Veículo Elétrico Com Interface para A Rede Elétrica Grid Interface For An Electric Vehicle

Faculdade de Engenharia da Universidade do Porto
Veículo elétrico com interface para a rede

elétrica
Grid interface for an electric vehicle
Luís Miguel Faria Miranda
Dissertation carried out under the

Master in Electrical and Computers Engeneering
Branch: Automation
Supervisor: Prof. Doutor Rui Manuel Esteves Araújo
June 2011
© Luís Miguel Faria Miranda, 2011
Abstract
The field of isolated DC-DC converter is an important area of power electronics and it is
growing with electrical vehicles. The two main applications of these converters are related to
the need of isolation between the power electronics system and grid during the charging
period and the implementation of low voltage DC bus, used to supply auxiliary loads. This new
requirements lead to a demand for new conversion architectures and solutions, with high
values of efficiency and power density. These converters should avoid the usage of bulky low
frequency transformers in the near term for low and medium-low power chargers. Topologies
with soft-switching capability can be of particular interest for this application.
In this work is performed a study of the most common solutions for this purpose and a
selection is carried out based on previously defined criteria. The development of the power
converter is focused on the DC-DC stage whose topology is so called Dual Active Bridge. The
converter structure, operation modes and particular aspects are presented and discussed. To
conclude this analysis the converter was simulated in order to evaluate its performance. After
that the main components which constitute the converter are selected and the high
frequency power transformer is designed with special detail since it is the most critical part
of the power stage.
The systems for interface with the power stage were then developed and implemented,
i.e. a system to measure the variables and a system to act the converter according to the
phase-shift control method. With all components working correctly, several tests were
performed for modeling the power transformer and also in order to validate its operation as
Dual Active Bridge.
Resumo
O campo dos conversores DC-DC isolados é uma area importante da electrónica de

potência e está em crescimento com os veículos eléctricos. As duas principais aplicações
destes conversores estão relacionadas com a necessidade de isolamento entre o sistema
electronic de potência do carro e a rede eléctrica durante o periodo de carga e a
implementação de um barramento DC de baixa tensão, utilizado para alimentar as cargas
auxiliares. Estes novos requisites levam a uma procura por novas soluções e arquitecturas de
conversão, com elevados valores de rendimento e densidade de potência. Tais conversores
deverão evitar a utilização de volumosos transformadores de baixa frequência a curto prazo
para carregadores de baixa ou média baixa potência. As topologias com capacidade de soft-
switching podem ser de especial interesse para esta aplicação.
Neste trabalho é realizado um estudo das soluções mais comuns para esta aplicação e é
feita uma selecção baseada em critérios previamente definidos.O desenvolvimento do
conversor de potância é focado no andar DC-DC cuja topologia é chamade de Dual Active
Bridge. A estrutura do conversor, modos de funcionamento e particularidades serão
apresentadas e discutidas. Para concluir este análise, o conversor foi simulado com o
objvectivo de estudar o seu desempenho. Após isso o principais components que constituem o
conversor foram seleccionados e o transformador de alta frequência foi projectado de forma
mais detalhada visto que é o componente mais crítico do andar de potência.
Os sistemas para interface com o andar de potência foram então desenvolvidos, i.e. o
sistema para medição das variáveis e o sistema para comandar o conversor de acordo com o
método de controlo por esfasamento. Com todos os components a funcionar correctamente,
vários testes foram realizados para modelização do transformador de potência e também com
o objectivo de validar o funcionamento do conversor como Dual Active Bridge.
Agradecimentos
Em primeiro lugar, gostaria de agradecer ao meu orientador, Prof. Doutor Rui Araújo pela
confiança depositada nas minhas capacidades e por todo o seu empenho e disponibilade na
realização deste projecto e ao Engenheiro Ricardo Ferreira pela constante disponibilidade no
desenvolvimento das tarefas reportadas neste documento.
Ao INESC Porto por ter suportado a realização deste trabalho, e mais em particular à
equipa do projecto REIVE – Redes eléctricas inteligentes com veículos eléctricos com quem
trabalhei directamente durante este tempo.
Aos meus colegas de curso, em particular aos finalistas da especialização em Electronica e

Sistemas do major de Automação, com muitos dos quais partilho além do gosto pela àrea uma
grande amizade.
À Joana, cujo o amor, apoio e compreensão estiveram sempre presentes, principalmente

nos momentos mais dificeis.
E por último, mas de forma muito sincera, aos meus pais que sempre acreditaram em mim
e que tornaram possível concluir a minha frequência no Mestrado Integrado em Engenharia
Electrotécnica e de Computadores.
Contents
Abstract ...........................................................................................iii
Resumo ............................................................................................v
Agradecimentos ................................................................................ vii
Contents .......................................................................................... ix
List of Figures ................................................................................. xiii
List of Tables ................................................................................. xvii
Acronyms ....................................................................................... xxi
Chapter 1 ......................................................................................... 1
Introduction .....................................................................................................1
1.1 The Electric Vehicle ...................................................................................1
1.2 Motivations and Challenges in Electric Vehicle Integration ....................................3
1.3 The Vehicle-to-Grid (V2G) concept .................................................................4
1.4 Objectives of the work ...............................................................................5
1.5 Work plan ...............................................................................................5
1.6 Map of the document .................................................................................7
Chapter 2 ......................................................................................... 9
State of the art .................................................................................................9
2.1 Modern Electric Vehicles .............................................................................9
2.1.1 Storage systems .............................................................................. 10
2.1.2 Traction Motors ............................................................................... 11
2.1.3 Motor drives ................................................................................... 12
2.1.4 Auxiliary Loads Supply ....................................................................... 12
2.2 Electric Vehicles Market ............................................................................ 12
2.2.1 V2G Market forecast for US ................................................................. 14
2.3 Standards .............................................................................................. 15
2.4 Charging systems .................................................................................... 19
2.4.1 AC Charging Levels 1 and 2 ................................................................. 19
2.4.2 DC Fast Charging ............................................................................. 20
2.4.3 Commercial systems ......................................................................... 21
2.5 Systematization of power converters ............................................................ 22
2.6 Summary .............................................................................................. 32
Chapter 3 ........................................................................................ 33
Topology selection........................................................................................... 33
3.1 Basic topologies ...................................................................................... 33
3.1.1 AC-DC........................................................................................... 33
3.1.2 DC-DC .......................................................................................... 35
3.2 Association of AC-DC and DC-DC stages ......................................................... 37
3.3 Criteria ................................................................................................ 38
3.4 Comparative table ................................................................................... 40
3.5 Discussion ............................................................................................. 40
3.6 Proposed architecture .............................................................................. 41
3.7 Summary .............................................................................................. 41
Chapter 4 ........................................................................................ 43
Dual active bridge converter .............................................................................. 43
4.1 Introduction .......................................................................................... 43
4.2 Steady-state analysis ............................................................................... 45
4.3 Boundaries for zero-voltage switching ........................................................... 49
4.4 Simulation............................................................................................. 50
4.4.1 Model and parameters ...................................................................... 51
4.4.2 Simulation results ............................................................................ 53
4.4.3 Maximum leakage inductance estimation ............................................... 54
4.4.4 Capacitors dimensioning .................................................................... 57
4.5 Summary .............................................................................................. 59
Chapter 5 ........................................................................................ 61
Prototype design ............................................................................................. 61
5.1 Power switches ...................................................................................... 61
5.2 Drive circuit .......................................................................................... 63
5.2.1 Parameter setting ............................................................................ 65
5.3 Interface board ...................................................................................... 70
5.4 Capacitors............................................................................................. 72
5.4.1 Electrolytic capacitors ...................................................................... 73
5.4.2 Metalized Polypropylene Film Capacitors (MKP) ........................................ 74
5.5 Transformer design .................................................................................. 76
5.5.1 Core selection ................................................................................ 78
5.5.2 Number of turns .............................................................................. 79
5.5.3 Core losses .................................................................................... 81
5.5.4 Wire ............................................................................................ 81
5.5.5 Windings ....................................................................................... 83
5.5.6 Mean turn length ............................................................................. 85
5.5.7 Total wire length ............................................................................. 85
5.5.8 Wire resistance ............................................................................... 85
5.6 Summary .............................................................................................. 86
Chapter 6 ........................................................................................ 87
Control system and results ................................................................................. 87
6.1 Measurements ........................................................................................ 88
6.2 Acquisition ............................................................................................ 91
6.2.1 Error analysis.................................................................................. 93
6.3 Phase-Shifted square wave generation .......................................................... 94
6.4 Transformer parameterization .................................................................... 95
6.4.1 Transformer model .......................................................................... 95
6.4.2 DC winding resistance measurement ..................................................... 97
6.4.3 Short circuit test ............................................................................. 98
6.4.4 Open circuit test ............................................................................ 101
x
6.4.5 Model and validation ....................................................................... 103
6.5 Experimental results ............................................................................... 106
6.5.1 Transformer load tests ..................................................................... 106
6.5.2 Converter operation ........................................................................ 108
6.6 Summary ............................................................................................. 111
Chapter 7 ...................................................................................... 113

Conclusions and future work .............................................................................. 113
7.1 Conclusions .......................................................................................... 113
7.2 Future work .......................................................................................... 114
References .................................................................................... 115
Annex A – Workplan ......................................................................... 119
xi
List of Figures
Figure 1.1 Basic diagram of Plug-in Hybrid Electric Vehicle (PHEV) ..................................2
Figure 1.2 Basic diagram of Battery Electric Vehicle (BEV) ............................................2
Figure 1.3 EV interface evolution ...........................................................................4
Figure 1.4 V2G based system ................................................................................5
Figure 2.1 Conceptual architecture of a battery electric vehicle [8] .............................. 10
Figure 2.2 Annual global EV and PHEV sales in BLUE Map scenario [10] ........................... 13
Figure 2.3 Projected US Annual Sales of Light Duty PHEV/HEV’s [11] ............................. 13
Figure 2.4 Battery costs through 2020 [10] ............................................................. 14
Figure 2.5 US V2G Unit Forecast [11] .................................................................... 14
Figure 2.6 US V2G Vehicle Market Value Forecast (US billions) [11]................................ 15
Figure 2.7 Mennekes® connector drawing and pinout [20] ........................................... 18
Figure 2.8 Commercial charging station [24] ........................................................... 21
Figure 2.9 Residential EVSE [24] .......................................................................... 21
Figure 3.1 Full Bridge PWM rectifier/inverter .......................................................... 34
Figure 3.2 Single-phase multilevel rectifier/inverter ................................................. 35
Figure 3.3 Half-Bridge converter (Buck/Boost) ......................................................... 36
Figure 3.4 Dual active bridge converter [39] ........................................................... 37
Figure 3.5 Chart of alternatives for the converter architecture .................................... 38
Figure 3.6 System architecture ............................................................................ 42
Figure 4.1 Circuit schematic of a single phase dual active bridge dc-to-dc converter [39] .... 44
Figure 4.2 Fundamental model [41] ...................................................................... 44
Figure 4.3 Equivalent circuit 1 ............................................................................ 45

Figure 4.7 Fundamental waveforms [41] ................................................................ 47
Figure 4.8 Output power vs. phase-shift [42]........................ Error! Bookmark not defined.
Figure 4.9 Soft-switching boundaries [42] ............................ Error! Bookmark not defined.
Figure 4.10 Soft switching region vs. Phase-shift and M .............................................. 50
Figure 4.11 Simulation model ............................................................................. 51
Figure 4.12 Input bridge voltage, leakage inductance current and output bridge voltage in
boost mode (=45º) ................................................................................... 53
Figure 4.13 Input bridge voltage, leakage inductance current and output bridge voltage in
buck mode (=45º).................................................................................... 53
Figure 4.14 Output current vs. phase-shift ............................................................. 55
Figure 4.15 Battery SoC, Current and Voltage with Input 400V, Output 360V, of 90º Lσ =
1 µH ...................................................................................................... 55
Figure 4.16 Battery SoC, Current and Voltage with Input 400V, Output 400V, of 90º Lσ =
1 µH ...................................................................................................... 56
Figure 4.17 Battery SoC, Current and Voltage with Input 400V, Output 355V, of -90º Lσ =
1 µH ...................................................................................................... 56
Figure 4.18 Output current vs. Phase-shift - analytical and simulation results .................. 57
Figure 4.19 Output voltage ripple ........................................................................ 58
Figure 5.1 IGBT / MOSFET selection chart [43] ..................... Error! Bookmark not defined.
Figure 5.2 Considered equivalent circuit ................................................................ 68
Figure 5.3 5 to 15V circuit ................................................................................. 71
Figure 5.4 Internal equivalent circuit of HALT signal ................................................. 71
Figure 5.5 HALT signal interface circuit ................................................................. 72
Figure 5.6 Simplified equivalent circuit diagram of an electrolytic capacitor [49] .............. 72
Figure 5.7 Transformer simplified equivalent model (primary referred) .......................... 76
Figure 5.8 Transformer design flowchart ................................................................ 77
Figure 5.9 Voltage waveform applied to the transformer ............................................ 80
Figure 5.10 Coil former drawing .......................................................................... 83
Figure 5.11 Mean turn length identification ............................................................ 84
xiv
Figure 6.1 Power converter modules ..................................................................... 88
Figure 6.2 Multiplexed analog inputs ..................................................................... 91
Figure 6.3 Parallel analog inputs .......................................................................... 91
Figure 6.4 Phase-shift generator .......................................................................... 95
Figure 6.5 Practical transformer equivalent circuit (adapted from [57]).......................... 96
Figure 6.6 Practical transformer equivalent circuit (primary referred) (adapted from [57]) .. 96
Figure 6.7 Simplified transformer equivalent circuit (primary referred) (adapted from
[57]) ..................................................................................................... 97
Figure 6.8 Equivalent circuit for DC resistance measurement ....................................... 98
Figure 6.9 Equivalent circuit for shorted secondary ................................................... 99
Figure 6.10 Primary winding voltage and current with sorted secondary ......................... 99
Figure 6.11 Step response with shorted secondary. .................................................. 100
Figure 6.12 Equivalent circuit for open secondary ................................................... 101
Figure 6.13 Input voltage and current with open circuit secondary ............................... 101
Figure 6.14 Magnetization current identification (dashed) ......................................... 102
Figure 6.15 Slope estimation of magnetization current ............................................. 102
Figure 6.16 Transformer model with parameters ..................................................... 104
Figure 6.17 Transformer transfer function ............................................................. 104
Figure 6.18 Simulated current with shorted secondary .............................................. 105
Figure 6.19 Simulated magnetization current ......................................................... 105
Figure 6.20 Experimental setup for load tests......................................................... 106
Figure 6.21 Light load test (120 W) ...................................................................... 107
Figure 6.22 Heavy load test (1.5 kW) ................................................................... 107
Figure 6.23 Converter schematic with external inductor ............................................ 108
Figure 6.24 Transformer primary current =5º ........................................................ 110
Figure 6.25 Transformer primary current =10º ...................................................... 110
xv
List of Tables
Table 2.1 Main characteristics of common battery types [9] ........................................ 10
Table 2.2 NIST priority areas for standardization purpose [12] ..................................... 15
Table 2.3 SAE Standards [13] .............................................................................. 16
Table 2.4 SAE J2847 parts [13] ............................................................................ 16
Table 2.5 Off-Board Standards from UL [14] ............................................................ 17
Table 2.6 On-Board Standards from UL [14] ............................................................ 17
Table 2.7 IEC Standards [15-18] ........................................................................... 17
Table 2.8 Control pilot signal states [19] ................................................................ 18
Table 2.9 IEC 61851-1 Charging Modes [19] ............................................................. 19
Table 2.10 Yazaki single-phase AC charging connector specifications [22] ....................... 20
Table 2.11 Yazaki DC charging connector (CHAdeMO) ................................................ 21
Table 2.12 Main features of GE WattStation™ [25] .................................................... 22
Table 2.13 AeroVironment™ MODEL EV50-FS specifications [26] .................................... 22
Table 2.14 Desired features for V2G capable power conversion system ........................... 23
Table 3.1 Components list for rectifier/inverter ...................................................... 34
Table 3.2 Components list for multilevel rectifier/inverter ......................................... 35
Table 3.3 Voltage range for Buck/Boost ................................................................. 36
Table 3.4 Components list for Buck/Boost .............................................................. 36
Table 3.5 Voltage range for dual active bridge ........................................................ 37
Table 3.6 Components list for dual active bridge ...................................................... 37
Table 3.7 SUR calculation .................................................................................. 39
Table 3.8 Output voltage ranges .......................................................................... 39

