Home

Design and Implementation Process for Controls - e

image

Contents

1. 240V z Brusa AG v 9919 AWG 8 V 80A AWG 8 Underside ESS Brusa AC INE Lives line 1 77 41772 600 ___ 240V 1 cond AWG 1 V 400A Brusa 01 AWG 80A Atine N Neutral p pen gt uv 0 cells 2 wires 4 iss ee ne 4 v Front ESS HV Control Board 176V 42cells Discharge AWG 1 we p S Tyco EV200 600V 400A em Precharge 1600hms lant 600V 100W E 2 5 pa soov 2 aN XC MDC ___ AAN E s 12 lana ADD E pa v Test Point gt ne Der DON Sure ee cnaige 0 U I c Tyco LEVI00 HU OE n Brusa DC 1 wire AWG 8 MS 4 06 p GFI aun sine m Or T rusa 5 Na _ EN e D 22 DC DC 2 M IU Figure 36 High Voltage Schematic Chapter 4 82 The 52 cell underside battery pack is connected to the one underhood by 1 gauge wire A normally open Tyco EV200 contactor is connected to the underside negative terminal which is also the most negative line of both battery packs combined Therefore this contactor is considered the main negative contac
2. 65 EPON AETA OT A AAE eem EEE d E 66 Chapter 4 Electrical and Control System Integration 67 EIUS 67 4 2 Conversion Methodology 0 0 ee aeter d et Re Si pe Se e ar dd gae 67 4 3 Electrome Control nti posa tese ti 69 2 3 MCMV BUS asd eta x be Me Wee 69 BCU change So ete boa e i e ee Heine tenu Coane Stag NO eu i e is tase 73 2 4 a 78 4 4 1 Powertrain Control 880 78 4 4 2 Discharge Control Strategy idee e e Pe ER ESTO ERE BRE TRACES 8l 4 4 3 High voltage schematic 82 2 5 Control LAW a esiti pte oes pota hi nita ine S tint E e e tip 84 4 5 1 Battery Management 85 453 0n Board harser Module 89 4 5 3 Motor COMO MET e atit or co pr 92 4 6 Safety Control Strategy 94 4 6 1 Acceleration Fault 44 4 22202 00000 0000 94 2 6 2 94 4 6 3 Voltage Current and Temperature Management 95 4 6 4 High Voltage Interloe k 95 4 6 5 Ground Fault Interrupt and Inertia 5 96 4 7 On board Debugging and User 97 4 7 1 Driver Notifications oie REG e HY RAR
3. AC DC ACK ACCM Ah ALDL ANL AOS APM Avg BCM BEV BMS BPCM Byte CAN CANH CANL CARB CD CGM CIDI CIS CRC DC DC AC DC DC DE DLC DLC DOD List of Abbreviations Ampere Air Conditioning Alternative Current Accessory AC to AC Converter AC to DC Converter Acknowledgement Air Conditioning Compressor Module Amp hour Assembly Line Diagnostic Link Argonne National Laboratory Automatic Occupant Sensor Auxiliary Power Module Average binary Body Control Module Battery Electric Vehicle Battery Management System Battery Pack Control Module byte 8 bits Controller Area Network CAN High CAN Low California Air Resources Board Charge depleting Communication Gateway Module Compression ignition direct injection Chassis Inertial Sensor Cyclic Redundancy Check decimal Direct Current DC to AC Converter DC to DC Converter Discrete Event Data Length Code CAN bus Data Link Connector OBD II Depth of Discharge xvili DSML DTC E EBCM ECC ECM EcoCAR ECU EMI EOF EPA EPS EREV ESD ESS EV FCV F FE FFEV FSCM FSM FTP GCC GHI GHG GM GMLAN GUI HCP HEV HIL HS HV HVCM HW HWFET Department of Energy Domain Specific Modeling Language Diagnostic Trouble Code Electronic Brake Control Module Electronic Climate Control Engine Control Module EcoCAR The NeXt Challenge inter university competition Electronic Control Unit Electromagnetic I
4. sss 119 2005 23 On Board Carvers 121 5 6 2 4 Datalogger and Instrument Cluster Screen 122 BET P c PR Xr 123 Chapter 6 Tests and Results 124 6 1 Introduction 124 6 2 Validation Calibration and Testing essere 124 6 2 1 Controls Development 124 6 2 2 Performance Vesting at o 128 6 2 3 Performance Testing at Year 2 129 6 2 4 Performance Testing at the Environmental Protection Agency 131 Zo Day Cn QU RA 131 0 2 4 2 Day 2 de OER ot de du mo 132 QUA E MS A 133 6 2 5 Vehicle Integration at Year 3 Competition eese 133 6 3 EcoCAR Project Results eise AE RIVE EXE RE VEA PER RE 134 6 3 1 Competition Result Overview 222000 0 nennen 134 6 3 2 Environmental Protection Agency Testing Results 135 o 3 Day D mE ones 135 Day and e Ra de MR es M 136 6 3 2 3 Ground Fault 155 6 5 139 6 3 2 4 140 6 3 2 5 SUMMA Y iiio HORRORE E SATA Med Sd RS NO Gane 141
5. spend considerable amounts of money to develop environmentally friendly alternative propulsion technologies It is projected that hybrid electric vehicles HEV and battery electric vehicles BEV will share 30 of the passenger vehicle market by 2020 2 Below is a list of historic and current production electric vehicles from the late 90s to 2014 e GM EVI and S10 EV Toyota RAV4 EV e Ford Ranger EV e Chrysler and EPIC Mini van e Tesla Roadster and Model S e Nissan Leaf e GM Chevrolet Volt and Spark EV GM Cadillac ELR Ford Focus BEV Mitsubishi BMW 13 city Fiat 500e Honda Fit EV Smart Fortwo EV Chapter 1 2 The method of building electric vehicles be divided in 3 categories a ground up design a factory conversion and a hobbyist conversion Having been engineered from the drawing board explicitly as an electric vehicle the famous EV1 and now Telsa Model and BMW 13 are from the first group The Toyota RAV4 EV the Nissan Leaf and GM Volt are based of existing internal combustion engine ICE vehicle platforms respectively from a RAV4 Nissan Versa and Chevrolet Cruze These can be classified as factory converted EVs as are most other products on the market Hobbyist conversions are normally gasoline or diesel powered vehicles to which an electrical powertrain is added most often using lead acid batteries and DC motors though this is now changing Hobbyist conversions need to r
6. 0200000 206 xi Appendix L UOIT CAN bus Dictionary 210 xii List of Figures Figure 1 UOIT EcoCAR Prototype in the Climatic Wind Tunnel at the Automotive CONGR ob Excellence ae rssh oe voa A Us sap acne rade ce sd 2 Figure 2 Degree of 1 catre dd ii edo tees 7 Figure 3 Series Configuration sii Dat 9 Figure 4 Parallel Configuration 10 Figure 5 Power Split Configuration 10 Figure 6 EV ECUs in the UOIT EcoCAR EV Prototype 9 18 Figure 7 Cell Comparison 10 terret ai 20 Figure 8 Cell Stack including cooling frames and voltage monitoring boards 2l Figure 9 Front ESS b Underside ESS 22 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 600 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Figure 23 Figure 24 Figure 25 Figure 26 Figure 27 Figure 28 Figure 29 Figure 30 Figure 31 Figure 32 Figure 33 Figure 34 Figure 35 Figure 36 Figure 37 Figure 38 Figure 39 Figure 40 Figure 41 Figure 42 BMS14C Ill C 23 In vehicle BMS layout a Front ESS b Underside ESS 24 a Brusa NLG513 3 3kW b 3 of 5 Brusa NLG513 3 3kW Integ
7. Figure 18 Bender IR470LY 20 Chapter 2 33 2 7 5 4 Custom instrument cluster screen Some off the shelf screens were first considered for integration to the dashboard near the instrument panel cluster IPC such as the Woodward embedded display series the RM Display 1001 and even an upgraded version of the navigation system of the vehicle However having a customizable screen able to display a broad variety of information being read from the CAN bus was a priority That s why the team selected an in house LCD screen and controller designed implemented and programmed in C code by the author Furthermore the small size of the screen allowed it to be integrated directly in the IPC Examples of information displayed are the battery voltage current temperature state of charge SOC errors and warnings It was also planned to display the car ready signal and GFI status on it but re using the chime and lights of the IPC was successfully accomplished and more intuitive for the driver This screen as it was being programmed on a test bench before integration in the IPC is shown in Figure 19 Custom Display Screen Chapter 2 34 2 7 6 Summary of Selected Subsystems Table 13 lists the key devices selected to implement UOIT s new architecture into the 2009 Saturn VUE Table 13 Main EV subsystems selected Components Models Battery Cell 94 Kokam SLPB 160460330 Battery Manageme
8. Unpack Software CAN Unpack Block Reference Vehicle Network Toolbox Library CAN Communication Matlab 2009b 49 http www race technology com wiki index php CANInterface B yteOrdering visited 13th December 2011 50 SAE Vehicle Architecture For Data Communication Standards Website http www sae org servlets works documentHome do comtID TEVEES 12 visited on 30th March 2012 51 National Instruments visited in February 2011 website http www ni com pxi Website 52 Wiring Controlsoft RS 485 networks KeyMaster Systems Hardware Installation Guidelines Controlsoft 2002 retrieved from www powerrichsystem com Downloads RS 485wiringNetworks pdf 53 ISO Committee Draft Road vehicles Controller area network CAN Part 1 Data link layer and physical signalling ISO CD 11898 1 ISO TC 225 3 October 1999 54 ISO Committee Draft Road vehicles Controller area network CAN Part 2 High speed medium access unit ISO CD 11898 2 ISO TC 225 3 September 1999 55 ISO Committee Draft Road vehicles Controller area network CAN Part 3 Fault tolerant medium access unit ISO CD 11898 3 ISO TC 22SC 3 2001 06 07 56 Wikipedia On board diagnostics Website http www en wikipedia org wiki On board diagnostics visited on 28th March 2012 57 Microchip Section 21 Enhanced Controller Area Network datasheet Microchi
9. 1 0 1 0 Gnd Sensor 1 1 1 1 Park Neutral 1 0 1 0 Several modeling techniques such as look up tables and Stateflow diagrams could have been used to program the transmission decoder in the VIM however simply using model based logic gates was judged adequate for this task Figure 86 in Appendix J shows the memory access blocks of the A B C and D inputs filtered by logic gates generating the park reverse neutral and drive output signals These are used to update the motor controller inputs pins 7 8 and 20 of Figure 39 respectively DRIVE REVERSE and PARKING and dynamically emulate GMLAN signals It should be noted that for simplicity reasons the VIM ignores the ground signal and the park neutral switch signals represented at the bottom of Figure 43 in the transmission decoder logic degrading the robustness of the input redundancy verification during the decoding process However these decoded values control the motor controller but do not control the gearbox The gears are mechanically unlocked by the shift lever This additional logic was omitted in the program for time saving reasons and does not affect the prototype functionality however should be programmed in order to use the transmission encoder to its full potential It would also create a safer vehicle by reducing the risk of having a mismatch between the mechanical lock released in the transmission and the condition the inverter believes the vehicle is in In addition to t
10. 27 Schultz Katrina GMLAN In Vehicle Communication Networks Presentation General Motors 1 October 2010 28 Linton Becky Communication for NI Dynamic Testing Software Hardware in the Loop Presentation National Instruments 29 http www can cia org visited on 13th December 2011 30 CAN in Automation Controller Area Network CAN Website http www can cia org index php id systemdesign can 31 CAN in Automation CAN physical layer Website http www can cia org index php id systemdesign can physicallayer 32 CAN in Automation CAN history Website http www can cia org index php id systemdesign can history 33 CAN in Automation CANopen Website http www can cia org index php id systemdesign canopen 34 CAN in Automation DeviceNet Website http www can cia org index php id 48 35 CAN in Automation CAN made easy Basic information on the CAN physical and data link layer CiA retrieved December 13th 2011 from www can cia org pg can additional can_basics_print pdf 36 Passemard Michel Atmel Microcontrollers for Controller area Network CAN Atmel Corporation 4069A CAN 02 04 http www atmel com Website 37 National Instruments Automotive Communication Bus Overview Website http zone ni com devzone cda tut p 1d 3352 visited 28th March 2012 38 FlexRayTM About FlexRay Website http w
11. Charger GFI Datalogger ICS Chiller Heater The baseline is a 2 mode hybrid this constitutes an internal combustion engine and 2 electric motors which are mechanically coupled through a planetary gear box controlled by the TCM As explained in Section 2 7 an induction motor controlled by a power inverter controller propels the converted vehicle This not only simplifies the powertrain but also renders the stock components obsolete for the converted vehicle Therefore the entire powertrain was removed along with its controllers i e ECM TCM and TPIM Chapter 4 74 For energy storage the hybrid utilized 63 22 L fuel tank and 2 3 kWh NiMH battery pack These are managed by a fuel system control module FSCM and a battery pack control module respectively Both systems were replaced by a 83 5 kWh lithium polymer battery pack and a battery management system BMS A vehicle integration module VIM including a high voltage control module HVCM and supervisory control module SCM was also added to fulfill the tasks executed by the TPIM and not covered by the motor controller and BMS respectively The controls the high voltage contactors whereas the VIM manages the vehicle integration In a hybrid the battery pack is charged by a combination of regenerative braking and engine power An on board charger is added as the primary charging device for the electric vehicle Regen
12. IDE is Dominant Logical SID1 SIDO 5 is Dominant Logical is Dominant Logical 0 11 bit Identifier Figure 53 Standard data frame 57 Arbitration Control Data one End of Field Field Field Frame _ IDENTIFIER IDENTIFIER 11 Bits 18 BITS 2 IDE is Recessive Logical 1 E SRR is Dominant Logical 07 RTR is Dominant Logical is Dominant Logical 2 29 bit identifier RB1 is Dominant Logical 27 Figure 54 Extended data frame 57 n Field ontrol Field lc Field EMEN Field Frame IDENTIFIER 11 Bits Recessive Logical 1 is Dominant Logical 27 11 bit Identifier Figure 55 Standard remote frame 57 End of Arbitration Field Control Field CRC Field ACK Field Frame IDENTIFIER IDENTIFIER 11 Bas 18 Bits i IDE is Recessive Logical 17 4 1 SRR is Recessive Logical 1 Jeo ee mare tog is Dominant Logical 0 EEE RB1 is Dominant Logical 0 29 bit Identifier Figure 56 Extended remote frame 57 Appendix B 163 5 Error frame An error frame has the two following fields error flag and error delimiter 22 The active error flag 6 dominant bits and the passive error flag 6 recessive bits are the 2 existing flags used 2
13. Specifications for Motorola Forward Isb Start bit Isb 11 Length 9 bits Specifications for Motorola Backward Start bit Isb 51 Length 9 bits For the most significant part of the signal the start bit is always the first bit of the following byte and the length is the remainder of the total length not used by the least significant part of the signal Thus a detailed expression of the specifications for each display format is Chapter 3 53 Specifications for Motorola Forward lsb Start bit Isb 16 Length 4 bits Start bit Isb 11 Length 5 bits Specifications for Motorola Backward Start bit Isb 40 Length 4 bits Start bit Isb 51 Length 5 bits The graphical representation of the conversion from Motorola Forward 156 and from Motorola Backward can be found in Appendix F 3 5 3 Message and Signal Definitions The difference between messages and signals was established in Section 3 5 1 This section describes in detail both messages and signals and how to filter them in order to receive and send CAN messages 3 5 3 1 Message Definition A CAN message is defined by the following characteristics e Message name e ID D mask DLC payload size Payload filter payload value e Payload mask e Periodic interval rate in ms Chapter 3 54 The message name is name of the message associated with a data frame remote frame meeting the filtering characteristics
14. 6 3 3 Vehicle Technical Specification 143 6 3 4 Consumer Acceptability SG RR 144 6 4 Thesis bed eda 145 esed 146 Chapter 7 ConelusiOmi adde OR SOARES 147 TI Thesis sce Pe pu 147 7 2 Conclusions ende dedu De 147 2 2 Original Contributions 150 TA MR COMMMETIG AU 10 151 TAT EFut re 151 qui Best 152 7 2 COVCIUSION M 154 deco 155 Appendix A Detailed Performance Simulation Results eese 159 Appendix B Additional CAN Theory 161 Appendix C CAN Physical Waveforms nennen enne 166 Appendix D Standard CAN eene nennen 168 Appendix EC ANN DransSCelVebs o eie RARE 170 Appendix Conversion of Display Format eene 173 Appendix G Post Office Analog ys 174 Appendix H Comparing LIN CAN ec Ic AGA sued ideas 176 Appendix I Detailed Subsystem 222222 212 0 178 Appendix J Vehicle Integration Module Models eene 194 Appendix Instrument Panel
15. ti pls NOS REOR PERO Deren tad M Lieu anna Rea d iii TERN iv Dedication reenn e ada se E N E ES 5 o dc sass ieee p ha sa so a a ed vi List of Fig fes PEE xiii Tables xvi Last of Abbreviation S sssini tein cott o xviii Chapter I Introduction veio tes reb tope wei lu WI UE 1 IL BOOC AR The Next Challenge 1 1 2 Previous 2 1 3 Goal ot THESIS e 4 1 4 Research Contributions iii b sodas REESE HARI E IE 4 T ssumimar y of THESIS SCCHOMS SML RS ERE EUM POR RADO RI 6 Chapter 2 Architecture Selection 7 ZN Introductio T E 7 2 2 Powertrain Selection and 5 7 2 2 1 Degree Hy bridization eet See tees ined eee 7 2 22 Fecho logie Sage pev out Det wets gece gees Qo p 8 2 2 3 Powertrain Configurations 9 9 DOS M ICI cd TTC 9 2 2 3 9 POW CLASP sy cle AURI Se ICE CUR 10 ZO uel ENE CLIO ase rao a LGR ET 11 2 4 Well to Wheel Influence and Architectures Considered 11 2 5 EV EREV E10 PHEV E10 Performance Simulations 13 2 6 Selected Architecture sitet cb nisi tdi vd da ead cens 15 21 Selected C ONIONS INS d aucto 17 2 To 19 2 7 1 1 Ba
16. 101 Chapter 5 Vehicle Integration Module 5 1 Introduction The current chapter details the controller at the center of the electrical and control system integration strategy described in the previous chapter An overview of the MotoTron controller is presented followed by an explanation of the model based program created The encoders and decoders required for a successful vehicle integration are detailed along with the high voltage controller This chapter concludes by describing how the vehicle integration module links the main ECUs through CAN communication 5 2 Overview of the MotoTron As described in the previous chapter removing several important ECUs from a vehicle requires the addition of new ones It is unlikely that all the new controllers will be compatible with one another or with the remaining stock ones The purpose of a programmable controller like the MotoTron is to integrate all these ECUs together as a functioning system by interconnecting them and rebuilding the dismantled system Its multiple CAN ports allow the MotoTron to communicate with controllers spread throughout different CAN buses and also act as a gateway between them Some CAN signals need to be translated before being sent back on the destination CAN bus while others must be emulated Emulated signals are sometime required to insure proper functioning of original ECUs which expect communication from removed ones Such a controller also has the task of contro
17. 300h Module ID 3FFh 580h Module ID 400h Module ID 500h 600h 618h 630h 650h Module ID 680h 690h Module ID 300 301 1 A 303 303h bo oo CA N W oo 5 gt 5 CA CA NIN Wl ON N Oo ON oo ON ON N BRUSA Tx Datalogger Tx OFEh 61ih 0FAh OFFh OFBh _ OFCh 3041 Appendix L 210 DN ON OD 790 Module ID Appendix L 211
18. 40 47 32 39 24 31 16 23 8 15 0 7 Forward Backward Terminology used to define the bit numbering is discussed next When representing 64 bits over the 8 bytes of the data field the expression sawtooth seems to not only define the numbering within a byte but also the message progression for the forward message progression In a forward sawtooth message the bit numbering increases from right to left while the message progression is from left to right creating discontinuities in the overall bit numbering This expression kind of loses its sense for the big endian byte order since the bit numbering increases from right to left and the message progression is also from right to left creating no discontinuities as implied by the word sawtooth The overall bit numbering of a backward sawtooth message has no discontinuities and looks like a forward sequential message starting the counting from the end of the message Chapter 3 49 Being inconsistent with the continuous and discontinuous overall bit numbering the bit numbering expressions are inadequate to visualize the overall bit numbering This is one of the reasons why it is essential to specify the message progression in a message to avoid confusion from the user when defining the overall bit numbering 3 5 2 4 Start Bit of Signals As shown in Table 20 signals are variables or pieces of data contained by
19. BMS ChrgCurrPct BMS ChrgCurrent 11 BMS ChrgMainCurrent BMS ChrgVoltage BMS DisChrgCurrPct BMS Slave2Frame1 Age BMS Slave2 V1 5 Slave2 V2 BMS Slave2 V3 7 5 Slave2 V4 5 Slave2 V5 1 Slave2 V6 7 BMS Slave2 V7 1 5 Slave2 V8 7 Slave2 V9 5 Slave2 V10 7 5 Slave2 V11 1 Slave2 V12 1 BMS Slave2 V13 1 5 Slave2 1 14 BMS Slave2 Vtot 1333700 1333 00 1160 00 1361200 j000 14208 14233 14248 14277 15852400 5 Slave3 Vtot 13439 i ic Ignition ON 1104 Probe IgnitionPot 5 i Probe Start 1 ParkOut 1 Probe ReverseOut 1 Probe DriveOut Probe NeutralOut 3 1 Charger ActVal AgeC 13514 128 00 Charger OutputCurr 0 00 Charger OutputVoltag 0 00 1 Charger SrcCurr 0 00 1 Charger SrcVoltage 000 188 4145 BMS Slave3 V1 4150 4150 BMS Slave2Framel AgeCount 55 14140 Bms Slaves V3 E 14150 1 14140 14145 14155 14165 14179 BMS Slave3 V4 1 BMS Slave3 V5 BMS Slave3 V6 1 BMS Slave3 V7 BMS Slave3 V8 1 5 Slave3 V9 1 5 Slave3 V10 BMS Slave3 V114 BMS Slave3 V12 1 5 Slave3 V13 BMS Slave3 V14 4155 4135 4145 4155 4165 4169 4199 4208 4233 4248 4267 4194 BMS Master V1 BM S Master V2 BMS Master V3 BMS Master V4 BMS Master V5 BMS Master V6 BMS Master V7 BMS Master V8 BMS Master V9 BMS Master 1 10 BMS Master_V11 BMS Master V12 BMS Master V43 BMS Master V14 BMS Master
20. Moon roof Engine control Fans Non time critical sensors 3 8 1 2 High Speed Applications For applications requiring near real time communications high speed CAN buses are used With the complexity of car controls constantly growing it is not infrequent to find more than one high speed bus in a car Examples of applications using high speed CAN buses are enumerated in Table 31 based on reference 24 Table 31 High speed applications ECU programming Diagnostic interface Engine management Electric motor controller High voltage battery management Adaptive Brake System ABS Body controller Accident avoidance system Fuel system Chapter 3 64 3 8 1 3 Diagnostic Interface Programming There are 2 other key applications of CAN in vehicles 24 e Diagnostic interface e ECU programming Most electronic control units ECUs save diagnostic information that can be sent on a CAN bus as diagnostic trouble codes DTCs to other ECUs and a diagnostic tester The diagnostic tester generally connects to the vehicle s On Board Diagnostic I OBD II port to read DTCs present on the CAN network in order to make a diagnostic Calibration of variables in ECUs and software updates are inevitable when developing a vehicle ECUs can usually be programmed through a CAN bus to simplify these tasks since not every ECU of a vehicle is easily accessi
21. Probe PRNDE i cE EngOilChnglO EngOilPrsLwlO r VehSpdAvgDrvn i VehSpdAvgDrvnSrc __ MehSpdAvghDrvn 2 PPEI_Engine_General_St N A False WA TransOutSpee i HVBatChrgCrdConnlO Valid Probe_ParkOut i Probe ReverseOut _ Probe DriveOut i Probe NeutralOut Valid _ Figure 80 VIM interface Appendix I 185 Figure 81 HVCM interface Appendix I 186 4 TIM600 Motor Controller RS232 Interface Table 69 Enabling Keys P60 95 99 82 Figure 82 Display Log Tab Appendix I 187 Table 70 Parameter list 125Hz 155V Default ara e 4 meter Description Range EcoCAR Value Unit Value Accelerator pedal gain factor for 14 bit analog reference 1 400 INP 1 100 235 6 100 0 Accelerator pedal offset factor for 14 bit analog reference 1 16383 P2 AN_INP_1 100 71998 EN i P3 Gain factor for 14 bit analog reference 2 INP 2 ye v 2 100 0 100 0 4 Offset for 14 bit analog reference 2 INP 2 d e um 400 5 Gain factor for 14 bit analog reference 3 3 100 100 0 100 0 6 Offset for 14 bit analog reference 3 INP 3 T P 8 9 Filter time constant for analog torque reference value 0 20 20 0 0 0 ms P21 CW acceleration time 0 01 199 99 1 00
22. devices and makes the overall design more complex Figure 29 illustrates this topology Battery Motor Supervisory High Voltage On Board Management Controll Control Control Module Charger System Module Module CANH CANH 1200 Controller Controller Controller Controller Controller 1200 Note The resistors can be internally integrated to the farthest devices circuits Figure 29 Daisy chain with short studs Another method of creating a daisy chain is to simply twist two wires together and screw them in place into a single slot of a connector as shown on Figure 30 However these types of connectors are not found in vehicles for reliability reasons Figure 30 Daisy chain with twisted wires in the connector 52 Chapter 3 60 3 7 Commercial Devices Some commercial devices are listed under commercial controllers in Table 17 where CAN transceivers CAN protocol controllers microcontroller with CAN and commercial controllers are enumerated This section focuses not only on commercial controllers but also on commercial devices For the purpose of this document commercial controllers are considered commercial devices but the opposite is not true A commercial device is the hardware portion of a CAN tool and requires a software application to interface and interact with the user It is important to distinguish commercial devices from their software applications since some sophisticated software
23. i e GMLAN SW Therefore despite its higher cost and more complex programming method the more powerful and versatile CANlog4 was selected Table 12 CANlog4 Specifications 19 CAN ports 2 Baby board HS CAN 3 Baby board Single wire CAN 1 Wake up HS CAN 1 External flash card 64 Mb Programming language Dag MUR Programmable LEDs 4 Status LEDs 2 I O expansion boards Yes Chapter 2 32 Figure 17 CANIog4 19 2 7 5 3 Ground Fault Interrupt Bender IRA70LY This unit is connected to the high voltage and low voltage buses to verify that there are no leakage current in the electrical system The unit distinguishes between low and high leakage scenarios by setting flags that send signals to the VIM so that the appropriate actions can be taken If a low leakage current is detected the motor controller can be placed in an emergency mode that limits the output current of the motor and a warning LED is turned on to inform the user If a high leakage current is detected the HVCM in the VIM sends a signal to open the main contactors disabling the motor controller output to stop the vehicle A ground fault interrupter of the Bender brand was mandatory according to the competition requirements They are specifically designed to detect ground faults in floating ungrounded AC systems The Bender IR470LY series 470 20 employed is shown in Figure 18 i 0000000000 LA jJ
24. key conclusion drawn from the testing was that the supply voltage from the temporary battery pack was inadequate for the inverter which was struggling to get much past 200 Vac The inverter requires at least 360 Vac on the DC mains to attain the motor s rated 240 Vrms For this reason the number of cells in the final pack was raised to that planned in the initial design That fact had long been known and was designed into the final pack configuration more than a year prior to these tests Chapter 6 135 During the previous year the team damaged two 510 motors overheating them while doing on road testing To avoid damaging a third motor an infrared temperature sensor was added to the motor housing in order to measuring the actual motor temperature in addition to the windings Therefore the motor temperature was closely monitored while performing the inverter calibration at EPA To the team s delight the motor overheating issues struggled with throughout the previous year appear to have been overcome through a fine tuning of the motor controller calibrations 6 3 2 2 Day 2 and Day 3 A major concern of the EcoCAR organizers was that UOIT s full function electric vehicle would not meet the competition range requirement A complete depletion of the 74 6 kWh temporary Li ion battery was therefore undertaken on the second day at EPA For this test the regenerative braking was deactivated since its calibration was only planned for the nex
25. the bus load is associated with CAN transceivers and depends on internal resistors inside the transceiver and line capacitance whereas the maximum speed and length are dependent on the acknowledgment bit since it has to travel from the receiving ECUS to the transmitting one in a timely manner The field of applications of each of these 3 specifications is broad Section 3 8 1 details applications in the automotive industry To assure robust arbitration when designing a network the bandwidth should ideally be kept below 7096 of the maximum bandwidth 26 Bits travel on the bus using a non return to zero NRZ bit encoding and decoding This means the bit level is maintained over a full bit time and changes at the following bit time if a complementary bit is transmitted 31 Chapter 3 40 For synchronization purposes maximum of 5 consecutive bits of equal logic level is allowed before the insertion of a complementary bit 22 This technique is called bit stuffing As explained previously the 2 bit levels are defined as dominant and recessive and are associated with the logic level in Table 15 Table 15 Bit logic level Bit Logic Level Dominant Recessive 1 The physical waveforms on a CAN bus associated with these 2 logic levels be found in Appendix B and the standard connectors used for CAN buses are covered in Appendix D 3 2 4 Data Transmission There are four different fra
26. voltage curve 141 Arbitration mechanism 36 161 Standard data frame 57 eese ete tere sten 163 Extended data frame 57 sse eene 163 Standard remote frame 57 nnne 163 Extended remote frame 57 sese enn 163 Error frame 22 164 Overload frame 22 3 45 5 foe ede eoe tee eet 165 High speed CAN bus waveform ISO 11898 2 222 166 Fault tolerant CAN bus waveform ISO 11898 3 166 Single wire CAN bus waveform SAE 1 167 DBO CODDSCIOT tbi oda bri ben cr tbt 168 Spi connector E591 esce dad duet dete 168 SAE T1962 CONMECIOE AU oso t t ea t RE E IER 169 CAN transceiver MCP2551 block diagram 41 170 Conversion from Motorola Forward 15 173 Conversion from Motorola Backward 173 LIN s b b s 24 5 tnnc iniu 177 BRUSA Charging profile Mode 1 Preconditioning 178 BRUSA Charging profile Mode 2 Constant Current 179 BRUSA Charging profile Mode 3 Constant Voltage 180 Booster v A I ORO datis Ata ew Ute UL SEN ARR TM EUIS 180 CAN Operation Mode and Extras Options 181 CAN Configuration WindoW A R
27. 112 156 Processor MHz 80 MHz 800 MHz Memory Mb 2 8 CAN ports 3 4 Inputs 33 analog digital 16 analog and 16 digital 9 Logic Level 5V 8 analog and 10 digital Outputs 14 Low Side Driver 5 or 12V up to 12V and 7A Contactor Compatibility No low side drivers sinking Yes External I O board up to 5 A required 5000 Cost controller harness 2 000 4 000 30 000 35 000 software Automotive Production Yes No Grade Even though the dSPACE controller has more inputs and outputs I Os and processing power the MotoTron was selected for its better automotive packaging allowing the team to directly mount it in the engine compartment along with its capabilities for controlling the battery and charger contactors Also 3 CAN ports are sufficient to interact with all the vehicle communication networks The controller is shown in Figure 15 It should be noted that a less powerful MotoTron ECM 0555 80 was first selected and used during the second year of the competition but was updated to the more powerful ECM 5554 112 during year 3 Chapter 2 30 Figure 15 MotoTron Controller 5554 112 0904 00 16 2 7 5 Other subsystems 2 7 5 1 DC DC MES DEA 400 V 1000 W The 2 main criteria used to select the DC DC controller were whether it was automotive grade and had a power output greater than the estimated requirement of at least 400 W Meeting both criteria being affordable and even leaving enough extra pow
28. 2 the 3 powertrain configurations also known as drivelines may be series parallel and power split parallel series 2 2 3 1 Series Series hybrids have their axle mechanically coupled to an electric motor providing a full electric propulsion In other words only the electric motor is connected to the drivetrain They also have an ICE coupled to a generator to charge the battery which acts as an energy buffer between the ICE and the electric motor Figure 3 illustrates this linear driveline configuration from which the name derives Motor controller Traction motor Tractive Effart Vehicle speed Power p Traction Battery charge Figure 3 Series Configuration 5 2 2 3 2 Parallel In parallel hybrids the electric motor and the ICE mechanical outputs are both coupled to the drivetrain hence parallel configuration In this case the engine is not directly coupled to the battery only the electric motor is This principle is shown in Figure 4 Chapter 2 9 Final drive and differential Controller Traction charger Sees Battery charge Figure 4 Parallel Configuration 5 2 2 3 3 Power Split Also known as series parallel hybrids power split HEVs feature the combination of series and parallel HEV characteristics In this configuration a power splitting device PSD couples the ICE generator and e
29. 5 Total pack energy 24 kWh Available energy 19 2 kWh 80 DOD Vehicle Curb Weight 2043 kg Appendix A 159 Table 61 Simulation Results of the EREV 60 5 Parameters 005 o o mw CD 539 _ 392WMawCDiSSmpeCS _ Charging efficiency 90 upstream energy use included Table 62 Powertrain Specification of the PHEV 30 5 Table 63 Simulation Results of the PHEV 30 5 memes Accel 0 60mph 9000100 AweSTomph 52 Upps 2540WWkm CD SI5mppeCS 2985Whkm CD 369mpseCS US06 3948 Whkm CD 2S1mppeCS Towing on grade _ Charging efficiency 90 upstream energy use included Appendix 160 Appendix Additional CAN Theory 1 Arbitration Mechanism As explained in Section 3 2 2 1 when a dominant bit 0 and a recessive bit 1 are simultaneously transmitted the dominant bit overrides the recessive one This is the fundamental principle under the bitwise arbitration mechanism using identifiers Figure 52 shows an example of arbitration by using 3 nodes that attempt to send a message simultaneously and gain access to the CAN bus k of Frame K Arbitration Field Node1 TxD Node 2 TxD Node 3 TxD CAN Bus Node 2 Node 3 loses loses Arbitration Arbit
30. Applications Vendor Devices not exclusively 9 fe O 8 z CANView USB X X RM CAN Device Monitor RM Leaf Light HS X X Kvaser CanKing Kvaser CANcaseXL x x Vector tools CANoe Vector CANcardXL x x Vector tools CANalyser Vector X X x x Vehicle Spy Intrepid CANLog 4 X x G I N Configuration Program Vector CANcaseXL Log X x x Vector tools CANoe Vector MotoHawk MotoTron simulitik Mgtor ne Woodward Mico Autobox x X X Control Desk Simulink dSPACE Labview PEDEXUPOMEDAS Simulink NI VeriStand bi CAN AC2 PCI X X x Matlab Simulink xPC Target Softing dSPACE X X x Control Desk dSPACE Limited capabilities 3 8 Applications for Vehicles So far this chapter has covered the theory involved in the CAN bus protocol the hardware required to support it its bus configurations how to define CAN messages in a non ambiguous manner and introduced a few devices supporting CAN Representing a rugged serial bus CAN has a wide range of applications from providing a network in building and factory automation to connecting controllers in ships to its use in aircrafts for navigation systems and sensors to elevators forklifts and numerous other applications in industrial controls 24 29 CAN bus is also used as an in vehicle Chapter 3 62 communication network by most vehicle manufacturers in North America Europe and Asia 24 In this section
31. Chapter 2 12 the most environmentally friendly prototype the UOIT EcoCAR team could build would be a purely electric one Nevertheless the decision making was not only influenced by the competition rules but also by the university s field of expertise and the infrastructure available Having no hydrogen refuelling infrastructure required of fuel cell vehicle teams these choices were eliminated from the potential architectures E85 and bio diesel were also excluded since the university does not have sophisticated emissions testing equipment Having a large electric vehicle fleet UOIT s expertise resided in the domain of electric vehicles which heavily influenced the decision matrix This fleet comprises a total of 3 Chevrolet S10 EVs 2 Ford Ranger EVs a RAV4 EV Chrysler mini van 2 restored electric buses a 2007 Toyota Prius PHEV and a 2003 Honda Insight HEV At this point in the powertrain selection process only 3 architectures remained HEVs using reformulated gasoline with 10 ethanol E10 PHEVs using E10 and the full electric implementation Among these the full electric scored best especially from the energy consumption assessment The next section further analyzes UOIT s top 3 architectures which were EV EREV E10 10 2 5 EV EREV E10 and PHEV E10 Performance Simulations The 3 following architectures were studied in detail e EV e EREV E10 e 10 Analysis was done through modeling and
