is designed specifically for 12 V deep cyc
187. t need to construct them from scratch or through software which is less reliable Third an expandable peripheral set including UART LCD LED and even video VO drivers was also an advantage for the kit so that 1t could be easily expandable to provide a more robust Ul and more communication protocol options The software package SDK including IDE and driver libraries was also considered in the selection process Finally the amount of certainty we had that the kit would drive our particular DC motors was considered heavily Cost was certainly a factor especially with a limited budget but since the choice of development kit is critical to the overall design even expensive kits were considered providing that they fulfilled the given criteria A summary of the pros and cons for each development kit alternative is shown in Table 5 4 1 1 1 1 TMDSIMTRPFCKIT This is a motor control development board from Texas Instruments combining both digital and analog hardware on a single board including a single 32 bit MCU designed to control two motors This board uses the F28035 controlCARD programmed by the Code Composer v3 3 software development package It is also designed to power two 200 W permanent magnet motors 4 1 1 1 2 RDK BDC24 This is a motor control reference design kit from Luminary Micro that serves as a variable speed motor controller for both 12 V and 24 V brushed DC motors at up to 40 A continuous current This module is powered by th
188. t temperature Also if the ambient temperature is above 32 C 1 e 24 C 8 C the fans will turn on regardless of if the motors are running If the temperature of the device falls below 0 C the motor controller will shut down due to an error condition Similarly if the temperature reaches 50 C and keeps rising regardless of the fan the device will promptly shut down This is a simple control system but it helps keep the motor controllers at a safe operating temperature 4 3 Future Work Production Design The idea of a production design has been discussed a few times so far but this design exceeded the scope of the project However this section briefly describes the steps that would go into creating a production design 4 3 1 Hardware Topology Section 4 1 3 describes the electronic hardware topology used in the delivery design However as mentioned this topology is not necessarily the most cost effective design because it involves three individual MCUs on three individual PCBs Therefore the topology of the delivery design will most likely not be used in the production design The selection of the MCU hardware topology for the production design will be based on cost size functionality simplicity of implementation and reliability For this design reliability and functionality will hold the greatest weight Most of all we want the customer to trust the performance of this device and this is reflected by the weighting of the sele
189. ted Finally the spacing around all pads were increased so that it would make soldering easier A diagram of the revised schematic is shown below in Figure 25 Team 3 Achieving Mobility Final Report Page 38 of 160 May 11 2011 Figure 25 Revised EAGLE Schematic 4 1 8 11 Revised EAGLE Board Design The revised EAGLE board design 1s shown below in Figure 26 TAS Yt el Figure 26 Revised EAGLE Board Design 4 1 8 12 PCB Fabrication The printed circuit board was fabricated by Johnson Controls Incorporated in Holland MI The picture below shows version printed circuit board Team 3 Achieving Mobility Final Report Page 39 of 160 May 11 2011 Figure 28 Final Version of Power Regulation Circuit 4 1 8 13 Flyback freewheeling diodes Flyback protection is used to protect the circuitry from Back Electromagnetic Flux BEMP voltage that could harm the circuit The electric motor has an inductive nature and thus builds up energy If the power 1s suddenly stopped 1 e blown fuse the energy from the electric motor could travel through the circuit in the opposite direction that 1t was designed for and possibly damage components Team 3 Achieving Mobility Final Report Page 40 of 160 May 11 2011 A diode would normally be placed across the motor so that the current would travel through the diode instead of the circuit This would release the built up energy without sending the current through the circuit in the opposite