Table 4.1 Leakage inductance voltages in each time instant ........................................ 47
Table 4.2 General simulation parameters ............................................................... 51
Table 4.3 Transformer parameters ....................................................................... 51
Table 4.4 Battery pack parameters ...................................................................... 52
Table 4.5 Considered scenario ............................................................................ 54
Table 4.6 Capacitor dimensioning ........................................................................ 58
Table 5.1 Input data for switch selection ............................................................... 62
Table 5.2 SKM100GB12V IGBT Half-Bridge main specifications [44] ................................ 63
Table 5.3 Main driver types comparison [45] ........................................................... 63
Table 5.4 Driver features................................................................................... 64
Table 5.5 Skyper 32 PRO R features [47] ................................................................ 65
Table 5.6 Minimum dead time estimation............................................................... 66
Table 5.7 Dynamic short circuit protection parameters .............................................. 68
Table 5.8 Input data for calculation [44] ........................................................... 69
Table 5.9 Gate resistors calculation ..................................................................... 69
Table 5.10 Electrolytic capacitors minimum requirements .......................................... 73
Table 5.11 Epcos B43456A9158M Electrolytic capacitor specifications [50] ...................... 74
Table 5.12 Epcos B32656S8155 MKP capacitor specifications........................................ 75
Table 5.13 Core selection guide for a specific usage [54]............................................ 78
Table 5.14 N27 UU 93/152/30 core magnetic characteristics [54] ................................. 79
Table 5.15 Primary voltage parameters ................................................................. 80
Table 5.16 Core losses estimation [54] .................................................................. 81
Table 5.17 Skin depth for Copper at 100ºC [37] ....................................................... 82
Table 5.18 Coil former characteristics [54] ............................................................. 83
Table 5.19 Turn segment dimensions .................................................................... 84
Table 5.20 Designed transformer full specifications .................................................. 86
Table 6.1 Measured variables and nominal ranges..................................................... 89
Table 6.2 LEM® LV 25-P voltage transducer data [55] ................................................ 89
Table 6.3 LEM® HY 15-P current transducer data [56] ............................................... 90
Table 6.4 Analog to digital converters specifications ................................................. 91
xviii
Table 6.5 Current acquisition parameterization ....................................................... 92
Table 6.6 Voltage acquisition parameterization ....................................................... 92
Table 6.7 Voltage measurement error ................................................................... 93
Table 6.8 Current measurement error ................................................................... 93
Table 6.10 Phase-shift generator resolution ............................................................ 95
Table 6.11 Transformer secondary to primary modification ......................................... 96
Table 6.12 Simplified transformer equivalent circuit parameters .................................. 97
Table 6.14 Results from DC winding resistance measurement ...................................... 98
Table 6.13 Parameters from the step response experiment ........................................ 100
Table 6.15 Peak values of magnetization current .................................................... 103
xix
Acronyms and symbols
AC Alternating current
ACIM Alternating current induction motor
BJT Bipolar junction transistor
BLDC Brushless direct current (motor)
BMS Battery management system
CAN Controller area network
CHAdeMO Charge de Move
DAB Dual active bridge
DC Direct current
DSP Digital signal processor
ESL Equivalent self inductance
ESR Equivalent series resistance
EV Electric vehicle
EVSE Electric vehicle supply equipment
FPGA Field programmable gate array
GE General electric
HF High frequency
IC Integrated circuit
ICE Internal combustion engine
IEA International energy agency
IEC International Electrotechnical Commission
IGBT Insulated gate bipolar transistor
LF Low frequency
MCU Microcontroller
MOSFET Metal–oxide–semiconductor field-effect transistor
NIST National Institute of Standards and Technology
NTC Negative temperature coefficient (thermistor)
PCB Printed circuit board
PHEV Plug-in electric vehicle
PLL Phase-locked loop
PMSM Permanent magnet synchronous motor
PTC Positive temperature coefficient (thermistor)
PWM Pulse width modulation
RMS Root mean square
SAE Society of automotive engineers
SoC State of charge
SUR Switch utilization ratio
UL Underwriters Laboratories
UPS Uninterruptible power supply
V2G Vehicle-to-grid
XOR Exclusive OR (logic gate)
ZVS Zero voltage switching
Phase-shift angle
Lσ Leakage inductance
Lm Magnetization inductance
RL Winding resistance
VDC1 DC voltage of the input bridge (left side)
VDC2 DC voltage of the output bridge (right side)
ω Angular frequency
I0 Average output current
P0 Average transferred power
td(on) Turn-on delay
td(off) Turn-off delay
tr Rise time
tf Fall time
VCEstat Collector-emitter threshold static monitoring voltage
VCEref Collector-emitter reference voltage
Tbl Blanking time
Cies Input to emitter capacity
Ls Equivalent series inductance
ψ Flux linkage
B Induction
Φ Magnetic flux
Magnetic relutance
δ Skin depth
le Effective magnetic path length
Ae Effective core cross section area
µa Amplitude permeability
Q Quantum
xxii
Chapter 1
Introduction
This chapter represents a first approach to the subject under study. Three subchapters
provide a successive approximation to the subject with a brief exposure of main parts.
At first, the basics about electric vehicles and the ones with interest for this work will be
presented. Those are battery electric vehicles and plug-in hybrid electric vehicles.
The second part will be focused on one of the major issues related to widespread of
electric vehicles – the grid interface. Some solutions will be presented and challenges will be
identified as motivation for the proposed work.
Lastly the most promising interface system (Vehicle-to-Grid) will be presented with a
rather deeper level of detail.
1.1 The Electric Vehicle
Along decades, cars with internal combustion engines (ICE) powered by fossil fuels have
been widely used by almost all transportation segments like family cars, trucks, bus and even
trains. This propulsion system found the success in a remarkable robustness and ease of refill
the energy source – fuel tank.
The concern to find alternatives existed since the beginning but little was done until a
few years ago. The conscience of sustainable growing and the approximation of the end of oil
in a few decades were the two main concerns that boosted the demand for new technologies.
Most promising technology for now and for the near future is surely electricity. The
knowledge gained over the years in electric power conversion (from and to electricity or
between electricity and another energy types) have provided valuable input for success of
Plug-In Hybrid Electric Vehicles and Battery Electric Vehicles.
Electric vehicles are usually divided in two main groups. The first one represents the Plug-
In Hybrids, cars that combine at least two propulsion systems one of them electrical.
These cars are not viewed as the future of mobility but only the transition between ICE
powered cars and electric vehicles. They are the current cost competitive solution to get
2 Introduction
better fuel economy, lower emissions and the ones that can achieve interesting ranges,
especially for huge and heavy cars [1].
The forthcoming solution is the other group, battery electrical vehicles. These cars are
designed to be only electrical-powered, with energy stored in batteries or any other electrical
storage system. One of the main advantages of electric cars is the fact that they can recharge
the battery during braking. Anyway, battery should not receive pulsed power from abrupt
braking what makes necessary the need for another energy storage system like super
capacitors or flywheels.
Figure 1.1 Basic diagram of Plug-in Hybrid Electric Vehicle (PHEV)
Figure 1.2 Basic diagram of Battery Electric Vehicle (BEV)
Most suitable electric motors for cars are AC induction motors (ACIM), synchronous motors
such as permanent magnet synchronous motors (PMSM) or brushless DC motors (BLDC) and
reluctance motors [2]. All of them have the great advantage of being brushless type requiring
minimal maintenance. Very advanced variable frequency drives (VFD) and control schemes
already developed for industrial applications are now used to improve performance from
these electrical machines, with high efficiency levels.
Overall the control chances related to this motors are very promising when compared with
ICEs. The ability to produce maximum torque in a wide speed range is a feature with special
interest for electrical vehicle (EV) applications. Another advantage is related with noise and
vibration generated by classic cars which are not present in modern EVs. Also in relation to
Motivations and Challenges in Electric Vehicle Integration 3
the traction system other advantages arises such as reduced external components needed
(like cables, motor drives, cooling system) and even the gearbox can be dispensed.
Lack of pollutant emissions is certainly one of the advantages of electric vehicles because
electricity is used only as mechanical power or wasted as heat.
Alongside these advantages other comfort-related can be found, whereas nowadays a car
is not just an engine and four wheels. Ease in feeding the growing vehicle auxiliary loads like
air conditioning or multimedia sets, electronic stability programs or traction control may be
other factors with high potential.
Anyway, the widespread use of such vehicles is still bound by some problems. The main
focus of research was around the storage system – batteries. They need to have high power
density, able to fast charge and an expected lifetime that is near the duration of an ICE car.
Expected range for cost competitive cars is still quite low which limits the utilization of EVs
only for short distance trips. The charging systems are not yet standardized as well as the
large-scale integration of this kind of vehicles with the existing power grids is still under
study.
1.2 Motivations and Challenges in Electric Vehicle Integration

The usage of electric vehicles brings naturally to a question – how to recharge battery?
People are used to recharge their mobile phones, notebooks and some other gadgets with no
problem. In fact these loads do not represent anything significant to the grid due to their low
stored energy.
With EVs this consideration cannot be done. Tens of kilowatt-hour batteries to be charged
in a few hours or even less than an hour represent large amount of power to recharge each
battery. With thousands of these loads coupled to grid serious problems would arise.
The results of such charging method are branch congestion and unacceptable voltage
drops in some branches due to an uncontrolled peak of consumption that coincides with the
time that everyone puts the vehicle to charge.
This problem creates a new research area and, above all, a new challenge in widespread
EV integration. In this scope the lowest level charging method was called dumb charging. This
method doesn’t allow large scale integration with the existing power grid, motivating the
appearance of increased intelligence systems such as smart charging or vehicle-to-grid (V2G)
interface.
These more advanced systems claims for new power conversions systems with
competence to exchange data with a centralized smart grid controller. With this system,
users can set the desired time for car to be charged and the controller will choose the
adequate period and possibly the charging ratio, therefore sends the command for the
charger. These functionalities should be available in a smart charge system.
The next level consists in a full-featured vehicle to grid interface. In this case, system
should be able to operate in inverter mode feeding power back to the grid. Proper control
scheme will adjust charging/discharging ratio with respect not only to battery requirements
but also grid constraints.
This feature can provide ancillary services like residential UPS system when Loss of Mains
occurs or helping in recovery from a failure.
4 Introduction
Figure 1.3 EV interface evolution
1.3 The Vehicle-to-Grid (V2G) concept

The V2G concept has been studied and can be seen as the future of charging systems. It is
applied to on-board chargers or charging stations. It can offer great benefits for electric grid
providing stability and load management but also for Plug-In owners.
Considering a first approach only grid would benefit from this strategy with owners being
affected with longer charging periods and wasted battery cycles when a discharge takes
place.
But on the other hand the vehicle owner can be able to be part of an open energy market,
buying or selling energy with very different tariffs. The economical point of view can make
this system attractive for users making money while the car is stopped.
Considering all these aspects, the system can become profitable to the owner and very
helpful to the grid. The already identified benefits considering electric power system point of
view are maximization of intermittent sources integration, reactive power compensation,
branch congestion avoidance, auxiliary frequency regulation and assistance in failure recovery
[3-5]. From the vehicle owner side some interesting opportunities have yet been identified
such as residential UPS system or potential revenue from electricity market [6-7].
The data exchange with a smart grid controller allows this system to have remote
operation and even as monitoring system sending local parameters back to smart grid
controller if the charger has the ability to send back data.
In the simplest case, the grid might just switch the vehicle charger on and off in response
to grid load. Grid could also send a message to the vehicle inhibiting it of charge during the
peak period, which can be related to technological aspects like branch congestion or
economical aspects like waiting for the cheaper tariff. In the most advanced scenario, grid
controller might setup a periodic data exchange of messages to the charger, setting charging
Objectives of the work 5
rate or even changing charger operation mode to flow of energy back into the grid. These
decisions can depend on several factors like statistical forecasts based on historical
information or real-time data including grid load, amount of renewable generation, state of
charge of the vehicle and energy pricing.
Figure 1.4 V2G based system
As seen so far this technology has a high potential representing a near term researching
area. A full-featured interface can be configured to provide several services with interest for
all stakeholders. Nevertheless some barriers like reliability and cost of overall communication
system and impact on vehicle battery lifetime throw the feasibility of this solution to a
medium term future.
1.4 Objectives of the work
The first objective for this work is to study relevant published knowledge about electric
vehicles and respective charging systems, especially cars with grid interface (Plug-In).
Next, a suitable power electronic conversion solution must be defined with appropriate
criteria in order to build a laboratory prototype. This hardware should allow getting valuable
data and establishing a comparison between the proposed system and other currently
available systems, which stands for the main objective of the work.
1.5 Work plan

To meet the objectives proposed for this work was developed a list of major tasks. These
tasks were divided into subtasks and then were scheduled in a Gantt chart (Annex A). The
justification and the results that are intended to achieve with each task are also presented.
6 Introduction
1. Analysis of project requirements. Selection of the power structure and operation

modes
This task has the main purpose of understand the main requirements that system should
have, to implement a vehicle-to-grid interface. When completed, it should result in a
technical specification of the power structure to be used, providing a comprehensive and
feasible solution. The modes of operation required should also be defined at this stage to be
contemplated in the prototype design.
2. Study, design and implementation of interface circuits for control and protection of
power semiconductors. Study of different technologies available to implement electrical
insulation.
At this stage should be considered different technologies to be used and chosen the most
suitable for the requirements of the project. After that shall be designed and built a robust
and modular hardware that allows the proper interconnection between the control and power
modules. The various protections should also be built and tested. The result should be a drive
and protection system that ensures a safe operation of all system components without
compromising their operation.
3. Study, design and implementation of the measurement system
The measurement systems are of crucial importance for the proper functioning of the
converter. Since it is a prototype, this system should be built in a modular way to ensure the
future possibility of reorganization. The resulting hardware should have characteristics that
do not compromise the closed-loop operation of the converter.
4. Assembly and testing of the power converter.
To accomplish this task the various power components must be assembled in a
comprehensible structure resulting in a user friendly experimental setup. User safety must be
ensured and it should be possible to collect data and make changes to the configuration
easily.
5. Study and implementation of a control system based on digital signal processors,
operating in real time.
The start of this task assumes that all the hardware is already assembled and tested. In
this stage should be done a research of various algorithms and computing techniques and
selected those with special interest for the controller to implement. After that, it should be a
study of the capabilities of digital control system and implemented closed loop control
scheme.
6. Exercise and develop the algorithms chosen based on numerical simulations.
The algorithms chosen should be implemented in numerical simulation and its results
should be analyzed to verify if they correspond to the expected. Once validated,
modifications can be made for the purpose of improving the performance of the control
system. The completion of this task should allow a transition to a physical implementation
with predictable results and stable operation.
7. Comparative testing between the developed algorithms and other technical
solutions based on conventional methodologies.
The purpose of this task is to assess if the developed or modified algorithms permit better
results than conventional or previously proposed. Performance measurements should be prior
established in order to evaluate different solution comparatively. The task should allow
produce valuable conclusions about this comparison.
Map of the document 7
8. Drafting of the dissertation.

Finishing the work, a detailed technical document should be produced with the aim of
structuring and presenting the developed knowledge so as to present the methodologies
adopted, the results achieved and relevant technical details.
1.6 Map of the document

This document is divided in seven chapters.
The first is a presentation of work. In it are first introduced the main concepts and
objectives / tasks of this dissertation. In this chapter there is room for a brief introduction
about the electric vehicle, for the presentation of the research motivations on grid
integration of electric vehicles and to introduce the vehicle-to-grid concept.
Chapter two starts with a brief description of modern electric vehicle key components
followed by an outer view of charging systems and its standards. Two commercial systems will
be presented with main characteristics listed and finally an analysis of power converters to
deal with vehicle-to-grid will take place.
In Chapter three are identified the basic topologies with interest to this application and
possible associations in multiple stages. Afterwards, criteria are defined and one of the
options is chosen. The proposed architecture for the system is then presented.
Chapter four is divided in two parts. The first aims to provide a description of the
topology to design in terms of its constitution and how it works. In the second part is
presented and described a simulation model for this converter and several simulations are
made with clear objectives.
In Chapter five are reported the main aspects concerning the design of the converter
power stage. The transformer design is presented with more detail, for being the most
critical component.
Chapter five can be split in two parts. The first presents the development of the
measurement and acquisition system and the converter command module. In the second part
experimental results are obtained with the aim of parameterize and test the transformer and
also results of the converter working as dual active bridge.
In the final chapter the main conclusions are taken and pointed orientations for future
developments.
Chapter 2
State of the art
This chapter can be split in three significant parts. First, a brief description of modern
electric vehicle key components will take place with some general remarks listed and the
statistical data about electric vehicles market focusing on V2G market forecast with a given
example for United States of America.
Second part will be focused in an outer view of charging systems with a short survey
considering the main standard with interest for this purpose such as safety and connection
requirements and after that the currently used charging levels will be described and a
discussion of their features, constraints and applicability will take place. Lastly two
commercial systems that together perform the three charging levels considered will be
presented with main characteristics listed.
Finally the third part will be based in an inner view of the charger, i.e. considering
electronics aspects only. A selection of recently published works matching this application
will be analyzed based on standardized table with brief appreciation for each one. Based on
these works will be performed an overall analysis identifying major trends and technological
options for topologies and another critical hardware with relevance for system design.
2.1 Modern Electric Vehicles

Prior to moving to a detailed discussion of the various power converter topologies it is
important to understand the reasons behind the ongoing development of electric vehicles.
In this context, the structure and key components of modern electric vehicles will be
presented. The discussion will focus on battery electric vehicles since they are the most
demanding in terms of performance and capabilities for all electrical components.
10 State of the art
Figure 2.1 Conceptual architecture of a battery electric vehicle [8]
2.1.1 Storage systems

The main power storage systems for full electric vehicles are batteries. For this
application several types of batteries can be used having rather different characteristics. In
any way the batteries are sensitive equipment and do not allow peak charging power that is
naturally generated by regenerative braking of vehicles. Due to this fact other technologies
emerged to deal with this situation such as ultra capacitors or flywheels [8].
Referenced in Table 2.1 are the main characteristics of a battery and its values for
different technologies. The theory associated with the batteries will not be presented
because it does not fall within the scope of this work.
Table 2.1 Main characteristics of common battery types [9]
Lead-Acid Ni-Cd Ni-MH Li ion
Cost Low Medium High Very High
Specific Energy (Wh.kg-1) 30 - 50 50 - 80 40 - 100 160
Voltage per cell 2 1,25 1,25 3,6
Charge current Low Very Low Moderate High
Cycle number (charge/discharge) 200 - 500 1000 1000 1200
Autodischarge per month (% of total) Low (5%) Moderate-High (20%) High (30%) Low (10%)
Minimum time for charge (h) 8 - 16 1 - 1,5 2-4 2-4
Activity requirement 180 days 30 days 90 days None
Environmental warning High High Low High
One of the critical issues related to battery usage is the charging process. The task of
charging only one cell individually would be easy since the state of charge was known. A
regulated power source with controlled output current or voltage can be used in order to
Modern Electric Vehicles 11
charge battery. In practice, battery packs are grouped in similar modules, with cells
connected in series to provide a proper voltage level for car DC bus. Although the current is
the same in all the cells, it does not mean that all were charged in the same way after a
certain time. Different constructive parameters or even differences in cell temperature
during charging can cause this situation. This phenomenon can also be observed in discharge
process.
In practice the present charging systems use a normal rate of charge until the best-
charged cells reach the full charge and thereafter switch to a reduced rate of charge mode
and go on until the worst-charged cells are estimated to be fully charged. This method usually
leads to a long time to go from near full charge to a real 100% state of charge in every cell
[8].
There are various alternatives of energy storage devices including the flywheel and super
capacitors. As a general rule both of these devices have increased specific power. However,
the amount of energy they can store is currently rather small.
Flywheels are devices that are used for storing energy. A plane disc spinning about its axis
would be an example of a simple flywheel. The kinetic energy of the spinning disc is released
when the flywheel slows down. Connecting an electrical generator directly to the disc can
capture the energy. The flywheel can be re-accelerated, acting as a regenerative brake.
An interesting alternative to flywheels is the use of large energy storing capacitors, so
called super capacitors. In many ways the characteristics of super capacitors are like those of
flywheels. They have relatively high specific power and relatively low specific energy. They
can be used as the energy storage for regenerative braking. Super capacitors are inherently
safer than flywheels as they avoid the problems of mechanical breakdown and gyroscopic
effects [8].
2.1.2 Traction Motors

Electric motor is obviously one of the key parts of electric vehicles. His characteristics
have serious influence on car weight, performance and reliability. The classic DC motor was
widely used in traction applications over decades, with many control opportunities especially
in separately excited configuration. The main problem associated with this motor is in its
principle of operation – the brushes-collector set. For intensive usage the need for regular
maintenance became a problem and other motor types emerged.
The alternative electrical machines used for electric vehicles are all brushless to not
compromise the vehicle overall reliability. High power density and robustness are another key
features pursuit for this application.
Two of them are very similar – the brushless DC and the permanent magnet synchronous
motor. Based on a permanent magnet rotor and a wounded stator these motors have high
power densities but also are more expensive than other options. The main difference
between them is the voltage waveform applied to the stator windings – trapezoidal in the
BLDC and sinusoidal in the PMSM. It is this fact that leads to the first being considered a DC
motor and the second an AC motor. Disregarding the cost this motors have been widely used
in modern all-electric and hybrid vehicles and represents a good solution for this purpose.
Another type of motor is called switched reluctance motor. This machine has basic
operation principle of minimize the air gap of a magnetic circuit. The rotor can be only
composed by a soft iron plate and two opposed stator windings. In practice a salient pole
rotor and multiple stator coils are used in order to attenuate the torque pulsations and the
12 State of the art
produced noise. The control of this motor is critical since it needs a measure or a precise
estimation of the rotor position to produce torque and control speed correctly.
At last, the well known alternate current induction motor (ACIM) must be referred. The
low cost and high reliability make this machine a very interesting solution in all kind of
industrial or residential applications. Originally designed to be feed by a three-phase system,
with the appearance of power electronic converters the application range has grown even
more and now the motor can be supplied by AC or DC source easily using proper electronic
drive [2, 8].
The control systems for this kind of motor are in a high state of development being this,
one of the main advantages for this alternative. Despite this, the motor design has special
design requirements for electric vehicle applications such as low weight and high power
density compared to industrial applications.
2.1.3 Motor drives

Several parts of electrical machines were originally proposed to be supplied by a specific
voltage system and many times at fixed speed. Early, mechanical systems were used to deal
with speed regulation needs. In the last decades modern automation systems such as robotics,
positioning systems etc have required precise speed regulated or torque regulated systems,
working in closed loop. The power converters used are known as variable frequency drives
and have the ability to deliver power at variable amplitude and variable frequency of
current/voltage. This type of control is now very usual on ACIM and PMSM.
The motor drive is another critical component of vehicles because the power stored in car
batteries are now very limited and the motor drive efficiency in both ways (acceleration or
regenerative braking) should be very high. The technology available today allows building
these drives with advanced control methods and power ratings in accordance with car needs
(50-150kW) with reasonable size, weight, volume and reliability.
2.1.4 Auxiliary Loads Supply
Power converters in electric vehicles could have many other applications apart from
battery charging or motor drive.
Currently high power density converters are studied to deal with this task. Loads like
advanced multimedia systems, air conditioning, sunroof, electric windows, heating seats,
cooling system pump, power steering and many other represents together a large amount of
power and its operating voltage is by far lower than car DC bus. Some of these loads are also
very sensitive, which leads to a converter with high power rating and well regulated output.
These converters are usually DC-DC converters that create a low voltage DC bus from the car
main bus, sometimes with electrical isolation due to safety aspects.
2.2 Electric Vehicles Market