32. Configuration 12 9 86 Chargers a High Voltage Parallel Configuration b I Os 89 ENSIS ET 92 Interlock configuration nasan ann ded eus 96 Instrument Cluster a Before b After without the custom screen 98 Motor Controller Controls 106 xiii Figure 43 Figure 44 Figure 45 Figure 46 Figure 47 Figure 48 Figure 49 Figure 50 Figure 51 Figure 52 Figure 53 Figure 54 Figure 55 Figure 56 Figure 57 Figure 58 Figure 59 Figure 60 Figure 61 Figure 62 Figure 63 Figure 64 Figure 65 Figure 66 Figure 67 Figure 68 Figure 69 Figure 70 Figure 71 Figure 72 Figure 73 Figure 74 Figure 75 Figure 76 Figure 77 Figure 78 Figure 79 Figure 80 Figure 81 Figure 82 Figure 83 Figure 84 Figure 85 Figure 86 Figure 87 Figure 88 Motor Shift Lever 5 107 Custom Faceplate of the Instrument Cluster eene 112 HVEM VOS 114 Temporary lead acid battery Box iiec 125 Set up when the wheels spun for the first time 1 127 Temporary Li ion battery pack rete 130 UDDS Cycle Portion with REGEN 139 Inverter with shorted rege e NAP 140 Experimentally derived battery pack energy vs
33. Current cR Charge Voltage mV EOCC cR Charge Hysteresis mV Min EQ Voltage mV Higher Cut Off mV Higher Temp dC Lower Cut Off mV 10 Lower Temp dC 11 Nominal Capac 12 Shunt mV 13 Shunt A 14 Shunt zero mV 15 HallSens AZV 16 HallSens zero mV 17 Max Dch current cf 48000 18 Init Chg current cR 24000 19 Shut Down mV 3700 20 Curr Meas ID 400h WO ACTS Figure 77 REAP BMS Parameters Setting Mode B REAP HyperTerminal caie a File Edit View Call Transfer Help Dees 0A E 0 Module ID 1 Module Count 7 2 Module Cell Count 14 3 Module Temp Count 7 4 Battery Cell Count 94 Figure 78 REAP BMS Configuration Mode Appendix I 183 3 MotoTron CAN Interface ai H E System 19 5 SlavetFramet Age BMS Slavet V1 21 BMS Slavet v2 ERE BMS Slavet V3 BMS Slave1 V4 BMS Slave1 V5 BMS Slave1 V6 ram BMS Slavet V7 BMS Slave1 V8 BMS_Slave1_V9 22 BMS_Slave1_v10 30 BMS_Slave1_v11 31 BMS Slave1 V12 BMS Slave1 V13 33 5 Slave1 V14 34 BMS 1 Vtot 133120 10 00 10 00 10 00 10 00 14277 00 14135 00 129 00 184 StatusProblem Ag H BMS_Summary_AgeCou BMS ExternalCtrl AgeCo DutyCycle 1351132 10 11 5 MasterFrame9 Age BMS ChargerCtri AgeCo BMS ChraCtriFlags
34. Description Appendix J Vehicle Integration Module Models Table 71 VIM Port description HVCM and SCM ECM5554 Pin Colour Used Inputs Outputs Simulink Description TIM 20 EST1 EST1 EST1 Discrete Ouput Park TIM 21 EST2 EST2 EST2 Discrete Ouput Reverse TIM 22 EST3 EST3 EST3 Discrete Ouput Drive TIM 23 ESTA ESTA ESTA Discrete Ouput Brake TIM 24 EST5 EST5 EST5 Discrete Ouput Traction En TIM 25 EST6 EST6 EST6 Discrete Ouput Emergency Moto 3 Org ANO1 01 Analog Input Ign pot Moto 4 Blk wht ANO2 ANO2 Analog Input Brake pot Moto 5 Red Blk ANO3 ANO3 Analog Input Accel pot Moto 6 Wht BIk ANOS ANOS Digital Input Interlock E Stop Moto 7 Wht 6 ANO6 Digital Input Interlock Lid Moto 8 Wht BIk ANO7 ANO7 Digital Input Front ESS Moto 9 Wht ANO8 ANO8 Digital Input Alarm 1 10 9 ANOS Digital Input Alarm 1 Moto 11 Blu AN10 AN10 Digital Input Charger AC On 16 Blu wht AN15 AN15 Digital Input Changer Proximity Moto 18 Grn AN22 AN22 Digital Input Encoder A Grn Moto 19 MM AN23 AN23 Digital Input Encoder B Moto 20 Grn BIk AN24 AN24 Digital Input Encoder C Moto 21 DkGrn AN25 AN25 Digital Input Encoder D Moto 30 Red BIk 1504 1504 Discrete Ouput Contactor E55 Precharge Red Conta
35. High 40 Table la lO sie evel 41 Table 16 SAE specifications for CAN buses applications in vehicles 50 42 Table T3 Commercial 44 Table 18 Device and message ide nt 45 Table 19 S b identifierS 45 Table 20 Message and signal examples RN Ne eA E Eae 46 Table 21 Byte numbering and 47 22 pyre ORCAS d sese pde ipd rbd d de e 48 Table 25 Bit mim bein oboe ot e pH ED pP quan Dit ME 48 Table 24 Message progression i rte tie i EXER GR PARE 49 Table 25 Displas ooi ette can etti are dades Dev reete iei 51 Table 26 Compact form of the 3 main CAN display 2 11 51 Table 27 ID filtering example teret ra tete xa Ee eee ay ea AS YR 56 Table 28 Payload filtering EUM pe edt 56 Table 20 C ommercidlde VICOS ou aoc ena Som 62 Table 30 Low speed applications DDR FIORI E d Pg adem 64 Table 31 High speed appli atiOTis PR RI Me URS 64 Table 32 List of ECUs on HS amp PT Ey ieri ne 71 Table 33 List of ECUs GMLAN S Wt exo era ah a Ro eas 12 Table 54 Hardware C anges uerus de EU o Sues tetas cep Um 74 T
36. IPC VTD RFA ECC AOS SDM ROS SDARS OnStar ACCM N DLC BCM c SW GMLAN Figure 33 Stock CAN buses a GMLAN HS b GMLAN PTE c GMLAN SW Dark grayed ECUs were removed from the vehicle see Section 4 3 2 and Table 35 4 3 2 ECU changes Among the ECUS present in the Saturn VUE some became obsolete after the conversion others were mounted on components being removed For example the TPIM unit contained the HCP MCPA and MCPB thus to remove the TPIM module three ECUS had to be removed at once It is important to understand the function of each ECU on the baseline vehicle to determine which ones are not required by an electric powertrain conversion An understanding of the additional modules is also fundamental when converting a vehicle This sub section compares the original ECUSs to those Chapter 4 73 replacing them in order to explain the identification process of which ECUs to remove The stock ECUS and their related hardware are compared to their replacements in Table 34 Table 34 Hardware Changes Hardware Changes Saturn VUE UOIT EcoCAR Engine and Electric Motor with Electric Motor integrated gear box ECM TIM TCM TPIM VIM HVCM Fuel tank NiMH Battery BPCM FSCM Li Ion Battery BMS VIM SCM Alternator APM 2 DC DCs
37. Nominal battery voltage VJ 384 0 NLG operation mode Auto v Temperature coefficient of the gassing voltage XC Reference temperature for voltages Active temperature sensos 8 1 2 3 Maximum mains 02160 Switch off immediately if Charging profile gt gt gt Proceed to the next section if gt gt gt Voltage rise in 5 Min below V Temperature above Relative amount of charge in this section above 2 UN 00 Temperature rise above Charging current below Amount of charge in this section above Ah fo 60 Charging time of this section above Min 4 Amount of charge in this section above Ah Relative amount of charge in this section above Charging voltage above V Charging time of this section above Min 240 00 00 on AND Figure 70 BRUSA Charging profile Mode 2 Constant Current Appendix I 179 3 BRUSA hargeStar NLG5 D Program Files BRUSA C profilesiTestBench LeadAcidBattery Profile Connection Extras Help Connected to unit NLG 513 General settings Operation mode as MSmAR as Tempel coefficient ofthe gassing vologe RC T 2227 22 Anourtof charge above AH Charging profile Switch off immediately if KPHASE che gt gt gt Proceed to the next section if gt gt gt Charging voltage above V in Chargi
38. Parameters BRUSA Charger RS232 Interface REAP BMS 5232 Interface MotoTron CAN Interface TIM600 MotorController RS232 Interface CANLog4 USB Interface qe 1 BRUSA Charger RS232 Interface 1 1 Final Vehicle Design The description of the interface fields is self explanatory 23 BRUSA ChargeStar NLG5 D Program Files BRUSA profiles TestBench LeadAcidBattery pf O69 Profile Connection Extras Help Connected to unit NLG 513 peration mode NLG operation made Auto General settings Switch off immediately if Amount of charge above Ah 2760 Charging profile _ fg Next gt gt 5 U PHASE Maximum charging current 4 oso Status Maximum charging voltage m 4544 Status 902 029 gt gt gt Proceed to the next section if gt gt gt Charging voltage above V Charging current below 1 00 Voltage rise 5 Min below V Temperature above C Temperature rise above C Charging current below Charging time of this section above Min Amount of charge in this section above Ah Relative amount of charge in this section above Figure 69 BRUSA Charging profile Mode 1 Preconditioning Appendix I 178 23 BRUSA ChargeStar NLGS D Program Files BRUSA C _profiles TestBench_LeadAcidBattery pfl 1 2 Profile Connection Extras Help Connected to unit NLG 513 General settings Operation mode
39. Removed ECUs Remaining ECUs GMLAN HS GMLAN PTE GMLAN HS GMLAN PTE BCM BCM ECM CGM CIS FSCM TPIM EBCM CGM TCM EPS EBCM BPCM APM ECM HCP OBD II port OBD II port OnStar Including the 1200 termination resistor for each bus Table 36 gives an overview of the modifications made related to the ECU architecture Table 36 Main ECUs After Conversion Removed ECUs Retained ECUs Added ECUs Added Components APM BCM BMS for Li Ion battery Motor Controller for S10 EV induction PPCM FM TIM motor ECM EBCM Charger OBCM J1772 power inlet VIM ee m SCM TCM OnStar DC DC for 120V power outlet Accessory center console TPIM DC DC for Thermal Battery Chiller and HCP MCPA amp B Management Heater CAN ports in the glove compartment Terminations Datalogger Instrument Cluster Screen ICS LCD Screen GFI Chapter 4 76 An extra CAN bus named UOIT CAN bus was first added for segregation Nevertheless having no overlap in CAN identifiers and being far from the maximum loading capacity of the PTE bus the additional ECUs from the UOIT bus and the ECUs from the original PTE bus were combined for simplicity into one CAN bus now referred to as UOIT CAN bus Technically the migration was never fully completed Some ECUs are still allocated to the CAN port 3 of the VIM instead of CAN port 2 on which the PTE bus is connected The VIM program ne
40. Tests and Results 6 1 Introduction The previous chapter explained the functionalities of the vehicle integration module The objective of the present chapter is to highlight the vehicle results obtained by the UOIT EcoCAR team with a focus on the author s accomplishments The validation calibration and testing methodology followed by the UOIT EcoCAR team general results are presented with the individual results provided by the author 6 2 Validation Calibration and Testing 6 2 1 Controls Development Process The vehicle integration module VIM is able to interact with the battery management system the motor controller and the charger The approach used to integrate each new ECU to the vehicle was to prove that the component was communicating and reacting as expected with the VIM and via its RS232 interface in normal and critical situations on a test bench Tests were conducted on all inputs and outputs of the component to better understand the subtleties of each feature and confirm integrity The motor and its controller were tested on the bench using a 400 V lead acid battery pack and an in house control panel to manage the motor controller inputs Each module of the BMS was tested and configured using a smaller capacity Li ion battery pack borrowed from a different project At this stage of the conversion the vehicle s Li ion battery was not fully assembled and having safer and more stable battery chemistry the 400 V lead acid battery
41. VV A MotoHawkT2ole1 4 Data Wrne TransEstGear AJ TransComaGear m urns 2 MotorsawkTable4 4 Data oma DMSnRCntrITrOtGear fa 42 2 22 if it 2 2 MotoFaukT ale C4 2 2 Figure 87 Shift Lever Translator Model 2 2 iz iz Volatlie Data TransEstGesr AU TransComeGear 1925 Volatlie Data On SnCntrITrgtGear uita Volatile Data TmsS fLvrPos uinea Volatiie Data ulta Appendix J 198 6 YEL GRY WHT BLK WHT Transmission Range Indicator Switch Park Input I 3 PIN Transmission Power Hor pd CONNECTOR Range Switch FOR H V INPUT 772 INPUT B A DE G RN D 773 INPUT C a 776 PARITY is GRN 470 SENSOR GROUND C EL gt NotUsed PRK L BLU Mas S10EV PRNDL ENCODER Figure 88 Schematic of States service manual Note In the UOIT vehicle GRND 3 pin connector and E 6 pin connector are not used as a ground reference but are powered with 5V instead N and PRK 3 pin connector are connected to the shift lever whereas A B C and D 6 pin connector are connected to the MotoTron analog inputs for decoding Once decoded the MotoTron discrete outputs are updated and the TIM sees the decoded value of the shift lever Appendix J 199 100 1 StarPulse Figure 90 Igniti
42. Viet Vss VIM VTD VTS Wh WTP WTW State of Charge Start Of Frame State of Health Substitute Remote Request Sport Utility Vehicle Single Wire Transmission Control Module Temperature Traction Inverter Module Traction Power Inverter Module Transistor Transistor Logic Transmit Transmit Digital Universal Asynchronous Receiver Transceiver Urban Dynamometer Driving Schedule drive cycle Utility Factor University of Ontario Institute of Technology United States United States Department of Energy Volt AC voltage DC voltage Supply voltage Reference voltage Ground Vehicle Integration Module Vehicle Theft Deterrent Vehicle Technical Specifications Watt Watt hour Well to Pump Well to Wheel xxii 7 ZEV Zero Emission Vehicle Symbols Q Ohm xxili Chapter 1 Introduction 1 1 EcoCAR The NeXt Challenge EcoCAR The NeXt Challenge also called EcoCAR Challenge debuted in the Fall of 2008 and ended 3 academic years later in the Summer of 2011 1 The objective of this inter university competition was to re engineer the powertrain of a donated vehicle into a more eco friendly prototype while maintaining consumer appeal in the areas of performance utility and safety Powertrain architectures that reduce fuel consumption well to wheel greenhouse gas emissions and energy consumption and tail pipe emission are considered eco friendly In 2008 the University of Ontario Institute of Technology UOI
43. Vtot BMS Slave4Framet BMS Slave4 V1 1 5 Slave4 V2 BMS Slave4 V3 5 Slave4 V4 BMS Slave4 V5 BMS Slave4 V6 BMS Slave4 V7 BMS Slave4 V8 BMS Slave4 V9 BMS Slave4 V10 BMS Slave4 V11 BMS Slave4 V12 5 Slave4 V13 BMS Slave4 V14 5851900 BMS Slave4 Vtot 83 4145 4160 4155 4140 4150 4155 4155 4169 4174 4194 4213 4233 4252 58572 00 i BMS_MasterTempNb i rj BMS MasterVmin BMS Slave5 V3 BMS Slaves V4 BMS Slave amp V5 BMS Slave5 V9 BMS V40 BMS Slaves 11 BMS Slaves V13 BMS Slave5 1 14 1 5 Slave amp Vtot MasterCellNb BMS MasterFrame0 Age ux BMS MasterFrame1 Age 114 BMS_ MasterFrame2 _Age 113 BMS MasterFrame3 _ 111 BMS MasterFrame4 Age 109 BMS MasterSOC 199 _ 5 165 4140 BMS Slave4Frame4 Age 82 BMS Slave5 V1 4140 14140 14145 14140 14155 14145 14155 14165 14174 14189 14213 14228 14243 14272 1 58504 00 BMS Slave5 V2 Slave5 V6 BMS Slave5 V7 BMS Slave5 V8 BMS Slave5 V12 Figure 79 BMS supervisory interface Appendix I 184 MotoTune EcoCAR Display Explo X Display 2 CAN PRNI i Prelnspectio H A System 277 HVBatSOC SvcHybChrgSyslO HVBatChrgCrdConniO HVBatChrgPow HVBatCrnt HVBatCrntV HVBatDschrgPow Probe Ignition ON Probe IgnitionPot Probe A Propp PREDIN
44. a robust in vehicle communication network not only to connect these controllers but also integrate with ECUs in today s vehicles This efficient and robust field communication protocol is termed the controller area network also known as CAN bus 29 This section constitutes one of the few non confidential documents available that details CAN bus for vehicles It covers the fundamentals of the communication protocol theory along with hands on application useful to most CAN users In summary an overview on CAN protocol fundamental theory a description of the hardware required to create a CAN environment list of devices supporting CAN available on the market a rigorous definition of the nomenclature used to accurately define CAN messages and CAN signals different CAN bus configurations a variety of techniques to use identifiers and a description of automotive applications 3 2 Overview 3 2 1 History Robert Bosch GmbH started developing the controller area network protocol in 1983 but only officially released it at the Detroit s Society of Automotive Engineers SAE congress in 1986 32 At that time the automotive industry was looking for a reliable in vehicle communication system CAN was developed for use in industrial environments and for in vehicle networks but was not the only networking technology competing in the race for the automotive market Volkswagen Group was developing the A bus while another European group was developing
45. as the energy consumption determination motor tuning and regenerative braking calibration and was fixed for the final competition 6 3 3 Vehicle Technical Specification Even though the vehicle did not partake in any of the dynamic events at the final competition the vehicle technical specification VTS met all the competition requirements as summarized in Table 57 The 0 60 mph and 50 70 mph accelerations along with the 60 0 mph braking had been tested at the second year one and EPA The performance values for the final design were extrapolated from the data collected at EPA and can also be found in Table 56 The vehicle can seat as many passengers as the stock vehicle configuration and has an increased cargo space through an in floor compartment The UF weighted FE and GHG emissions were calculated extrapolating the results recorded at EPA A range of 482 km was estimated UOIT s vehicle scored 100 at the towing capacity event at year 2 competition Using the same motor and inverter with a larger battery and no significant mass increase the towing capacity of the vehicle should not be affected At the final competition the vehicle mass and ground clearance were measure at 2130 kg or 2112 kg and 178 mm respectively The starting time of the vehicle is only limited by the high voltage precharge circuit calibrated to last less than 2s In summary the vehicle performance met all the EcoCAR competition requirements Chapter 6 143 Tab
46. can be used with different hardware while other less elaborate software is specific to a device Commercial devices for CAN can be categorized in 4 major groups e Scopes e Loggers e Controllers e Hardware In the Loop HIL Scopes are basically CAN to USB gateways allowing one to visualize the identifier and raw 8 bytes of data CAN message A logger s purpose is to log traffic on a CAN bus Some can be programmed and operate without constant computer interaction using embedded memory in other words they are standalone devices with memory whereas other devices have no memory but still allow logging through a software application by using the computer s memory As discussed in Section 3 3 controllers are programmable microcontrollers with circuitry allowing communication on one or several CAN buses Hardware in the loop simulation devices are machines able to interact in real time with an external system and update the measurements in a model running on a computer or on the device itself to test real time embedded systems 59 Chapter 3 61 list of commercial devices available on the market is extensive Some of the most common ones are summarised in Table 29 with their features software and vendor information is extracted from 64 65 66 67 68 69 70 71 72 Table 29 Commercial devices Features 55 8 Sof t Commercial v Fee ES 2 515
47. completed the integrated powertrain on the test bench was taken and installed as is in the car The test bench custom control panel and terminal blocks fixed on a piece of wood were first used but quickly the pedals and other controls were re routed to control the car from the driver seat Static and dynamic testing followed This methodology resulted in an efficient vehicle conversion while minimizing the risk of controls errors The outcome of this controls conversion process is discussed in more details in Section 6 2 1 4 3 Electronic Control Units 4 3 1 Original ECUs The network architecture of the hybrid 2009 Saturn VUE consists of 3 controller area networks CAN buses using the higher layer protocol called General Motors Local Area Network GMLAN This protocol is also compliant with 2 SAE specifications for CAN buses applications in vehicles SAE J2284 3 high speed CAN for vehicle applications at 500kb s and SAE J2411 single wire CAN for vehicle applications at 33kb s A brief description of the 3 CAN buses present on the vehicle follows HS GMLAN High speed GMLAN is the main dual wire CAN bus on GM s vehicles and provides a communication network for most chassis and powertrain modules i e TPIM EBCM EPS BCM CGM OnStar BPCM FSCM DLC Data is transmitted at 500 kb s and is compliant with the SAE J2284 3 standard Chapter 4 69 PTEGMLAN GMLAN powertrain expansion is also dual wire CAN b
48. current is detected the motor controller is placed in an emergency mode that limits the output current of the motor and an orange warning LED is turned on to inform the user If a high leakage current is detected the MotoTron sends a signal to open the main contactors disabling the motor controller output to stop the vehicle In this case a red GFI LED is also turned on in the instrument cluster An inertia switch from First Technology is calibrated to get triggered when the vehicle sustains an impact greater than 8 G s It acts as a safety feature used to open the main contactor relays in the event of a collision to protect passengers from high voltage exposure In other words it is a resettable crash switch that disconnects the power in the high voltage distribution box during a significant collision The specifications on this component are reviewed in Table 44 Table 44 Resettable Crash Sensor Specifications Response to impact 360 Rated load 10A Calibration range 8 30 G s Resettable Manually Inertia Mechanism Magnet restrained mass 4 7 On board Debugging and User Interfaces 4 7 1 Driver Notifications The instrument panel cluster or dashboard informs the driver through gauges lights and chimes on the vehicle status The information required by someone driving an electric vehicle is different than the one needed for driving a conventional vehicle To address this need the instrument cluster was
49. customized by installing a new layout back panel modifying the logic behind multiple LEDs gauges and chimes and also integrating an in house screen within the original dashboard Figure 41 shows a picture of the stock cluster and one of the new layout A detailed model of the instrument panel cluster is shown in Appendix K Chapter 4 97 b Figure 41 Instrument Cluster a Before b After without the custom screen The speedometer remains unmodified the fuel gauge now displays the battery voltage as an approximation of its state of charge SOC and the empty light turns on at low SOC A bidirectional ampmeter gauge was created using the original 12 V battery needle to indicate the current flow while monitoring the amount of regenerative braking The yellow oil pressure light was replaced by a red LED for ground fault interrupt alarms and any issues related to the powertrain A chime is also triggered when the GFI Chapter 4 98 light is activated cue the driver via visual and audible indicators for critical situations For a lower current leakage the malfunction indication light MIL or commonly called the check engine light is enabled and no chime is audible A green vehicle ready light replaces the yellow oil change indication light The vehicle ready light indicates that the main HV contactors are closed in other words that the powertrain is enabled The stock screen occupying the center of the cluster
50. depletion strategy within the aforementioned SOC window gives UOIT s electric vehicle a range of over 400 km Chapter 6 discusses this topic in detail Regenerative braking controlled by the motor controller is blended with the conventional hydraulic braking system The brake blending algorithm is proprietary to the TIM The hydraulic pressure is controlled by the displacement of the brake pedal and the electronic brake control module EBCM whereas the motor controller manages the amount of regenerative braking to apply according to calibration values For example a low regenerative braking is applied while coasting to emulate an internal combustion engine ICE feel Also regenerative braking is reduced at high SOC to avoid overcharging the lithium battery which would result in triggering a protection mechanism Another aspect of the control strategy for discharging the battery is the emulation of the typical creep torque of an ICE A built in function of the motor controller was calibrated to emulate creep torque when having the shift lever in the drive position Chapter 4 81 4 4 3 High voltage schematic The high voltage schematic of the UOIT EcoCAR architecture is detailed in this section and shown in Figure 36 Charge Port
51. door lock receiver Roll over detection for airbag deployment XM Radio Remote Function Actuator Roll Over Sensor Satellite Digital Audio Radio Service Sensing and Diagnostic Module Vehicle Theft Deterrent The physical interconnections between the ECUS described in Table 32 and Table 33 are shown on a layout diagram in Figure 33 The ECUS on the 2 high speed buses utilize a hybrid daisy chain topology this occurs when the majority of components are daisy chained together with a few connected in a star topology Nevertheless the single wire bus is split in 2 buses using a star topology which are daisy chained together by the body control module BCM HS and GMLAN are both terminated by a 120 resistor at Chapter 4 72 both ends of their CAN bus The terminates both networks at one end and independent resistors of the same value adjust the impedance at the other end The 120 Q resistor for the HS network 15 located near the fuel tank and the one for the can be found underneath the vehicle at the front passenger seat position TPIM TCM EARS CGM H OnStar H BPCM H FSCM EPS DLC HS GMLAN gt ECM TPIM H EBCM 3CGM CIS DLC b PTE GMLAN IRC
52. e LE 206 Figure 97 Custom Faceplate of the Instrument Panel Cluster without the 2 display ERR ER T E 207 Figure 98 Custom Faceplate of the Instrument Panel Cluster Warning Symbols Hidden without the 2 display SCrEENS o osedsentot ens d oeste xime A oe 208 Figure 99 Stock Faceplate of the Instrument Panel 22112 209 XV List of Tables Table 1 Well to Wheel energy and GHG production 7 eese 12 Table 2 Performance Simulation Summary 7 essen 14 Table 3 EcoCAR Competition Performance Requirements 5 14 Table 4 Targeted Vehicle Technical Specifications 71 16 Table 5 Kokam SLPB 160460330 11 20 Table 6 BMS14C Specifications ous 23 Table 7 BRUSA NLG513 3 3kW Specifications 13 25 Table 8 Delphi S10 EV Delco System 110 400 A 240 Via 14 28 Table 9 MES DEA TIM600 15 ar 28 Table 10 Comparison MotoTron vs MicroAutoBox 16 17 00 30 Table 11 MES DEA DC DC Converter 1 kW Specifications 18 31 Table 12 24 19 32 Table 13 EV subsystems selected te ae Cale 35 Table 14 Differences between Low Speed and
53. for the cascading shut off sequence The traction inverter module uses a RS232 interface named Powertrain Inverter Series This interface was very useful for motor calibration and behaviour analysis For example graphs can be plotted in near real time and motor data can be recorded for post analysis as shown in Appendix G Figure 82 Relevant calibration values programmed in the UOIT EcoCAR traction inverter can also be found in Appendix G Table 69 and Table 70 An auto tuning feature is also available to automatically tune the motor with the motor controller Additionally approximately 300 parameters can be modified to customize this drive Unfortunately these parameters have poor documentation making it difficult to calibrate properly In conclusion the physical connections required to enable the communication of these 4 controller interfaces were made accessible from the front passenger seat through the glove compartment for convenience Chapter 4 100 4 8 Summary architecture of the original and modified vehicle was presented The control strategy implemented by the high voltage controller and the 3 main powertrain controllers were detailed A more in depth explanation of the safety control strategy was provided A description of the on board diagnostics system and troubleshooting tools developed was reviewed The next chapter addresses the implementation of this control strategy via programming the VIM Chapter 4
54. ignored Table 28 Payload filtering example Binary Value Results ID 600 0b110 0000 0000 ID mask 7FF Ob111 1111 1111 1 care 0 don t care ID acceptance filter 0b110 0000 0000 Only ID 600 accepted Payload filter 01 00 00 0b00000001 00000000 00000000 Payload mask FF 00 00 Ob11111111 1 care 0 don t care Payload acceptance filter 0b00000001 Incoming Message x any value allowed Sub ID 01 in byte 0 ID 400 0b100 0000 0000 ID acceptance filter 0b110 0000 0000 Message ignored Payload 0100 00 Payload acceptance filter ID 600 0b110 0000 0000 ID acceptance filter 0b110 0000 0000 ID accepted Payload 03 00 00 0b00000011 Payload acceptance filter 0b00000001 Message ignored ID 600 0b110 0000 0000 ID acceptance filter 0b110 0000 0000 ID accepted Payload 0100 00 0b00000001 Payload acceptance filter 0b00000001 Message accepted Chapter 3 56 3 5 3 2 Signal Definition A CAN signal is described by the following characteristics 28 47 48 Signal name Start bit Length in bits Byte order little endian or big endian Data type signed unsigned float double boolean etc Units mV A C km h etc Scale also called factor or gain Offset Range minimum and maximum Scaling Signals are often co
55. it is programmable for different output voltages and currents or charging time and is compatible with a 120 input as well as 240 Moreover multiple charging phases can be programmed and are used in UOIT s charging strategy to coordinate the operation of the 5 independent BRUSA chargers in the prototype Table 7 details the characteristics of 1 BRUSA charger along with the specifications of the UOIT EcoCAR prototype charging profile 13 Figure 12 displays an individual charger and the layout on the engine craddle while Figure 13 shows the SAE J1772 charging plug located behind the front licence plate Table 7 BRUSA NLG513 3 3kW Specifications 13 1 module Final Charger Design of modules 1 5 Vin rms ac max 230 V 230V Vin rms ac min gt 120 V gt 120 at reduced output power 1 kW 1 kW Efficiency 90 to 9396 90 to 9396 I 12 5 A 58 5 260 V 282 V V 520 394 8 6 25 41 8 Max output power 3 3 kW 16 5 kW Chapter 2 25 b Figure 12 a Brusa NLG513 3 3kW b 3 of 5 Brusa NLG513 3 3kW Integrated Chapter 2 26 Figure 13 41772 Charging Inlet 2 7 3 Motor and motor controller Delphi S I0EV and MES DEA TIM600 The electric traction motor is controlled by a traction inverter module TIM also referred to as motor controller Only the combination of a motor and gearbox assembly with differential can be dee
56. on TXD input The thermal shutdown block disables the driver control outputs controlling the transistors when they overheat Transistor overheating can be caused when conducting an excessive current due to a short circuit on the bus An external signal Rs for Figure 65 can control the rise and fall times of CANH and CANL in order to reduce electromagnetic interference also called radio frequency interference RFI This signal also controls the CAN transceiver sleep mode In sleep mode the discriminator or receiver operates at a lower current and the driver control is turned off Filters can be placed between the CAN transceiver and the CAN bus to help reduce noise In the case of a single wire CAN bus EMI can be reduced by adding an inductor next to the CAN transceiver on the CAN wire For a dual wire CAN bus high speed CAN bus a common mode choke and capacitor can be added to improve radiated emissions In both cases electrostatic discharge transient suppressor techniques also known as ESD protection techniques can be used to prevent permanent damage to the CAN transceiver related to undesired voltage transients on the bus One of these techniques involve the connection of back to back zener diodes such as mmbz27vclt1 61 or pesd24vs2uat 62 as close to the CAN transceiver as possible between CANH and the ground and CANL and the ground Other techniques use ESD capacitors or metal oxide varistors MOV or other suppression de
57. the French Vehicle Area Chapter 3 36 Network VAN 1992 Mercedes Benz built the first car integrating CAN bus helping CAN to finally win this networking technology race in the mid nineties 32 Since 1996 the on board diagnostics II OBD II standard is a mandatory vehicle diagnostics standard 56 Five communication protocols are included in that standard CAN being the most important With the arrival of the CAN based protocols DeviceNet and CANopen in the mid nineties CAN rapidly found applications in factory automation Since 2000 this robust field bus kept expanding to other industries such as elevators and forklifts connections between subsystems in ships flight status sensors and navigation systems in aircrafts factory and building automation automatic door control for monorails different types of industrial controls medical systems as well as laboratory and operating room automation Some of the important CAN features are its multi master capability its built in error detection and correction capability as well as its unique fault confinement 22 3 2 2 Properties of CAN Protocol 3 2 2 1 Basic Concepts CAN bus is a serial communication protocol level supporting real time systems with a high reliability It handles the detection of collisions the detection of errors the retransmission of corrupted messages and the prioritization of sent and received messages The identifier length can be 11 bits standard form
58. the current entering and exiting the ESS SOC estimations are made and the CAN messaging is heard by the MotoTron controller which can take actions like limiting current to from the TIM or chargers or disabling the system Similar to the HVCM the BMS has 2 main modes normal and charging mode These modes are controlled using a hardware connection from the HVCM to each BMS unit The normal mode of the BMS can be subdivided into ignition off and ignition on When the ignition is off cell balancing occurs When the ignition input is activated the BMS expects current to be drawn As soon as AC is detected at the charging inlet the Chapter 4 87 VIM exits the run mode described in Figure 35 regardless of the shift lever position and vehicle speed The VIM activates the charger input of the BMS enables the chargers closes the contactors and starts charging once the ignition key is in the OFF or ACC position and all the other charging conditions are met It is possible to charge with the ignition key ON if it has not been cranked in the current key cycle In other words if the driver cranked the ignition key or attempted to crank it while charging it has to be turned back to ACC or OFF before charging is enabled again Additional safety could be combined to the ignition key logic and AC detection using the shift lever position and the vehicle speed The BMS also provides regular battery information and updates to the VIM using the modi
59. the focus is primarily on CAN bus as an intra vehicle communication environment In addition Appendix H compares CAN to LIN a cheaper communication system for vehicles 3 8 1 Types of Applications There are usually multiple CAN buses in vehicles Up to 5 CAN buses have been seen for example 4 high speed buses and a single wire one The types of applications differ whether it is a single wire bus or a high speed one 3 6 1 1 Low Speed Applications As defined in Table 14 33 3 kb s is the speed of a single wire CAN communication system Transmission rate is the primary limit of single wire buses Fault tolerant buses and high speed buses configured with a low baud rate can also be used for non time critical purposes however requiring only one wire and being slightly cheaper to implement than high speed and fault tolerant buses single wire buses are generally used for these applications Table 30 provides a list of examples of low speed applications expanded from reference 24 Chapter 3 63 Table 30 Low speed applications Dashboard Instrument cluster panel Cabin temperature controls Light sensor Steering wheel electronics Entertainment controls Infotainment controls Air Conditioning controls Door control Mirror Window Door locks Seat control Seat position Seat heater Occupancy sensor Roof control Interior lights Visor lights
60. was spent routing signals from one terminal block to another along with the BMS installation and configuration This required testing and replacing some cell inline fuses needed for cell voltage monitoring connecting and routing the fuse strip board cables to the BMS etc No dynamic performance tests were achieved during the final competition since the vehicle was not driven The UOIT team ran their vehicle for the last time after the EPA event in road testing of the motor controller s tuning parameters prior to the final competition event and before the temporary Li ion battery pack situated in the vehicle cargo space was disassembled and rebuilt into the intended battery pack design 6 3 EcoCAR Project Results 6 3 1 Competition Result Overview UOIT placed 6 at the second year competition and was one of the few teams with a running vehicle Similarly UOIT was one of the few teams having their prototype ready for dynamometer testing and taking advantage of the sophisticated EPA facility Extra dynamometer time was even given to the team for a lack of usage from the other universities Unfortunately a very risky and lengthy step remained to be achieved the Chapter 6 134 transition from the temporary Li ion battery pack stored in the trunk of the SUV to the final battery packs one located underneath the vehicle and the other in the engine bay Delays in the assembling process combined with the short period of time between th
