Compact AC/DC-module for Electric Vehicle Charging Dissemble, Evaluation and Design Development of a Portable Battery Charger Master’s Thesis in Electric Power Engineering MARTIN ALERMAN THERESE STENBERG Department of Electrical Engineering CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2018 Master’s thesis 2018 Compact AC/DC-module for Electric Vehicle Charging Dissemble, Evaluation and Design Development of a Portable Battery Charger MARTIN ALERMAN THERESE STENBERG Department of Electrical Engineering Division of Electric Power Engineering Chalmers University of Technology Gothenburg, Sweden 2018 Compact AC/DC-module for Electric Vehicle Charging Dissemble, Evaluation and Design Development of a Portable Battery Charger MARTIN ALERMAN THERESE STENBERG © MARTIN ALERMAN THERESE STENBERG, 2018. Supervisor: Erik Ahlqvist, at CEVT Examiner: Prof. Yujing Liu, at the department of Electrical Engineering Master’s Thesis 2018 Department of Electrical Engineering Division of Electric Power Engineering Chalmers University of Technology SE-412 96 Gothenburg Telephone +46 31 772 1000 Cover: The proposed off-board charger for electric vehicles, created in Tinkercad and Visio. Typeset in LATEX Printed by Chalmers Reproservice Gothenburg, Sweden 2018 iii Compact AC/DC for Electric Vehicle Charging MARTIN ALERMAN THERESE STENBERG Department of Electrical Engineering Division of Electric Power Engineering Chalmers University of Technology Abstract At present, the automotive industry has equipped their electric vehicles with the on-board charger, where it converts the power in order to charge the high voltage battery. However, there is a demand to remove the on-board charger from the vehicle since it requires large space and weight. Therefore it is desirable to investigate if it is feasible to relocate and redesign the on-board charger to an external portable charger. Since this charger is suppose to be portable, it is essential that the design is lightweight and compact. This research completed an analysis of a dissembled on-board charger, topolo- gies review of converters, components selection, loss calculations, electric circuits simulation and thermal simulation studies. In this study it was demonstrated that it was feasible to design an off-board charger for electric vehicles. The charger was design for a current level of 4 A, which has a size of 1024 cm3, a weight of 2.8 kg and an efficiency of 94.42 %. Comparing only the module itself with the OBC, there was a weight reduction of 71 %. Forced cooling has been implemented, to prevent overheating of the power electronic components. Keywords: On-board charger, Off-board charger, AC/DC converter, DC/DC con- verter, PHEV, Boost, Totem-pole, Full-bridge, Half-bridge and Cooling system. iv Acknowledgements First of all, thanks to you who found and will read this master thesis. A special thanks to our examiner Prof. Yujing Liu at Chalmers University of Technology for all his help and suggestions. We would also like to thank Erik Ahlqvist, including the entire power conversion team at CEVT for their guidance and support during this project. Thanks to the people at the Electric Power Engineering division at Chalmers for their technical support. With special thanks to Daniel Pehrman for his contribution to the dissembling process and consultation. Also thanks to Alessandro Acquaviva for his guidance through the initiation phase of this project. Last but not least, we are thankful for the endless support from our families and friends during this time. Martin Alerman and Therese Stenberg, Gothenburg, June, 2018 vi Abbreviations CAN Controller area network CCS Combined charging system CEVT China Euro Vehicle Technology CCM Continuous conduction mode CM Common mode CP Control pilot CrM Critical conduction mode DSP Digital signal processor DCM Discontinuous conduction mode EV Electric vehicle EVI Electric vehicle inlet EMI Electromagnetic interference ESR Equivalent series resistance GaN Gallium nitrate G2V Grid-to-vehicle HV High voltage IEC International electrotechnical commission IGBT Insulated gate bipolar transistor IP International protection marking LV Low voltage MOSFET Metal oxide semiconductor field effect transistor MCU Microcontroller unit OBC On-board charger PCB Printed circuit board PHEV Plug-in hybrid electric vehicle PFC Power factor correction PP Proximity pilot PWM Pulse width modulation SiC Silicon carbide TIM Thermal interface material THD Total harmonic distortion ZVS Zero voltage switching V2G Vehicle-to-grid VDDM Vehicle dynamics domain master viii Contents List of Figures xii List of Tables xv 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.4 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.5 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2 Method 6 3 Safety and Sustainability 8 3.1 Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.2 Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 4 Theory 10 4.1 Plug-in hybrid electric vehicle . . . . . . . . . . . . . . . . . . . . . . 10 4.2 On-board charger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 4.2.1 AC/DC converter . . . . . . . . . . . . . . . . . . . . . . . . . 11 4.2.2 PFC controller . . . . . . . . . . . . . . . . . . . . . . . . . . 11 4.2.3 DC-link . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 4.2.4 DC/DC converter . . . . . . . . . . . . . . . . . . . . . . . . . 12 4.3 Performance and size . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 4.3.1 Standard regulations . . . . . . . . . . . . . . . . . . . . . . . 13 4.3.2 Switching technology . . . . . . . . . . . . . . . . . . . . . . . 13 4.3.3 EMI filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 4.3.4 Switching frequency . . . . . . . . . . . . . . . . . . . . . . . 16 4.4 Power factor correction . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.4.1 LT1248 - PFC controller . . . . . . . . . . . . . . . . . . . . . 19 4.5 Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 4.5.1 AC/DC converter with PFC . . . . . . . . . . . . . . . . . . . 20 4.5.2 Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.5.3 DC/DC converter . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.5.4 Number of components . . . . . . . . . . . . . . . . . . . . . . 26 4.6 Cooling implementation . . . . . . . . . . . . . . . . . . . . . . . . . 27 ix Contents 4.7 OBC from Geely Group . . . . . . . . . . . . . . . . . . . . . . . . . 28 4.7.1 Design specification . . . . . . . . . . . . . . . . . . . . . . . . 28 4.7.2 3 separated connectors . . . . . . . . . . . . . . . . . . . . . . 29 4.7.2.1 AC connector . . . . . . . . . . . . . . . . . . . . . . 29 4.7.2.2 HVDC connector . . . . . . . . . . . . . . . . . . . . 30 4.7.2.3 LV system connector . . . . . . . . . . . . . . . . . . 30 4.7.3 Circuit diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.7.4 Data communication . . . . . . . . . . . . . . . . . . . . . . . 31 4.7.5 Operation of OBC . . . . . . . . . . . . . . . . . . . . . . . . 31 4.7.6 Heat losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 4.7.6.1 Transformer losses . . . . . . . . . . . . . . . . . . . 33 4.7.7 Weight and volume . . . . . . . . . . . . . . . . . . . . . . . . 34 4.7.8 Thermal management . . . . . . . . . . . . . . . . . . . . . . 34 4.8 Charging cable from Mennekes . . . . . . . . . . . . . . . . . . . . . . 35 4.9 Thermal calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.9.1 Power losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.9.2 Thermal calculation and design . . . . . . . . . . . . . . . . . 42 5 Results 44 5.1 OBC - disassembling . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 5.1.1 Layout of components . . . . . . . . . . . . . . . . . . . . . . 45 5.1.2 Component characteristics . . . . . . . . . . . . . . . . . . . . 47 5.2 Charging cable - disassembling . . . . . . . . . . . . . . . . . . . . . . 49 5.3 Off-board charger - design characteristics . . . . . . . . . . . . . . . . 50 5.3.1 Design requirements . . . . . . . . . . . . . . . . . . . . . . . 50 5.3.2 Components selection . . . . . . . . . . . . . . . . . . . . . . . 51 5.3.2.1 AC/DC converter with Boost PFC . . . . . . . . . . 51 5.3.2.2 AC/DC converter with Totem-pole PFC . . . . . . . 53 5.3.2.3 Full and Half-bridge DC/DC converter . . . . . . . . 54 5.3.2.4 EMI filters . . . . . . . . . . . . . . . . . . . . . . . 54 5.3.3 Weight and size . . . . . . . . . . . . . . . . . . . . . . . . . . 55 5.3.4 Summary of weight and size . . . . . . . . . . . . . . . . . . . 59 5.4 LTspice simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 5.4.1 AC/DC converter with Boost PFC . . . . . . . . . . . . . . . 60 5.4.2 AC/DC converter with Totem-pole PFC . . . . . . . . . . . . 63 5.4.3 Full-bridge and Half-bridge DC/DC converter . . . . . . . . . 64 5.4.4 Summary of LTspice results . . . . . . . . . . . . . . . . . . . 67 5.5 Power loss calculations . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.5.1 AC/DC Boost . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.5.2 AC/DC Totem-pole . . . . . . . . . . . . . . . . . . . . . . . 73 5.5.3 DC/DC - Half and Full . . . . . . . . . . . . . . . . . . . . . 74 5.5.4 Micro controller unit . . . . . . . . . . . . . . . . . . . . . . . 81 5.5.5 Summary of loss calculations . . . . . . . . . . . . . . . . . . . 81 5.5.6 Topology selection . . . . . . . . . . . . . . . . . . . . . . . . 82 5.6 Cooling system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 5.6.1 Heat calculations . . . . . . . . . . . . . . . . . . . . . . . . . 84 x Contents 5.6.2 Summary from heat calculations . . . . . . . . . . . . . . . . . 86 5.6.3 Heatsink selection . . . . . . . . . . . . . . . . . . . . . . . . . 86 5.6.4 Design suggestion . . . . . . . . . . . . . . . . . . . . . . . . . 87 5.6.5 Heatsink simulation . . . . . . . . . . . . . . . . . . . . . . . . 87 5.7 Final design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 5.8 Nozzle design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 6 Discussion 94 6.1 Electrical performance . . . . . . . . . . . . . . . . . . . . . . . . . . 94 6.2 Heat and thermal losses . . . . . . . . . . . . . . . . . . . . . . . . . 95 6.3 Weight and size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 7 Conclusion 98 7.1 Future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Bibliography 100 A Appendix - Circuit diagrams I B Appendix - Components, operation mode, topologies III C Appendix - LTspice simulation VIII D Appendix - Power losses IX E Appendix - Heatsink simulation XIII xi List of Figures 2.1 Transformer design flowchart. . . . . . . . . . . . . . . . . . . . . . . 7 4.1 The energy conversion system inside a PHEV [20]. . . . . . . . . . . . 10 4.2 Electric vehicle battery charger architecture [23]. . . . . . . . . . . . 11 4.3 Receiving and distribution signals from the PFC controller to an AC/DC converter [27]. . . . . . . . . . . . . . . . . . . . . . . . . . . 12 4.4 Comparison of semiconductors regarding frequency, voltage and cur- rent [39]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 4.5 EMI filter for single stage power supply [39]. . . . . . . . . . . . . . . 16 4.6 Output power from the DC/DC converter in relation to the efficiency at various switching frequency levels [54]. . . . . . . . . . . . . . . . . 17 4.7 Illustrating the voltage- and current characteristics waveforms by not having a PFC to the left, and by having a PFC [59]. . . . . . . . . . 18 4.8 Schematic of LT1248 PFC controller in LTspice. . . . . . . . . . . . . 19 4.9 AC/DC converter with Boost PFC configuration [66]. . . . . . . . . . 21 4.10 AC/DC converter with Totem-pole PFC configuration [66]. . . . . . . 22 4.11 Isolated Full-bridge DC/DC converter [39]. . . . . . . . . . . . . . . . 24 4.12 Isolated Half-bridge DC/DC converter [39]. . . . . . . . . . . . . . . . 24 4.13 Sealed module enclosed with thermoelectrically enhanced heat rejec- tion [80]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 4.14 OBC Geely 8888003014 G BQ8AA from a PHEV. . . . . . . . . . . . 28 4.15 AC connector with pin-configuration [37]. . . . . . . . . . . . . . . . 30 4.16 AC connector with pin-configuration [37]. . . . . . . . . . . . . . . . 30 4.17 Losses in the winding in transformer at 100 kHz [88]. . . . . . . . . . 