Home

Final Report on Robotic Manipulator Project

1. 60 CONTR01L5S M 61 AND CONTROLzS ERAT EGY 61 102 VOPIC AND CODE DEVELOPMENT eaten 62 FUTURE CONSIDERATIONS 65 J AL y 65 65 Eu Serine Mai UN QUEM eU 65 IMP XO E rr 66 APPENDIX 67 SHOULDER LOWER ARM AND FOREARM BILL OFMATERIALS 67 GRIPPER MATER ALES uu kn ua u a has hu 69 A CON TROES BIELOFMA TERA 5 uu 2 __ _ 70 SUMMARIZED COSI 71 AO SSUPPETER INFORMATION 2 acon toads 12 APPENDIX P O 73 Dy CALCULA TIONS __ _____ _____ Me nU RR 73 B 2 GEAR RATIO CALCULATIONS FOR FIRESTORM 18V DRILL MOTOR 74 APPENDIX u s 78 UN O EA T fan matos tcu 78 CZ MASTERKROOPI CONTROL CODE uQ Sh aqa ete tese dd erae 81 SLAVE OOPIC CONTR ORC ODE
2. VII i20 91020 PROJECT HISTORY amp OVERVIEWN 1 1 1 GATEWAY COALITION PROGRAM BACKGROUND essent eene ee ee eene ree a 1 E2 VD 1 INO NES p IE ID ed anes cece 4 5 n rnm 5 1 6 2000 GRADUA STUDENT DESIGN eects 6 J y 3 8 L u 7 COALITION DYNANIICS 10 PAL EC IPART A 10 2 2 COMMUNICATION 11 23 PROJECT TASK ASSIGNNIENIS L 1 u 11 2000 2001 ARM OVERVIEW 12 TL OBJIECTIVES uu l ua i a 12 3 2 DESIGN ede 12 3 3 CONETG URATIOJDN Girl aset 13 14 S PERPORMANCEHIIIGIIBIGI DS a a 17 BASE
3. Lower Arm Tube 130 200 390 20 typ All Talerances 0005 Unless Otherwise Stated Material Aluminum The Ohio State University Gateway Coalition Motor Bit 02 06 01 131 UNC Size 43 holes T Ll The Ohio State University Gateway Coalition 650 All Tolerances 2005 Unless Otherwise Stated Motor Flate 132 L1 L Mounting Brock et 133 8 18 UE ee Shoulder f LILL I ME 5 sas 134 E i E LUC 2 shoulder Mount a a 135 136 ia NEUES Shoulder Shatt mar a 137 UNT ea Size 43 p 25 128 E 250 125 825 4500 2000 025 4 comers 2500 All Tolerances 0005 The Ohio State University Gateway Coalition Unless otherwise Stated Side Flote Buemome __________ zoo 138 2000 000 Im k Bugs 12 15 07 05 m 123 nai drug nl siZ h Bole id She dum R REK Bales 1E 0n ciae dung 3 mi Zh bni a tS Bras dum Boles EZ 25 m FE rius Dales ZA 27 JIE dung Pu nma Stotiomry Plate re 13
4. 04 08 2001 list p l6F84 radix hex _ config XT Tz on WOT off HS oscillator no code protection Set up custom file registers Input equ 0 ontime equ offtime equ 110 w_tmp equ Stat_tmp equ 0x10 joint equ 0x11 include lt 16F84reg inc gt org 0x000 goto init Interrupt Handler Ong 0x004 movwf w tmp movf STATUS w Save current state movwf stat_tmp ee btfsc INICON RBIF Check for joint select goto btn_intr This simply counts down the PWM on time 1f ontime gt 0 restore state goto off_intrl on_intr movf ontime w iorlw d O ontime ontime 1 Dor se SALUS goto off_intrl return decf ontime f else call restore retfie el These count down PWM off time but at certain points other tasks are performed offtime offtime 1 restore state goto off_intr2 moviw d 20 if offtime gt 20 subwf offtime w btfss STATUS C 2 goto off_intr2 decf offtime f d call restore NS retfie During this segment of off time the ADC is prepared for outputting the serial signal OTT moviw d 16 subwf offtime w 62155 STATUS C goto off intr3 movlw b 00000010 xorwf PORTB f decf offtime f call restore retfie d return d m else ssp dp orbe T6 Cycle AVE CLK 122 Clioe
5. Entr On IMROU Intr Ons RB Temporarily disable interrupts Start with first input at the center Activate Joint 0 see data table for name Re enable interrupts Clear Reset ADC if input lt 100 goto rev if input gt 155 goto fwd else ontime 0 PWM signal 1 on Loop until ontime runs out Turn off PWM signal ADC CLK Low Tell ADC to start conversion process se eS 50 i l e 15 OT Loop until offtime runs out 113 goto off_loop goto main Do this to make motor run in forward direction fwd movlw d 145 E s subwf input w ontime input 145 movwf ontime NOTE 11 lt ontime lt 111 bsf PORTB 5 Turn on forward PWM signal goto on_loop Do this to make motor run it reverse direction rev movf input w ES sublw d 110 ontime 110 input movwf ontime NOTE 11 lt ontime lt 111 bsf PORTB 4 Turn on reverse PWM signal goto on_loop Read in one bit of data from ADC bit_read bs f PORTB 1 ADC CLK High ADC clock signal high DOT SATUS C btfsc 2 Carry Bit bit read from ADC bsf STATUS al Rotate read bit into input bert PORTB 1 ADC CLK Low return Joint selection data table joint tbl addwf Pe Look retiw b 00000001 Activate Shoulder Joint 0 retiw b 00000010 Activate Elbow Joint 1 retiw b 00000100 Activate Wrist
6. 88 APPENDIX D uu 91 5 A 91 APPENDIX Evier ar ER EE LE ESER ELERESE E EEE EE E 92 PET TALTERNA TUVE CGONTROL 92 BZINTRODUC TION AND GOAL 92 92 CONTROL DESIGN GENER AL DESCRIP TION Lone eet nonton 93 BS CONTROLS DESIGN DETAILED DESCRIPTION u u l Q k u 95 E GHARDWARE CONS TION 105 J C Bc Su 108 J LIP SDI eles m S use D Lu 108 N rer 110 APPENDIX E 115 MOTOR TESTING PROCE DURE R 115 APPENDIX 117 y A WING cR 117 REFERENCES SS u ADDENDUM DESIGN MODIFICATIONS vi TABLES 1 1 Previous Attempts at Marketing Robotic Arms 2 2000 2001 Gateway Coalition Team Members 8 1 Power Consumption of Typical Appliances 10 1 Control Modes Shoulder Lower Arm and Forearm Raw Materials Costs A 2 Shoulder Lower Arm and Forearm Component Costs A 3 Shoulder Lower Arm and Forearm Machining Costs 4 Gripper Bill of Materials A 5 Controls Bill of Materials A 6 Summarized C
7. WRIGHT STATE ST EE ue __ o m eee 121 en Ae ELBOW MOTOR STOP PLATE WRIGHT STATE UNIVERSITY REVISED 4 1001 ERIC A 122 a Ail 0 0 o ud SHAFT FEMALE WRIGHT STATE PERS NEC 123 250 128 1158 ALTERNATIVE MALE ELBOW SHAFT REVISED 4 1 G00 WEIGHT 124 xdi 7850 525 003 400 M 100 Seo mW R 1257008 002 200 888 BOO 500 All Tolerances State University Gateway Coalition 005 in Unless stated Round Top Finger Material Aluminum 2 24 00 gt 125 FIK B MTIANI WAHT STATE LEPERSTY E _ L l 126 14 20 LINE 2 4 40 Luc 4 Places cis g Placez ANNE 500 R 1500 250 F REARM REVISED 313 200 127 29297 ee E rom 128 Link 100 Link 2X 125 0005 2X 125 0005 250 79901 0001 250 750 005 1000 2005 003 2X 1257004 003 2X m i25 State Ohio State University Gateway 00 Coalition 75 Link amp 100 Link Material Aluminum 129 i IE E M
8. register because the main program may have been in the middle of using them for something restore movwf mov f DC Der return mov f STATUS tmp w INTCON TOIF INTCON RBIF selected Loop until button is released Select the next joint Restore state Return the state of the SIATUS register when the interrupt occured Stat 112 Initializaion Chip configuration init bsf mov lw output movwf mov lw movwf mov lw movwf mov lw movwf bcf bcf Cer mov lw movwf 051 051 STATUS RPO DE 0004100 3 TRISB 0x00 TRISA 0 00001000 OPTION REG 0 10100000 STATUS RPO INTCON 7 PORTA PORTB input 0 INTCON 7 Main Program Body 051 mov f sublw btfsc goto mov f sublw btfss goto Cher main PORTB O input w d 100 SIATUS C rev input w 1155 fwd ontime set initial values Teach Port B bits 2 3 input all others Teach Port A all bits output 1 1 Pre scaler for Global e Do this stuff while the on_loop iorlw ptfss goto bcf bcf bcf bcf mov lw movwf Do this stuff while the PWM signal off loop iorlw btfss movf ontime w 4 0 STATUS 7 4 5 PORTB 1 0 050 offtime movf offtime w STATUS 2
9. 2 GEAR RATIO CALCULATIONS FOR FIRESTORM 18 DRILL MOTOR Necessary Dimensions 1 Stage se weas Teeth mm Sun Metal 18 14 Planet 3 gud Stage Teeth mm Sun Metal 21 16 Planet 3 3H stage maera 7 Teeth mm Sun Metal 12 10 Planet 3 Important Equations 2N where ring planet 74 N N of teeth on the planet gear planet of teeth on the ring gear ring of teeth on the sun gear 5 2 OO arm 2 W sun N planet where Rotational speed of planet gear Rotational speed of arm gear Rotational speed of sun gear 3 Ding J arm _ on W sun N ring 1 Stage Calculations From motor literature operational speed of motor 10702rpm P sun since amp ring 0 Lo NN an _ 18teeth 10702rpm 07 2918 73rpm 48 18 teeth 2 Stage Calculations am becomes sun the next set of calculations N 5 W arm 2lteeth 2918 73rpm _ 028 7rpm 45 2 teeth iis 3 Stage Calculations T arm becomes the next set of calculations sun3 sun N N ring 2 75 124 928 7 rpm 07 185 747 48 12 teeth Gear Ratio Input d Gear Ratio E spera Output speed Gear Ratio 185 74rpm Power loss in Gear Train of Firestorm 18V
10. p i 15 7 j A 1 N Y Ee KT Figure 12 Control Box top Amplifier Box middle H bridge Stack bottom For this prototype design it was decided to mount the modules in sheet metal boxes Sheet metal boxes inexpensive easy to make and easy to customize In particular sheet aluminum was used along with aluminum rivets To hold all of the boards in place inside their respective boxes standoffs were used These standoffs are hex side 1 4 male female threaded standoffs For the control module the standoffs were steel 12 long and had 6 32 threading There were fourteen all for on the top and bottom of each corner and another three on the top and bottom around the joystick to prevent the board from flexing and breaking when the joystick button is pushed The amplifier module used 4 40 threaded long aluminum standoffs between the boards 24 standoff in all along with 8 4 40 threaded 1 5 aluminum standoffs on the top and bottom of the board stack For both the control box and the amplifier box The boards are attached to their respective boxes with nuts on the bottom and machine screws on the top which also holds the lid 4 40 threading for the amplifier box 6 32 threading for the control box 105 Inside the amplifier box the main power connector 1s 2 pin Molex plug female that enters through the back of the
11. stationary plate causes the two places where the rear mounting bracket 15 bolted to the plate to be at different locations This meant that one side had to be shortened to ensure that the fit was correct A new rubber insert was also selected for the mounting brackets The new insert is thinner and can be attached by adhesive This change makes the machining of the brackets less expensive since the tapped holes are eliminated from the cutout where the rubber inserts are placed The thinner inserts changed the cutout dimensions slightly but not extensively Figure 4 9 New Rear Mounting Bracket with Twist Motor The front mounting bracket was not affected by the new motor or change in the stationary plate The only changes to the initial design of the front mounting bracket was the elimination of the tapped holes for the rubber insert and the dimensional changes for the cutout where the insert is placed Figure 4 10 New Base Assembly The new twist motion utilizes a belt and pulley system instead of gears see Figure 4 10 This change was made because the new motor moved the center of the motor shaft further from the center of the bearing hole A benefit of this change 15 the pulley also acts as a safety device The pulley will slip and prevent the motor from burning up if the arm is constrained from twisting for any reason This limitation on the twist torque may also prevent people from getting hurt by inadver
12. while jl 0 pwm2 value 10 OOPic delay 20 pwm2 value 0 m2f value 0 m2e value 0 s2 value do l wb value 1 E Oy wb value 0 s3 value do wl value 1 while jl 0 wl value 0 f Joystick right Sub void jrgo_code void nogo 0 sl value s2 value nogo 1 sl value s3 value nogo 1 s2 value s3 value nogo 1 if 0 86 sl value m2b value 1 m2e value 1 pwm2 value 35 OOPic delay 10 do pwm2 value 80 while jr 0 pwm2 value 10 OOPic delay 20 pwm2 value 0 m2b value 0 m2e value 0 s2 value do Wt lt value 1 while jr wf value 0 if s3 value do wr value 1 while jr wr value 0 87 C 3 SLAVE OOPIC CONTROL CODE SLAVE CODE FOR CONTROL OF THE ARM DATE LAST MO L4 AUTHORS opwm pwm3 opwm pwm4 odiol wf odiol wb odiol wl odiol wr 5 08 01 DIFIED new opwm new opwm new odiol new odiol new odiol new odiol DU LOU E MICHAEL STEVENS AND AARON WEAVER odiol m4f new odiol odiol m4e new odiol odiol m4b new odiol odiol m5f new odiol odiol m5e new odiol odiol m5b new odiol ogate wfgate new ogate ogate wbgate new ogate
13. 2 m5b direction m5e ioline 2 m5e direction cvoutput 7 o l N cvoutput Output link wfgate wfgate wfgate wbgate wbgate wbgate wlgate wloate wloate wrgate operate 1 output link operate 1 operate 1 inputl link wr operate 1 wrgate wrgate EIU TAA EAR p TTT TT IT FS Wrist bend forward Sub void wfgo_code void m4f value 1 m4e value 1 mb5bf value 1 mb5be value 1 do pwm3 value 102 pwm4 value 102 while wf 1 pwm3 value 10 pwm4 value 10 OOPic delay 20 pwm3 value 0 pwm4 value 0 m4f value m4e value m5f value m5e value O O O lt Wrist bend backward Sub void _ void m4b value 1 m4e value 1 mob value 1 m5e value 1 do pwm3 value 102 pwm4 value 102 whale wb se pwm3 value 10 pwm4 value 10 OOPic delay 20 pwm3 value 0 pwm4 value 0 m4b value m4e value l O O lt lt mbb value mbe value 0 0 Wrist twist left Sub void wlgo_code void m4f value m4e value m5b value m5e value pwm3 value pwm4 value e errre lt while wl pwm3 value pwm4 value OOPic delay pwm
14. Joint 2 retlw b 00001000 Activate Gripper Joint 3 end 114 APPENDIX MOTOR TESTING PROCEDURE The motor testing procedure requires the following equipment Purpose Procedure Magtrol Dynamometer Model HD 710 8 Magtrol Model 5210 Power Supply Magtrol Model Number BL 001 Cooling Fan 5V DC Power Supply 22V DC Power Supply Oscilloscope Voltmeters 2 Ammeter Establish a lab procedure to obtain necessary data from certain motors in order to plot a torque versus speed curve l 2 8 9 Obtain motor Connect necessary output wires to motor These wires may be used to collect data from the motor such as voltage and current draw Place motor on test stand For this step a test stand may need to be fabricated specifically for the motor being tested Connect motor to brake via a shaft Brake should be attached to Magtrol Dynamometer Model HD 710 8 Supply 5V DC to the encoder See Magtrol Dynamometers manual for details supply 22V DC to the torque output See Magtrol Dynamometers user s manual for details Turn on cooling fan Magtrol Model Number BL 001 Caution must be taken while performing the test Brake may overheat and cause permanent damage See Magtrol Dynamometers user s manual for horsepower versus time curves for maximum allowed testing times Turn on brake current supply Magtrol Model 5210 Power Supply This 1 used to vary the torque applied by the b
15. OOPic delay 10 do pwml value 90 while jf 0 pwml value 10 OOPic delay 20 pwml value 0 mof value 0 lt moe value Joystick back Sub void jbgo_code void nogo 0 sl value s2 value nogo 1 sl value s3 value nogo 1 s2 value s3 value nogo 1 84 1 0 if sl value mlb value 1 mle value 1 pwm2 value 153 wf value 1 while jb 0 pwm2 value 10 OOPic delay 20 pwm2 value 0 mlb value 0 mle value 0 wf value 0 if s2 value m3b value 1 m3e value 1 pwml value 30 OOPic delay 10 do pwml value 90 wf value 1 while jb 0 pwml value 10 OOPic delay 20 pwml value 0 m3b value 0 m3e value 0 wf value 0 s3 value m6b value 1 m6e value 1 pwml value 30 OOPic delay 10 do pwml value 90 he 45 0 pwml value 10 OOPic delay 20 pwml value 0 m6b value 0 m6e value 0 I Joystick Sub void jlgo_code void nogo 0 sl value amp s2 value nogo 1 85 sl value s3 value nogo 1 s2 value amp s3 value nogo 1 1 0 1 sl value m2f value 1 m2e value 1 pwm2 value 30 OOPic delay 5 do pwm2 value 90
16. The only way to limit activity to one H bridge at a time 1s to provide a way to disable all of the others This is done by the use of the mode enable lines from the control module along with the enable transistor found on each H bridge By sending a logic 0 0 volts to the enable transistor the transistor is turned off making the H bridge an open circuit by disconnecting the ground When a logic 1 1s applied to the enable transistor 5 volts the H bridge becomes grounded and 1s therefore active This is how the control system limits the motion to two motors at a time 24 Q1 IRF5305 IRF5305 R1 FORWARD 212 1 1k Q2N2222 M2 NDP6060 NDP6060 M4 ENABLE pa Q2 REVERSE A Iu t 1k 0232222 Figure E 10 H bridge Schematic There are two motors that get both an x axis signal pair and a y axis signal pair These are the wrist motors Because of the differential system used the wrist both motors have to run together to move the wrist in either a twist or bend manner For twist x axis both motors 103 run in the same direction For bend y axis the motors run in opposite directions do this the router board sends x and y axis signal pairs to both motors However one of the motors receives a reversed y axis signal pair This corresponds to output set number 4 on the router board schematic and 1 also the same set on the actual board when looking at it from above Now w
17. bracket are also deeper than the height of the tubing so a bolt can be placed through the bottom of the bracket This bolt can be tightened to squeeze the mounting brackets around the frame of the chair and secure the entire arm to the wheelchair Two of these mounting brackets are used to mount the arm to the frame Both are bolted to the stationary plate discussed the next section of the arm and clamped to the frame of the wheelchair see Figure 4 2 There is a small difference in the two mounting brackets Since the frame of the wheelchair 1 not parallel to the ground the front mounting bracket was designed to 18 be slightly taller than the rear bracket This difference ensures that the stationary plate of the arm is parallel to the ground This type of clamp design offers several improvements over last year s design The first improvement is that one person can mount the entire arm The tight fit of the mounting brackets around the frame allows someone to slide the arm onto the wheelchair Figure 4 2 Front and Rear Mounting Brackets and the arm will stay in position while the person inserts and tightens the squeeze bolts in each of the mounting brackets Another improvement 15 the addition of the rubber inserts between the brackets and the frame The rubber not only helps clamp the arm to the frame but also protects the frame from scratching during installation and removal The rubber also eliminates the m
18. elecro motive force that works against the input voltage The faster the motor spins the more back EMF it generates until the motor reaches a speed where the back EMF and the input voltage are approximately 98 equal At this point the motor has reached a stable speed that it will maintain regardless of the load placed on it This 15 all assuming that there is a enough current available for the motor to draw from and b the motor is not approaching its stall current where it will eventually slow down and stop Now there are two ways to control the voltage level being input to the motor The first 15 to place a variable resistor in series with the motor at the motor s input By increasing the resistance the voltage drop across the resistor increases thereby decreasing the input voltage to the motor This is a rather inefficient and unreliable means of motor speed control Plus it can drastically limit the current being provided to the motor The other way to control the motor s input voltage is through a process called pulse width modulation or PWM PWM is nice because the motor 15 always supplied with the maximum voltage available By inputting the voltage in pulses of varied length the effective voltage changes and creates the same response from the motor as applying a constant voltage of the same level as the effective voltage This is much more efficient than using a variable resistor to change the input voltage and the current delive
19. gearbox and clutch pack The gearbox consists of three sets of planetary gears The two sets closest to the motor that are under the lowest torque have plastic idler gears The speed rating on the drill is 0 400 rpm for low range and 0 1400 rpm on high For our application low range would be tested The low range has a gear ratio of 21 4 1 The clutch packs of the drill enable a maximum torque limit As a safety factor the packs could be set to reduce the maximum force provided to the arm For all testing the packs were set to the drill mode which delivers the maximum torque available The motor was rewired at its terminals to provide current and voltage readings for testing A Magtrol HD 710 model dynamometer was used for testing This dynamometer uses a strain gage to calculate torque input to a brake unit This strain gage has a voltage output that 1s converted back into a torque output An encoder produces a pulse frequency and an oscilloscope was used to convert to rpm A stand had to be designed and built so the drill could be effectively coupled to the brake unit The stand was designed to be adjustable the horizontal and vertical axes to accommodate different drills for future testing 52 m gt a gt m nd l BC EN E C TNI 1 1 M 7 igure 8 1 irestorm 18V drill motor and accessories See Figure 8 2 Since the drill had a 3 8 inch chuck