So far, electric cars have not yet gained a significant market share. Despite the incentives
of some governments, the price, range and uncertainty about the durability of the batteries
fail to convince buyers. IEA has defined a technology roadmap for upcoming years with sales
Electric Vehicles Market 13
forecast to next 40 years in a BLUE Map scenario for light duty vehicles. It is expected that
sales will hit a substantially volume in 2015 and began to grow rapidly from there.
Figure 2.2 Annual global EV and PHEV sales in BLUE Map scenario [10]
According to the US Energy Information Agency‘s Annual Energy Outlook 2010, annual
annu unit
sales of hybrid electric vehicles (HEVs) and plug-in
plug in hybrid electric vehicles (PHEVs) are
expected to grow from 287,000 in 2010 to 2.2 million in 2035.
This represents a annual growth rate of 14.7% during this time period. The Annual Energy
Outlook
ok 2010 expects that PHEV/EV unit sales represent 2.8 percent of all light duty vehicle
(LDV) sales in 2010 and 11.6 percent in 2035. [11]
Figure 2.3
3 Projected US Annual Sales of Light Duty PHEV/HEV’s [11]
IEA have already identified the high battery packs cost to be the main obstacle for a
large-scale
scale appearance of these cars on the streets. Due this medium term target prices have
been established in order to produce large battery packs with affordable cost. The
achievement of these targets must enable to produce cost-competitive
cost competitive vehicles with ranges
similar to those non electric vehicles on road today.
14 State of the art
Figure 2.4 Battery costs through 2020 [10]
2.2.1 V2G Market forecast for US
Important data was recently published in a report made by Zpryme, where a market
forecast
ast for 2020 related to V2G was presented. United States, China, Germany, UK, South
Korea and Denmark are the countries in this report.
United States market was presented has an example of V2G market potential.
Figure 2.5 US V2G Unit Forecast [11]
Apart from the fact that being only forecasts the presented data with just a medium term
range (ten years) should be able to make V2G an emerging market with significant value as
can be seen in Figure 2.6 where within just ten years (2020) representing a market value of
8.2 billion dollars. [11]
Standards 15
Figure 2.6
6 US V2G Vehicle Market Value Forecast (US billions) [11]
The above data should be able to encourage major players in areas such as power
electronics and communication systems to increase their investments in this technology in
coming years.
2.3 Standards
Standardization always represents a first step to guarantee the interoperability of whole

systems and so the integration of plug-in
plug vehicles onto the grid. Several organizations have
been working on standards for electric vehicles.
Published in January 2010 by National Institute of Standards and Technology (NIST), the
NIST Framework and Roadmap for Smart Grid Interoperability Standards set the priorities for
standardization (Table 2.2
2).
Table 2.2
2 NIST priority areas for standardization purpose [12]
Demand Response and Consumer Energy Efficiency
Wide-Area Situational Awareness
Energy Storage
Electric Transportation
Advanced Metering Infrastructure
Distribution Grid Management
Cyber Security
Network Communications
The main organizations focused on writing standards for electric vehicles are the Society
of Automotive Engineers (SAE),
(SAE) the International Electrotechnical Commission (IEC) and the
Underwriters Laboratories (UL).
16 State of the art
SAE Standards
One of the most popular standards is the SAE J1772 standard for conductive charge
coupler which is now supported by GM, Chrysler, Ford, Toyota, Honda, Nissan, and Tesla
Motors [12]. This standard allows the appearance of normalized connector which is now
supplied by several manufacturers and can be easily found in commercial electric vehicles.
The SAE organization is working on other relevant standards (Table 2.3).
Table 2.3 SAE Standards [13]
J2293/1 & /2 Energy Transfer System for Electric Vehicles
Part 1 Functional Requirements and System Architectures
Part 2 Communication Requirements and Network Architecture
J2836 Use Cases & General Information
J2847 Detailed information (messages, state diagrams, etc.)
The J2847 standard is of great relevance for the purpose of this work, especially parts 1
and 3 (Table 2.4).
Table 2.4 SAE J2847 parts [13]
J2847/1 Communication between Plug-in Vehicles and the Utility Grid
J2847/2 Communication between Plug-in Vehicles and off-board DC Chargers
Communication between Plug-in Vehicles and the Utility Grid for Reverse
J2847/3
Power Flow
J2847/4 Diagnostic Communication for Plug-in Vehicles
J2847/5 Communication between Plug-in Vehicles and their customers
In January 2011 only the part 1 of the SAE J2847 standard was published with the
remaining parts reported as work in progress status, according to the SAE international
website.
UL Standards
In this context, UL standards need to be referred too. The specifications are split in Off-
Board and On-Board related. They are essentially safety standards for connectors, cables,
power converters, traction motors, chargers and batteries.
Standards 17
Table 2.5 Off-Board Standards from UL [14]
UL 62 Standard for Safety of Electric Vehicle Cable
UL 2202 Standard for Safety of Electric Vehicle (EV) Charging System Equipment
UL 2231 Standard for Safety of Personnel Protection Systems for EV Supply Circuits
UL 2251 Standard for Safety of Plugs, Receptacles, and Couplers for EVs
UL Subject 2594 Subject Standard for Safety of Electric Vehicle (EV) Supply Equipment
UL Subject 2735 Subject Standard for Safety of Electric Utility (Smart) Meters
Table 2.6 On-Board Standards from UL [14]
UL Subject 458A Subject Standard for Safety of Power Converters/Inverters for Electric
Land Vehicles
Subject Standard for Safety of On-board Electric Vehicle Equipment

UL Subject 1004-1
Traction Motors
UL 2202 Standard for Safety of Electric Vehicle (EV) Charging System Equipment
UL Subject 2733 Subject Standard for Safety of Surface Vehicle On-Board Cable
Subject Standard for Safety of Connectors for Use with On-Board Electrical
UL Subject 2734
Vehicle (EV) Charging Systems
IEC standards
At last the well-known International Electrotechnical Commission (IEC) also published the
some standards for electric vehicles. The most relevant standards for this work are the ones
related to charging system, especially conductive charging.
The first standard is the IEC 61851/part 1 and describes the general requirements for on-
board and off-board equipment used to charge electric road vehicles with up to 1000 V a.c. or
up to 1500 V d.c.. Part 21 gives the electric vehicle requirements for conductive connection
to a a.c. supply up to 690 V or a d.c. supply up to 1000 V. Part 22 gives the requirements for
a.c. electric vehicle charging stations for conductive connection with a.c. supply voltages up
to 690 V. [15-18]
Table 2.7 IEC Standards [15-18]
IEC 61851-1 Electric vehicle conductive charging system – Part 1: General requirements
Electric vehicle conductive charging system – Part 21: Electric vehicle

IEC 61851-21 requirements for conductive connection to an a.c./d.c. supply
Electric vehicle conductive charging system – Part 22: AC electric vehicle

IEC 61851-22
charging station
Plugs, socket-outlets, vehicle couplers and vehicle inlets – Conductive charging

IEC 62196-1 of electric vehicles – Part 1: Charging of electric vehicles up to 250 A a.c. and
400 A d.c.
18 State of the art
The IEC 62196/part 1 is applicable to plugs, socket-outlets,

socket outlets, connectors, inlets and cable
assemblies for electric vehicles, intended for use in conductive charging systems which
incorporate control means, with a rated operating voltage not exceeding [18]:
– 690 V a.c., 50 – 60 Hz, at a rated current not exceeding 250 A;

– 600 V d.c., at a rated current not exceeding 400 A.
A
A connector/inlet set according to the IEC 62192-1 was proposed by MENNEKES® (Figure
2.7). The connector is suitable to supply on-board chargers with three-phase
phase or single-phase
single
input. It is composed by 5 power pins (3 phases, neutral and protective earth) and two
additional control pins. These two control pins are the “plug present” pin which immobilizes
the vehicle when the connector is attached and the “control pilot”. The communication
through this pilot pin is used to exchange information between the electric vehicle on-board
on
charger and the EVSE. The physical signal is pulse width modulated with a frequency of 1 kHz.
IEC 61851-1
1 standard defined six distinct states depending on the average value of this signal.
sig
Table 2.8 Control pilot signal states [19]
Vehicle state Voltage (avg./DC Description of vehicle state

designation level)
State A 12.0 ( a ) Vehicle not connected
State B 9.0 ( b ) Vehicle connected / not ready to accept energy
State C 6.0 ( b ) Vehicle connected / ready to accept energy / indoor charging

area ventilation not required
State D 3.0 ( b ) Vehicle connected / ready to accept energy / indoor charging

area ventilation required
State E 0 EVSE disconnected, utility power not available, or other EVSE

problem
State F -12.0 ( a ) EVSE not available, or other EVSE problem
Notes:
a) Static voltage
b) Positive portion of 1 kHz square wave, measured after transition has fully settled
Figure 2.7 Mennekes®

Mennekes connector drawing and pinout [20]
Charging systems 19
2.4 Charging systems
Due to its limited range the electric vehicle charge is a serious concern for users.
Desirable features for this task are a fast charge, to preserve battery lifetime and have the
best tariff for energy. In fact, several options are currently offered by car manufacturers but
often by third-party companies such as General Electric.
In this context the system that stands between the car and the power grid is called
electric vehicle supply equipment (EVSE).
Table 2.9 IEC 61851-1 Charging Modes [19]
Maximum voltage (V)

Maximum
Level/Mode Single- Three- DC Earth conductor
current (A)
phase phase
1 16 250 480 - Required
2 32 250 480 -
connection of the EV to the a.c. supply network (mains) utilizing dedicated
3 EVSE where the control pilot function extends to control equipment in the
EVSE, permanently connected to the a.c. supply network (mains)
connection of the EV to the a.c. supply network (mains) utilizing an off-board
4 charger where the control pilot function extends to equipment permanently
connected to the a.c. supply.
2.4.1 AC Charging Levels 1 and 2
The first and at the same time less powerful systems are tagged as Level 1 charging
systems, according to IEC 61851-1. This solution provides an up to 250V AC single-phase or
480V three-phase connection with current up to 16 A. The power rating is insufficient
especially for 100% electric vehicles since nowadays battery packs have tens of kWh the
charging time will be unacceptable. It is a good solution for cases of appeal, neighborhood
vehicles, two-wheelers and hybrids with battery packs up to 10 kWh.
These EVSE systems can be supplied on the simplest form being just a cable with a
connector on car side and a standard outlet on the other plus an additional very simple pilot
circuit with specific safety/protection hardware between two connectors.
Single-phase connectors
The currently used connector (known as SAE J1172 connector) for single-phase charging is
available from some major suppliers in the specific area of vehicle connectors and cabling.
For instance, Yazaki corporation proposes a product rated up to 30 A (Table 2.10). A similar
1-Phase Connector is offered by REMA with 16 A and 30 A versions. As can be seen these
connectors cover the AC Levels 1 and 2 full power range [21-22].
20 State of the art
Table 2.10 Yazaki single-phase AC charging connector specifications [22]
Certification authority JET / UL / TÜV Rheinland
Certification standard Electric Appliances and Material Safety Law / UL2251 /

IEC62196-1
Rated voltage 100 to 250V AC 120/240V AC
Rated current 20A (JPN/EU) / 30A (JPN) 20A / 30A (USA)
Cabtyre cable POWER 2, GROUND 1, SIGNAL 1
Operation temperature -40° to +60°

range
Reference charging time 200V AC, 15A : 7h 100V AC, 15A : 14h
(100%) * In case of a standard electric car
Three-phase connectors
The use of on-board chargers is not limited to single-phase inputs. IEC 61851 standard
predicts three-phase connections for AC charging levels 1 and 2. As seen so for a connector
was proposed by Mennekes® according to this standard. Other manufacturers, such as REMA®,
currently offer this connector. This solution may be of limited utility in a near term as three-
phase outlets are more difficult to find than the single-phase outlets. On the other hand it is
expected that with widespread use of electric vehicles, home or office parking places have
available three-phase connections or even EVSE equipments.
There are no major differences between the single and three-phase connectors, apart
from the two additional pins used to connect power lines. Connectors of such type are
available with current ratings up to 63 A and usually split in three options: 16 A, 32 A and 63
A. [20-21]
2.4.2 DC Fast Charging
As seen the previous two levels have the capacity to meet different charging needs.
Nevertheless these two levels do not meet all the user needs. Imagine if a trip away in the
distance is beyond the range of the vehicle in which case you need to charge extremely
quickly at least a significant portion of the battery. This failure to deliver enough power led
to the emergence of the DC fast charge.
This is only possible using an external battery charger connected to the battery through a
larger connector that came up together with the charging method presented by CHAdeMO.
This Japanese association is formed by The Tokyo Electric Power Company, Nissan, Mitsubishi,
Fuji Heavy Industries (the manufacturer of Subaru vehicles) and Toyota. Charging stations
with the availability of DC charging currently in use provide CHAdeMO compliant connectors
as well as the newer vehicles provide these connectors too. This connector is now available
from industry-leader Yazaki Corporation.
Given that in this mode the charger is out of the vehicle, the on-board battery
management system (BMS) should establish communication with the charger in order to
transmit to it the required parameters for the charging process. This communication can be
Charging systems 21
performed by both analog signals and digital signal using CAN. Each connector has four analog
signal pins, two CAN digital
al pins and a ground pin. [23]
Table 2.11 Yazaki DC charging connector (CHAdeMO)
Certification authority UL / TÜV Rheinland
Certification standard UL2251 / IEC62196-1
Rated voltage 500V DC
Rated current 120A
Cabtyre cable POWER 2, SIGNAL 7, LED power supply 2
Operation temperature range -40° to +60°
Reference charging time (100%) 500V DC, 120A : 30min

* In case of a standard electric vehicle
2.4.3 Commercial systems

Several players from electric/electronic systems market have already purposed their
solutions for electric vehicle charging. With WattStation™ trademark General Electric offers a
range of chargers including charging stations and public EVSEs residential.
Figure 2.8 Commercial charging station [24] Figure 2.9 Residential EVSE [24]
These charging system are designed for Level 2 charge including a conductive coupled in
accordance to SAE J1772 intended to be the commonly accepted solution for Level 1 and
Level 2 AC charging. The built-in
built in bidirectional communication system enables this system to
be integrated in a Smart Grid infrastructure.
22 State of the art
Table 2.12 Main features of GE WattStation™ [25]
UL Certified
Charger and vehicle communication in accordance with NEC 625
Cord connection in accordance with SAE J1772
Indoor and outdoor enclosures
Wall and pedestal mount
Commerce and card reader capability
User authentication
Network communications
Revenue grade metering with AMI communications
LED lights and display
Level 2 (208-240 VAC) charging capability
This proposal from General Electric fails to include even the capacity for DC fast charging.
Other manufacturers like AeroVironment™ offer more powerful charging stations with power
ratings of 30, 50, 60, 125 and 250 kW with public and fleet types. CHAdeMO association have
certified until now 17 charging stations from multiple vendors. [23] One of these certified
stations is the (fleet type) with a power rating up to 50 kW (Table 2.1).
Table 2.13 AeroVironment™ MODEL EV50-FS specifications [26]
Power Rating 50kW
Power Factor 0.95
Input Power 480V, 3 Phase AC (Other inputs available)
Input Current 68A
Frequency 50Hz / 60Hz
Efficiency Rating >90%
Max. Output DC Current 120A
Max. Output DC Voltage 50-500V
Voltage Accuracy 1%
Current Accuracy (Lesser of the two) .5A or 5%
A remarkable case study can be found in Portugal. The MOBI.E project is implementing a
national charging network which provides all the services to allow the utilization of EVs. Users
can subscribe this network and get access to 118 charging points throughout country.
Systematization of power converters 23
2.5 Systematization of power converters

Prior to a first analyze of power converters do deal with V2G task the definition of some
requirements should be done. These minimal requirements shouldn’t be seen as constraints
on the bellow presented research but rather as suitability measurements of the various
alternatives.
Table 2.14 Desired features for V2G capable power conversion system
ID Feature Relevance
1 Bidirectional power flow Needed
2 Close to unity power factor in rectifier or inverter mode Desirable
3 Insulation between grid and car battery Needed
4 Current control mode for battery charging Desirable
5 Synchronous operation in inverter mode Needed
For this purpose a standard table was created in order to bring a better overview of all
topologies and especially to provide a comparative analysis in an easy way.
Alongside with the verification of the previous features in each topology some other
parameters are evaluated such as power and control scheme (if available), insulation method,
the author’s own conclusions about the work, control hardware used and the provenance of
the results - analytical, numerical simulations or prototypes.
At the end of each table a brief analysis about each converter takes place.
24 State of the art
Author and year Weise, N.D., K.K. Mohapatra, and N. Mohan, 2010
[27]
Power schematic
AC-AC Stage Single/Three Phase matrix converter
AC-DC Stage Full bridge synchronous rectifier
Results provenance (s) Simulation
Insulation method High Frequency
Control scheme
Control platform/hardware NA
Implemented features (1)(2)(3)(4)(5)
Approximated models used are very acceptable, with

errors studied and observed to be small.
Conclusions of the author (s) All the V2G requirements are satisfied.
Usable with single or three phase source.
Hardware being built to compare results.
Summary
This solution is surely one of the most suitable for this application having potential to
reach high power density levels with HF insulation but practical results awaited to
proof the feasibility of this solution.
Low order harmonics present in current drawn from the source still need to be studied.
Interesting approach with control schemes and PWM generation done with low
computational costs.
Author and year Bilgin, B., A. Emadi, and M. Krishnamurthy, 2010

[28]
Power schematic
AC-DC Stage Single phase PWM rectifier/inverter
DC-DC Stage Bidirectional Buck/Boost
Insulation method None
Control scheme
Implemented features (2)(4)
Tests with maximum and minimum inputs (250V/50Hz

Conclusions of the author (s)
and 90V/60Hz) shows good performance
Summary
Critical passive components have been designed based on analytical background.

Design of feedback controller was done carrying the influence of DC bus capacitor and
valuable observations relating feedback parameters and harmonic distortion generated
and output current ripple.
26 State of the art
Author and year Rei, R.J., et al., 2010

[29]
Power schematic
AC-DC Stage Three-phase three-level PWM rectifier/inverter
DC-DC Stage None
Insulation method Low Frequency
Control scheme
Implemented features (1)(2)(3)(4)(5)
Technical feasibility of charging strategies confirmed

Conclusions of the author (s) Developed a method handling coexistence of voltage
and frequency droops
Summary
Interesting topology but voltage output with limited range due to inexistence of DC-DC
stage. Suggested low frequency insulation transformer can create a bulky final system.
All features of the proposed control system validated with good results. Local grid
condition estimation seems to be a promising short term solution for V2G feasibility.
Author and year Grenier, M., M.G. Hosseini Aghdam, and T.