61. was used to test the charger The VIM inputs and outputs were verified mainly with a multimeter switches and contactors whereas the CAN ports were tested by establishing communication with the main powertrain controllers Chapter 6 124 Figure 46 Temporary lead acid battery box Once the independent testing of each module was completed they were integrated and tested two by two on the test bench For instance the main powertrain components were tested as follow the BMS and the VIM then the charger and the VIM and finally the TIM and the VIM Methodically each interacting set of controllers were tested 2 by 2 then 3 by 3 and so forth After numerous tests the system was ready to be migrated systematically into the vehicle The migration from the test bench to the vehicle was initiated by integrating the VIM Two of the VIM s main tasks once installed in the vehicle were to act as a gateway between the high speed and powertrain GMLAN buses the latter being also used as a custom CAN bus plus to emulate all the removed modules i e Engine Control Module ECM Fuel System Control Module FSCM Battery Pack Control Module BPCM Traction Power Inverter Module TPIM and Accessory Power Module APM Unfortunately the emulation of the removed ECUS could not be tested outside of the vehicle Therefore the first tests performed were to connect the VIM to the HS and the PTE GMLAN buses to establish communication with the vehic
62. 00 sec 9 1 mna Exit 100000 sec 55 0 0 _J iKey I IL 0 0 ChargeMode entry Status lgnKeyss IL 0 IL_ESSNeg 0 IL_ESSPos 0 0 Init entry Relay ESSNeg 0 entry Relay 55 5 0 entry Relay_Chrg 0 entry TractlonEN 0 entry HV_Status 0 1 5 2 19 1 amp StartPuls e StartPuls e 0 Stop entry Relay ESSNeg 0 entry Relay 55 5 0 entry Relay Chrg 0 entry TractionEN 0 entry HV_Status 0 L lgnKe y 0 IL 0 lgn Ke y 0 IL 0 IL_ESSNe g amp IL 1 amp 0 afte r 1000000 Sec y Chrgzo entry TractionEN 0 entry HV_Status 8 55 0 IL ESSPos 0 after 1000000 sec 11 ESSNeg 0 PrechrgRun entry Relay ESSNeg 1 entry HV_Status 1 gt 2 1 amp IL_ESSNeg 1 ESSPosEnRun entry Relay ESSPos 1 entry HV_Status 2 IL ESSPosz 1 TractionEnable entry TractionEN entry HV Statusz3 Rs 1 StartPulse 0 amp after 10000 s ec RunMode entry HV Statusz4 0 IL_ESSPos 0 AC 1 Figure 95 HV Control Module Stateflow Appendix J 204 3 Software Versions Used The software versions used with Windows 7 to program the VIM are Matla
63. 10 00 S P22 s 23 5 24 5 40 I nom AC P42 P43 51 P52 Phi nom P54 Monitor sampling period 0 1999 9 ms Appendix I 188 VOTI Default etis Description Range EcoCAR Value Unit Value P55 Points memorized after monitor trigger 1 2000 100 1 526 Monitor trigger level 400 2 E 0 Access Code for Reserved Parameters P61 P99 key 95 ais P62 V nom motor P63 Hz P64 Vnom P65 RPM n max P67 Number ofmotorpoks 4 4 Integer P69 Integer P91 Degree C P95 ohms P96 Thermal trip threshold on logic output 14 50 0 100 0 of 99 Access key to TDE parameters 92 0 key 82 P100 Value of acces key P60 to reserved parameters 95 95 key 95 P101 PWM frequency 2500 9000 7500 5000 Hz P103 I nom P105 106 volts P107 volts P113 A P115 Multiplication factor for motor PTC NTC PT100 analog 0 200 200 01 200 01 g reference value Appendix I 189 UOIT EcoCAR Value Acceleration time for tests 3 and 4 during auto tune 0 3 1999 9 Smart brake voltage cut in level 300 850 750 0 Enable alarms 100 100 Maximum torque in DRIVE Max RGEN torque in DRIVE 200 200 Max RPM in DRIVE 12000 12000 Time to reach torque value P200 in DRIVE 100 100 Time to decrease the torque value from P200 to zero in Time to reach
64. 2 Bit Timing 500 kbaud TX Queue 16 messages RX Queue 16 messages MotoHawk Function Trigger FGND RTI PERIODIC Priority Order 0 Total FLASH Total EEPROM Total RAM App FLASH App EEPROM App RAM MotoTune Protocol Enabled City ID 0x0B PCM 1 oooooo Foreground MotoHawk CAN Definition existence of a Target Definition block somewhere in the model sets it up as MotoHawk project capable of being built to a MotoTron target module Name 3 Bus CAN 3 Bit Timing 500 kbaud TX Queue 16 messages RX Queue 16 messages MotoHawk RTW Code Coverage Test Bit System ALWAYS GENERATE CODE COVERAGE CODE The Main Power Relay block monitors the module power and controls the Main Power Relay also performs startup and shutdown procedures Doroci Emablud City ID Ox0B PCM 1 MotoHawk RTW Tool Chain Definition necessary for EEPROM storage Most designs should contain this block Name MyToolChain Tool Chain goo powezpcceabispa 4 4 All actions must be performed inside of a function call subsystem triggered by a MotoHawk entry point trigger block in order to be included in the generated code MotoHawk RTW Fault Manager Definition Storage FLASK WY Data Type uint amp Read Access 1 Wrte Access 1 Clear Access 1 Figure 84 Top Level Subsystem Appendix J 195 Model Browser FinalController 2
65. 22 1 Total miles driven 232 967 Average 347 2 22277 SOC 9 End 4 8 Table 52 EPA Energy Consumption Results Battery Depletion Summary Averaged Drive Schedule Estimated ASOC Wh mi Range mi UDDS Highway 60 mph Battery Depletion Schedule W O REGEN Table 53 EPA Energy Consumption Results Regenerative Braking Activated Miles Cumulative Energy Energy for Consumption Extrapolated Schedule Driven Use KWh Schedule KWh Wh mi Range mi Start of Schedule UDDS 9 Final tune regen UDDS 10 UDDS 39 10 estimate UDDS 9 Final tune regen UDDS 10 GFI UDDS 9 10 estimate UDDS 29 10 estimate Table 54 EPA Energy Consumption Results Regenerative Braking Activated Summary 1 1 9 Estimated Drive Schedule Consumpion Estimated SOC Wh mi Range mi Start of Schedule UDDS with REGEN averaged from both clamps Highway consumption with regen 8 60 mph Battery Depletion Schedule with regen engaged Chapter 6 138 Table 55 Influence of Regenerative Braking on the Range w o REGEN w REGEN 356 3 km 401 1 km 222 7 mi 250 7 mi Range UDDS Cycle Portion with REGEN Operative 30 120 100 20 2 2 40 d 15 20 o 8 9005 speed 2 10 i 20 3 current REGEN 40 60 0 80 2300 2400 2500 2600 2700 2800 Time scale seconds Figu
66. 3 It should be noted that the bit stuffing rule does not apply to these flags The way to recover from an error is by automatically re transmitting the faulty message from the transmitter Each node has an error counter determining whether it is in active error mode passive error mode or bus off mode When a station is in active error mode it sends an active flag on error detection The opposite applies for a node in passive error mode In other words by sending flags with dominant bits nodes in active error mode have priority over the ones in passive error mode The general structure of an error frame is illustrated in Figure 57 Data ERROR FRAME Interframe Space or Frame Error Flag Overload Frame superposition of Error Flags Error Delimiter Figure 57 Error frame 22 The station or node sending the first error flag sends 6 dominant bits when in active mode breaking the bit stuffing rule When the bit stuffing error is detected the other nodes in either active or passive modes start sending active and passive flags respectively In this case passive error flags are ignored because they are superscribed by the active flags dominant bits As a consequence the value of the bus becomes a superposition of error flags varying from 6 to 12 dominant bits see Figure 57 This superposition does not affect the length of each nodes error flag but does affect the overall error active flag o
67. 38 Description of Powertrain Control Strategy by Operational Modes Main contactors are open controllers are in sleep mode HV discharge circuit is active to remove remaining energy from the motor controller Accessory controllers are functional main contactors are open discharge circuit is still active Instrument cluster turns on all display screens are functional main contactors are open discharge circuit is still active Transition state The car must be in park and the driver must press the brake pedal while cranking the ignition key to start the vehicle This starts the precharge cycle closes the precharge contactor and the battery negative terminal contactor Main contactors are closed motor controller output is enabled car is ready to drive the car is ACC ON Ignition OFF Ignition Starting o d gt E E Z Ignition ON Regen Disable in charge depletion mode Regen Enable MotoTron sends regen enable signal and brake pedal ON OFF signal to the motor controller Charging Mode Charging is enabled when the charge port detects an AC source plugged in At this time the main contactors are closed charging contactor is closed motor output traction is disabled Table 39 Power States of Main Controllers by Operational Modes ACC ON Ignition OFF Powered HV output disabled MotoTron Powered Functional ON Ignition ON Power
68. 4 1 Powertrain Control Strategy The FFEV provides some simplification in the controls strategy to ensure functionality and operation of the vehicle Complex controls to switch between various modes of operation are not required The single charge depletion mode isolates most of the controls complexity in the BMS coding to ensure safe charging procedures and the implementation of various algorithms for internal safety systems are used to protect users and promote extended battery range and life The control system strategy consists of charge depletion mode regenerative braking charge assist and charge mode Complementary to these modes the controls strategy is classified based on the states of vehicle operation which dictates what control system architecture is necessary to fulfill the requirements of each state Chapter 4 78 f Y Initialization Stop Mode Start Run Mode N E N Charge Mode Run Mode gt 2 Start Charge Mode Figure 35 HV Control Module flow diagram The flow diagram provided in Figure 35 describes at a high level the high voltage control module HVCM the detailed Simulink Stateflow can be found in Appendix J Figure 95 The points gathered in Table 38 give a brief description of the states of the vehicle operation and their control functionality whereas Table 39 details the power states of the main controllers Chapter 4 79 Table
69. CM 0555 was powerful enough for the second year competition it was missing an important feature to accomplish the prototype s final design a 3 CAN port Being more powerful overall and having the required extra CAN port are the main reasons why the ECM 5554 was chosen over the ECM 0555 in the final year and the unit is displayed in Figure 15 Section 2 7 4 Also the VIM pinout can be found in Appendix J Chapter 5 103 5 3 Model based design 5 3 1 Top Level Subsystem The top level subsystem of the VIM model contains the standard MotoTron blocks required to prepare the model for code generation and to configure the basic function blocks Extra blocks are also added to initialize functions related to the program modeled Appendix J Figure 84 represents the top level subsystem of the VIM The MotoTron controller series has many different controllers sharing the same Simulink library thus a target definition block must be added shown by item 1 in Figure 84 Appendix J This block is used to set up the model as a MotoHawk project and to specify the target i e the MotoTron device As mentioned in the previous section the MotoTron chosen for the VIM is the ECM 5554 112 0904 C00 M Along with the target the compiler for the Matlab Real Time Workshop RTI needs to be defined As a reference the compiler required by this MotoTron is the gcc powerpc eabisps 4 4 identified by item 2 in Appendix J Figure 84 The main power relay and
70. Connected SI HEV E85 0 6 0 69 L CIDI Vehicle BD20 1 13 Grid Independent CIDI HEV BD20 Grid Connected CIDI HEV BD20 Electric Vehicle FCV G H2 Using the EcoCAR energy mix for electricity electric vehicles ranked 1 for energy consumption and grid independent CIDI HEVs using biodiesel were best in emissions However fuel cell vehicles seem to be the best choice overall in terms of energy use and GHG production for the EcoCAR energy mix for electricity Grid dependent and grid independent SI HEVs using E10 were respectively 3 and 4 in energy use for the EcoCAR and Ontario energy mixes thus they were considered as potential architectures by UOIT When only considering emissions with the mix electric vehicles ranked 6 showing that they are not the logical architecture to score high at this competition in regards with emissions However they score best in energy use and considering the points weighting were second only to FCVs The hydrogen based architecture was not an option due to infrastructure requirements so the full electric vehicle architecture placed first on UOIT s list Additionally electric vehicles are clearly more eco friendly than any other propulsion technologies when using electricity generated in Ontario Competition rules aside and considering that the electricity generated in Ontario is greener than the energy mix defined by the EcoCAR challenge
71. Design and Implementation Process Controls Integration using CAN bus on a Full Function Electric Vehicle Conversion by Hugo Provencher A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Applied Science in Automotive Engineering Faculty of Engineering and Applied Science University of Ontario Institute of Technology March 2014 O Hugo Provencher 2014 Author s Declaration I hereby declare that I am the sole author of this thesis This is a true copy of the thesis including any required final revisions as accepted by my examiners I understand that my thesis may be made electronically available to the public The literature reviews used mainly Controller Area Networks For Vehicles were written by the author The EcoCAR reports used were written by many authors As a co author I declare that what is used in the thesis was my own work unless otherwise referenced iii Abstract From the electrical engineering perspective this thesis addresses the design and implementation of the conversion process from a hybrid electric to a full function electric vehicle FFEV The architecture selection process and main components of an electric vehicle EV are described and an exhaustive literature review on the controller area network CAN is presented The electrical and control system integration strategy is explained along with the model based algorithm programmed in
72. E UE oad HAE EON RR 97 4 7 2 Troubleshooting Interfaces for Engineering Development 99 255 SUMMAT modd e S de rad teen dp d Ga de 101 Chapter 5 Vehicle Integration Module por Rer 102 rer 102 5 2 Overview of the Moto ron 4nd eiii oed ch ia utei cette edi editors 102 1 104 3 3 L Top Level Subsystem 104 5 3 2 Model Struct re 105 5 4 Encoders and 106 5 4 1 Motor Controller 106 106 106 34 2 2 LOI OTI 6 E 109 5 4 3 Acceleration and Brake Pedals ise e tice a m LR NR RI NETS 109 5 4 4 Radiator Fatis cta cns e b ee EO CR S SER EO aA 110 Jo pecdo metei eere pos nce p URS 110 2206 Didone sties and Safe Py daas 112 5 461 DIAT BOSLUICS au ee I 112 ou D PAUCI PME 113 3 9 ute 113 114 am M wees 115 SOCAN Gommutueatloli ooi toi p Eni EDD NE uisu t IER 115 5 6 1 Emulated CAN M6SSIEBS lees oed NU te enge 115 SOZ OFT ACs 117 2 62 T Mot r Controller 118 5 6 2 2 Battery Management System
73. Foreground _ BMS Rx BPCM Charger Datalogger EBCM Rx ECM ECM PTE Controller 103 NewFeatures PRND IGN Pedal Faults TCM Tx PROPRIETARY AND CONFIDENTIAL THE INFORMATION CONTAINED IN THIS DRAWING IS THE SOLE PROPERTY OF WOODWARD ANY REPRODUCTION IN PART OR AS A WHOLE WITHOUT THE WRITTEN PERMISSION OF WOODWARD IS PROHIBITED 2009 BER BPCM ECM HCP TCM Custom bus Po BMS Rx Charger Rx TIM Rx Description FinalController Foreground EBCM Rx Tx Dstslogger Ios HV Controller Pedsl Faults NewFestures PRND_IGN Figure 85 Main program Woodward 1000 East Drake Road Fort Collins CO 80528 mcs woodward com 866 588 0494 Path FinalControllerriForeground REV 027 MotoHawk RTW Code Coverage Test Bit Appendix J 196 Prose PRND Prope PRND B Prose C Prope PRND D Probe E 1 5 Fale 0 059 Prove_ParkOut Data Wrte AND EST ver pu Fan L3 g ReverseOut R Data Wrne EST2 var EX lt lt gt Z pu eese Prooe NeutralOut NOT Prooe DreOut Data Wre AND EST3 pnel Figure 86 Shift Lever Decoder Model pu 8 197 Pang Reverse Neutra umts ums H V V
74. Highway FTP 2 UDDS 6 60 mph until 30 SOC UDDS 7 UDDS 8 60 mph until depletion Chapter 6 132 Similar to the day prior the battery was again charged upon dynamic test completion Due to time constraints caused by the interdiction of charging overnight the vehicle was only charged to approximately 3096 SOC 6 2 4 3 Day 3 For the last day of testing an additional current clamp was installed to verify earlier results but no significant difference was found However the focus on that 34 day of testing was on calibrating the regenerative braking parameters to obtain a smooth blending with the hydraulic brake system and a smooth deceleration while coasting similar to an internal combustion engine vehicle To achieve a natural throttle response the team engaged the help of the professional EPA driver in tuning a progressive regenerative braking level on the fly to a point where the car naturally coasts to a stop as the driver releases the throttle and with very light application of the brakes 1 before the hydraulics activate This makes the car almost drive itself down the UDDS hills to a stop giving a very natural braking performance while recharging the battery The testing program during day 3 ended about 1 hour prematurely while the vehicle was being run with more aggressive REGEN braking levels The ground fault indicator GFI went off while the car was holding during a 30
75. ID ID mask payload filter and payload mask The data length code DLC can be between 0 and 8 and describes how many bytes of the data field are used for a given message Messages can be sent when triggered by an event such as a remote frame or can be transmitted periodically The periodic interval value is commonly expressed in milliseconds Filtering CAN communication follows a producer consumer model where all the nodes are masters Messages give no information on the destination consumer only on the source producer 30 Therefore each node must read all the messages on the bus When a message is sent produced by a node it is up to the other nodes on the network to decide whether to receive it consume or to ignore it In other words nodes need to sort or filter the messages they are interested in receiving When using microcontrollers having CAN integrated features also called CAN chips filtering is done by hardware using CAN buffers and by software using a custom dispatcher In microcontrollers without specific CAN buffers sorting is accomplished by software only and is less efficient The sorting process has the two following filtering stages e ID filtering using the ID and the ID mask e Payload filtering using the payload filter and the payload mask In the first case messages are filtered according to their identifier The filter is the identifier mask defining which bits of the identifier to care about Th
76. IT Pre Competition Report UOIT EcoCAR Report 4 University of Ontario Institute of Technology 2009 6 Hybrid Center page consulted in October 2013 Website http www hybridcenter org hybrid center how hybrid cars work under the hood 2 html 7 Greg Rohrauer Pierre Hinse Helen Qin Mike Maduro Year 1 UOIT Final Technical Report UOIT EcoCAR A Full function Electric Vehicle Design with 400KM Range University of Ontario Institute of Technology 2009 8 Greg Rohrauer Pierre Hinse Helen Qin Mike Maduro Joseph Brennan Gavin Clark Hugo Provencher EcoCAR Challenge Year 2 Progress Report 3 University of Ontario Institute of Technology 2010 9 Gavin Clark Samantha Hazell Year 3 Technical Presentation Presentation University of Ontario Institute of Technology June 2011 10 Axsen Jonn Burke Andrew Kurani Ken Batteries for Plug in Hybrid Electric Vehicles PHEVs Goals and State of Technology circa 2008 Institute for Transportation Studies University of CA Davis May 2008 11 Cell Specification Data SLPB 160460330 Kokam Co Ltd 12 REAPsystems Ltd Manual release 1 02 BMS 12C s w version 0 12 www reapsystems co uk 13 BRUSA USER s MANUAL Battery Charge NLG5 02 2003 NLG5xx_108_bis 24 4 www brusa biz 14 Rating plate on the Delphi S10 EV motor 15 MES DEA Traction Inverter Module TIM300 TIM400 TIM600 User Manual Installa
77. Mike Maduro Lesley McLelland Dr Greg Rohrauer Hugues Marceau Helen Qin and Shawn Sandham It was a great pleasure and an honour getting to know you and working with you Eco Fellows A huge thanks to Dr Greg Rohrauer my thesis adviser for trusting me and believing in my electrical engineering skills and also for giving me the opportunity of building an electric car and high responsibilities inaccessible to young engineers outside academia I am indebted to Gavin Clark who played an important role during my graduate studies as a friend and also as a dedicated colleague who spent hours proof reading my EcoCAR reports literature reviews and this thesis It is an honour for me to have worked with and learned from Pierre Hinse which I consider a knowledgeable mentor He has made available his support in a number of ways in the design and implementation of the UOIT EcoCAR I would like to show my gratitude to Amanda Kalhous for being an awesome mentor first as a GM mentor for the EcoCAR project and after helping me joining her at GM as a life and work mentor Thanks to my family to whom my work might be a mystery but still supported me despite the distance between us I would like to thank Dr Richard Marceau for his advice on the writing process of this thesis Finally thanks to all the people who worked with me closely or sporadically on the EcoCAR Challenge vi Table of Contents Author s Declaratio custos
78. OLING FAN ENABLE OUT6 PIN 32 REGEN ENABLE PIN IO INIO pe BRAKELIGHT _OUT9 INI3_ PIN 36 BRAKE CONTACT IN2 6 26 IN3 zt 12 PIN 19 PIN 12 INI TRACTION ENABLE 5Kohm PIN 33 REC _ PIN 40 IGNITION KEY KEY a V_POT__ PIN 27 PIN 31 POWER PIN 13 5 Kohm POT PIN4I PIN 15 FUSE PIN 14 PIN 2 PIN 42 5 Kohm gt PIN28 Figure 39 TIM I Os 15 Chapter 4 92 Table 43 TIM I Os Description 15 RS232 Configuration of TIM s software CAN Rx Receive commands from the MotoTron Tx Broadcast data to the MotoTron Cooling fan enable Cooling fan activated when needed Brake pedal Directly connected to the stock brake pedal Accelerator pedal Directly connected to the stock accelerator Gear selector shifter lever Park Reverse Drive and Economy mode Hand brake contact Activated when hand brake engaged AC net connected Charger connected Regen enable Always enable PWM safety cut off is used if battery at full charge Brake contact Activated when brake pedal is pressed Emergency Shut off HV output when activated Traction enable Activated when on drive or reverse gear Ignition key Activated by the stock ignition key Voltage sensors Measure the battery voltage Current sensors Measure and limit the battery charging discharging current Temperature sensors Measure internal and external temperatures PWM safety
79. PHEV can charge from the grid grid dependant whereas mild full hybrid electric vehicles HEV do not grid independent Extended range electric vehicles EREV use a full electric propulsion for a limited range and revert to hybrid operation for the extended range Hydrogen fuel cell hybrids also fall into this category It is to be noted that PHEVs can be considered as specific category of HEVs having a stronger electric drive and the ability to charge from the grid Similarly EREVs are PHEVs with a motor powerful enough to sustain the electric mode regardless of the vehicle speed as long as the battery holds charge 3 2 2 2 Technologies To increase fuel efficiency hybrids commonly use the following 3 advanced technologies 2 Automatic engine idle stop The engine is stopped at low speeds and started back up during faster acceleration This feature is possible due to a powerful electric motor driven by the high voltage battery Regenerative braking The vehicle kinetic energy is recaptured and transformed to electrical energy by the electric machine and power inverter and stored in the battery during deceleration e Efficient ICE operation electric motor drive assist In parallel or power split HEVs the ICE be decoupled from the axle while the electric motor still powers the vehicle in an attempt to keep the ICE operating only in efficient regions Chapter 2 18 2 2 3 Powertrain Configurations As shown on Figure
80. S to shut off the charger s HV cut off output Charger enabled Enable signal is sent to the BMS and TIM when charger is output ready to charge using the MotoTron PWM safety cut off not implemented in UOIT s design see Section 4 4 2 This charging system consists of five BRUSA chargers connected in parallel and each charger has an individual charging profile Initially it was planned to connect the chargers in a master slave configuration to allow the BMS to send commands over the UOIT CAN bus to the master which would then control the slaves However being limited to 3 chargers this configuration was changed to a parallel one when additional chargers were added to facilitate a faster charge time to meet VTS charging expectations The maximum charging power each BRUSA can deliver is 3 3 kW therefore a maximum charging power of 16 5 kW can be supplied to the batteries thus 47 4 A continuous at the nominal battery voltage of 347 8 Vac The acceptable grid voltage range is single phase 100 250 Vac and current limits can also be set to avoid tripping breakers The power plug on the vehicle at present is a SAE J1772 80 A and adaptable via the power cord to whatever supply is available for charging The AC pilot line is constantly read by the VIM however the PWM pilot line limiting the power flow is held static and fixed to 100 using a diode and resistor Each charger unit monitors the grid voltage and current the battery v
81. SA Chargers 5 J wo 22 45 A C Compresso Front ESS Thermoelectric unit Frame Rail Encapsulator rcd 2i r amp 1 4 12V Battery Underside ESS Thermoelectric unit Figure 6 EV ECUs in the UOIT EcoCAR EV Prototype 9 Rear ESS Chapter 2 18 2 7 1 Battery Batteries are known as the main detractor causing a late arrival of electric cars on the market At first heavy lead acid batteries were tried then it was nickel cadmium NiCd followed by nickel metal hydride NiMH with a higher energy density but still too heavy for automotive applications Today lithium batteries are the main prospect for EVs PHEVs and HEVs However they are fragile expensive and flammable Therefore they require a protection and management system known as a battery management system BMS Since they re also a source of energy for EVs and PHEVs it s essential for a BMS to monitor them and send information to the driver This subsection presents the battery cells and battery management system BMS selected for the EcoCAR project at UOIT 2 7 1 1 Battery Chemistry Lithium Polymer The main criteria for the battery were highest volumetric and gravimetric energy density Although 18650 format cells as used in laptops and also used in Tesla products have the highest energy density by both gravimetric and volumetric measure they are extremely difficult to integrate due to the quantity required The approach of using the largest singl
82. T was amongst the 16 universities across North America awarded participation in the EcoCAR Challenge Selection was based on a competitive RFP to which over 50 top engineering institutions responded across North America During the first year of the challenge 2008 2009 collegiate students had to design and simulate their greener powertrain The following year 2009 2010 competing universities were given their vehicle a 2009 GM 2 mode hybrid Saturn VUE They had the year to implement their design before competing in dynamic events at the second year competion taking place at GM s Proving Ground in Yuma Arizona The third year of the challenge 2010 2011 was intended for testing calibrating and optimizing the vehicle s new and greener architecture before the final competition taking place at the Milford Proving Grounds in June 2011 The main sponsors and organizers of this competition were the US Department of Energy DOE General Motors GM and Argone National Laboratory a division of the DOE From the sponsor standpoint the primary objective was to prepare the next generation of engineers for careers in the green vehicle industry 1 The UOIT EcoCAR prototype is shown on Figure 1 Chapter 1 1 Figure 1 UOIT EcoCAR Prototype the Climatic Wind Tunnel at the Automotive Center of Excellence 1 2 Previous Research To satisfy the constantly growing demand for green vehicles and comply with new government policies
83. T EcoCAR team selected a full electric powertrain while all the other universities taking part in the competition chose an architecture with 2 energy sources i e mostly plug in hybrids The different levels of hybridization are first detailed Then typical fuels are listed and their well to wheel WTW influence is analyzed After having established the powertrains and fuels available component sizing and vehicle performance simulations were run To conclude this chapter the components selected to implement the chosen architecture are described 2 2 Powertrain Selection and Configurations 2 2 1 Degree of Hybridization The degree of hybridization of a vehicle can be defined by its powertrain electrification level in other words the capacity of the energy storage system ESS and the electric motor powertrain In Figure 2 the different degrees of hybridization are categorized in ascending level of electrification Non hybrid ICE vehicles Mild hybrid Series Full hybrid gt Parallel e PHEV e Full electric FFEV BEV Figure 2 Degree of Hybridization Power Split Chapter 2 7 In Figure 2 the expression non hybrid refers to internal combustion engine ICE vehicles Vehicles having a single electric propulsion source are deemed full function electric vehicles FFEV also called battery electric vehicles BEV In between ICEs and EVs are various levels of hybrids Plug in hybrid vehicles
84. a message They are the actual information i e they can represent a voltage a current a state of charge a temperature a speed a status errors warnings etc The start bit of a signal is usually its 156 when discussing CAN bus however this is not always the case Therefore when positioning a signal in a message the start bit used must be specified as the e Jsb ofthe signal e msb of the signal If a signal is defined using the Isb as its start bit the bit wise progression is towards the left but the byte wise progression depends on the byte order 3 5 2 5 Display Formats This section details 6 display formats for messages using the terminology explained in the above sections In an attempt to further clarify these display formats a column describing visually the bit wise and byte wise progression is added to Table 25 It summarizes the 6 common display formats in CAN communication expanded from 47 45 49 Chapter 3 50 Table 25 Display formats Bit progression Display Byte Order Bit Msg Ser from the start bit to full Format Numbering Progression of signals length S ee d Little endian Sawtooth Forward Isb 2 Hi ids Little endian Sequential Forward Isb eee id um Nun dios b Big endian Sawtooth Forward Isb Pus n E n ieri Bien sewn Fw me piwie nc Big endian Sawtooth Backward Isb avs re S EL dor Big endian Sequential Forward msb Rut r
85. able 35 Removed and Remaining 95 76 Table 36 Mam BCUS After Conversion e uie eie e RR ER eve Dna su ERR eH IRR Ru RIS EUR NN 76 Table 37 Modified HS and UOIT buses Hp e 77 Table 38 Description of Powertrain Control Strategy by Operational 80 Table 39 Power States of Main Controllers by Operational Modes 80 Table 40 BMS I Os Description 12 87 Table 41 OBCM I Os Description 13 PA as 90 Table 42 Charging Phases Sete 91 Table 43 1 Description 15 93 Table 44 Resettable Crash Sensor Specifications 97 xvi Table 45 Comparison between ECM 0555 and 5554 103 Table 46 Transmission Shift Lever Encoder 108 Table 47 Emulated EGUS cubo on 115 Table 48 Dynamically Emulated CAN Signals eene 117 Table49 Test Schedule at sore eed e 131 Table 50 Drive Schedule Sequence 132 Table 51 EPA Energy Consumption Results Battery Depletion 138 Table 52 EPA Energy Consumption Results Battery Depletion Summary 138 Table 53 EPA Energy Consumption Results Regenerative Braking Activated 138 Table 54 EPA E
86. age controller manages the high voltage strategy using a Stateflow controller PRND IGN decodes the transmission shift lever position and the ignition key position Pedal Faults detects and mitigates faults on the acceleration and brake pedals whereas the NewFeatures subsystem controls other features such as the front radiator fans Chapter 5 105 5 4 Encoders and Decoders 5 4 1 Motor Controller Controls As seen in section 4 5 3 numerous inputs are required by the motor controller to operate Where the acceleration and brake pedals are directly wired to it the other features are controlled by the VIM Using 0 to 5V TTL outputs the MotoTron ports couldn t be directly connected to the motor controller Therefore a voltage translator was added between the VIM and the motor controller to achieve voltage boost to 12V Figure 42 is a high level schematic of the controls configuration between the VIM and using a voltage translator and the UOIT CAN bus Park Park Reverse Reverse Drive Drive Hand Brake Hand Brake AC Net Connect Ed AC Net Connect VER Regen Enable Regen Enable i Brake Contact Brake Contact Motor Module Traction Enable Traction Enable 2 UOIT CAN bus A Other ECUs Figure 42 Motor Controller Controls 5 4 2 PRND and Ignition 5 4 2 1 PRND The shift lever is connected through a cable to the motor transmission switch encoder which also m
87. aneously and are off at the same time The voltage drop through such a transistor and a diode is about 1V to 1 5V depending on their impedance and the current going through them Assuming a supply voltage of 5V when the transistor connected to CANH 15 conducting CANH is pulled up from 2 5V to only 3 5V due to a 1 5V drop between the 5V source and CANH caused by the transistor and the diode When CANH is pulled up CANL that was idling at 2 5V is pulled down and reaches 1 5V due to a voltage drop of 1V across its transistor and diode The diodes between the lines and the transistors are also protection against high voltage transients The receiver is basically a discriminator circuit having as inputs the CANH and CANL wires and for output the digital receive signal A discriminator circuit compares the voltage between its two inputs and sets its output to 1 when both signal are the same and to 0 when they are different In other words the digital output is based on differential voltage levels between the lines Since interference is generally induced almost equally in both twisted wires CANH and CANL and knowing that the receiver reacts to a difference in voltage level between the two lines this circuit provides high noise immunity Appendix E 171 Several other features are implemented basic CAN transceivers 41 The TXD dominant detect block is used for ground fault protection equivalent to transmitting only dominant bits
88. aning of programmable analog output 1 63 64 _ current velocity C17 Meaning of analog input A I 1 14 bit O speed ref 1 E BEREE _ torque ref 2 torque limit ref 0 2 C32 Motor thermal switch Block drive Motor thermal switch Block drive O1 Motor thermal switch Block drive O1 46 Enable temperature sensor of 026 1 Thermal _ switch 22NTC 3 TS NTC CAN Baud rate C48 021M 12500K 22250K 3 125K 44100K 5450K 6 20K 7 10K C52 CANbusenable 0 disable 1 enable 56 of Overload 0 120 x 30 sec 1 150 x 30s _ 2 200 30 3 200 3 155 30 64 Enable Current Control C69 Enable second order filter on speed governor 01 1 0 C74 Enable managing incremental encoder overtime 1 0 _ 1 75 Disable autotuning starting from the defaults Appendix I 192 5 CANLog4 USB Interface System CANITiming Timing500K CAN2Timing Timing500K CAN3Timing Timing500K CAN4Timing Timing40K CAN 1Output OFF CAN2Output OFF CAN3Output OFF CAN4Output OFF Output LEDI 1 0 LED3 0 LED4 1 Recordfilter exclude CANI 000 7ff exclude CAN2 000 7ff exclude CAN3 000 7ff exclude 000 7ff include CANI OFA CANI CANI OFC END Figure 83 Log Task Language LTL Code Programmed in the CANlog4 Appendix I 193 1 VIM Port