33 4.18 Back side of OBC, where the cooling system is equipped with cooling pipes mounted to remove heat dissipation from the components. . . . 34 4.19 Charging cable with control and supervision module. . . . . . . . . . 36 4.20 Thermal resistance of a heatsink model [39]. . . . . . . . . . . . . . . 43 5.1 To the left: under the plastic safety shield the circuit board over the low voltage system is seen inside the OBC. To the right: the base of the OBC casing, where the cooling system is located, are mounted with TIM to dissipate thermal heat generated from the components. . 44 5.2 Back side of circuit board, where the high voltage system and its power electronic components are implemented. . . . . . . . . . . . . . 45 xii List of Figures 5.3 Layout overview of the OBC, where an AC/DC converter with PFC circuit, and a DC/DC converter are included. The flow of the incoming AC and supplying DC are illustrated with blue arrows. . . . . . . . . 45 5.4 Identification of the segments that are included in the high voltage system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 5.5 To the left: under the plastic safety shield the circuit board over the high voltage system is observed inside the module of the charging cable. To the right: the base of the module’s plastic casing is shown. 49 5.6 Dimension of the selected transformer and its core. . . . . . . . . . . 54 5.7 Simulation schematic in LTspice of the AC/DC converter with Boost PFC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 5.8 The waveforms of input voltage-and current from the AC grid with a PFC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 5.9 The waveforms of output voltage-and current from Boost PFC with 230 V (RMS) and 4 A supplied from grid. . . . . . . . . . . . . . . . 62 5.10 Simulation schematic in LTspice of the AC/DC converter with Totem- pole PFC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 5.11 The waveforms of AC voltage input and the pulse sequence of Q1 and Q2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 5.12 The waveforms of output voltage-and current from Totem PFC with 230 V (RMS) and 4 A supplied from grid. . . . . . . . . . . . . . . . 64 5.13 Simulation schematic in LTspice of the Full-bridge DC/DC converter. 65 5.14 The waveforms of output voltage-and current from the DC/DC con- version stage when 4 A supplied from the AC grid. . . . . . . . . . . 66 5.15 Simulation schematic in LTspice of the Half-bridge DC/DC converter. 66 5.16 Peaks currents levels with higher frequency. . . . . . . . . . . . . . . 67 5.17 THD level at AC input current for the Boost-and Totem PFC. . . . . 68 5.18 PF level at AC input current for the Boost-and Totem-pole PFC. . . 68 5.19 Comparison of theoretical efficiency of the Boost PFC and Totem-pole PFC at different output power [155]. . . . . . . . . . . . . . . . . . . 73 5.20 Proposed heatsink design with mounted components. . . . . . . . . . 87 5.21 Dimension of selected heatsink with mounted components. . . . . . . 87 5.22 Heatsink simualtion without fan. . . . . . . . . . . . . . . . . . . . . 88 5.23 Heatsink simulation with fan. . . . . . . . . . . . . . . . . . . . . . . 90 5.24 The proposed module of off-board charger. . . . . . . . . . . . . . . . 91 5.25 Comparison between OBC and the off-board charger module. . . . . . 92 5.26 Feasible nozzle design connector with the pin configuration for the off-board charger. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 5.27 Data communication in the off-board charger integrated between a wall socket and a PHEV. . . . . . . . . . . . . . . . . . . . . . . . . . 93 6.1 Comparison between size design. From left to right: OBC, off-board charger and module from Mennekes. . . . . . . . . . . . . . . . . . . 96 A.1 Circuit diagram of the OBC Geely 8888003014 G BQ8AA from Geely Group. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I A.2 Circuit diagram of the proposed off-board charger. . . . . . . . . . . . II xiii List of Figures E.1 Simulation of natural cooling on a larger heatsink size. . . . . . . . . XIII xiv List of Tables 4.1 Standards regulations for the performance of the electrical parameters. 13 4.2 Comparison of MOSFET and IGBT [40]. . . . . . . . . . . . . . . . . 15 4.3 Comparison of Classic Boost PFC and Totem-pole Bridgeless PFC. . 20 4.4 Comparison in operation modes [69]. . . . . . . . . . . . . . . . . . . 23 4.5 Advantages and disadvantages of the Full-bridge converter [76],[77], [78]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.6 Advantages and disadvantages of the Half-bridge converter [76], [77], [78]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 4.7 The number of power electronic components in the selected circuits. . 27 4.8 Design specification of OBC - Geely 8888003014 G BQ8AA [37]. . . . 29 4.9 Type of connectors with configuration arrangement of the contacts [86]. 31 4.10 Efficiency and power dissipation of the OBC with various currents. These data was measured with a cooling temperature of 40 ◦C [87]. . 33 4.11 Design specification of Mennekes charging cable [92]. . . . . . . . . . 36 5.1 Volume and weight of the power electronic components. . . . . . . . 48 5.2 Total weight and volume for the power electronic components. . . . . 48 5.3 Other weight of the OBC. . . . . . . . . . . . . . . . . . . . . . . . . 49 5.4 Other weight of the charging cable. . . . . . . . . . . . . . . . . . . . 50 5.5 Target design specification of off-board charger. . . . . . . . . . . . . 50 5.6 With various current ratings supplied to the AC/DC converter Boost PFC, gives different component value, volume and weight. . . . . . . 56 5.7 Output capacitor for the AC/DC converter with Boost PFC. . . . . . 56 5.8 With various current ratings supplied to the AC/DC converter Totem- pole PFC, gives different component value, volume and weight . . . . 57 5.9 Output capacitor for the AC/DC converter with Totem-pole PFC. . . 57 5.10 Various current ratings supplied to the EMF filter, gives different component value, volume and weight for the inductor. . . . . . . . . 57 5.11 Various current ratings supplied to the EMF filter, gives different component value, volume and weight for the filter capacitor. . . . . . 58 5.12 Weight and volume of the components that are included in the full- bridge DC/DC converter. . . . . . . . . . . . . . . . . . . . . . . . . 58 5.13 Weight and volume of the components that are included in the Half- bridge DC/DC converter. . . . . . . . . . . . . . . . . . . . . . . . . 59 5.14 Total weight-and volume of the different topologies. . . . . . . . . . . 59 5.15 Input and output parameters of the AC/DC converter for 4 A. . . . . 69 xv List of Tables 5.16 Power losses for Boost at different current levels. . . . . . . . . . . . 72 5.17 Totem-pole estimation power losses. . . . . . . . . . . . . . . . . . . . 73 5.18 Power losses for Totem-pole at different current levels. . . . . . . . . 74 5.19 Input DC/DC values taken from Totem-pole calculation. . . . . . . . 74 5.20 Calculated transformer parameters for different configurations and current levels for Totem-pole. . . . . . . . . . . . . . . . . . . . . . . 81 5.21 Efficiency - Boost. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 5.22 Efficiency - Totem-pole. . . . . . . . . . . . . . . . . . . . . . . . . . 82 5.23 Loss comparison of the four configurations. . . . . . . . . . . . . . . . 82 5.24 Efficiency - DC/DC - Totem. . . . . . . . . . . . . . . . . . . . . . . 83 5.25 Total power loss - Totem-pole + DC/DC alternative. . . . . . . . . . 83 5.26 Thermal resistance - Half-bridge. . . . . . . . . . . . . . . . . . . . . 86 5.27 The temperature of the components with natural cooling. . . . . . . . 89 5.28 Temperature of the components with forced cooling. . . . . . . . . . . 90 B.1 OBC - list of the power electronic components. . . . . . . . . . . . . . III B.2 Charging cable - list of the power electronic components. . . . . . . . IV B.3 Comparison between switching devices MOSFET and IGBT [41], [42], [43]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V B.4 Comparison of different operation modes such as CCM, DCM and CrM [71],[72],[73]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI B.5 Comparison of different PFC topologies [62],[63][64],[65]. . . . . . . . VII B.6 Total capacitance at the DC-link for both Boost-and Totem-pole PFC.VII C.1 Output result from LTspice simulation with Boost PFC. . . . . . . . VIII C.2 Output result from LTspice simulation with Totem-pole PFC. . . . . VIII D.1 Component selection for the different devices in Totem-pole, 4 A. . . IX D.2 Comparison between the two core of transformers. . . . . . . . . . . . X D.3 Power loss and efficiency comparison, B=0.15 T. . . . . . . . . . . . X D.4 Power loss and efficiency comparison, B=0.1 T. . . . . . . . . . . . . XI D.5 Totem + DC/DC half-bridge power losses. . . . . . . . . . . . . . . . XI D.6 Totem + DC/DC Full-bridge power losses. . . . . . . . . . . . . . . . XII D.7 Device characteristics, 4 A, Totem + DC/DC. . . . . . . . . . . . . . XII E.1 Temperature of the components for natural cooling with a larger heatsink size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XIII xvi 1 Introduction 1.1 Background With the automotive industry moving towards electrification, the number of electric vehicles (EVs) has increased in the market in recent years [1]. Therefore, it is vital to have a sufficient and accessible charging infrastructure in the society for EVs, in order to recharge the batteries in the vehicle and mitigate the driver’s range anxiety. Simultaneously, the batteries have become larger to extend distance range further, which has led to an increase in weight of the vehicle [2]. A feasible solution for this could be to relocate the on-board charger (OBC) from the vehicle to a charging cable to reduce weight in the vehicle, and to provide customers with an applicable charging alternative. This cable is for household applications, where utility power is being used, which allows the driver to bring the charger with them to charge at other locations equipped with standard wall sockets. The Propulsion unit at CEVT (China Euro Vehicle Technology) is responsible for the development of future energy transmissions for Geely Group as well as the development of electric propulsion systems including all high voltage components. At this unit, the power conversion department is interested in investigating and finding a portable off-board charger solution for plug-in hybrid electric vehicles (PHEVs) directed toward households. The assignment is to redesign the existing on-board charger and relocate it in the charging cable as an off-board charger. Simultaneously, the AC/DC-module should be designed to be more compact by reducing size and number of the power electronic components and also adding a minimized cooling system. The AC/DC- module in the OBC includes an AC/DC-converter and DC/DC-converter. The aim of this project is to make design improvements regarding efficiency, losses, size, and weight. One issue with having an OBC-system inside the vehicle is that it requires a lot of space, which can be used for other things, i.e. control system, larger battery or cooling system. By having more space for a larger battery, it would lead to less stress on the battery’s system and longer travel distance range. 1 1. Introduction 1.2 Aim The aim of this project is to develop a compact mechanical design of an AC/DC- converter module, including thermal and electrical sizing, for an off-board charging cable for PHEVs. The device choice shall be based on various factors such as efficiency, volume, and losses. 