Thiringer, 2010
[30]
Power schematic
AC-DC Stage PWM rectifier/inverter
DC-DC Stage Bidirectional Full-bridge
Insulation method High Frequency
Control scheme
NA
Control platform/hardware FPGA
Implemented features (1)(2)(3)(4)
Better efficiency with duty cycle control than phase

shift control
Conclusions of the author (s) Efficiency increases as battery voltage increases
Opting for an HEV could allow savings of 51% in
energy costs
Summary
Significant data about efficiency study demonstrating linear relationship between

drawn current and power losses.
Estimation of weight and volume for the converter resulting in 4.75 Kg and 2.70 liter.
Expected power density of 1.37 kW/liter with power rating considered in this paper
Lack of details about resulting final system and also about control implementation
28 State of the art
Author and year Zhou, X., et al., 2009

[31]
Power schematic
AC-DC Stage single-phase PWM rectifier/inverter
DC-DC Stage Not specified
Control scheme
Implemented features (1)(2)(5)
The PR and selective harmonics compensation

method achieves good rejection of dominant
harmonics in V2G mode. Capacitor current feedback
method and PR controller guarantee low THD of
output voltage for different types of loads in V2H
mode. The proposed converter can greatly improve
the performance of PHEV’s integration with power
grid.
Summary
Remarkable results with the proposed proportional resonant + selective harmonic

compensation controller when compared to classic PI
Good study about harmonic distortion in charger and V2G mode
Proposed vehicle-to-home operation mode (residential UPS) working as sinusoidal
voltage source tested with several types of loads achieving good results due to
capacitor current feedback strategy adopted
Author and year Xiaohuz, Z., et al., 2009

[32]
Power schematic
AC-DC Stage Single phase PWM rectifier/inverter
Results provenance (s) Simulation and experimental setup
Control scheme
Control platform/hardware TMS320F28335
Implemented features (1)(2)(5)
Simulation and experimental results verify the

functions and performance of the proposed charger.
A household based PHEV charger can be used to
Conclusions of the author (s) implement multiple functions including battery
charging, vehicle to grid and vehicle to home, which
make the proposed charger promising for PHEV
applications.
Summary
Lack of galvanic insulation studied and interesting techniques are presented to deal with
possible safety issues. The small-scale prototype enables feasibility proof about the
proposed topology and control strategies and acquisition of valuable data such as drawn
current waveform and low order harmonics magnitude.
30 State of the art
Author and year Jaganathan, S. and G. Wenzhong., 2009

[33]
Power schematic
AC-DC Stage Single-phase three-level PWM rectifier/inverter
Insulation method
Control scheme
Implemented features (1)(2)(4)(5)
The converter draws unity power factor from the

grid. Also the harmonics of the system has shown to
be greatly reduced and the charging system is shown
to be utility friendly.
Summary
Option for a multilevel NPC resulting in low THD but practical implementation issues
ignored. PI based control with detailed description of parameter calculation and
converter model determination. Synchronization method not described.
The presented tables describe recently published work in charging systems technology for
electric vehicles. Many options were discarded for not being minimally framed in the basic
requirements defined earlier. Several simulation works has been presented recently but there
are no many published experimental results. Some aspects like grid synchronization methods,
passive components design and physical parameters like weight, size or even thermal
management systems are still currently ignored by most authors. Apart from this some
technological options start to become usual.
In AC-DC conversion is agreed the adoption of a force commutated rectifier with modern
switches (especially IGBTs) and some PWM command technique. The variations are between
single-phase and three-phase, with some alternative solutions for multilevel. If the topology
does not expect much discussion, the control opens up many possibilities but always based on
digital systems. The usage of controlled grid-connected systems implies the need for
synchronization information like instant grid angle (Θ) commonly required for rotating frame
based control systems or synchronized PWM generators. The most promising solution is the
phase locked loop (PLL) block, prior known from analog communication systems. This method
has become interesting for the control of synchronous electronic converters especially those
based on digital control.
Another relevant issue is the passive component design. Some published works presents
mathematical expressions essentially for line current filter inductor and DC bus filter
capacitor and explicit their influence for system behavior. The signal acquisition implies some
design cares but is not a particularity of this application and is not usually referred in this
scope.
The analyzed solutions in some cases are built with just a single conversion stage but
important features like galvanic insulation or battery pack voltage level matching are not
possible unless a second DC-DC conversion stage is added. The discussion about the topology
to use on this stage is still very open.
The simplest and more obvious option is the bidirectional buck/boost converter. This
converter has a simple operation with only two controlled switches and an output voltage
range between zero and DC bus voltage. Despite the continuous output current a large
amount of power drawn from this leads to a high current ripple which could not be supported
by batteries. This ripple needs to be attenuated with a filter inductor which certainly
represents a large volume due to winding with large wire and the need for a huge core to
handle magnetic saturation effect. In reverse mode the circuit operates as boost converter
ensuring proper DC bus voltage to feed the AC-DC stage functioning as inverter.
Anyway the insulation requirement cannot be fulfilled with this solution. Considering this
feature required prior to converter design the demand for a new topology has to be
considered limiting the search to insulated DC-DC converters. Such converters are always
based on magnetic elements (transformers) used to transfer power from one side to another
through a magnetic circuit. Even with the need for bi-directionality options might include
choice of half or full-bridge with several options for the control system such as controlled or
uncontrolled rectification.
No relevant work has yet been proposed considering such insulated converters possibly
due to a high level of abstraction used by the major part of authors when the feature is
suggested to be implemented with a grid frequency transformer which not represents a
practical solution to a physical system considering the large size and weight involved.
32 State of the art
Apart from the conventional topologies a more sophisticated system using a AC-AC matrix
converter was proposed in [27] to match the insulation requirement through a high frequency
transformer with output rectified by a controlled full-bridge. The option for this converter
will not be considered in this work due to control complexity and to not compromise the
feasibility of the purposed work.
The main above referred alternatives are classified by [34] split in AC-DC stages ranging
from single-phase half-bridge to three-phase multilevel full-bridge and DC-DC stages including
dual active bridge, conventional bidirectional buck-boost and some other modified buck-boost
designs.
An additional discussion is focused in which hardware should be used for power circuit -
dedicated switches or the integration with motor drive using the existing power converter. It
could be seen as a solution to reduce costs taking advantage from the high performance and
power rating typical in such converters. In some approaches the motor windings are also
properly arranged to perform a powerful input current filter or even used as power
transformer providing insulation to charging system through some modifications in motor
construction. The switching between charging/motor drive operation modes can be done
easily through relays (solid state or mechanical) [35-36].
2.6 Summary
In this chapter the basic structure and components of modern electric vehicles have been
described. The standards were also listed as they are essential in the evolution of electric
vehicles, especially in charging systems.
After it, there were presented currently available charging systems. In the end it was
analyzed recently published articles relating to electronic on-board charging, concluding with
a discussion of the conversion topologies that have been used by several authors to the most
relevant trends.
Chapter 3
Topology selection
The purpose of this chapter is to do a comparative analysis of suitable power converter

structures to deal with vehicle-to-grid technology in the future. Due to the high amount of
recent proposals and possibilities, an in-depth review of the state-of-art in bidirectional
charger system was performed in the previous chapter.
After a survey of more appropriate solutions and analysis of current trends, is now
important to identify the basic topologies of interest to this application and identify possible
associations in multiple stages.
For the purpose of this work will be considered a single-phase connection to the grid.
This restriction should not make the project less valuable since all the grid connected
converters can be found in single and three phase configurations.
3.1 Basic topologies

3.1.1 AC-DC
Single Level
Surely the simplest way to make a bidirectional connection with grid is through a full
bridge. This topology is far more interesting when compared with the conventional line
commutated rectifiers due to its ability to shape a sinusoidal current waveform on grid
interface and impose a unitary power factor.
Besides the active components (switches) in its simplest configuration also includes a
filter on the DC side voltage generally consisting of one or more capacitors in parallel and a
filter on the AC current, which can be a single inductor in its simplest version or a tuned LCL
filter, with particular interest to major powers where the inductor size is very critical.
One of its main limitations is to have a minimum threshold for the DC bus voltage as a
function of input voltage. This way can only be used as battery charger with high rated
voltage batteries or there is a need to add another converter with voltage step down
characteristic.
34 Topology selection
Figure 3.1 Full Bridge PWM rectifier/inverter
Table 3.1 Components list for rectifier/inverter
Component Quantity Observations
Switch 4 Fully-controlled
Inductor 1 Large for high currents or low switching frequencies
Large for a DC bus with low ripple
Capacitor 1
Many times split in series/parallel associations
The control of this converter is also strongly studied and described in literature [37-38].
The vector control techniques based on rotational frame are the most sophisticated at this
level although the computational resources used nowadays for the control of power
electronics can already perform it with relative efficiency.
Multi Level
In systems with considerable power where AC power quality is a relevant factor, the
option for a multilevel topology can be considered interesting. Multilevel converters make use
of switches associated in series that allow beyond the usual levels + VDC,-VDC and 0, put on the
AC terminals other intermediate voltage levels, thus making it easier to shape a sinusoidal
waveform and reducing the harmonic content for a similar conditions of operation when
compared with single level configuration. These intermediate voltage levels can be produced
by voltage sources isolated from each other or by series association of capacitors in the DC
bus.
Their main difficulties are: first, the added complexity in the control circuit due to the
need of twice the control signals and a more complex interlock system and the second relates
to the balancing of voltages in capacitors responsible for generating the intermediate voltage
levels. This management is usually done through the redundant states so that for long periods
of time the energy extracted from various capacitors is the same. In a low power converter
can be difficult to justify the choice for this solution.
Basic topologies 35
Figure 3.2 Single-phase multilevel rectifier/inverter
Table 3.2 Components list for multilevel rectifier/inverter
Capacitor 2 At least but can be rated to half DC bus voltage
3.1.2 DC-DC
Half-Bridge converter (Buck/Boost)
This DC-DC converter can be viewed as nearly universal as long as viewed from both
sides. It consists of half-bridge with two fully controlled switches and an inductor connected
between the two switches and the load (battery pack in this case).
Thus when the power is transferred from the DC bus to the load it works in buck mode,
i.e. allows charging any battery whose maximum voltage during charging process does not
exceed the DC bus voltage. This operation is possible by controlling the top switch and with
bottom switch keeps functioning as free-wheeling diode.
Inversely enables to discharge the battery transferring energy to the DC bus and
operating the circuit as a boost. This mode is achieved by switching the bottom switch and
holding the top switch to act as a diode. So in this mode only voltages higher than the battery
voltage can be achieved on the DC bus.
For purposes of this application can be pointed out two obstacles to the use of this
topology. First the battery is not electrically isolated from the DC bus or from the mains,
which violates a requirement of the project. Secondly a very large inductive filter would have
to be used in series with the batteries due to magnetic saturation phenomena since the
current has a strong DC component at this component.
Figure 3.3 Half-Bridge converter (Buck/Boost)
Table 3.3 Voltage range for Buck/Boost
Limit Value
Minimum 0

Maximum
(DC bus voltage)
Table 3.4 Components list for Buck/Boost
Dual Active Bridge converter (Buck/Boost)
To meet the insulation requirement many DC-DC topologies could be considered but the
majority (Forward, Flyback, etc.) are not suited to power levels that rise above 1 kW. The
only common options that can be taken into account for these levels are the half-bridge and
full-bridge in combination with a high frequency transformer. These topologies are widely
used in commercially available power supplies to power levels over 500 W but cannot be
entered into the project by not having characteristics bidirectionality.
An interesting variant of the full-bridge called dual active bridge can be introduced to
meet these two requirements: isolation and bidirectionality. This topology is certainly the
more versatile DC-DC converter as it can be controlled in order to operate as a buck or boost,
in either direction. Another of its interesting characteristics it’s the low ripple produced at
DC terminals. Still an advantage should be pointed out that relates to their characteristics of
ZVS (zero-voltage switching) over nearly the entire operating range. Unlike Buck/Boost has a
Association of AC-DC and DC-DC stages 37
perfect symmetry, i.e. the control can be performed exactly the same way to transfer power
in either direction.
Figure 3.4 Dual active bridge converter [39]
Table 3.5 Voltage range for dual active bridge
Limit Value
Minimum 0
Maximum ∞
Table 3.6 Components list for dual active bridge

Transformer 1 High-frequency. Expected to be small and light
Capacitor 1 At least one on the batteries side. The other is already on the DC bus
3.2 Association of AC-DC and DC-DC stages

As can be seen the total matching of all requirements is not possible with any of the basic
topologies presented previously. The solution lies in involving two or more of them with the
aim of forming a conversion architecture with multiple stages so that all requirements can be
fulfilled.
One of the more immediate solutions is the combination of PWM inverter / rectifier with a
buck / boost that would implement all the requirements except for the isolation of which
cannot renounce. If this hypothesis were taken would be a need to put a low frequency
isolation transformer in the grid interconnection.
Another solution would be to use the dual active bridge again associated with the PWM
inverter / rectifier. This solution although more technically complex offers electrical isolation
through a high frequency transformer which is certainly more interesting than the solution of
the low frequency isolation.
A variant of these solutions would be to use a multilevel inverter / rectifier which would
optimize the interface
face for the network and even if the option does not fall on it at this stage
will always be considered as a future improvement.
To make all this range of options more comprehensive a diagram was made with available
options for each stage. Note that with regard
r to the DC-DC
DC converter the hypothesis of no use
at all is also taken into account.
Figure 3.5 Chart of alternatives for the converter architecture
3.3 Criteria
After studied a range of alternatives for the converter implementation is time to establish
criteria for a justified choice. Although at this stage still a conceptual design the criteria will
have the most practical concern since the project is to be realized in fact. It is still necessary
to take into account the final application of the system, which is an on-board
board charger so the
volume available for such a device is limited being the power density of a winning criteria.
Following will be established the considered criteria as well
well as brief remarks on each.
Power density – the amount of power (time rate of energy transfer) per unit volume. A
key criterion in this choice taking into account the application. In this field high values can be
achieved by two factors: first the demand for converters with very high efficiency and
secondly to avoid the use of large passive elements.
Cost – final cost of the whole system. Always an important factor when there is the
possibility of becoming a commercial product. In this study,
st only a comparative
parative evaluation
can be made since at this stage the project still far from the physical system design.
Control complexity – a comparative measurement considering amount of computational
requirements involved and the control circuit complexity. A very high level may affect the
system on several factors such as cause the need for a higher performance control hardware
(more expensive), add components to the measurement and act systems increasing the total
Criteria 39
cost of the product and also last but not less important increase the engineering costs due to
longer development and debugging time of both software or hardware, which sometimes can
represent a significant part of total cost.
Switch utilization ratio – is the total switch utilization ratio defined as
=

× ×
,
where is the number of power switches in the circuit, and are their peak voltage
and current. Can be understood as the ratio between the total converter power and the sum
of the power ratings of all switches used by the topology. This criterion serves as a measure
of wasted capacity of switches and usually prejudice complex topologies with large numbers
of switches. The calculation of this parameter for the different solutions is shown in Table
3.7.
Table 3.7 SUR calculation
Solution Calculation Result Rating
×
4 × ×
1-ph PWM Rectifier 0.25 ++
1-ph PWM Rectifier + Buck-Boost ×

6 × ×
0.17 +
1-ph PWM Rectifier + Dual active bridge ×

12 × ×
0.08 -
×

1-ph multilevel PWM Rectifier
8× ×
2
0.25 ++
×

8× × + 2 × ×
1-ph multilevel PWM Rectifier + Buck-
2
0.17 +
Boost
×
12 × ×
1-ph multilevel PWM Rectifier + Dual
0.08 -
active bridge
Output voltage range – range of the output voltage at which the convert has the capacity to
deliver power to the battery. Since this design is not targeted to any particular battery pack
is important to have a voltage range as wide as possible. At this point the battery packs that
can be found in electric vehicles are not yet stabilized at a particular value which makes the
project more difficult. This way to create a charger with high practical value is important
being compatible with several packages of batteries. Even for a given battery, voltage may
vary considerably for different state of charge values and a good regulation is desirable in the
whole process.
For this purpose was an estimate of each converter output voltage range as is described
above. These data are presented in Table 3.8.
Table 3.8 Output voltage ranges
DC-DC stage Range Rating
None (output of the AC-DC stage) 0

Buck-Boost 0 +
Dual active bridge 0 ∞ * ++
* It means that there is no theoretical limit.

3.4 Comparative table

Once established the criteria it was considered as very useful to put all data in a table to
allow a choice for the most interesting solution for this application. In fact a few options
were left out for various reasons but the most rational choices are the ones that do have
interest in this phase. The purpose of this table is to support a comparative analysis and allow
proposing a solution.
Switch Output
Power Control
Solution Utilization Cost voltage Total
density complexity
Ratio range
Single-phase
LF* transformer + PWM ++ - - ++ 0 +2
rectifier
LF transformer + PWM + -- - + + 0
rectifier + Buck-Boost
PWM rectifier + Dual - ++ ++ 0 ++ +5
active bridge
LF transformer + PWM ++ - - - 0 -1
multi-level rectifier
LF transformer + PWM + -- - - + -2
multi-level rectifier +
Buck-Boost
PWM multi-level - ++ + - ++ +3
rectifier + Dual active
bridge
*LF stands for low frequency.
3.5 Discussion
This study has covered a few different power converter topologies for bidirectional
battery charger for electric vehicle application. According to the criteria defined so far the
chosen topology is the association of a conventional single-phase PWM inverter/rectifier with
the dual active bridge DC-DC converter.
The feature of providing insulation without penalizing the power density and cost was the
main advantages of this solution. A reasonable control scheme can also be design to this
converter with any of the conversion stages handling the task of controlling the rate of
charge/discharge.
A proper voltage output level can be achieved not only with the transformation ratio but
also with the proper control of any of the converters.
For power density classification the LF transformer was the determinant factor, but not
the only one. In buck-boost the need for an output filter built through large and heavy passive
elements may still penalize more the classification.
In relation to costs was understood that the LF transformer would represent a
substantially larger amount than all other components.
All multilevel topologies got a bad rate essentially on control complexity. The need for a
monitoring of half-bus capacitors to keep they balanced along operation are a critical
Proposed architecture 41
requirement. Considering the switching scheme these converters appear again with more
switching states and an inherent more complex control algorithm. They need twice more
drive circuits and a more complex protection circuit.
The rest were ranked in descending order starting with only a single-phase converter,
after associated with the Buck-Boost and finally the association with the Dual Active Bridge.
A wide range of output voltages is important not only to make the solution applicable to a
wide range of batteries, but mainly to allow for a proper control of the load current in all
states of charge of the battery. Either the buck-boost or the dual active bridge allows a wide
range of output voltage levels. In theory the buck-boost is upper limited and will charge the
battery with output voltages ranging from the maximum voltage DC bus down to zero. On the
other hand the dual active bridge doesn’t have a theoretical upper limit and can the
minimum voltage output is also zero.
In summary, the problem of transferring energy between a Grid and a high voltage DC bus
in Electric Vehicle has been addressed. This work seeks to answer the following question:
What is the best structure of the power converters to charge the batteries of electric vehicles
with smart grid concept? The answer is of course not unique, given that not all requirements
are known. However, this study points to a direction based on the single-phase Dual Active
Bridge topology. The dual active bridge converter have attractive features in terms of
reasonable number of components, low switching losses, high-power density and simple first-
order stable dynamics. The main design issue is the transformer design.
3.6 Proposed architecture

This section provides the proposed structure for the charging system. For a better
understanding of the framework a diagram is presented in Figure 3.6.
The first level is only conceptual and presents the main components of the car electrical
system. This level is important to understand where the charging system takes place into the
overall electric system. Then the proposed architecture for the charger is presented
according to the selection made above. In the last level appears an electrical circuit of the
converter that is proposed to be developed in this work.
3.7 Summary
In this chapter were presented basic topologies considered more relevant to the
implementation of this charging system. Its voltage ranges as well as an assessment of their
feasibility and complexity are also discussed.
After that are defined criteria for a reasoned selection and are data unified to make way
for a discussion and a consequent choice for one of the alternatives. The converter
architecture is presented and detailed for three levels.
Figure 3.6 System architecture

Chapter 4
Dual active bridge converter
This chapter is divided into two parts. The first merely theoretical in nature aims to
provide a description of the converter to design in terms of its constitution and how it works.
It also examined the operation in steady state supported with basic waveforms and equivalent
circuits. The most basic expressions are also presented as well as defining the boundaries for
soft-switching operation. The limits for soft switching are also mentioned.
In the second part is presented and described a simulation model for this converter and
several simulations are made with clear objectives. The first is to demonstrate the proper
functioning of the converter in buck and boost modes presenting its key waveforms. The
second objective is to determine a maximum value for the leakage inductance and perform
various simulations for comparison with analytical results. In this phase should also be
evaluated the suitability of the converter to the specific application. Finally there will be an
estimate of the minimum capacitor to be placed on the batteries side to meet predefined
ripple amplitude.
4.1 Introduction
The Dual-Active Bridge (DAB) topology is inserted in the group of bidirectional isolated
DC-DC converters. It was firstly proposed by Kheraluwala et al. in [39-40] and consists
essentially in two full bridges and a high frequency power transformer.
The ac terminals of each bridge are connected to each side of a high frequency
transformer and the dc sides connected to different DC bus. Bridges are controlled in order to
generate a symmetrical bipolar square wave with high frequency. The converter has a total
symmetry, i.e. whether seen from primary to secondary or from secondary to primary, is
exactly the same. This feature allows controlling power flow on both directions in the same
manner and can operate in buck or boost mode.
44 Dual active bridge converter
Figure 4.1 Circuit schematic of a single phase dual active bridge dc-to-dc converter [39]
The most common control method for this converter consists in control both bridges so as
to create phase-shifted waveforms. When phase-shift is null there is no active power transfer
in any direction and with 90º of phase shift the maximum power flow is achieved. In a range
of 0 to 90º power flow can be controlled being always delivered from the bridge with leading
square wave. The amount of power flow not only depends on the phase-shift but also on other
parameters such as the DC voltages on both sides, the switching frequency and leakage
inductance of power transformer. This last one, the leakage inductance, is generally seen as
a transformer parasitic and implies the transformer design to be very careful with the aim of
obtaining a particular value.
In order to realize its principle of operation is usually presented a fundamental model
with two square wave energy sources connected by means of an inductor. This is a classic
model has been used to represent the basic principle of energy transfer between to
alternating sources, but with the difference of sources producing sinusoidal waveforms.
Figure 4.2 Fundamental model [41]
One of the most interesting features of this converter is that it allows the switches to be
operated in zero voltage switching (ZVS) mode for a wide load range, which allows
implementing converters with high efficiency and power density. Another advantage of DAB
over other topologies is the potential for size reduction of the passive components, by using
Steady-state analysis 45
integrated magnetics. These characteristics are of great added value when application is an
electric vehicle battery charger.
4.2 Steady-state analysis
Equivalent circuits
One of the methods to analyze switching circuits is representing the equivalent circuits
for each state. In this case the analysis will be conducted only for half cycle due to the
characteristics of symmetry of its operation.
For this analysis we assume that the switching sequence begins with current flowing
through the devices DA+, Ds3 and Ds2 as shown in Figure 4.3. As the current is flowing by the
diodes, the voltage drop on the switches is zero. Thus the switching of TA+ and after a certain
phase-shift of Ts2 and Ts3 will take place on zero voltage.
Figure 4.3 Equivalent circuit 1
When current reverse will begins to flow by TA+, DS2 and DS3 as can be seen in Figure 4.4.
After that is the time for the switches and Ts2 Ts3 turn off and forcing the current flow and
by Ds1 and Ds4. This is represented in Figure 4.5.
When the switch TA+ is turned off the energy stored in the inductance is transferred to the
snubber capacitors. The snubber capacitor placed across TA+ will take over the current, and
the switch turn-off occurs under ZVS conditions and when the device is carrying a certain
minimum current. Additionally, the snubber capacitor connected across the TA− will be
discharged and will force the diode DA− to be forward-biased. Therefore, the inductor current
starts flowing through the diode DA− and diodes Ds1 and Ds4, as shown in Figure 4.6.
Similarly, the switch TA− is turned on at zero voltage and the commutation sequence will
be repeated as above[42].
An alternative analysis can be made to this circuit through its main waveforms. This
analysis is then performed and allowed to identify the modes of operation of the converter.
Voltage across the leakage inductance terminals (VLσ) can be seen as the difference
between the ac terminals voltage in at any moment.
Recalling that the bridges are switched with a bipolar pattern only two different values of
output voltage at the ac terminals can be obtained (+ VDC and - VDC).
Steady-state analysis 47
Figure 4.7 Fundamental waveforms [41]
Table 4.1 Leakage inductance voltages in each time instant
Instant Primary bridge Secondary bridge Leakage inductance

(Figure 4.7)
1 + V1 - V2 V1+V2
2 + V1 +V2 V1–V2
3 - V1 + V2 -(V1+V2)
4 - V1 - V2 -(V1-V2)
From Figure 4.7 can now be identified two operation modes, each consisting of two
instants that the current has the same behavior but with inverted slope.
In this way will be established mode 1 that includes the instants 1 and 3, where voltages
have different polarity, and mode 2 in which are inserted instants 2 and 4 and where voltages
have the same polarity.
Mode 1:
0 ≤ ( ≤ ()
(4.1)
/ + 0
*+, -(. = ( + *+, -0.
(12 (4.2)
Mode 2:
() ≤ ( ≤ (/
(4.3)
/ − 0
*+, -(. = -( − () . + *+, -() .
(12 (4.4)
Considering that the steady state current waveform is perfectly symmetrical the current
at the end of a half-cycle can be expressed as:
*+, -0. = −*+, -(/ .