89. antenna In order to reduce the electrical noise the two 120 termination resistors should be placed next to the 2 farthest controllers leaving only short leads between the resistors and the controllers thus eliminating the two branches acting as antennas This method is still considered a star topology It retains simplicity and flexibility for the addition or removal of devices on the bus Figure 27 gives an example of such a topology A bus topology can be defined as nodes connected by passive links through a single cable allowing transmission in both directions 40 Hubs are generally used on the main bus to easily add and remove devices They can be passive or active Since the length of a CAN bus is short in automotive applications hubs do not need to regenerate or amplify the bus signal therefore only passive hubs will be considered Active hubs require power and act as repeaters Chapter 3 58 A star topology is a network where all the nodes connected to a central one 40 The addition of a hub on a bus configuration creates a node to which several other nodes are connected This distribution node is therefore a centralized point for these other nodes modifying the architecture of the network When a hub is inserted to a bus structure the network architecture becomes a combination of the bus and star topologies Networks using more than one topology are called hybrid networks A CAN network using one or more hubs as sho
90. arious reasons such as competition restrictions availability or cost A typical EV is composed of a motor an energy storage system ESS a battery charger a DC DC converter high voltage control module and a vehicle integration module VIM This collection of devices requires a sophisticated controller in order to function properly Motors used in EVs are not simple DC motors They are usually powered by a 3 phased high voltage AC source However the energy storage system provides DC high voltage thus a motor controller also named traction inverter module is required to effectively convert DC voltage to 3 phased AC The ESS is composed of a battery and a management system BMS that monitors the current voltage state of charge temperature etc A battery charger is required to convert the AC from a power outlet to DC current at the battery A high voltage control module HVCM is in charge of the high voltage safety and the power flow paths by controlling contactors on the high voltage bus located in the high voltage distribution box A vehicle integration module VIM is useful for doing general controls and for bridging devices with one another An EV does not have an alternator to power the accessories in the car thus a DC DC converter must be added to the vehicle For high voltage safety a ground fault interrupt GFI monitors current leakage between the accessories grounded to the chassis and the floating high volta
91. at or 29 bits extended format while the data length can vary between 0 and 8 bytes The bitwise arbitration using the identifier ID gives a static message priority to the protocol This property is possible because of the binary logic used by the protocol either dominant logical 0 or recessive logical 1 When a dominant and a recessive bit are simultaneously transmitted the dominant bit supersedes the recessive one 22 see Appendix B This protocol uses a producer consumer multi master message model instead of the more common client server model Its principal characteristic 1 the interpretation of the identifier in a message 25 It does not inform about the destination of the message but rather indicates where it was sent from in other words providing the source of the data Chapter 3 37 CAN messages be transmitted periodically on request or a state change at a uniform and fixed bit rate up to 1 Mb s within a CAN bus or CAN network As explained in Section 3 2 3 the speed or bit rate can have different values in different networks A similar concept applies for the message format standard or extended The CAN communication protocol shows configuration flexibility by requiring no hardware or software modification for any node when a new node is connected to the CAN network The quantity of nodes on such a network is theoretically unlimited however delay times and electrical loads will place an empirica
92. at the Environmental Protection Agency In March 2011 the UOIT EcoCAR team and the competing universities were given the opportunity to test their prototypes for 3 days at the National Vehicle and Fuel Emission Laboratory EPA Although the general test methodology followed at the EPA facility was similar in some ways to what was normally done at the university garage the additional instrumentation dynamometers and certified drivers provided the unique advantage of testing in a controlled environment The UOIT test schedule at EPA is detailed in Table 49 Table 49 Test Schedule at EPA Motor Controller Calibration Day 1 ES Charge Monitoring to 100 SOC Us Range test without regen to 0 SOC a Charge Monitoring 0 to 30 SOC Regenerative Braking Calibration Day 3 Battery Current Monitoring 6 2 4 1 Day 1 Initially the vehicle underwent experiments on the dynamometer with different motor controller tuning parameters that had been initially set at the university garage Different steady state speed power consumption and acceleration runs were performed for each parameter under study After analysis the team converged on a set of motor tuning calibrations for subsequent tests Also a number of full throttle runs were performed to evaluate the vehicle acceleration performance Once the inverter calibration was completed the vehicle was unstrapped from the dynamometer and was brought to the charging
93. b Simulink 2009b MotoHawk 2009b ML7p5 to ML7p9 exe MotoServerRuntime RELEASE 8 13 7 140 exe MotoTune RELEASE 8 13 7 140 exe gcc powerpc eabispe 4 4 0 SPl exe kvaser drivers w2k w7 exe Appendix J 205 Appendix Instrument Panel Cluster Figure 96 Custom Faceplate of the Instrument Panel Cluster without the 2 display screens Appendix 206 These are hidden warning lights unless illuminated Warning lights requiring All dials are visible with and backlighting to be visible without backlighting from the rear with an LED Symbol is always Warning lights Warning lights visible but does not requiring backlighting requiring backlighting have any backlighting to be visible to be visible Figure 97 Custom Faceplate of the Instrument Panel Cluster without the 2 display screens Appendix K 207 Figure 98 shows what the faceplate looks like when no warning lights are these areas are visible during the daytime when the lighting is not effective and also when the backlight is ON at night As for Figure 97 the 2 display screens are not represented on Figure 98 Figure 98 Custom Faceplate of the Instrument Panel Cluster Warning Symbols Hidden without the 2 display screens Appendix 208 Figure 99 Stock Faceplate of the Instrument Panel Cluster Appendix 209 UOIT bus dictionary Table 72 UOIT CAN bus Dictionary Appendix L BMS Tx
94. ble and or removable This programming method has the advantage of requiring no modifications to the vehicle s hardware and can generally be done using the OBD II port 3 6 1 4 Gateways between CAN buses As mentioned earlier in this section it is not uncommon to find multiple CAN buses interconnecting electronic control units ECUs within a vehicle The need for having multiple CAN environments comes from the constantly increasing number of ECUS requiring more information thus using more bandwidth 24 When the bandwidth of a CAN network is saturated it is common practice to add a new CAN bus to allow the addition of new controllers and features These new ECUS present on the new CAN environment might need information transmitted on another CAN bus This is why most ECUs have multiple CAN ports i e 2 to 4 and some ECUS are used as gateways between different CAN buses Gateways are also required to interconnect ECUs from CAN buses using different high speed frequencies or to interconnect ECUs from a single wire bus to a high speed one Chapter 3 65 3 9 Summary This chapter provided overview on CAN protocol fundamental theory a description of the hardware required to create a CAN environment lists of devices supporting CAN available on the market a rigorous definition of the nomenclature used to accurately define CAN messages and CAN signals Different CAN bus configurations as well as different techniques for using identif
95. cessive bits as the overload delimiter are sent End of Frame J OVERLOAD FRAME 9 Zik Nan A Inter rror Delimiter or Frame Overload Delimiter Space or Overload Flag Overioad Frame superposition of Overload Flags Overload Delimiter Figure 58 Overload frame 22 Appendix B 165 Appendix CAN Physical Waveforms Below is an example of a CAN high speed transmission showing physical bits on a CAN bus and therefore the NRZ coding method Recessive bit Dominant bit Recessive bit Figure 59 High speed CAN bus waveform ISO 11898 2 CAN L Recessive bit Dominant bit Recessive bit Figure 60 Fault tolerant CAN bus waveform ISO 11898 3 Appendix C 166 4 5 3 6 1 8 0 9 single wire Recessive bit Dominant bit Recessive bit Figure 61 Single wire CAN bus waveform SAE J2411 Appendix C 167 Appendix D Standard CAN Connectors There are 2 standard types of connectors to access a CAN bus DB9 and 5 pin M12 respectively Figure 62 and Figure 63 Pin Signals Description 1 Low 2 CAN L dominant level CAN GND C AN L m 3 CAN GND Ground CAN H 4 CAN GND 5 6 CAN GND uen CAN Vct High 7 CAN_H dominant level 8 Power 9 CAN_V optional Figure 62 DB9 connector 3 Ground 7 2 12V Power 1 Shield Ground A position Figure 63 5
96. communications between the vehicle s major components Vehicle Integration Module VIM Battery Management System BMS On Board Charger Module and Traction Inverter Module TIM 6 2 2 Performance Testing at UOIT The previous section described the process followed and tests performed on the prototype before and during the integration This section focuses on some of the tests realized with the integrated vehicle The methodology used to safely test the vehicle was as follows the motor tuning and controls were tested with the vehicle off the ground thereafter short stints of mild driving acceleration and braking were done followed by longer tests at higher speeds maximum power output and sustained duration running For safety the Vehicle Integration Module VIM was generally re flashed with the vehicle off the ground to avoid any unintended acceleration and preliminary verifications performed then The motor tuning was one of the first powertrain calibrating attempts done on the integrated vehicle Having it on a hoist with the wheels free spinning the motor was initially tested with the same in house control panel used for preliminary verification and debugging on the test bench before controlling it via the VIM The team ran the inverter s built in auto tuning function a few times until successful completion of the test sequence After connecting the acceleration pedal to the inverter instead of using the potentiomet
97. crocontroller CANL Transceiver ax Controller CANH a X ED Figure 23 CAN Transceiver CAN Protocol Controller and Controllers Table 17 Commercial devices CAN CAN Protocol Microcontrollers Commercial Transceivers Controllers with CAN Controllers High Speed High Speed Microchip MotoHawk MCP2551 HS e B dsPIC33FJ256GP710 ECM 5554 112 TJA1041 HS SJA1000 HS PIC24HJ256GP610A ECM 0555 080 TJA1050 HS Atmel ECM 0565 128 MAX3050 HS 128 Instruments MAX3057 HS 89 51 03 6660 HS Vector Semiconductor Fault Tolerant MPC5554 CANlog4 TJA1054 LS MPC555 CAN Case XL MC33388 LS MPC565 Others Single Wire CAN AC2 PCI TLE6255G SW CANview USB Chapter 3 44 3 4 Use of Identifiers In a CAN network a technique sometimes used to easily identify a device is the division of the identifier into two fields the device identifier and the message identifier This can be useful when several similar devices are connected on the same network Table 18 shows an example of a Tritium WaveSculptor 58 configuration using a 6 bit device identifier of 0b100 00 xxxx and 5 bit message identifiers Table 18 Device and message identifiers Motor Controller TX ID 0x400 Message Identifier sd Device Msg re Ki Hen ig pa gm m cm _ n SerialNumber Some devices use
98. ctor ESS Moto 31 Wht LSO5 LSO5 Discrete Ouput DC DC1 DC DC2 Moto 32 Blu Wht 1506 1506 Charger Moto 35 LSO9 LSO9 PWM Ouput Radiator Fans Appendix J 194 2 Simulink Model Model Browser x PROPRIETARY Woodward AND CONFIDENTIAL 1000 East Drake Road Fort Collins CO 80525 APM_PTE THE INFORMATION CONTAINED IN THIS mcs woodward 866 588 0494 331 BMS Rx DRAWING IS THE SOLE PROPERTY OF Description WOODWARD ANY REPRODUCTION IN i BH BPOM PART OR AS A WHOLE WITHOUT THE Fina ontroser 3j Charger WRITTEN PERMISSION OF WOODWARD Top level root of model FinalController mdl 2H Datalogger IS PROHIBITED 2009 _ Path FinalController _ HCP HCP PTE MotoHawk CAN Definition 2 HV Controler FinalController le Name CAN 1 10 FinalController mdl rr El NewFeatures 2 IGN Copyright 2009 All Rights Reserved patch crac 2 Pedal Queue 16 messages gt MotoTune Protocol Enabled City ID 0 0 PCM 1 TIM Tx Main Power Relay On Delay 100 MotolGwk Off Delay 250 Target ECM 5554 112 0904 xD DEY Floating Point single 32 bits Stacks FGND 3072 BGND 2048 IDLE 1024 IRQ 1536 Heap Size 4096 DLL Filename FinalCont 027 SRZ Filename FinalController 027 MotoHawk CAN Definition Name CAN 2 Bus CAN
99. cut off Input controlled by the BMS to shut off the TIM s PWM lim power HV output Light reverse Activated when on reverse gear PWM safety cut off not implemented in UOIT s design see Section 4 4 2 The reverse light built in function of the inverted is not used but the vehicle reverse and brake lights dynamically emulated by the VIM The vehicle acceleration and brake pedals have their respective potentiometers directly hardwired to the motor controller input pins throttle and power limiter The accelerator position commands the torque output of the motor when the gear selector is not in PARK position The brake pedal position governs the regenerative braking amount blended to the hydraulic braking when regenerative braking is enabled and brake contact is activated The VIM controls these 2 inputs along with the GEAR selector HAND BRAKE CONTACT AC NET CONNECTED EMERGENCY TRACTION ENABLE and IGNITION KEY The TRACTION ENABLE input of the TIM is the main enabler of the powertrain and as explained in Section 4 4 is managed by the HVCM residing inside the VIM Chapter 4 93 Similar to the BMS and charger 232 interface is used to program and calibrate the motor controller the CAN communication provides information on the motor and the drive itself and sensors read the battery voltage and current along with the temperature of the TIM 4 6 Safety Control Strategy 4 6 1 Acceleration Fault Manage
100. d sending CAN commands The 3 phase induction motor is hooked directly via three short and shielded 1 gauge wires The vehicle HV inlet is compliant with the SAE J1772 standard and located behind the front license plate which is mounted on a sprung hinge From the power inlet the 240 Vac flows to a breaker box splitting the input power between the 5 BRUSA chargers using 8 gauge wire individually protected by 240 V 15 A breakers however only the Chapter 4 83 240 Vac 80 A breaker is shown on Figure 36 In addition to the charger internal fast acting 15 A fuse the DC output of each charger is protected by 600 V 20 A fuses and merged using a positive and a negative bus bar From the positive bus bar 8 gauge wire is connected to a normally open Tyco LEV 100 contactor independently managing the charger s logic and the HV DC connection between the chargers and the vehicle s HV line As detailed in Appendix I 3 operation modes are available on BRUSA chargers auto master booster and CAN Due to using 5 BRUSA chargers a master booster configuration was not possible Although the CAN identifiers used by each BRUSA charger could have been modified using RS232 communication as shown in Appendix I CAN control of multiple chargers would have required designing complex controls within the VIM However this would have allowed the VIM to monitor and control each BRUSA charger individually and dynamically For simplicity and as a competition time
101. device specifications The preclusion of disabling the CAN receiving command while keeping the status transmission activated caused delays in the control strategy implementation and forced the author to find temporary solutions to some issues for the second year competition and testing at EPA Vehicle speed determination and motor information recording and monitoring were affected The importance of creating and respecting a predefined timeline cannot be emphasised enough According to the author a couple months of delay in the battery build process and the inconvenient timing of the testing at the EPA facility resulted in a lack of time for thoroughly testing the new controls and high voltage systems Consequently the UOIT EcoCAR prototype was not ready on time to compete in dynamics events at the final competition resulting in component failure and an overall 13 place finish The last results from dynamic testing were gathered at EPA a few months prior An impressive range of over 400 km was demonstrated on a dynamometer with a temporary Li ion battery pack of 74 6 kWh while the final battery has a capacity of 83 5 kWh and an estimated range of 482 km with REGEN activated Also the vehicle 0 60 mph acceleration is estimated at 14s limited primarily by the drive motor assembly permitted to the team The available ETX 101 unit is the preferred choice for continued work Chapter 7 149 7 3 Original Contributions The list below organi
102. duction 2009 2 mode hybrid Saturn VUE In chapter 2 the architecture selection process that led the UOIT EcoCAR team to choose a full electric powertrain for their vehicle prototype along with listing the main powertrain controllers selected was described Chapter 3 probed deeper in the theory behind the most important in vehicle communication protocol used today the controller area network CAN The electrical and control system integration strategy was then detailed in chapter 4 An exhaustive study of the model based algorithm programmed in the vehicle integration module VIM which is the controller at the heart of this integration strategy was provided in chapter 5 Chapter 6 discussed the tests and results of the vehicle conversion to full electric propulsion both at a general team level and individually for the author s specific accomplishments and contributions 7 2 Main Conclusions The first main conclusion is that the tight packaging of the 83 5 KWh in house built Li ion battery in an existing chassis was the primary accomplishment of the UOIT EcoCAR team The competition organizers were among the skeptics thinking this could not be achieved with a battery having enough capacity to meet the range requirement of the competition The UOIT team proved them wrong by exceeding the competition range requirement when doing a complete battery depletion test at the EPA facility Building and packaging a battery of this magnitude while dealing wi
103. e An estimation of the energy consumption for the final parameters is also present in the aforementioned table In Table 54 the energy savings with REGEN activated was estimated at 896 for the highway and 0 for the steady speed of 60 mph since no brake are applied In summary activating regenerative braking increases the extrapolated range by 44 8 km for a maximum range of 402 km as stated in Table 55 Figure 49 illustrates some results of the energy regenerated One should note the smooth transition in REGEN current every time the vehicle comes to a stop Chapter 6 137 Table 51 EPA Energy Consumption Results Battery Depletion Drive miles Cumulative Energy Energy for consumption EXtrepotated Estimated SOC Schedule Driven Use KWh Schedule KWh Wh mi Range mi UDDS 1 7 443 2 874 2 874 386 1 198 18 SOC Start 96 0 UDDS 22 7 450 5 688 2 814 37757 202 60 92 6 9005 3 7 437 8 440 2 752 370 0 206 80 89 2 UDDS 4 7 442 11 260 2 820 378 9 201 95 85 9 5 60 46 118 26 236 14 976 324 7 235 66 82 6 9 UDDS 5 7 430 29 020 2 784 374 7 204 23 64 7 8 Highway 10 243 32 032 3 012 294 1 260 24 61 4 gt Highway 10 244 34 984 2 952 288 2 265 56 57 8 UDDS 6 7 435 37 728 2 744 369 1 207 35 54 3 a 60 mph 60 533 56 663 18 935 312 8 244 64 51 0 UDDS 7 7 433 59 324 2 661 358 0 213 76 28 5 100 8 7 407 62 039 2 715 366 5 208 78 25 3 60 mph 46 352 76 526 14 487 312 5 244 85
104. e testing at EPA and the final competition resulted in shipping an unfinished vehicle to the event Numerous hours were spent at the competition at the Milford Proving Ground finishing the two battery pack modules leaving minimal time for troubleshooting the newly made high voltage distribution module and its high voltage controller the vehicle integration module When the organizers gave the ultimatum to the team the ignition of their newly modified vehicle was turned on prior to an exhaustive testing of each component in the new high voltage system resulting in the explosive failure of the main precharge resistor Therefore the UOIT full function electric vehicle prototype did not run during the very last competition of EcoCAR The NeXt Challenge Not competing in the dynamic events severely affected the UOIT score card After the static events taking place in Washington DC the team ended up finishing in 13 place of 15 entries 6 3 2 Environmental Protection Agency Testing Results 6 3 2 1 Day 1 On the first day of testing steady state and acceleration runs were performed with different inverter tunings It was discovered that the vehicle exhibited a surging at around 35 mph i e a noticeable torque peak seen by the vehicle occupants and the supervisory team This caused the wheels to slip on the dynamometer rolls 0 60 mph data indicated acceleration times of around 19 seconds possibly faster but roll slippage is difficult to quantify
105. e Rm Table 26 is a compact form of Table 25 using the 3 main CAN display formats Table 26 Compact form of the 3 main CAN display formats Bit progression Display Byte Order BH Msg star t Bit from the start bit to full Format Numbering Progression of signals length Intel Little endi n bit wise to the left Standard byte wise to the right Decreasing Forward Motorola from left to 156 Forward lsb 2 bit wise to the left Big endian right S Motorola byte wise to the left Backward Backward 3 5 2 6 Examples Representation in the 3 main display formats In this example adapted from reference 45 a signal is represented in the 3 main display formats and is defined by its start bit and length The data length code DLC of the message must be greater or equal to 7 and it is assumed that the start bit 1s the least significant bit of the signal in each of display formats Chapter 3 Specifications Start bit 156 11 Length 9 bits Intel Standard Byte 0 Byte 1 Byte 2 Byte 3 Byte 4 Byte 5 Byte 6 Byte 7 Byte 0 Byte 1 Byte 2 Byte 3 Byte 4 S Lu 5 I 6 Byte 7 63 56 55 EET m Figure 24 Representation of a signal in the 3 main CAN display formats This means that three controllers using the same signal specifications but having three different display formats will send their signal in a completely different space of the data fi
106. e cells of highest energy density for packing simplification component count reduction and volume minimization of the whole ESS was pursued The Kokam 240 Ah cell using NCM lithium ion chemistry met these objectives in the best regard among all the cells surveyed and available at the time Figure 7 compares the different cells analysed and Table 5 reviews characteristics of the Kokam cell chosen Chapter 2 19 10000 Goal Pack Goal 3 Sample Cell Sample Pack Cell Power Density Total 100 Wikg 10 1 0 20 40 60 80 100 120 140 160 180 200 Cell Energy Density Total Wh kg Figure 7 Cell Comparison 10 Table 5 Kokam SLPB 160460330 11 1 cell Final Battery Design of cell 1 94 V min 2 7 V 282 0 V 3 0 V cell Vnom 3 7 V 347 8 3 7 V cell V max 4 2 V 394 8 V 4 2 V cell competition limit 400 V Size mm 325 x 455 16 Operating temperature Charge 0 40 Charge 0 40 degC Discharge 20 60 Discharge 20 60 Max charge current 1C T eon masta charger limit Max discharge current 2C 40 m inverter limit Peak discharge current 480 A 400 A inverter limit Cycle life 80 DOD gt 800 cycles 1500 cycles Weight 4 9 kg 460 1 kg Capacity 240 Ah Energy 0 888 kWh 83 5 kWh Chapter 2 20 The battery design was based on using a single string of c
107. e second type of filtering is used when a portion of the data field is used as a sub identifier Messages are sorted using a mask on the payload named payload mask indicating which bits of the data field to care about Usually the payload mask defines which bits of the data field to use as sub ID allowing the demultiplexing of signals in messages using modes as first explained in Section 3 4 The payload filter is more likely the sub ID of a message but Chapter 3 55 could be a specific payload value It should be noted that the ID is to the ID mask what the payload filter is to the payload mask Since the payload filtering is used to filter the sub IDs of a specific ID the ID filtering is processed first When sub IDs are not used the payload mask is an all pass filter Table 27 presents an example of an ID filtering process while Table 28 details a payload filtering example using sub IDs based on reference 26 Table 27 ID filtering example Filtering Configuration ID 400 ID mask 7 0 Binary Value Results 0b100 0000 0000 Ob111 111 0b100 000 1 0 don t care any value allowed ID acceptance filter Incoming Message ID 403 ID acceptance filter ID 404 ID acceptance filter 0b100 000 0b100 000 0b100 000 0b100 000 Message accepted Message accepted ID 600 0b110 000 ID acceptance filter Filtering Configuration 0b100 000 Message
108. e to reach P221 BRAKE value 7 10 100 s P223 Enable complementary BRAKE logic 0 brake on 0 0 logic P224 Minimum Acceptable value of Accelerator pedal 2 0 2 0 P225 Maximum Acceptable value of Accelerator pedal P226 Threshhold for Throttle Brake transition point 200 200 _ Accelerator P249 Enable Analog Regenerative logic P250 Proportional gain control in overload 100 100 multiplier P251 Maximum IGBT junction temperature in overload P252 Velocity threshold switch more or less filtered P253 Time filter for accelerator pedal C 150 0 150 0 ms P266 CAN Transmit enable TXO 1 1 logic ee P267 Transmit message ID TXO 25 25 268 Periodic transmission rate TXO 7 1000 1000 P269 CAN Transmit enable TXI o ooo do ooo _1__ o OE P270 Transmit message ID TXI 1 24 254 271 Periodic transmission rate TX1 80 50 ms P274 Receive Message ID to stop transmission 785 785 Appendix I 191 UOIT EcoCAR Value Default Value Para Description Enable Rx message for Vmax Battery EN Second order filter time constant for Bus Voltage HENCE ee P277 Time allowance for PWM signal inversion 1000 7500 21 P278 Time to decrease the regenerative BRAKE torque to zero 0 50 0 50 8 15 Me
109. ear 2 Competition After a few days of last minute modifications the full function electric vehicle of the UOIT team passed the qualifying static and dynamic safety technical inspections and was ready to compete in the dynamic events of that second year s competition The event took place at the GM Yuma Proving Ground a facility that provided a great opportunity to realize dynamic testing in a controlled and safe environment Furthermore extra monitoring instruments were added to the vehicle provided by the organizers i e a GPS antenna a display gauge and a battery current sensor interconnected to a small MotoTron controller via a dedicated CAN bus The list of dynamic events performed is as follow e Acceleration 0 60 mph e Acceleration 50 70 mph Chapter 6 129 e Stopping Distance 60 0 mph e AVL Drive drive quality test e Autocross serpentine cone lined course e Towing 1500 Ibs load up a 3 5 grade at 45 mph for 15 miles e Lane Change vehicle safe handling limits Even though a temporary Li ion battery pack was used and installed in the cargo space of the vehicle as shown in Figure 48 the control strategy and most of the powertrain devices in place were that from the final design Therefore the results of these dynamic events gave the team a good preview of the performance achievable by the final version of their prototype Figure 48 Temporary Li ion battery pack Chapter 6 130 6 2 4 Performance Testing
110. eas hardware reliability decreases over time 7 5 Conclusion This thesis documents the state of the art in comprehending architecture choices components selection and controls development in an electric vehicle conversion designed for competition This thesis was also written in the hope of providing a solid high voltage and controls basis to students engineers and even hobbyists working on green vehicle prototypes such as electric vehicles to successfully accomplish their own vehicle conversion This manuscript should also be one of the starting points for future students involved in the continuation of the UOIT EcoCAR project if the latter is re launched one day Chapter 7 154 References 1 EcoCAR The NeXt Challenge page consulted in October 2013 Website http www ecocar2 org ecocarchallenge 2 Miguelangel Maduro WELL TO WHEEL GREENHOUSE GAS EMISSIONS AND ENERGY USE ANALYSIS OF HYPOTHETICAL FLEET OF ELECTRIFIED VEHICLES IN CANADA AND THE U S University of Ontario Institute of Technology 2010 3 Leon Zhou Jeremy Wise Shaun Bowman Curran Crawford Zuomin Dong Design Modeling and Hardware Implementation of a Next Generation Extended Range Electric Vehicle University of Victoria 2010 2011 4 Argonne National Laboratory page consulted in April 2014 Website http www transportation anl gov modeling_simulation PSAT 5 Greg Rohrauer Pierre Hinse Helen Qin Mike Maduro Year 1 UO
111. echanically takes the motor out of park pawl locking clutch Chapter 5 106 lever according to the PRND lever position In order to vary the transmission signal in the controls the output of the encoder originally connected to the 510 engine control module is rerouted to the custom vehicle integration module VIM The VIM decodes the shift lever position translates the position for numerous dynamically emulated GMLAN signals and generates the hardwired inputs required by the motor controller The logic of the encoder is depicted by Figure 43 and Table 46 explicitly details its logic A complex encoding system is used to avoid unintended gear shift Two of the A B C and D encoder inputs need to be high to enable a specific gear along with the ground sensor Additionally a park neutral switch needs to be low to enable reverse or drive modes 6 PIN CONNECTOR A BLK WHT R 3 PIN CONNECTOR FOR H V A B N C PRK L BLU YEL Transmission Range Indicator Switch Park Input GRY PARK 771 INPUT A WHT BLK eter 772 INPUT 773 INPUT C 776 PARITY 470 SENSOR GROUND Et rF NotUsed S10EV PRNDL ENCODER Figure 43 Motor Shift Lever Encoder Chapter 5 107 Table 46 Transmission Shift Lever Encoder Logic P R N D A 1 1 0 0 0 1 1 1 0 0 0 1
112. ed HV output enabled Not Powered HV output disabled Powered HV output disabled Not powered Powered HV output enable Powered Powered HV output disabled Powered or not Functional or not Powered or not HV output disabled Not powered Chapter 4 80 4 4 2 Discharge Control Strategy Having only one energy source the UOIT discharge control strategy is charge depleting 100 of the time to nearly full depletion Normal operation limits are between 5 and 9596 SOC with a reduced power mode below 5 SOC Regenerative braking mode is a normal function of the motor controller and a regenerative brake blending was calibrated for an intuitive brake feel The final configuration of the ESS contains 94 cells nominally 83 5 kWh The BMS maintains the cells in a 95 to 5 state of charge SOC window thus 90 ASOC gives 75 15 kWh useable energy When the battery hits 576 SOC the TIM is switched into a progressively reduced power mode controlled by CAN messages between the VIM and TIM A built in PWM safety cut off to stay within voltage limits is available in the BMS and could have been used to trigger and control the TIM reduced power mode However the team preferred having the flexibility of implementing their custom program in the VIM for a smoother integration It should be noted that the PWM safety cut off could be implemented as a second safety in case the VIM fails A full
113. eds to be updated accordingly However the CAN network of the vehicle is not affected since these 2 CAN ports of the VIM were physically connected together as a temporary solution In other words they physically communicate over the same wires even though the signals originate from different ports Table 37 lists the ECUS present on the 2 high speed buses and Figure 34 shows their physical layout in the respective network A hybrid topology is used for both CAN buses leaving the stock ECUS in their original configuration while the additional ECU s are connected to their CAN bus using a star configuration External 120 2 terminations were added to compensate for the ones removed with the ECM Table 37 Modified HS and UOIT buses GMLAN HS UOIT CAN BCM BCM CGM CIS EBCM CGM EPS EBCM OBD II port OBD II port OnStar BMS Motor Controller Charger VIM VIM SCM HVCM SCM HVCM Datalogger Instrument Cluster Screen Chapter 4 77 H EBCM f BCM OnStar VIM EPS DLC a HS GMLAN mm EBCM CGM cis HS VIM BMS TIM Datalogger ICS b UOIT CAN DLC Figure 34 Modified CAN buses a GMLAN HS b UOIT CAN bus 4 4 Control Strategy 4
114. el gauge is now a state of charge gauge on the UOIT prototype The SOC received by the VIM from the BMS is a value between 0 and 100 Having the same range as the original GMLAN signal the SOC can be dynamically emulated on the HS GMLAN without any conversion As discussed in section 4 6 5 and 4 7 1 the GFI indication lights are using the check engine light MIL and the oil pressure light for any level 2 and level faults evaluated The oil change indication light is now used as the vehicle ready light However the LED colour was changed to green to make the notification intuitive for the driver The vehicle ready indication light illuminates when the high voltage control module HVCM enables the powertrain s traction The vehicle ready signal is also hardwired to the Chapter 5 111 traction enable input on the motor controller pin 33 of Figure 39 representation of the customized faceplate of the instrument panel is shown on Figure 44 In addition to the modifications aforementioned most of the original backlight white LEDs were replaced by light blue ones and a green one was used for the ready light Figure 44 shows the custom faceplate of the instrument cluster Additional figures can be found in Appendix K voy Ree dy Figure 44 Custom Faceplate of the Instrument Cluster 5 4 6 Diagnostics and Safety 5 4 6 1 Diagnostics Several diagnostics run at different interva
115. eld Conversion from Intel Standard As shown by the previous example in order to decode a signal it is essential to know its display format and apply an appropriate conversion when reading it with a controller using a different format This example shows how a signal sent from controller using an Intel Standard display format is read by two controllers one using a Motorola Forward lsb display format and the other Motorola Backward Specifications for Intel Standard Start bit 156 11 Length 9 bits Chapter 3 52 Intel Standard Figure 25 Conversion from Intel Standard Figure 25 shows how the signal in the Intel Standard display format is represented in the other two formats In both cases since the byte order is big endian instead of little endian the signal must be rebuilt by concatenation of two signals According to the little endian byte ordering the most significant part of the signal is in Byte 2 and the least significant part in Byte 1 Therefore the concatenation is signal in Byte 2 signal in Byte 1 In the case of the Motorola Forward Isb the start bit does not need to be converted since it is using a forward message progression like the Intel Standard does However a conversion needs to be applied to the start bit for Motorola Backward because the message progression used is backward The graphical representation shown in Figure 25 is an easy method to find the new start bit in a different display format