190. terial Pound for pound steel is the obvious choice when 1t comes to strength however aluminum will provide more than enough strength for the structure based on the calculations shown in Appendix B Stress Calculations A finite element analysis FEA was also performed to determine 1f the strength and deflection were sufficient for our design and 1s also shown in Appendix B Stress Calculations Based on the FEA analysis the max stress in the stroller is far below the Team 3 Achieving Mobility Final Report Page 63 of 160 May 11 2011 yield stress therefore strength will not be a problem The yield strength of each material analyzed was found in The Engineering Toolbox and is shown in Table 11 Table 11 Strength properties for each material Material Yield Strength Young s Modulus psi Mpsi 304 Stainless 42100 1020 HRS 53700 6063 Aluminum 25000 5 1 2 5 Deflection Deflection is the degree to which a structural element is displaced under a load The deflection for each material is based on the formula 5wL v 384EI where w is the load distribution L is the length to the center of the tube E is Young s Modulus and I is the moment of Inertia Since w L and I are the same for each material the deflection is inversely proportional to E Aluminum will therefore have three times the deflection of steel Based on the current design each material will have the deflections shown in Table 12 based on a distance between sup
191. tex M4 will be backwards compatible with the M3 series JTAGISWD O input and l Reset switch Debug out LED Speaker Switches Powa LED Be T nd A Slalus LED 1 Li Selec switch OFM Graphics muc leder ter ei o riim Fe NL DII H pin LO Piu i 3 Tis break out 4 pn io PEGA 43 AI header break cut pieni E JE header atellarisd LM 3352365 u Microcontroller Ir circurt Deoug Interface CAN bus connector Figure 6 Labeled Picture of the LM3S2965 CAN Board Team 3 Achieving Mobility Final Report Page 22 of 160 May 11 2011 4 1 3 Hardware Topology One of the most important design considerations of the electronics system is the topology of the hardware The main factors in designing the hardware topology are the number of MCUs and how they interface with each other Since the MCU s must control and calculate the speed correction factors two motors independently the number of MCUs and their layout are critical decisions such that no MCU is given too much work 1 e such that no MCU 1s over utilized at any given time However the chosen topology for the delivery design had additional constraints stemming from the selection of motor control development boards as discussed in the following sections The hardware topology for the delivery design is shown below in Figure 7 Touch Button Input LED Output ECU LM3S2965 CAN Bus 32 bit
192. that is strong enough to keep the magnets separated Once this voltage is cut the magnets are quickly attracted to a metal plate that clamps down on the axle preventing any further rotation This device will be used for the emergency braking system in parallel with an on off toggle switch Whenever this switch is flipped off or the batteries run out power to the wheelchair as a whole is cut including the brake release which in effect immediately stops the motors Figure 66 shows a picture of the switch used in the delivery design Figure 66 On off toggle switch for emergency braking Team 3 Achieving Mobility Final Report Page 89 of 160 May 11 2011 5 6 4 Financials Table 23 and Table 24 shows the costs of the production and delivery design respectively Several of the costs were based on pricing from other companies to obtain a rough estimate however for the production design most of these parts will be purchased from manufacturers reducing the price pre mark up Table 23 Production Design Costs for the Braking Systems Production Design Cost Rim Brakes 39 98 Based on actual purchase of Sunlite MX brake set from Kentwood Cycling and Fitness 19 99 per brake set 0 00 Included in the above cost Parking Brakes 74 86 Based on pricing of an Invacare FDX parking brake assembly 37 43 per wheel lock Electronic 0 00 Cost included in controller and motor costs Brakes Emergency 0 00 Cost included in motor costs Brakes
193. the speed going up and down hill were calculated to be just within the required speed window However as mentioned earlier the terrain was 20 degrees steeper than the requirement so the speed will stay even more constant with the suggested terrain of 10 degree slopes Table 38 Time Trials to Calculate Speed of Stroller Trial 1 sec Trial 2 sec Trial 3 sec Average sec Speed mph Tus a 89 s a CT 30 degree incline 30 degree decline 59 Jes 62 62 33 6 2 1 2 Electric Brake Testing Since the requirements state that the stroller should come to a complete stop within 2 feet after the user releases their finger from the touch button this functionality was thoroughly tested This was tested by having a user in the device traveling at full speed and then releasing the touch button Since measuring the stopping distance proved to be inaccurate the testing was done by performing calculations with the UART debugging data Since the stroller has a top speed of 2 5 mph the amount of time to brake was calculated to be Team 3 Achieving Mobility Final Report Page 116 of 160 May 11 2011 2 5 mph 0 545 s 2 ft Thus the maximum braking time was 0 545 seconds The following data was measured using a 200 Ib occupant on flat ground Motor Deceleration From Full Speed Output Voltage V o 99999 55 55 5 56 56 5 57 57 5 58 Run Time s Figure 91 Motor Deceleration from 2 5 mph as a Function of