20. the same individual would marginally increase their manipulation ability The profile was redefined as a person having occasional and then frequent use of reaching handling and fingering skills An individual with occasional use of these skills was then shown to be capable of performing approximately 300 jobs The individual with frequent use of these skills was shown to be capable of performing over 1100 jobs The results of this study alone demonstrate the impact a rehabilitation robot would have on the lives of potential beneficiaries and validates the attempt to design such a device There have been a number of attempts thus far to create a rehabilitation robot that is both affordable and effective A list of such products is shown in Table 1 below Only three of the commercial endeavors shown in the table Rehab Robotics Exact Dynamics and Rehabilitation Technologies are actively marketing and supporting their product The Raptor by Rehabilitation Technologies began sales in 2000 and has not had enough marketing time to measure its sales performance As can be seen from the table many of these products have not been successful and no one product has had overwhelming success There are many factors which contribute to the failure of these previous attempts including poor user interface isolation from clinical reality the cost benefit 1 not justified lack of portability poor organization and lack of capital funds of these fact
21. uuu 18 4 1 MOUNTING u 18 BASEPLATE DESIGN E E A 19 VENRTIqpu x 29 ERJON I u xu 29 s uu 31 91721719218 gt 33 FOREARM u _______________ _ __ _ ___________ _ 38 OILFOREARM TUBING uuu cc teat crete 38 62 WRIST DIFFERENTIA L Lu uuu uuu u 40 u u usu s 42 TAVRESBARCH AND CONCEP US 42 72 DESIGN AND ACTUATION u u u ENE 44 MOTOR TESTING uu u gt 49 8 1 MOTOR RESEARCH AND 8 2 0 22 49 MOTOR TESTING eg Mea 51 CONTROLES HARDWARE un sS 58 DOPNHICROCONCDROELDLER 58 DO 5 5 9 9 8 94 E esto 59 9 3 ANGLE MEASUREMENT FOR CLOSED LOOP FEEDBACK
22. 1 3 4 show some different orientations of the completed 2000 2001 robotic arm Figure 3 1 Arm mounted on Wheelchair Figure 3 2 Home Position Frame Figure 3 3 Reaching below the Floor Figure 3 4 Same Position but Different Orientation 13 3 4 TERMINOLOGY Early in the project 1 was realized that describing the different motions parts and joints of the arm was difficult Therefore members of both teams decided to create a standard terminology in order to make communication easier and less confusing The basis of the Lower Arm Forearm Gripper Figure 3 5 Exploded View of Arm terminology was the human arm The main parts of the arm were chosen as the base lower arm forearm and the gripper see Figure 3 5 The joints were defined as they would be on the human arm shoulder elbow and wrist The shoulder is the joint between the base and the lower arm The elbow is the joint between the lower arm and forearm Finally the wrist is between the forearm and the gripper Since these three joints account for 5 of the 6 degrees of freedom the motion of each of the joints were added to the terminology The shoulder joint was broken into the shoulder twist and the shoulder bend motions see Figure 3 6 The shoulder twist 15 the motion of the arm rotating around a vertical axis as viewed from the side of the chair The shoulder bend is therefore the motion of the lower arm rotating around a horizontal axis Since
23. 12 x 6 x 3 4 Green Oil Impregnated Cast Nylon Sheet 23 24 12 x 12 x 0 125 6061 T6 Aluminum Sheet 8 x 4 x 0 25 6061 T6511 Aluminum Plate 12 x 6 x 0 5 6061 T6511 Aluminum Plate 8 x 4 x 1 6061 T6511 Aluminum Plate 89155 72 McMaster Carr 19 38 1 19 38 x 3 4 OD x 6 6061 T6511 Aluminum Rod 2 x 1 5 x 1 6061 76511 Aluminum Rectangular Bar 8975K253 McMaster Carr 22 03 1 22 03 Subtotal 81777 _ Part number is for ordering twice the amount of material actually used in constructing the robotic arm Table A 2 Shoulder Lower Arm and Forearm Component Costs 0 625 OD 0 25 Bore Oil Lubricated Ball Bearing S9912Y E2562FS2 SDP SI 10 33 20 66 Wrist Motor GM8724S027 150 22 300 44 Wrist Differential SDP SI S9570A TS3 SDP SI 304 93 304 93 ii Bearing Distributors Inc 2 OD 0 75 Bore Gear GSS486Y G Boston Gear Distributor 23 30 46 60 Bearing Distributors Inc Wrist Pinion GSS486Y P Boston Gear Distributor 19 27 38 54 1 125 OD 0 5 Bore Flanged Ball Bearing E7 S3F PIC Design 25 46 101 84 3 6 V 180 RPM Cordless Screwdriver 9074 Black amp Decker ____ Catalog 20 1 2 20 Pitch 20 tooth 1 Pitch Diameter 0 5 Bore Steel Item 46133 Boston Gear Catalog Catalog YA60A m 20 Pitch 60 tooth 3 Pitch Diameter 0 5 Bore Steel Item 10550 Bearing Distributors Inc 25 00 pacis Boston Gear Catalog Boston Gear Distributor Screws Nuts Bolts Washers 242
24. 125 250 195 2505 MATERIAL STEEL Lo qegver o ELBOW MOTOR SHAFT WRIGHT STATE UNIVERSITY REVISED 7200 ERIC duces 162 250 196 3 750 MATERIAL STEEL ELBOW SHAFT WRIGHT STATE UNIVERSITY REVISED 2001 ERIC CHES 163 1574 _ due ______ L ELBOW THRUST BEARING BLOCK WRIGHT STATE UNIVERSITY REVISED 6 7 2001 a EN Tom 164 4x 250 2 112 132 LI cee Ce LI I eee i i errare a 1575 5 000 FOREARM TUBE REVISED 2 001 750 165
25. 14 lt Shoulder Twist 4 Shoulder Bend 4 Figure 3 6 Shoulder Twist and Bend the elbow only has one degree of freedom its motion was simply defined as the elbow bend See Figure 3 7 The wrist like the shoulder has two degrees of freedom See Figure 3 8 The wrist bend motion refers to the rotation of the gripper around an axis that 1 located at the end of the forearm and runs along the width of the forearm tube The wrist twist motion 1s the rotation of the gripper about an axis that runs along the length of the forearm The sixth degree of freedom is the clamping action of the gripper and 15 referred to as the gripping motion 15 Figure 3 7 Elbow Bend Wrist Twist Wrist Bend Figure 3 8 Wrist Twist and Bend 16 3 5 PERFORMANCE HIGHLIGHTS The final design of the 2000 2001 robotic arm has the following performance and functional characteristics o Freedom of Motion Shoulder twist 360 o Shoulder bend 210 max Elbow bend 255 o Wrist twist 360 o Wrist bend 140 o Shoulder to Elbow 15 5 in o Elbow to Wrist 14 56 in o Wrist to Finger Tips open 11 05 in Full Extension 41 11 in o Width Extension beyond the wheelchair width 3 25 in Wheelchair width with the arm 27 75 in Vertical Reach o Above the ground 61 50 in max Below the ground plane wheelchair is sitting on 10 25 In max o Horizontal Reach o Extension from front wheel of the wheelc
26. 2 r6 cos ie Equations 8 and 9 were rearranged as follow 79 0 sin O rO cos 0 10 cos 0 0 IO 10 cos 0 7 0 cos 0 r 5 Finally equations 10 and 11 were set equal to each other and solved for the derivative of angle 01 The equation 1s as follows Xcos 0 0 ysin O O 12 9 r cos 0 0 80 C 2 MASTER OOPIC CONTROL CODE MASTER CODE FOR CONTROL OF THE ARM FILE NAME CREATED LAST MODIFIED AUTHORS MASTER 1 5708701 y MAP AO RU MICHAEL STEVENS AND AARON WEAVER THIS CODE WILL OPERATE IHE ARM USING FOUR DIRECTIONS ON THE YSTICK ANOTHER OOPIC WILL BE USED TO OPERATE THE WRIST WILL BE SENT FROM THIS MASTER CODE TO THE JO SLAVE LINES MODES OF MOTORS ARE NUMBERED AS FOLLOWS 4 NOTES I F 7 1 Ly 2 EI 3 4 iy 6 opwm pwml opwm pwm2 odiol mlf odiol mle odiol mlb odiol m2f odiol m2e odiol m2b odiol m3f odiol m3e odiol m3b odiol m 6f odiol odiol odiol jf odiol jb odiol jl odiol jr ogate jfgat ogate jbgat ogate jlgat ogate jrgat oevent jfgo oevent jbgo oevent jlgo oevent jrgo es odiol s2 odiol s3 SHOULDER BEND SHOULDER TWIST COMMUNICATION odio
27. 256 1338 Fax 330 723 2012 614 861 2776 Fax 330 995 9600 dayton bdi usa com info pittmannet com rockford ohio net ele sales memaster com Web Site www bdi usa com www pittmannetcom _www sdp si com www pic design com www mcmastercom Company Name Diverse Electronic Services Savage Innovations NF NS Rep Address 1202 Gemini Street Nanticoke PA 18634 3306 Sales Rep Phone Number Phone 570 735 5053 Fax 603 691 7688 EMail 0 carlk3jml bigfoot com SavagelInnovations OOPic com Web Site members tripod com divelec 72 APPENDIX B 1 TORQUE CALCULATIONS This section consists of the torque calculations for motor and gearing requirements The initial calculations were determined using revised weights and measurements from the 1999 2000 teams design These calculations were then updated as the design progressed and changed The calculations take into consideration the length center of gravity and weight for each part of the arm These calculations are somewhat approximated but accurate enough for motor and gear selection Weights Ibs Lengths in A 15 0 a 5 B 4 00 b 25 5 C 3 00 C 33 0 X 15 50 y 14 56 Weights Ibs Lengths in B 4 00 b x 10 0 3 00 17 5 Shoulder Motor 494 0 Ib in Elbow Motor 214 0 Application Factor 1 5 Load 3 3 Ibs Figure B 1 Schematic for torque calculations
28. 6 Degrees of freedom e Maximum weight of entire assembly less than 30 165 e Fully functional control system with working user interface e A maximum linkage movement speed of 0 5 m s e Lifta 1 5 kg 3 3 Ib mass e Maximum cost of 4000 including the controls 3 2 DESIGN CONSTRAINTS Due to the physical layout of the wheelchair many constraints had to be placed on the design to ensure that no clearance or interaction problems would occur The first constraint placed on the design was that no part of the assembly could extend more than 6 inches beyond the furthest edge of the wheelchair see Figure 3 1 This constraint was placed to make sure that with the arm mounted on the wheelchair the wheelchair would still be able to fit through a standard doorway without too much difficulty Another constraint was that in the home position see Figure 3 2 the arm was not taller than and does not interfere with the armrest of the wheelchair or with the rear tire The placement of the robotic arm was also limited by cross member of the frame of the wheelchair This cross member had to be taken into consideration when selecting the placement and mounting widths of the mounting brackets The final constraint of the design was that no part of the assembly would interfere with the swivel and rotation of the front tire This constrained the mounting height of the arm and extension of any parts below the base of the arm 3 3 CONFIGURATION Figures 3
29. Furthermore angle 0 represents the angle between link 1 2 and 05 represents the angle between links 2 and 3 The kinematics analysis is detailed in the following paragraphs and equations The following relationships were determined by analyzing the geometry of the linkage 78 1 6 180 2 0 180 6 0 The position of the end of link 3 1 e the tip of the gripper was evaluated analytically using open loop kinematic analysis of the linkage In the following equation x 1s the position of the gripper in the horizontal direction 3 x r cos 0 r r Equations 1 and 2 were substituted into equation 3 to obtain position equations in terms of the angles between the links instead of from the horizontal In the following equations y represents the position of the end of link 3 in the vertical direction 4 r 0 r 0 5 sin r sin O r sin O 0 assumption was made that angle must remain constant Equation 2 was then differentiated yielding the following equation 6 0 0 0 Solving 6 for the derivative of 05 yielded the following relationship 7 6 6 Equations 4 and 5 were then differentiated incorporating the simplifications from 6 and 7 This yielded the following equations 8 7 0 sin O 7 sin O 9 y
30. a shaft was designed to transmit torque to the brake Three milled flat lands were placed on the chuck end of the shaft to eliminate the possibility of slippage of the chuck A keyway was placed on the brake end to drive the dynamometer See the shaft in Figure 8 3 The shaft completed the mechanical changes to the stand See the final setup in Figure 8 4 53 Figure 8 3 Shaft Figure 8 4 Test Setup Magtrol HD 710 54 The Magtrol required both 22 volt and 5 volt power supplies to operate the data outputs An automotive 12 volt battery was wired In series with an adjustable voltage supply to deliver the 22 volts This was done because the voltage supply had a maximum voltage of 17 volts A separate 5 volt supply was used for the 5 V circuit The connector shown in Figure 8 5 was wired to provide the input power to the data acquisition of the Magtrol The testing procedure which was followed to carry out the testing can be found in Appendix E Figure 8 5 Power Input Wiring to Magtrol HD 710 Testing of the setup proved very brief Unfortunately there was a short between the 22 V and the 5 V circuits Two loud pops were heard and electrical smoke emanated from the stand We think the short occurred in the orange end of the connector The Magtrol case was opened 15 voltage regulator was destroyed along with a capacitor See Figure 8 6 for a photo of the board components These parts were inexpensive 1 45 for
31. and also improves upon the first design in a number of ways The use of the pulleys not only acts as a safety device but they are also cheaper than the gear design used initially These pulleys are easily machined from stock aluminum and do not have to be specially ordered The new design also has a more desirable gearing ratio for the twist motion The pulleys produce a speed ratio of 9 1 This ratio produces 25 more torque to overcome the friction in the bearings and also slows the rotational speed considerably The initial bearing design was 0 50 inch thick bearing The outer diameter was 3 50 inches and the inner diameter was 3 00 inches This 15 a very simple and effective design The choice of material was that of oil impregnated brass because of its wear and frictional characteristics This was an improvement upon the previous year s design because it was cheaper and reduced the material required The redesign of the base also brought about a redesign of the twist bearing A three piece bearing assembly was designed to take the place of the single bearing see Figure 4 13 and 4 14 The main piece of the assembly is the center bearing This piece is bolted to the twist pulley from the top and 1 the only moving component in the bearing assembly The lower part of the Figure 4 13 Cutaway of New Twist Bearing Assembly bearing has a diameter of 3 375 inches This is slightly smaller than the hole in the stationary plate where it 1s
32. and spring presentations to unify and prepare ideas Finally several individual meetings were held to swap components and information 2 3 PROJECT TASK ASSIGNMENTS Following the arrangements made by previous teams the tasks for developing the robotic arm were divided among the three participating schools Ohio State University was in charge of developing the gripper and electronic controls Wright State University was responsible for designing the base lower arm and forearm Drawings for the completed arm components were sent to Sinclair Community College for manufacturing and assembly 11 2000 2001 ARM OVERVIEW 3 1 OBJECTIVES At the initial meeting October 2000 representatives from both Ohio State and Wright State Universities met to discuss the project and how the design of the 2000 2001 robotic arm would be achieved After reviewing the previous years work on the project the group decided on some design objectives and constraints that would make sure that the 2000 2001 robotic arm would be an improvement upon the previous designs The 2000 2001 design needed to be cheaper and lighter than that of Chris Fearon and the 1999 2000 team Other improvements that were to be implemented were to increase the lift capacity of the arm and to have a fully functional control system By the end of the meeting the 2000 2001 team members from both universities decided on the following parameters for the new robotic manipulator e
33. equ 4 RB5 equ equ 6 RB equ 55 LNTCON REGISTER reo m D Q gt BW PF OPTION REGISTER PSO equ 0 51 equ 1 PSZ equ 2 equ 3 05 equ 4 TOCS equ 5 INTEDG equ 6 RBPU equ REGISTER RD equ 0 WR equ WREN equ 2 WRERR equ 3 EEIF equ 4 destination bits W equ 0 T equ SPECIAL PROGRAM ADDRESSES 109 PROGRAM_START equ 0 INTERRUPT_START 4 2 CONIGURATION FUSE FOSC Es FOSCU LPOSC 0 XTOSC equ 1 HSOSC equ 2 RCOSC equ 2 gt 02 Watchdog timer enable WDTENABLED equ 4 D3 Power up timer disable PWRUPTMRDISABLED equ 8 D4 Code protection disable CODEPROTDISABLE equ 16 endi f of 16F84REG INC E 7 2 MAIN PROGRAM 5 le controll asm 2 This program has two functions The first is to provide a variable speed motor control via pulse width modulation based on an 8 bit digital conversion of analog voltage from a potentiometer Note that the conversion is done externally by an analog to digital converter National Semiconductor s ADCO831CCN other function of this 15 to provide the ability to select the active joint on the arm The joint is selected by pressing a button connected to bit 7 of Port B The joint selection is cycled with each press until the desired joint activated Note that no motor on the arm will function while the button 15 depressed
34. motor the same motor as originally chosen for the shoulder twist and bend motions Ordering the same motor for all joints would decrease the difficulty of obtaining all necessary motors but this proved not to be the best design for the robotic arm Combined with a controls encoder the motor extended approximately 8 inches and weighed over 1 76 pounds Initially the design team sought to place the motor at the baseplate and transfer the power up to the elbow joint through transmission pulleys minimizing torque requirements at the shoulder bend motor However since the motor encoder assembly extended nearly the entire length of the lower arm tubing the drill motor was mounted within the tubing directly aligned to the elbow joint as seen in Figure 5 8 Bevel gears reduced the speed of the elbow rotation to below 12 rpm Figure 5 8 Preliminary Elbow Motor Without Encoder Flanged bearings were press fit into the lower arm tube to support the elbow shaft Outside of the bearings snap rings were fastened to the shaft to secure it from sliding and to lock the bearings place Upon further research the motor from a Black amp Decker Model 9074 3 6V cordless screwdriver was selected to replace the drill motor The screwdriver motor was 33 smaller and lighter than the drill motor the complete motor assembly weighed approximately 1 pound reducing the motor weight by approximately 40 However the 3 7 inch long motor still produc
35. ogate wlgate new ogate ogate wrgate new ogate oevent wfgo oevent whgo oevent wlgo oevent wrgo new oevent new oevent new oevent new oevent sub void main void wf io wf direction wb io wb direction wl io wl direction wr 10 wr direction line line line line 5 4 s 64 CVIDDUL CV ID UU evinput pwm3 ioline 17 pwm3 prescale 3 pwm3 period 255 pwm3 operate 1 pwm3 value 0 pwm4 ioline 18 pwm4 prescale 3 pwm4 period 255 pwm4 operate 1 pwm4 value 0 m4f ioline m4f direction m4b i m4e i oline oline 0 m4b direction 9 3 2 m4e direction m5f i oline 28 m5f direction cvoutput cvoutput cvoutput cvoutput PWM 3 PWM 4 Command from MASTER to Command from MASTER to Command from MASTER to Command from MASTER to for m4 for m5 bend wrist forward bend wrist backward twist wrist left twist wrist right Motor Motor Motor Motor Motor Motor 4 4 4 5 5 5 forward line enable line backward line forward line enable line backward line Gate Gate Gate Gate Event Event Event Event to link wfgo to to link wbgo to to link wlgo to to link wrgo to for wrist bend for wrist bend for wrist bend for wrist bend wt wh wl wr forward backward left righe 88 m5b ioline