1.3 Objectives In order to design a compact AC/DC-module charger, the following main objectives will be investigated: • A literature study of different topologies will be examined to find one that applies for this project. • The electrical performance of the converter shall be taken into consideration. The power electronic components are selected to achieve high power factor (PF), high efficiency, and low THD (total harmonic distortion). These per- formance parameters are obtained by the simulation software LTspice and calculations procedure. • Within the thermal aspect, the selected components will be hand calculated regarding the losses in order to evaluate if it is feasible to design an off-board charger for low-power charging application. Furthermore, a cooling system is suggested to reduce the thermal losses and will be simulated in an application software for thermal analysis. • The components and cooling system selection will decide the final size of the off-board charger, regarding weight and volume that will be estimated. • An investigation regarding the safety and sustainability aspects of charger module for PHEVs will be conducted. 1.4 Problem The purpose of investigating an off-board charger for PHEVs, with design modifica- tions regarding efficiency and size, is to create more compartment capacity within the PHEV that can be utilized for larger batteries in the first place. In order to design an effective and compact converter for the off-board charger, diverse topologies of power electronic circuits will be analyzed. These topologies are created to find an appropriate circuit that can be used for external portable charging. By studying these topologies, a comparison in requirements can be made in regard to the type of converter, efficiency, level of power application and low number of power electronic components (capacitors and inductors). These requirements also 2 1. Introduction include how the cooling implementation for the converter will be executed, where it is expected to be large in size, since liquid cooling from the vehicle is not used. Thus, it is vital to select material and components that are sufficient to meet the demands within the project. Since the focus is mainly on weight and size, existing integrated circuit (IC) PFC will be utilized. In relation to the size of the converter and the make the module portable, the size and number of components needs to be reduced and the cooling system mini- mized. Having a portable module allows the driver to bring the off-board charger and charge the PHEV at locations where wall-sockets exist. Simultaneously, there must be a balance between delivering sufficient charging power to the vehicle and having the charger light weighted. Low weighted is desired since the potential customer should be able to lift the charger without much effort. The components temperature depends on the thermal design and the loading conditions. These components will be required to handle temperatures preferably below 150 ◦C, when the conversion process occurs inside the module. Regarding the losses, it is favorable to find a converter that can perform sinu- soidal current consumption, where a unitary power factor is achieved [3]. The issue in using an AC/DC converter is that it contains non-linear loads, in forms of rectifiers. These loads will provide undesirable frequency to the current and thereby increase THD. This leads to that a lower power quality will be delivered to the load [5]. To avoid this, linear loads (capacitors and inductors) are needed to compensate this, where a higher power factor can be obtained to achieve lower THD [3]. To reduce the size of the components, the switching frequency can be increased [4]. However, a higher switching frequency leads to higher losses caused by the switching elements turning on and off in a high-speed releasing transition energy per second [6]. If a conventional DC/DC converter will be used, it is favorable to implementing a galvanic insulation in the transformer of the converter, to provide safety measures [38]. Since AC/DC-conversion creates high frequency electronic noise, there is a need for control or reduction in the electromagnetic interference (EMI), using filters. 1.5 Scope Since this project is about to design a portable battery charger exclusively for PHEVs, other types of EVs will not be acknowledged. For PHEVs, the most commonly design specification for OBC that is existing today, is designed for 230 V, 50 Hz, 16 A rates, giving an output power of 3.4-3.6 kW, 92-94 % in efficiency and a PF of > 0.99. Additionally, a standard OBC weights approximately 4 kg with a switching frequency of 98.3 kHz both for the power factor correction (PFC) and DC/DC converter. The charging optimization is considered not as important as to ensure that the AC/DC-converter module operates in the charging range, which is the main focus of this project. After consultation with CEVT, reference values were settled to proceed from. Since 3 1. Introduction the off-board charger will be operating within low power range, it is more reasonable to follow international standards. The efficiency is expected to be at least 90 % with a minimum PF of 0.9, according to international standards [7]. The power density is assumed to be approximately 1 kW/kg. Moreover, the overall weight of the module should be minimized from 4 kg, with aiming to reach a higher switching frequency than 100 kHz for both the AC/DC- and DC/DC converter since it is desirable to reduce the components size. Since that the charger is for household applications, single-phase system is used. Other systems, such as 2-or 3-phase system will not be investigated. These house- holds are assumed to have European standard shucko-wall sockets, limited with 16 A and 3.68 kW [8]. However, since EVs usually draw a maximum current of 10 A when connected to a wall socket, the input current level is assumed to be maximum 10 A [9]. The calculations will be based from a current level of approximately 4 A since low-power charging is desired. The charger should also have a unidirectional flow configuration, which implies that the power from the grid is transferred one way to the battery, also known as grid-to-vehicle (G2V) [10]. The charging time is a non-issue, considering that the charger will be operating with normal charging, where it takes approximately 6-8 h to fully charge an uncharged battery [11],[12]. The high voltage (HV) battery inside the PHEV have a voltage charging range between 320-380 V. The duty cycle for the pulse sequence is assumed to be fixed. Since the project aims towards designing an off-board charger, a thermal man- agement system for the power electronic components inside the charger is essential to study. This is since it cannot have used the cooling system inside the vehicle. The cooling system will most probably be a heatsink, installed and mounted on the AC/DC-converter module. The module itself will assumed to have international protection marking IP67 standard [93]. In this project, robustness and impact re- sistance are excluded and not prioritized, due to that it is more related to material design. A construction of a prototype model and measurements are not included in this project. The calculations will give an estimated power loss and heat values, since com- puting a proposed solution involves a lot of work with many choices to be assumed and motivated. Regarding the transformer in the DC/DC converter, some simplifica- tions will be conducted. A first size of a transformer core will be selected to proceed calculations from. A heatsink suggestion will be presented to illustrate an example of what can be used. No comparison between different heatsinks will be made, due to the thermal simulation software offer is limited. Additionally, other cooling implementation will not be taken into consideration. Since LTspice offer a limited range of components, it will be difficult to find same 4 1. Introduction components from datasheets. Therefore, the selection of components will give an approximately result. Additionally, the datasheets that are lacking information will be complemented with values from similar datasheets. The selection of specific manufactures of the components will not be considered. Conventional hard-switching technique will be used instead for soft-switching, ZVS (zero voltage switching), since it simplifies the LTspice simulation. For the OBC, a controller area network-bus (CAN-bus) communication is required to start the OBC. Then it is possible to do voltage- and current measurements. However, this bus communication is not available since there is no laboratory set up, which means that the OBC cannot be up-started. Thus, no measurements will be conducted. Considering the size, the data communication system is assumed to fit the size of future compact off-board charger. Therefore, a narrow investigation will be carried out to understand what data communication is required for the charger. Depending on what communication channel is needed, for an example control pilot (CP), prox- imity pilot (PP), CAN-bus or power line communication (PLC) will decide how the pin configuration on the DC nozzle interface will look like. The electromagnetic-interference (EMI) will not be considered. However, the size of the input and output filters should be taken into consideration as they may have an impact on the weight and volume. Other circuits that are not a main part of the converters are neglected. The economic aspect will neither be a priority during this project. The project will be conducted over a period of five months, thus restrictions within this project are essential to pursue. 5 2 Method A literature review of converter topologies were carried out, and necessary informa- tion and data were collected from related work as well. Furthermore, CEVT provided this project with an experimental OBC and a charging cable for disassembling and examination of its components and design. An examination of the OBC was be conducted to obtain weight and size of the power electronic components. This aided to make an easier comparison between the components of the OBC and the selected components for the off-board charger. The module of the charging cable was a reference size to aim for for the off-board charger. Since LTspice was not able to measure real power for reactive components and measure reasonable power dissipation data, calculations were carried out regarding efficiency and power losses. Simulation models in LTspice were created to verify that these values are reasonable and to obtain data about PF and THD. The switching frequency was according to standards regulations. The power density was estimated from the final charger. Additionally, to reduce the size of the components in the circuit, different topologies of power electronic designs models were investigated. These topologies were analyzed for various current levels. The first sizing of the components was done with calculations and later verified with simulation. The most applicable topology was selected to advance with for further analysis. A comparison regarding electrical performance, thermal aspects and size was made between the off-board charger and the OBC. Decision matrix were utilized to make an informed choice by listing performance objectives, showing which factors are most important for this project and adding a weight number. The summation of the weight number and the point system together, gave the total points for the different subject and could be compared with each other. The lowest value of the total points is the selected alternative. In order to obtain the heat and power losses, following steps procedure were con- ducted: • Step 1, is to select a general solution and with reasonable components. Calcu- lations will be based on current level variant 4 A. This will not be included in the report. • Step 2, means that more variants are calculated based on formulas and details from step 1. In this step, power losses and heat calculations will be made. 6 2. Method These calculations and results are an important part of how a configuration is chosen. • Step 3, is to sum up the calculations and values that result from the selected configuration. Choosing suitable cooling can be suggested and calculated for the chosen configuration. Finally, a reasonable design can be displayed. A method of how to design a transformer is shown in Figure 2.1. Figure 2.1: Transformer design flowchart. Based on the selected components, the cooling system could be designed. A heatsink was designed by hand calculations to obtain the physical size. To analyze if the heatsink was sufficient enough to reduce the high-temperature levels, a simulation model was created in Heatsinkcalculator software to observe thermal radiation and critical temperatures. By considering the choice of topology, components, size of the cooling system, including connectors, enclosures etc. the finalized size and weight of the module can be estimated. The collected information and results were recorded in a thesis report and was presented at both the company and the university. 