194. the team since they will be living in close proximity to the customer 3 2 5 PC Interface The PC interface allows a technician to connect the device to a PC in order to install future software and firmware upgrades For the delivery model the firmware installation can be done using the LM Flash Programmer application provided in the evaluation kit This is a PC application provided by Luminary Micro that allows software downloads to the device through JTAG interface 3 3 Mechanical Hardware 3 3 1 Frame The frame provides support for the entire stroller and is strong enough to hold the weight of the equipment and passenger It provides adequate areas for storage as well as mounting capabilities for all necessary equipment 3 3 2 User Interface The mechanical user interface consists of a touch button and a system of LEDs which allows the user to control the motion of the wheelchair 3 3 3 Bed The bed provides the user with a comfortable place to lie while operating the device 3 3 4 Storage Storage provides the user with a location to store his or her medical equipment so it is easily accessible and always nearby For this prototype medical equipment includes a feeding bag as well as a suction device 3 3 5 Wheels The wheels allow the vehicle to facilitate movement They provide the necessary traction and strength to support and transport a load Team 3 Achieving Mobility Final Report Page 17 of 160 May 11 2011 3 3 6 M
195. tions on the stroller to hook to 2 4 5 Electrical Shock Safety Since the device uses electricity the proper precautions must be taken in order to protect people from electrocution Large currents of electricity flowing through one s body can burn tissue freeze muscles and fibrillate one s heart This has a potential to occur when a voltage difference is applied between two points on that body Therefore the user should not be able to contact two different nodes of the wiring and all open nodes should be covered Team 3 Achieving Mobility Final Report Page 7 of 160 May 11 2011 2 5 Design Norms Design norms are a set of moral guidelines that try to incorporate ethical issues with technical engineering issues Ethical judgments are often made in engineering and design norms heavily affect this decision process There are several different design norms that the team has focused on throughout the course of this project These design norms include caring integrity trust and transparency 2 5 1 Caring The design norm of caring should show care towards all individuals affected by the end product The final product should take into account the social physical and psychological effects of all individuals involved This product shows caring because its main function is to give mobility to those in need This product is specifically designed to give individuals namely Isaac Postma a new perspective on life The goal is that individuals
196. to find a motor control development kit or set of kits that would work for the project with a fairly small learning curve Therefore a large amount of research was conducted on these development kits and several alternatives were considered before making the final decision The selection of development kit options was limited severely because the vast majority of kits were made for brushless DC motors There were four main options for development kits each with their own pros and cons These development kits were evaluated with regard to the needs of this project and the relative ease of implementation and integration with the rest of the stroller Table 4 shows a list of the development kit options that were considered along with the associated MCU FPGA Table 4 List of Development Kit Alternatives Instruments RDK BDC24 LM3S2616 Cortex M3 DM163029 Microchip PIC16F917 PicMICRO Mid Range 16F91x EP2C35F672C6 Cyclone Il 2C35 There were several criteria that were pivotal in the development kit decision First it was desirable for the kit to include the analog power components necessary to drive the motor s These components Team 3 Achieving Mobility Final Report Page 19 of 160 May 11 2011 include but are not limited to an H bridge with power transistors and protection diodes refer to section 4 1 12 for the more information on these components Second built 1n hardware based PWM output drivers were an advantage so we would no