36. placed so that it can rotate freely without contacting the stationary plate The upper section of the center bearing piece has a 2 75 1nch diameter This part extends up through the center of the upper bearing piece and the bearing plate This is the end that is bolted to the twist pulley 26 Figure 4 14 Cutaway and Exploded View of Bearing Assembly The other two parts of the bearing assembly are very similar and only have one difference Both are 0 125 inch thick and have a diameter of 3 50 inches The difference is that the upper bearing piece has a 2 75 inch diameter hole cut through it Both pieces are pressed into the stationary plate and do not rotate The lower piece goes under the center piece and the upper piece fits around the upper section of the center bearing piece Figure 4 15 Complete Redesign Base Assembly All three pieces of the twist bearing assembly are made of oil impregnated nylon Nylon was chosen over brass due to its availability and lower cost The bearing assembly is designed such that all points of friction are nylon against nylon The nylon to nylon decision was made to try and minimize wear and noise These contact areas are the entire bottom of the center piece against the lower piece the ledge of the center piece against the upper piece and the OD of the upper section of the center piece against the ID of the upper piece 2 The bearing assembly is held in place by beari
37. right traces with a conductive silver pen To improve the physical durability of the otherwise delicate traces you can cover them with a green overcoat pen or overcoat spray In the case of the H oe 200 4 OOD 9000 90 399 do OU UPA 0200099 2090039099 OO OO gd od i SENI CODO 999 Figure E 9 H bridge Top above H bridge Bottom below bridges all of the large current carrying connections were made with short pieces of 12 AWG wire rather than the traces to avoid burning up the traces The components on the H bridge include 2 p channel MOSFETs 3 n channel MOSFETS 2 NPN bipolar junction transistors BJT 9 1 kQ resistors 4 PCB mount male spade connectors and one 3 pin receptacle Note that all of the part numbers are listed on the H bridge schematic in the figure 11 10 below Starting from the bottom of the H bridge top image the spade connectors go as follows ground voltage supply forward out to motor reverse out to motor The 3 pin input will only plug in one direction and the color convention is same as mentioned above for the H bridge 102 outputs The H bridge design is relatively standard with the exception of the enable transistor added just before the ground of the board bottom most transistor on the image All of the H bridges that are controlled by the same axis of the joystick receive input from the control module at the same time
38. saving measure the same thinking applied to AC motors In the current arm design there are five main motors to get an idea of the size of inverter required Statpower a company that specializes in power conversion maintains a list of typical appliances that can utilize their power conversion equipment This list is shown below in Table 8 1 Table 8 1 Power Consumption of Typical Appliances Cell Phone Charger 10 20 Watts Camcorder Video Games Fax Machine VCR Soldering Iron Laptop 19 100 W Work Light 3 8 Drill 500 Watts According to Statpower and several other inverter manufacturers small motors like those found in power tools require between 400 and 500 Watts of power This being said the Gateway arm fitted with five AC drill would require between 2000 and 2500 Watts of power from the AC inverter Because of the price of this size inverter 1 was determined that the cost outweighed the benefits MajorPower a large supplier of power inverters carries several models in the 2500W range that retail for between 1 800 and 2 000 Other drawbacks to the power inverter option are the dimensions and weight of the inverter The 2500W inverter from MajorPower weighs 32 pounds and has overall dimensions of 20 x 15 x 5 5 which would make it difficult to position the inverter in a suitable place on the wheelchair for accessibility to the occupant or for the protection of the device The inverter also adds unnecessary complex
39. shoulder joint holes were to be machined through the centers of the spacer and mounting plate so that a 0 4375 inch shaft could be run from the shoulder motor up to the shoulder joint A 1 00 inch OD bearing was placed in the center of the twist gear to support the shaft but also allow it to rotate independently of the twist gear A 0 75 inch beveled gear was used to drive the bend of the shoulder This beveled pinion drove a 2 00 inch bevel gear which was fitted to a 0 50 inch diameter shoulder shaft which was supported by 1 00 inch diameter 21 bearings mounted inside the shoulder mounting brackets The shoulder shaft was also connected to the lower arm tubing which was placed between the shoulder mounting brackets and was held in position by snap rings on each end As previously mentioned both the twist and bend motions were to be driven by the same type of motor The common use of motors allowed for a lower overall cost fewer part numbers and easier repairs The Firestorm drill motor applies enough torque that after a small amount of gearing the torque applied at the shoulder for the bend motion would easily meet the nearly 500 in lbs required to move the arm under maximum loading conditions The torque needed to twist the arm is more than enough to over come the friction of the bearings which is easily done by the motor even before gear reduction The Firestorm also operates at a fairly low rpm so that gearing can slow the rotation of ea
40. the voltage regulator and hopefully they will fix the board Magtrol was contacted and a replacement board costs 175 The team installed a new control board and the dynamometer was recalibrated At this time it was determined that the tachometer encoder was also damaged preventing accurate speed dynamometer speed measurements This component was order and replaced at a cost of 55 dollars The only data obtained the brief test was that the motor drew a maximum current of 35 Amps at the 18 0 V battery voltage At this high torque load testing the 1 amp hr drill battery was drained in less than three minutes After all the part replacements and recalibration the dynamometer did not produce the required data for motor torque speed curves since the dynamometer only produced usable data up to 3 4 N m 30 19 in Ib 55 Figure 8 7 shows the torque speed curves for the motor alone provided by the motor manufacturer Johnson Electric Because the maximum torque provided by the motor and 57 6 1 drill gearbox combination was approximately 250 in lb this motor could be used in the shoulder joint with gear reduction to increase torque and decrease rotational speed Gear ratio calculations and gearbox efficiency calculations are shown in Appendix E A further 80 1 reduction would be required to reduce speed from 400 to 5 rpm as required by the arm design the cost of this gear reduction would negate any savings gained by the use of motors obtained from ma
41. 3 value pwm4 value m4f value m4e value m5b value m5e value Wrist twist right Sub void wrgo_code void m4b value m4e value mof value m5e value do pwm3 value pwm4 value while wr pwm3 value pwm4 value OOPic delay pwm3 value pwm4 value m4b value m4e value m5f value m5e value lt lt lt 102 1025 90 APPENDIX D ELECTRICAL SCHEMATICS 1 LSDA OOPic 1 2 2 2 GND Q Q U WON Uw TO ro r W Q C2 Q CO TO SHOULDER BENI TW ELBOW 2 M rit 24 MOTOR 53 MC6 12 MOTOR 52 MC6 12 MOTOR 5 CONTROLLER CONTROLLER i 9 ILE T6 ENABLE PWM T8 CNTRLGND T7 SYSGND 13 8 CNTRLGND 7 GND 13 OR 12 M e M e f 28 8 MC6 18 MOTOR EN od FOREARM CR e FOREARM GRIPPER MC6 19 MOTOR lt 6 12 MOTOR CONTROLLER CONTROLLER T9 B T9 B CONTROLLER 6 ENABL
42. 4 holes 80 All Tolerances 0005 The Ohio State University Gateway Coalition Unless Otherwise Stated Motor Plate NENNEN 0206 LT 150 125 125 R for 0 250 ghafi 125 4 000 1900 The Ohio Stole The Ohio State University PLATE 151 540 2 1220 F 062 0001 125 X 4 340 2 25 i 357 0001 1220 X 2 950 0001 X 2 All Fillets R 063 2 X4 DETAIL 100 1 200 125 350 550 All Tolerances 2005 The Ohio State University Gateway Coalition Unless Otherwise Stated 6 Mam 152 R 8 5 125 Ohio State University Gateway Coalition 325 00 153 7 050 625 003 2X 125003 4003 100 2X 125 4003 500 All Tolerances 005 in Ohio State University Gateway Coalition unless stated Round Top Finger Moteriak Aluminum 154 B26 4 UNC Hex head 287 T 29 Size 43 125 125 I L C 30 125 25 DEC 4 500 2000 i 025 4 comers 2500 All Tolerances 0005 The Ohio State Unversity Gateway Coalition 750 Unless Otherwise Side Plate sins Stated 155 REFERENCES ADC0831 ADC0832 ADC0834 ADC0838 8 Bit Serial I O A D Converters with Multiplexer Options Datasheets National Semicon
43. 530200 30 00 Subtotal Bearing Distributors Inc Boston Gear Distributor 20 00 1 20 00 Table A 3 Shoulder Lower Arm and Forearm Machining Costs Posty Hs Prototype Cost 25 2 500 00 68 2 GRIPPER BILL OF MATERIALS Table A 4 Gripper Bill of Materials Part Description Part Number Vendor ida Part Cost quantity Prototype Lead Screw 17 070 25 20 2 00 12 50 Moving Plate 222 5530 50 00 Gripper Motor Plate 25530 25 00 Differential Plate 5 30 25 00 30 30 1 30 30 Palm 5 30 Side Plates for Gripper Motor Fingers 52 00 Spacer betwn Diff Gear and Wrist Motor 0 35 12 50 12 85 1 12 85 1 00 Finger Link 0 16 6 25 6 41 3 19 23 Iso triangular Finger Link 1 60 s1875 wes 3 Straight Middle Finger Link 1 60 18 75 20 35 3 61 05 Finger Tips Torsional Springs for Motor Bit Screwdriver Motor Gearbox 9072 Thrust Bearing for Gripper 727 0512 Pins 1 5 ft used S121M2150 Retaining Clips 1 0424 Screws Nuts Bolts Subtotal 59426 Machining was estimated at 25 hr A 3 CONTROLS BILL OF MATERIALS Table A 5 Controls Bill of Materials em Part Cost Prototype Part Description Part Number Unit Quantity 8 Position Joystick 1 RadioShack 18 00 Switches Radio Shack 1 49 3 4 47 3 6V Black amp Decker Model VP750 Black and Decker 2 4V Black amp Dec
44. 9 etl CNN peering UE paneer 140 L cuo enter 141 Aruri dps r en beanng 72038 142 mE 3 n Anari uo pem 143 S IE Cass maa 21028 144 154 438 ax WRIST MOTOR PILLOW BLOCK WRIGHT STA TE UNIVERSITY REVISED 98 2001 ERIC vU CHES 145 13 WS Material Alwminurm TOQ m The Ohio State butewdy Coldition Differential Spacer 4 17 00 146 75 Link 100 Link rm 2 125 0 005 2X 12520005 280 40 00 28050 1750 0 05 1000 0 05 _ 1 2X R 125 003 000 Ohio State University Gateway Coalition L75 Link amp 100 Link ns Inches Material Alurninurn M Hunt 2 24 00 2 147 25 Ohio Stole University Galeway Coalition Middle Finger DETAIL iy 32300 s 148 R 100 250 230 230 20 typ 200 300 All Tolerances 0005 The Ohio State University Gateway Coalition Unless Otherwise Stated Motor Bit __________ Tozer 149 4 UNC Hex head Top Size 43 All
45. COMPUTER PRINT OUT NOMINAL MOTOR CURVES Performance and charactertics are measured based on limited motor samples only Reference no 32687 57 CONTROLS HARDWARE 9 1 MICROCONTROLLER To provide adequate control of the six different motors located the arm the selected motor controller is required to meet several requirements These requirements include cost ease of programming number of I O lines Analog to Digital A2D converter and processing speed The most important of these considerations is cost Since the project has a limited budget and power constraints a commercially available motor control system would be out of the question These products would utilize a disproportionate amount of the available budget The goal of the project is to keep the costs low making the device available to a greater percentage of the populous While cost was an important consideration ease of programming was also a priority The team members were eager to learn new programming languages and techniques but the limited time frame required the use of familiar languages Most controllers on the market utilize assembly language which was unfamiliar to all of the team members This limitation narrowed the list of viable options to two microcontrollers the BASIC Stamp and the OOPic When comparing these two options some obvious advantages developed The first major advantage of the OOPic over the Stamp is the multitasking capabilit
46. Differential 40 Although the differential gear set purchased from Sterling Instruments was one of the most expensive components purchased for the arm the mechanical advantages were deemed to appropriately offset the high cost of the product Providing two degrees of freedom within a small space was essential to maximizing the movement and functionality of the robotic arm design 41 GRIPPER 7 1 RESEARCH AND CONCEPTS The previous Gateway teams invested much time into gripper design and some very interesting ideas were developed Two designs were mainly discussed the 1999 2000 design and Chris Fearon s graduate design Other previous designs were quickly determined not to meet the goals of this year s team Chris Fearon designed a gripper that was very efficient at picking up large objects 1 5 to 3 inches but had some trouble with small objects He used three fingers as opposed to the previous Gateway teams two finger wrench designs which allowed for more stability and a greater variety of objects that could be handled Much of the work went into the analysis of the fingers and the loads that they could handle Chris did extensive research and testing using ANSYS to pick the best finger design that would limit finger deflection and stress points Figure 7 1 Fearon s final gripper design Aluminum was chosen for the major components of the fingers and moving pieces of the palm The motor was also locked into pl
47. E T6 ENABLE T7 SYSGND 13 PWM CNTRLGND 7 SYSGND 13 PWM CNTRLGND T7 SYSOh 3 91 APPENDIX 1 ALTERNATIVE CONTROLS One member of the design team developed an alternate control scheme and this control scheme is described in the following sections E 2 INTRODUCTION AND GOALS This alternative control system has been developed in attempt to further lower cost as well as gain some practical knowledge in designing controls for the arm This chapter includes details on this control system as well as the knowledge gained throughout the controls design As mentioned the first goal of this control system was to reduce cost It was earlier estimated that the current control system in development would be in the ballpark of 1000 2000 dollars making the controls cost a significant part of the total cost of the arm Any reduction in that cost would certainly help with the goal of creating an arm with a practical price tag The second goal was simply to have a control system that would operate the arm in its entirety The arm has 6 degrees of freedom therefore the controls need to be able to actuate all 6 degrees without rearranging any wires hooking and unhooking various joints motors A third goal was to have a control system with variable speed capabilities Variable speed makes fine tuned motion easier to operat
48. Final Report on Robotic Manipulator Project June 1 2001 By The Ohio State University Corey Johnson Tim Kocher Curt O Donnell Michael Stevens Aaron Weaver and Jeff Webb Advisor Gary Kinzel Wright State University Shawn Riley Jason Ruge Lawrence Thomas Eric Yu Advisor James Menart Sinclair Community College Brad Cutting Chris Shirkey Tim Trepanier and the Step II Machining amp Manufacturing Class Advisors Beth Johnson and Scott Hawkins Sponsored by NSF Gateway Coalition ABSTRACT The 2000 2001 Gateway design team engaged in a nine month design process to develop a robotic arm that is mounted to a wheelchair with the specific goal of helping paraplegics and quadriplegics function in their immediate environment This year s team consisted of students from The Ohio State University Wright State University and Sinclair Community College and was sponsored by the Gateway Engineering Coalition The educational object of this project was to teach the participants the skills to solve an engineering design problem in a cooperative environment This report presents a general background of the project along with a description of the overall group dynamics The report is organized into two main sections the design and manufacture of the arm which was the major task for Wright State and Sinclair and the controls system of the arm for which Ohio Sate was responsible The design and manufacture sections descr
49. K offtime restore state offtime 1 goto off_intr3 For this segment of off time the bits are read from the ADC and stored to input Because only 111 one 256 S he off_intr3 subwf ADC btfss goto call decf call retfie bit is read for every timer interrupt every clock cycles the process is slow enough for ADC to create nice stable output movlw d 8 offtime w SJ ATIS C OTT SHIP bit read offtime f restore 9 e e e det o call bit read read store one bit from offtime offtime 1 restore state return else b SOLO OFT BE 1 This section simply counts down all of the remaining off time movf offtime w off_intr4 iorlw btfss decf call retfie This is the algorithm qa SATUS offtime f restore 9 selection procedure DEM Wer bcf Dpef Clem CIE incf mov f xor lw DET SC clrf 0 btfsc goto mov f call movwf call retfie clrf PORTA PORTB 4 PORTB 5 ontime offtime T joint w 4 4 STATUS f joint 7 10 joint w joint tbl PORTA restore offtime offtime 1 restore state return that handels the joint e e e Turn off all joints gt Turn off reset PWM signal Determine next joint to be e e e This is used to restore and the Working
50. ace using aluminum plates The side cover plates were made from Acrylic simply for aesthetics These two materials offered high strength to weight ratios and were fairly low in cost Actuation was handled through the use of a stepper motor leadscrew system The NEMA size 17 stepper motor from Applied Motion and a lead screw and power nut assembly from Precision Industrial Components Corp were the products purchased The thrust bearing was used to eliminate the thrust force that would be placed on the motor due to heavier objects The high torque size 17 stepper motor provided 31 4 in oz of torque at 300 rpm which 15 more than enough to overcome the joint friction and operate the gripper The overall cost of material and machining for Chris gripper was about 620 for a 42 prototype and about 430 for production of 50 or more assemblies A solid edge model of Chris Fearon s gripper is shown Figure 7 2 Power Nut Stepper Motor Strips Rubber x Lead Thrust Bearing pager Figure 7 2 Fearon s Solid Edge Model The 1999 2000 Gateway team took a different approach to the design They took Chris Fearon s design to the next level by implementing underactuated fingers A mechanism 1 underactuated if it has fewer actuators than degrees of freedom This type of design allows the fingers to wrap around an object as it closes and to pick up a wide variety of objects small and large The group then decided t
51. all objects Fixing this problem would require a complete redesign of the fingers and palm Another suggestion is to somehow attach a hooking device to open things such as drawers and cupboards 11 4 CONTROLLER For the next academic year the Gateway team may want to look into controllers other than the OOPic Although the OOPic is very convenient to use in some respects it does have some shortfalls First there are other chips available now that are much faster than the chip on the OOPic Most of the controllers with these chips would need to be programmed in the assembly programming language which is not as desirable Also the OOPic should be able to handle all of the necessary calculations but there are some problems in getting the OOPic to do all of the math properly The control system durability should also be improved by eliminating large pulse width modulation amplitudes on lower rated motors 1 e not using 12 volt motor control boards with 2 4 volt motors Another suggestion 15 to implement a control code that is able to run the arm only when the wheelchair 15 turned off 66 APPENDIX SHOULDER LOWER ARM AND FOREARM BILL OF MATERIALS Table 1 Shoulder Lower Arm and Forearm Raw Materials Costs 88875 46 McMaster Carr 53 08 53 08 1 4 Diameter 399 mm go Neoprene O Ring Round 8060K35 McMaster Carr 4 98 4 98 12 x 6 x 1 4 Green Oil Impregnated Cast Nylon Sheet 84845K61 11 01