116. ells individually monitored for voltage and temperature A modular cell frame design was evolved that incorporated liquid cooling in response to organizer s demands for active temperature control The frames were stacked into different height modules among 2 battery boxes one under the vehicle the other replacing the IC engine space underhood The designs had to be validated for crush space and puncture resistance an involved and lengthy exercise for the mechanical engineering structural team The cells were also chosen for their cyclic life safety with respect to thermal runaway acquisition price and delivery Voltage Monitoring HV Wires Figure 8 Cell Stack including cooling frames and voltage monitoring boards Chapter 2 21 4 b Figure 9 a Front ESS b Underside ESS 2 7 1 2 Battery Management System REAP Lithium batteries need to be charged and discharged with caution in order to avoid hazardous situations and ensure the battery life does not degrade During the charge or discharge processes a battery management system accomplishes the supervisory and Chapter 2 22 control tasks required to maintain the battery in optimal condition Some examples of these tasks are to e Monitor individual cell conditions max and min cell voltage max and min cell temperature and battery current e Calculate the State of Charge SOC and State of Health e Protect the cells from over discharge e Prolo
117. emain affordable and may lead to many consumer acceptability concessions However hobbyists converting cars are often ready to accept several trades offs and compromises to see their vehicles powered by a greener energy Source The Tesla s structure is designed around the battery the weight is minimized and it is practically the only long range capable EV on the market as of today Having the range of a Tesla and the look of a production cross over vehicle without being designed from scratch but rather along the lines of a factory conversion is what makes the vehicle described in this thesis so special and that it existed 1 5 years before Tesla s Model S came to market It describes the best electric vehicle the UOIT EcoCAR team could come up with not starting from a ground up design The vehicle visually and functionally looks like it came out of the factory as shown in Figure 1 and Figure 41 respectively It can be driven without any speed grade or cabin climate restrictions for more than 400 km In other words the average consumer would never be challenged for autonomous range requirements in any suburban setting Chapter 1 3 1 3 Goal of Thesis The goal of this thesis is to explain with an electrical engineering focus the design and implementation of the process employed to convert an existing HEV into an EV having over 400 km of range with similar look and functionality to a production vehicle i e a full function electric
118. ending on the ignition key position The on state voltage is connected to the motor controller ignition key input pin 31 of Figure 39 Nevertheless none of the BCM outputs could be deciphered to detect when the driver actually cranks the ignition to start the powertrain To overcome this issue an additional wire was tapped to the ignition potentiometer and connected to the vehicle integration module The VIM reads the ignition potentiometer and feeds the converted signal to the high voltage control module Figure 89 and Figure 90 in Appendix J shows the hardware dependant input block debouncing logic calibrations converted signal and debugging probes hooks 5 4 3 Acceleration and Brake Pedals The acceleration and brake pedals both have 2 potentiometers one connected to the motor controller and the other one to the vehicle integration module Originally these potentiometers were all wired directly to the engine control module for torque determination and fault management In the UOIT prototype the fault management is done in 2 different ECUs i e and VIM and the torque is only determined in the TIM The motor controller has built in fault management functions and will enter an Chapter 5 109 emergency mode when a fault is detected Under normal conditions it manages the vehicle acceleration torque demand and the regenerative braking blend The supervisory control module of the VIM utilizes these potentiometers to dynamicall
119. ents were accomplished at that competition except for the vehicle mass these projections are considered as the final performance results of the UOIT EcoCAR project Table 56 Performance projection based on EPA Results EPA Competition Battery Capacity kWh Battery Charging Efficiency AC to DC 96 UDDS Consumption Wh km Highway FTP Wh km Steady 60 mph Wh km L 100 km MPGGe Fuel Consumption GHG g km PEU Wh km 0 60 mph s 50 70 mph s Braking Distance m Weight kg EV Range 55 UDDS 4596 HW km Acceleration Further the BMS signal emulator datalogger and dashboard displays were all performing nominally No mechanical issues were uncovered and only an unrated relay in the charger circuit was discovered at technical inspection With a replacement already in hand this concern was resolved immediately CAN communication issues with the motor controller were discussed in Section 5 6 2 1 and were still present at EPA since messages were only being read from the inverter and none were sent This issue was fixed before the final competition by sending CAN commands from the VIM to the TIM enabling proper functionality of the inverter s CAN communication feature This lack of functionality restricted the motor information Chapter 6 142 the team was able to record during testing at EPA however it did not affect the results of the targeted tests such
120. eo 181 REAP BMS Screen e oe 182 REAP BMS Screen Mode 182 REAP BMS Parameters Setting 183 REAP BMS Configuration 1120 183 BMS S pervisory interface 184 NAMATIGEPACE 185 HVCM interface oot cii decedente e t dde ode ede ves 186 Display eral ic acc is suse cht esc 187 Log Task Language LTL Code Programmed in CANIog4 193 Top Level cdi i or velie d beo Edd vede deis 195 EN 196 Shit Lever Decoder Model Mee 197 Shift Lever Translator Model 198 Schematic of States service manual 199 xiv Figure 89 Ignition Key Decoder Model ate 200 Figure 90 Ignition Cranking Conditions see 200 Figure 91 Pedal Decoder Model a Acceleration b Brake 201 Figure 92 Radiator Pan Controls eee 201 Figure 93 Pedals Fault Management Model 202 Figure 94 High Voltage Controller Inputs and Outputs eene 203 Figure 95 HV Control Module Stateflow 204 Figure 96 Custom Faceplate of the Instrument Panel Cluster without the 2 display 5 6
121. er for unexpected power draw made the MES DEA 1 kW a logical choice for the vehicle s DC DC converter The specifications of this converter are listed in Table 11 and it is shown in Figure 16 Also a second unit was later purchased to independently supply the energy consumed by the battery thermal management subsystem Table 11 MES DEA DC DC Converter 1 kW Specifications 18 MES DEA Vin ac V 190 400 Vout ac V 13 3 14 4 Efficiency 96 gt 90 Tout dc max A 70 Maximum power output W 1000 Weight kg 3 6 Chapter 2 31 Figure 16 MES DEA DC DC Converter 1 kW 18 2 7 5 2 Datalogger Vector CANlog4 The competition required the teams to record specific variables from the vehicle CAN buses while competing in dynamic events thus the presence of a datalogger was mandatory As discussed latter in Section 3 7 dataloggers can be either standalone by having embedded memory or can constantly stream the information to a computer hard drive Non standalone loggers were allowed during the second year but not for the final competition Consequently the selected datalogger had to be standalone 2 devices were considered a CANcaseXL log and a CANlog4 both made by Vector Due to its two CAN ports limitation two separate CANcaseXL log would have been required to record information from the 3 high speed CAN buses i e GMLAN HS GMLAN PTE and UOIT added bus and also if necessary the single wire CAN bus
122. er of the in house control panel the vehicle was drivable Unfortunately the Chapter 6 128 auto tuning feature allowed a lot of leeway in the setup and two S10 EV motors were eventually damaged in the following months Hours were then spent trying to understand the poorly translated and utterly incomplete inverter manual by modifying each major control parameter individually Section 6 2 4 1 explains how the final calibrations were obtained during dyno testing Once the creep torque calibration variable was found among the several parameters under study a small acceleration was tuned to provide an intuitive drive feel before pressing the gas pedal while engaging the drive or reverse gear This simulates a torque present on automatic transmission internal combustion engine vehicles making the drivability of this electric vehicle a little more intuitive Most of the dynamic testing was performed in the shop s parking lot such as short stints of mild driving braking and even hard acceleration as the team s confidence in the hardware increased Thereafter the vehicle was driven on public roads near the garage for longer periods to test the vehicle at higher power demand However 0 60 mph accelerations could not be evaluated accurately in the limited sized parking lot or safely on public roads and were left for the end of the third year during performance testing at the EPA lab and the final competition 6 2 3 Performance Testing at Y
123. erative braking is available through the traction inverter module TIM A typical vehicle powers its ancillaries with a 12V lead acid battery charged by an alternator coupled to the engine A DC to DC converter DC DC reducing the high voltage from the traction battery to 12V is utilized to charge a 12V battery and power the accessories As a result the auxiliary power module APM integral to the TPIM block was also removed A second DC DC converter is added to satisfy the power requirements of the battery thermal management system which is only enabled at temperature extremes On the baseline vehicle the battery pack has a smaller liquid thermal management system consuming less energy thus avoiding the need for extra power on the 12V system A datalogger is added to record data from the HS and UOIT buses Even though the dashboard cluster screen is replaced by a custom one managed by a proprietary controller the instrument panel cluster IPC is still needed to manage other functionalities of the dashboard Although it was not necessary for the conversion the baseline radio was replaced by a combination of navigation and rear seat entertainment system Chapter 4 75 the removed ECUs i e FSCM TPIM HCP were on the HS and GMLAN SW GMLAN remained unmodified Table 35 lists the removed and remaining stock ECUs Table 35 Removed and Remaining ECUs
124. erence between Data Frame and Remote Frame 42 3 2 5 CAN Standards for Vehicle 0221 140 42 3 3 CAN Transceiver and Protocol Controller eene 43 45 3 CAN Message Defintiodnh 46 3 501 Message and Signal ete teas etes 46 3 5 2 Mapping and Positioning of Signals eese 46 3 5 2 1 Byte Order CBIHOIaHUESS ou er P 47 9 5 2 2 Bit coa edo 48 35 298 Message RUN ERR NEU SUE 48 3 5 2 4 Start Bit of Signals eee ree en ova PER eee RR SIN ea SENSE ER UG 50 3 9 2 5 Display 50 3 2 2 6 UE EROS RU RR UTOR RR ER 51 3 5 3 Message Signal Definitions 54 3 9 9 b Message De BOTE oae e mtt aeu us 54 3 5 3 2 Signal i no eee reo rid e Ed Mori 57 58 3 7 Commercial Devices iiss oda ema nien ormai ote a dad d eds 61 viii 3 8 Applications for Vehicles 52 23 62 Sub D Type S OP AppDLICRUO TIS uoa es S os RR cae du qat 63 3 8 1 1 Low Speed 5 63 3 8 1 2 High Speed Applications 64 3 8 1 3 Diagnostic Interface and ECU 65 3 8 1 4 Gateways between CAN buses
125. ergy vs voltage curve 6 3 2 5 Summary The dynamometer results were encouraging and in line with the prediction made from PSAT simulations for energy consumption and range by the team in the first year of the competition Although the temporary Li ion pack was not fully balanced before testing at EPA it managed to deliver slightly more energy than the nominal 74 6 kWh expected 375 km over 230 mi were covered through the various drive cycles thus easily demonstrating competition requirements for autonomous range were being met With the data subsequently gathered from the two UDDS cycles run with REGEN active projections were made for energy consumption in final tune This computes to approximately 178 Wh km 286 Wh mi for the UDDS cycle and 167 kW km 268 Wh mi on the highway cycle If the entire test sequence were re run with REGEN active a 400 km 250 mi range should have resulted Not to mention that these results were based on a temporary battery pack An additional 10 cells were installed in the final Chapter 6 141 battery modules to raise the voltage but also provide for approximately 8 9 kWh extra energy capacity thus 83 5 kWh total Using a 55 UDDS and 45 HW cycle mix and regenerative braking the range with the temporary battery was 442 km whereas the estimation for the final one is 482 km Table 56 reflects the performance data collected at EPA and projections thereupon for the year 3 competition Since no dynamic ev
126. expressions are often referred to as Intel and Motorola respectively since Intel processors use the little endian method and Motorola processors use the big endian one Table 22 explains this nomenclature Chapter 3 47 Table 22 Byte ordering Referred Byte order Endianness to as Byte 0 Byte 1 2 Byte3 4 Byte 5 Byte 6 Byte 7 Little endian Intel LSB MSB Big endian Motorola MSB LSB In summary the expression little endian and big endian define the byte ordering of data LSB to MSB and MSB to LSB respectively In other words they inform on the ordering of significance in the data field of a CAN message but give no information on the numbering associated to each bit stuffed within a byte of the data field 3 5 2 2 Bit Numbering The bit numbering indexing of bits within a byte can be defined by the following 2 methods 47 49 Decreasing from left to right Sawtooth e Decreasing from right to left Sequential or Monotone Table 23 Bit numbering Byte Bit Numbering Referred toas msb 5s n 5 5 6 Bit Bit Bit Bit Bit Bit Bit Bit g e t left to right Decreasing from Sequential 0 1 2 3 4 5 6 7 right to left Monotone The bit numbering also called bit counting is independent of the bit order which is referring to the msb to Isb bit progre
127. f the dynamically emulated CAN signals can be found in Table 48 Chapter 5 116 Table 48 Dynamically Emulated CAN Signals ECUs BPCM HV battery voltage State of charge Temperature Auxiliary voltage limits 13 6V battery current Engine run active Vehicle speed Accelerator actual position Brake pedal status Transmission shift lever position Engine oil change indication on Ready light Check engine indication on MIL GFI Level 1 light Engine oil pressure low indication on GFI Level 2 light Fuel level percentage HV battery voltage 12V battery gauge battery current Cruise Control off Air conditioning compressor command off Air conditioning refrigerant high side fluid pressure off HCP Transmission shift lever position EE TCM Transmission shift lever position The dynamic emulation of the cruise control and air conditioning was intended but was never implemented At present static emulation keeps these 2 functionalities disabled 5 6 2 UOIT Added ECUs As shown by Figure 34 Chapter 4 the UOIT bus physically interconnects battery management system BMS on board battery module traction inverter module TIM datalogger and instrument cluster screen ICS to the vehicle integration module VIM Each of these controllers communicates directly only with the VIM Even though they are on the same network they d
128. fied CAN bus whether it is in charging or normal mode Chapter 4 88 4 5 2 On Board Charger Module The UOIT Full Function Electric Vehicle can be equipped with up to 5 BRUSA 0513 onboard chargers connected in parallel as shown in Figure 38 Their inputs and outputs are detailed in Table 41 HV AC HV DC Charger 1 AC line 1 Live Charger 2 M LA HV Breaker Charger 3 Distribution Box AC line 2 Neutral Box Charger 4 17 Charger 5 T a High Voltage Parallel Configuration RS232 5x 3 cond UOIT CAN bus Charger 1 AC line 1 Live Charger 2 AC line 2 Neutral Breaker Charger 3 Box Charger 4 Charger 5 AC on Enable MotoTron b Figure 38 Chargers a High Voltage Parallel Configuration b I Os Chapter 4 89 Table 41 OBCM I Os Description 13 RS232 Configuration of each charger s software CAN Rx Receive command from the BMS to drive the charge cycle Tx Broadcast data to the MotoTron Voltage sensors Measure the battery voltage Current sensors Measure and limit the battery charging current Temperature Internal and external temperatures Sensors PWM safety Input controlled by the BM
129. flect the needs of an electric vehicle The emulation of CAN signals was used to take control of the instrument cluster gauges and indicator lights Features controlled by the VIM are e Speedometer e Current gauge Chapter 5 110 e State of charge gauge e and vehicle ready lights e Custom screen The remaining stock ECUs on board the vehicle expect a speed signal which originally emanated from the ECM and used wheel speed sensor values broadcast by the electronic brake control module EBCM The VIM has the 2 following options to determine the vehicle speed process the speed sensor values or use the speed value sent by the motor controller Both options were explored however experiencing conversion inaccuracies in the speed sensor value processing lead to using the velocity output provided by the motor controller in km h In this case the VIM just acts as a gateway between the UOIT bus and the HS GMLAN since no scaling or offset is required The 12V battery gauge was modified as a current gauge The VIM receives the battery current from the BMS and TIM However the value from the BMS was judged more representative This value is converted in the VIM to use the full range of the 12V battery gauge and allow negative values for regenerative braking The conversion is necessary as the instrument cluster expects a CAN signal having a scaled value of 0 to 30V while the VIM wants to send a current ranging from 200 to 200 A The fu
130. g the last charging phase without reducing the charging time The control logic in the chargers is static and programmed via RS232 communication and using the supplier s software The VIM only enables and disables the chargers along with controlling the contactors by monitoring the CAN messages sent by the charger configured to turn off last and on first Chapter 4 91 4 5 3 Motor controller The available I Os on the TIM are shown in Figure 39 For clarity the high voltage wires and sensors are omitted Nevertheless among the available controls I Os several were not required by the control strategy used for the powertrain thus their pin numbers were dashed An exhaustive description of the TIM s I Os is found in Table 43 GEAR CAN BUS HIGH PIN 39 PIN 7 DRIVE CAN BUS LOW PIN 11 PIN34 INS ECONOMY _ PIN 8 ING REVERSE Serial RXD_2 24 SESE PIN20 __IN7___ PARKING _ Serial TXD_3 PIN 38 Serial GND 5 PIN 23 PINI6 VBATT_U 12 AA TRACTION READY ES PIN 3 PIN 18 IN PWM LIM POWER SPEED BELOW P50 UT3 PIN 17 HAND BRAKE CONTACT v2 PIN 9 INS gt LIGHT REVERSE OUTS 25 NOT USED PIN 5 IN12 POWER STEERING RENAULT OUT 30 x1 NOT USED 35 IN9 CONNECTOR gt IU L TACHOMETER SIGNAL OUT acho _ PIN 4 FIRMWARE UPDATE ENABLE PIN37 INIG PARKING ENABLE OUT4 PIN I AC NET CONNECTED PIN21 INII a PIN 29 ALARM RESET PIN 22 INIS CO
131. ge systems Extra gauges and displays are normally added to visually output information specific to EVs It is also common to find a datalogger in EV prototypes for research purposes Chapter 2 17 the controllers mentioned previously interact with the car through sensors input and actuators output Examples of sensors would be potentiometers for the accelerator the brake pedal and the ignition key voltage and current sensors in every high voltage device speed sensors on the wheels the shifter lever or the interlock between high voltage components On the other hand actuators would be contactors on the high voltage bus firing signals for MOSFETs and IGBTs in power electronic devices or a signal to control the small DC motor powering the speedometer needle in the instrument cluster This subsection described the following additional key components added to the production VUE to implement the electric architecture battery cells battery management system BMS charger motor motor controller vehicle integration module VIM along with the secondary devices such as a DC DC datalogger ground fault interruptor GFI and dashboard graphical user interface These components are positioned in the UOIT EcoCAR EV prototype as shown in Figure 6 HV distribution box 2 A deis Integration Module I 4 5 42V distribution box 22 lt W 4 Aw aig W 4 TAA gt Ey 74 E BRU
132. gn proven to be final For example a lot of pressure was put on the author to cut the VIM wiring harness by 2 meters but the team was relieved to have the extra length when the cables needed to be rerouted to accommodate the front battery pack Test the equipment regardless of whether it is new or used For example the brand new CANlog4 a 2000 device had 2 defective CAN transceivers and a few of the BMS boards reused from another project had blown surface mount fuses Archive a new version of the software each time a small improvement is made Make release notes for each software version The use of an online software release management tool might be useful when working within a group of programmers Keep the same file names when creating new releases to avoid bugs only modify the folder name Avoid spaces and special characters in file and folder names along with long names 32 bits Keep a compiled version of the working software release to avoid having to rebuild it again i e the dll and srz files for MotoTron controllers Compiling and building a program can be time consuming Ideally always flash 2 controllers to have a spare in case the flashing of one fails and leaves the other controller in a useless state The high voltage lead and the controls lead should be 2 different electrical engineers to spread the work load Chapter 7 153 Once embedded system software is proven reliable it will stay reliable wher
133. h 2 termination resistors of 120 ohms each Figure 20 shows a typical CAN bus topology when two differential wires are used 25 Description CAN High wire CANH CAN Low wire CANL Voltage between CANH and CANL Termination resistors 1200 Node i e ECU 1 Node i e 2 Node i e 3 DA NN BY Go Figure 20 Topology Depending on the configuration of the network and the environment the transmission distance can reach up to 1 km Table 14 summarizes 3 common specifications of the protocol Chapter 3 39 Table 14 Differences between Low Speed High Speed Parameters ISO 11898 2 ISO 11898 3 SAE J2411 fault tolerant Name high speed single wire 33 3kb s Baud rate up to 1Mb s up to 125kb s up to 83 3kb s Number of 2 2 1 wires Dominant AN H iV 3 6V 5 bit level 3 2 33 Recessive 1 75 bit level 1 5 1V OV Each node needs to Termination 2 resistors of 1200 terminate both CAN 9 09kQ load resistor lines individually Limited by busload and data rate Length 40m Mb s 100m 250 kb s Ikm 50kb s Limited by busload and data rate Number of Limited by busload nodes and data rate up to 32 up to 32 2000 msg s 250 kb s Maximum 4000 msg s bandwidth 500 kb s 8000 msg s Mb s As further explained in Appendix E
134. h voltage powertrain is ready to power the motor i e when the ESS and ESS contactors close after the key was cranked and the high voltage precharge circuit is complete 5 6 CAN Communication 5 6 1 Emulated CAN Messages Among the 8 main ECUs removed from the vehicle 5 of these were still expected to be sending CAN messages either on GMLAN HS or PTE by the remaining controllers A summary of the emulated ECUs is provided in Table 47 ECUs expecting inputs were determined by looking in the CAN database provided by General Motors checking if the remaining controllers were normally receiving messages from the removed ones Table 47 Emulated ECUs GMLAN HS GMLAN PTE X X X X X X X Chapter 5 115 Expected CAN messages Removed ECUs APM In an attempt to avoid as many issues as possible related to removing ECUs all messages typically transmitted by the remaining ECUs were broadcast unless the original ECU expecting the message was removed from the vehicle Most emulated signals are sent statically meaning that a fixed value for each signal was chosen by scrutinizing the original dynamic vehicle CAN bus traffic and is now constantly being transmitted Common examples of these signals are validity and fault active bits being emulated to 0 letting other ECUs believe everything is working properly However some CAN signals need to be updated frequently and are referred to as dynamic ones The list o
135. he encoding safety a mechanical latch in the shift lever box prevents the PARK position to be engaged while the vehicle is in motion to protect the gearbox This feature is reused form the stock vehicle and is complementary to pressing the brake pedal to engage the PARK position Once the PRND signals are decoded 4 TTL transistor transistor logic outputs of the VIM are used to update the motor controller Before being sent as CAN signals the decoded shift lever signals needed to go through a second signal processing phase After Chapter 5 scrutinizing the GMLAN database and considering the removed ECUs 4 CAN messages were found to require the PRND information and all 4 were using a different encryption Appendix J Figure 87 illustrates the shift lever translator model making the appropriate conversion for each CAN message As the vehicle controls complexity increases additional CAN messages must be transmitted between controllers leading to extra CAN buses Sending multiple redundant CAN signals becomes a waste of bandwidth and vehicle manufacturers ought to avoid this type of inefficient situation but it probably results because new hardware is developed piecemeal and is deployed over existing configurations 5 4 2 2 Ignition The ignition key has 4 positions off accessory ACC on and cranking Each of these positions are decoded by the body control module BCM which activates different 12V power modes and CAN signals dep
136. he vehicle but only used as an emulator and a diagnostic tool via MotoTune There was no main high voltage controller or distribution box The car ignition was still connected to the stock ECUS but not to the VIM or the other added powertrain controllers The in house control panel was used to drive the motor controller inputs such as the acceleration and brake pedals ignition traction enable and regenerative braking enable With the vehicle off the ground this is when the team saw the front wheels of their electric vehicle prototype spin for the very first time A few minutes after the acceleration and brake pedals were routed to the terminal blocks while the other signals were still originating from the control panel Figure 47 Set up when the wheels spun for the first time Chapter 6 127 As mentioned previously most of the components added to the vehicle functioned properly virtually the first time when migrating from the test bench to the vehicle This can be attributed to an efficient test bench connected to the VIM and was realized during previous months by following a meticulous debugging process Debugging on a complete test bench is much more efficient and easier to follow because one can verify each module individually and analyze its reaction to known stimuli as opposed to trying to debug it in a complex network of controllers The methodology employed has helped improve the control system and ensured proper functioning and
137. hts gt Speed SOC and current gauges e Design and implementation of an in house instrument cluster screen with CAN capabilities Chapter 7 150 Description of an unambiguous terminology to explicitly define CAN messages and map their signals Implementation of a strategy to control five BRUSA chargers connected in parallel e Implementation of 2 GFI fault levels Explanation of the removal and emulation of ECUs Explanation of the electrical aspects of a vehicle conversion to electric 7 4 Recommendation 7 4 1 Future Work Even though the UOIT EcoCAR team spent countless hours developing their EV prototype mainly due to competition deadlines some engineering issues were left unresolved temporarily fixed or ignored As described in chapter 6 after dynamic testing of their prototype at EPA the team undertook major modifications to the high voltage system by upgrading the Li ion battery high voltage distribution box and control module along with the migration to a more powerful controller for the VIM Unfortunately the totality of these changes could not be accomplished in time for the final competition and the vehicle was left in a non running state Future research and development should focus on completing this migration phase to bring the vehicle back to a drivable state Below is a list of main tasks to accomplish in an attempt to reach that objective and also improve the vehicle e Precharge circuit analysis b
138. icant bit lsb at the rightmost position The byte numbering defined as the number assigned to each byte of the data field 1s also fixed but is increasing from left to right 45 46 47 48 The first byte sent is defined as byte O and the last byte is defined as byte 7 These concepts are illustrated in Table 21 Table 21 Byte numbering and bit order Data Field Message Byte 0 Byte 1 Byte 2 Byte 3 Byte 4 Byte 5 Byte 6 Byte 7 msb 186 msb lsb 186 msb lsb msb lsb msb lsb msb lsb 15 Knowing that the bit order of significance and byte numbering are fixed to map a CAN message and to place its signals in the data field the following 4 pieces of information are required Byte order of significance endianness Bit numbering Message progression and overall bit numbering Start bit of signals It should be noted that the length of a signal is needed for its definition but not to position it within the message 3 5 2 1 Byte Order Endianness When discussing byte order 2 methods of sending the bytes of a message exist little endian and big endian 47 Little endian means sending the least significant byte LSB first and the following bytes in increasing order of significance Intel Big endian means sending the most significant byte first MSB first and the following bytes in decreasing order of significance Motorola These
139. iers and numerous automotive applications are covered In other words it was shown how the CAN communication protocol allows ECUS to interact with one another The next chapter probes deeper into these interactions by explaining the control strategy managing the powertrain ECUs selected by the UOIT EcoCAR team now implemented in the EV prototype Chapter 3 66 Chapter 4 Electrical and Control System Integration Strategy 4 1 Introduction The in vehicle communication technology linking on board ECUS through a robust network was covered in the previous chapter The control strategy for integrating these additional components to the baseline production vehicle is detailed in the current chapter Removal and addition of ECUS is discussed first along with their layout on the CAN communication network An overview of the general control strategy is then presented followed by the control details of the main powertrain modules Additionally the safety control strategy is explained followed by a description of the on board diagnostics system 4 2 Conversion Methodology The EcoCAR challenge suggested competing teams follow the GM validation process diagram shown in Figure 31 8 Define Validate Control Control System System equirement In Vehicle Systems System Testing Engineering Calibrate Vehicle rchitectur amp M E Rapid Control Prototyping Validation Algorithm Requirements Develop
140. ifier hex Baud Rate Transmit Messages Identifier hex Necs joen Set Defaults INLGS ACTII ijosi _ NLGS ERR 10614 Control Pilot Figure 74 CAN Configuration Window Appendix I 181 2 REAP RS232 Interface REAP HyperT Ele Edit View Call Transfer Help Da 1935 009 M MJ AJ AJ AJ AJ RJ I RC 3 1 1 14 0 12 35 5 4 4 4 4 4 4 4 4 4 4 4 4 4 4 150 Figure 75 REAP Screen Mode Current Measurements Cell Cell a Shunt1 Reading A Number Voltage LTEM gt Hall Effect Sensor Reading Sum of both inputs Discharging current are negative Charging Currents are positive Capacity Measurements d Remaining Capacity Ah Total Capacity Ah Total Cycled Capacity of the battery This value is accumlated over the entire life of the battery 3 3 3 3 3 3 3 3 3 3 3 Status and Errors BMS status see Table titled BMS Status Error Status see Table titled Error Messages Last temperature value is average value over previous weeks Figure 76 REAP BMS Screen Mode description Important values V I SOC and errors warnings are displayed Appendix I 182 REAP HyperTerminal File Edit View Call Transfer Help 0265 28 PID Voltage PID Reset Empty
141. instead Furthermore operational commands can be received by the master but are not used since the BMS module s basic settings were configured manually via RS232 communication The summary message periodically broadcast by the master gives an overview of the battery by providing the e State of charge e State of health Maximum temperature e Maximum cell voltage Minimum cell voltage In addition the delta cell voltage between the maximum cell and the minimum cell was computed by the VIM It is a useful indicator to determine how well balanced the battery pack is After a request by the competition organizers this value was also recorded with the datalogger Even though the overall current going in and out of the battery is not present in this summary it is sent by the master on a dedicated CAN message Unfortunately the total battery voltage is simply not provided by this BMS Being an important metric to monitor when using a battery pack the VIM computes the battery voltage by adding the voltage of every single lithium cell To accomplish this computation the VIM needs to read 28 additional messages from the UOIT CAN bus Chapter 5 28 different messages would typically require the same quantity of CAN IDs However the BMS manufacturer decided to use the first byte of the 8 data bytes available in CAN messages as a sub ID to minimize the use of IDs In other words the first byte of the payload is used to identify the content
142. is defined in Table 66 Table 66 Post Office Analogy Post Office CAN Letters CAN messages Postman Hardware Software dispatcher Address ID Filtering Mail box of the sender ID mask Slot Desired N Payload filtering Desired CAN letter n Payload msg of the sender mask Doorbell Interrupt or a function trigger Imagine a world where letters are CAN messages and the postman is the hardware and software dispatcher In this world the mail system works according to the sender s address and the name instead of the destination information Also a house owns a mail box for each type of sender in other words for each type of desired letter Several people can send letters from the same address and a mail box is only limited to one sender s address but may contain letters from numerous people from that address A letter is Appendix G 174 desired when its sender s address and name match the ones the mail box In this analogy a desired letter represents a filtered CAN message i e a CAN message having an identifier and a data field fulfilling the criteria of the identifier and payload filters For priority mail the postman notices the mail box but delivers the urgent letter directly to the front door and rings the doorbell This is the equivalent to setting a desired CAN message to an interrupt or a trigger function Appendix 175 Appendix H Comparing LIN and CAN Table 67 com
143. l as the transition to the vehicle Safety features were successfully implemented in the VIM such as the acceleration and brake pedal fault management cell protection HV interlock system inertia switch and voltage current temperature management The instrument panel cluster was customized to display the vehicle powertrain status i e ready and ground fault interrupt lights relevant information about the FFEV on an in house screen and stock gauges were reprogrammed to suit the needs of a FFEV Also the cluster background panel was modified accordingly by another team member In addition to the stock OBDII CAN bus access under the dashboard and on the left of the steering wheel RS232 and CAN access points to every powertrain EV controller were made available through the glove compartment Troubleshooting interfaces were created to diagnose the vehicle using variables in the VIM The VIM inputs and outputs were all tested first to insure proper functioning of the controller leading to a successful test bench integration phase In the 3 year competition the VIM model based program was translated to migrate to a more powerful MotoTron controller i e from the ECM Chapter 6 145 0555 to the 5554 voltage translator was also added with success to link the and motor controller controls signals The 3 main tasks of the VIM i e emulation decoding sensors and controlling actuators along with managing the HVCM functio