7 3 Safety and Sustainability 3.1 Safety In the recent years, the increase in sales of EVs and various charging methods has made it easier to charge the EVs in the comfort of people’s own home [14]. Statistically, home charging tends to be popular, especially since it is time-saving and comfortable. Although, some caution is required when using a charging cable that comes with when buying the EV, since it might draw more current from the wall-plug (around 16 A) than the outlet is designed for. This could be due to lack of educational level on home charging, since a charging cable is included in the deal of buying an electric vehicle the customer assumes this cable can be safely utilized. Another reason why the customer chooses to use the charging cable that includes with the vehicle can be that buying a wall-box is too expensive to invest in [15]. These factors point out that the home-charging is not simple enough to use and can lead to hazard consequences for the user and others. Since February 2018, the Swedish government decided to fund home charging installation of wall boxes (mode 3) by giving cost contribution for each household. The reason for this is to encourage more people to switch to electric vehicles. Another reason why people tend to utilize home charging could be simply that it is more convenient, where schuko outlets are having a vast distribution geographically [16]. However, considering charging for household’s applications, where the customers directly connect the charger to a wall-socket, entails the possibility for cable fires to occur, which could lead to fatal outcome. This happens due to that the cables are not dimension for high power charging during a longer amount of time [14],[15]. To avoid this hazard, people are encouraged to ensure that they have a valid European standard wall-box installed in their households. Insurance companies have pointed out that the household insurance would not cover the damages of the fire if there was a cable fire with an invalid wall-box installed [14]. Considering the off-board charger, it is feasible to assume that this charger could utilize the schuko outlets. This is due to fact that it is operating as a low power charger, which would reduce the load. Thus, the wall-socket is not able to reach high temperatures, and thereby the risk for cable fires are minimized. This points out that this charger would be a safer alternative for home-charging applications. However, even if this charger is not able to reach higher temperatures, it is suggested 8 3. Safety and Sustainability that the shucko outlets must be equipped with a residual-current device that have the ability to break the current instantly to avoid hazards events from occurring [15]. 3.2 Sustainability In relation to the sustainability, the power system should endure multiplicity con- nection to the wall-sockets simultaneously, especially in the evening. This is the main issue rather than if the amount of power production is sufficient enough to sustain an operational society with power [17]. One solution could be to implement smart controllers where the loads are charged at different time points during the day depending on what needs to be prioritized, for instance, the grid or the vehicles. Smart control of the power system could ensure that the system is not under heavy load at the same time. However, a scientist team from VTT Technical Re- search Centre of Finland, investigated an imaginable scenario if 5 million EVs was connected to the grid, and concluded a rise of power demand of 3.8 GW was needed if the smart control was not implemented. If the smart control was implemented, only 1 GW more of power demand was required [17]. Yet, researchers at Chalmers University of Technology predicts that this still could be an issue for smaller local grids, where the transformer capacity may not be enough to supply additional loads connected. Simultaneously, while moving towards a renewable energy generation, it is essential to find resources that can replace coal and nuclear power in the future to achieve the extra power demand needed for the EVs. A suggestion is to implement unused charged EVs with bidirectional function, to utilize V2G technique, where the vehicle can operate as batteries to supply the grid [17]. This could benefit the grid to avoid a power outage. This off-board charger, which is a low power charger, would not affect the power grid to the same extent as an OBC would, as it is a high-power charger. Thus, by having a lower load for the power grid to handle would likely prevent the grid from falling into power outage. Other solutions for the future could be to redesign the current AC grid to a DC grid, thus the wall-outlets provides DC instead. This grid could provide all elec- tronic devices including PHEVs to charge directly from the grid. Thus, the power conversion process within the vehicle or outside the vehicle is not needed anymore, leading to space-and cost savings for the PHEVs. This would likely put an end for the development stage for the OBC and the off-board charger, where these would be considered to be obsolete. 9 4 Theory This chapter is presenting essential background theory regarding various technologies, approaches, and implementations to reduce overall size for OBCs. The provided OBC from CEVT is also presented with function and design specification. These descriptions will aid to find the most applicable solution for a functioning compact off-board charger for PHEVs. 4.1 Plug-in hybrid electric vehicle A PHEV is a hybrid vehicle that belongs to the electric vehicle (EV) family. It differs from other EVs since it is equipped with both an internal combustion engine and an electric motor (EM) in its propulsion system. The battery in the PHEV can be recharged by being plugged into a power source by using a charging cable. This cable is either connected to a charging station or power outlet in a household. The range anxiety that may exist with battery electric vehicles, does not exist with the PHEVs since they will not run out of power even if the battery is uncharged [18]. 4.2 On-board charger A charger for the HV battery, known as the OBC, is located inside of the PHEV as a part of an EV’s powertrain. The OBC consist of three main components; AC/DC converter, PFC controller and DC/DC converter shown in Figure 4.1. Additionally, some filters are included in the input and the output of the charger to reduce EMI. The functionality of the OBC is to convert the AC voltage from the supplying grid to DC voltage to the HV-battery of the vehicle. The stored electrical energy in the recharged battery is inverted from DC to AC voltage, which is transformed to mechanical energy via an EM to initial movement of the wheels [19]. Figure 4.1: The energy conversion system inside a PHEV [20]. An overview of an OBC is shown in Figure 4.2. The AC/DC converter rectifies 10 4. Theory the AC voltage from the power source to a DC voltage, usually integrated with a Boost PFC converter circuit [21]. The DC/DC converter is isolated to give galvanic insulation ability, where the DC voltage is transformed to high frequency AC voltage. Next, this voltage is transferred, controlled and insulated through a transformer. Finally, the voltage is rectified to DC and remove high-frequencies components from the voltage [21], [22]. Both AC/DC- and isolated DC/DC converter is supervised by a controller [22]. Figure 4.2: Electric vehicle battery charger architecture [23]. 4.2.1 AC/DC converter The AC/DC converter consists of a Full-bridge rectifier and Boost PFC converter, where this type of converter has the benefit to provide a low THD for the input current that includes harmonics, simple converter design and can sustain a moderated power factor. However, the Boost PFC converter size is increasing in proportion to the delivered power. This implies that if the high-power output is needed to be supplied, the converter becomes larger [22]. 4.2.2 PFC controller To achieve a sinusoidal AC output current with minimal-phase-shift between the current and voltage, it is vital to have an operational control system in the circuit. This control system is typically a closed-loop control designed as a PI- or PID controller that is controlling the gate at the switch with PWM signals [24]. The gate is alternating between on and off mode, to ensure that the inductor current is reaching the settled reference values. In other words, the PFC controller is forcing the current to be drawn in phase with the input voltage [25]. Since the error amplifier inside the controller is operating with a slow pace, it may take a certain number of cycles of the power line before the resulting output is stable [26]. The controller receives measurements of the output voltage, reference voltage, AC input voltage, and chopped average inductor current to compare these with their reference values in order to decide the amplitude and duty cycle of the PWM [24]. If there is a deviation between the reference value and measurement values, the PWM signal is modified by adjusting the duty cycle depending on the level of Boost that 11 4. Theory is required [24]. This is illustrated in Figure 4.3, where a typical conventional Boost configuration is showed. Figure 4.3: Receiving and distribution signals from the PFC controller to an AC/DC converter [27]. 4.2.3 DC-link A DC link, also known as a capacitor bank, is an intermediate stage between an input stage and output stage. This DC link commonly consists of a capacitor that is connected between two other capacitors with positive and negative ends. These stages are connected with each other by this DC link, which is shown in Figure 4.2. A common input stage is AC/DC converter (rectifier) with PFC circuit, where the DC link is operating as an output filter to prevent voltage ripple. To diminish the voltage ripple and improving the PF by reducing the reactive power, it is favorable to parallel connect several capacitors with a proper dimension [28]. For the output stage, a converter with switching ability or an inverter is usually implemented. If the AC/DC converter is operating for high power mode, the capacitance value is as low as possible to increase the storage capacity for storing energy, while converters with low power mode, needs to have low capacitance to ensure that the voltage ripple is not interfering [30]. 4.2.4 DC/DC converter A DC/DC converter is an electrical circuit that converts from a DC value to another DC value. Specifically, first, the incoming voltage is rectified to a DC, followed by switching devices that convert the voltage to a high-frequency AC voltage. Then, this AC voltage is converted back to a desired DC output voltage [30]. Within an OBC, an isolated Full-bridge DC/DC converter is used. This converter contains a transformer that provides an isolation in high-power applications, which is in this case important during battery charging. On the primary side of the trans- former, there are four switches, in this case metal-oxide-semiconductor field-effect transistor (MOSFET) are used to make a Full-bridge. While on the secondary side, 12 4. Theory rectifying diodes are used. This topology can be equipped with zero-voltage-switching (ZVS), which means that there will be lower switching losses and therefore higher efficiency [31]. The DC/DC converter design is dependent on different factors such as efficiency, switching losses, and low stress. These factors need to be considered when creating a small converter size as possible [32]. 