52. arry out the operation At first it was built into the code to send a quick ramp up or down of the PWM signal being sent to the motor control board for starting and stopping the motor respectively It was later found that the motor control boards would automatically ramp the motor speed up as long as the motor enable switch was reset every time the motor was started This was integrated into the single motor control code along with a PWM step down for the motor slow down event The final phase of the programming considered trying to get all of the motors to act together through the kinematic equations and feedback using potentiometers The first test codes written were to test the potentiometers connected to the OOPic This was done by turning on various input output lines depending on potentiometer position The potentiometers were then used to vary the PMW output to the motors and thus affect the motor speed Both of these codes worked very well Unfortunately the time that 1 would take to implement a system with the potentiometers and kinematic equations was too long and the team ran out of time to make an attempt at this Next year s Gateway team should have an easy time starting where this year s team left off and be able to construct a completed speed control system with potentiometer feedback fairly easily 63 In light of these problems it was decided that multi motor control to keep the arm operating in one plane at a time was not feasib
53. ate jbgate jbgate jbgate jlgate jlgate jlgate jlgate jrgate jrgate jrgate jrgate p pi p O p U AT ATT LAETITIA TIAA SAS l d lt lt lt H I OOO O lt lt lt lt lt lt dco RC ays o invertout 1 DUE operate 1 THnputl lvn ki b s invertout 1 Link Togo operate 1 L invertout 1 outputs 9 operate 1 ct 2 Pp lt CIEL invertout 1 s sk operate 1 Joystick Sub void jfgo_code void nogo 0 sl value s2 value nogo if sl value 53 15 if s2 value amp s3 value nogo if 0 1 sl value mlf value 1 mle value do lt 83 pwm2 value 153 wb value 1 while jf 0 pwm2 value 10 OOPic delay 20 pwm2 value 0 mlf value 0 mle value 0 wb value 0 1 s2 value m3f value 1 m3e value 1 pwml value 30 OOPic delay 10 do pwml value 90 wb value 1 while jf 0 pwml value 10 OOPic delay 20 pwml value 0 m3f value 0 m3e value 0 wb value 0 s3 value mof value 1 m6e value 1 pwml value 30
54. bled gripper that is shown Figure 7 7 Figure 7 7 Physical Closing Sequence of the Gripper The final prototype cost of the gripper was determined to be 524 26 and 414 88 for a production volume of 50 Notice that the cost of the gripper is very similar to that of last year s design 547 05 for 99 00 design prototype However the lack of cost savings is made up by the ease of control that was created by using the screwdriver motor The difference in price for a production volume of 50 or more assemblies is mostly due to the reduced cost of machining larger lots of parts This was estimated in the calculations by reducing the machining time by 25 from 17 5 hours to 13 125 hours The cost of the gripper components material and machining is summarized in Appendix A Bill of Materials A disadvantage of this year s gripper design 1s the increase length from 4 60 inch side plate length to 7 25 inch side plate length However the arm design allowed for a maximum gripper length of 12 inches which we are still well below The final Solid Edge gripper assembly is shown in Figure 7 8 47 Figure 7 8 Final Solid Edge Gripper Assembly 48 MOTOR TESTING 8 1 MOTOR RESEARCH AND SELECTION PROCESS Through the initial literature search and a trip to Motoman in Dayton Ohio several motor options were defined The overriding decision was whether to proceed with AC or DC motors Previous design teams have
55. box refer to Figure E 12 Inside there are two 12 AWG wires that run under the board stack and up the front with quick splice female spade connectors for five of the six H bridges The ends of the wires are topped off with right angle female spade connectors to connect power to the top H bridge The connection to the motor from an H bridge starts as two female spade connectors attached to two 16 AWG wires On the other end of the wires 15 2 pin Molex plug female which sticks out through the front of the box There are six in all From the front of the box the power to the motor then goes through a resistor block and then on to the motor It s not important which H bridges are hooked up to which motors because all of the bridges are the same It is however very important that each motor s Figure 13 Amplifier Box Internal Power Connector H bridge be hooked up to the correct output from the router board depending on which degree of motion the motor is actuating The following table 1s a list of the router board output number found on the schematic of Figure E 7 and the corresponding motor it must be connected to Remember that the actual router outputs physically correspond to those the schematic when looking at the router from the top Table E 2 Router Board Output Hookup ROUTER BOARD OUTPUT NUMBER JOINT MOTION MOTOR Shoulder Bend Shoulder Twist Elbow Bend Wrist Motor 1 y axis reversed Wrist Mot
56. ces needed by each type of motor The main advantage to any of the different types of DC motors is the ability to produce many different speed torque relationships that can be easily tailored to the needs of the project Most DC motors when properly configured can also produce as much as three to five times their rated torque for short bursts Wound DC motors permanent magnet DC motors and brushless DC motors were three types considered for the Gateway arm Stepper motors were not considered since the power requirements for the arm were above those provided by reasonably priced stepper motors Brushless DC motors were considered briefly since they react more like AC motors than DC motors They also exhibit more durability than conventional wound DC motors since they contain no commutator or brushes which wear over time This advantage is also a disadvantage since the added circuitry increases the overall cost of the motors This leaves permanent magnet motors and wound DC motors Permanent magnet DC motors provide many of the same characteristics of wound DC motors but variability in the magnetic material that goes into construction can lead to inconsistencies from motor to motor The PM motors are also susceptible to shock vibration and temperature variations that are not concerns for wound DC motors The final DC motor option considered are wound DC motors These motors come in several configurations that lead to varying control options and motor cha
57. ch of the two motions to a manageable speed One problem with the Firestorm drill motors is that even though the gearing can slow down the operation of each of the two joints they will still be much faster than the targeted 5 rpm speed 0 5m s linkage speed The gearing ratios selected were governed by space limitations of the base and shoulder joint The twist gearing was 2 5 1 and the bend gearing was 8 3 These gear ratios were chosen to produce the largest torques and the slowest rotational speed without changing the configuration and design of the rest of the components For example the bevel gear in the shoulder could not be any larger without interfering with the lower arm tubing and the shoulder pinion could not be any smaller because of the size of the motor shaft it is attached to Further investigation into electric motors led to the acquisition of less expensive more powerful and slower motors that better suited our arm than the previously selected Firestorm motors This change in motor selection required an immediate and extensive redesign of the base of the arm The new twist motor is a window lift motor see Figure 4 6 and the new bend motor is a high torque low rpm motor Due to the size and orientation of the new bend motor it could no Figure 4 6 Shoulder twist motor longer be mounted in the base and was moved to the lower arm The twist motor 1 still the base 22 but numerous changes have been made The
58. ded complexity to the design without much benefit to the customer In this year s design the team was able to locate a source of surplus motor controllers with favorable characteristics to the surplus gear motors selected for this project These model MC6 motor controllers provide 30 amps continuous control by a Pulse Width Modulation signal supplied by the OOPic microcontroller The motor controller has a built ramp function for a smooth startup both forward and reverse directions The inputs required from the OOPic are 5V enable signal to engage the motor control board two direction signals and a PWM signal for speed control As an added benefit these boards are available both 12 VDC and 24 VDC configurations Problems surfaced when using these boards to control the screwdriver motors located in the gripper and elbow and the window motor used in the shoulder rotate Electro Magnetic Interference EMI was caused by the noisier motors This EMI feedback caused the OOPic to randomly turn on and off I O lines We made an attempt to filter this noise using 0 1 capacitors mounted between the motor power leads and each lead to the motor casing This solution fixed the noise originating from the window motor To reduce the noise to a level that would allow the 1 to control the screwdriver motors a separate power source was required The battery used was a 6 V lantern battery with a common ground to the wheelchair This elim
59. drill Input power from motor Tn 9549 Power where T Motor torque in Nm n Motor speed in RPM _ 0 40126Nm 10702rpm 9549 Power 45kW 450W The resultant force between the sun and planet gears can be calculated at each mesh assuming 2 losses per mesh A free body diagram of the gear train is displayed in Figure B 2 From static analysis of this diagram the relationship between the force F and the force transferred to the arm Fa The relationships for and are as follows Ts Fs 098 75 where T Torque in the sun gear r Sun gear radius Fa 2 76 F amp Ring Sun Planet X k w Figure B 2 Planetary Gear Train Free Body Diagram We calculate that 1 24Nm is transferred to the 2 sun gear 3 63Nm is transferred to the 3 sun gear and that the output torque is 13 5Nm Since the output speed 1s 185 74 rpm the power at the exit of the gearbox 15 263 Watts POWT 2997 Input power 450W zu APPENDIX C 1 KINEMATICS Figure 1 Kinematic Model In the linkage displayed in the figure the links are represented by lengths r2 and r3 These lengths correspond to the link lengths of the arm where 15 the length of the lower arm is the length of the forearm and is the length of the gripper The angles 01 02 represent angles of the links from the horizontal
60. ductor August 1999 http www national com ds AD ADC083 1 pdf 2001 Design of a Robotic Manipulator for a Wheelchair 1999 2000 Gateway Coalition Multiuniversity Project Ohio State University 2000 Farr J M et al O Net Dictionary of Occupational Titles U S Department of Labor 1998 ISBN 1563705109 Fearon C M Design of a Robotically Assistive Orthotic Device for Children in Wheelchairs Master s Thesis Ohio State University 2000 IRF5305 HEXFET Power MOSFET Datasheets PD 91385B International Rectifier http www irf com product info datasheets data irf5305 pdf 2001 Kinzel G L Waldron K J Kinematics Dynamics and Design of Machinery New York John Wiley amp Sons Inc 1999 ISBN 0471583995 NDP6060 NDB6060 N Channel Enhancement Mode Field Effect Transistor Datasheets Fairchild Semiconductor March 1996 http www fairchildsemi com ds ND NDP6060 pdf 2001 PIC16F84A 18 Pin Enhanced Flash EEPROM 8 Bit MicroController Datasheets Microchip Technology Inc http www microchip com download lit pline picmicro families 1 6f8x devices 16f84a 35007a p df 2001 Raptor Wheelchair Robot System Rehabilitation Technologies Applied Resources Corp http www appliedresource com RTD Products Raptor index htm 2001 Scherz P Practical Electronics for Inventors New York McGraw Hill Professional Publishing 2000 ISBN 0070580782 Shigley J E and Mischke C R Mechanical Engineering Design 5th ed N
61. dule Circuit Bottom regulator At is the only thing the supply voltage is connected to The rest of the circuit 1s connect to the 5V output of the regulator which provides a nice steady voltage that is more to the digital circuit s liking With the help of a small filtering capacitor the voltage regulator also filters out most of the electrical noise created by the motors 97 The and y axis control circuits both consist of a microcontroller and analog to digital converter ADC and a single 20MHz clock oscillator which is shared by both microcontrollers The microcontroller chose was Microchip Inc s PICI6F84 20MHz variety they also make an older 4 MHz version The ADC 1 National Semiconductors ADC0831 The PIC16F84 1 an 8 bit microcontroller that has 13 I O lines It has the capability to do simple multitasking with the use of its timer overflow interrupt feature The internal clock speed of the chip is 1 4 input clock speed In this case that means chip is essentially running at 5 MHz It 15 a relatively efficient chip taking only one cycle to run an instruction with the exception of program branches such as goto s and subroutine calls One of the most attractive features of this microcontroller is the price At about 6 apiece they provide a good deal of functionality The ADC this control circuit is an 8 bit serial converter This particular converter uses a conversion technique known as successive app
62. e even a small amount the motor s effective resistance due to the back increases dramatically So the 5 8Q is much less than the motor resistance and therefore most of the power is dissipated across the motor The only real downside to the setup is that the effective current is limited to no more than about 2 75A That s not strong enough to move the larger motors when they are loaded Figure 11 Resistor Block On a final note about amplification there are several companies that make motor driver integrated circuits ICs One company in particular Allegro Microsystems makes the motor 104 driver 3952 This chip is single inline package SIP which has some nice features like current limiting temperature protection and dynamic breaking One chip alone is only capable of delivering up to 2 A of current However these chips can be wired in parallel to deliver more current For example four chips in parallel would be able to deliver 8 A of current which 1 enough to drive the shoulder bend motor Another plus 15 that an amplifier made of these chips would potentially be very small Also it is likely that it would only take some minor programming changes to the PICs order to use these motor drivers with the current control module 6 HARDWARE CONSTRUCTION T it T 14 j H HIN wih M me HL
63. e It 1s clear to see how important slow motion 1s for a manually controlled robotic arm which may at times be used to do delicate tasks that would be difficult or impossible to do without speed variability Finally a fourth goal was to create a control system that was reasonably easy and intuitive to use Because this is an interface between a machine and a human being human factors becomes an important issue The controls need to make sense be easy to operate and perform a reliable and predictable manner ASSUMPTIONS Like most design problems some reasonable assumptions can be made in order to get some clear direction for development These assumptions turned out to be a crucial part in the shaping of this control system The first assumption was that no positional feedback from the arm is required at any joint Unlike robotic arms used in industry which are largely completely automated the wheelchair arm is manually controlled by a human being who can literally see the position of the arm providing the only positional feedback necessary This assumption helped to simplify the controls a great deal and reduce cost The next assumption was that no motor speed feedback was required again simplifying the controls Motor speed feedback usually in the form of a tachometer 1 costly and used mainly for systems that require precision movements Because this arm is manually operated the user can see the speed and vary it accordin
64. e other side of the switch the output lead goes to the microcontrollers but at the same time is pulled low by 1000 resistor Because the input of the microcontrollers have nearly infinite input impedance there is no current flowing along the output line of the switch Therefore when the switch is open the output is connected to ground through the resistor and is referred to as being pulled low When the switch is closed the input voltage is passed along to the output and separated from ground by the resistor This means that the input to the microcontroller 1s whatever the input voltage to the switch was making the switch behave like a binary input high or low The supply voltage coming into the control module 1s the very same supply voltage that 1 TIE m 4 2 bj it 1 it 9 uz 4 ii LT 4 val x M 34 253139 2 55579 TE 3 of a za r ta 47 runs the amplifier module and in turn the motors The supply comes from the two 12V car batteries hooked in series that are mounted on the wheelchair This makes a total of 24 volts However because the control module is mostly a digital circuit it is not designed to run at such a high voltage To remedy this the 24V supply that is input to the control module is run through the C power button straight to a 5V voltage Figure E 6 Control Mo
65. ed 40 in lb of torque enough to operate the elbow Furthermore the screwdriver operated at a maximum of 180 rpm thus requiring less gear reduction than the drill motor Figure 5 9 Elbow Motor Mounting Block The gearing for the screwdriver involved two planetary gear sets contained within a plastic chuck To maintain the lightweight compact internal gearing arrangement and to avoid machining additional parts the entire chuck and gear assembly was removed from the screwdriver mounted to a fabricated motor mounting block The mounting block seen in Figure 5 9 was designed to hold the gears within the chuck properly mate the screwdriver motor to the gearing and secure the entire drive assembly to the lower arm tubing The block was made as thin as possible to minimize weight Two machined retaining plates were attached to the motor mounting block to secure the motor and to prevent unwanted motor rotation seen in Figure 5 10 34 Figure 5 10 Elbow Motor Assembly The adapter for the screwdriver bits was removed from the press fit clamp within the screwdriver gearing and replaced with a steel 0 25 inch diameter shaft upon which the elbow worm was aligned and pinned A 20 1 gear ratio was selected to drive the elbow joint The 16 pitch 0 625 inch diameter elbow worm was chosen from the Boston Gear catalog for its strength small size and bore diameter which permitted the worm to be attached to the screwdriver motor assembly T
66. efit associated with the use of this product over the old stepper motor The Ht17 070 cost 50 but the entire cordless screwdriver only cost 15 This cost savings was offset by the addition of new components but the main reason for choosing this motor is for consistency and ease of control One issue that we wanted to solve this year was that we wanted to use a consistent type of motor The stepper motors were very difficult to program using the OOPic controller A stepper motor offers the ability to know the exact position of the motor at all times However we do not need to know the exact position of the motor for this application The opening and closing of the gripper can be controlled visually with sufficient accuracy The screwdriver motor has the advantage of not being back drivable This eliminates the need for any braking and allows the gripper to maintain a tight grasp on an object without drawing any current Several components were re designed and several more were modified to complete our final product One drawback of the new design 15 that it is about 2 7 inches longer and slightly heavier than the previous year s gripper This is a tradeoff that we were willing to accept to achieve our previously discussed motor control goals The length of the motor and gearbox was what caused the major design changes This required lengthening the side plates to house the new equipment Since these plates were now longer they needed to be even more