197. tors were not overloading the motor controllers with too much current draw Motor Sampling Data as a Function of Time Vout A Currentl T i e E al E di fe a ue m q A Current 2Z 32 Time s Figure 43 Example Graph of Motor Status Data 4 2 7 Temperature Control Maintaining the motor controller s temperature within an operable range as dictated by the components specifically the H Bridge transistors 1s necessary for safe operation of the device The electronic components most likely to overheat are the transistors in the H Bridge circuit This is because there can be large spikes of current flowing through them when the direction of the motor changes which can greatly impact the heat dissipated by the transistors Therefore it will be necessary to measure the temperature of the electronic components especially the H Bridge The temperature control system resides in the individual motor controllers and is therefore not represented in the block diagram shown in Figure 34 The design of the temperature control system involves a closed loop feedback mechanism However since the system does not need to maintain a constant temperature a simple on off mechanism with hysteresis was used The target temperature is Team 3 Achieving Mobility Final Report Page 57 of 160 May 11 2011 currently set at 24 C with a hysteresis of 8 C If the motors are running the fans will turn on regardless of the ambien
198. tructure was one of the largest mechanical milestones and was dependent on several other components The size and weight of all electronic components as well as mechanical components needed to be determined before the frame could be complete The design of the frame changed as other components of the system changed This milestone was completed March 2 2011 8 3 2 7 Final Brake Design This milestone includes all three braking systems and was another large mechanical milestone that was achieved The final brake design occured simultaneously with the structure design since a change in one affected the other This milestone was completed March 2 2011 8 3 2 8 Final Component Mounting Design The final component mounting was one of the later milestones since all components had to be chosen before the mounting could be designed This milestone was completed March 12 2011 6 3 2 9 Velocity Control Software Complete This software controls how fast the user will travel in the product An open loop control system was used to regulate the speed to make sure that it moves at an appropriate speed T This milestone was completed April 1 2011 8 3 2 10 Printed Circuit Board Designed The printed circuit board was for a more professional look to the final product The circuit board was designed for proper cooling as well as fitting all of the components in a small space This milestone was completed April 22 2011 Team 3 Achieving Mobility Final Report
199. turing the device to the end user Wheelchair Manufacturer Medical Equipment Supplier Insurance Doctors Company recommend which pay for wheelchair supplier and brand End User parents of end user Figure 93 Product Flow Chart Both insurance companies and doctors have a high influence in which wheelchair supplier and which wheelchair brand the end user usually purchases The insurance companies usually pay for the product not the end user and they often have a maximum amount of money that can be spent on any medical item The device therefore must be marketed to the insurance company as a quality product that fits within the price boundaries of their maximum allowable If the insurance companies approve to fully or partially cover the device the product looks more appealing to the end user The end user also decides on a particular wheelchair based on the recommendation of a doctor Doctors understand best the user s needs and will recommend a wheelchair with features based on those needs Doctors therefore need to be made aware of this product and its special features and convinced that it better meets the needs of his her patient than anything else available Figure 28 shows the breakdown of revenue in 1996 for the wheelchair market by source Medicare private insurance and Medicaid make up 80 percent while commercial institutional doctors hospitals make up 9 percent of the revenue This mean roughly 11 percent of th
200. ty is about 6 A The wires from the battery to the motor are 10 AWG because they are rated for 30 A Therefore this wire 1s rated to handle the high current demands of the motors A fuse was also added into the circuit in order to provide additional protection to the circuit If for some reason the circuit 1s short circuited a fuse needs to be available to break so that the electrical components are not destroyed An 80 A fast blow fuse was also included because the motor controller should never handle more current than this even for half a second A slow blow 20 A fuse was included within the wiring because no more than 20 A of continuous current should be used for longer than a few seconds Shown in Figure 15 is a graph for the average time it takes to break fuses at a particular current The graph shows that at 40 A the 20 A slow blow fuse would blow in about 10 5 seconds Average Time Current Curves 15 1004 4A 5A TA 8A 10A 12A 15A 20A 25A 30A a A II nd um mi RE TENIS NN NR E ULL ER LI demnm Ilo nd oS S Sd o nd ol al LL LI e LL A LHP D 3A LAT ee E LII y esee LL LILLL E E AA II o O iii e AM AA il AAA AAA T IT xm ee m EE EE REN UU MECO AER EEC NANA AAC ll QW 1 CRA L l a i A PP ee so ee ee E LLLLL zT LT l1 llIlLL LI DIL DIDPILLLe ErL id mu Ll ZT uu Lb Lm Lm EET LL i eee TTI A ll Y Lai i Lamm LL aL LL