4.3 Performance and size To make the overall design of the charger as compact as possible, it is essential to reduce the size of the components and maintain high performance. In this case, the reactive components, inductors, and capacitors are desired to be reduced in size. This is because their size and volume are the largest inside the power electronic equipment and filters. In this subsection, it is presented how this can be carried out. 4.3.1 Standard regulations EV charging applications are obliged to be in accordance with international standards regulations. These regulations show in what span the electrical performance requires to be within. For the switching frequency for the PFC, the minimum and maximum frequency is 70 kHz respectively 150 kHz, before too high distortion is obtained [33],[34]. For the DC/DC converter, the frequency range is between 140 to 350 kHz [35]. THD is accordance to EN 50160, where the supply voltage is required to be lower than 8 % [36]. The PF is desired to be at least 0.9 at full load according to IEC 61000-3-2 and IEEE-519 [24], [25]. Lastly, the majority of the OBCs have in general an efficiency of at least 90 % [37], [34]. In Table 4.1, the regulations are shown. Table 4.1: Standards regulations for the performance of the electrical parameters. Parameters Standard values Unit AC/DC fsw 70-150 kHz DC/DC fsw 140-350 kHz THD ≤ 8 [%] PF ≥ 0.9 - Efficiency η ≥ 90 [%] 4.3.2 Switching technology For the selection on what switching device to use, MOSFET and insulated-gate bipolar transistor (IGBT) are investigated since they are both within the correct range (assumed to be approximately 150-200 kHz, 600 V, and 4-8 A). From Table 4.4, it can be seen that both IGBT and MOSFET is within the right parameters for this project scope compared with other switch technologies. 13 4. Theory Figure 4.4: Comparison of semiconductors regarding frequency, voltage and current [39]. Therefore a comparison between MOSFET and IGBT is shown in Appendix B in B.3. From datasheets, the components are desired to have low resistance of the MOSFET as possible, since it will require less cooling [41], [42], [43]. Both MOSFET and IGBT transistors can be used, and in order to select one of the transistors for this project, a points-matrix was made. The factors that were considered most important are high switching frequency, temperature, efficiency, low RDS(on) and losses at high frequency. RDS(on) is the drain-source on resistance inside a transistor during on-state. This resistance decided the maximum current of a transistor. Since temperature and RDS(on) are linked together, high temperatures will increase the resistance [44]. The transistor that contains the best out of these factors has been given the grade ’1’ and the second best transistor has been given the grade ’2’ seen in Table 4.2 [40]. 14 4. Theory Table 4.2: Comparison of MOSFET and IGBT [40]. Factors MOSFET IGBT High switching frequency 1 2 ZVC benefits 1 2 Higher efficiency at low voltage 1 2 Lower thermal impedance, hence better power dissipation 1 2 Elimination of current tail 1 2 Higher voltage and current capabilities 2 1 Cost at low power 2 1 Losses at high frequency 1 2 Low Rds(on) 1 2 Diode recovery behavior 1 2 Strong gate driver 1 2 On-state losses 1 2 Performance at low power 1 2 Body drain diode 1 2 Can perform fast switching applications with little turn-off losses 1 2 Lower conduction loss 2 1 Total point: 19 p 29 p From this table, the MOSFET received 19 p while the IGBT got 29 p, hence MOS- FET is selected for this project. A MOSFET is active when supplied voltage is between the source-gate. Not until then, a drain current is conducting from the drain to source. Typically, MOSFET mitigate the switching losses since it is able to operate with a high switching frequency [6]. 4.3.3 EMI filters EMI is undesirable electrical signals, occurring in power electronic equipment due to frequent changes in both voltage and current [6]. In order to sustain a high-power quality, the arising EMI equipment needs to be mitigated, which is why it is crucial to conceal the interference [48]. This is achieved by implementing EMI filters, that eliminates high EMI by ensuring to decrease high frequency. An EMI filter consists of passive electronic components (inductors and capacitors) that mitigates and increase the resistant against interference for the devices that need to be shield [48]. A basic EMI filter can either be configured as π-filter, L-filter or T-filer [49]. A typical EMI filter for single-phase system is shown in Figure 4.5 [51]. 15 4. Theory Figure 4.5: EMI filter for single stage power supply [39]. EMI filters can be designed to handle two types of noises: common mode (CM) and differential mode (DM) interference [51]. The capacitors C1 and C2 seen in Figure 4.5 is suppressing DM noise, known as the DM filter, while L1, L2, C3, and C4 are mitigating the CM interference, known as CM filter [50]. CM noise is the measurement of voltage or current between the ground and power lines, while DM noise is the measurement of voltage or current between the power lines, either as line-to-line or line-current [6]. The size of the filter can be decreased by increasing the switching frequency. However, this reduction is only true of the DM filters, meanwhile, the CM filters are slightly increased [52]. Furthermore, the inductor chokes can be mounted on the top of the filter capacitors in order to save space on the circuit board [53]. For the off-board charger, it is assumed that two EMI filters are included. Since EMI is not investigated, it is not possible to conclude how many filters are required. Thus, in order to include possible weights from the filters to the AC/DC module, this assumption is made [37], [34], [38]. 4.3.4 Switching frequency According to an IEEE article, an experimental attempt was made to make a compact high efficiency 3.3 kW OBC. The switching frequency was selected to 70 kHz for a PFC circuit and 200 kHz for a ZVS Full-bridge DC/DC converter [34]. For the PFC circuit, 70 kHz was selected to fulfill the EMI regulations, which implies that a switching frequency should be below 150 kHz. For the DC/DC converter, a switching frequency of 200 kHz was selected to ensure that the losses are minimum at full performance of the OBC. This is based according to Figure 4.6, where the ratio of the efficiency and output power is shown for an output voltage of 300 V [34]. This is reasonable to assume for this project since the HV battery requires an operation range of at least 320-380 V in order to charge. 16 4. Theory Figure 4.6: Output power from the DC/DC converter in relation to the efficiency at various switching frequency levels [54]. As shown in Figure 4.6, the efficiency is increasing with higher frequency, and increasing further with higher power output from the DC/DC converter. However, a too large switching frequency is undesirable to use since the inductors- and capacitors components are manufactured within a limited range of size [4]. With the case of the off-board charger, it is suggested to use a switching frequency of 150 kHz to reduce the components size for the PFC circuit. Since this project is about to design a low-power charger, an output power range between 800 W and 1200 W is more appropriate to expect from the DC/DC converter, seen in Figure 4.6. From this range, 900 W is selected for further research. Considering the switching frequency, a too high frequency would lead to higher switching losses in the switching devices. Therefore, a frequency of 200 kHz is selected to obtain a high reasonable efficiency of 92 %, avoiding too large losses and ensuring that the size of the components is reduced. 4.4 Power factor correction In order to obtain the highest power quality while designing AC/DC converters, it is fundamental to establish a reasonable PF and THD, that follows international standards. PF is described as the ratio between real power (P) and the apparent power (S), or as if there is phase-shift between a sinusoidal current- and voltage waveforms, the equation can be expressed as [55]. PF = P S = Kθ × Kd. (4.1) 17 4. Theory With a phase shift, the cosine of the phase angle between the current and voltage is established. This known as the displacement factor Kθ = cos(Φ). The second parameter, distortion factor Kd = IRMS(1) IRMS , determines how sinusoidal the waveforms is [7]. Since the real power is the transferred component, the reactive power (Q) must be as low as possible. This would give PF closer to 1.0, known as unity, where the active power is equal to apparent power. A sinusoidal current and voltage that are in phase with each other imply that PF is close to 1.0 [24]. This is known as a linear load. If the waveforms are sinusoidal but not in phase, the PF is not equal to 1.0 [55]. This is known as non-linear load [58]. This is illustrated in Figure 4.7, where the current spikes representing harmonic currents. Figure 4.7: Illustrating the voltage- and current characteristics waveforms by not having a PFC to the left, and by having a PFC [59]. In order to reduce the distortion and ensure that the input current is in phase with the input voltage, known as a resistive load, a power factor correction (PFC) is needed. This would improve the overall power quality, where reduction of current harmonics, low output voltage ripple, increase efficiency, multiple output voltage levels, fast output dynamics and good load regulation. PFC is also a require- ment within AC/DC-converters according to international standards, such as IEC 61000-3-2 and IEEE-519 [55]. For a power supply with PFC, the PF is expected to be between 0.95-0.99, while a power supply without PFC has a PF of 0.70-0.75 [60]. THD is defined as the level of distortion at the input current in this case shown as [6] THDi = 100% × √ 1 K2 d − 1 (4.2) Considering a linear load in equation 4.1, thus Kθ=1, gives PF = Kd. This implies in PF = 1√ 1 + (THD)2 (4.3) 18 4. Theory Since PF is linked to the total harmonic distortion, an increase in THD will result in a decrease in PF. This is not desired since harmonic distortion can cause damages to cable, create overheating, circulating currents, fire risk, equipment malfunction and component failures [57]. 4.4.1 LT1248 - PFC controller The IC LT1248 is a universal controller with power factor correction. It can regulate to a maximum power of 1500 W and is able to operate in both continuous conduction mode (CCM) and discontinuous conduction mode (DCM) [61]. An overvoltage protection circuit is located at pin 8, where the output voltage is decided based on the equation as follows Vout = Vref × ( R1 + R2 R2 ) (4.4) where Vref is the reference voltage, set as 7.5 V for this IC. The schematic model of the LT1248 in LTspice is shown in Figure 4.8. Figure 4.8: Schematic of LT1248 PFC controller in LTspice. It has maximum switching frequency of 300 kHz internally. At pin Rset and pin Cset of the IC, the switching frequency can be determined. Since the switching frequency is equivalent to the oscillating frequency, it can be calculated according to the forumla fsw = 1.5 RSET × CSET (4.5) where RSET is the the set resistor and CSET is the set capacitor [61]. These values are selected from a frequency graph found in the datasheet for this PFC. For this project, several IC designs were considered, but since this PFC controller had the highest maximum output power, this controller will be selected for the LTspice simulation. 19 4. Theory 4.5 Topologies Different topologies for both AC/DC and DC/DC circuits are presented and one of each is selected for this project. 