67. entire arm to fold into a compact rest position Each bracket includes a jog to accommodate the flanges from the elbow shaft 36 Figure 5 13 2000 Graduate Student Elbow Bracket Figure 5 14 2001 Elbow Bracket 37 if 7 6 1 FOREARM TUBING The forearm design was closely modeled after the 2000 graduate student design once again incorporating the square aluminum tubing The tubing with a 1 8 inch wall thickness provides a compact lightweight supporting structure that completely encloses all forearm components protecting both the equipment as well as the user from potential injury The preliminary design employed a 2 5 inch square tube similar to the preliminary tube used in the lower arm Uniform tubes reduced the costs for obtaining raw materials and permitted a compact rest position for the arm The tubing supported two mounting plates which secured the motors driving the differential gears at the wrist Seen in Figure 6 1 one motor was offset behind the other motor permitting both to fit within the enclosed tube An aluminum drive shaft supported by a bearing connected the offset motor to its differential gear while the pinion from the other motor was mated directly to its respective gear Figure 6 1 Preliminary Forearm Tubing with Offset Wrist Motors 38 When the lower arm tubing was expanded for the final design to accommodate new motors and because the 2 5 inch square tubing was expensive and di
68. ere are several items that need improvement and they should become part of a new team s objectives 1 Improve the center of gravity of the arm by using lighter components and moving heavy objects such as the motors further down in the arm The use of belts has been examined in the past and was found to be rather expensive but this would greatly reduce the center of gravity of the arm 2 Make the arm lighter by selecting lighter materials and components or further redesigning the arm 3 Investigate other ways to draw power off of the two twelve volt batteries This year both batteries were wired in series to get the 24 VDC but all of the current was drawn off of the grounded battery This caused this battery to drain at a high rate 11 2 MOTORS This year s team tried to be consistent with motor selection to simplify controls and design This 14 seemed to work very well and should be improved upon in the future The goal should be to use as many of the same motors as possible Also DC motors should be the focus of study for the future teams because they tend to be very compatible with the controls and the wheelchair 11 3 GRIPPER This year s gripper design worked fairly well but there are several things that could be changed for improved performance The first problem is that the gripper is extremely long This is not necessarily a problem in operation but it looks awkward Another thing that the gripper has trouble with is sm
69. es in 3 D space so a natural instinct was to attempt to analyze its movement in three dimensions However this method of analysis would have been very complicated and unnecessary The analysis was simplified to two dimensions because only the movement in the plane of the arm needed to be considered The arm s angle of rotation about its base was not important to the kinematics In other words the orientation about the z axis vertical axis was ignored Further simplification of the kinematics analysis procedure was still desirable The control strategy was to control the movement of the gripper in one direction at a time either horizontal or vertical Therefore the kinematics needed to address the speed of the gripper moving linearly one direction at a time In other words the arm wouldn t rotate about its base and extend the gripper at the same time Considering this from a kinematics standpoint the kinematics of the arm was simplified to a three link open chain The arm is depicted as a simple linkage in Figure C 1 in the Kinematics section of the Appendix The analysis is shown in detail in that section The team wanted to make the arm as simple to operate as possible for the user Optimally we would have implemented speed control at the shoulder elbow and wrist joints Incorporating the ability to adjust motor speeds using the results of the kinematics analysis would have allowed the user to move the gripper in one direction either hori
70. etal to metal noise while the wheelchair is in motion 4 2 BASEPLATE DESIGN The governing characteristic to the design of the base of the arm assembly was the limitation of width As stated previously the arm assembly was not to increase the width of the chair by more than six inches Since the stationary plate was the part that extends the farthest from the wheelchair the width of the stationary plate was the limiting factor its design Initial brainstorming and discussion Figure 4 3 Stationary Plate led to the decision that both the twist and the bend motors of the arm should be mounted in the base of the arm to save weight in the lower arm With this in mind initially the stationary plate and corresponding components were designed to accommodate both motors in the base 19 The initial design of the stationary plate was 0 50 inch thick piece of aluminum that measured 5 00 x 7 00 inches see Figure 4 3 The large 3 50 inch diameter hole in the plate was for the placement of the twist bearing and the smaller 1 00 inch diameter cutout supports the bearing for the twist motor shaft An 18 volt Black and Decker Firestorm cordless drill motor was selected for both the twist and bend motions of the arm The twist motor was mounted directly to the bottom of stationary plate in a vertical orientation see Figure 4 4 drive gear for the twist motion was linked to the motor by a 0 4375 inch shaft supported b
71. ew York McGraw Hill Book Co 1989 ISBN 0 07 056899 5 156 ADDENDUM DESIGN MODIFICATIONS Following completion of the 2000 2001 design two flaws were discovered but could not be addressed before the completion of the academic year First the belt driving the shoulder twist pulley was found to slip It appeared that the Neoprene belt stretched under the stress of the pulley system Second binding occurred between the elbow gearing due to a lack of thrust support for both sides of the worm Additionally the two part aluminum elbow shaft was too weak to support the forearm The aluminum yielded from the stress of the steel pin securing the elbow worm gear to the shaft To eliminate slipping at the shoulder twist pulley a stranded or V shaped belt should replace the Neoprene belt Because a stranded or V shaped belt is less likely to stretch under tension the belt will maintain high pressure around the pulley eliminating slip To eliminate binding components for the elbow were redesigned The steel drive shaft for the screwdriver motor is elongated to enable placing a 0 434 inch stainless steel thrust bearing at the free end A small aluminum block houses the thrust bearing and mounts within the lower arm tubing The elbow motor is repositioned to accommodate the additional thrust bearing See Figure AD 1 for the cutaway view of the elbow joint incorporating the thrust bearing 157 Figure AD 1 Cutaway View of Redesigned Elb
72. fficult to obtain the forearm tubing was also enlarged to maintain uniformity The 3 0 inch square tubing which is readily available was chosen in place of the 2 5 inch square tube The 3 inch tubing permits the two wrist motors to be placed next to each other eliminating the need for a second motor mounting plate bearing and drive shaft Figure 6 2 illustrates the parallel position of the wrist motors Fewer components within the forearm reduced not only materials and machining costs but also the overall weight of the design Figure 6 2 Final Forearm Tubing with Parallel Wrist Motors Two oil lubricated ball bearings are press fit into two wrist pillow blocks machined from 0 25 inch aluminum plate and attached to the forearm tubing to support the wrist differential The bearings and wrist pillow blocks can be seen in Figure 6 3 39 Figure 6 3 Wrist Pillow Block and Bearing 6 2 WRIST DIFFERENTIAL The wrist twist and bend motions are controlled by a differential gear set identical to the gearing on the 1999 2000 undergraduate students and the 2000 graduate student designs The bevel gear configuration provides two degrees of freedom within a compact volume In the final design placing the wrist motors side by side increased the distance between the motor shafts Spacers appearing as red Figure 6 4 were fastened between the differential and the outer bevel gears to accommodate the repositioning Figure 6 4 Wrist
73. for a rehabilitation robot even though it may be limited Furthermore if one looks at the number of assistive robotic products sold as compared to the number of people that could benefit from such a product it is obvious that only a very small portion of these individuals are currently benefiting from rehabilitation robots This suggests that there is still a need for such a product if 1 could be designed to be affordable and efficient Since the product 15 aimed at individuals who have a deficiency manipulation ability the primary focus of such a device 1s to provide the user with a device that aids them in performing day to day manipulation tasks Other groups have conducted research to determine the impact on the life of the user by such a robotic device A study was performed by creating a profile of an individual with a severe manipulation disability This profile was defined as a person having sedentary strength and no use of reaching handling and fingering The Dictionary of Occupational Titles which defines all jobs in terms of different levels of manipulation abilities was then used to find the number of possible types of jobs that an individual with this particular profile could hold The study found 40 job descriptions that this type of individual would be able to perform These jobs consisted of primarily professional technical and managerial jobs The study then made the assumption that with the aid of an assistive robotic product
74. ful The important characteristic 15 whether the robotic aid will meet the needs of the consumer COALITION DYNAMICS 2 1 PARTICIPANTS The participants in the 2000 2001 project at each of the schools are listed below next to their respective institutional logo Table 2 1 2000 2001 Gateway Coalition Team Members OHIO STATE UNIVERSITY Corey Johnson Tim Kocher Curt O Donnell Michael Stevens Aaron Weaver Jeff Webb WRIGHT STATE UNIVERSITY Shawn Riley a Jason Ruge WRIGHT STATE Lawrence Thomas UNIVERSITY Eric Yu SINCLAIR COMMUNITY COLLEGE Brad Cutting Chris Shirkey Tim Trepanier Step Machining amp Manufacturing Class 10 2 2 COMMUNICATION TECHNIQUES Throughout the course of this project three primary modes of communication provided the means for exchanging information and successfully collaborating between the teams E mail was an effective way for the groups to relay data and ideas Weekly teleconferences were scheduled to maintain regular communication between groups and to address ideas and concerns in a more personal manner The Gateway Coalition web pages were extremely helpful in transferring large data files and for keeping the other schools informed on progress Additionally the groups met together at the beginning of the academic year to exchange initial design ideas formulate project goals and establish communication arrangements The teams also assembled before the fall
75. gly The final assumption was that the user as limited dexterity in the hands This is a reasonably assumption because the arm 1s being designed in the hopes that it can be used by a paraplegic or quadriplegic Limited hand dexterity in this case means that the user is able control the movement of his her hand but may not be able to grip things As a result a joystick with yaw control 3 D joystick would not be a viable option This limits us mainly to buttons switch and a 2 axis joystick similar to the one used to control the chair itself E 4 CONTROL DESIGN GENERAL DESCRIPTION The controls consist of two system not including the motors themselves There is the actual control module and there is the amplification module The control module is what the user interfaces with to actuate the arm and the amplification module amplifies the control module signals in order to power 5 the motors The user operates the control module Figure E 1 Control Box with a 2 axis analog joystick which has a built in button The x and y axis of the joystick are used to allow variable speed control of two different motors either direction at the same time The button is used to select one of four available operation modes 93 Because the arm has 6 degrees of freedom and the joystick has only two not all arm motors can be controlled at the same time without the use of a very high level math processor and some rather complicated programmi
76. gn The shoulder joint bearings are model E7 S3F flanged ball bearings from PIC Design They have an outside diameter of 1 125 inches and a bore of 0 5 inches This accommodates the shoulder shaft perfectly The bearings are rated much higher than they will ever be tested in practical use The shoulder shaft is a simple D shape design with a diameter of 0 5 inches see Figure 5 3 The D shape shaft fits into a similar hole cut into one end of the lower arm tube This effectively mounts the shaft to the lower arm so they move as a single unit The shaft rotates in the mounting brackets by way of the shoulder bearings The motion of the arm is caused by Figure 5 3 Shoulder Shaft 30 the shoulder bend motor driving the shoulder pinion around the shoulder gear see Figure 5 6 5 2 LOWER ARM TUBE In trying to reduce the overall weight of the robotic arm in comparison to previous designs the structure of the lower arm was rethought and redesigned In the 1999 2000 Gateway team design the lower arm primarily consisted of two 0 5 inch thick aluminum plates see Figure 5 4 Though these plates were very strong they added an unnecessary amount of weight to the arm By using square aluminum tubing see Figure 5 5 we were able to keep the structural integrity of the arm while reducing the weight significantly This design 15 similar to that of recent OSU graduate student Chris Fearon Figure 5 4 Arm Plate from 1999 2000 Desi
77. gn Figure 5 5 Lower Arm Tube Figure 5 6 Preliminary and Final Shoulder Joint Designs 31 In our preliminary design the shoulder bend motor was attached to the base of the arm Using a set of bevel gears the power was transferred to the shoulder joint This kept the weight of the motor off the arm thereby reducing the required torque to rotate the shoulder joint Upon changing the motor for this application this mounting position was no longer feasible for the shoulder bend motor due to space limitations The solution was to house the motor in the lower arm tube and transfer power to the joint using a pair of spur gears Although this position adds to the necessary motor torque it also reduces the cost of gearing significantly over the initial design The large size of the motor chosen to power the bending motion at the shoulder joint required the lower arm tube to be machined extensively Since the motor could not be placed inside the tubing completely a large contoured hole was machined out of the tube to accommodate the protruding parts of the motor To provide for easy servicing a slot was cut into the tubing to allow the motor to be slid into place and bolted securely This allows for easy assembly and service of the shoulder motor and joint see Figure 5 7 Figure 5 7 Exploded View of Shoulder Joint 32 5 3 ELBOW MOTOR The motor initially selected for the elbow joint was the 18 volt Black amp Decker Firestorm drill
78. h respect to the ground to remain fixed while the elbow or shoulder bend 15 actuated This has many benefits to the user For example the operator could use the arm to pick up a glass of water and the controls would make sure that no water was spilled by keeping the gripper level with the ground There are several disadvantages to the control strategy we used however For one the gripper cannot be made to move only a horizontal or vertical direction but will always move a combination both horizontally and vertically Also the gripper s linear speed 15 not fixed this control strategy and does not remain constant as the arm moves 10 2 OOPIC AND CODE DEVELOPMENT The OOP in OOPic stands for object oriented programming In this scheme objects are created and various properties of these objects are manipulated order to achieve the desired output In the OOPic all of the embedded controller s hardware circuits are grouped into objects Now instead of each of the controller s circuits having to be called individually circuits that will act together in the code are given one object name This once the programmer is familiar with the objects can greatly simplify how the code 1 written This section recounts the process of learning object orient programming and how the code evolved as the team s knowledge increased It also touches on why control of the arm using the kinematic equations was not used and what multi motor contro
79. hair 38 11 Extension from side of the wheelchair frame 45 11 o Total Weight 22 165 Lift Capacity 3 165 The above values are actual values measured from the final arm once it was completely assembled and mounted on the wheelchair The design criteria as mentioned before were not only met but also exceeded in some areas 17 BASE 4 1 MOUNTING BRACKETS Due to the limitation in mounting positions and the structural stability of the frame of the wheelchair it was decided that the arm would once again be mounted to the right side of the frame of the wheelchair The frame 15 a rectangular cross sectional beam that runs from the front to the back of the wheelchair see Figure 3 1 This provides for a strong mounting position between the front and rear tires of the wheelchair The only limitations that needed to be considered with this mounting location were clearance of the front and rear tires location of the cross beam of the frame and height to the wheelchair arm rest The brackets see Figure 4 1 are custom machined plates of aluminum that fit over the frame Rubber inserts are attached to the inside of the bracket cutouts and are Figure 4 1 Rear Mounting Bracket the only parts of the entire arm assembly that contact the wheelchair The selection of the size of the mounting bracket cutouts is such that when the rubber inserts are added the brackets fit snuggly around the frame The cutouts in the
80. he associated 1 25 inch pitch diameter gear provided the necessary gearing reduction while permitting enough room for the worm motor assembly to be completely mounted within the lower arm tubing The assembly was mounted at a 21 53 offset from horizontal as seen in Figure 5 11 Figure 5 11 Elbow Motor Mounted at Angle 35 The elbow shaft was modified both to secure the worm gear and eliminate the retaining clips The 0 5 inch shaft was revised into two mated components each with a 0 75 inch diameter flange which when assembled in the lower arm were secured between the elbow bracket and elbow bearings effectively preventing the shaft from moving laterally The worm gear with a 0 25 inch bore diameter was clamped between the two elbow shaft components and pinned upon the male member The entire shaft and gear assembly is seen in Figure 5 11 and 5 12 Figure 5 12 Elbow Shaft and Gear Assembly D Elbow Brackets Connecting the forearm and the lower arm are two 0 125 inch thick aluminum elbow brackets Incorporating a 135 degree bend the brackets significantly increase the range of rotational motion of the forearm Combined with mounting the brackets on the outside of the aluminum tubing bracket design increases rotational motion from 180 degrees in the 1999 2000 undergraduate students and 2000 graduate student s designs to 255 degrees The angle also allows the forearm to rest directly atop the lower arm permitting the