144. l limit on a bus having too many nodes Sleep mode and wake up are available options for each node to reduce power consumption 41 Since the reliability of a communication system is paramount each node accomplishes powerful safety tasks such as error detection signalling and self checking Each node is able to identify the difference between temporary errors and permanent failures of a node In the event of a permanent failure the other nodes on the network automatically command the faulty node to switch off 23 3 2 2 2 Properties In summary the priorities of the CAN communication protocol are 22 e Configuration flexibility e Prioritization of messages bitwise arbitration using the identifier e Simultaneous reception by multiple nodes with time synchronization Multi master system e Guarantee of latency time e Error detection and signalling by each node e Automatic retransmission of corrupted messages once the bus is idle again e Error distinction between temporary errors and a permanent failure of a node e Fault confinement automatic switch off of defective nodes Chapter 3 38 3 2 3 Physical Characteristics The CAN specification does not define how the single channel which carries bits is implemented A common way to implement a CAN bus is by using a single wire and a ground However the most typical implementation and the main focus of this work is using 2 twisted differential wires CAN high and CAN low wit
145. le and Chapter 6 125 thereafter verifying if the main car modules would still operate properly after the systematic removal of ECUs It was difficult to know exactly what to expect since the emulation program of the VIM does not respond exactly like the car modules in every situation As mentioned previously some CAN messages are emulated dynamically whereas others are only static There are so many possible scenarios that it is very difficult to determine with certainty whether a CAN signal needs to be static or dynamic However the numerous hours studying the GMLAN dictionary and logs of the unmodified CAN buses paid off since the car turned on without any indication of warnings on the instrument cluster the first time Also the presence of diagnostic trouble codes DTC was checked by the team s GM mentor using a NeoVI and the GMLAN database Again no unexpected trouble codes were found The shift lever acceleration and brake pedals were rerouted to the VIM and dynamically emulated with success Proper emulation of the brake pedal and shift lever was shown by functioning brake and reverse lights along with the dynamic update of the PRND position on the instrument cluster display Power steering was enabled by the emulation of the shift lever signal and a few other specific signals The power steering also worked immediately Unfortunately a modification made under competition pressure in the VIM programming during the second yea
146. le 57 UOIT Team VTS 14s Requirement Accel 0 60 mph 16 lt 48 14s O lt 9s 8 3 L 100 km 7 4 L 100 km 2 45 L 100km Towing gt 680 kg 3 5 gt 680 kg 3 5 680 kg 1500165 20 min 72 km h 45 mph 20 min 72 km h 45 mph Height 457 mm 18 226 Cargo Capacity 0 83 m Depth 686 mm 27 i d Connon Width 762 mm 30 Passenger 7s 5 Braking 38 m 43 m 123 60 0 mph 140 ft 51 8 m 170 ft 44 m 144 ft 1758 kg 3875 Ib lt 2268 kg 5000 Ib 2112 kg Ground gt 198 mm 7 8 in 7178 mm 7 in 178 mm 7 in Range gt 580 km gt 320 km 482 971 360 mi 200 mi 300 mi 6 3 4 Consumer Acceptability The UOIT prototype was designed to improve consumer acceptability by maintaining the stock features of the vehicle and adding new ones The vehicle can seat 5 occupants The cargo space was maintained and an in floor trunk compartment was even created and carpeted to make use of the free space left by the removal of the stock NiMH battery Having no engine and an electric powertrain this vehicle is quieter than the original one The audible noises now come from the electronic brake control module pump when activated and the rolling sound of the tires against the road A DVD player with a screen integrated to each front seat head rest was installed for rear seat passenger entertainment For infotainment the do
147. lectric motor Therefore the ICE can charge the battery through the generator as in series hybrids and also mechanically power the drivetrain as in parallel hybrids This concept is displayed in Figure 5 GASOLINE ENGINE GENERATOR POWER 5 j CC 2 4 ELECTRIC BATTERY ELECTRONICS MOTOR Figure 5 Power Split Configuration 6 Chapter 2 10 2 3 Fuel Selection In vehicles the onboard energy is usually referred to as fuel This advanced vehicle engineering competition allowed participating teams the following choice of fuels E10 a reformulated gasoline with 10 ethanol E85 a reformulated gasoline with 85 ethanol BD20 a reformulated diesel with 20 biodiesel H2 hydrogen required for fuel cell to operate Electricity electricity in a rechargeable battery The onboard energy consumed by a vehicle is not the only energy having an environmental impact This energy is referred to as pump to wheel PTW whereas the energy required by the generation or transformation process to render it usable by vehicles is called well to pump WTP When analyzing the overall energy and greenhouse gas GHG production of vehicles it is important to use the well to wheel WTW analysis which includes both the well to pump and pump to wheel energy For example electric cars have no pump to wheel GHG emissions but are still accountable to those released by the generating plant producing the electricity
148. lling the additional devices This is accomplished by controlling their inputs and supervising their outputs These could be hardwired as well as transmitted information A detailed list of the MotoTron I Os and their functionality can be found in Appendix J while a list of the relevant ports follows e Analog inputs e Discrete inputs e Discrete outputs Chapter 5 102 e PWM outputs Low side outputs current sinking drivers CAN communication ports MotoTron is a controller series made by Woodward MotoTron Control Solutions thus a few controllers with different features could have been chosen as the vehicle integration module As a reference MotoHawk is a the software development tool which uses the Matlab Simulink MotoHawk library for model based design and the MotoTune program for flashing and calibrating MotoTron controllers The software versions used to program the VIM are in Appendix J Table 45 compares the MotoTron controller used for the second year of the competition to the one chosen in the final design Table 45 Comparison between ECM 0555 and ECM 5554 0555 5554 Freescale Freescale Processor MPC555 MPC5554 40 MHz 80 MHz Flash memory 448 kB 2 MB RAM memory 26 kB 64 kB of CAN ports 2 3 Analog inputs 19 33 TTL level outputs 8 8 4 contactors 4 contactors 16 others 19 others Cost with harness 1000 1300 Even though the E
149. ls in a vehicle often at start up They verify the proper functioning of most sensors actuators and controllers of the vehicle When a diagnostic fails a diagnostic trouble code DTC is generated on the CAN bus and logged These DTCs are convenient to quickly know the status of a vehicle when serviced Requiring a considerable programming time creating custom diagnostics to determine the status of the electric powertrain was not in the scope of this vehicle Chapter 5 112 conversion However obsolete CAN messages reserved for DTCs could have been reused to populate on the OBDII port the status of the BMS TIM OBCM and VIM along with the added sensors and actuators Each controller sends its status to the VIM able to emulate corresponding DTCs For example the motor controller sends its status to the VIM The VIM acting as a translator and a gateway could emulate the motor controller status by reusing existing DTCs reserved for the engine For the added sensors and actuators not monitored or diagnosed by any of the added ECUs complete diagnostics would have had to be programmed in the VIM For the purpose of this prototype diagnostics and DTC emulation is not managed and integrated by the VIM The status of each added controller is available for troubleshooting using CAN on the MotoTune interface of the VIM and RS232 communication on controller specific software directly wired to each of them For convenience the connectors re
150. main switch can be found in the wiring harness used to power and access the MotoTron inputs and outputs The main switch is a button for the user to manually turn on and off the controller whereas the main power relay is controlled by the MotoTron and needs to be initialized in the top level subsystem This initialization block is marked by item 3 in Appendix J Figure 84 The non virtual subsystem named Foreground contains the whole program of the VIM A virtual subsystem is only a subsystem grouping blocks for visual convenience whereas a non virtual subsystem in Simulink is a subsystem triggered by a function call These function calls can be triggered by actions or at periodic time intervals The latter can also be termed function call subsystem or triggered subsystem For MotoHawk projects each non virtual subsystem located in the top level must be triggered by a MotoHawk entry point trigger block to allow the compiler to generate compatible c code Multiple non virtual subsystems can run at different rates and interact with one another However it was elected to design the VIM s program in a single subsystem Chapter 5 104 triggered every 5 ms The MotoTron trigger block and the triggered subsystem represented by item 4 in Appendix J Figure 84 The 4 items mentioned are required for any MotoTron controller regardless of its purpose Using CAN communication and fault management additional definition blocks related to
151. me types 22 Data Frame standard or extended Remote Frame standard or extended Error Frame generated by the protocol manager Overload Frame generated by the protocol manager The data frame can be divided in 7 subsections 22 Start of Frame Arbitration Field Control Field Data Field CRC Field ACK Field End of Frame Chapter 3 41 3 2 4 1 Difference between Data Frame and Remote Frame The remote frame is a data frame having a payload equal to 0 it carries no data It is used to request data from specific nodes Another difference between these 2 frames is the value of the remote transmission request RTR bit in the arbitration field dominant for data frame and recessive for remote frames Figure 21 and Figure 22 detail the general structure of data frames and remote frames translated from 25 Additional info on the different frame types and formats can be found in Appendix B on CAN theory Start of Arbitration Control Data field CRC Ack End of Frame field field pea field field Frame Mbit 12bits or 32bits bits 198655 16bits bits 7bits Figure 21 Data frame general structure 25 Start of Arbitration Control CRC Ack End of Frame field field field field Frame ibit 12bits or 32bits 6bits 16bits Qbits 7615 Figure 22 Remote frame general structure 25 3 2 5 CAN Standards for Vehicle Applications Table 16 groups the 3 different
152. med useful Its gear ratio and power rating need to match the application A few choices were available with caveats listed below e GM e drive GMT101X denied use 110 kW EcoStar Siemens drive heavier but only rated at 67 kW Delco System 110 s SIO EV rated 85 kW 350 MES DEA 30 kW continuous peak unknown no gearbox e Powerphase 125 no matching gearbox unobtainium price In the end the S10 EV overdriven to 400 A as per the specifications was chosen for its compactness lowest weight and package size plus availability The team selected the MES DEA TIM600 which is designed to run asynchronous induction machines can be fully calibrated and auto tune according to the motor s dynamic characteristics It accepts torque commands directly from the acceleration and brake pedals along with digital inputs for gear selection and ignition position Regenerative and hydraulic braking can also be easily blended with a brake pedal potentiometer Chapter 2 27 Table 8 Delphi S10 EV Delco System 110 400 A 240 Vms 14 Type 3 phase induction motor Power 102 kW 137 hp 7000 rpm Torque 150 Nm 110 ft Ibs gt Transmission Gear reducer and differential Table 9 MES DEA 600 15 Service voltage 12V Vin de 80 450 Vac Tout ac 266 Arms Tout ac max 400 Arms fout 0 500 Hz Maximum power 160 kW 400 A 400 Vac inp
153. ment Throughout the years of design a lot of concerns were raised by the competition judges about fault mitigation of the acceleration pedal The accelerator has 2 potentiometers of opposite travel logic to determine its position One is connected to and managed by the motor controller s throttle input and the second one is linked directly to the supervisory controller which has fault management capabilities The TIM has a basic built in fault detection and management short to ground or power supply and arbitrates between control signal inputs if there is any conflicts on the throttle or brake favoring the brake Furthermore it sends the pedal status on the UOIT CAN bus The VIM is also programmed to detect and mitigate faults to ground or power on the acceleration pedal potentiometer directly connected to it The supervisory controller prevents unintended acceleration by detecting faults determined by the redundancy between the pedal status sent from the TIM and the second potentiometer locally monitored Therefore the acceleration pedal is read by 2 devices TIM and VIM having fault mitigation safety features able to open the main contactors to prevent unintended acceleration 4 6 2 Cell Protection Another major safety concern on electric vehicles is the cell voltage and temperature of the lithium battery cells Here again redundancy is used The BMS monitors for over charge over discharge and temperature conditions I
154. ment Validate Controller System Controller Testing HIL Unit Algorithm amp Software Behavioral Design Test Sub System Testing SIL Unit Testing Figure 31 Validation Process Diagram 8 Chapter 4 67 Even though the general approach of the UOIT team can be explained using this validation model the conversion methodology is also accurately represented by Figure 32 Study CAN bus Bench Testing of EV Hardware In Veh Integration of VIM ECU Removal and Emulation On Bench Integration of EV Hardware In Veh Integration of EV Hardware In Veh Routing of pedals Ign etc Static Testing Troubleshooting Dynamic Testing Troubleshooting Figure 32 Conversion Methodology During the conversion process 2 phases took place in parallel i e one on the original vehicle and one on the test bench and were followed by the vehicle integration phase In the vehicle phase the CAN buses were studied message by message signal by signal and even bit by bit to learn the communication pattern between the original ECUs Then the author started communicating with the car using the VIM and finally removed and emulated the ECUs Chapter 4 68 In the test bench phase the author tested and calibrated all the powertrain components such as the battery battery management system motor motor controller chargers and vehicle integration module first individually then integrated When both phases were
155. n be found in Appendix G Figure 79 to Figure 81 the BMS supervisory VIM and HVCM interfaces respectively Chapter 4 99 Troubleshooting of the 3 other devices was possible using RS232 serial communication The BMS can be programmed using a generic serial terminal however the UOIT team had more success employing the program HyperTerminal From the serial port 3 modes can be accessed monitoring of the cell voltages and general errors parameter settings and configuration mode Screen captures can be found in Appendix G Figure 75 to Figure 78 The serial communication has the disadvantage of being limited to only 1 BMS unit at a time Therefore each module needs to be accessed individually to be monitored calibrated and configured Most of the variables are monitored through MotoTune and the calibration and configuration are infrequent Additionally BMS modules need to be configured individually but the calibrations in the master module are automatically shared with the slaves Similar to the BMS each BRUSA unit has to be programmed individually using RS232 communication and the specialized program Charger Star NLGS It allows programming the 3 charging phases and sets multiple thresholds such as the nominal battery voltage charging and mains current time limit and maximum voltage output to name only a few Figure 69 to Figure 71 found in Appendix G represent the 3 charging modes programmed into the last unit to switch off
156. n properly A functional power steering the customized instrument cluster features and the absence of DTCs are examples of the successful emulation of CAN messages The VIM makes decisions based on the acceleration and brake pedals ignition potentiometer and shift lever position values along with actuating the radiator fans The subsystem of the VIM responsible of controlling the HV module actuates as intended the main contactors precharge and discharge relays charger contactor and the traction enable signal sent to the motor controller To accelerate the integration process and troubleshooting of the vehicle terminal blocks were used instead of automotive connectors to route the wires between the VIM and the other controllers 6 5 Summary The validation calibration and testing methodology was presented with specific examples carried out both at UOIT and EPA The overall results of the vehicle performance and the team s success at the EcoCAR competition were then discussed Finally results of projects lead by the author were described The next chapter concludes by highlighting the author s contributions to the UOIT EcoCAR project recommending future work to be considered and presenting a closing argument Chapter 6 146 Chapter 7 Conclusion 7 1 Thesis Summary The primary goal of this thesis was the electrical design and implementation of a full function electric vehicle powertrain architecture based on GM s existing pre pro
157. n the bus which can be of a maximum length of 2 active error flags Using recessive bits a passive error flag is much simpler and is always 6 recessive bits long regardless of the superposition of other passive error flags from other nodes that might overlap with the error delimiter However it can be overwritten by an active error flag 23 Appendix B 164 The error delimiter is 8 recessive bits following an active error flag or a passive error flag In the first case there is a superposition of passive error flags and error delimiters that are overwritten by active error flags Once the last active error flag is completely sent after 6 to 12 dominant bits the first recessive bit of the overall error delimiter appears In the second case no node is sending dominant bits resulting in the start of the overall error delimiter after exactly 6 recessive bits during the superposition of passive error flags 23 6 Overload frame There are 2 causes resulting in the generation of an overload frame 23 e A receiver requiring a delay before being able to decode the next data frame or remote frame A node detecting a dominant bit in the first and second bit positions during the intermission field of the interframe space In both cases an overload frame represented in Figure 58 has the general form of an error frame containing active error flags In other words 6 to 12 dominant bits as superposition of overload flags and 8 re
158. nergy Consumption Results Regenerative Braking Activated SUMMAE iS 138 Table 55 Influence of Regenerative Braking on the 139 Table 56 Performance projection based on EPA Results 142 Table UOIT Team VIS ies ci 144 Table 58 Powertrain Specification of the Full Electric 5 159 Table 59 Simulation Results of the Full Electric 51 159 Table 60 Powertrain Specification of the EREV 60 5 159 Table 61 Simulation Results of the EREV 60 5 160 Table 62 Powertrain Specification of the PHEV 30 5 160 Table 63 Simulation Results of the PHEV 30 51 160 dentlfie 162 Table 162 Table 66 Post Office Analogy ec nere nte 174 Table 67 Comparing LIN and CAN 24 176 Table 68 Typical LIN localized support areas eene 177 Table 69 Enabling nay ea 187 Table 70 Parameter list 12504 193 obe Tte 188 Table 71 VIM Port description HVCM and 194 Table 72 UOIT CAN bus dictionary 210 xvii gt gt gt
159. ng System Electronic Stability Control Control throttle position spark plug and fuel injection Electric Power Steering Fuel System Control Module Amplify steering effort according to driving conditions Control fuel pump and fuel level sensor OnStar system Communication for security remote diagnostics hands free calling and navigation Transmission Control Module Automatic transmission Traction Power Inverter Module e Hybrid Control Processor Motor Control Processor Motor Control Processor High voltage and motor controller Dark grayed ECUs were removed from the vehicle see Section 4 3 2 and Table 35 Chapter 4 71 Table 33 List of ECUs GMLAN SW ECUs on GMLAN Single Wire SW Body Control Module Communication Gateway Module Main controller Gateway between HS PTE and SW buses Not an ECU SAE 11962 defines the connector used for the OBD II interface Give access to all CAN buses for vehicle diagnostics Communication for security remote diagnostics hands free calling and navigation On board diagnostics II OnStar system Air Conditioning Compressor Module Control air conditioning Automatic Occupant Sensor Occupant seat sensors Electronic Climate Control Cabin climate control panel Instrument Panel Cluster Control the dashboard Integrated Radio Chassis Radio and GPS Navigation Remote control
160. ng battery life by terminating the charge early e Balance the cell voltages Although there are many BMS systems available today in 2009 the choices were limited and one of the most advanced designs employing a flying capacitor cell balancing strategy CAN bus communication and programmability was available thru a British company named REAP Systems The selected BMS model employed was the REAP 514 and its specifications are listed in Table 6 12 It should also be mentioned that one of the 7 BMS modules was reconfigured by the manufacturer to be compatible with 10 cells instead of 14 for a battery totalizing 94 cells A picture of the BMS board is displayed in Figure 10 and an in vehicle layout in Figure 11 Table 6 BMS14C Specifications 12 1 module Final BMS Design of modules 1 7 of cell monitored 14 94 of thermistor 7 49 Weight g 68 476 Size mm 132 x 72 x 19 7 132 x 72 x 19 Figure 10 BMS14C 1 module Chapter 2 23 I g 4 i ed 4 a 4 dha U AE CHA 4 b Figure 11 In vehicle BMS layout a Front ESS b Underside ESS Chapter 2 24 2 7 2 On board charger 5 Brusa NLGS13 3 3kW BRUSA chargers were required by the competition organizers This charger is robust and has very flexible configuration settings making it versatile enough to accommodate the needs of 16 universities each having a unique architecture For example
161. ng cheaper hardware LIN is often used as an alternative to single wire CAN to control functions in a localized area However single wire CAN networks are not limited to localized areas In such Appendix H 176 applications the LIN master node acts as gateway between the LIN network and CAN environment to extend communications to other ECUs spread all over the vehicle In other words LIN is often used in vehicles as a sub bus via a LIN master node to manage devices in a localized area such as the roof or a seat providing several complementary networks to the main CAN buses Figure 68 illustrates a typical application of a LIN sub bus in a vehicle while Table 68 lists common LIN localized support areas based on reference 24 Table 68 Typical LIN localized support areas Doors Mirror control Mirror switch Window lift Door locks Engine Sensors Small motors Roof Moon roof controls Light Sensor Interior light Visor lighting Seat Occupancy sensor Seat position motor Seat heater Steering column Wheel tilt position control motor Cruise control switches Wiper control Turn control A Mirror motor 4 7 Door lock d 88 LIN master node CAN body Control system Figure 68 LIN sub bus 24 Appendix H 177 Appendix I Detailed Subsystem
162. ng current below 1 00 Voltage rise 5 Min below V Temperature above Temperature rise above Charging current below Charging time of this section above Min Amount of charge in this section above Ah Relative amount of charge in this section above Figure 71 BRUSA Charging profile Mode 3 Constant Voltage 1 2 Booster Mode Not Used BRUSA ChargeStar NLG5 Profile Connection Extras Help R Not connected to unit 211200 4020 30 0 2 3 150 Derating 50 50 0 Figure 72 Booster Mode Operation mode INLG operation mode Ext mains current for Booster OFF m 5 0 Ext mains current for full Booster Power 4 100 General settings lt L Switch off immediately if Appendix I 180 1 3 CAN Mode Not Used Although it has not been tried by the author as shown in Figure 73 and Figure 74 it appears that each BRUSA charger CAN identifiers can be configured allowing dynamic control of multiple chargers from a vehicile integration module BRUSA ChargeStar NLG5 Profile Connection 272277 Help General settings Options Properties cms oan wolage 7020 Resencetempesueirvkeger TC 2790 772222 TH Switch off immediately if Figure 73 CAN Operation Mode and Extras Options Parameters Receive Message Ident
163. nt System BMS 7 REAP BMS14C Charger 5 BRUSA NLG513 Electric Motor Delco S10 EV Motor Controller MES DEA TIM600 Vehicle Integration Module MotoTron 5554 112 00 2 DC DCs MES DEA 1kW Datalogger Vector CANlog4 Ground Fault Interrupt Bender IR470LY Dashboard Graphical User Interface In house 2 8 Summary The different powertrain configurations and practical fuel choices for implementation in today s vehicles to increase fuel economy and reduce GHG emissions were considered by EcoCAR organizers The well to wheel energy use and GHG production were then studied and analysed by the UOIT team for several different architectures The modeling and simulations of the 3 most promising ones lead to the selection of a full function electric vehicle The last section gave an overview of the sophisticated hardware required by modern electric vehicles along with a list of the key electrical components selected for implementation in the prototype designed In the subsequent chapter the theory and vehicular application of the controller area network CAN communication protocol is presented Chapter 2 35 Chapter 3 for Vehicles 3 1 Introduction The powertrain architecture selection process the specialized hardware found in a typical electric car and a description of the components selected for implementation on the UOIT EcoCAR are described in the preceding chapter The goal of this section is to focus on establishing
164. nterference End Of Frame Environmental Protection Agency Electronic Power Steering Extended Range Electric Vehicle Electrostatic Discharge capacitor Energy Storage System Electric Vehicle Fuel Cell Vehicle Fuel Economy Full Function Electric Vehicle Fuel System Control Module Finite State Machine Federal Test Procedure drive cycle GNU Compiler Collection Ground Fault Interrupt Greenhouse Gas General Motors General Motors Local Area Network Graphical User Interface hexadecimal Hybrid Control Processor Hybrid Electric Vehicle Hardware In the Loop High Speed High Voltage High Voltage Control Module Highway drive cycle Highway Fuel Economy Test drive cycle xix ICS ICE ID IDE IGBT IPC IRC ISO I O J K km kW kWh L LCD LED Li ion Li Battery LiFePO LIN LS Isb LSB M MCPA MCPB mi MIL MIS MOSFET MOV mpgge mpg MPG msb MSB Msg Instrument Cluster Screen Internal Combustion Engine Identifier Identifier Extension Insulated gate bipolar transistor Instrument Panel Cluster Integrated Radio Controller Independent System Operator Input Output kilometer kilowatt kilowatt hour Liquid Crystal Display Light Emitting Diode Lithium ion battery Lithium Battery Lithium Iron Phosphate Local Interconnect Network Low Speed least significant bit Least Significant Byte Motor Control Processor A Motor Control Processor B mile Malfunction Indication Ligh
165. nverted before being transmitted over the CAN network The purpose of the scale and offset of a signal is to convert the raw value of a signal to its engineering value The physical value or unit value is the signal value having a scientific or physical meaning by being associated to a unit The raw value or bus value is the signal value transmitted over the CAN bus The scale and offset are adjusting the raw value of the signal to obtain its physical value expressed in the desired unit The following basic linear equation is used to express the relation between the physical value and raw value based on 46 48 physical value scale raw value offset y mxt b Chapter 3 57 3 6 Bus Configurations The basic configuration of a CAN bus is the bus topology Figure 26 illustrates this concept using devices found on an electric vehicle CAN network CANH 1200 1200 High Voltage Control Module Motor Controller Battery Supervisory On Board Management Control Charger System Module Module Figure 26 Bus topology In this topology a main bus of 2 wires exists and devices are connected to it in parallel The length of the wires connecting each device to the bus can vary While this technique works well for small networks electrical noise can be experienced on large networks since every single branch connected to the main bus acts as an
166. o not exchange CAN messages Chapter 5 117 directly but indirectly through the VIM acting as a bridge between them The main reason for this resides in the control strategy chosen for the 3 main powertrain controllers i e BMS OBCM The best solution was to use the VIM to integrate them individually to the vehicle architecture and use RS232 communication to define the initial settings of the devices The motor controller is the only added ECU configured to receive as well as transmit CAN messages The BMS and OBCM were configured to only broadcast messages while the datalogger and ICS were set to only receive messages The next sub sections describe in detail the communication between the VIM and each of these controllers Also a complete list defining their CAN messages can be found in Appendix L 5 6 2 1 Motor Controller The motor controller was the only main powertrain ECU to receive CAN messages as it was only able to function as desired with the CAN communication enabled and receiving periodic CAN commands Three CAN messages need to be sent at 135 ms intervals to the motor controller for proper functioning The first one specifies on the maximum voltage and current discharge rate while driving The second one focuses on regenerative braking and motor activation using these 4 signals maximum regenerative voltage and current regenerative braking enable and traction enable The latter 2 signals are also provided by hardwi
167. of the 7 remaining data bytes Each BMS board sends information on a predefined CAN ID for a total of 7 different identifiers and uses sub IDs to differentiate messages It should be noted that the cell voltages are sent using 2 bytes and only 7 bytes remain in each message Therefore some cell voltages are split between 2 CAN messages and need to be concatenated by the VIM when received In theory when concatenating to a single variable 2 signals of 1 byte received from 2 different CAN messages it is recommended to verify that both signals contain a new value before overwriting both bytes of the existing variable Otherwise a mismatch could occur between the high and low byte of the merged variable However considering the order of voltage magnitude of the high byte it was determined to change slowly enough compared to the low byte of the same variable that such a verification could be neglected The temperature values measured by the NTC thermistors negative temperature coefficient inside the battery module and connected to the BMS boards are transmitted using the same sub ID system Each BMS module also sends a status and problem message The information expressed uses the boolean flags listed below Chapter 5 120 Status flags Start up Control charge Control discharge Equalize cells Monitor battery Do active control Cooling on Heating on Do shut down 5 6 2 3 On Board Charger Problem flags Overvoltage p
168. oltage branch output current and internal temperature In addition to CAN messages a hardwire output signal indicates whether or not the unit is ready for charging Due to the parallel configuration the total charging current is the summation of all 5 charger current outputs and is only monitored by the BMS current sensor Therefore the output current of each Chapter 4 90 unit must be limited to calibrated values and adjusted for each charging phase Table 42 specifies the voltage of each charging phase Table 42 Charging Phases Phase 1 Preconditioning V V lt 282 3 0V 94cells 01 Max Charging Current Available Phase 3 Constant Voltage V gt 394 8 4 2V 94cells 0 07 Phase 2 Constant Current 282 lt lt 394 8 When the battery state of charge is very low i e a voltage under 282 V the first phase slowly charges the battery to a safe voltage level using a low constant current 0 1C Once the safe voltage threshold is reached the charger enters the constant current phase and uses the maximum charge current available For the cell technology used the nominal battery limit is 1 0 C without cooling When the battery reaches the maximum voltage of 394 8 V the last mode is started In this phase the voltage remains constant and the current slowly drops until it tappers off 0 07C Moreover the chargers are programmed to turns off sequentially to increase the accuracy durin
169. on Cranking Conditions Appendix J 200 67 83 417 1023 05 0949 10235 100 0 17V 171102355 350 67 V 049 10255 100 TV 171 10235 350 Volatile Data 2 10015 2 BrakeON boolean P b 12V 0 25 gt 912 7 75 DutjCycle 0 50 gt 12 775 DutyCycle 075 gt 12 7 75 HSOL10r HSOL2 can sink no free wheellng diode HSOL1 57 YeliowOrange 275 YellowPink Figure 92 Radiator Fan Controls Appendix J 201 MetoHawkTaniel 101 FauR Status AcelOORHIgn Acte 2 fa Apes fa Fask Status AccelOORLONw m 2 113 Operator BrkPedalHiIFauk 2 Fal 113 Operator2 Figure 93 Pedals Fault Management Model Appendix J 202 StartPulse H i i data TractionEN Finite State Macrice HV Controller Deta Read AC Test Disabled 0 umm Prox_Test Disabled 0 NEL Figure 94 High Voltage Controller Inputs and Outputs Appendix J 203 lgnKe 0 amp IL StartChrgMode PrechrgChrg entry Relay ESSNegz entry Relay Chrg 1 entry HV_Status 5 z amp IL_ESSN 1 after 10 ESSPosEnChrg entry Relay ESSPos 1 entry HV Statusz IL ESSPos 2 after 10000
170. or frames overload frames the start of frame SOF the acknowledgement ACK and the end of frame EOF Most CAN users for instance application developers only need to understand the general idea of the arbitration field standard or extended identifier ID the control field encoding method of the data length code DLC and the data field endianness 26 The CAN protocol controller takes care of the details such as applying the protocol specifications It automatically handles the protocol and errors However a CAN transceiver is required to convert the bus voltage to a logic level value and to carry out the opposite conversion Even though most microcontrollers supporting CAN have an integrated CAN protocol controller improving performance they still need CAN transceivers sometime called line drivers Nevertheless commercial controllers have these two devices integrated into their circuitry Figure 23 summarizes these 3 cases and Table 17 gives examples of different products on the market Additional theory explaining at an electrical level how the bidirectional analog digital conversion is achieved inside a CAN transceiver chip is covered in Appendix E Chapter 3 43 Microcontroller featuring embedded CAN Controller can aa CANL Transceiver i CANH Controller featuring a CAN Transceiver and CAN Controller embedded in its circuitry Rx ow LEAN HEHE Mi
171. owing Capacity 680 kg 1500 Ib 20 min 72 kph 20 min 72 kph 45 mph 45 mph 457 mm 18 457 mm 18 Cargo Capacity 0 83 m D 686 mm 27 D 686 mm 27 W 762 mm 30 W 762 mm 30 Passenger 38 m 43m lt 51 8m 51 8 m 00 5 mph 123 140 ft 170 ft 170 ft 1758 kg lt 2268 kg Ground Clearance 198 mm 7 8 gt 178 mm 7 210 8 25 Range gt 580 km gt 320 402 971 360 mi 200 mi 250 mi PSAT values were scaled back 6 to arrive at VTS charging efficiency 90 upstream energy included utility factor UF weighted fuel economy FE 2318 kg 5111 Ibs with optional rear callipers GHG green house gas emissions are CO2 warming equivalents on a well to wheel basis mpgge miles per gallon gasoline energy equivalent Chapter 2 16 2 7 Selected Components The examination of the state of the art architectures for green vehicles achieved so far in this chapter led to choosing a battery electric vehicle design along with establishing the performance goals for the prototype After defining the architecture into which the GM donated vehicle was to be converted the component selection process began The purpose of this section is to acquaint the reader with the advanced components required by EVs and explain their selection process Unfortunately for the UOIT EcoCAR team not all the ideal components or first choices could be acquired for v