(4.5)
The initial current (iLσ) can now be calculated. From (4.2) iLσ(ωt0) is determined by
considering ωt = ωt0. From (4.4) iLσ(ωt1) is determined by setting ωt = ωt1 and the iLσ(ωt0)
found on (4.2). Rearranging and defining 4 = 567 in combination with the (4.5) results:

568
/
*+, -0. = 9-1 − 4.-(/ − () . + -1 + 4.() :
2(12 (4.6)
In a similar way, *+, -() . can now be determined from (4.2) by insertion of *+, -0.
expression found in (4.6).
/
*+, -() . = 9-1 + 4.() − -1 − 4.-(/ − () .:
2(12 (4.7)
From Figure 4.7 is visible that (/ − () = ; − and () = , being the phase-shift of
two voltage waveforms in radians. Rewriting the equations:
/
*+, -0. = 9-1 − 4.-; − . + -1 + 4.:
2(1, (4.8)
/
*+, -. = 9-1 + 4. − -1 − 4.-; − .:
2(1, (4.9)
The converter output power is then given by [39]:
2/ 0 ||
<= = 4 >1 − @
(1, ; (4.10)
Average output current:
<= 2/ ||

= = = >1 − @
0 (1, ; (4.11)
Boundaries for zero-voltage switching 49
4.3 Boundaries for zero-voltage switching

As mentioned earlier, this conversion topology has a large load range where the switches
can be operated under zero-voltage switching and ways to identify the limits of this area is
crucial when high efficiency is pursued. Thus it is necessary that a certain minimum current is
circulating through the circuit ensuring that every time a switch occurs in a particular
switching cell, the current is circulating through anti-parallel diodes in that moment.
These conditions can then represented analytically by:
*+, -0. ≤ 0
(4.12)
*+, -. ≥ 0
(4.13)
Gathering information from, (4.8), (4.9), (4.12) and (4.13) the soft switching region can
be delimited as a relationship between 4 and .
;
4≤
; − 2 (4.14)
2
4 ≥1−
; (4.15)
An important conclusion can be made based on (4.14) and (4.15). For M=1 the soft-
switching conditions are fulfilled with any phase-shift angle.
This feature can gain a special interest when the converter is used in combination with
another such as a grid-connected inverter / rectifier. This way you can adjust the input
voltage to meet the output voltage of the converter and thus ensure a soft-switching
operation throughout the operating range.
2,5 Soft Switching Region
1,5
M
0,5
0
0,00 0,17 0,35 0,52 0,70 0,87 1,05 1,22 1,40 1,57
Phase-shift (rad)
Figure 4.8 Soft switching region vs. Phase-shift and M
Once plotted the region where the converter operates in soft-switching mode is proven
the above conclusion and can still be made other comments. The higher the phase-shift
(which means a higher power transferred) is the widest area of soft-switching which is very
favorable since losses increase with power level.
power point ( = ), zero voltage transitions are guaranteed.

In the limit can be observed that for a maximum phase-shift corresponding to maximum
B
0
4.4 Simulation
After studied in detail the topology of the converter to be used and their modes of
operation it was felt interesting perform some modeling and numerical simulation tasks. The
simulation will allow easily gaining skills in the operation of this topology and assessing their
potential for integration into the complete system.
At this stage equipment to be connected on both sides of the prototype are already
known, so it is of great interest to include them as part of this simulation.
Another added value of simulation is to allow easily evaluate some parameters of passive
components especially and point to a provisional acceptable values at this stage.
The simulation model is thus constituted by an ideal DC voltage source that represents
DC-bus interconnection with the converter AC-DC, two full bridges with ideal switches, an
ideal transformer with a turn’s ratio of 1:1 and externally added leakage inductance, and a
Lead-Acid battery pack with almost 400 V. A capacitor was also added in parallel with the
battery pack in benefit of reducing the ripple voltage.
Open loop control system for validation of the convert is implemented by a square wave
generator and the variable time delay. This solution can only be used for simulation because
it’s of very limited phase-shift resolution. When the system is implemented in a real-time
Simulation 51
platform, the phase-shift

shift generating mechanism must be done by dedicated hardware, so that
the phase-shift
shift step can be much faster than the cycle time of the control program.
The simulation will be carried out in Simulink® from MathWorks. The choice of this
software is based on the fact of its large availability of blocks, whether for modeling
m the
control system or power system, high control over the simulation process and also due to its
comprehensive graphical user interface. For power blocks SimPowerSystems library will be
used.
Measurements selected at this stage are very important. Measure the voltage output of
the ac terminals of each bridge is essential as well as the current in the leakage inductance.
This form will then be possible to establish a comparative analysis with the waveforms shown
in the theoretical introduction previously
previo made to this converter. Another critical component
is the battery pack. The battery model provided by Simulink® already has a port of measures
to that purpose.. At that port is provided a vector with three signals: state of charge, current
and voltage
e on battery terminals. Monitoring these signals is essential to assess whether this
converter delivers energy in a way appropriate to charge / discharge batteries.
4.4.1 Model and parameters
Figure 4.9 Simulation model
Table 4.2 General simulation parameters
DC voltage (left side) 400 V
Switching frequency 10 kHz
Table 4.3 Transformer parameters

Turns ratio 1:1
Primary winding resistance 5 mΩ
Secondary winding resistance 5 mΩ
Magnetization reactance 50 mH
Equivalent Leakage inductance 1000 µH
Table 4.4 Battery pack parameters
Battery type Lead-Acid
Nominal voltage 384 V (12 V x 32)
Rated capacity 7 Ah
Simulation 53
4.4.2 Simulation results

As mentioned earlier the primary objective of the simulation was to verify and evaluate
its proper operation in association with the DC power supply and battery pack. Figure 4.10
and Figure 4.11 representing the typical waveforms of this converter in both buck and boost
modes allow to ensuring correct operation of the converter. These first results allow also
verifying the proper functioning of the command system implemented.
Figure 4.10 Input bridge voltage, leakage inductance current and output bridge voltage in boost mode
(=45º)
Figure 4.11 Input bridge voltage, leakage inductance current and output bridge voltage in buck mode
(=45º)
4.4.3 Maximum leakage inductance estimation

At this point a critical task must be performed. As previously seen the power transmitted
by this converter as well as the current is limited by several parameters such as input
voltages and output phase-shift, frequency switching and leakage inductance of the
transformer.
Some of these parameters are already established at this stage being them the DC bus
voltage, minimum allowed voltage for the battery pack and the switching frequency.
The goal is to charge or discharge the batteries with a current of 10 A at the point of
maximum power in any conditions. Since the phase-shift is the system control variable, there
remains only the leakage inductance. The challenge now is to find the maximum acceptable
value for this parameter to ensure the desired level of power transferred.
Table 4.5 Considered scenario
Parameter Value
Input voltage (DC bus) 400 V
Battery pack voltage (Minimum) 360 V
Switching frequency 10 kHz

;
90º E
2
Phase-shift (maximum power transfer)
2/ ||
= = >1 − @
(1, ; (4.16)
2 × 400 ; 1
10 ≤ × G1 − H
2 × ; × 10 × 10F × 1, 2 2 (4.17)
800;
10 ≤
80; × 10F × 1, (4.18)
800;
1, ≤
80; × 10F × 10 (4.19)
1, ≤ 1 × 10IF J
(4.20)
After determined a maximum value for leakage inductance a plot based on Table 4.5
scenario is now presented in Figure 4.12. As expected the maximum current is 10 A,
corresponding to a shift of 90º.
Simulation 55
10
9
8
Ouput Curent (A) 7

6
5
4
3
2
1
0
0 0,3925 0,785 1,1775 1,57 1,9625 2,355 2,7475
Phase-shift (rad)
Figure 4.12 Output current vs. phase-shift
Achieved a value is now inserted into the simulation model to validate the calculation. It
is import to refer that this value is an upper limit for the leakage inductance. Lower
inductance values can be used to fulfill the desired power range. For instance, an equivalent
leakage inductance of 500 µH allow to achieve a maximum output current of 20 A. This way
controlling the phase-shift angle between zero and a certain value below 90º, a (0 - 10) A
range can be performed. Such characteristic can avoid the need for an external inductor.
B
battery at a voltage near 360 V and a phase-shift of 90º E .
The first simulation will be carried out exactly to the proposed dimensioning, i.e. with the
Figure 4.13 Battery SoC, Current and Voltage with Input 400V, Output 360V, of 90º Lσ = 1 µH
One of the important conclusions that can be taken from (4.16) is that the average output
current does not depend on the output voltage. Thus the output voltage was then modified by
establishing a new initial value for the SoC of the battery, which increased from 20% to 80%.
The aim is to verify that conclusion based on the analytical expression.
Figure 4.14 Battery SoC, Current and Voltage with Input 400V, Output 400V, of 90º Lσ = 1 µH
Another important task to be undertaken during the simulation is to verify the

characteristics of symmetry relative to bidirectionality. This will be confirmed expeditiously
by changing the phase shift of 90 degrees to 270 degrees (which corresponds to -90°).
Figure 4.15 Battery SoC, Current and Voltage with Input 400V, Output 355V, of -90º Lσ = 1 µH
Simulation 57
At this stage were already checked some important conclusions about the operation of
this converter in steady state. Although it is still important to characterize the converter as a
function of phase-shift for a particular operating point and compare with the analytical values
of Figure 4.12.
To this purpose several simulations were carried out and will not appear obviously the
plots of them all. Instead the current values in the battery for several values of phase-shift
will be plotted on a new chart together with the analytical values. These values are measured
in steps corresponding to five degrees.
12
10
8
Ouput Curent (A)
0
0 0,314 0,628 0,942 1,256 1,57
Phase-shift (rad)
Analytical Simulation
Figure 4.16 Output current vs. Phase-shift - analytical and simulation results
Obtained these results the converter was characterized for steady state operation in
simulation aspects. The values are approximate although affected by an offset error. This
same procedure could be carried out for a variation of output current vs. input voltage, but
with a reduced interest by two factors: first the input voltage is assumed that does not vary
significantly during the operation and second the relationship between input voltage and
output current is linear.
4.4.4 Capacitors dimensioning

Another important task that can be carried out during the simulation is a first
approximation to a reasonable value for the capacitors. This procedure can be divided in two
phases: the first one will justifiably determine what the worst scenario for a particular ripple
signal is and the second set a maximum allowable ripple (usually in percentage).
Besides all this, find a reasonable value cannot have only a lower limit, but also upper.
Usually in applications of power electronics capacitors come with high capacity values
generally associated to high voltage ratings, which means large capacitors. In this phase must
have sense and balance all these factors to achieve an appropriate solution.
Surely the worst scenario for the ripple in the output voltage is at the moment when the
voltage is lower and output is coupled to heaviest load.
For this specific application, that scenario occurs when the battery pack is at minimum
state-of-charge and the allowable output current is the highest, since as previously shown the
output current is independent of voltage.
One of the features that make this topology very feasible for charge / discharge batteries
is its low ripple on the output. Enjoying this fact may be given a very low limit for the ripple.
In the simulation model was placed a resistive component (ESR) in series with the
capacitor otherwise ripple observed would be unrealistic. This value was approximated at 100
mΩ by excess given that electrolytic capacitors in this range and of high quality should be a
value somewhat less.
The maximum value for the ripple was then set at ±1% of the average output voltage for
this stage and already includes that resistive component for the capacitor.
338,5
338
337,5
337
336,5
336
335,5
335
334,5
334
0,0135 0,01355 0,0136 0,01365 0,0137 0,01375
Battery Voltage Ripple
Figure 4.17 Output voltage ripple
Considering the conditions mentioned above were then performed simulations with an
output of about 10 A and a phase-shift of 90 ° and typical values of commercially available
capacities (470µF, 1000µF, 1500µF and 2200µF). The lowest value capable of fulfilling the
requirement of ±1% of the DC voltage was the value of 1000µF and the signal for that capacity
is plotted in Figure 4.17.
Table 4.6 Capacitor dimensioning
Parameter Value
Output voltage (average) 335,7 V
Output voltage (minimum) 334,2 V
Output voltage (maximum) 338,0 V
Ripple voltage 3,8 V
1,13 %
Ripple in %
±0,57 %
Summary 59
This value should ensure safe operation of the batteries. Considering the total size of the
converter this capacity should not represent a high volume. If a safety margin is desired the
capacity can be increased without major problem for a 1500µF capacitor to deal with higher
ESR or even to reduce the voltage ripple.
4.5 Summary
This chapter is an introduction to the topology to be implemented, describing the basics
of his theory. Their average expressions are also presented as well as the limits for operation
in soft switching.
In the second part a simulation work is performed with the objectives to understand the
operation of the converter as well as assess the influence of various parameters in its
operation.
Chapter 5
Prototype design
All aspects of the hardware power are described in this chapter being the discussion
always focused on practical issues. In each of the parts there is a brief introduction to the
most relevant options for this application and then are there defined criteria / measures of
performance in which the choice is based. The procedure will always result in choice of a real
component to be used in the prototype.
In the choice of switches there is a brief comparison of the most common types used in
power electronics for this application and the specification of minimum requirements that the
devices must have. Lastly, the chosen power module is presented taking into account all the
aspects discussed. After choosing the switches a study is made about the appropriate drive
circuits and a list of requirements is presented.
For capacitors the design procedure starts with a brief discussion of the types with
relevance for this application and the reference to performance measures for each type with
a choice of specific components in the end.
The transformer design is discussed in the last section, with a greater level of detail.
Issues such as physical structure, dimensioning of magnetic and electric circuit, estimation of
losses and other parameters and choice of all the components are presented.
5.1 Power switches

First step in active components selection is to find out the voltage and current ratings
involved, as well as the switching frequency range.
In this specific case the frequency used for the converter switching is constant and
defined as 10 kHz. The voltage rating is pointed to be twice the maximum voltage that each
switch will block. Considering the input data presented in Table 5.1 the selected switches
must have a blocking voltage of at least 800 V. Major types of switches are designed with just
a few values of blocking voltages. For instance IGBTs are easily found on the market with
typical levels of 600, 1200 or 1700 V.
Another critical parameter is the current supported by the switch under continuous
operation. This value should be oversized respecting to the one calculated during circuit
62 Prototype design
analysis or measured in simulation environment, especially in a first prototype. If too short an

accurate thermal protection must be designed and the transient overload intervals have to be
studied in order to prevent critical overheating or possibly destruction. Considering the
typical applications a scale factor of two can be inserted to the defined nominal value,
resulting in a quite robust solution. In this converter the nominal power is set to 4 kW which
means a DC bus current up to 10 A. Using the scale factor mentioned above concludes that
the current rating of the switches to be used must be at least 20 A.
Table 5.1 Input data for switch selection
Parameter Value
Maximum voltage 2 × 400
Rated power 4 KL
Maximum current (continuous) 2 × 10 M
Once defined the minimum values of the most basic parameters (Table 5.1), the switch
type must be selected. The main types with interest to implement such converter are MOSFET
and IGBT.
MOSFET - They have very high input impedance (capacitive) which allows to no power
consumption under steady state conditions – power is only required to switch the device. The
on-state resistance has no physical limit so that devices with very low on-state losses can be
found. A positive temperature coefficient prevents thermal runaway and makes them suitable
to use in parallel.
IGBT – They combine the advantages of the MOSFET including the ease and efficiency of
the command with high voltage and current ratings of bipolar transistors. Its negative
temperature coefficient makes its utilization in parallel difficult to implement. Commercially
available devices provide a built-in anti-parallel diode [43].
Considering these aspects the most suitable switches type for this application are IGBTs.
The voltage and current ratings offered along with an easy command make this technology
particularly interesting for such applications.
Another relevant issue is the case where the semiconductor is supplied. IGBT are available
in cases ranging from a small D2-PAK for PCB surface mount to large non-PCB devices that can
be small blocks to mount on heatsinks with standard 80 or 93 mm rails or manufacturer
proprietary case types. These large devices are usually found on the market with a single
IGBT on each case or in power modules. These modules can be half-bridges, full-bridges or
six-packs. Some of these modules integrate a temperature sensor.
The selected device is a block-type IGBT with two transistors per case. Its main
advantages are high mechanical robustness, ease of mount due to standard distance between
fixation screw holes (80 mm), ease of electrical connection in power (M5 screw) and
Drive circuit 63
command terminals (Fast-on) and the high availability of accessories like DC bus bars and
snubber capacitors. Unlike PCB assemblies this one can be easily reconfigured to future work.
The selected component was SKM100GB12V IGBT Half-Bridge from Semikron®.
Table 5.2 SKM100GB12V IGBT Half-Bridge main specifications [44]
Parameter Value
Blocking voltage (VCES) 1200
Nominal current (ICnom) 100 M
5.2 Drive circuit

In power electronics, the so called driver is an electronic circuit that has the functionality
to operate the main power switch in a correct manner. In most cases this circuit includes
protection functions. These protections may be over and under voltage, over current, short
circuit and over temperature. Drivers for multiple switches can include also interlock and
dead time features.
Nowadays there is a wide commercial availability of drivers for power switches. They can
range from small integrated circuits (IC driver) to high-end products known as hybrid drivers
that include many of the above described features. The main differences between both types
are listed in Table 5.3.
Table 5.3 Main driver types comparison [45]
Hybrid driver
Driver type Driver IC
Midrange High-end
Single, Halfbridge, Sixpack

Available configurations Single, Halfbridge, Sixpack etc.
etc.
Insulation no galvanic insulation galvanic insulation
Level shifter Pulse transformer

Signal transmission Opto-coupler
(mainly for high side) Fibre optic
Bootstrap-circuit
Energy transmission Switch mode DC-DC converter
(for high side)
low power medium power high power (>100

Application
(<5 kW) (5...100 kW) kW)
The most interesting devices are undoubtedly the hybrid drivers due not only to features
which typically include, but also its performance. For that reason this type of drivers will be
chosen for this design, improving the robustness and safety of the prototype and reducing the
development time.
64 Prototype design
In order to select the most interesting product the performance measurements have to be
identified as well as the required/desired features. Its performance measurement is usually
made based on some essential characteristics which are:
Peak output current: Should be as high as possible. This characteristic can be achieved
with very low output impedance on the secondary side. Currently available drivers reach
values as high as 50 A (Skyper 52 R from Semikron®) [46].
Maximum switching frequency: At least the converter maximum switching frequency. In
most cases this frequency is dependent on the gate charge of the power switch and ambient
temperature where the component is placed. Its value is typically in the range of tens of kHz,
depending on the range. Higher frequencies are of limited interest for high power converters,
resonant converters unless.
Table 5.4 Driver features
Feature Relevance Explanation
Implemented generally by continuous VCEsat monitoring