81. he base Figure 1 6 1998 1999 Final Design 1 5 1999 2000 DESIGN Seen in Figure 1 7 the 1999 2000 design utilized six degrees of freedom in the arm and was mounted to a wheelchair The large range of motion was made possible by two motors located at the shoulder controlling the twist and bend motions as well as a motor controlling the elbow bend The arm also featured a three point underactuated gripper connected to a compact differential gear set which permitted both twist and bend at the wrist The final design was not without drawbacks however The entire arm assembly was heavy weighing over 40 pounds and requiring at least two people to mount the arm to the wheelchair Extensive machining also increased the manufacturing costs Additionally the arm could only lift a 1 kilogram 2 2 pound payload Figure 1 7 1999 2000 Final Design 1 6 2000 GRADUATE STUDENT DESIGN Ohio State University graduate student Chris Fearon produced a completely enclosed design in the summer of 2000 as seen in Figure 1 8 His design incorporated 2 5 inch square tubes throughout the length of the arm and the shoulder bend motor was located within the tubing The motor for the shoulder twist rotation was placed beneath a compact mounting plate The smaller design was lighter overall weight than the 1999 2000 design The primary drawback to the design was its cost requiring over 10 000 for extensive machining and fabricating Furthermore mo
82. ibe the design of the base lower arm forearm and gripper The controls portion is broken into sections covering motor testing controls hardware and controls software and strategy Several conclusions that were drawn and could be used to improve the arm include 1 the center of gravity of the arm should be lowered 2 an attempt should be made to make the arm lighter 3 electrical current should be drawn off of both wheelchair batteries instead of just one 4 all of the motors should be of the same type preferably DC to simplify the control scheme 5 a new controller other than the OOPic should be considered and 6 the gripper should be examined and possibly modified so that it can grip small objects ACKNOWLEDGEMENTS The 2000 2001 multi university design team would like to thank the Gateway Coalition and Invacare the primary sponsors for the robotic arm project Without their support this project would not be possible Special thanks to the faculty advisors Dr Gary Kinzel of The Ohio State University Dr James Menart of Wright State University and Beth Johnson and Scott Hawkins at Sinclair Community College Their guidance and leadership have made this project a highlight in the college experience of the participating students The participants at Ohio State would like to thank Dr Giorgio Rizzoni for his knowledge and assistance in motor testing Ted Harper for fabrication of the motor test stand mount Joe West for motor
83. iew of Redesigned Elbow Joint AD 2 Final Redesigned Elbow 58 74 71 78 PROJECT HISTORY amp OVERVIEW 1 1 GATEWAY COALITION PROGRAM BACKGROUND The Gateway Coalition is an organization comprising seven institutions dedicated toward advancing engineering education Supported by the Engineering Directorate of the National Science Foundation the Gateway Coalition sponsors several projects including this multi university senior design project The system developed for this project 1s a wheelchair mounted robotic arm to assist paraplegics and quadriplegics in their daily lives Three institutions are currently collaborating to develop the robotic arm Ohio State University Wright State University and Sinclair Community College The multi university project has continued since 1995 and the robotic arm for a wheelchair has been used as a project since 1996 A brief description of the previous years efforts is given in the following sections This is followed by the details on the project for the 2000 2001 academic year 1 2 1996 1997 DESIGN The 1996 1997 final design 15 illustrated in Figure 1 1 This device was a 3 link 6 degree of freedom device with a transmission system consisting of cables and transmission pulleys The purpose of the transmission pulleys was to keep the motors at the base of the robot thus decreasing the torque on the arm Unfortunately a mistake was made in the initial torque analysis so inappropriate motor
84. inated power supply line noise To eliminate this extra power source reduction of noise in the power lines would be required An inexpensive fix would be to use ferrite cores around the power lines leading to the OOPic 59 9 3 ANGLE MEASUREMENT FOR CLOSED LOOP FEEDBACK Two types of measurement devices were investigated for determining the joint angles Both encoders and potentiometers were considered Absolute encoders were considered more favorably over incremental encoders since they did not require an initial home position for angle measurement The cost of the absolute encoders at around 1000 for each of the encoder assemblies made them an unacceptable option Incremental encoders are a much cheaper option about 100 per encoder but require the arm to return to a home position prior to any angle measurement This characteristic as well as the increased processing requirement made incremental encoders an unacceptable option The next type of measurement device explored was the use of potentiometers Initial analysis of the potentiometer precision needed for the feedback control required more processing than the 8 bit A2D available from the OOPic Upon further investigation all angle measurements were limited to less than 360 thus spreading the 256 divisions over a smaller angle range 60 CONTROLS 10 1 KINEMATICS AND CONTROL STRATEGY The team studied the kinematics of the arm to understand its motion The robot arm mov
85. ith all of these components we have the makings of the amplifier module Hook up a motor and a power supply input the signals from the control module and we have a motor with variable speed There is however one problem that can be devastating to the H bridges This problem occurs just before the motor starts moving The problem is that when the motor 1s not turning it 1 not generating any back and behaves like a simple piece of wire a short circuit This H bridge design has no current limiting features So during motor startup if the motor does not start to rotate immediately there is a voltage drop of 24 volts across the three transistors in the active circuit path All of these transistors have an operating resistance in the 1075 of milli ohm range meaning that the power they are dissipating is in the kilowatt region The transistors cannot dissipate much power for more that a short time before they blow The simplest way to fix this problem is to put a low value high power resistor 1n series with the motor In this case the best resistance found on short notice was a 5Q 50W resistor and a 0 80 25W resistor After PWM the effective voltage across the motor is about 16V That makes the power dissipation across the 5 8Q resistance approximately 44W re v 16 44W R 5 8 This may seem like a waste of power but the only time this much power 1 dissipated by the resistors 1s before the motor starts moving Once the motor starts to mov
86. ity to the design For example operating the controls for the inverter would be difficult for the disabled user and any malfunctions or blown fuses would disable the arm Although AC drill motors are less expensive for similar capacities the large cost of the inverter all but negates this advantage The second disadvantage to AC motors 1s the need to run the motor at high speeds to obtain the rated torque for a given motor AC motors produce most of their torque at higher speeds and this torque drops off rather quickly as the motor speed is reduced The Gateway design does not require high speeds any of the main motors so additional gearing would be required to obtain the necessary torque at lower speeds This gear reduction adds weight and additional cost to the design Therefore these drawbacks outweigh the gains AC motors have over DC motors Once DC motors were determined to be the choice for the Gateway arm the next decision relied upon which type of DC motor to use In previous Gateway designs several 50 different types of DC motors were used because the motor s specific advantage provided the best result for a specific arm function This allowed the arm to be optimized with these different motors but made controls a more tedious job trying to balance the various inputs and outputs required by the assortment of motors The decision was made to settle upon a single type of DC motor simplifying the controls and the different interfa
87. ker Model 9072 Black and Decker Screwdriver Motor 24V Motor Diverse Electronics 12V Car Window Motor Mendelson Liquidation 12V Wrist Motors 2 87245027 Pittman Reynolds Electronics RadioShack 10K Resistors Radio Shack U shaped Pin Connectors jm Radio Shack Quick Disconnects RadioShack Breakout Boards 7 RadioShack __ ProtopingBoards RadioShack Terminal Posts 1 RadioShack 5 Pin Connectors WM2003 ND Digikey WM2000 ND Crimp Terminals Subtotal 4 SUMMARIZED COST OF ARM Table A 6 Summarized Cost of Arm Subassembly Prototype Cost Base Lower Arm and Forearm 3605 78 524 26 828 13 Subtotal 4958 17 71 5 SUPPLIER INFORMATION Table 7 Supplier Information Stock Drive Company Name Bearing Distributors Inc Products Sterling PIC Design McMaster Carr Instrument OH Sales OH Sales Representative Representative Branch 52 Dayton Mitchell Contact Person Niese Branch Manager Joe Batdorf Office Manager ROCKFORD JOHN n OLSEN CONTROLS OHIO Address 343 Godshall Drive 815 W Liberty 2017A Lublin Drive P O 94930 P O Box 4 Dayton 45401 0761 Harleysville 19438 0003 Medina OH 44258 d i d 44101 Phone 614 861 Phone 215 256 6601 or Phone 800 572 Phone Number Phone 513 224 1537 1 feb 1 8626 0479 2776 800 270 330 995 5500 FAX 513 224 5868 215
88. l odiol odiol odiol wt wh wl wr obit nogo ELBOW BEND amp 5 FOREARM GRIPPER new opwm PWM 1 for m3 amp m6 new opwm PWM 2 for ml amp m2 new odiol Motor 1 forward line new odiol Motor 1 enable line new odiol Motor 1 backward line new odiol Motor 2 forward line new odiol Motor 2 enable line new odiol Motor 2 backward line new odiol Motor forward line new odiol Motor 3 enable line new odiol Motor 3 backward line new odiol Motor 6 forward line new odiol Motor 6 enable line new odiol Motor 6 backward line new odiol Joystick forward line new odiol Joystick backward line new odiol Joystick left line new odiol Hf Joystick eight Line new ogate if Gate bo linke Jfgo to 9f e new ogate Gate to link jbgo to jb new ogate Gate to Link 190 Eo Jl e new ogate i Gare bo tink Jbgo t Jr new oevent Event for joystick forward new oevent Event for joystick backward new oevent Event for joystick left new oevent Event for joystick right new odiol l Switch Xl new odiol 7 Switch 2 new odiol Switch 3 new odiol Command SLAVE to bend wrist forward new odiol Command SLAVE to bend wrist backward new odiol Command SLAVE to twist wrist left new odiol Command SLAVE to twist wrist right new obit sub void main void 81 pwml ioline 17 pwm
89. l is implemented using a simplified kinematic approach The first task was to become familiar with the basics of OOPic programming and operation First off it was decided to program in C because that is the language that the team 62 had the most experience with Then as test a sample code that was provided by Savage Innovations the OOPic s manufacturer was downloaded into the OOPic This code was to simply have a LED flash on and off once a second The next task was to modify the sample code by integrating a Virtual Circuit This again was provided in the OOPic s manual It became evident early that Virtual Circuits would be the most important part of the control code so it became vital to become familiar with their operation early in the programming phase A Virtual Circuit links a property of one object to the property of another This allows the code to constantly monitor a specific value and have this value continually updated in the code This is important for monitoring things such as joystick input and system feedback Virtual Circuits also allow the code to enter events or subroutines which are vital to the system s operation After becoming familiar with the operation of the OOPic Virtual Circuits and event driven programming the team began to develop code for single motor control This code involved reading a joystick input determining whether the input was for motor forward or backward and then entering an event to c
90. l prescale 3 pwml period 255 pwml operate 1 pwml value 0 pwm2 ioline 18 pwm2 prescale 3 pwm2 period 255 pwm2 operate 1 pwm2 value 0 mlf ioline 31 mlf direction cvoutput mlb ioline 30 mnibdrrectroH Gvoubsputs mle ioline 29 mle direction cvoutput m2f ioline 28 m2b ioline 27 m2b direction cvoutput m2e ioline 26 m2e direction cvoutput m3f ioline 25 lt artrece Lom m3b ioline 24 m3b direction cvoutput m3e ioline 23 m3e direction cvoutput wf ioline 5 wE direction cvoutput wb ioline 4 wb direction cvoutput wl ioline 7 wl direction cvoutput wr ioline 6 rection mof ioline 3 mot drrecCion oQvouuput m6b ioline 2 m6b direction cvoutput m6e ioline 1 m6e direction cvoutput jf ioline 13 jf direction cvinput jb ioline 14 JD OL rOGCctloH 11 1011 12 jl direction cvinput jr ioline 11 jr direction cvinput sl ioline 8 Sl direction covVvLmnpub s2 ioline 9 s3 ioline 10 82 SS reer Lon ccgvrnpub mlf value mlb value mle value m2f value m2b value m2e value m3f value m3b value m3e value wf value wb value wl value wr value 0 mof value 0 m6b value 0 m6e value 0 jfgate jfgate jfgate jfgate jbg
91. lained here We will still just focus on one axis of motion because the other axis functions the same way The process starts when the PIC requests a joystick reading from the ADC After the ADC has finished with the conversion the PIC begins reading it in one bit at a time Meanwhile the PWM is in last stages of the off cycle When the entire conversion 15 read in the PIC determines whether the joystick 15 In reverse forward or the dead zone The dead zone 15 the area where the joystick 15 centered that 15 about 1 5 the width of the joystick s entire range of motion input number will be between 100 and 155 If the joystick is this location the PIC produces duty cycle goes back to the beginning of off cycle and starts all over again If the joystick is reverse the PIC subtracts the input number between 0 and 99 from 110 and the result is the length of the duty cycle in timer interrupts When the joy stick is in the forward position the PIC subtracts 146 from the input between 156 and 255 and the result 15 the length of the duty cycle in timer interrupts After determining the duty cycle length the PIC turns on the forward line for forward or the reverse line for reverse Notice that only one of these lines can be on at a time During the duty cycle the PIC basically just continues to count down the timer interrupts until the duty cycle is to end At the end of the duty cycle which ever line was turned on is now turned
92. le to accomplish by the end of this academic year It was decided though that a simplified multi motor control scheme could be implemented with little problem This scheme would not require the use of a feedback system Basically the simplified control scheme ensures that the gripper will stay level with the ground so that the orientation of the item in the gripper would not change after it has been grasped This is particularly useful when the user does not want objects such as a glass of water to be tipped when he or she 1s manipulating it One final feature that was implemented was a one second delay of the gripper actuation This feature is intended to protect against an inadvertent movement of the joystick that would normally cause gripper movement The benefit being that an accidental movement of the joystick will not allow the object in grasp to be released In order to implement the control of 6 motors a user friendly fashion 3 joystick modes were used There are a total of 12 operations that are controlled by one four way joystick and three switches A list of the mode operations 1s shown in Table 10 1 below These operations were assigned an intuitive fashion to simplify use of the arm Note that the shoulder bend and elbow bend commands incorporate the wrist leveling algorithm Table 10 1 Control Modes MODE LEFT RIGHT FORWARD BACK Shoulder Twist Shoulder Bend 64 FUTURE CONSI DERATI ONS 11 1 GENERAL Th
93. ne exception The y axis systems mode enable lines power a set of four LEDs which indicate the current mode selection In contrast the x axis systems enable lines are the very same lines that run down to the amplifier module The joystick is essentially two 10 potentiometers and a normally off contact switch push button Each potentiometer 15 one axis of motion A variable output voltage 1s achieved by applying a low and T high reference voltage to the potentiometer Using its center tap as the output the potentiometer is varied by moving the ji joystick This changes resistance of Ni NB eu 25 Figure E 5 Control Module Circuit potentiometer and as a result varies the 96 voltage between the low and high reference voltages This is very common way of achieving analog control in electronics The contact switch built into the joystick is a normally off switch meaning that it is the off position until pushed The difference between a contact switch and a standard button is that a contact switch wasless play between the on state and the off state With an ordinary button there 1 some gray area between on and off This causes problems with a digital circuit that 1 trying to interpret whether the switch is sending a high or low signal This particular contact switch is used to select the desired operating mode The input lead of the switch is connected to the supply voltage On th
94. new stationary plate has been extended to be 5 00 x 8 00 inches and no longer has a uniform thickness see Figures 4 7 and 4 8 To mount the twist motor under the stationary plate length wise the plate is 0 85 inches thick at one end and 0 625 inches at the other This mounting configuration is also why the stationary plate had to be extended out to 8 00 inches The change 1s so that the motor will fit tightly against the bottom of the plate and a minimum amount of weight would be added to the base 3 5 inch diameter circle is cut 0 5625 inches deep for the placement of the twist bearing assembly Also a 1 125 inch diameter hole is cut to allow clearance Figure 4 7 New Stationary Plate for the new twist drive pulley A 4 00 x 4 00 inch Square area concentric with the bearing cutout is cut 0 125 inches deep so a bearing retaining plate can be mounted flush to the top of the stationary plate Figure 4 8 Bottom of Stationary Plate with and without the Twist Motor Orienting the motor across the bottom of the plate and having two different thicknesses required a change in the rear mounting bracket also see Figure 4 9 The placement of the rear bracket 1 still the same and so are the majority of the dimensions except now the rear bracket has to go around the motor Therefore a cutout was made to the old design see Figure 4 9 to make sure the mounting bracket would fit around the motor Also the step on the bottom of the 23
95. ng To fix this issue the arm s operation is broken into four different modes each controlling a different combination of motors joints The different modes and their joint combinations are shown below in Table E 1 Table 1 Controller Mode Setup MODE JOINT MOTION JOYSTICK AXIS Should Twist Should Bend Should Twist Elbow Bend Wrist Bend Gripper Open Close This control module uses a technique called pulse width modulation or PWM to vary the speed of the motors This process will be explained in greater detail in the next section The signals from the control module are amplified by H bridges in the amplifier module This too will be explained the next section 94 5 CONTROLS DESIGN DETAILED DESCRIPTION As mentioned in the previous section the controls system consists of two modules the control module and the amplifier module The two modules are connected with a 10 conductor cable that carries all of the control signals as well as the power supply and common ground for the control module User Input SUPPLY VOLTAGE GROUND MODE ENABLE MODE 2 ENABLE MODE 3 ENABLE MODE 4 ENABLE X AXIS REVERSE X AXIS FORWARD Y AXIS REVERSE Y AXIS FORWARD POWER Figure E 3 Basic Signal Flow Chart The signals going to the amplifier module consist of the mode enable signals the x axis forward and reverse and the y axis forward and reverse The PWM signal is sent across the f