172. p Technology Inc 2007 0570185 http www microchip com Website 58 Tritium WaveSculptor CAN Bus Communications Protocol Specification Datasheet 20 August 2007 http www tritium com au index html Website 59 Provencher Hugo Embedded Real Time System Development for Electric Vehicles Faculty of Engineering and Applied Sciences UOIT ENGR 5910G Embedded Real Time Control Systems 2011 02 16 60 CAN in Automation Safety in field bus systems Website http www can cia org index php id 50 61 ON Semiconductor MMBZ15VDLT1G MMBZ27VCLTIG SZMMBZISVDLTIG SZMMBZ27VCLTIG 40 Watt Peak Power Zener Transient Voltage Suppressors Datasheet Semiconductor Components Industries LLC 2012 http www onsemi com pub_link Collateral MMBZ15VDLT1 D PDF 62 NXP Semicondutors PESDxS2UAT series Double ESD protection diodes in SOT23 package Datasheet 2004 February 18 http www nxp com documents data sheet PPESDXS2UAT SER pdf 63 Vector Vehicle Networks An Introduction to CAN Webinar Vector CANtech Inc 2012 64 RM Michaelides software amp elektronik Connecting PCs to mobile or stationary CAN networks Website http www rmcan com index php id273 amp L 1 References 157 65 Kvaser Advanced CAN Solutions Home Products CAN USB Website http www kvaser com en products can usb html 66 Vector VN1600 Website http www vector com vi_vn1600_en h
173. pares major characteristics of LIN and CAN Table 67 Comparing LIN and CAN 24 Feature LIN CAN Network topology Bus Bus Number of wires 2 1 Maximum data rate 20 kbps 1 Mbps Communications UART based Controller based method Network access xn Hated Non destructive transmission 1 master up to 64 128 typically limited by physical Node support 15 slaves layer or higher layer protocol The number of nodes a CAN environment can support shown in Table 67 is different from the details provided by Table 14 The literature does not agree on a specific limit of nodes allowed on a CAN bus Some documents define the maximum number of nodes as 30 23 others as 128 24 The important fact to remember is that the quantity of nodes supported by a CAN bus is dependant and limited by its load and data rate Regardless of these 2 criteria a CAN network can support more nodes than a LIN environment Being controller based CAN is a more expensive solution compared to a universal asynchronous receiver transmitter based local interconnect network i e UART based LIN However according to Table 67 CAN is 50 times faster more efficient more flexible and more robust than LIN 24 Moreover LIN is only designed for automotive applications whereas CAN has numerous uses beyond vehicles The data rate of a LIN environment might be 50 times slower than a high speed CAN bus but is similar to a single wire one Havi
174. pecific signals recorded at 1 Hz after each competition event Having designed and built their own high voltage battery pack the UOIT EcoCAR team was asked by the organizers to record 3 signals in addition to those defined in the rules The list of recorded signals is Motor temperature Vehicle speed km h Motor speed rpm Motor torque Nm Acceleration pedal position 96 Brake pedal on off Transmission shift lever position P R N or D Battery current A Battery voltage V Battery temperature State of charge 96 Battery maximum cell voltage mV only UOIT Battery minimum cell voltage mV only UOIT Battery maximum delta cell voltage mV only UOIT Here s the list of signals also read and displayed by the instrument cluster screen Transmission shift lever position P R N or D State of charge 96 Battery current Battery voltage V Chapter 5 122 5 7 Summary As indicated in the present chapter the vehicle integration module MotoTron controller plays a key role in the electrical and control system integration strategy A brief overview of the controller hardware and software was given Its various features were identified and described such as encoders decoders high voltage control module CAN signal emulator and gateway function The final results of UOIT s electrified vehicle prototype are summarized in the next chapter Chapter 5 123 Chapter 6
175. pin M12 connector 58 Appendix D 168 It is also interesting to note that in the automotive industry the connector used to access the CAN buses on a vehicle is the SAE J1962 on board diagnostics OBDII connector as shown in Figure 64 based on reference 50 Pin Description 1 Single wire CAN Chassis ground Signal Ground 2 3 CAN High ELE 4 Chassis ground 5 Signal ground 6 CANI High 7 8 9 10 s 11 CAN2 Low 12V Battery 12 2 L CAN1_L E reserved _ 14 CANI Low 15 16 Battery 12V Figure 64 SAE J1962 connector Appendix D 169 Appendix E CAN Transceivers Figure 65 represents the block diagram of a basic high speed CAN transceiver the MCP2551 manufactured by Microchip 41 It is equivalent to the TJA1050 made by NXP Semiconductors known as Philips Semiconductors before 2006 43 The TJA1041 is more sophisticated and will not be discussed in this paper although its basic functionalities are similar to the MCP2551 and TJA1050 The digital signals are on the left of the figure while the analog signals are on the right The three primarily blocks in this diagram are the 2 5V voltage source the driver control and the receiver both interfacing between the digital and analog world 5v TXD Thermal Dominant Shutdown VDD Detect 5 TXD Drver X lt Control y Slope Po
176. quired to troubleshoot these ECUs were gathered in the glove compartment 5 4 6 2 Safety Having left unmodified the sensing and diagnostic module SDM and stock crash sensor along with the single wire GMLAN the airbags of this prototype will deploy in the event of a crash As explained in Section 4 6 5 an additional inertia switch was added and is used as an interlock to disable the high voltage powertrain when triggered The airbag CAN messages could also have been used by the VIM as redundancy to this additional crash sensor but was not programmed due to time constraint 5 5 HVCM The high voltage control module modes which were detailed when discussing the powertrain control strategy in section 4 4 1 are programmed in the VIM mainly using a Simulink development tool named Stateflow Figure 45 shows the high level inputs and outputs of the HVCM subsystem of the VIM Simulink model The detailed Simulink model and Stateflow can be found in Appendix J Chapter 5 113 Ignition ESS Contactor Ignition Cranking ESS Contactor Interlocks HVCM ue High Voltage a x 7 Control Module Precharge Delta Voltage gt Traction Enable ESS Contactor Feedback HVCM Status ESS Contactor Feedback NS Y Figure 45 HVCM I Os 5 5 1 Inputs In Stateflow the inputs are used as conditions to enter and exit states wherea
177. r competition events disabled that functionality for the following few months until the controls team spent an evening troubleshooting the problem It is also important to mention that the turning signals and hazards were not impacted by the conversion The car ready and GFI lights were implemented in one of the VIM software release during the last year of design updates By converting the vehicle into an EV a number of control modules were no longer required for standard vehicle operation and needed to be removed The fuel system control module FSCM was judged to be the easiest module to emulate since it only broadcast CAN messages and did not receive any It was therefore selected to be the first ECU to be removed from the vehicle and emulated by the VIM This emulation was Chapter 6 126 required to ensure no error codes were passed into the vehicle s ECUs Through successful emulation and removal of the FSCM the remaining modules were also removed i e the stock battery pack control module BPCM followed by the engine assembly where the TPIM ECM TCM and APM were mounted The second key step for the migration from the test bench to the vehicle was to hook the temporary 400 V lead acid battery pack without BMS to the accelerator brake pedal motor controller and the MotoTron controller VIM Also for this step the motor and its controller were connected and installed in the engine bay At that point the VIM was integrated to t
178. rated 26 SAE JU7 72 Charing Mlet 2 a Delphi 510 with inverter chargers and DC DC b MES DEA err ER DE A 29 MotoTron Controller ECM 5554 112 0904 C00 M 16 31 MES DEA DC DC Converter 1 kW 181 32 CANIOS4 19 suiit or De dte e VR EO a o ERG 33 Bender IR470 LY 20 d etc et et eu eee ace 33 Dashboard Graphical User interlace sivas ea et et e Rt DR 34 TODOlO 39 Data frame general structure 25 i erred e 42 Remote frame general structure 25 2 rrr etnies 42 CAN Transceiver CAN Protocol Controller and Controllers 44 Representation of a signal in the 3 main CAN display formats 52 Conversion from Intel Standard seen 53 BSUS topology NT 58 Bus topology minimizing NOISE 59 Hybrid ER ORGAN I iE 59 Daisy chain with short SUGS 60 Daisy chain with twisted wires in the connector 521 60 Validation Process Diagram 67 Conversion Methodology mp d ene 68 Stock CAN buses GMLAN HS b GMLAN PTE c GMLAN SW 73 Modified CAN buses GMLAN HS b UOIT CAN bus 78 HV Control Module flow diagram 2 79 High Voltage Schemalic dh Co recto dept e tis edt em 82 BMS I Os and Master Slave
179. ration Figure 52 Arbitration mechanism 36 The grey area on the left overlaps the 3 bit of the arbitration field of each signal Node 1 and 3 have a dominant bit 0 while node 2 is attempting to send a recessive one 1 Since the 39 bit of node 2 s identifier is overridden node 2 realizes that one or several messages of higher priority are trying to access the bus Thus it will stop transmitting and re attempt to send its message as soon as the bus is free again A similar scenario happens to node 3 on the 6 bit of the identifier shadowed by the right grey area Trying to send a recessive bit while node 1 sends a dominant one node 3 ceases sending its message Node 1 wins the arbitration thus gets access to the bus and finishes transmitting its message As detailed under Appendix E by sending a dominant Appendix B 161 bit physically node 1 is pulling up CANH and pulling down CANL while node 3 is trying to maintain both CANH and CANL at 2 5V and loses arbitration Node 3 detects the voltage difference on the line then gives up and relinquishes till the bus is idle again During the arbitration field whenever a node is attempting to transmit a bit that is different than the physical result on the CAN bus which is illustrated as the bottom signal on Figure 52 the node in question stops transmitting and awaits the end of frame before re attempting to send it once more From another perspective the node winning the arbi
180. re 49 UDDS Cycle Portion with REGEN Activated 6 3 2 3 Ground Fault Interrupt Issue Upon inspection back at our shop it turned out the main fuse had blown so there was no ground fault present rather the indicator was triggered by a loss on continuity on the battery circuit This feature is a little known function of the ground fault interrupt unit used from Bender The blown fuse was traced to a shorted IGBT leg in the inverter Since the vehicle was not under any high demand situation through the UDDS cycle it is theorized that increased high voltage spikes in REGEN caused the IGBT to avalanche To remediate this problem the high voltage bus capacitance was increased the high voltage leads were re routed for reduced inductance and the inverter signal lines were isolated from ground noise by floating its supply and opto isolating signal lines Noise on the vehicle ground had been a consistent problem and may have caused false triggering of the IGBT taking away all dead time and causing shoot through Figure 50 shows the inverter apart with the shorted IGBT on the bench Chapter 6 139 gt pp Figure 50 Inverter with shorted IGBT In any case noise issues had been at the root of proper RS232 communications with the inverter and frequent loss of communications with the device while recording data It is believed but not proven that the majority of ground plane noise came from capacitive coupling between the batter
181. red inputs The arbitration between the CAN messages and the hardwired inputs is done as an OR gate i e as soon as one of the values whether CAN or hardwired is a logic 1 the motor controller consider the signal to be a logic 1 More testing needs to be done on the inverter to better understand the arbitration between the CAN messages and the discharge charge limits programmed using the RS232 communication However it is seems logical to the author that the CAN messages would have priority since they are dynamically sent The third message defines the maximum acceptable voltage while using an AC charger For the UOIT prototype the maximum regenerative voltage and the maximum charge voltage were set to the same value 395V which is 4 2 V cell Chapter 5 118 Two CAN messages broadcast at 12 ms rate the motor controller One contains boolean variables for the status of all the inputs outputs and alarms present whereas the other one provides the input battery voltage vehicle speed motor speed and motor torque 5 6 2 2 Battery Management System Numerous CAN messages are sent by the BMS to the VIM from a battery metrics summary to detailed individual cell information A status and problem frame is also sent along with commands for a charger Due to the presence of 5 on board chargers the built in charging commands broadcast on the network by the BMS could not be used and a custom charging strategy had to be employed
182. roblem Undervoltage problem Temperature too high for charging Temperature too high for discharging Hardware overvoltage problem Hardware undervoltage problem Charge current too high CAN response failure CAN receiver buffer overflow Configuration error may be set during start up Doing start up is always on during start up Similar to the BMS even though it can receive messages such as commands from a BMS it is not part of the control strategy and the on board charger only transmits information Since most of the troubleshooting was done manually using a RS232 communication only 1 message from the exhaustive list of CAN messages sent by the charger at a 100 ms rate was used It provides the AC voltage and current coming into the charger and the DC voltage and current coming out of the charger i e charging the lithium battery Although it was intended to program each of the 5 chargers to communicate with the VIM only the one shutting off last in the cascade sequence as explained in Section 4 5 2 ended up being programmed due to competition time constraints Chapter 5 121 5 6 2 4 Datalogger and Instrument Cluster Screen The datalogger and the ICS are grouped under the same sub section since the ICS only displays some of the datalogger s recorded CAN messages and does not require any additional messages from the VIM For the EcoCAR competition it was mandatory for each team to provide a log of s
183. s built The mechanical integration of the prototype should be finished about 2 months prior to a competition to give the controls and high voltage teams enough time to test and troubleshoot their systems A lack of testing equals a lack of results Stick to the timeline If it does not exist create one The important is not to have the best or perfect design for the competition but a reliable prototype Make priorities and do not waste time on little things in other words do not let the perfect be the enemy of the good At one point but not too early the team need to stop cuddling their car and push it to its limits more likely resulting in parts breaking This step is essential to learn and test the robustness of the prototype however thorough debugging and calibrating need to be completed first Chapter 7 152 Skip no steps in the design integration testing or calibration phase Log all CAN buses in the different ignition key modes before disassembly Insure the car can be debugged conveniently For example put all the communication connectors at the same location i e the glove compartment Know your teammates and learn the best way to deal with each of them considering their strengths and weaknesses Know your personal limits when you are productive and when to rest When building a prototype the design is in constant evolution Wires should not be shortened until the system has been thoroughly tested and the desi
184. s the outputs are action variables updating the high voltage hardware of the powertrain It is important to distinguish the difference between the ignition ON ignition key and ignition cranking start pulse signals The ignition ON signal is 1 when the key is either in the ON or cranking positions while the ignition cranking signal is only 1 when the key is held in the cranking position As explained in section 4 6 4 the interlock signal is generated using an and gate employing the vehicle s hardwired interlocks as inputs The AC connect signal equals 1 when either high voltage is detected at the charging inlet or the SAE J1772 charging connector is physically plugged in The precharge delta voltage signal turns to 1 after the precharge is completed i e when the voltage difference between the voltage read at the BMS and the one measured at the motor controller is less than 10 volts Interlock ESS and ESS are feedback signals from the vehicle s main contactors indicating whether they are opened or closed Chapter 5 114 5 5 2 Outputs The ESS ESS and charge contactor outputs used to ground the negative side of each contactor s coil using low side outputs of the MotoTron These outputs are rated for sinking up to 7 A which is enough to sustain the current peak created by the coils of these high voltage contactors As discussed previously the traction enable signal is fed to the motor controller and is activated when the hig
185. saving measure the auto mode was therefore chosen and each charger was configured through RS232 communication for autonomous operation The control strategy for this is covered in Section 4 5 2 The discharge circuit is designed to have the same performance as the precharge circuit thus employs similar hardware i e a RY4S UL DC12V normally closed relay from IDEC normally open for the precharge circuit 160 Q 100 W resistor and 6 gauge wire There is no protection required for the GFI since the HV leads are only probes and no current is drawn Both DC DCs utilize 600 V 10 A fuses and 12 gauge wire The air conditioning is connected to 600 V 35 A fuses through 10 gauge wire whereas the heater uses 12 gauge wire and 600 V 25 A fuses 4 5 Control Law A MotoTron controller model ECM 5554 112 is used in order to integrate new systems within the GM donated SUV It emulates messages required by the stock controllers remaining on the CAN network The MotoTron emulates these along with providing a communications bridge to the CAN networks for the added HV systems Thus the vehicle has 3 CAN buses HS GMLAN High Speed General Motors Local Area Network PTE Power Train Expansion GMLAN and SW Single wire GMLAN Chapter 4 84 As explained in Section 4 3 2 it was planned to implement extra CAN bus for the systems but for wiring simplicity the UOIT CAN bus was merged to the PTE GMLAN The combined bus also doubles as
186. second stop sign pause about halfway through a UDDS cycle The subsequently scheduled highway cycle with REGEN the last in the test sequence planned for that day was thus not run No quick fix was found to trace the problem so testing ended 6 2 5 Vehicle Integration at Year 3 Competition Unfortunately for several different technical reasons the UOIT team did not successfully complete the safety scrutineering at the final EcoCAR competition thus was not allowed to compete in any of the dynamic events However during the week of Chapter 6 133 competition lots of work was accomplished the vehicle As an example with 2 new battery modules just finished and assembled overnight prior to shipping the vehicle to the competition at Milford Proving Ground the team had to spent most of the competition integrating them to the vehicle Beside the battery the integration involved a more powerful MotoTron controller a junction box interconnecting most of the controls wires for the conversion and a new high voltage distribution box Even though the high voltage control module operation was programmed in the new MotoTron controller and had been tested first through Simulink simulation and then on a test bench it had never been operated with the new high voltage distribution box Since the controls connected to the junction box needed to be functional first debugging of the HV distribution box was kept for last Therefore a lot of work
187. simulation using software designed by Argonne National Laboratory known as PSAT Powertrain System Analysis Toolkit recently renamed Autonomie 4 The powertrain specifications i e motor and generator power battery energy capacity and vehicle weight and the simulation results of acceleration energy consumption for different drive cycles and vehicle range are Chapter 2 13 detailed in Appendix A Table 2 presents a summary of the simulation results using the EcoCAR North American energy mix Table 2 Performance Simulation Summary 7 PHEV 30 EREV 60 Full Electric Avg Eco mpgge 60 6 75 4 93 5 Electricity Wh km 246 241 222 GHG g km 178 162 155 Utility factor 0 512 0 739 0 971 Acc 0 97 km h s 8 9 8 7 9 5 Weight kg 2027 2043 2139 Electric Range km 47 95 444 Total Range km 600 600 444 It is to be noted that all 3 architectures meet the competition performance requirements which are specified in Table 3 Table 3 EcoCAR Competition Performance Requirements 5 Specifications GHG Emission lt 217 g km Accel 0 97 km h 14s Weight lt 2268 kg Range gt 320 km The battery capacity was necessarily chosen to achieve a range greater than 320 km since the test procedures for range evaluation at competition were not clearly outlined by the organizers in the rules These 3 simulations indicated the full electric architec
188. situated between the main 2 gauges could not be reprogrammed or modified due to its embedded firmware Instead a custom one was installed at the hybrid gauge position on the left The in house screen controller communicates with the MotoTron over the UOIT CAN bus and enhances the information displayed to the driver It is located behind the instrument cluster The PRND maximum and minimum cell voltage SOC based on energy consumption battery temperature and expected range are examples of variables displayed on this backlit LCD Screen 4 7 2 Troubleshooting Interfaces for Engineering Development Engineering a prototype having the complexity of an electric vehicle requires a lot of debugging and good troubleshooting tools In the design of the UOIT EcoCAR the 4 programs interfacing with the 4 main controllers of the powertrain i e the MotoTron BMS charger TIM were key to the project s success A specialized program was provided by each manufacturer except for the BMS Being the integration module the MotoTron is connected to all CAN buses and in communication with most modules Therefore it was used as a diagnostics tool via its user interface MotoTune This gives access to any desired variable of the controller for monitoring calibrating and testing Having a visual layout similar to an Excel spreadsheet variables could be dragged and dropped easily to create displays to diagnose the vehicle Three main interfaces were created and ca
189. ssion within a byte and is fixed in CAN communication as explained previously 3 5 2 3 Message Progression Now that the byte ordering bit ordering byte numbering and bit numbering are defined the bit numbering progression from one byte to another called message Ch apter 3 progression must be defined The message progression assigns a bit number from 0 to 63 to each bit in the message The message progression can be 47 46 Forward numbering from the first byte Backward numbering from the last byte In other words regardless of the bit numbering technique used forward means the bit numbering begins from the first byte sent bit 0 is in byte 0 and backward means the bit numbering begins from the last byte sent bit O is in byte 7 In order to send and receive CAN data in a coherent manner it is important for CAN users to know the bit number of each bit stuffed in a message i e the overall bit numbering This results in 4 possibilities expanded from references 47 48 49 Table 24 Message progression Msg Bit Data Field Message Progression Numbering Byte 0 Byte 1 Byte2 Byte3 Byte4 Byte5 Byte6 Byte 7 Sawtooth 7 0 15 8 23 16 31 24 39 32 47 40 55 48 63 56 Sequential 0 7 8 15 16 23 24 31 32 39 40 47 48 55 56 63 Sawtooth 63 56 55 48 47 40 39 32 31 24 23 16 15 8 7 0 Sequential 56 63 48 55
190. st the added powertrain electrical subsystems and the design and implementation of an in house instrument cluster screen with CAN capabilities The knowledge gained by the author through years of developing electric vehicle prototypes is captured by a list of hands on personal best practices Chapter 1 5 1 5 Summary of Thesis Sections Chapter 1 introduces objectives of this research Chapter 2 explains the UOIT EcoCAR powertrain architecture selection process and provides an overview on the EV powertrain components chosen Chapter 3 covers the fundamentals of the CAN communication protocol along with hands on theory and applications useful to many CAN users in the automotive industry Chapter 4 describes in detail the electrical and control system integration strategy Chapter 5 examines the vehicle integration controller more specifically the vehicle integration module VIM making this strategy possible as well as its model based program Chapter 6 presents with an electrical focus the tests and results of the UOIT EcoCAR project Finally chapter 7 concludes by stating the author s contributions highlighting directions for future research and development and giving hands on best practices Chapter 1 6 Chapter 2 Architecture Selection Process 2 1 Introduction In the previous chapter the goals of the research and development efforts described in this document were introduced This chapter addresses why the UOI
191. station where the battery pack was charged in preparation for the full depletion test occurring the next day Due to some remaining cell imbalance the battery pack was charged to an estimated state of charge SOC of 96 instead of 100 The AC input voltage current and power of the charger Chapter 6 131 along with their DC output values were measured and logged to study the charger efficiency total battery capacity and charge profile 6 2 4 2 Day 2 The second day a full discharge to approximately 096 of the pack capacity was sought while running a variety of drive schedules To accomplish this level of discharge different drive cycles and steady speed driving were selected by EPA and UOIT team in consultation The EPA crew was mainly interested in gauging performance nominally at 100 80 5096 and 20 states of charge The UOIT team did not expend effort in developing unique drive cycles rather tried to run the standard EPA cycles as much as they were allowed US06 was not permitted although it would have been indicative and helpful to test the vehicle on a more demanding drive cycle and prove the robustness of the vehicle However the team was allowed to run the vehicle at sustained speeds of 60 mph The schedule devised to perform this battery depletion test is shown in Table 50 Table 50 Drive Schedule Sequence Drive Schedule UDDS 1 UDDS 2 UDDS 3 UDDS 4 60 mph until 65 SOC UDDS 5 Highway FTP 1
192. subdivisions of the high speed CAN specification for vehicle applications SAE J2284 and the specification of the single wire version SAE J2411 50 These specifications are based on the CAN definition of the physical and data link layers of the open system interconnection OSI reference model However they define them more precisely i e with more restrictions For example these SAE specifications are specific to using non shielded wires Table 16 SAE specifications for CAN buses applications in vehicles 50 Name Description Status SAE J2284 1 High speed CAN for vehicle applications 125kb s Issued 07 Mar 2002 SAE J2284 2 High speed CAN for vehicle applications 0 250kb s Issued 07 Mar 2002 SAE J2284 3 High speed CAN for vehicle applications 0 500kb s Revised 02 Mar 2010 SAE J2411 Single wire CAN for vehicle applications 33kb s Issued 14 Feb 2000 Chapter 3 42 SAE J2284 3 providing a higher baud rate and SAE 72411 presenting a cheaper solution seem to be more common in vehicles then SAE J2284 1 and SAE J2284 2 Being the only CAN specification updated and re published by the Society of Automotive Engineers since their first release it is safe to assume that SAE J2284 3 will continue to be used for automotive applications in the foreseeable future 3 3 CAN Transceiver and Protocol Controller It is not really necessary to know all the details in the transmission management of err
193. t check engine light Manual Interrupt Switch Metal Oxide Semiconductor Field Effect Transistor Metal Oxide Varistor miles per gallon gasoline energy equivalent miles per gallon Milford Proving Ground Most significant bit Most Significant Byte Message XX NiCd NiMH NRZ NTC OBCM OBD II ODE OEM OnStar OSI PbA PDE PHEV PRND PSAT PSD PTE PTW PWM RAM RFA RMS ROS Rs RTI RTR RXD SAE SCM SDARS SDF SDM SI Nickel Cobalt Manganese Nickel Cadmium Nickel Metal Hydride Non Return to Zero Negative Temperature Coefficient On Board Charger Module On Board Diagnostic Ordinary Differential Equation Original Equipment Manufacturer OnStar System made by General Motors Open Systems Interconnection Lead Acid Partial Differential Equation Plug in Hybrid Electric Vehicle Park Reverse Neutral Drive Powertrain System Analysis Toolkit Power Split Device Powertrain Expansion Pump to Wheel Pulse Width Modulation Random Access Memory Remote Function Actuator Root Means Square Roll Over Sensor Slope Control Real Time Workshop Remote Transmission Request Receive Receive Digital Society of Automotive Engineers Supervisory Control Module Satellite Digital Audio Receiver System Synchronous Data Flow Sensing and Diagnostic Module Spark ignition SOC SOF SOH SRR SUV SW TCM Temp TIM TPIM TTL Tx TXD UART UDDS UF UOIT US US DOE Vac Vac Vpp
194. t day Table 51 shows the detailed results of each drive schedule i e the number of miles driven cumulative energy energy for a schedule energy consumption per mile the extrapolated range and the estimated state of charge SOC Table 52 summarizes the results found in Table 51 and displays an extrapolated range of 222 7 miles for one charge 96 to 4 8 SOC with an enabling regenerative braking and with the temporary Li ion battery pack It should be observed that the extrapolated range is more conservative than the total miles driven of 232 967 miles On the third day regenerative braking was tested and calibrated It should be noted that a second current clamp was added for more accuracy As explained in Section 6 2 4 3 and reflected in Table 54 a ground fault went off during the UDDS cycle 10 and the subsequent highway cycle with REGEN activated could not be run Therefore a combined estimation of UDDS 9 and 10 was made and can be found on the last row of Table 53 The REGEN calibrations were changed on the fly within a single UDDS cycle Therefore the energy consumption in Table 53 for UDDS 9 and 10 is a mix of different calibrations of REGEN However these calibrations mainly affected the drivability thus the energy savings achieved from one calibration to another were similar That is why Chapter 6 136 the results of UDDS 9 and 10 were still compared to the energy consumption of the cycles ran without REGEN the day befor
195. t sends this information over CAN to the VIM controlling the charge and main contactors along with the signal enabling Chapter 4 94 and disabling the charger high voltage DC output The BMS built in charging and discharging relay switches could also be used as additional safety but were not implemented in UOIT s prototype The VIM reads the information provided by the BMS over CAN bus and calculates the delta between the minimum and maximum cell voltage under load conditions to detect defective cells It activates the TIM s emergency mode when a defective cell is established to limit the output current of the motor controller thus reducing the stress on the cell A similar safety feature exists in the VIM for cell temperature As it is explained in Section 4 6 5 there are 2 levels of safety reduced current and OFF HV contactors open 4 6 3 Voltage Current and Temperature Management The powertrain has 4 main controllers the BMS charger TIM and MotoTron The first 3 monitor voltage current and temperature and can enter different safety modes depending on the values sensed They also send the information to the VIM which can take additional actions when abnormal situations are detected 4 6 4 High Voltage Interlock HV interlock lines are installed on the HV hardware i e all the HV box lids HV connectors cables HV controllers and E Stops and are monitored by the VIM Figure 40 shows the interlock configuration of the EV pro
196. th an existing chassis came with challenges such as cooling electrolyte leaks high voltage control and cell monitoring and balancing From an architectural standpoint it seems less challenging to build an EV from the ground up to accommodate a battery able to provide a range greater Chapter 7 147 than 400 km However low sales volumes of electric vehicles generally preclude developing bespoke chassis designs thus UOIT demonstrated what lies in the realms of possibility The end results of the VIM integration prove that the model based design approach and upgraded controller used for rapid prototyping were adequate It provided a fast programming method the flexibility to calibrate variables on the fly compatibility with a broad range of sensors and actuators the capability of communicating on multiple CAN buses and was even left with enough controller resources for future additions to the control system if required Results also indicated it is entirely possible to remove a considerable number of ECUS mainly powertrain ones from a vehicle while avoiding DTC generation if a proper CAN emulation method is employed and the physical CAN buses are reconstructed The combination of having the GM CAN dictionary alongside the in depth analysis of the vehicle CAN logs made an accurate static and dynamic CAN message emulation possible Not being provided with the CAN database static emulation would have potentially been possible by st
197. the first byte of the data field as a sub identifier often referred to as a mode When the identifier is not already divided in two the first byte might also be called message identifier This method usually requires more coding in the controller to having a fixed identifier of 600 and using sub IDs Table 19 Sub identifiers decode the information Table 19 details a battery management system BMS message BMS TX ID 0x600 Sub Identifier 5 0 Byte Byte Byte Byte Byte C Sub ID 1 3 6 7 ns Problem Fags 7 600 Enor Ganse ActiveEwor Waring Measar 4 001 Curent Voltage SOC Temp Master nso Temperature 602 Average Mm These techniques not only useful to rapidly structure custom CAN buses but are also standardized and used as a basis for communication protocols based on CAN Chapter 3 toon Susie ni Fiss Enor tags Act Rese 2 0000010 Bus Measur 8 Bus Voltage Bus Cument 3 Velocity Motor Velocity Vehicle Velocisy are ena aa 45 3 5 CAN Message Definition This section defines some common terms used to discuss CAN messages A proper definition and understanding of these expressions is essential when creating custom CAN databases to interconnect controllers using different transmission display format This is likely to be the case when converting a vehicle
198. the programmed logic are present in the top level subsystem A CAN definition block is added for each of the 3 CAN buses and a fault manager definition block is require for the acceleration and brake pedal fault detection and mitigation 5 3 2 Model Structure The main program of the VIM is found within the Foreground non virtual subsystem and is arbitrarily divided into 18 virtual subsystems Note this does not mean non virtual subsystems can not be used but rather means their use was not deemed necessary considering that the main program the Foreground is evaluated every 5 ms Consequently every sections of the main program is evaluated at a 5 ms periodic rate i e the CAN message emulation HVCM I O decoding and actuating fault management etc For example the acceleration pedal potentiometer is read every 5 ms and the fault manager outputs an updated status indicating whether or not the accelerator is faulty within 5 ms in other words before the next accelerator potentiometer value is read Appendix J Figure 85 shows the core model of the VIM Most of these subsystems are for CAN communication The emulated ECUs on high speed and powertrain expansion GMLAN have their own subsystems and so do the added UOIT powertrain controllers The I Os subsystem contains the hardware dependent blocks reading the MotoTron input pins connected to sensor outputs and actuating the MotoTron output pins connected to actuators The high volt