201. uating with a degree in engineering with a mechanical concentration For the past two summers Matt has been interning at Sensis Corporation as a System Integration and Test Engineer working on testing the components of a radar system used in airports Currently he has accepted a full time job at Bechtel Plant Machinery in Pittsburgh PA He will be working as a project engineer on nuclear power plants for submarines and air craft carriers used in the Navy After working a couple years he hopes to obtain a master s degree in business administration Rob VanderVennen is from Grandville MI and will be graduating with a degree in engineering with a mechanical concentration Rob interned for two years at Progressive Surface Inc as a mechanical design engineer working on pneumatic blasting machines Currently he works at Gentex Corporation as a Production Support Engineer in the outside electrochromic mirror division He has dealt heavily with manufacturing and machine design over the last several years Rob plans on going back to school within a couple years to start working on his MBA Matt Rozema is from Zeeland MI and will be graduating with a degree in engineering with an electrical and computer concentration Matt currently holds a part time position in software engineering at Johnson Controls and will be working there full time after graduation He is currently working on analyzing debugging and maturing software that will be used in 2012 Mazda H
202. ulation circuitry The following section provides a brief overview of each major piece of electrical hardware Team 3 Achieving Mobility Final Report Page 10 of 160 May 11 2011 3 1 1 Microcontroller The microcontroller can be thought of as the brain of the electronic hardware A microcontroller MCU is needed in the design because the control algorithms needed for this device exceed the capabilities of a simple digital circuit As seen in Figure 2 the MCU processes all the information coming in from sensors and the touch button and determines the proper response through software algorithms It also controls the LED user interface 3 1 2 Touch Button Since the customer 1s incapable of lateral movement with his fingers and wrists a touch button was used as the user s method of control for the stroller A joystick system is more typical for electric wheelchairs and 1s more convenient to navigate however moving a joystick 1s impossible for the customer 3 1 3 Sensors The sensors in the system include a circuit board temperature sensor and battery voltage sensor The circuit board temperature sensor is used to regulate the temperature to make sure the electrical components maintain a safe operating temperature The battery voltage sensor is used to display the percent of battery power left to the customer so they know how much longer they may operate the vehicle 3 1 4 Battery The battery is needed to supply the necessary pow
203. used to get a rough idea of overall size and weight Team 3 Achieving Mobility Final Report Page 63 of 160 May 11 2011 Figure 47 Preliminary design used for base case analysis Since this design had to be made longer as well as taller a brand new design was developed as shown in Figure 48 This design was mainly used for overall length and height purposes but was discarded since it had an undesirable look Figure 48 Second design used for overall height and length purposes The final frame design is shown in Figure 49 This design simply shows the frame and does not include all of the mounted components The final design of the frame was dependent on many other factors and was constantly being changed to incorporate new components Team 3 Achieving Mobility Final Report Page 66 of 160 May 11 2011 Figure 49 Final design of the stroller frame Some of the major factors that influenced the design of the frame were the overall size requirements First the frame had to be taller than his previous stroller per customer request Therefore the frame was made 3 taller to incorporate this design request If the frame was made too high the risk of tipping over would increase therefore a 3 increase in height was chosen Next the frame also had to be longer to incorporate a longer bed per customer request The customer s original bed was 52 long and the delivery design was made to incorporate a 57 75 bed The customer