Short-circuit detects when this value goes above a previously defined
Needed
protection voltage. It’s indispensable to protect the integrity of power
hardware.
Although it can be managed by the control platform, cannot

Top/bottom
Needed be sufficiently robust to ensure that commands will never
interlock
short-circuit the DC bus.
It is necessary to have access to an error signal so the control

Error output Needed
platform can manage it in right manner.
At system start-up or after a recovery from a failure is

Reset input Needed required to reset the drivers to resume normal converter
operation.
Required to ensure that the power switch limits are

Over current
Desirable respected. Needs to be configured according to the power
protection
modules or other hardware limitations.
Required to ensure that the power switch limits are

Over and under
Desirable respected. Needs to be configured according to the power
voltage protection
modules or other hardware limitations.
It can be as simple as a voltage comparator with an

Over temperature adjustable limit to allow a sensing with an NTC or it can be
Desirable
protection configurable to use PTC or NTC. This feature can be
implemented by an external dedicated circuit.
Holding these all the selected driver is the model Skyper 32 PRO R from Semikron®.
Besides all the features mentioned above, this driver include some more advanced. All of
them are represented on Table 5.5.
Drive circuit 65
Table 5.5 Skyper 32 PRO R features [47]
Two output channels
Integrated potential free power supply
Under voltage protection
Drive interlock top/bottom
Dynamic short circuit protection
Half logic signal (reset + error signals)
Failure management
Soft turn-off
External error input (secondary side)
Some of these features are partially implemented on an evaluation board that interfaces
between the driver core and power hardware. For the same driver core there are several
different boards for direct connect to proprietary power modules or alternatively a universal
adapter for a wired connection to any other power module is available.
This driver solution supplied it’s almost ready to use. Although, it’s necessary assemble a
few components to configure some parameters. A technical document related to the device
has detailed information on how to determine the values of these components.
5.2.1 Parameter setting

The parameter setting and board configuration is made through some components
soldered on the evaluation board and based in [48].
5.2.1.1 Dead time setting

Dead time is an important parameter since the two or more fully controlled switches are
series associated in the same DC bus. In most cases they are turned ON/OFF in a
complementary way. This process is not instantaneous and transient short circuits may occur
on the DC bus.
Merging the data from the driver core and the evaluation board technical documents is
possible to find that the dead time is controlled by means of four external pins, one to
enable/disable dead time e three to select from 8 different values.
The simplest approach to establish a value for the dead time is to ensure that a transistor
is turned on right after the process of shutting down the other is fully completed. In other
words, the dead time will be equal to or greater than the sum of the turn-off delay with the
fall time.
66 Prototype design
Table 5.6 Minimum dead time estimation
Parameter Value
Turn-off delay (td(off)) 418 N
Fall time (tf) 62 N
Minimum dead time 0,48 µN
Considering this value any of the values available from the drive could be used but in fact
is not so, as the parameters present in the technical document refer to tests carried out with
a gate resistance of 1Ω, which is quite low. A higher value of gate resistance will lead to a
higher fall time and this made a need to have a value of dead time also higher. For this
reason, it will be used for the dead time a higher value until the gate resistance is set and fall
time is accurately determined. The value used in this phase will be the default on the
evaluation board, 3.3 µs.
Drive circuit 67
5.2.1.2 Dynamic short circuit protection
This protection feature is implemented through a parallel connection of a resistance and

a capacitor. These two basic components are intended to generate a time constant that has a
behavior similar to the VCE voltage during normal operation of the circuit. The analytical
expressions to determine these values are given in [47].

QVWW + Q ×
Q 9KΩ: = −15,5KΩ × T U1 − KΩ X
8 (5.1)
µN
\] 9µN: − 2,1 − 0,11 × Q
YQ 9Z[: = Ω
µN
0,00323
Z[
(5.2)
Expressions (5.1) and (5.2) have two input parameters and they are the Collector-emitter
threshold static monitoring voltage -QVWW . and the blanking time -\] ..
The VCEstat is the maximum allowed VCE value in steady state mode when the switch is
turned on that is not seen as a short-circuit. In the selected transistor the maximum value of
VCE with the nominal current is 2.65 V, which means that the value of VCEstat expected to be
always above this value. If this parameter has a value too low the transistor will be better
protected but the protection can be activated unexpectedly during transient overload. On the
other hand, if the value is too high the transistor may be subject to weak short-circuits,
which do not act the protection.
The tbl is the time between a turn-on command and the first monitoring of the transistor
VCE voltage. To correctly choose this value the switching characteristics have to be studied in
detail. When the switch is turned on the VCE voltage falls from the prior blocking voltage to
VCEonsat value with an exponential behavior. To handle with this characteristic is required that
VCE reference fall more slowly than the true value of VCE, or at limit at the same speed to do
not trigger the comparator and hence the protection.
For being beyond the scope of the work was decided to use a reference value for a typical
1200V IGBT application, suggested in [47]. These values for VCEref and tbl are presented on
Table 5.7.
68 Prototype design
Table 5.7 Dynamic short circuit protection parameters
Parameter Value
VCEref 5,5
tbl 5,1 µN
RCE 18 KΩ
CCE 330 Z[
5.2.1.3 Collector Series Resistance
As seen earlier, short-circuit protection is done by monitoring of the voltage VCE, which
implies the need for sensing the voltage at the collector of the transistors. The connection
between the drive circuit pin which makes this measure and the collector of the transistor is
done with a reverse biased high voltage diode, which is already assembled on the evaluation
board.
When using a 1200 V IGBT only the diode should be placed in series and with 1700 V IGBT
should be placed the diode in series with a resistance of 1 kΩ with a minimum power
dissipation of 0.4 L. The place should be assembled with a shunt or a 0Ω resistor.
5.2.1.4 Adaptation Gate Resistors
The gate resistance is a very critical parameter for the operation of the circuit, especially
with concerns to the switching characteristics. A very low value should able to fast
commutation of the transistor, but with higher dv/dt and di/dt values and a higher current
absorbed from the driver. In most cases the semiconductor has some different turn on and
turn off times. This leads to the need to use different resistors to switch on and off the
transistor. This driver provides such optimization and contemplates separate spaces for
assembling turn on and turn off resistors for each transistor.
Considering the transistor datasheet and collecting the values of the gate input capacity,
rise and fall time and approaching the command circuit for an RC network is possible to
estimate an optimized value for the gate resistance.
Figure 5.1 Considered equivalent circuit

Drive circuit 69
Table 5.8 Input data for calculation [44]
Parameter Value
Cies 6.01 [
tr 38 N
tf 62 N
As a series RC circuit has a time constant ` given by × Y. The Y value is given in the
datasheet of the transistor, and approaching the rise and fall times by 5` is possible to
determine the resistance value to place in series with the IGBT gate.
The on resistance is calculated using the rise time and the OFF resistance is based on the
fall time.
Table 5.9 Gate resistors calculation
Parameter Value
38
à = = 7,6 N
5
Rising time constant
62
`c = = 12,4 N
5
Falling time constant
YghV
def = = 0,79Ω
à
Gate resistor (ON)
YghV
deii = = 0,48Ω
`c
Gate resistor (OFF)
The gate resistors are not only limited by the characteristics of the transistor. The value
of peak current supplied by the driver must also be considered as a restriction on the
minimum resistance that can be used.
The driver has the output voltage level of +15 V when to turn on the transistor and -7V
when to turn it off. This represents an almost instantaneous change of 22 V when the
command changes. Recalling that the peak current provided by the driver is 15 A, and
considering the transition to be instantaneous, the resistance should ensure that this current
limit is never exceeded.
Despite the current limit of the driver is high it was understood that at this stage would
not be necessary to achieve its maximum value. Thus the value of desired maximum current
was set at one-third the maximum current, i.e. 5 A.
d ef Ieii 00
jkf l mnjop
l lr,r s
q (5.3)
70 Prototype design
Considering all these factors the selected value was 4.7 Ω, since it is the closest value
above 4.4 Ω found in E12 resistor table.
5.2.1.5 Soft Turn-Off
The functionality of soft turn off is the one that distinguishes this version (PRO) of the
normal version of the same product. This mechanism doesn’t allow an abrupt turn off the
transistor in case of a short circuit. Thus when a fault occurs the transistor is switched off
smoothly (gradually increasing the resistance in series) through its transient operation in the
linear region. This form voltage spikes due to very high di/dt will not be generated, especially
in the presence of inductive elements with very high inductance values.
The speed of this process of soft turning off the transistor is related to a resistance that
can be mounted on the evaluation board. At this point this functionality is disabled.
5.2.1.6 Over Temperature Protection
Together with all the features above, there is an extra feature of temperature protection
provided by the evaluation board. This protection system needs to be configured to specific
parameters of the temperature sensor and for the sensor type, either PTC or NTC. The
evaluation board uses the external input error to put the driver in HALT mode if the set
temperature limit is exceeded.
This feature is disabled because the power hardware does not provide any temperature
sensor, but can easily be added in the future.
5.3 Interface board

It is common to the inputs of driver circuits waiting to receive signals with a value of
0/15V, especially in industrial drivers. This is not the typical voltage level supplied by the
output pins of any control platform whether it is based MCU, DSP or FPGA. The commonly
used values are 0/3,3V or 0/5V either for specific internal PWM generator pins or any other
general purpose pins.
Thus it is then necessary in most cases using an auxiliary circuit that makes this
adaptation of voltage levels, which in this case is 0/5 V to 0/15 V. This circuit should be as
simple as possible not to add complexity to the project and to facilitate the hardware
debugging process. The circuit should also be designed to have minimal rise and fall times and
also low propagation delay.
The solution has to use a low-power bipolar transistor, operated always in switch mode.
The collector and base resistors are calculated in order to ensure that is in saturation mode
when turned on. This solution allows to obtain a waveform that switches between the desired
values (~0/15 V) but with an inverted logic. In addition the output values are not exactly 0 or
Interface board 71
V, which cannot be well interpreted by the driver circuit. To solve these two
exactly 15V,
problems was put forward an inverted buffer, which thus returns to signal a non-inverted
non
logic and switches between values very close to 0 and 15 V, which will be interpreted
correctly by the driver.
Figure 5.2 5 to 15V circuit
This driver also includes a signal with particular characteristics that is the HALT signal.
This signal is active low and both can be activated internally by the driver and forced
externally by another circuit. The control system is also of interest to monitor the value of
this signal, to have information about the state of operation of the driver. These
characteristics imply the need for a bidirectional 5 <->
< 15 V circuit.
Figure 5.3 Internal equivalent circuit of HALT signal
The solution to handle this signal

signal is to implement the two functions separately. The
function of forcing the driver to enter the HALT state is made by a bipolar transistor placed in
parallel with the internal one, represented in Figure 5.4,, so that when turned on will trigger
the protection by forcing the signal to a logic low level. The signal reading is done through a
buffer or voltage follower in order to provide access to this value without interfering with
72 Prototype design
their behavior. In the buffer output a resistive divider is placed to convert the signal voltage
from 15 V to 5 V.
Figure 5.4 HALT signal interface circuit
5.4 Capacitors
Capacitors are passive components of high importance in this design. Its function is to
stabilize the voltage at the two DC buses to prevent the DC sources to be operated under a
high voltage ripple. In the preliminary design of the converter and for the simulation purpose
they can be considered as ideal components, i.e. simply specifying the value of its capacity.
In fact they are not ideal components
components as any other. Unlike the inductors or transformers
which are often designed specifically for a particular converter, the capacitors are available
on the market in wide ranges of capacities and rated voltages.
Figure 5.5 Simplified equivalent circuit diagram of an electrolytic capacitor [49]
At this stage the focus will not be projecting any component but to establish the minimum
specifications to match the converter needs and identify performance measures that will
allow choosing the most adequate component.
The types of capacitors with interest for this work are aluminum electrolytic capacitors
and metalized polypropylene film capacitors. The electrolytic capacitors
capacitors are with high
capacity per unit of volume and unit of weight and are best suited to stabilize voltage within
large time constants. The other type, Metalized Polypropylene Film Capacitors (MKP) are
especially suited to handle high dv/dt, i.e. they provide or absorb large amounts of energy
Capacitors 73
during very short periods of time without opposing great resistance. When used in DC bus of a
switch-mode converter they attenuate the oscillations generated by switching.
5.4.1 Electrolytic capacitors
The minimum values of capacity and rated voltage for electrolytic capacitors were
defined in simulation (Table 5.10). Afterwards it was considered important list the
parameters that influence its performance and discuss their influence in the converter
operation.
Table 5.10 Electrolytic capacitors minimum requirements
Parameter Value
Capacity (value from simulation) ≥ 1000 µ[
Rated Voltage ≥ 400
Equivalent Series Resistance (ESR) - Extreme importance in a careful design. This value
should be as low as possible since it represents on capacitor terminals a voltage ripple that is
a fraction of current. On the other hand also have a direct influence on the power dissipated
by the capacitor.
Self-inductance (LESL) - With particular importance especially when dealing with high
current gradients, as this generates parasitic voltage variations proportional to the variation
of the current, which is undesirable in a DC bus.
Maximum RMS current - Related to previous is also of great importance. Safe capacitor
operation within the nominal voltage rating is only valid in limited temperature range so that
the power dissipation is also limited. Maximum power dissipation can be approached as the
product of ESR by the square of maximum RMS current.
Operating temperature - The nominal voltage is typically set to a certain maximum
temperature. The higher the lower the thermal management concerns of the component. Yet
the lower the temperature at which the capacitor can be better maintained since its lifetime
increases and can be operated safely with higher voltages.
Useful life - It is generally referred to a specific temperature. In this case it comes as a
prototype is not of critical importance but a final product is always a factor to be taken into
account. This may become significantly longer operating the capacitor at lower temperatures
by decreasing its RMS current or using a cooling method.
Protection - An important detail especially in bench top equipment. This type of
capacitor is polarized and when connected in the wrong way there is a risk of explosion. In
large components can be interesting the existence of a protection system that avoids the
consequences of neglect.
Considering all these aspects was a choice of a component to use. Attempt to be a
solution with an acceptable volume compared to the size of the converter and with screw-
74 Prototype design
type terminals, so as to ease the cabling for connection to the buses. The selected part is
B43456A9158M from Epcos®.
Table 5.11 Epcos B43456A9158M Electrolytic capacitor specifications [50]
Parameter Value
Rated Capacity 1500 µF
Rated Voltage 400 V
Typical ESR 45 mΩ
(100 Hz 20 ºC) ~0.6xESR100Hz
(10 kHz 20ºC) 27 mΩ
Maximum ESR (100 Hz 20 ºC) 67 mΩ
Self-inductance LESL 20 nH
Maximum Impedance (10 kHz 20 ºC) 46 mΩ
Maximum RMS current (100 Hz 40 ºC) 17 A
Rated RMS current (100 Hz 85 ºC) 6.0 A
> 12000 h @ 85º

Useful life
> 250000 h @ 40º
5.4.2 Metalized Polypropylene Film Capacitors (MKP)
These capacitors are considered as of great importance for good performance of the
converter, especially with high power levels.
As for the electrolytic, before proceeding to the selection of a specific component it was
considered useful to establish performance measures and their influence on the converter.
Rated DC Voltage - Must be at least equal to the maximum rated voltage of the bus
because they have to withstand this voltage during most of the time.
Rated AC Voltage - A very important parameter. As already mentioned these capacitors
are designed to withstand the oscillations (spikes) generated by switching on the bus. This
way is so important that they be prepared to deal with voltages above the nominal value. The
particular value of this voltage depends on several factors such as the nature of the load or
the speed of switching and at this stage is difficult to determine accurately.
RMS current - Another important parameter as in the electrolytic. In this case the values
are generally referred to the high frequencies in the range of kHz. The limitations are set not
only by thermal aspects but also by the terminals used, which are usually large. It is
understood that a modest value of RMS current values may correspond to very high peak
current, due to the high crest factor that the current waveform have in these capacitors.
Equivalent Series Resistance (ESR) – In this type of capacitors this parameter is of utmost
importance, together with the self inductance. In an ideal situation these capacitors should
provide or absorb any current without any voltage change. In fact this assumption is limited
by the existence of such parasitic effects. The commercially available components have very
Capacitors 75
low ESR values (only a few mΩ) due to the materials used and the high contact area of the
terminals especially in large capacitors.
Some cases are specially designed for the power modules to be used in this prototype and
will be given preference to such devices. Thus it is possible to put in contact the entire area
of the terminal from the power module with the capacitor terminals ensuring a minimum
added equivalent series resistance.
Self-inductance (LESL) - Indeed the major quality characteristic of this type of capacitors
since they will be subject to surges of high amplitude but also with high variation rates. As
the ESR inserts a voltage drop proportional to current value, this parasitic inserts a drop
proportional to the derivative of the current.
yz{
Maximum permissible voltage change per unit of time t x =
uv
uw |
- Of great importance
because the capacities are generally low in this type of capacitors and the current peaks are
expected to fairly high which will result in high voltage gradients. The higher was the ability
of capacitor to support them better.
Table 5.12 Epcos B32656S8155 MKP capacitor specifications
Parameter Value
Rated Capacity 1500 nF
Rated DC Voltage 850 V
Rated AC Voltage (f ≤ 1 kHz) 450 V
RMS current (100 kHz) 13 A
ESR (100 kHz) 5.0 mΩ
Self-inductance LESL 17 nH (5.5)
Maximum dv/dt 400 V/µs
Maximum k0 680 000 V2/ µs
As can be seen on Figure 5.5 a realistic model of a capacitor can be represented by four
components. At this step the parallel resistance will be neglected since it has a large value.
The reactive components form a LC series circuit that has a particular resonance frequency
given by:
1
}ahV =
2; ~Y1V (5.4)
The manufacturer provides information for this range of products about the resonance
frequency vs. the capacity of the component.
For the selected capacity (1500 nF) the natural resonance frequency has a typical value of
1 MHz [51].
1 1 0
1 1 0
1V = G H = G H = 1,7 × 10I J
Y 2;}ahV 1.5 × 10I 2; × 10 (5.5)
76 Prototype design
5.5 Transformer design

The transformers are widely used components in all areas of electrical engineering. Its
function is to transmit energy from one circuit to another through a magnetic circuit. For this
discussion the transformers with special interest are the high frequency component that
distinguishes the isolated conversion topologies. This type of transformers is usually operated
in the area of large kHz to few MHz and with a non-sinusoidal voltage waveform, which leads
to the use of specific materials, especially for the core.
Unlike the low-frequency transformers used in electrical network for which many
manufacturers offer full commercial ranges, those for power electronics typically have very
specific requirements for each proposed converter, implying the need for skills in the project
magnetic components such as transformers and inductors.
The most common material in the transformer core for low frequencies is magnetic steel,
most the times laminated to reduce eddy currents due to high electrical conductivity. It is for
this fact that is not suitable for use at high frequencies.
For use in power electronics are another type of material being ferrite the most common
core for these applications [37, 52].
Figure 5.6 Transformer simplified equivalent model (primary referred)
Several approaches to the design of such transformers are presented in the literature [37,
52-53], some with the aim of being fast and other supposed to be more accurate. In fact, like
in any other design some simplifications have to be made because of input data that may not
be complete and certain parameters can be neglected for a given application.
The considered transformer model is presented in Figure 5.6. To make clearer the design
procedure followed, a flow chart is presented in Figure 5.7.
The transformer models are discussed in section 6.4, where the parameterization of such
component is performed.
Transformer design 77
Input Output
VA rating Core characteristic Core material,

Switching frequency size parameter shape and size
Core dimensions, peak Determine:

induction Flux Number of turns
Waveform parameters Flux linkage
Determine: Magnetization
Relative permeability
Relutance inductance
Core specific power losses Determine:

Magnetic losses
and volume Magnetic losses
Copper skin depth Determine:

Switching frequency Wire type and Wire size
Maximum current diameter limits
Wire lenght
Determine:
Copper DC
Winding characteristics Mean turn lenght
resistance
for both windings
Copper losses
Figure 5.7 Transformer design flowchart