4.5.1 AC/DC converter with PFC To access an overview of some different common (active) PFC topologies, such as the Classic Boost PFC, Dual Boost PFC, Totem-pole Bridgeless PFC and Interleaved Boost PFC, a few of their characteristics are presented in Appendix B in Table B.5. In order to decide between what PFC topology to use in the off-board charger, a comparison table of characteristics between different PFC topologies is made. From the most important factors like the number of components and suitable for low power rating, the classic Boost, and Totem-pole Bridgeless PFC are both selected for further investigation. Furthermore, to be able to select an optimal PFC for this project, a compari- son between only classic Boost PFC and Totem-pole Bridgeless PFC is made, which can be seen in Table 4.3. From the table, some of the factors have been given higher priority such as the number of components being used, efficiency and power rating, marked with a �. In Table 4.3 there is a point system, where the grade "1" means the best char- acteristics and the grade "2" means secondary. Table 4.3: Comparison of Classic Boost PFC and Totem-pole Bridgeless PFC. Factors/Topology (PFC) Weight Classic Boost Totem-pole Bridgeless Low EMI 1 1 Power rating � 1 1 Number of components � 2 1 Efficiency � 2 1 Switching losses 2 1 Inherent ZVS 2 1 High frequency noise 2 1 Simple control 1 2 Cost 1 2 Common-mode interference 1 2 Can use Si, GaN, SiC (or only possible with GaN or SiC) 1 2 Total 16 15 The total points when looking at these factors are shown in Table 4.3, where it is seen that the Totem-pole receives the lowest amount of points, which is the preferable one. 20 4. Theory However, the score is even between the classic Boost and Totem-pole. Therefore, simulation and calculations regarding the electrical performance of both Boost and Totem-pole will be conducted. However, with an additional weight of ’×2’ at the prioritized factors marked with stars, the total scoring becomes 21 points for Boost and 18 points for Totem-pole. Thus, if other selection methods between these two typologies are not conducted, Totem-pole is selected. The classical Boost PFC is one of the most common PFC application on the market. It can receive a peak line voltage between 0-375 V, and provide an output voltage of more than 380 V. It is equipped with a Boost inductor after the rectifier bridge, which generates a continuous smooth current. This gives the benefit that no more filter is required to be implemented, thus, lower cost for the topology [45]. The circuit design for the AC/DC converter with Boost PFC topology is shown in Figure 4.9. Figure 4.9: AC/DC converter with Boost PFC configuration [66]. Compared to the Boost PFC, the Totem-pole PFC have fewer number of components and low conduction loss [67]. Additionally, it provides a higher power density, higher efficiency and low CM noises [68]. For the PFC operation, the Totem-pole is active during the positive half and negative half cycles of the AC input waveform. The current flow is adjusted by determining the switching frequency for the transistors. The AC/DC converter with Totem-pole PFC topology is shown in Figure 4.10. 21 4. Theory Figure 4.10: AC/DC converter with Totem-pole PFC configuration [66]. Operation mode It is required to select an appropriate operation mode to calculate values for the components in the selected PFC. Hence a comparison has been conducted between the operation modes CCM, critical conduction mode (CrM) and DCM which can be seen in Appendix B in Table B.4. To select an operation mode to use, a comparison points-matrix was created. The factors that were considered most important are stress on components, losses, low di/dt, these factors are also marked in the weigh column with a �. The grade ‘1’ in Table 4.4 represents the mode with the best abilities for the different factors. Since the DCM operation mode has the highest peak current compared to both CCM and CrCM without any major advantages over CrCM, as shown in Appendix B, the comparison will be between CCM and CrCM in Table 4.4 [69]. A weight ‘×2’ has been added to the most important factors shown with(�) in Table 4.4. 22 4. Theory Table 4.4: Comparison in operation modes [69]. Factor Weight CCM CrCM Switching frequency � 1 (easier to filter) 2 Switch stress � 1 2 Inductor stress � 1 2 Diode stress � 2 1 Losses (at medium-high power) � 1 2 Cost 2 1 Large inductor value � 2 1 Low energy 2 1 Performance 1 1 Power saving (low power) 2 1 Improving power density (low power) 2 1 Low di/dt, can reduce EMI � 1 2 Turn-on losses (with MOSFET) 2 1 Total points: 29p 30p As shown in the table, the total points were even, where the CCM received 29 p while the CrM collected 30 p. From Table 4.4, operation mode CCM is selected. 4.5.2 Material With the selected transistor to be MOSFET from Table 4.2 and the selected operation mode to be CCM from Table 4.4, there will be turn-on losses to take into consideration. To avoid too high losses, a low value of the reverse recovery charge (Qrr) is needed. Hence to obtain a low-value Qrr, ultra-fast diodes or silicon carbide Schottky diodes can be utilized at CCM mode [45]. Ultra-fast diodes improve the efficiency and can be implemented in AC/DC conversion equipment [46]. Schottky diodes are commonly used for hard-switching implementation and EV applications [47]. 4.5.3 DC/DC converter The majority of OBCs on the market are equipped with a Full-bridge converter configuration for the DC/DC converter stage [34], [74], [37]. Thus, the DC/DC con- verter for this project is investigating the Full-bridge. Furthermore, the Half-bridge topology is also analyzed since it might have a less impact on the size and weight. There are two types of Full-bridge converters, Full wave bridge rectifier and Full wave center tapped rectifier. The main difference is that the Full wave center tapped has a bulky center tapped transformer, which is also costly. Therefore, the full wave 23 4. Theory bridge rectifier is preferable to investigate since it is smaller in size and less costly. From here on out, the Full wave bridge rectifier will be referred to as only Full-bridge converter [75]. A Full-bridge converter consists of four switches and four diodes, divided into two legs, shown in Figure 4.11 There are two switches on one leg and they do not conduct at the same time. However, to avoid short-circuit on the DC input, the switches are turned-off during a short ‘blanking’-time. If ideal switches are assumed then the switches are not turned-off simultaneously [6]. Figure 4.11: Isolated Full-bridge DC/DC converter [39]. The Full-bridge converter is designed with four switches, where the switch pair T1 and T2 are switched on together and T3 and T4 are then off until they shift. The Full-bridge converter also has an electrically isolated transformer. Some advantages and disadvantages for this Full-bridge converter are presented in Table 4.5 Compared to the Full-bridge, the Half-bridge DC/DC converter only has one leg with two switches and two diodes or capacitors. This is illustrated in Figure 4.12. Figure 4.12: Isolated Half-bridge DC/DC converter [39]. The capacitors C1 and C2 are arranged to create a voltage potential point. For the 24 4. Theory switches, T3 and T2 are operating separately from each other. The diodes connected in parallel from the switches are used to protect the switches [6]. The advantages and disadvantages of the Half-bridge are listed in Table 4.6. Table 4.5: Advantages and disadvantages of the Full-bridge converter [76],[77], [78]. Full-bridge DC/DC converter Advantages Disadvantages Utilized for high power applica- tions (>1kW) High number of components is re- quired. The output voltage of a Half- bridge is twice as large, giving higher power rating. This gives the possibility to reduce the num- ber of windings as well, and yet remaining with the same output voltage. Less economical alternative, in re- lation to high conductive losses if this converter is operating for low power rating. Utilizing four diodes to rectify and polarity change the incoming wave. Higher losses than the Half-bridge converter, due to it have more switches components. An efficiency between 90-98 %. More voltage ripple than a Half- bridge. 25 4. Theory Table 4.6: Advantages and disadvantages of the Half-bridge converter [76], [77], [78]. Half-bridge DC/DC converter Advantages Disadvantages Utilized for high power applica- tions (250W - 1kW). At the DC bus, the supply poten- tial is half-sized, thus the output voltage is reduced. This implies that the current needs to double to obtain the same power output. Thus, the transformer core needs to be large. Air gap of the magnetic path is not needed. Current mode control is not appli- cable for this converter. Cheaper and simpler to design. Two capacitors connected in series are required at the input of the bridge and needs to have a large design due to the capacitance is halved. Lower switching losses than the Full-bridge converter, due to that it has fewer switches components. Since only two transistors are op- erating, they are required to han- dle twice as large conducting cur- rent, resulting in high losses. An efficiency between 88-96 %. Since a comparison between different output power from 900 W, 1300 W and 1700 W are conducted, both Half-bridge and Full-bridge may theoretically be suitable candidates for external charger application. Simulation and calculations of both Full-bridge and Half-bridge will be conducted to analyze which one is most preferable to utilize. 4.5.4 Number of components To see the difference clearer between the number of power electronic components used for the selected topology circuits, a list is summarized in Table 4.7. The Boost and Totem-pole circuits represent the selected AC/DC converters, while Full-bridge and Half-bridge represent the selected DC/DC converters. EMI-filter components are not included in the list since they will be included in the final topology selection for all cases. 26 4. Theory Table 4.7: The number of power electronic components in the selected circuits. Circuits/ Components Diode PFC Inductor Switch Trafo Bridge- diodes Capacitor Total Boost 1 1 1 0 4 1 8 Totem-pole 0 1 2 0 2 1 6 Half-bridge 0 1 2 1 4 3 11 Full-bridge 0 1 4 1 4 1 11 4.6 Cooling implementation In this project, the off-board charger needs a cooling system to reduce high tempera- tures generated at the heat loads. This cooling system should have a size that can be encapsulated into the ‘module of the charging cable’ structure. Since the structure is enclosed, it might be difficult find a balance between having an efficient thermal management system and simultaneously prevent dust and moisture from infiltrating. The aim is to find a cooling system for a sealed or semi-sealed module [79]. The cooling system that is used in this project is natural or convection cooling, where the heat is transferred by air flow. To use the natural air flow efficiently, there is a need to add vents to the en- closed electronic equipment including a cooling fan. To prevent dust contamination while using vents, air filter units can be used [81]. A filter is necessary to use since dirt and moisture in combination can become conductive and thus can cause intermittent operation. For high power applications, convection cooling may not be practical to have due to the cost and size required. However, for this project it might be appropriate to implement cooling channels or pipes where the incoming air can be equally distributed over the module. An alternative to this could be to transfer the heat from the components to outside of the enclosed module. This can be achieved by using a heatsink as a roof of the charger module, that transfer and remove the heats. The high temperature producing components ought to be mounted directly in contact with the heatsink and evenly spaced to make the heat transfer as efficient as possible. An idea for a starting point with an enclosed module is shown in Figure 4.13 [82]. 27 4. Theory Figure 4.13: Sealed module enclosed with thermoelectrically enhanced heat rejection [80]. 4.7 OBC from Geely Group A specific OBC, named Geely 8888003014 G BQ8AA, was provided by CEVT for this project. This OBC was designed in 2017 by Geely Group and assembled here in Sweden. However, at the moment, a new model is under developing stage and is scheduled to be complete in 2018, where this model includes equivalent electronic components as in the 2017 model. The circuit board for the data communication system inside the OBC is manufactured by Kongsberg Automotive [37]. In Figure 4.14, the physical OBC is shown. Figure 4.14: OBC Geely 8888003014 G BQ8AA from a PHEV. 4.7.1 Design specification According to design specification of the OBC, this battery charger is a 1-phase 3.4 kW charger for PHEVs, where the power electronics and cooling system are packaged in a 4 liter metallic module. This gives a total weight of approximately 4 kg [37]. The maximum current rate of 16 A can be received from the grid to the OBC. For 28 4. Theory protection against environmental contamination, IP67 and IP6K9K standards are implemented into the shield design. The design specification for the OBC is listed in Table 4.8. Table 4.8: Design specification of OBC - Geely 8888003014 G BQ8AA [37]. Specification Values Units AC input voltage range 85-264 Vac AC input frequency range 44-65 Hz AC input current 2-16 A DC output voltage range 200-410 Vdc DC output power 3.4 kW PFC - switching frequency 98.3 kHz DC/DC - switching frequency 98.3 kHz Efficiency 92-94 % PF > 0.99 - Power density 1 kW/kg Dimension (H x W x L) 6.8 x 19.2 x 25.5 cm Weight ≤ 4 kg Volume 3329.28 cm3 Ingress protection IP67 - 4.7.2 3 separated connectors In order for the OBC to operate properly, there are three separated inbuilt connectors. These connectors are one AC connector, DC connector, and LV connector. The AC connector receives the incoming power from the charging cable, supplied from the AC grid, which is plugged into the power outlet. The DC connector delivers the converted DC power to the battery inside a PHEV. The LV connector receives and sends signals channel through a communication CAN-bus between the battery and the charging cable to establish sufficient and correct current level for the battery [37]. 