204. ustomer is sleeping during the night 2 3 3 LCD Monitor There shall be one color LED monitor mounted to the device that will show the front view of the device The LCD shall provide an image that is readable in direct sunlight as well as in dim lighting The LCD monitor must be compatible with the video camera have a colored display and a resolution of at least 300 000 pixels The monitors must also measure at least 4 5 diagonally so that the user can easily see the display 2 34 Camera There shall be one camera that captures the view of what is in front of the device The camera needs to output the same video format that the LCD receives The frame rate should be over 25 frames per second Autofocus is not necessary but would be preferred 2 3 5 Software The software shall directly control the speed of the vehicle by making real time calculations It shall also provide an informational display to the customer by means of LEDs and a character LCD screen separate from the color LCD in section 2 3 3 Each LED shall remain illuminated for 2 seconds before cycling to the next The software shall contain self test functions to determine the state of the device upon start up If there is a problem the user will be notified via a sequence of LED flashes which is documented in the Team 3 Achieving Mobility Final Report Page 6 of 160 May 11 2011 user manual Finally software shall ensure that the stroller maintains a speed of 2 3 1 0 m
205. utator also has another crucial function reversing the direction of current flow The polarities of the stator and rotor magnetic fields are misaligned and the rotor will rotate until they become aligned Once they align the rotor will stop rotating The commutator reverses the direction of the current flow through the coil windings providing a magnetic field in the opposite direction continuing to rotate the rotor 5 5 3 2 Shunt Wound Direct Current SWDC Motor Permanent Magnents a Stator Coil Windings Brushed Rotor m Coil Windings Stator Body PMDC Motor SWDC Motor Figure 57 Comparison of PMDC and SWDC motors Team 3 Achieving Mobility Final Report Page 77 of 160 May 11 2011 A SWDC motor works the same way as a PMDC motor except instead of permanent magnets in the stator there are coil windings Figure 57 When current is applied to these coil windings a magnetic field is produced providing the necessary force to rotate the rotor 5 5 3 3 Brushless Direct Current BLDC Motor DC Controller Coil Windings Figure 58 Parts of a BLDC motor A BLDC motor is similar to a PMDC motor except the rotor and stator are switched BLDC motors have rotating permanent magnets in the rotor and stationary coil windings in the stator Figure 58 Because the permanent magnets are in the rotor and not the coil windings no current is required to the rotor and no brushes are necessary The magnetic fields produced b
206. ve program bottom and the system boot procedure top Once the device is turned on the initial configuration is completed Once complete the software continues to the cyclic executive which is shaded in a gray box This cyclic executive involves continually checking for interrupts servicing the state machine and servicing the motor controller If there are no interrupts to service the ECU sits idle and does not consume processor time This is the Team 3 Achieving Mobility Final Report Page 45 of 160 May 11 2011 highest level software flowchart for the system the direction controller and velocity controller modules are described in more depth in sections 4 2 2 and 4 2 4 respectively Initialize global settings Hardware Initialization No nitialization Complete imm No Poll touch button state i Query Motor Controller Status Yes AN Status Dat No Ready Service Velocity Controller 2 second timer m Service Directional Controller Figure 35 System Boot and Cyclic Executive Software Flowchart 4 2 2 Directional Control A system of direction control is necessary to provide the user with a means of navigating in a desired direction This control system must allow the user to turn in any of four possible directions Since the directional user interface of this device consists of a single
207. venience to the user Two alternatives considered for the LCD include a multi pivot system and a gooseneck system The multi pivot system consists of several linkages with various degrees of freedom to provide all the necessary positions movements that are required for the LCD screen A goose neck system on the other Team 3 Achieving Mobility Final Report Page 102 of 160 May 11 2011 hand consists of a flexible tube that can be bent into any configuration Figure 79 show an example of a multi pivot system and a goose neck system Ja Mutli Pivot System Goose neck System Figure 79 Comparison of LCD Screen mounting alternatives gt Table 33 compares the advantageous and disadvantageous of each alternative for the LCD screen mounting system Table 33 Comparison of LCD screen mounting system Advantages Disadvantages Flexible e Unlimited positions e More difficult to store Tubing e More expensive e Easier to store e Only certain positions Multi pivot Less expensive e Capable of being produce in house A multi pivot system was chosen for the production design because it has many more advantages over the flexible gooseneck most importantly the ability to make it in house and not pay more to have it built and shipped somewhere else For the delivery design we did not have the resources to build a multi pivot system ourselves and we could not find one that worked with our wheelchair layout to purchase therefore t