78 Prototype design
5.5.1 Core selection
As mentioned this type of transformer usually has to be built specifically for a

particular application. Yet there is an extensive range of cores, coil formers and all kinds of
accessories for this purpose.
The ferrite cores can be found in many different shapes and types of material. The most
common formats are toroids, pot cores, and U, I and E cores, all of them in several different
sizes. The designations and frequency ranges of the different types of materials suitable for
power transformers are presented in Table 5.13.
Table 5.13 Core selection guide for a specific usage [54]
Usage Frequency range Material Shapes
N 27 E, EC, ETD, U, RM, PM
1 to 100 kHz N 41 Pot cores, RM
N 61 Small ring, E cores
N 62
Power E, U, UR, ETD, ER
transformers, up to 200 kHz N 67
Chokes
N 72 E, ETD
up to 500 kHz N 87 ETD, EFD, RM, ER
0,3 – 1 Mhz N 49
EFD, ER, RM (low profile)
0,5 – 1 Mhz N 59
So far we can conclude that most of the materials are perfectly suited to the desired
frequency (10 kHz). In fact they usually have limitations on maximum frequency or in other
cases are optimized for a certain range (e.g. N 49 and N 59).
The selected material is then type N 27 because fits the desired frequency. Their wide
availability of shapes and sizes and the ease with which can be found on the market are also
factors in favor.
Once selected the material is now necessary to find the appropriate shape and size to
handle the power level of this converter. The size is generally select by the core
characteristic size that is directly connected with the sum of VA ratings, i.e. the sum of
apparent power from all windings.
In this specific case this quantity can be expressed as [52]:

WW = aV aV = 400 × 10 + 400 × 10 = 8000 M
]] ggV (5.6)
In accord with [52] the relationship between this sum of powers and the characteristic
size can be expressed by:
/

WW
= G H
M (5.7)
Where:
is an exponent, characterizing the material and shape of the core, = 3
M is a coefficient; for ferrites, M is in the range M = 5–25 if is in cm
In remarks concerning to prior expression is noted that for low-frequency design (20–30
kHz) and low-frequency materials or presence of high DC current components, low values of M
(5-20) are applicable.
In this case the frequency is low (10 kHz) as well as N 27 material type is also for low
frequency, but a continuous component isn’t present. Balancing these factors was pointed out
a M coefficient with a value of 10.
= 9,28 = 92,8

(5.8)
So is necessary to find a core which characteristic size is at least 92.8 mm. The U-shape
93/76/30 core with a characteristic size of 93 mm is a correct choice for the design.
Due to high availability and adequate power rating a UU 93/152/30 (mm) N27 ferrite core
set, composed by two U 93/76/30 cores, will be used. According to datasheet this set is
designed for power transformers above one kW, considering a 20 kHz switching frequency.
The magnetic characteristics of this core are presented in Table 5.14.
Table 5.14 N27 UU 93/152/30 core magnetic characteristics [54]
Effective magnetic path lenght Th

354 mm
Effective magnetic cross section Mh 840 mm2
Effective magnetic volume h

297 000 mm3
Core mass 1500 g
Relative effective permeability µh 1800
Peak flux density ,

370 mT @ 100ºC
5.5.2 Number of turns
After selecting a core is important to determine the most appropriate number of turns
to be wound. This calculation is usually made for the primary and the number of turns of the
secondary then determined by the transformation ratio, which in this case is 1:1. The method
80 Prototype design
of calculation used is commonly found in the literature, and starts

st from the integration of a
half period of the maximum voltage applied to the primary. These data are presented in
Table 5.15 and Figure 5.8.
Table 5.15 Primary voltage parameters
Waveform Square bipolar
Negative peak -400 V
Positive peak +400 V
Duty cycle 50 % (fixed)
Frequency 10 kHz
Period 100 µS
Figure 5.8 Voltage waveform applied to the transformer
Flux linkage:
W0
= -. = 20 × 10IF - − N.
W/ (5.9)
Peak-to-peak induction:
= 2 × , = 740 × 10IF -.
(5.10)
Peak-to-peak
peak magnetic flux:
= Mh × = 621,6 × 10I -L.

(5.11)
Number of turns:

= ≅ 33 EN
(5.12)
To achieve a turns-ratio of 1:1 the primary and secondary windings must have the
same number of turns. After the known number of turns and the core to be used it is possible
to estimate the magnetization inductance. The first step is to determine a value for the
reluctance of the magnetic circuit.
µ) = 4; × 10I J/
(5.13)
Th 354 × 10IF
= = = 98,6 × 10F EN/J
µ) µ Mh 4; × 10I × 3400 × 840 × 10I (5.14)
Once determined the value of the magnetic circuit reluctance is possible to obtain a
value for the magnetization inductance.
0 330
1 = = = 11,04 × 10IF H
186,3 × 10F
(5.15)
5.5.3 Core losses

The estimation of the total core losses are based on two parameters from the component
technical data. These parameters are the specific core power losses that is given for two
different temperatures and the effective volume of the core. The total core losses are
calculated for the two different temperatures (Table 5.16).
Table 5.16 Core losses estimation [54]
Specific core power losses (PS) 150 kW/m3 @ 20 ºC ; 100 kW/m3 @ 100 ºC
Effective magnetic volume (Ve) 297 x 10-6 m3
Total core power losses (PT) 44,5 W @ 20 ºC ; 29,7 W @ 100 ºC
5.5.4 Wire
Once completed the design of the magnetic circuit the windings must now be specified.
One of the most important factors in this phase is to select the most suitable type of
82 Prototype design
electrical conductor for the windings. The main types used are solid copper conductor, with
round or rectangular cross section, and Litz wire.
The round solid copper conductor is widely used in transformers and coils for frequencies
up to some kHz. Its high availability in the market along with the ease to be wound makes it
the best choice for most applications. For windings with very large cross section the
rectangular conductors are a good option even for frequencies of a few kHz.
The stranded conductors, such as Litz wire, are of particular interest especially for high
frequencies above 100 kHz where the skin effect causes a poor use of the copper section.
Table 5.17 Skin depth for Copper at 100ºC [37]
Frequency (Hz) Skin depth δ (mm)
50 10,6
5k 1,06
10 k 0,750
20 k 0,530
500 k 0,106
In this project the frequency is fixed and was set at 10 kHz. Since it will be used a copper
conductor with solid round in section an estimate of the maximum value for its diameter
should be made in order to take advantage of the whole cross-section.
/) = 0,75

(5.16)
á = 2 × /) = 1,5

(5.17)
¢á = ; × /) 0 = 1,8 0

(5.18)
The values found correspond to the maximum diameter / cross section that the driver
must have to be totally negligible skin effect. The use of a larger conductor should take into
consideration that the resistance will be different for alternating current to DC resistance.
Considering the AWG measurement system, the driver who best fits this application is the
AWG14, which makes a diameter of 1.63 mm and a cross section of 2.05 mm2.
From the data available for this section we can determine that is enough for the
maximum current in the circuit (10 A).
5.5.5 Windings
After choosing the type and diameter of the electrical conductor to be used will be
important to determine how and where it will be wound. Since the core has been selected
previously is now necessary to define the coil former to be used. The manufacturer of ferrite
core (EPCOS) suggests the use of B67345 former together with UU 93/152/30 core, which
drawing and available characteristics are presented in Figure 5.9 and Table 5.18 respectively.
Table 5.18 Coil former characteristics [54]
Material GFR 6-polyamide
Number of sections 1
Winding cross section AN

1052
(mm2)
Average turn lenght lN

195
(mm)
Resistance factor;
6,4
AR = RCu/N2 (µΩ)
Figure 5.9 Coil former drawing
At this stage are already selected the number of turns, the wire used for winding and the
coil former. Considering these data it is possible to determine how many layers of winding are
necessary, to determine the mean turn length and therefore the required length of wire and
finally estimate its resistance.
Winding length (Figure 5.9):

L]hW = 86,2
(5.19)
Wire overall diameter:

∅ ¤ = 1,63 + 2 × W = 1,83
(5.20)
W − NT* T¥¦E ℎ*K¦NN ≈ 0,1

(5.21)
Maximum number of turns per layer (considering close wound assembly):
L]hW 86,2
= = ≈ 47 EN
∅¤ 1,83 (5.22)
84 Prototype design
Considering the number of turns previously calculated – 33 turns - is found that only one
layer is enough to achieve the desired winding.
At this stage it is already determined all the physical structure of the windings of the
transformer. As was done for the magnetic circuit is now necessary to estimate the power
losses for the electrical circuit. For this it is necessary to know the maximum rms current and
have an accurate estimate of the resistance of the windings.
Known the
e conductive material to be used (copper) and its cross-section
cross section becomes
necessary to know also the total length to determine its resistance.
The first step is to determine the mean turn length which is represented in Figure 5.10.
Figure 5.10 Mean turn length identification
Table 5.19 Turn segment dimensions
Segment Dimension
∅¤
36,4 + 2 × ∅ ¤ + 2 × = 36,4 + 3 × 1,83 = 41,89
2
A
∅¤
36,4 + 2 × = 36,4 + 1,83 = 38,23
2
B
33,6 + 2 × ∅ ¤ + 2 × = 33,6 + 3 × 1,83 = 39,09

∅©
0
C
∅¤
33,6 + 2 × = 33,6 + 1,83 = 35,43
2
D
5.5.6 Mean turn length
Primary winding
Tª-. = 2 × + 2 × « = 147,32 = 0,14732
(5.23)
Secondary winding
Tª-V. = 2 × M + 2 × Y = 161,96 = 0,16196
(5.24)
5.5.7 Total wire length
Number of turns
= 33
(5.25)
Primary winding
T ¤-. = Tª-. × = 0,14732 × 33 = 4,86
(5.26)
Secondary winding
T ¤-V. = Tª-V. × = 0,16196 × 33 = 5,34
(5.27)
5.5.8 Wire resistance
Copper resistivity at 20ºC

¬0)º = 1,68 × 10I Ω.
(5.28)
Wire cross section

¢ = 2,05 0 = 2,05 × 10I 0
(5.29)
Total resistance for primary winding

T ¤-. 4,86
g.-. = ¬ × = 1,68 × 10I × = 39,8 Ω
¢ 2,05 × 10I (5.30)
Total resistance for secondary winding

T ¤-V. 5,34
g.-V. = ¬ × = 1,68 × 10I × = 43,8 Ω
¢ 2,05 × 10I (5.31)
86 Prototype design
Once estimated the resistance of the windings and knowing the maximum rms current
that will circulate in the conductors is possible to estimate the losses by Joule effect.
<-. = ®¯° 0 × g.-. = 100 × 39,8 × 10IF = 3,98 L
(5.32)
<-V. = ®¯° 0 × g.-V. = 100 × 43,8 × 10IF = 4,38 L

(5.33)
<-¤. = <-. + <-V. = 3,98 + 4,38 = 8,36 L

(5.34)
In fact, the losses in the windings should be somewhat higher than calculated, due to skin
effect mentioned above. Once completed the transformer design their main characteristics
will be gathered and presented in Table 5.20.
Table 5.20 Designed transformer full specifications
Parameter Value
Core material N27
Core shape UU 93/152/30
Peak-to-peak induction 740 × 10IF -.
Magnetization inductance 5,85 × 10IF H
Primary 33
Number of
turns
Secondary 33
Transformation ratio 1:1
Wire type Round solid copper
Wire gauge AWG14
44,5 W @ 20 ºC
Core losses
29,7 W @ 100 ºC
Copper losses 8,36 W @ 20 ºC
Total losses 52,86 W @ 20 ºC
Frequency 10 kHz
5.6 Summary
The details of the prototype development were presented in this chapter. Switches were
selected and parameter setting of the hybrid driver was described in detail as it was the first
contact with such components.
For the passive components is also given special attention, particularly in the transformer
design. The transformer design was made taking into consideration the most important
aspects, resulting in the Table 5.20.
Chapter 6
Control system and results
In the previous chapter were designed the main components comprising the power stage
of the conversion system. Now it’s time to design the command and acquisition system to
correctly interact with the converter. The first step was to correctly identify the functional
modules and how they are interconnected with the power stage previously designed.
The modules defined as essential were voltage and current transducers, the acquisition
system, and the mechanism to generate phase-shifted square waves. Initially will be chosen
the current and voltage transducers to be used for the measurements. It is now a
precondition that is electrically isolated, as this is a bench top prototype and safety of the
experiments must be guaranteed. The other selection parameters are described at the time
of choice. Then will be studied the acquisition system already available in the laboratory. In
the end, it should be possible combining the characteristics of the transducers and this
acquisition system to characterize all of the measurement system in terms of ranges,
resolution and error analysis. A mechanism for generating phase shifted square waves with
high resolution will be described in detail and its resolution in degrees will be assessed.
The parameterization of the transformer will be performed based on a simplified model
and finally the first experimental results of the dual active bridge converter were presented.
In Figure 6.1 are diagramed the basic architecture of the system for better understanding.
The modules to develop and whose design is reported in this chapter are in the shaded
region.
88 Control system and results
Figure 6.1 Power converter modules
6.1 Measurements
The measures are of central importance in the design of a digitally controlled conversion
system. Most of the time signals with high bandwidth have to be measured under good
conditions, without affecting the operation of the converter and for many applications the
insulation is also a key requirement. With the growth of power electronics manufacturers
designed particular components with these characteristics and nowadays they are adapted to
perform this function.
Before choosing a specific component is important to understand what distinguishes them,
i.e. their performance measures. It is also necessary at this time to establish the
characteristics desired for these transducers.
Bandwidth - should be as high as possible. Even when it comes to measure DC signals, the
rapid variations in the measured signal should have an effect on the output of a transducer
without significant delay. The transducers to be used should also provide the correct
measurement of DC signals.
Insulation voltage - even with electrically isolated transducers, the dielectric strength
has limits. In the selected transducers this value should be above the maximum voltage that
can be found in the power stage (400 V neglecting transient overvoltages).
Measurement range - this range should include the whole range of variables to measure,
at least. A reasonable extra range can be added for greater flexibility and also providing a
possible operation of the converter in overload conditions for short periods.
Measurements 89
Table 6.1 Measured variables and nominal ranges
# Variable Range
1 Input current −10 + 10 M
2 Output current −10 + 10 M
3 Transformer primary current −10 + 10 M
4 Input voltage 0 + 400
5 Output voltage 0 + 400
Sensibility - Represents the relationship between the physical quantity measured and
sensor output. It is one of the basic characteristics of any transducer. Must be large to
measure small variations in the measured quantity with ease.
Linearity - It means that the transducer output varies linearly with the measured
quantity. If the sensor has a good linearity is reasonable to relate the input and output
through a gain. This is a very desirable feature because it avoids adding complexity to a
possible signal conditioning circuit.
Offset - represents the presence of a constant component in the output of the transducer,
regardless of what exists at the entrance. In most cases it is too low since the component was
designed so that this value was zero. If not negligible, should at least be a well-known value.
Some components with internal signal conditioning insert a specific value to adapt the output
range to typical input ranges found in common analog to digital converters.
Power consumption - Here it is necessary to refer to two types of transducers and they
were the active and the passive. Passive transducers extract energy only from the measured
quantity and in such cases it is necessary to assess whether the extracted energy is negligible
compared to the magnitude of the variable. In Active transducers the energy extracted from
the measured quantity, tends to be smaller, since there is a second power source. The
components with internal processing or conditioning are of this type preferentially. For this
case it is also important to assess if the total energy consumed by the transducer is low
compared with the magnitude of the power stage variables.
Taking into account all these parameters were then defined the current and voltage
transducers to be used.
Table 6.2 LEM® LV 25-P voltage transducer data [55]
Parameter Value
Bandwidth «Y to 10 KJ±
Insulation voltage 4,1 K
Measurement range Configurable from ±10 to ±1500
Output range up to ±12
Sensibility Depending on range from 5,72 / to 858 /
Linearity error < 0,2 %
Offset < 0,15 M
Power consumption 1,05 L
The range of input voltages and output of this component needs to be set by two resistors
placed on the outside, one in series with the primary and another in series with the
secondary.
The calculation of the resistance of the primary will be done already, since the maximum
voltage to be measured it is already known. From [55] is found that the maximum rms current
for the primary side is 10 mA.
So, this resistor will be determined in order to pass this current or less when the variable
being measured is at its maximum rms value.
aV -. 400

/ ≥ = = 40 KΩ
µ 0,01 (6.1)
Considering the most common range of resistor values (E12) the resistor selected will be
the first option above the minimum value determined, being that value 47 KΩ.
The resistor to put in series with secondary will be determined in the next step, as it is
necessary to know the input range of analog to digital converter.
Table 6.3 LEM® HY 15-P current transducer data [56]
Parameter Value
Bandwidth «Y to 50 KJ±
Insulation voltage 2,5 K
Measurement range ±15 MaV ; ±45 M
Output range ±4
4

= /M ≅ 266/M
15
Sensibility
Linearity error < ±1 % } µ

Offset < ±40
Power consumption 0,3 L
Unlike the voltage transducer, current transducer has all its parameters fixed. Therefore
there is no need for any calculation. If the output range is not suitable for the input range of
the acquisition system will be designed later a conditioning circuit.
Both components are to mount on printed circuit boards. Because this is a prototype was
considered appropriate to make a board for each transducer. Thus it is possible to easily
reconfigure the system of measures to completely match any control scheme.
The developed boards have the same form factor, so they can be mounted on the same
holders.
Each one also has three extractable connectors. In the high voltage side connector the
conductors are fixed with screws to ease future modification, for the transducers supply there
is a three way connector with voltages of (-15, 0, +15 V) and for the output signal was placed
a BNC type connector. These features ensure a truly modular measurement system. The
resistors needed for the operation of the voltage transducer are also mounted on the voltage
measurement boards.
Acquisition 91
6.2 Acquisition
In closed loop digitally controlled systems the quantification of the signals is the first
essential procedure. In fact for the control mechanism have a proper operation and a good
performance implies that the acquisition system is well implemented.
In the laboratory where it was developed this prototype there is a platform for rapid
prototyping which is already equipped with a analog signal acquisition system. The main
characteristics of this system are in Table 6.4.
After known these characteristics and selected the transducers, it is now appropriate to
determine whether it is possible a direct connection of signals. Given that all the necessary
data is available at this stage is important to assess if the resolution is enough to perform the
desired functions and how much the acquisition error represents in absolute terms of the
variables measured.
The platform has four parallel channels of analog to digital converters that can be
sampled simultaneously, and four groups of multiplexed acquisition each with four input
channels and one analog to digital converter. The system allows acquiring up to 8 channels
simultaneously, thus not being necessary to use the analog multiplexers since the measured
variables are five.
Table 6.4 Analog to digital converters specifications
Parameter Specification
Muxed ADCs (Figure 6.2) Parallel ADCs (Figure 6.3)
Number of Channels 16 4
Resolution 16 bit 16 bit
Input voltage range ±10 V ±10 V
Overvoltage protection ±15 V ±15 V
Conversion time 1 µs 800 ns
Offset error ±5 mV ±5 mV
Gain error ±0,25 % ±0,25 %
Offset drift 40 µv/K 40 µv/K
Gain drift 50 ppm/K 50 ppm/K
Signal-to-noise ratio > 83 dB > 83 dB
Figure 6.3 Parallel analog inputs
Figure 6.2 Multiplexed analog inputs

Table 6.5 Current acquisition parameterization
Parameter Value
Resolution 16 *
Input voltage range ±10
− g 20
ADC
¶= = = 305,18 µ
2 − 1 65535
Quantum
4

= /M ≅ 266/M
15
Sensibility
20
¶ 65535
¶ = = = 1,14 M
Current

4
Current “Quantum”
15
transducer
Output range ± 4 - ≤ ±10 .
Table 6.6 Voltage acquisition parameterization
Parameter Value
Resolution 16 *
Input voltage range ±10
− g 20
ADC
¶= = = 305,18 µ
2 − 1 65535
Quantum
Maximum voltage to measure aV -. = 400
Primary series resistor / = 47 KΩ
400
µ-. = = 8,51 M
47 × 10F
Primary current
2500
°µ-. = × = 21,3 M
1000 µ-.
Secondary current
10 ≥ ¯ × °µ-.
¯ ≤ ; ¯ ≤ 469,5 Ω *
/)
Voltage Maximum measuring
·f-.
transducer resistance

= × × 330;
= 17,55 /
/ 0q))
r×/)¸ /)))
Sensibility
20
¶ 65535
¶ = = = 17,4

17,55 × 10IF
Voltage “Quantum”
Output range + 7,03 - ≤ + 10 .
* The selected value is 330 Ω due to sensor restrictions referred on respective datasheet
As expected the resolution of analog to digital converters (16 bit) is quite enough to
acquire these signals. This can be proved by quantum obtained in absolute value of the
measured variable, both for the current or voltage transducer (¶ and ¶ ).
Finished this procedure can be concluded that the performance of this measurement
system is enough to perform the desired functions.
Acquisition 93
6.2.1 Error analysis