4.7.2.1 AC connector The AC connector consists of five input pins. PIN1, PIN2, and PIN3 are corre- sponding to L1, protective earth (PE) to the chassis of the vehicle and neutral (N). Between the one single-phase conductor L1 and the neutral connector, receives up to 230 V. PE is the protective earth that protects equipment and people. PIN4 and PIN5 are interlocked connections that receive 15 mA. These connections are implemented for safety measurements, to ensure it is only possible to charge when there is a connection between the cable and AC connector at the electrical vehicle inlet (EVI). It has a weight of 103 g [89]. The pin-configuration for the AC connector is illustrated in Figure 4.15 [37]. 29 4. Theory Figure 4.15: AC connector with pin-configuration [37]. 4.7.2.2 HVDC connector The DC output consists of four output pins. PIN1 and PIN2 are corresponding to DC+ and DC-, where it is supplying between 200-410 V output voltage for the HV battery. Similar to the AC connector, the DC connector is also having PIN3 and PIN4 as interlocked connections. It has a weight of 106 g [89]. In Figure 4.16, illustrating the pin-configuration for the DC connector [37]. Figure 4.16: AC connector with pin-configuration [37]. 4.7.2.3 LV system connector The low-voltage (LV) connector receives and sends signals through a communication CAN-bus between the battery and the charging cable. This communication line is established in order to ensure sufficient and correct current level is delivered to the battery to avoid any damages to the battery. In regard to safety aspects, other signals are being sent to locking the engine during charging process [37]. This connector has a weight of 82 g [89]. 4.7.3 Circuit diagram According to the available documents from CEVT, the OBC is suggested to have a typically HV board architecture. This architecture is constructed with an AC/DC converter with interleaved PFC circuit coupled together with an isolated Full-bridge DC/DC converter. Both the interleaved PFC circuit and DC/DC converter are having a switching frequency of 98.3 kHz. Additionally, various type of filter is implemented at the front and end of the charger to mitigate the level of occurring ripples and thereby establish a higher level of power quality. A schematic diagram of the OBC is shown in Figure A.1 Appendix A. 30 4. Theory 4.7.4 Data communication In order for the charger to know when the battery is fully charged or needs to be recharged, a data communication line is required between a vehicle and power supply station. At present, the connector is equipped with two separated communication contacts: CP and PP. A CP is used for transmitting significant information using pulse width modulation (PWM) signals, while PP ensures that the engine startup sequence is deactivated, the cable is attached to the EVI, and maintaining ampacity during the complete charging process [83]. Ampacity is the maximum rated current that is conducted through a cable to withhold a safe temperature level [84]. Both CP and PP detects if the cable is attached to the vehicle, and are being transmitted through the EVI. The return signals back to the charger is going by the PE. When the battery is fully charged, the charging process is aborted inside of the vehicle and release the connectors by a break apparatus [37]. Depending on the geographic location in the world, the configuration arrangement of the connector has different styles to benefit the market and customers. CP and PP are often arranged symmetrically from each other either combine with the phases. For single-phase function, CP and PP are combined with L1, N, and PE, and for single-and three-phase as L1, L2, L3, N and PE [85]. In Table 4.9, the different configuration arrangement is described, where GB is a Chinese abbreviation for Guobiao and CCS stands for combined charging system [85]. Table 4.9: Type of connectors with configuration arrangement of the contacts [86]. Mode/type Type 1 (US Type 2 (EU) GB (China) Japan AC SAE J1772/ IEC 62196-2 IEC 62196-2 GB Part 2 IEC 62196-2 DC Tesla US Tesla EU GB Part 3/ IEC62196-3 CHAdeMO/ IEC62196-3 CCS (AC+DC) SAEJ1772/ IEC62196-3 IEC 62196-3 4.7.5 Operation of OBC The OBC Geely 8888003014 G BQ8AA follows a specific operation pattern before, under and after the charging process. This is regulated and controlled by the micro- controller unit (MCU) SPC560B54L3B4E0 located on the low voltage board system. This operation pattern is shown and listed as following [37]: 1. Charging system is blocked when the propulsion system (electrical- and com- bustion engine) is running. 2. Propulsion system is blocked when the charging system is about to charge. 3. Data communication through the EVI, where CP and PP detect if the charge cable is attached to the vehicle. If the cable is attached, OBC will wake up from it sleeping-mode and begin to perform CAN communication with the HV battery. 31 4. Theory 4. The OBC receives essential data, maximum voltage, and current limitation, from the HV battery. This is done to know how much generation the OBC is allowed to produce and to avoid any damages on the battery. 5. While the CP and PP limit the maximum current from the AC grid. 6. As soonest the levels from the grid and HV battery are verified, the motor is locked and cannot be ignited. Simultaneously, the cable is attached to the vehicle. At this stage, the LEDs in the EVI illuminates yellow. 7. Through the EVI, OBC updates the charger with previous status information if the cable connected, charging rate, error etc.) 8. When PP, CP and AC voltage from the grid is approved, a digital signal processor (DSP) inside the OBC activates the charging process with a signal. Now, the LEDs in EVI illuminates with green lights 9. Through a CAN communication, the vehicle dynamics domain master (VDDM) permits for unlocking the cable from the vehicle and abort the charging process. 10. As soon as the cable is detached from the vehicle, the motor is being unlocked, and the propulsion system can be ignited again. LEDs in the EVI illuminates with yellow light. 11. When the cable is unplugged from the EVI, the LEDs illuminates with white light. 12. CAN Communication from the OBC begins to cease, and goes back to sleeping- mode. 4.7.6 Heat losses The only measured data over the thermal losses for the OBC as a component itself was available. These data are presented in Table 4.10, where the efficiency from the OBC for different currents ratings is presented. The rest of the efficiency are thermal dissipation. This is measured with a 230 V (RMS) and 50 Hz supply as an input. The output voltage is selected to be 375 V since this level is close to the battery start charging point, 380 V. For this output voltage it requires a cooling temperature of 40 ◦C. As shown in the table, various input current and output voltage ratings affect the amount of heat dissipation. By having a higher current, leads to that lower heat losses are obtained. Vice versa, by having a lower current- and voltage rating gives higher heat losses [87]. 32 4. Theory Table 4.10: Efficiency and power dissipation of the OBC with various currents. These data was measured with a cooling temperature of 40 ◦C [87]. Input current [A] Output voltage [V] Efficiency [%] Dissipation [%] 2 375 88 12 4 375 92 8 6 375 93 7 8 375 93 7 10 375 94 6 12 375 94 6 14 375 94 6 16 375 94 6 4.7.6.1 Transformer losses According to a test report from 2016, the highest power losses occurring internally of the OBC is located in transformer [88]. These losses are generated by listed as: • DC losses • Skin effect losses • AC losses due to proximity effect These parameters are based on winding losses in transformer shown in Figure 4.17. Figure 4.17: Losses in the winding in transformer at 100 kHz [88]. The Litz wire is assumed to have 32 strands, giving a cross section of 1 mm2. It should be noted that transformer is provided with thermal interface material near 33 4. Theory the surface of the transformer and capacitors [88]. 4.7.7 Weight and volume From a datasheet about assemblies and materials related to the Geely OBC, in- formation about power electronic components serial number, quantity and their weight was given [89]. However, the given datasheet is not the latest version, which suggests weight and size estimation needs to be conducted for each power electronic component inside the OBC. Their weight and size are described in subsection 5.1.2. For the off-board charger, the weight and volume of the components are being estimated based on data from datasheets. These components will be compared with the components from the Geely OBC. 4.7.8 Thermal management In order to reduce heat losses occurring at the heat loads, such as combustion engine inside the vehicle, requires a thermal management system. This type of system typically uses liquid cooling through pipes which is highly effective in removing heat dissipation. This type of cooling system is also implemented in PHEVs, where one of the heat loads is the OBC. The cooling system for the OBC is shown in Figure 4.18. Figure 4.18: Back side of OBC, where the cooling system is equipped with cooling pipes mounted to remove heat dissipation from the components. The cooling system has an estimated weight of 893.11 g. The heat load or charging- load is the only load occurring during the charging process when the vehicle is not moving, it implies that the drive-load can be excluded. To avert higher volume and size of the cooling system, one single heatsink may be sufficient to utilize for both loads [90]. In order to protect the OBC from environmental contamination, such as rain, dust, and sand, it is packed inside an impenetrable enclosure. To ensure to avoid high temperatures that are damaging the OBC, the walls of the enclosure is con- nected to the heat loads to diverting the heat to the outdoor air. Additionally, while selecting the insulation material to the walls of the enclosure, it needs to have a 34 4. Theory tolerable range for the electrical insulation and thermal conduction [90]. Considering an OBC with a charging rate of mode 1, delivering an efficiency of approximately 94 % with a power dissipation of 250 W [90]. In relation to this project, water cooling pipes is not an option since the charger will be outside the vehicle. Therefore, a heatsink implementation would be used. If a heatsink would not be sufficient to reduce high-temperature levels, air cooling in combination with a heatsink is another alternative to utilize. Additionally, to reduce the thermal further and saving weight for the OBC, heat pipes can be scattered over the base of the heatsink. This transport and condense the heat towards the flanges of the fan, and simultaneously increasing the power density. Installation of a liquid cooling plate for the power electronic devices can mitigate the temperature lower. These plates are often customized in regard to the power density, heat loads, and type of material that is selected. The heat loads are often transmitted over to the protected package where the heat is highly concentrated [90]. The pad size of each component on the printed circuit board (PCB) of the heatsink, is recommended to be larger than the component itself [34]. 