208. watt power supply while the motor terminals were connected to a 100 watt power supply The brake was confirmed to disengage from the motor axle because when power was applied a single click noise was heard When power was applied to the motor terminals the wheel on the motor began to rotate The wheel rotating on the motor confirmed that the motors functioned properly when voltage was applied to them This test determines whether the motors function and not their performance 6 1 7 Brakes The braking system consists of four subsystems user controlled brakes attendant controlled brakes parking brakes and emergency brakes Component testing only determined the functionality of these brake systems and not their performance Braking performance was determined under integration testing 6 1 7 1 User controlled Braking System For the user controlled brakes the electrical brakes must be able to stop the motor when the touch button is depressed To determine this the motor terminals were connected to a 100 watt power supply and the brake release terminals were connected to a 25 watt power supply Once the wheel had been rotating for approximately 2 minutes the power to the motor terminals was decreased to O watts in about 2 seconds SOW sec This simulated the user letting go of the touch button The wheel on the motor then came to a gradual stop therefore passing this test This test determines whether the brakes function and not their perfor
209. will be able to experience life with a new appreciation for the gift of mobility The device will also affect those around Isaac as it will make their life easier so taking care of Isaac is not as burdensome Each member of this team felt a special calling to help out Isaac and completing this project will impact the design team just as much as him 2 5 2 Integrity The design norm of integrity should promote completeness and portray harmony of form and function The end product should promote human values and relationships and be pleasing to the end user The final product will show integrity because it is designed specifically for Isaac Postma and the final design will be a working reliable and safe product Even though this project targets Isaac specifically any future product will be tailored specifically to an individual Therefore strong relationships will be formed between the customer and the team This product promotes human values because even though it is a product that will likely never see a large demand the team is designing it to dramatically improve the lives of a few individuals 2 5 3 Trust The design norm of trust should promote a design that is dependable and reliable for the customer Trust can be seen by steps that are being taken in order to provide a safe product to the customer This is a very important design norm for this project since an individual could be harmed if the product is not designed with safety and trust in
210. will purchase it each year increasing 50 each year as the product gains popularity Table 52 shows the projected net income at the end of each year Based on the assumed number of units sold the business will break even around the end of year two and turn a large profit at the end of year three Variable costs of goods sold are based on labor and material to make the product and fixed costs of goods sold are yearly utilities warehouse and equipment rental and IP insurance Variable operating costs are commissions based on sales and fixed operating costs are salaries R amp D and advertising Depreciation is from the purchased equipment and is based on a 7 year recovery MACRS Interest expense is calculated based on the amount of debt carried through the current year based on a 6 interest rate Team 3 Achieving Mobility Final Report Page 132 of 160 May 11 2011 Table 52 Projected net income for first three years of business operation Team 3 Achieving Mobility Pro Forma Statement of Income Units Sold Sales revenue Variable Cost of Goods Sold Fixed Cost of Goods Sold Depreciation Gross Margin 313 986 511 468 803 908 Variable Operating Costs Fixed Operating Costs Operating Income 275 015 97 968 382 908 Interest Expense 4 500 5 820 955 Income Before Tax 279 515 92 148 381 953 Income tax 40 36 859 152 781 Net Income After Tax 279 515 55 289 229 172 Since this business will not be able to get started without som