199. the regenerative torque value P201 in DRIVE 20 200 Time to decrease the regenerative torque value from P201 to zero in DRIVE Default Value Para Metis Description P123 P163 P206 P207 Maximum torque in ECONOMY 100 0 100 0 Tnom P208 Max RGEN torqueinECONOMY 200 200 P209 1200 1200 RPM P210 Timetoreachtorque value P207 in ECONOMY 100 100 s Time to decrease the torque value from P207 to zero in eh ECONOMY o Time to reach the regenerative torque value P208 122 o oae Time to decrease the regenerative torque value from P208 Ram to zero in ECONOMY 0 20 5 P214 Maximum torque in REVERSE 150 0 150 0 9o Tnom P215 Max RGEN torque in REVERSE 200 200 9 Tnom P216 RPM in REVERSE 2000 2000 RPM Appendix I 190 Default gd Description Range EcoCAR Unit Value Value P217 Time to reach torque value P214 in REVERSE 10 100 S Time to decrease the torque value from P214 to zero in Pele REVERSE o Time to reach the regenerative torque value P215 in 1 REVERSE o Time to decrease the regenerative torque value from P215 122 to zero in REVERSE i 09 vaN P221 Maximum regenerative BRAKE torque 70 0 70 0 Tnom P222 Tim
200. the supervisory safety controller In addition to CAN communication RS232 is also used but only to upload and download configuration files and data to or from the added controllers These include the REAP BMS system the MES DEA 600 induction motor controller and NLG513 chargers along with a Vector CANlog4 data logging device As explained in Section 4 4 the MotoTron outputs also command the high voltage contactors and set the precharge sequence thus it acts as an integration module and a high voltage controller so it is referred to as a vehicle integration module VIM 4 5 1 Battery Management System This system consists of 7 units responsible for monitoring 14 cells each using a master slave configuration as shown in Figure 37 To save space only 4 modules are depicted on the schematic Table 40 describes the inputs and outputs of a REAP BMS unit Chapter 4 85 1 1 Digital Outputs i PWM output for Charger LI Charger Discharger Daisy Ignition Charger Ignition Charger Inputs SS Po EISSN PWM Safety Cut off Charger Motor Controller Current Measuring per fiat a oos Jen ef Figure 37 BMS and Master Slave Configuration 12 Chapter 4 86 Table 40 BMS I Os Description 12 RS232 Configuration of each BMS s software CAN Normal mode Broadcast data to the Mo
201. tion guide of Electric Powertrain MES DEA version 7 2 ENG MES DEA Alternative Energy Division 16 Woodward ECM 5554 112 0902 C F Engine Control Modules Calibratible Flash datasheet 36757 MotoHawk Control Solutions 17 dSPACE MicroAutoBox IT Website http www dspace com en pub home products our solutions for flexray development with ds dspa ce products applic microautobox flexray cfm 18 MES DEA DC DC CONVERTER 1000W 100V 400V 12VDC Output Feb 11 2005 19 G i N mbH CANlog 4 User Manual Vector 20 Bender IR470LY User Manual www bender org References 155 21 Greg Rohrauer Hugo Provencher Gavin Clark Joseph Brennan 3 Progress Report 3 Revision H University of Ontario Institute of Technology April 2011 22 Bosch CAN Specification Version 2 0 Robert Bosch GmbH 1991 23 Kadionik Patrice Le bus CAN Ecole Nationale Sup rieure Electronique Informatique amp Radiocommunications Bordeaux 2001 24 Held Gilbert Inter and intra vehicle communications Auerbach Publications 2008 25 Provencher Hugo Introduction au protocole de communication CAN Presentation ELE4202 Commande des processus industriels D partement de g nie lectrique Ecole Polytechnique de Mtl Automne 2009 26 MotoHawk Training Controller Area Network Presentation Woodward MotoTron Control Solutions 28 October 2008
202. tml 67 Vector CANlog 3 and CANlog 4 Website http www vector com vi_canlog_en html 68 New Eagle Welcome to New Eagle Learning Center Website http www neweagle net support wiki index php title Main_Page 69 dSPACE MicroAutoBax Website http www dspace com en inc home products hw micautob cfm 70 National Instruments Controller Area Network CAN Website http www ni com can 71 Softing bus Interface Card CAN AC2 PCT Website Softing your connection to excellence http www softing com home en automotive electronics products can bus interface cards can pci 2 php 72 Intrepid Control Systems inc neoVI FIRE 6x CAN 4x LIN neoVI RED 2x CAN 2x LIN Website http intrepidcs com neovifire index html 73 Vector Introduction to CAN Website http www vector elearning com vl einfuehrungcan portal en html 74 Vector Vector Training Worldwide Website http www vector com vi training en html References 158 Appendix A Detailed Performance Simulation Results Table 58 Powertrain Specification of the Full Electric 5 Siz Motor GMT101X 110 kW peak 90 cells 240 Ah Li pol Total pack energy 80 kWh Max available energy 80 kWh 100 DOD Vehicle Curb Weight 2139 kg Table 59 Simulation Results of the Full Electric 5 Charging efficiency 90 upstream energy use included Table 60 Powertrain Specification of the EREV 60
203. to electric drive using generic controllers from multiple suppliers 3 5 1 Message and Signal Each identifier is assigned a maximum of 8 bytes data field referred to as a message regardless of the division and order of the data A message is usually broken into signals which are subdivisions of the data field and therefore of the message 26 46 47 The only rule in creating a signal is it has to be made of consecutive bits Its length varies from one bit to a few bytes and can overlap 2 bytes of the data field even if it is a 2 bit signal There are no rules concerning the length of signals however most controllers do not allow signals longer than 32 bits If a custom microcontroller is used for instance a 16 bit microcontroller signals should not exceed 16 bits to avoid extra coding for storage and arithmetic operations Table 20 shows examples of a few messages and their signals Table 20 Message and signal examples Data Field Message Mode o info Errors Passives Warsmg Unit NE Batt Voltage State of Charge ES Avg Cell Temp Max Cell Temp Min Cell Temp 3 5 2 Mapping and Positioning of Signals The bit order or bit order of significance within a byte is always increasing from right to left as specified by the CAN communication protocol 22 In other words the most significant bit msb is always at the leftmost position of a byte and the least Chapter 3 46 signif
204. to the vehicle integration module VIM Emulating electronic control units ECUs from the original powertrain and controlling additional ones for the electrical drivetrain through CAN bus along with keeping the same functionalities of a typical production vehicle makes this vehicle conversion similar to a factory built model Finally the tests and results originating from this conversion to a full electric powertrain are discussed The vehicle features a 83 5 kWh Li ion battery built in house resulting in an estimated range of 482 km Keywords CAN EcoCAR electric vehicle transportation energy lithium battery lv Dedication I dedicate this thesis to Amar El Tarazi and Loic Moquin L ger I remember meeting you and joining my first solar car team the second week of my undergrad at Polytechnique Over those 2 years of preparation and through the World Solar Challenge 2007 in Australia you both transferred me your passion for solar cars and electrical engineering and taught me the fundamental technical basis ever need in my career I have been using these skills on a daily basis while working on my second solar car project and on the EcoCAR challenge Amar first introduced as controls and high voltage lead of Esteban 4 and Loic as team lead I now consider both of you as good friends of mine Acknowledgements I owe my deepest gratitude to the main core of the UOIT EcoCAR team Joseph Brennan Gavin Clark Pierre Hinse
205. toTron Charge mode Broadcast commands to the charger and data to the MotoTron Ignition Charger Ignition input When car is in ON and ignition ON mode Input Charger input When the charger is connected When AC is detected at the charger inlet Current Measuring Measure the current entering or exiting the battery PWM safety cut off Normal mode Shut off the motor controller output Charge mode Shut off the charger output Digital Outputs N A PWM safety cut off not implemented in UOIT s design see Section 4 4 2 Not used Each BMS unit monitors the voltage and temperature of its cells to ensure that they are within set ranges Each cell is connected to a BMS module through a lead protected 125 V 1 A fuse This allows the BMS system to balance the battery pack The system achieves balancing via a flying capacitor scheme and all modules are networked via CAN A temperature sensor is fitted between every 2 cells and cell voltages are individually monitored and balanced When an individual cell voltage lies outside the specified range a signal is sent by the BMS to the MotoTron to open the main contactors and disable the motor controller When an individual cell temperature exceeds the specified range the BMS sends a signal to the MotoTron to limit the output current of the motor controller A hall effect current sensor iSensor L 1 1 400 from REAP System is place on the main HV line to monitor
206. tor of the powertrain The main positive contactor is located in the front ESS The underside ESS positive terminal is protected by a 600 V 400 A fuse The fuse and the contactor provide a safe unpowered outlet when the module is disconnected from the vehicle for maintenance and when the vehicle is turned off Containing most of the high voltage control hardware the front battery pack located in the engine bay employs a more complex wiring strategy The front battery negative terminal is connected to the underside ESS positive terminal using a 1 gauge wire A manual interrupt switch MIS having a built in fuse of 600 V 400 A is also on this high voltage line The precharge circuit and the battery main positive contactor in a parallel configuration are connected to the front battery positive terminal The other side of the main positive contactor is the main positive HV line whereas the main negative HV line is directly connected to the underside ESS contactor The following devices are connected to this HV line TIM chargers discharge circuit GFI DC DC 1 DC DC 2 A C Heater The TIM is directly connected by 1 gauge wire to the main HV lines and utilizes the two 600 V 400 A battery fuses as protection It gets connected and disconnected from the battery by the main contactors however the motor is not yet powered when high voltage is sensed at the TIM s input The traction is enabled by the VIM controlling many of the TIM inputs an
207. totype designed by UOIT Chapter 4 95 High Voltage Control Module MotoTron d N Front E Stop lt gt E 3j N Rear E Stop E N High Voltage AND Gate Connectors N 2 High Voltage Controller aX Front ESS Lid 2 GFI Level 1 Inertia Switch E sy Figure 40 Interlock configuration When one of these HV trace lines is not a closed circuit the VIM triggers the opening of the main contactors and deactivates the traction enable signal to the motor controller The main HV contactors are equipped with a feedback relay switch also monitored by the VIM to insure they are in the desired state The TIM broadcasts the traction enable status on the UOIT CAN bus Since the motor controller already has built in safety features designed to disable the motor power input it was judged unnecessary to incorporate it to the feedback verification made by the VIM However it could be implemented for extra safety 4 6 5 Ground Fault Interrupt and Inertia Switch The ground fault interrupt unit is connected to the high voltage and low voltage buses to verify that there is no leakage current in the electrical system The unit Chapter 4 96 distinguishes between low and high leakage scenarios by setting flags that send signals to the MotoTron so that the appropriate actions can be taken If a low leakage
208. tration is identical to the result on the CAN bus 36 2 Frame format Standard or Extended The main subdivision in the arbitration field is the identifier The CAN specification 2 0 offers 2 possibilities regarding the length of the identifier standard 11 bits or extended 29 bits 22 These are referred to as the frame format The substitute remote request SRR bit is a recessive bit in the remote transmission request RTR position of the extended format In the extended format the identifier extension IDE bit is recessive and part of the arbitration field while it is dominant and part of the control field in the standard format Table 64 Identifier length Frame Format Identifier Length Standard 11 bits Extended 29 bits 3 Frame type Data or Remote As explained previously a device can request another device to send back a specific data frame by sending a remote frame Such a frame contains no data and is recognized by the RTR bit located at the end of the identifier in the arbitration field 22 Table 65 RTR values Frame Type RTR Logic Level Data Dominant Remote Recessive 1 Appendix B 162 4 Format Summary This section depicts in detail the 4 possible combinations of frames discussed above Arbitration Control Data CRC End of EL Hm E ULT ME od rem _ keene Space Interframe Space IDENTIFIER 11 Bits uL
209. ttery Chemistry Lithium Polymer eee 19 2 7 1 2 Battery Management System REAP eee 22 2 7 2 On board charger 5 Brusa NLG513 25 2 7 3 Motor and motor controller Delphi 5 10 and MES DEA 600 27 2 7 4 Vehicle Integration Module MotoTron ECM 5554 112 C00 M 29 vii 2 143 Other SUuDSYSCefIIs cou oo coe ee Qus quie a ale 31 2 7 5 1 DC DC MES DEA 400 1000 21 2 7 5 2 Datalogger Vector 4 32 2 7 5 3 Ground Fault Interrupt Bender 4701_ 33 2 7 5 4 Custom instrument cluster screen 34 2 7 6 Summary of Selected Sabsysterisu setae de se 35 sae Qu ata unas Mu eu aus E 35 Chapter 3 CAN for Vehicles 36 Sc GEO 36 3 2 OVetVIeW uiuis td E EO RARE SOR OR SG ORS 36 SAM Pc 36 22 2 Properties Ob C ANCPEOROGDL a ee Rea Raa ah ie hn dede dads 37 2 2 2 1 Basic ONCE PIS aia hee 37 352 2 Properes estos needs 38 3 253 Physical Characteristics 255 ete de a Ha EVE c v a Iden 39 3 2 4 Data Es TT 41 3 2 4 1 Diff
210. ture still has the lowest GHG emissions 155 g km The estimated fuel economy was 2 51 L 100 km 2 67 L 100 km in the vehicle technical specifications to be conservative compared to 8 3 L 100 km for the production vehicle Chapter 2 14 2 6 Selected Architecture After considering the competition requirements and scoring criteria the list of 11 possible architectures was reduced Realistically fuel cells could not be supported at UOIT because of the lack of an hydrogen refuelling infrastructure Similarly requiring advanced emissions testing equipment bio diesel and E85 fuel conversions were also rejected With the fields of expertise at the university and the provincial energy mix the number of viable architectures was narrowed down to 3 In order to nominate the winner these 3 architectures were modeled and simulated Meeting all the competition requirements and having the lowest GHG emissions a full electric vehicle prototype was selected as the UOIT EcoCAR team s design The targets for this electric prototype were set by the team based on simulations and competition requirements and are summarized in Table 4 Chapter 2 15 Table 4 Targeted Vehicle Technical Specifications 7 Specification Competition UOIT Team EcoCAR Production VUE Competitimi 212 Requirement Target 8 3 L 100 km 7 4 L 100 km 2 67 L 100km 1 MEME 28 3 mpg 32 mpgge 88 mpgge 250 g km 224 g km 165 g km 7680 kg 3 5 7680 kg 3 5 T
211. uble DIN radio was replaced by a touch screen navigation system NAV The custom screen back panel and LEDs create a customized instrument cluster with relevant EV information providing immediate feedback to the driver such as speed SOC estimated range battery current and voltage motor and cell temperature along with vehicle ready and GFI lights As a maintenance feature the on board diagnostic port OBDII still works for diagnosing the remaining stock controllers Chapter 6 144 ECUs however DTCs for the new high voltage powertrain controllers were not programmed These ECUs can be diagnosed through CAN and RS232 communication connectors located in the glove compartment with the datalogger Another hidden characteristic of this prototype is the use of automotive grade and environmentally resistant electrical components and connectors 6 4 Thesis Results Where section 6 2 and 6 3 detailed the overall results of the EcoCAR project at UOIT this section focuses on the author s accomplishments After studying the CAN dictionary of the vehicle a list of ECUs to disconnect and CAN messages to emulate was made Then ECUs were individually removed from the vehicle taking care to rebuild the multiple CAN bus network after each removal Also selected CAN messages were successfully emulated by the VIM The integration of the high voltage and controls of the VIM BMS charger and motor controller went smoothly on the test bench as wel
212. udying CAN logs before and after an ECU removal and putting the missing signals back on the CAN buses with constant values to avoid DTCs However it would have made it difficult and time consuming to gain control of the instrument panel cluster requiring dynamic emulation Furthermore establishing CAN communication between generic controllers from different suppliers can require a series of trials if one does not understand the terminology defining CAN messages and signals or as it is too often the case incomplete documentation and poor terminology employed in the controller specification sheets A preliminary analysis made at the final competition on the precharge circuit resistor failure pointed towards a component selection with insufficient unspecified surge current rating however an in depth analysis might come to a different conclusion Chapter 7 148 The flawless migration of the electric powertrain from the test bench to the vehicle proved the benefits of a step by step methodology for designing testing and troubleshooting engineering systems In the engineering world a transition of this importance comes rarely without major engineering issues to solve The choice of the MES DEA TIM600 as a motor controller caused the team several problems and challenges The poor documentation resulted in poor calibrations damaging two S10 EV motors and even one of the motor controller IGBTs even though days were spent analyzing the
213. us based on the SAE J2284 3 specification thus having a baud rate of 500 kb s It is a secondary high speed network to accommodate the data flow increase caused by the components of the hybrid powertrain i e ECM APM TPIM EBCM CGM CIS DLC e SW GMLAN Single wire GMLAN also called low speed utilizes a transmission rate of 33 kb s and is compliant with the SAE J2411 standard ECUs not requiring near real time communication generally chassis modules i e IRC NAV IPC CGM VTD RFA ECC BCM AOS SDM ROS SDARS OnStar DLC are found on this network Table 32 enumerates the ECUS present on the high speed CAN buses and summarizes their main functions whereas Table 33 introduces those on the single wire one Chapter 4 70 ECUS High Table 32 List of ECUs on HS amp PTE Name Auxiliary Power Module Battery Pack Control Module Speed HS amp Powertrain Expansion PTE Manages the 12V auxiliaries Controls the lithium battery Body Control Module Main controller Chassis Inertial Sensor Inertial Y aw Sensor Communication Gateway Module Gateway between HS PTE and SW buses On board diagnostics II Data Link Connector Assembly Line Diagnostic Link Not an ECU SAE J1962 defines the connector used for the OBD II interface Give access to all CAN buses for vehicle diagnostics Electronic Brake Control Module Engine Control Module Automatic Braki
214. used to charge them i e well to pump emissions 2 4 Well to Wheel Influence and Architectures Considered Including universities from Canada and the United States the EcoCAR competition specifies a North American energy mix for electricity comprised of the assumed 13 generated in Canada and 87 produced in the U S 1 Knowing that almost all vehicles are deployed locally the UOIT EcoCAR team also thought it interesting to determine which architecture was the most eco friendly for Ontario That s why the EcoCAR energy mix was also compared to one for the North Eastern U S and for the province of Ontario Table 1 summarizes the energy use and the GHG production of 11 different possible architectures for 3 energy mixes for electricity non hybrids using either E10 or E85 HEVs and PHEVs using either E10 E85 or BD20 a compression ignition direct Chapter 2 11 injection CIDI vehicle using BD20 and a BEV 7 In this table the green to red colour scale means best to worst choice respectively Table 1 Well to Wheel energy and GHG production 7 Corrected Energy kWh km Corrected GHG g km Energy Mix EcoCAR NE US Ontario EcoCAR NE US Ontario Vehicle Energy Mix Energy Energy Energy Mix Mix Mix E10 baseline 0 69 0 69 EtOH FFV E85 1 18 1 17 Grid Independent SI HEV E10 0 47 0 47 Grid Independent SI HEV E85 0 79 0 82 Grid Connected SI HEV E10 0 46 0 45 Grid
215. ut 213 hp 16 5 165 Weight 7 5 kg Size mm 150 x 212 x 346 M 10 L min Liquid cooled min flow 50 50 EG Chapter 2 28 b Figure 14 a Delphi S10 EV with inverter chargers and DC DC b MES DEA TIM600 2 7 4 Vehicle Integration Module MotoTron ECM 5554 112 C00 M The constantly growing complexity of automotive controllers forces engineers to develop innovative programming solutions in order to efficienty design robust embedded software A model based programming approach seems to be a solution to overcome this challenge For that reason the competition offered a choice between 2 rapid prototyping Chapter 2 29 embedded controllers employing this coding technique to use as vehicle supervisory control module This controller is referred to as a vehicle integration module VIM throughout this thesis since it is far more than just a supervisory controller It also acts as a CAN message emulator a high voltage control module HVCM a communication bridge between new powertrain components and the rest of the subsystems inherited from the production VUE and is thus the central controller of the powertrain Table 10 compares a MotoTron to a MicroAutoBox from the 2 competition sponsors Woodward MotoHawk Control Solutions and dSPACE respectively Table 10 Comparison MotoTron vs MicroAutoBox 16 17 MotoTron MicroAutoBox Model ECM 5554 112 C00 M ds1401 of Pins
216. uthor and directed to a broad range of CAN users The author put significant effort to first understand the surrogate vehicles electrical software interface along with the new systems that had to be implemented to Chapter 1 14 convert it to a full function electric vehicle FFEV A methodology was then developed to provide a systematic approach to replace and emulate ECUs This methodology is outlined in chapter 4 and 5 Although team members helped with physical wiring and building of components the author generated all software interfacing with hardware The entire controls development and implementation of the UOIT EcoCAR was accomplished by the author along with a significant portion of the wiring layout tracing and debugging Chapter 6 summarizes project results that could not have been achieved without the author s significant contribution Section 6 4 highlights the author s accomplishments in more details such as the integration of the high voltage and controls of the vehicle integration module VIM battery management system BMS charger and motor controller Additionally the original vehicle CAN bus was thoroughly studied leading to the accurate static and dynamic emulations of removed ECUs which resulted in a diagnostic trouble code DTC free vehicle CAN communication was also used to integrate the full electric powertrain Chapter 7 outlines a list of original contributions such as the bench test created to te
217. vehicle FFEV An important part of such a project is to design and implement the high voltage powertrain by removing and emulating obsolete ECUS and replacing them with relevant ones in the new powertrain The center piece of these added controllers is the vehicle integration module VIM used to communicate with each controller emulate CAN messages and control the high voltage distribution hardware Finally this thesis is one of the very few detailed non confidential documents on the topic 1 4 Research Contributions The UOIT EcoCAR project was a highly complex engineering project that would not have been possible without the author The author who joined the EcoCAR after the first year of the competition acted as the controls and high voltage lead and was solely responsible for all the hardware and software interfaces A literature review of hybrid powertrains is presented in Chapter 2 Followed by a review of the architecture and component selection process made by his teammates and finally a breakdown of the components selected for the project Chapter 3 is one of the few non confidential documents on the controller area network protocol CAN fundamental theory applied for vehicles It summarizes in a single chapter a clear easy and exhaustive description of the CAN bus communication protocol for those interested in its fundamental theory along with relevant and useful hands on knowledge beyond the low level theory acquired by the a
218. vices instead of or in addition to back to back zener diodes Providing a good ground to CAN transceivers and connecting it only at only one end of the cable to avoid ground loops using a twisted pair wires to reduce high frequency noise and using shielded cables twisted pair and ground in high RF environments are efficient noise reduction solutions easy to implement Appendix E 172 Appendix Conversion of Display Format r Intel Standard Figure 66 Conversion from Motorola Forward Isb Intel Standard i Ie ae Figure 67 Conversion from Motorola Backward Appendix F 173 Post Office Analogy As mentioned before CAN messages are not destination oriented but source based However most nodes need only a small subset of CAN messages travelling on the bus Modules are required to filter the information transmitted on the bus to select the messages of interest A relevant and simple analogy to understand the sorting of CAN messages is the post office analogy adapted from reference 26 Filtering also called sorting can be done by hardware and or software Microcontrollers featuring integrated CAN controllers typically have a hardware layer involving programmable buffers acting as a first filter A custom dispatcher can be programmed in the target module s CAN chip software to refine the first sorting The post office analogy terminology
219. wer On DX cANH Nur EM Control Reset 25 0 5 VDD i55 1 lt GND ko x Receiver VREF X M Reference Voltage 54 Vss GND Figure 65 CAN transceiver MCP2551 block diagram 41 The CANH and CANL voltages are determined by the driver control reacting to the digital transmit signal and by a voltage source of 2 5V This voltage source is responsible for setting the CAN bus at 2 5V when a node is powered on This value corresponds to the recessive state voltage The 2 5V source is connected to each line through a high Appendix E 170 impedance resistor 25 kQ in the case of the TJA1050 43 The value of these resistors contributes to the busload and therefore the maximum number of nodes a CAN environment can support For example a high speed CAN bus on which each node uses a MCP2551 CAN transceiver can support up to 112 nodes 41 This source can be turned off by the power on reset block to disconnect the CAN transceiver from the CAN bus To obtain an analog dominant bit level the driver control achieves the bus voltage by driving two transistors One is connected between the voltage supply Vpp 5V and the CANH wire and acts as a pull up while the other one connects the CANL line to the ground Vss and acts as a pull down Even though the block diagram shows two outputs on the driver control both transistors are conducting simult
220. wn on Figure 28 is considered to have a hybrid star bus topology with a bus as the network backbone This topology is sometime referred to as a cascaded star topology A disadvantage of this technique is the addition of more branches and therefore more noise in the network CANH Motor High Voltage Controller Control Module Battery Supervisory On Board Management Control Charger System Module Module Figure 27 Bus topology minimizing noise CANH Battery On Board Management Charger System Supervisory High Voltage Module Control Control Module Module Figure 28 Hybrid star bus topology Chapter 3 59 To reduce the electrical noise to minimum another variation of the bus topology known as a daisy chain can be used This method brings the main bus directly to every device on the CAN bus removing the branches and therefore the antenna effect induced by them In this configuration the devices have an internal parallel connection design and each device acts as a hub for the following device in the chain The expression daisy chain topology can be viewed as a bus topology with short studs However the expression is appropriate regarding the method used to create the wiring harness for the configuration and has an influence on the hardware of the devices connected to the bus The absence of hubs in this configuration makes it more difficult to add and remove
221. ww flexray com visited on 28th March 2012 39 www freescale com files microcontrollers doc fact_sheet FLEXCOMMS YSTMAFS pdf Website 40 http en wikipedia org wiki Network topology visited on 10th December 2011 41 Microchip MCP2551 High Speed CAN Transceiver datasheet Microchip Technology Inc 2007 DS21667E Website http www microchip com 42 Microchip dsPIC33FJXXXGPX06 X08 X10 Data Sheet High Performance 16 Bit Digital Signal Controllers datasheet Microchip Technology Inc 2007 DS70286A http www microchip com Website 43 Philips Semiconductors NXP Semiconductors 1050 High speed CAN transceiver datasheet Product specification 22nd October 2003 www nxp com documents data sheet TJA 1050 pdf 44 Tech Note CAN Data Field Format Intel Motorola Forward and Motorola Backward Website www warwickcontrol com visited on 13th December 2011 References 156 45 T McLaughlin Richard Data Formats and X Analyser Functions Presentation 2011 Warwick Control Technologies retrieved December 13th from www testing expo com europe O5txeu conf pres mclaughlin pdf 46 Vector Academy Introduction to CANoe DENoe video v7 0 2009 03 10 CANoe v7 0 Vector CANtech Inc 47 Vector CANdb Online Help Mapping and Position Numbering of Signals in Frames Software Vector CANdb Editor Version 3 0 62 SP6 48 Matlab Product Help
222. y emulate GMLAN signals and also for fault detection Under faulty conditions it updates the emergency hardwired input of the motor controller pin 19 of Figure 39 EMERGENCY Therefore the vehicle enters an emergency mode even though the gas and brake pedal potentiometer connected to the VIM are not used directly to control the motor torque This conservative strategy was implemented to in an attempt to make the prototype being designed as safe as possible The acceleration and brake pedal decoder is modeled in Appendix J Figure 91 and the fault management model is shown in Figure 93 5 4 4 Radiator Fans In the original vehicle the 2 fans on the front radiators were controlled by the ECM The latter being removed from the vehicle these controls were rerouted to the VIM which drives the stock fan controller by using its PWM input to adjust the fan speed according to the shifter lever position and the motor controller cooling needs The intended design was to enable the fans and vary their speed based on temperature sensors At present however when the vehicle is in drive or reverse the fans are simply turned on to their maximum speed Therefore the control logic could be improved by implementing a more complex thermal management algorithm based on temperature feedback which would optimize energy savings The basic solution implemented is represented in Figure 92 Appendix J 5 4 5 Speedometer The instrument cluster was modified to re
223. y plates and the aluminum boxes they were contained in The noise on the HV is thus transmitted to the ground thus signal lines pick up ground reference problems This noise problem was experienced with the temporary Li ion battery using an aluminum box and located in the trunk of the vehicle With the new battery box such coupling was to be minimized due to the non metallic composite material employed in construction and the shorter high voltage leads 6 3 2 4 State of Discharge Since the charger input was instrumented it was possible to verify at least in part the cyclic efficiency of this process The state of charge is not an easy quantity to verify accurately because it depends on temperature and remaining surface charge on the cells from the last direction of current flow However from the battery depletion test the Chapter 6 140 team now has a fair idea of the applicable discharge curve and can estimate SOC to better than 5 even in the midranges The BMS offers accuracy of individual cell voltages down to 1 mV and temperatures to 0 1 C Figure 51 is the experimentally derived voltage SOC curve of the temporary battery pack based on the EPA results Pack SOD vs Voltage 360 0 340 0 320 0 300 0 _ 280 0 260 0 pack 84 actual Pack Voltage 240 0 220 0 200 0 0 20 40 60 80 100 State of Discharge of nominal rating Figure 51 Experimentally derived battery pack en
224. y troubleshooting the HV distribution box and HVCM Verify the polarity of the battery main contactors in particular if any e Further testing on the motor controller calibrations determined at EPA and tuning if necessary now with a higher voltage battery pack e Addition of proactive on board diagnostics for the EV powertrain controllers sensors and actuators to reduce the need for reactive troubleshooting when an issue or failure arises and tend towards a certified vehicle that is diagnostics compliant Fix battery cell leaks incorporate updated cell module packaging Chapter 7 151 Verify terminal block signal routing Fix the 12V battery draining problem that exist on the stock hybrid Saturn VUE but ignored by the team Improve the power reduction request controlling the inverter when a ground fault occurs 7 4 2 Best Practices The list following enumerates the hands on personal best practices gained by the author over the years while working on developing vehicle prototypes such as the EcoCAR project and 2 prior solar car projects They are recorded here for the benefit of readers contemplating similar endeavours Hardware is as important as software A good wiring harness drives good results in other words a bad harness equals bad results Learn how to make a proper wiring harness soldering crimping wire lengths loom etc A need for software testing time in the schedule once the hardware i
225. zes some of the author s specific contributions successfully implemented in this research and development project e Non confidential document on electric vehicle conversion e n vehicle implementation of the vehicle integration module VIM and its model based algorithm including gt Dynamic and static CAN message emulation of removed ECUs Acting as a CAN message gateway such as sending messages to be displayed on the in house IPC screen or recorded by the datalogger High voltage distribution box control through the high voltage control module Decoding and management of stock sensors such as acceleration and brake pedals ignition key and transmission shift lever Fault management of acceleration and brake pedal inputs gt Actuation and control of the radiator fans e Bench testing of the added powertrain electrical subsystems i e VIM BMS Li ion battery TIM motor charger In vehicle electrical integration of the EV powertrain subsystems e High voltage design of the powertrain architecture and Li ion battery packs e Modifications to the vehicle CAN buses GMLAN HS and PTE including gt Fixing of the broken daisy chain when removing stock ECUs gt Migration from daisy chains to hybrid configurations i e combination of star configuration and daisy chain gt Addition of several ECUs to the CAN networks e Gaining control of the dashboard by properly emulating CAN messages for example gt Ready and GFI lig

Download Pdf Manuals

image

Related Search

Related Contents

Estimating water requirements and water storage requirements for  MODE D`EMPLOI DE PHOTOVOLTACALC.XLS  Sending lasers around the world  Samsung ProXpress M4075FR Mono    MANUEL D`INSTRUCTIONS  vantag  fichier 1 - CRDP de Montpellier  AMT Lightning  

Copyright © All rights reserved.
Failed to retrieve file