4.8 Charging cable from Mennekes The mode 2 charging cable for EVs, designed by Mennekes, is produced to be utilized as an emergency charging cable only. An emergency scenario could be if there is no accessible charging infrastructure. For this reason, it is not designed to operate as a permanent charger for the load, compared to a charging station, which is a continuous load and is in accordance to the IEC 61851-1 standard [91]. The length of the cable is approximately 4 meters long [92]. The charging cable is connected to a certified commission for conformity testing of electrical equipment (CEE) wall sockets or households. To avoid high temperatures, the cable is equipped with a module, where an integrated heat-tracking device is included to ensure the cable itself, vehicle nor battery get damaged [91], [92]. When the temperature is too high, it breaks the current and thereby the charging process. As soon as a reasonable temperature level is reached, the charging process is initiated again. The module of the charging cable is presented in Figure 4.19. 35 4. Theory Figure 4.19: Charging cable with control and supervision module. Since it is connected to a power outlet, it is using single-phase operation, where the delivering current rate is having an adjustable range between 4-8 A. Considering the robustness of the cable, it can withstand a weight of 500 kg. One important safety feature, it is only operating and delivering power to the vehicle when it is connected properly to the EVI. In relation to contamination protection, it has an IP67 standard, implying it is fully dust protected and waterproofed [93], [91]. These design-specification is summarized and presented in Table 4.11. Table 4.11: Design specification of Mennekes charging cable [92]. Mennekes charging cable with Mode 2 charging for EVs Protection IC-CPD 1 protection (IP67) AC grid Single-phase operation, 230 V, 50 Hz, 30 mA Delivering current rate 4-8 A with maximum 8 A Dimension (H x W x D) 5.2 cm x 24.0 cm x 10.0 cm Temperature range: -32≤ Tamb ≤ 40 C◦ This module from Mennekes will be used as a reference design, where the aim is to design a module with similar dimension for the off-board charger. 4.9 Thermal calculations 4.9.1 Power losses In this subsection, the loss calculations will be discussed in more detail supplemented with relevant equations [69]. Diode To calculate power loss for the input rectifier diode bridge, the following equations are needed Pconduction = VF × Iavg (4.6) 36 4. Theory where the value of the voltage drop VF is obtained from datasheet. The switching loss of a diode can be calculated by Pswitching = (Eon + Eoff ) × fsw ≈ Eon × fsw (4.7) where Eon is the energy during on-state and Eoff the energy during off-state. The Eon can be known by using Eon = 0.25 × QRR × VR (4.8) where QRR is the reverse recovery charge, VRR is the reverse recovery voltage. By adding the conduction and switching losses together, the total diode loss becomes Ploss = Pconduction + Pswitching. (4.9) Duty cycle The duty cycle for Boost and Totem-pole is calculated by D = Uo − Uin Uo . (4.10) The duty cycle for Half-bridge converter is DHalf = Uo × N1 Uin × N2 (4.11) and the duty cycle for Full-bridge converter is calculated as DF ull = Uo × N1 Uin × N2 × 2 (4.12) where N1 and N2 are the numbers of turns on the primary and secondary side of the transformer [6]. Transistor The conduction losses are expressed as Pconduction = RDS,on × Idc 2 × D (4.13) where the RDS,on are the drain-source on-state resistance, the Idc is the dc current after the rectifying diode bridge and D is the duty cycle. The switching loss is calculated by Pswitching = 0.5 × Vdc × Idc × (tr + tdon + tf + tdoff ) × fsw (4.14) where tdon is the delay on time and the toff is the delay off time from datasheet [94]. The total loss will then become Ploss = Pconduction + Pswitching. (4.15) 37 4. Theory The gate drive loss and turn on loss are neglected since their values are small. Inductor The inductor losses are calculated according to the equation Ploss = Pcore + PDCR + PACR (4.16) where the ACR is the AC resistance of the inductor and the DCR is the DC resis- tance of the inductor. The PACR and PDCR are the wire loss caused by AC and DC resistance. The DCR value is taken from the datasheet, where the core loss and PACR are low and therefore will be neglected [157]. The Boost inductor losses are calculated based on a design guide of a Boost PFC converter [69]. The inductor current IL,RMS is calculated by using the CCM boundary condition IL,RMS ≥ ΔI 2 (4.17) where the current ripple is ΔI = VL × D L × fsw . (4.18) To obtain the inductor current can now be found IL,RMS ≥ ΔI 2 . (4.19) The inductor copper loss can be calculated by using PDCR = PL,cu = IL,RMS 2 × DCR. (4.20) Capacitor When calculating the DC-link losses, the reactance value is first needed Xc = 1 2 × π × fsw × C . (4.21) Then equivalent series resistance (ESR) can be calculated ESR = tanδ × Xc (4.22) where the tanδ is taken from datasheet. Now the losses can be calculated using Ploss = Ico,RMS 2 × ESR (4.23) where Ico,RMS is the RMS output capacitance current [69]. Transformer Inside the transformer, there are iron losses, copper losses, stray losses and dielectric losses. The copper losses consist of skin effect, proximity and frequency components. The skin effect is the current at high frequencies that is concentrated near the surface of the conductor. While the proximity effect is when the current in an conductor 38 4. Theory influence other currents [99],[96]. At higher frequencies, the skin effect and proximity effect increases [96]. Iron losses include both eddy and hysteresis losses. Since high frequencies are used, the material that is best suited is ferrites. Ferrites are a combination of iron-oxide (Fe2O3) and metallic materials, such as Zinc. If ferrites are utilized, then the eddy losses are neglectable [6]. Furthermore, hysteresis losses or core loss is the consumed power from altering the direction of the BH-curve, it is dependent on the core material and the frequency used. B represents the flux density and H is the magnetic field strength. Moreover, the dielectric losses can occur in the solid insulation. When the solid insulation takes its toll on the quality or gets damaged the efficiency of the transformers gets worse. Finally, stray losses are the stray inductance that occurs when leakage field is presence. Stray inductance losses are very small compared to the major losses hysteresis, and copper, it will be neglected [98]. In the following segment, the transformer power losses equations will be presented, where most of the heat is generated are from the transformer, transistors and diodes. The number of turns is obtained by first calculating the turns per volt that are needed Te = 1 4.44 × fsw × Bm × Ac [Turns/V olt] (4.24) where Te is 1 V ind , Vind is induced voltage. The 4.44 represents a sinusoidal waveform, Ac is the core area and the Bm is the maximum magnetic flux density [100]. Thus the number of turns in primary winding is given by N1 = Te × Vp (4.25) where Vp is the voltage in the primary winding. The skin depth (d) is calculated in order to see how large area of the wires should be designed when considering the skin effect. The skin depth is calculated by d = 1√ π × μ0 × σ × fsw (4.26) where μ0 is the permeability in vacuum and σ is the conductivity in copper [6]. The window area is the amount of space that the wireBundle can fit inside the core. This area is calculated as Aw = bw × hw (4.27) where width bw and height hw are selected from the datasheet. The maximum area that the copper wire was dimensioned for, is obtained with 39 4. Theory Acu,max = kcu × Aw Np × Ns (4.28) where kcu is the copper fill factor [6]. The copper wire area of the winding should be less than Acu,max. The core girth of the transformer is defined by Girth = Dcore × π (4.29) where the diameter Dcore is taken from datasheet. The length of the wire was calculated as L = Girth × (Np + Ns) (4.30) where L is the total wire length of the primary and secondary winding. The size of the transformer is calculated as V = l × b × h (4.31) where the transformer length, height, and width are from the datasheet. With an added I-core. When calculating losses, the focus is on the major losses of the transformer, lo- cated in the winding and the core. There are other losses as well, however, they are significantly lower in this case, that they are neglected. The losses are calculated by Ploss = Pw + Pcore (4.32) To calculate the winding losses Pw, the wire resistance is needed Rwire = ρcu × L Acu,bundle (4.33) where ρcu is the electrical resistivity of copper, L is the total wire length and the copper wire area of the bundle. While the Rdc can be known from the expression as follows Rdc = (MTL) × N × Rwire = girth × N × Rwire (4.34) where MTL is the mean turn length, which is assumed to be close to girth value [101]. N is both N1 and N2 together. With Rdc known, the winding loss can be calculated Pw = Rdc × I2 RMS (4.35) 40 4. Theory where IRMS is I√ 3 because of the triangular waveform [102]. To calculate the magnetic loss density PM , Steinmetz equation is implemented as PM = k × fa × Bd peak (4.36) where k, a and d are constants depending on what type of core material and frequency is used [98]. In the selected core, there is Ferrite material 3F3, thus the constants are k = 2.5 × 10−4, a=1.63, d=2.45 [6], [103], [104]. The volume of the core is calculated from and the values found in the datasheet [98]. Giving the volume Vcore = Acore × Height. (4.37) The core losses can be calculated by Pcore = PM × Vcore (4.38) which give the total losses Ploss = Pw + Pcore. (4.39) The efficiency is known as the standard equation η = Pout Pin (4.40) where the power input was calculated as Pout = Pin − Ploss. (4.41) Hence the new efficiency of the transformer is η = Pin − Ploss Pin . (4.42) Microcontroller unit The logic board consists of an LV board and DSP board, where the microcontroller unit (MCU) SPC560B54L3B4E0 is located on the LV-board. This controller is expected to generate highest losses in the low voltage system. Due to lack of infor- mation, the losses from the MCU needs to be estimated. The power dissipation for each pin can be roughly calculated with the equation follows as Ploss = V 2 × fsw × Cload (4.43) where it is known from data sheet that fsw = 279-285 kHz, supply voltage of V=5V and Cload= 3.8-5 pF [37], [105]. To obtain the complete losses of the MCU, a scaling factor is needed. This factor is assumed to be 2 for interconnection for pins [106]. 41 4. Theory This factor is multiplied with the losses for one pin, thus gives the total losses of the MCU [107]. The power losses for this controller is assumed to be the same for the controller in the off-board charger. 4.9.2 Thermal calculation and design When the total amount of losses in the circuit is known, the thermal resistance value can be calculated Rth,ja = Tj − Ta Ptot (4.44) where Tj is the maximum permissible operating temperature of the components according to their datasheets, Ptot is the total power dissipation and Ta is the highest assumed ambient temperature [6]. Since it is vital to analyze the temperature level of the components, it is desired to calculate the thermal resistance of a heatsink to ensure the temperature level is in a safe range. The total thermal resistance from the junction to ambient is presented as Rth,ja = Rth,jc + Rth,cs + Rth,sa (4.45) where the R values are representing the thermal resistance between the junction and the case, the case and the sink, the sink and the ambient temperature, and all together they make the thermal resistance between the junction and the ambient temperature. These thermal parameters can be represented as a thermal model of a heatsink shown in Figure 4.20 [6]. With the thermal resistance values from all the components separately are added together in parallel, to get one thermal resistance value as 1 Rth,sa,tot = 1 Rth,sa,1 + 1 Rth,sa,2 + ... + 1 Rth,sa,n . (4.46) In order to obtain the dimensions of a heatsink, the Rth,sa value needs to be known. The lower the Rth,sa value is, the larger heatsink will be required. In order to calculate the operating temperature for a