211. wolf net papers charging html gt 18 Average Height and Weight for Children Buzzle Intelligent Lift on the Web N p 2010 Web 6 Dec 2010 lt http www buzzle com articles average height and weight for children html gt 19 The Engineering Toolbox N p 2010 Web 22 Nov 2010 lt http www engineeringtoolbox com gt 20 Invacare Corporation Parts Catalog Invacare Corporation n d Web 5 Dec 2010 http www invacare com cgi bin imhaprd inv_catalog partsPHII_home jsp s O amp partsHome partsHome amp newSearch true gt 21 Metals Depot Shopping Cart Metals Depot 1999 Web 5 Dec 2010 lt http www metalsdepot com Cart3 viewCart1 phtml LimAcc amp aident gt 22 Kolczynski Jill Pride Mobilty Phone interview Oct 2010 23 Brushed DC electric motor Wikipedia n d Web 28 Nov 2010 http en wikipedia org wiki Brushed DC electric motor 24 Condt Reston Brushed DC Motor Fundamentals N p Microchip Technology Inc 2004 N pag Web 29 Nov 2010 lt http ww1 microchip com downloads en AppNotes 00903B pdf gt 25 Automotive Electronics DC Motors Clemson University Vehicular Electronics Laboratory n d Web 29 Nov 2010 lt http www cvel clemson edu auto actuators motors dc html gt 26 Electronics Tutorial about DC Motors Ed Wayne Storr Basic Electronics Tutorials n d Web 29 Nov 2010 lt http www electronics tutorials ws 10 10_7 html gt 27 Brushless DC Electric Motors Wikipedi
212. y William Mechanics of Materials 6th ed New York John Wiley amp Sons Inc 2007 Print Barett J T How Wheel Speed Sensors Work 23 September 2009 lt http www ehow com how does_5449960_wheel speed sensors work html gt Easys Rehabilitation Stroller Thomashilfen 2010 Web 12 Nov 2010 lt http www thomashilfen com th index php en gt Microcontroller UART Tutorial 2010 Retrieved from Society of Robots http www societyofrobots com microcontroller_uart shtml Williamson J C 2009 October 1 Lithium Ion Batteries for Powerchairs and Scooters Retrieved from Wheelchair Driver Team 3 Achieving Mobility Final Report Page 147 of 160 May 11 2011 12 Appendices 12 1 Appendix A Work Breakdown Structure Task Name Duration Estimated ds ERES days hrs EM Project Management ANON MIA NE Presentations ia us xm ons spon 100 Elevator Pitch Presenation 1 15 315 10 20 10 10 20 10 100 Final339Presentation 3 30 25 11 29 10 12 1 10 1005 _ 34OPresentation1 1 3 20 20 2 6 11 2 18 11 1006 34OPresentation2 3 20 35 427A1 4 29 11 1006 Banquet Night Presentation 3 3 30 35 5 8 11 5 20 11 1006 PPS ts 176 183 10 25 10 12 6 10 1005 Introduction 11 3 10 10096 Course Overview 3 2 2 13 0 19 5 10 1006 Problem CCE 3 2 2 j 13 0 19 5 10 1006 Project Statement 3 2
213. y of the metal underneath it For the delivery design the upper half of the bed must have an approximate 4 inch rise over the total length per customer request 5 4 2 Alternatives Several different companies were contacted to make a custom bed to fit our design Many companies said that they would be capable of making such a product for the team These four companies were the following Everything Upholstery Casey s Upholstery Services Perrin s Upholstery and Aacaway Bedding Barn All four of these companies are located in Grand Rapids MI After doing further research however it was determined that medical professionals would be the best to make the bedding The bedding needed to be designed to factor in various pressure points from the patient and the safety restraint system attached to the bedding had to be made under certain standards to make transportation safe and legal As a result the team partnered with Mary Free Bed Rehabilitation Hospital to make the bedding 5 4 3 Proposed Design The proposed design of the bed for the production design and delivery design was fabricated by Mary Free Bed Rehabilitation Hospital The final design of the bed depended largely on the final design of the rest of the stroller After meeting with several experts at Mary Free Bed it was decided that the bed would be made out of two layers of 1 thick memory foam wrapped in Rubatex fabric They highly Team 3 Achieving Mobility Final Report Page 73 of
214. y the permanent magnets and the coil windings provided a force to rotate the rotor Because the permanent magnets and the coil windings provide a stationary magnetic field the rotor will rotate until they are aligned and then stop To keep the rotor rotating the current through the stator coil windings must be varied This variation of current must be regulated by a special controller that alternates the DC in the coil windings 5 5 3 4 Engage Disengage Mechanism The motors will be outsourced to a manufacturer because there is not a high enough volume to make it cost effective to build them in house Most powered wheelchair motors are manufactured with an engage disengage mechanism so no further design is necessary The mechanism consists of a simple lever attached to the gearbox of the motor as shown in Figure 59 The lever is rotated clockwise 90 to engage the motor and counterclockwise 90 to disengage the motor Team 3 Achieving Mobility Final Report Page 78 of 160 May 11 2011 Disengaged Figure 59 Diagram of the motors engaged and disengaged 5 5 4 Selection Criteria A decision matrix comparing the three types of DC motors PMDC SWDC and BLDC considered for this design are shown in Table 16 Several different motor characteristics were considered and each was weighted based on its importance to the overall design Each type of motor was then scored in the different categories and then summed up to determine the best motor opti

Final Report - Calvin College

Contents

Download Pdf Manuals

Related Search

Related Contents