96. ng base This design was structurally and functionally better than the previous year s design However it was very expensive due to the large amount of machining required and it was never mounted to a wheelchair Another drawback of this design was that the gripper could only be actuated one direction It was spring loaded in the open position and closed by a cable which was wound with a small motor Figure 1 3 1997 1998 Final Design Figure 1 4 1997 1998 Final Design Folded Position Extended Position 1 4 1998 1999 DESIGN Figure 1 5 is a solid model of the 1998 1999 initial design This design utilized motor on joint control and square extruded aluminum tubing an attempt to reduce the complexity and machining costs of the arm Another advancement was the attempt to control the gripper motion in both directions Additionally this was also the first year that the robot was actually mounted to a wheelchair Figure 1 5 1998 1999 Initial Design Figure 1 6 is a photograph of the prototype that was built One of the main shortcomings of this design was that the gripper was very bulky and could not produce enough force to lift a payload The shoulder motor extended about 8 inches from the side of the joint severely inhibiting the passage of the wheelchair through standard doorways Finally although the arm was mounted to the wheelchair the base rotation failed to function due to both an undersized bearing and motor at t
97. ng plate This 4 00 inch square plate bolts to the stationary plate in the previously described cutout and holds the bearing assembly in the stationary plate The final base assembly is shown in Figure 4 15 with the stationary plate twist pulley and twist bearing plate all transparent so that the rest of the components can be seen Finally an exploded view is shown in Figure 4 16 This shows each component of the redesigned base assembly and how they are assembled Figure 4 16 Exploded View of Base Assembly 28 LOWER ARM 5 1 SHOULDER JOINT The shoulder joint was of primary importance in the design of our robotic arm Due to the weight of the arm and its overall length the imposing moment on this joint is substantial Preliminary calculations estimated this moment to be 483 Ib in This meant that all components of the shoulder joint needed to be designed to support this load This was especially crucial in choosing a motor to power the bending motion at this joint During early brainstorming sessions the thought of using cordless drill motors for high torque joints seemed feasible Upon testing however it was apparent that excessive gearing would be required to bring the rotational speed of the motor down to a reasonable speed It was for this reason that a gearhead motor was sought The chosen motor is a 24V DC gearhead motor that supplies 300 Ib in of torque and rotates at a maximum speed of 33 rpm see Figure 5 1 This motor
98. o use a six bar underactuated two degree of freedom linkage This design took much kinematic research and testing An example of a six bar linkage closing around an object 1s shown in Figure 7 3 Figure 7 3 Closing sequence of an underactuated system 43 This design could pick up cylindrical objects ranging in size from 1 to 4 inches Forces were studied and parts were designed to reduce stress concentration and machining costs This meant rounding edges to create more dumbbell shaped parts Most of the components were constructed of aluminum but the side plates and finger pieces were made of Lexan The gripper was again actuated by stepper motor leadscrew system The 7 070 stepper motor from Applied Motion Products and 4 inch leadscrew from PIC design were the products purchased 1 4 inch aluminum plate was moved by the leadscrew and slid in slots cut into the J F a sideplates This motion is what closed the Figure 7 4 1999 2000 Final Assembly fingers and actuated the six bar linkage This turned out to be a very cost effective design as well One of the main reasons that Chris Fearon did not invest much research into the use of an underactuated system is that he thought the cost would be too high After all of the kinematic analysis and design was finished the actual machining and material cost was only 547 This was actually a lower cost than Chris Fearon s prototype g
99. off and the PIC enters the off cycle state were it will begin the process all over again That procedure goes on indefinitely as long as the controller in turned on The only exception is when the mode select button is pressed Like the timer overflow the button pushing triggers an interrupt The PIC stops whatever it s doing shuts off both forward and reverse lines changes to the next mode and then stays in pause loop until the button 1s released At that point the PIC returns to the beginning of the off cycle and goes back to normal operation The code for both PICs 1 the identical It can be found in Section E 7 The operation and output of the control module is worthless without some way to amplify the signal to a point that 1s strong enough to actually power the motors This 1s the reason behind the amplifier module 100 The amplifier module has two major parts the router board and the H bridges The router board does exactly what the name implies it routes the signals coming from the controller module to the appropriate place It is a simple device consisting of 12 pin Molex plug for input a 2 pin power input and six 3 pin output All of this is connected with a series of wires tracers and diodes The other part of the amplifier module is the part containing amplifiers themselves The type of amplifier used 1s called an H bridge a device designed specifically for 2 direction motor control There are six H bridges one for each moto
100. or 2 Gripper Open Close A final bit of information that might be useful pertains to the making of the control module printed circuit board Mostly all PCBs are created using the same general process You start with a board completely plated with copper on one or both sides You mask the copper you 106 want to keep using some form of etch resistant covering Finally you etch the board using a chemical that dissolves the unmasked copper away The most common chemicals used are ferric chloride and sodium perchlorate The most significant difference between board making techniques lies in the way the board is masked A common form of masking especially in industry is the use of a photo sensitive board For this process a negative of the circuit design 15 created and then in a darkroom the board 15 flashed with the image exposing all of the parts that are to be masked The exposed parts turn black and the rest is cleaned off This process is very fast and precise but requires a lot of equipment and chemicals Another means of masking is the use of an etch resistant pen and or etch resistant decals Both of these masking tools are good for quick and dirty PCB creation but it is nearly impossible to make anything with any amount of precision or detail The lack of precision makes this method practically worthless when the circuit contains items such as microchips that have pins that are very close together 0 1 spacing Another rathe
101. ors must be taken into serious consideration when attempting to design a rehabilitation robot Table 1 1 Previous Attempts at Marketing Robotic Arm Product Name Approx Cost Approx Sold Where Sold Prab PRAB Vocational Worksites Command USA Robotics Workstation 48 000 20 Rehab Centers DeVar USA Vocational 100 000 3 Clinical Works Inc Workstation Evaluation Manus Nether Exact Wheelchair 35 000 50 lands Dynamics Mountable Handy 1 UK Rehab Mobile Base 6 000 140 Robotics Feeding Unit SA 0 UK K A Individual Users Clinical and Research Evaluation Dutch Kinetic Wheelchair Helping Hand U Rehabilitation Mountable 1 Instruments U Oxford Vocational Clinical RAID France Intelligent Workstation 55 000 Evaluation Sweden Machines Education Arlyn Arm USA Vocational 30 000 Workstation Robotic Sidekick USA Assistance Mobile Base Corporation Raptor USA Rehabilitation Wheelchair 11 950 Unknown Robotic Neil Vocational Clinical Assistive Canada Squire Workstation 23 000 7 Rehab and Appliance Foundation Industry It is also worth noting that the price and performance of a robotic aid is linked to N Papworth Wheelchair Clinical and Papworth Arm Group Mountable 8 000 5 Research Evaluation the complexity of its design For example fewer degrees of freedom will lead to a device with less capability However this fact alone does not mean that simpler robotic aids will not be use
102. orward and reverse lines while a steady logic on off signal 1s sent down each of the four mode enable lines Only one of the four mode lines 1s on at a time The control module houses a custom made printed circuit board PCB which contains the entire control module circuit minus the on off button located on the box lid In order to allow for the board s easy removal from the box all of its input output lines run to a 9 pin Molex plug which connects to the leads of the outgoing 10 conductor cable The tenth wire which is the supply voltage is routed separately through the power button via two sets of shielded spade connectors 95 JOYSTICK 3 mccum 1 i RS oe ee 277 gt CLK Apco834 aa poo poo 4 po E 2 ME 2 se vas ss pn ims on iW 4 1 1 MODE4 N a a a a REVERSE JU 1 FORWARD 1 MODES 1 MODE2 L REVERSE FORWARD Sut MODE 4 PIC1I6F84 C1 0 4u Figure 4 Control Circuit Schematic The circuit in the control module can be broken up into three main parts the joystick the x axis controller and the y axis controller Both the x and y axis controller contain identical hardware with o
103. ost of Arm Supplier Information E 1 Controller Mode Setup E 2 Router Board Output Hookup Fl Motor Testing Data Table AD 1 Elbow Revision Bill of Materials FIGURES 1 1 1996 1997 Final Design 1 2 Ohio State s 1997 1998 Initial Design 1 3 1997 1998 Final Design Folded Position 1 4 1997 1998 Final Design Extended Position 1 5 1998 1999 Initial Design 1 6 1998 1999 Final Design be 1999 2000 Final Design 1 8 2000 Graduate Student Design 3 1 Arm Mounted on Wheelchair Frame 3 2 Home Position 3 3 Reaching Below the Floor 3 4 Same Position but Different Orientation 3 5 Exploded View of Arm 3 6 Shoulder Twist and Bend 3 7 Elbow Bend 3 8 Wrist Twist and Bend 4 1 Rear Mounting Bracket 42 Front and Rear Mounting Brackets 4 3 Stationary Plate 44 Complete Base with Motors 4 5 Cutaway of Stationary Plate and Bearing Assembly 4 6 Shoulder Twist Motor LIST OF TABLES AND FIGURES 4 7 4 8 4 10 4 11 4 12 4 13 4 14 4 15 4 16 New Stationary Plate Bottom of Stationary Plate with and without the Twist Motor New Rear Mounting Bracket with Twist Motor New Base Assembly Drive Pulley Twist Pulley Cutaway of New Twist Bearing Assembly Cutaway and Exploded View of Bearing Assembly Complete Redesign Base Assembly Exploded View of Base Assembly Shoulder Bend Motor Left and Right Shoulder Mounting Brackets Shoulder Shaft Arm Plate from 1999 2000 Design Lower Arm Tube Preliminary and Final Shoulde
104. ow Joint Manufacturing the two part elbow shaft design with steel is not possible because steel cannot be machined to the tolerances required for the design Therefore the two part elbow shaft is replaced with a solid 0 25 inch diameter shaft The repositioned motor requires the elbow shaft and worm gear to be offset within the lower arm tube Finally the elbow brackets were revised to still permit the forearm tubing to rest upon the lower arm tubing in the home position Please see Figure AD 2 for the final elbow redesign Figure AD 2 Final Redesigned Elbow 158 While the design changes have not been evaluated for their feasibility experimental performance during the final presentation for the 2000 2001 arm indicates that the modifications will adequately support the forearm Since bearing support and the resizing of the elbow brackets were the only major modifications the modifications will eliminate the flaws and produce a successful design ELBOW REVISION BILL OF MATERIALS Table AD 1 Elbow Revision Bill of Materials Description Part Number Vendor Qty 02 5 PIC Design 159 500 185 71 48 250 ex R 250 LEFT ELBOW BRACKET WRIGHT STATE UNIVERSITY REVISED 6 7 2001 ERIC 160 44 8 250 E 185 a 500 24 250 er RIGHT ELBOW BRACKET WRIGHT TATE UNIVERSITY REVISED 647 200 ERIC WeHES 161
105. r All of the H bridges were made the same for simplicity sake and all of the components were chosen to be able to handle the largest motor of the lot All of the H bridges have one 3 pin input receptacle two lines to the motor a supply voltage line and a ground line This 3 pin input receives and enable signal white a reverse PWM signal green and forward PWM signal red These signals are all passed to the H bridges from the control module via the router board ENABLE INPUT OUTPUT es FORWARD I ORUUND d i i I i I 5501 I MODE T ENABLE I 3 I p _ REVERSE 1 l n U mo 1 1 1 HHH FORWARD I mia 8 Y REVERSE E zx d Figure 7 Router Board Schematic Both the router board and the H bridges were made in a similar manner using perf boards and a silver pen A perf board is a silicon board like a standard PCB with 0 1 spaced holes pre drilled in it Each hole has its own copper pad to which things can be soldered To make a this way you simply solder your components place and connect everything by drawing 101 Figure E S Router Board Top left 1 Ranier Board Bottom
106. r Joint Designs Exploded View of Shoulder Joint Preliminary Elbow Motor without Encoder Elbow Motor Mounting Block Elbow Motor Assembly Elbow Motor Mounted at Angle Elbow Shaft and Gear Assembly 2000 Graduate Student Elbow Bracket New Elbow Bracket Preliminary Forearm Tubing with Offset Wrist Motors Final Forearm Tubing with Parallel Wrist Motors Wrist Pillow Block and Bearing Wrist Differential Fearon s Final Gripper Design Fearon s Solid Edge Model Closing Sequence of Underactuated System 1999 2000 Final Assembly Comparison of Old New Motors Motor and Differential Plates Physical Closing Sequence of the Gripper Final Solid Edge Gripper Assembly Firestorm 18V Drill Motor and Accessories Motor Test Stand Shaft Test Setup Magtrol HD 710 Power Input Wiring to Magtrol HD 710 Damaged Dynamometer Control Board Drill Motor Torque Speed Curve 9 1 1 2 1 E l E2 E 3 4 E 5 E 6 E7 E 8 10 E 11 ET E 13 OOPic microcontroller Schematic for Torque Calculations Planetary Gear Train Free Body Diagram Kinematic Model Control Box Amplifier Box Basic Signal Flow Chart Control Circuit Schematic Control Module Circuit Top Control Module Circuit Bottom Router Board Schematic Router Board Top amp Bottom H bridge Top amp Bottom H bridge Schematic Resistor Block Control Box Amplifier Box H bridge Stack Amplifier Box Internal Power Connector AD 1 Cutaway V
107. r interesting masking technique the one used to make the control board it the use of toner transfer paper This paper when printed on or photocopied on with some sort of a toner based printing system such as a laser printer can be ironed onto the copper clad board just like an iron on transfer for a t shirt The board and paper are then put into water to dissolve the adhesive that holds the toner to the paper just like a decal The board is then etched and drilled 107 E 7 CODE The following code was use for the alternative controls system This code 1 to be used with the Microchip Inc 6 84 only The code is for both PICs used in the alternative controls system E 7 1 HEADER FILE LOF84REG INC ifndef 16C84REG INC 0p SPEGTAL REGISTERS INDF equ TMRO equ Pol equ STATUS FSR equ PORTA equ equ FEDATA FEADR equ PCLATH INTCON OPTION REG TRISA TRISB EECON2 RAM_BASE sf STATUS REGISTER 0x00 0 02 equ 0 04 0 05 0 06 equ 0x09 equ equ equ equ equ G equ 0 DC equ 1 2 PD equ 3 TO equ 4 RPO equ 5 RP 1 equ 6 equ PORTA Bits equ 0 equ 1 0 03 0 08 81h 85h 86h equ equ equ PIC 16184 Registers 1 88h 89h Och 108 2 equ 2 RA3 equ RA4 equ 4 TOCKI equ 4 PORTB Bits RBO equ 0 INT equ 0 RB1 equ 1 RB2 equ 2 RB3 equ 3 RB4
108. racteristics These configurations include series wound shunt wound and compound wound The series wound motor is useful for low speed high torque applications but has high starting torque and poor speed regulation The shunt wound motor is available in both long and short varieties and provides good speed regulation and a flat torque speed curve The compound motor is a combination of a series wound motor and a shunt wound motor It provides a compromise by offering better starting torque than the shunt motor but less accurate speed regulation The DC motor obtained from a cordless drill 1 most likely a series wound DC motor 8 2 MOTOR TESTING 51 The previous arm designs utilized expensive specialized actuators and gearboxes to provide movement A main goal for this year s design was to reduce cost Consumer products afford the cost savings of mass production After researching the available cordless products the cordless drill had the best combination of a motor torque rating and gearbox The drills researched ranged in price from 100 to 200 and the advertised torque ranged from 210 to 500 in lb In the interest of time saving a baseline inexpensive drill would be tested with future purchases based on how well the test drill faired The test drill purchased was a Black amp Decker 18 0 Volt Firestorm Model HP932K 2 The drill was disassembled and the key parts are shown in Figure 8 1 The main components are the motor speed control
109. rake With motor running raise current being supplied to the brake slowly until the torque level reaches the stall torque of the brake Record this value 10 Connect oscilloscope to encoder output See Magtrol Dynamometers user s manual for details This frequency observed corresponds to the speed of the motor 11 Connect voltmeter to torque output See Magtrol Dynamometers user s manual for details The voltage observed corresponds to the torque 100 mV N m 115 12 Connect voltmeter to voltage output on motor 13 Connect ammeter to current output on motor 14 Set brake current to value found step 9 15 Start motor 16 Record necessary data 17 Reduce brake current 18 Record data 19 Repeat steps 15 and 16 until brake current 1 zero 20 Create plot using gathered data Data The following table should be used to record data obtained from motor testing Table F 1 Motor x Data Table Braking Current Motor Speed rpm Torque Drill Voltage V Drill Current A 116 APPENDIX PART DRAWINGS This section contains all of the two dimensional part drawings for the robotic manipulator 117 en 250 LEFT ELBOW BRACKET WRICHT STATE UNIVERSITY REVISED 4 1 2001 118 RIGHT ELBOW BRACKET WRIGHT STATE UNIVERSITY REVISED 44 200 ERIC Yu 119 L 120 Lu Jos ERIC
110. requires very little gearing to achieve the high torque low rpm goal This motor also features a right Figure 5 1 Shoulder Bend Motor angle gearhead which makes transferring the power to the shoulder joint simple and eliminates the need for expensive bevel or miter gears The power from the motor 15 transferred through set of spur gears to reduce the speed by a 3 1 ratio The chosen gears were generously donated by the Boston Gear Corporation The gears are made of steel and have a diametral pitch of 20 and a face width of 0 5 inches The shoulder joint pinion and gear are Boston s YA20 1 2 and YA60A models respectively The pinion has a pitch diameter of 1 inch and a bore of 0 5 inches The gear has a pitch diameter of 3 inches and a bore of 0 5 inches 29 save space and weight the shoulder joint the gear was incorporated into the design of one of the mounting brackets see Figure 5 2 In previous designs the shoulder brackets served only to support the arm to mount the arm to the base and to provide a location to place the shoulder bearings In the new design one of the brackets also serves as part of the shoulder Figure 5 2 Left and Right Shoulder Mounting Brackets joint gearing By modifying the shoulder gear the new design allows the gear to accommodate a bearing for the shoulder shaft and be mounted to a small aluminum block to create a bracket that 15 equivalent In size to the previous desi
111. rew on the other end This allowed for the use of a coupler to join the two pieces Because the 1999 2000 gripper design was used as our baseline the components discussed above were the only parts that were fabricated from scratch The remaining parts of the gripper were re assembled and re used with the exception of the finger pieces These pieces were also re machined from aluminum instead of Lexan for consistency of appearance The mounting and dynamics of the gripper are identical to those of last year s design The gripper is mounted to the wrist differential by fastening the bottom plate by using six screws the differential plate can be seen above in Figure 7 6 This plate also serves as protection for the motor itself Because of interference between the gripper and the forearm housing a spacer had to be added to create a slight separation 46 The gripper 15 again actuated through the use of a leadscrew that moves plate that slides in slots milled into the side plates The motion of this plate 1s what begins the movement of the six bar underactuated linkage The upward motion of the plate forces the fingers of the gripper to close until they make contact with an object After contact has been made the joints the fingers will bend The bending of the joints 1s what allows for the fingers to wrap around an object to create a firm hold This idea is shown in Figure 7 3 but can also be seen in the closing sequence of the assem
112. rigid another reason for choosing aluminum over Lexan Also the old plate that housed the motor had to be modified and moved in order to support the new motor gearbox assembly This support plate fits 45 around the Johnson Electric motor and Gearbox rests against the bottom of the plastic 1 i Motor Plat housing of the gearbox This plate was then fixed into place using epoxy Fastening using epoxy was not the desired way to attach the motor to the plate but was used due to time constraints at the end of the project The original idea was to weld the support plate to the motor housing A gas Figure 7 6 Motor and Differential Plate tungsten arc weld or Tig weld would have been used because of the thin housing of the motor and the joining of aluminum parts The motor gearbox assembly was not modified as it came from the screwdriver The only modification was that the fixture on the end of the gearbox output shaft was reduced in length by of an inch Originally a screwdriver was disected to see what parts could be eliminated and how the output of the gearbox could be fastened to the leadscrew These screwdrivers were difficult to disassemble without damaging the components Therefore the decision was made to make use of the hexagonal shaped slot that was already part of the screwdriver A special coupling was machined with a hexagonal insert on one side and an identical match to the leadsc
113. ripper A production cost was not calculated for the 1999 2000 gripper but would have been even lower due to the reduced cost of machining 7 2 DESIGN AND ACTUATION This year s design team used most of the design from the 1999 2000 team but made several modifications to increase durability and ease of control The first design change was that we eliminated the use of Lexan We believed that this material was used more for visual purposes than for the mechanical benefits In examining the old gripper cracks had formed around the areas of high stress caused by the screws The Lexan is also much more flexible which allowed the side plates to bend and twist fairly easily Aluminum was chosen to replace 44 these parts Aluminum made the frame much more stiff and durable without adding very much weight The cost of aluminum 15 also quite low and it is easier to machine than Lexan Another design change came with the use of a cordless screwdriver motor The motor chosen was a 2 4 VDC Johnson Electric and was found in a Black and Decker Model 9072 screwdriver The screwdriver itself came with a gearbox that was also used in the new design The planetary gearing used provided an output of 20 in lbs of torque at 150 rpm This was very appropriate for our application because it allowed for the gripper to move its full range of motion in about 4 8 seconds There was Figure 7 5 Comparison of old new motors also a cost ben
114. roximation This means that the converter determines one bit at a time starting with the most significant bit bit 7 and working down to the least significant bit bit 0 The process takes about 30 us to complete after which it is ready to output the result The ADC requires two lines to control The first line tells the ADC to start the conversion the CS line The second line supplies the digital clock signal to the ADC in order to get it to send its output one bit at a time The ADC also receives a high and low reference voltage to which to compare its analog voltage input When the conversion process starts the ADC takes a snapshot of the input analog voltage and compares it to the reference voltages It then outputs an 8 bit number between 0 and 255 that indicates where the analog voltage is between the reference voltages For example if the input is the same as the low reference voltage then the output would be a binary 0 If the input is half way between the low and high references then the output would be 127 half of 255 This give the microcontroller a way to understand where the joystick 1s positioned Before explaining how this circuit works as whole it 15 important to understand the process of pulse width modulation and how the microcontroller counts time The speed of a DC motor is controlled by voltage provided that the motor is not approaching its stall current The spinning of the motor generates a backwards voltage or back EMF
115. ry capability 15 limited only by the power source The PWM signal consists of two parts the off time or off cycle and the on time or duty cycle The off cycle is the period of time that the pulse is off while the duty cycle is the period of time that the pulse is on To determine the effective voltage being generated by the PWM the formula is off cycle duty cycle eff sup ply 0 off For the PIC microcontroller to generate a PWM signal it has to have a way of counting on time and off time This is done by a process called timer interrupt counting The timer in the microcontroller is an 8 bit register that counts clock cycle the internal ones not the external ones After the timer has counted 256 cycles 51 2 us total 1 overflows back to zero and calls an interrupt When the interrupt 15 called the PIC drops whatever it was doing and goes to the interrupt handler in the code to find out what to do After the interrupt instructions have been carried out the PIC returns to where it was before 1t was interrupted and continues on This allows for a rudimentary form of multitasking because the PIC can be busy doing other things 99 while it counts off time To generate PWM signal the PIC as a set off time of 50 timer interrupts and it varies the on time between 10 and 110 timer interrupts depending on the position of the joystick How the entire circuit works together to control the arm is exp
116. s which were not strong enough to easily move the arm were selected Its overall size complicated manufacturing requirements and the predicted high maintenance costs also hindered the design Figure 1 1 1996 1997 Final Design 1 3 1997 1998 DESIGN A solid model of a design concept proposed by the 1997 1998 Ohio State team 15 seen in Figure 1 2 This design featured 5 degrees of freedom with vertical motion controlled by a lead screw and horizontal motion accomplished by a motor at the elbow joint This design differed significantly from the first design in that no cables or pulleys were used because the motors are mounted directly on each joint motor on joint control Lower assembly and maintenance costs were the primary advantages of this design over the cable and pulley system Another feature of this design 15 that it utilized more off the shelf parts drawback however was that the required off the shelf components were quite expensive It was also questionable whether the structure of the arm was rigid enough to support the torque produced by both the object to be picked up and the motors required to manipulate the arm Figure 1 2 Ohio State s 1997 1998 Initial Design The final 1997 1998 design pictured in Figures 1 3 and 1 4 incorporated a transmission system similar to that of the previous year s design Some of the improvements included the addition of a knuckle joint and a more compact rotati
117. ss produced consumer goods The cost prohibitive nature of the drill motor adaptation led to a search for alternative motor and gear drive combinations Upon further investigation a supply of surplus gear motors was located These motors operated on 24 VDC providing 325 in lb of torque at 35 rpm These motors utilize a worm drive mated to the electric motor to provide this reduction Figure 8 6 Damaged Dynamometer Control Board 56 JOHNSON ELECTRIC ENGINEERING LTD Project No Production Motor Curve HC785LP 70009 Date 04 27 1999 Model issued by Dept File Mo 70000 xs 5 12 90 Torque m Nm ya Figure 8 7 Drill Motor Torque Speed Curve 100 00 Amp 500 00 Watts 6 22 Dai Shun St Po industrial Estate N T Fax 852 2663 6106 Motor tested rapidly to prevent significant temperature rise At a constant voltage of 18 00 Volts with a circuit resistance 0 000 Ohm At the ambient temperature of 25 30 deg C At No Load Speed 21403 Rpm Current 2 150 Amp At Stall Extrapolated Torque 802 51 m Nm Current 98 57 Amp At Maximum Efficiency Efficiency 76 91 Torque 104 33 m Nm Speed 18621 Rpm Current 14 69 Amp Output 203 30 Watts At Maximum Power Torque 401 26 m Nm Speed 10702 Rpm Current 50 36 Amp Output 449 38 Watts Characteristics Torque Constant 8 3230 m Nm Amp Dy Resistance 0 1830 Ohms Motor Regulation 26 6700 Rpm m Nm
118. tently twisting the arm into a person standing close to the wheelchair 24 The drive pulley is very intricate piece that is machined from solid aluminum see Figure 4 11 The lower side of the pulley 15 designed to slide over the existing 0 562 inch square drive of the twist motor and is held in place by a Figure 4 11 Drive Pulley setscrew The upper section is a 0 75 inch diameter cylinder with a 0 125 inch radius groove cut around it The groove creates a 4 50 inch diameter pulley that accepts a 0 25 inch diameter belt The twist pulley see Figure 4 12 is similar to the twist gear originally designed in that it still supports the shoulder mounting brackets and therefore the rest of the arm The pulley 15 4 75 inch diameter circular plate of aluminum that is 0 4275 inches thick A 0 125 inch radius groove Is cut around it to create 4 50 inch diameter pulley that accepts 0 25 inch diameter belt Since the bend motor 15 now in the lower arm the hole for the bearing and motor shaft have been eliminated The twist pulley is mounted to the center twist bearing piece which is part of a 3 piece bearing design and will be discussed later The twist belt is a black neoprene belt Figure 4 12 Twist Pulley with a round cross section The diameter of the cross section of the belt is 0 25 inches and thus matches the grooves in the two pulleys The belt has a length of 399 mm This redesign accommodates the twist motor
119. testing troubleshooting and Chris Fearon and Fred Griesemer for their continued research support and guidance Carl Kollar of Diverse Electronic Services was also a great help in choosing electrical hardware and offering advice A special thanks also goes out to Gary Gardener and Keith Rogers for their project saving machining in the waning hours of the final presentation The Wright State team would like to thank Greg Wilt for his technical support Dennis Hance for his solid modeling expertise and Tim Leger and Sean Mortara for establishing an efficient computer network for which to develop the solid modeling Special thanks to Dr Nathan Klingbeil for his advice in mechanical matters and Byron Formwalt from Wright Patterson Air Force Base for his electrical engineering insight Last but not least we thank Boston Gear MA for generously donating the shoulder and elbow gears Motoman Dayton OH for their plant tour and advice on robotic arm development and KEA Components for donating an analog mini joystick ii TABLE OF CONTENTS E II ACKNOWLEDGEMENITS HHI TABLE LIST OF TABLES AND FIGURES VIT a0
120. unting to the wheelchair remained difficult due to the two piece clamping brackets Figure 1 8 2000 Graduate Student Design 1 7 MARKET ANALYSIS Rehabilitation robotics in general was studied by conducting literature and patent searches The patent search did not reveal any existing patents for rehabilitation robotics in particular but some specific components have been patented such as grippers The literature search revealed many interesting facts about the state of rehabilitation robotics today These findings are discussed the following paragraphs The market for such assistive robotic products was found to be somewhat limited as robots are an alternative only for individuals who may have a deficiency in manipulation ability Only about 10 of the population has some sort of handicap and much less have both lower and upper body mobility impairments The simple fact that only those individuals with both upper and lower body handicaps will use the product limits the market Those benefiting from a robotic arm must also not be so severely handicapped that they cannot reasonably control a joystick or other input device further limiting the market Therefore it is estimated that of the approximately 1 5 million people who are confined to electric wheelchairs in the United States between 100 000 and 500 000 could benefit from a robotic arm based on the type and extent of their disability These numbers indicate that there is a market
121. utilized DC motors since the wheelchair already had this power source available AC motors however presented some enticing options specific to the project that merited further investigation One such advantage is the wide spread use of AC motors in industry This is a two fold benefit the cost of these motors should be lower for a given set of characteristics over a similar DC motor due to their widespread use second any control specific questions might more easily be answered by those in industry since they are more likely to have faced similar questions their own designs Another advantage AC motors have over DC motors relates to the power consumption of the motor at its different states As pointed out by Michael Ondrasek an engineer from Motoman AC motors require the least amount of power in a holding position and the most power while running at full speed the case of DC motors the most power is required at stall and the least power is required while running at full speed The AC motor characteristic seems to better suit the needs of the Gateway design since the arm will spend a good deal of time in a rest or stationary holding position The AC motors are not without their drawbacks however Two main concerns are the need for an inverter and the speed torque characteristics of AC motors First the need for an AC inverter to convert the available 24 VDC to 120 VAC was investigated Since DC drill motors were being considered as a cost
122. y of the OOPic This advantage is important for simultaneous kinematic calculation while changing multiple motor speeds The OOPic also allows for more operations per second thru the use of virtual circuits and twice the I O line capacity of the Basic Stamp The OOPic has an internal Analog to Digital converter providing potentiometer input for closed loop feedback control This A2D converter operates at 8 bits providing 256 divisions over the range of measurement Additionally the OOPic provides networking capability between Figure 9 1 OOPic microcontroller multiple controllers Furthermore the OOPic was 58 among the microcontrollers proposed a future suggestion by previous Gateway teams Finally the OOPic compiler supports Basic C and Java languages while the Stamp has a proprietary PBASIC command set and Basic programming were among the skill sets possessed by several group members Figure 9 1 shows the OOPic microcontroller 9 2 MOTOR CONTROLLER In past years Gateway groups had designed and built special motor controllers to fit their needs This specialization was necessary due to the large variety of motors used in previous designs For 2001 one of the goals was to use one type of electric motor for the arm tasks Previous years designs utilized brushless DC motors stepper motors and DC servomotors Each of these motors required different programming and hardware schemes This assortment of motors ad
123. y the aforementioned bearing The drive gear was a 2 00 diameter simple spur gear that had a face width of 0 3125 inches Rotational motion was Figure 4 4 Complete Base with Motors transferred to the large twist gear via the drive gear The large twist gear was a major component in the operation of the arm The gear was a 5 00 inch diameter spur gear with a 0 3125 inch face width The diameter was limited to 5 00 inches because of the width of the stationary plate A larger diameter gear would hang over the edges of the plate and cause operational and safety issues The importance of the twist gear was that the rest of the arm was mounted to it The shoulder mounting brackets would sit on top of the twist gear and be bolted from the bottom of the gear 20 Attached to the bottom of the twist gear was the shoulder motor assembly see Figure 4 5 This assembly consisted of a spacer motor mounting plate and motor The 0 75 inch Shoulder Bend Pinion Twist Drive Gear Shaft Bearings Twist Gear Figure 4 5 Cutaway of Stationary Plate and Bearing Assembly aluminum spacer was to sit inside the twist bearing and support all of the moment experienced by the joint 0 25 inch plate which served as a mounting plate for the motor was located at the bottom of the of the spacer plate The motor was bolted to this plate in a vertical position This design allowed the shoulder bend motor to twist with the rest of the arm To bend the
124. zontal or vertical at a constant speed This would have incorporated equations 7 11 and 12 from the Kinematics section of the Appendix to regulate the motor speeds Using this control scheme would have required some way to measure the joint angles such as encoders or potentiometers This control strategy was not implemented due to time limitations Another strategy for controlling the arm lies at the other extreme for ease of operation for the user This strategy is to control each degree of freedom individually No speed control would be necessary and therefore there would be no need to measure the joint angles This control strategy was developed initially and maintained as backup However more user friendly strategy was implemented the final arm The control strategy used in the final design of the arm is a combination of the single motor control without speed control and the strategy of controlling three degrees of freedom simultaneously using speed control The team designed the controls similar to the single motor control structure However the ability to control two degrees of freedom was added Using relationship 7 in the Kinematics section of the Appendix the control code was structured so that the wrist bend would always function unison with either the shoulder bend or the elbow bend This is a great benefit over simply controlling each degree of freedom individually because it forces the grippers orientation wit

Final Report on Robotic Manipulator Project

Contents

Download Pdf Manuals

Related Search

Related Contents