This section will analyze the measurement error of the set composed by the transducer
and the sample & hold ADC. The main objective of this analysis is to evaluate if the error
value is acceptable for this application and thus decide if this acquisition system is suitable.
For this purpose it was collected all available information about the errors of both
transducers and the ADC's and estimated a total value. It was found important to also refer
this value in absolute units of variables to measure in order to make values more intuitive to
be related to the values really measured.
Table 6.7 Voltage measurement error
0,2 %
Parameter Value
±0,15 M
Voltage transducer Linearity error
±¯ × 0,15 M = ±49,5

Offset
±5
Offset in Volt
±0,25 %
Multiplexed ADCs Offset error
-±49,5. + -±5. = ±54,5

Gain error
-0,2 + -±0,25.. = 0,45 %

Total Offset error
± -54,5 + 0,45%.
Linearity + gain error
± -3,1 + 0,45%.
All errors
Error referred to measured
variable
Table 6.8 Current measurement error
< ±1 % } µ
Parameter Value
< ±40
Current transducer Linearity error
±5
Offset
±0,25 %
Parallel ADCs Offset error
< ±45
Gain error
< ±1,25 %
Total Offset error
± -40 + 1,25 %.
Linearity + gain error
± -0,15 + 1,25%. M
All errors
Error referred to measured
variable
The error obtained in absolute value of the measures has a reasonable value and it is
understood that should not compromise a reliable reading of variables.
Nevertheless some conclusions can be made. The voltage transducer has a significant
offset error, although its influence is weakened as the absolute value of the variable
increases and approaches the nominal value (400 V), which represent less than 1% error at
that point. The proportional component of the error has a rather low value (for 400 V is less
than 2 V). The main error in the gain of this transducer is not in the component itself, but in
the resistors mounted externally. In fact if common resistors with 5% tolerance were used is
essential that at least one of them has fine adjustment to calibrate the transducer before
use.
In relation to the current transducer offset error can be neglected since the sensor is used
in its entire measurement range. Otherwise if it is used to measure very low currents there is
a component of uncertainty that can make the measure meaningless. The error in proportion
is higher than for the voltage sensor. Considering that the component has a nominal current
rating of 15A, the error when measuring the nominal value should represent equal parts, i.e.
the offset error and gain linearity error should have identical values. In this case there are no
external components that can insert error.
6.3 Phase-Shifted square wave generation

The control method to be used in this system is phase shifting the two bridges. Thus it is
necessary to create a precise mechanism for passing an absolute value of phase-shift in the
range of -90 º to +90 º generated in the digital control system to control logic signals that will
act on the drivers of the power stage. When the simulation was introduced was presented a
basic method of generating two square waves with control over the phase-shift. As stated at
that time the method was not feasible for practical implementation.
The method created at this stage combines internal components (in Slave DSP) and
minimal external components. The main component is the internal PWM generator of four
independent channels. Outside the logic is implemented by two XOR gates. The method
benefits from a low complexity and also from the high time resolution of internal PWM
generator. The operation of the phase-shift generator will then be explained separately for
direct and reverse power transfer modes (Figure 6.4).
In direct power flow mode the phase shift angle can range from 0 to 90º. The “Power flow
mode” bit is set at FALSE. The PWM A channel generates a left-aligned square wave with 50%
of duty cycle. The signal for Bridge 1 produced by first XOR gate is exactly the same as PWM
A. The switch is set at top position and the phase shift angle is directly converted to a duty
cycle, with 360º corresponding to 100%. The computed duty cycle is connected to channel B
and in channel C is inserted the duty cycle from B plus 50%. The XOR logic operation results in
a square wave for Bridge 2, with a phase-shift referred to channel A. That phase shift is equal
to “Phase-Shift Angle” input.
For reverse flow the operation is similar. The “Power flow mode” bit is set at TRUE. The
waveform generated by PWM channel A is now inverted due to the constant TRUE value on
the other XOR input. The switch changes from bottom position and in this position the
“Phase-Shift Angle” input is subtracted to an offset of 180º. This way with an input of 0º the
output is 180º degrees offset is null, since the channel is constantly shifted by 180 degrees.
When the “Phase-Shift Angle” input value increases, the angle used to calculate the duty
cycles B and C decreases, so that there is a negative phase-shift between Bridge 1 and Bridge
2 outputs.
Transformer parameterization 95
Figure 6.4 Phase-shift generator
Once designed and validated, the system showed a correct operation. Since the
characteristics of the PWM generation module are known including their temporal resolution
will then be given the resolution of this system in degrees.
Table 6.9 Phase-shift generator resolution
100 µN
Parameter Value
50 N
Switching period
for periods lower than 3,2 N

PWM module time resolution
50 × 10I
× 360 = 0,18º
100 × 10I
Phase-shift generator resolution in degrees
As was mentioned at the beginning, it was expected a high resolution of this system. This
way we can have increments as small as 0.18 degrees. For example, considering a phase shift
range of 0 to 90 degrees and a current range of 0 to 10 A, the minimum increment for
regulation will be just 0.02 A.
6.4 Transformer parameterization

As already mentioned, the transformer is of critical importance in this converter. During
the project the main characteristics were specified in order to have a component functional
and appropriate to this application. Yet some parameters are difficult to calculate accurately
because it depends on many factors.
At this stage the other components of the converter are ready to operate. The phase-shift
mechanism and the measurement system are also built and validated, and allow to measure
the main variables with isolation.
A transformer set was then mounted respecting as closely as possible the specifications
resulting from the project. With this transformer were performed several tests with the aim
of extracting critical parameters for the operation of the converter, especially the leakage
inductance which has a direct influence on the operation of the converter.
With these tests will be possible to determine a fairly realistic model of the transformer
to better characterize it and use the parameters for simulation upgrades.
6.4.1 Transformer model

Prior to start the tests in order to find the transformer parameters and create a model is
important to understand the considered model and what each parameter represents in fact.
At this stage, only three parasitic will be considered: the magnetization inductance, the
leakage inductance and the copper resistance.
The immediate model that represents such non-idealities is presented in Figure 6.5.
Figure 6.5 Practical transformer equivalent circuit (adapted from [57])
The achieved model has parameters in primary side and secondary side. Using the
transformation ratio is possible to place all of them on primary or secondary side. In this work
was defined that a primary referred model will be used.
Table 6.10 Transformer secondary to primary modification
Parameter Value
1 0
120,/ = 120 × G H
Secondary leakage inductance (primary referred)
120,/ 1
1 0
+ 0,/ = + 0 × G H
Secondary resistance (primary referred)
+ 0,/ 1
The conversion expressions of Table 6.10 result in a primary referred “T” model. This
model is presented in Figure 6.6.
Figure 6.6 Practical transformer equivalent circuit (primary referred) (adapted from [57])
Using a simplified model with combined parameters for the leakage impedance is possible
to achieve the model used for transformer parameterization (Figure 6.7). The expressions for
the combined parameters are presented in Table 6.11.
Figure 6.7 Simplified transformer equivalent circuit (primary referred) (adapted from [57])
Table 6.11 Simplified transformer equivalent circuit parameters
Parameter Value
Total equivalent leakage inductance (primary referred) 12 = 12/ + 120,/

12
+ = + / + +
Total equivalent resistance (primary referred)
+ 0,/
Considering this model is very simple to determine a s-domain transfer function for the
transformer. From Figure 6.7 can be concluded that the voltage applied to primary winding is
equal to the sum of voltage drops across 12, 1 and 1. The component of input voltage
transferred to secondary winding is the voltage drop across 1.
1 ¹+ -N.
J-N. = ×
1 ¹+ -N. + ¹+, -N. + + -N. (6.2)
N1
J-N. =
N-1 + 12. + + (6.3)
6.4.2 DC winding resistance measurement
The first test performed was to determine the resistance value of the copper windings,
with the aim to estimate the losses by Joule effect and extract this parameter to the model.
Due to its extremely low value would not be possible to measure with a common multimeter
or ohmmeter. The solution found is to put a DC source in series with a resistance of 5 ohms
and raise the test voltage so that the circuit current reaches 10 A. This value of current was
measured with an ammeter and was simultaneously measured the voltage drop of the
transformer winding. The resistance value is thus obtained simply by Ohm's law. The test was
performed for both windings.
Figure 6.8 Equivalent circuit for DC resistance measurement
Table 6.12 Results from DC winding resistance measurement
Parameter Value
Test current 10 A
Voltage drop 295 mV
Primary Calculated DC resistance

+ /
29,5 mΩ
Voltage drop 386 mV
Calculated DC resistance
+ 0
38,6 mΩ
Secondary
Primary referred
+ 0,/
38,6 mΩ
Total equivalent resistance (Table 6.11)

+
68,1 mΩ
Such values do not correspond closely with the value of the ohmic resistance of the wire
only, since the transformer was mounted on an experimental board and should be inserted
the values of the solder and connectors used.
The estimated power losses based of these values should be seen as representative of the
whole experimental set and not just of the transformer.
6.4.3 Short circuit test

The second test performed was with the secondary short circuited. Considering the
approximate model of the transformer when the secondary is short circuited the equivalent
circuit of the primary is a RL series circuit (Figure 6.9). The step response of this circuit must
determine its time constant and thus determine the value of leakage inductance, knowing a
priori the value of the resistance.
Figure 6.9 Equivalent circuit for shorted secondary
In fact this test cannot be done by directly connecting the power supply to the
transformer primary. This limitation is due to the fact that in the current in steady state is
limited only by the winding resistance, whose value is just some mΩ. Moreover using this very
low resistance value could insert large errors in the inductance value determined; because it
is difficult to measure the resistance accurately. The test was then performed with an
external testing resistor (RT).
15
10
-5
-10
-15
0,00E+00 5,00E-05 1,00E-04 1,50E-04 2,00E-04
Time (s) Voltage x0.1 Current
Figure 6.10 Primary winding voltage and current with sorted secondary
Processing data from Figure 6.10 to get one of the rising edges so that the exponential
shape can start from t=0 and I=0, it’s possible to perform a curve fitting. This result is
presented in Figure 6.11. The expression of current vs. time for a series RL circuit is given in
(6.4).
14
12
10
Measured Interpolated
Current (A)
0
0,00E+00 1,00E-05 2,00E-05 3,00E-05 4,00E-05 5,00E-05
Time (s)
Figure 6.11 Step response with shorted secondary.
®
*-. = *9∞: × G1 − ¦ It + xW H
(6.4)
This interpolation was performed by a solver with the aim of minimizing the absolute
difference in relation to the set of points obtained from measurement, adjusting for this the
L parameter of (6.1). This can only be done if *9∞: and are known parameters. The current
for steady state is determined by the average current value after considered steady (20 to 48
µs). It is important to report that the voltage step applied to the winding have a small drop
with the increase of the current. Thus, consider magnitude of the square wave equal to the
DC bus voltage inserts an error on the measurement.
For the resistance value estimation was used a method consisting in apply a DC current to
all the circuit and measure the voltage drop across the used components (winding resistance,
external resistor, wires, etc.). This method is described and used in the next section. The
results obtained by this numerical method for the leakage inductance are presented in Table
6.13.
Table 6.13 Parameters from the step response experiment
Parameter Value
Measured resistance = W + + 12,26 Ω
Current final value i[∞] 13,0 A
Equivalent Leakage inductance 53,7 µH

At this time the transformer leakage impedance, referred to the primary, is already
determined.
6.4.4 Open circuit test
After determined the series impedance of the transformer, i.e. the leakage inductance
and the winding resistance, is now important to determine the magnetization inductance.
This test can be performed applying a square wave of considerable magnitude to primary
winding, and measure the current on the same winding.
Figure 6.12 Equivalent circuit for open secondary
In this case is expected that current have a small value and with a triangular shape, which
means that for the selected frequency the transformer is away from saturation. This
experiment should allow the estimation of the magnetization inductance.
-1
-2
-3
0,00E+00 5,00E-05 1,00E-04 1,50E-04 2,00E-04
Figure 6.13 Input voltage and current with open circuit secondary
In fact, not only the magnetization current appears on primary waveform. Considerable
oscillations after switching are visible and difficult the direct reading of the peak
pe current
values. For this purpose some segments were added to Figure 6.14 to identify the
magnetization current and allow the estimation of the peak values.
-1
-2
-3
0,00E+00 5,00E-05 1,00E-04 1,50E-04 2,00E-04
Figure 6.14 Magnetization current identification (dashed)
After identified the magnetization current is important to find a method to estimate the
peak values of the triangular waveform. An average method consisting in overlap several
segments and fit the curve with a linear function was used.
1,5
0,5
-0,5
-1
0,00E+00 1,00E-05 2,00E-05 3,00E-05 4,00E-05 5,00E-05
Time (s) Current Linear (Current)
Figure 6.15
15 Slope estimation of magnetization current
Table 6.14 Peak values of magnetization current
Parameter Value
Positive peak + 0,5 M
Negative peak − 0,5 M
Once determined the value of the peak-to-peak current is possible to estimate the
primary inductance, as it has a direct relationship with the slope of the line.
*-.
+ -. = 1
(6.5)
º*
+ = 1
º (6.6)
º × + 50 × 10I × 375
1/ = = = 18,75 × 10IF J
º* -500 × 10IF . − -−500 × 10IF . (6.7)
This value stands for the total inductance of the primary circuit, i.e. the sum of leakage
and magnetization inductances. Since the leakage inductance is previously determined, the
difference can be calculated to determine a value for 1 .
1 = 1/ − 1,
(6.8)
1 = 18,75 × 10IF − 53,7 × 10I = 18,70 × 10IF J

(6.9)
As can be seen the leakage inductance has a value negligible compared to the
magnetizing inductance. This means that the transformer makes a good magnetic coupling
between primary and secondary.
The Y» capacity was not determined in these tests. Completed this process and thus we
can consider transformer parameterized in its three main components, the leakage
inductance, winding resistance and magnetization inductance.
6.4.5 Model and validation
After determined an approximate model the parameters should be inserted on a

simulation model so the tests can be performed and results compared. This way the model
can be validated and subsequently used in other simulations.
The parameterized model is presented in Figure 6.16 and was implemented in Simulink®
to validate results. Those waveforms obtained are very similar to the measures in reality
using the oscilloscope.
Figure 6.16 Transformer model with parameters
N18,75 × 10IF
J-N. =
N-18,75 × 10IF + 53,7 × 10I . + 68,1 × 10IF (6.10)
Figure 6.17 Transformer transfer function
The first simulation was to be held with the secondary shorted. The waveform has the
same format, same time constant and an approximate saturation value. The current
magnitude difference is due to voltage drop present on laboratory DC power supply, which is
not present in simulation because an ideal DC voltage source was used. The test was
performed in conditions as close as possible from the practical test. Obviously the oscillations
are not present because in simulation was not considered any capacity.
15
10
-5
-10
-15
5,00E-05 1,00E-04 1,50E-04 2,00E-04
Time (s)
Voltage x0,1 Current (Simulated Model) Current (Measured)
Figure 6.18 Simulated current with shorted secondary
Once verified the waveform for the short circuit test will then observed the waveform for
open circuit. The waveform is expected to be a triangular shape without oscillation as before.
The values of current peak to peak amplitude should be identical to a practical test.
Oscillations are related to winding capacitive effects and are not represented in the
considered model. The test was performed with an identical voltage (±375V).
-1
-2
-3
0,00E+00 5,00E-05 1,00E-04 1,50E-04
Time (s)
Voltage x0,005 Current (Simulated Model) Current (Measured)
Figure 6.19 Simulated magnetization current

Completing this process is concluded that the model used for the transformer is not very
far from reality and that the selected tests allow getting valuable data. From the verification
through simulation it is possible to say that this model adds a high value to the development
process of the converter.
6.5 Experimental results

The experimental results are considered the ultimate aim of this work. The state of
development at this stage allowed the converter to extract some relevant waveforms. The
first to be presented are not related to the converter as dual active bridge, and are only for
testing the power transformer.
In the second section the command signals and the transformer primary current are
presented and analyzed, this turn operating as dual active bridge.
6.5.1 Transformer load tests
At this stage it's time to check how transformer responds to power levels for which it was
designed. The power of 4kW cannot be achieved due to limitations of the power supply,
which can only provide up to half of that power. The test was performed with a DC power
supply connected to one of the full bridges. The AC terminals of the bridge were connected
to the transformer primary and secondary to a resistive load (Figure 6.20). The tests were
performed for various input voltage levels, because it was considered interesting to see the
operation under light and heavy load conditions.
Figure 6.20 Experimental setup for load tests
The first test was performed with a voltage of ±80 V and a resistance of 48.8 Ω which
means a current with amplitude of 1.6 A. Resulting waveforms of the voltage and current in
the transformer primary are shown in Figure 6.21.
Experimental results 107
10
8
6
4
2
0
-2
-4
-6
-8
-10
0,00E+00 5,00E-05 1,00E-04 1,50E-04 2,00E-04
Figure 6.21 Light load test (120 W)
The oscillations observed during the parameterization procedures are still present in
these tests. This phenomenon should lead to a further study since its influence on the
converter is not properly evaluated. At this point it is important to observe if this
phenomenon remains limited for high power levels.
Several tests were then performed with different input voltages and the same load,
resulting in a quadratic increase of power. The waveforms obtained in all these tests are
obviously not presented since they not contain any additional information.
The waveform shown in Figure 6.22 is for the maximum power level for which the tests
were performed.
15
10
-5
-10
-15
-20
0,00E+00 5,00E-05 1,00E-04 1,50E-04 2,00E-04
Figure 6.22 Heavy load test (1.5 kW)

This time the oscillations appear with higher amplitude. In fact during the tests it was
found that their value has a variation proportional to the amplitude of the current and with a
limited value.
From these tests can be concluded that the transformer has a proper operation and can
operate on high power range and with high power values. Some critical aspects such as
vibration, noise and overheating have been verified as being stable and reasonable for a
laboratory assembly.
Some improvements especially in fixing the two core parts may be of interest to reduce
the vibration of the transformer.
6.5.2 Converter operation

Once determined the correct operation of all components constituting the converter is
the first time to carry out tests for its operation as phase-shifted dual active bridge. These
tests were performed with one of the bridges powered by a DC power supply and the other
one delivering power to a battery pack (load).
The tests might be performed not only putting the transformer between the bridges
(Figure 6.23). In reality, the leakage inductance of only a few µH would result in very large
growth rates of the current every time there was a phase-shift between bridges. An
inductance was added in series with the transformer primary to increase the value of the
equivalent inductance. The inductor used is air-core and was placed far away enough to not
have mutual influence with the transformer.
External
inductor
DC
Figure 6.23 Converter schematic with external inductor
The inductor value added was measured and has an inductance of 327,9 µH. Recalling
model value for the leakage inductance can now be determined the total equivalent leakage
inductance. This component was used for having an inductance value and a proper current
rating, due to limited availability of these components in the laboratory.
1, WW] = 1, + 1, hW. = 53,7 × 10I + 327,9 × 10I = 381,6 × 10I J

(6.11)
This value is still below the maximum value determined for the leakage inductance which
was 1 mH. On the other hand isn’t too low, not allowing the current slope to be very high.
Experimental results 109
This test was intended to observe the operation in buck and boost modes. This
observation is easily made if the input voltage is slightly above the output, since the voltage
drops due to current flow rapidly make the converter passing from buck mode to boost mode.
The tests were performed for various values of phase-shift.
Figure 6.24 Transformer primary current =5º Figure 6.25 Transformer primary current =10º
Summary 111
In Figure 6.24 to Figure 6.29 are presented the waveforms collected from the
experimental tasks performed. In the first two plots we can see that the converter is in buck
mode, i.e. the input voltage is greater than the output voltage. The transformer has no
influence in this relationship since it has a transformation ratio of 1:1.
With 15 degrees voltages are very close to each other. This obviously results in a zero
slope instants when the bridges are not phase-shifted.
In the following plots (Figure 6.26 to Figure 6.29) the converter is already operating in
boost mode, since the high current flow dropped the power supply voltage that is not
regulated. Due to the higher values of phase-shift is now easy to observe these periods and
see the current varying fast. Another important observation is that amplitude of the
oscillations does not increase proportionally to the current. This way can be considered that
this phenomenon loses its meaning as the power level increases.
We can then conclude that the converter works correctly in the way of buck and boost
operation and high power levels.
6.6 Summary
In this chapter are designed the phase-shift control system and the measurement system.
After validated were used to control the converter. Therefore tests were then conducted,
first for parameterization and load testing of the transformer and ultimately to operate as
dual active bridge converter.
Chapter 7
Conclusions and future work
7.1 Conclusions
From an analysis of the converter has been concluded that provides interesting features
for the implementation of this application. The simulations allowed viewing the main
waveforms and having the first contact with the operation of the converter. Features such as
the current output does not depend on the output voltage (ideally) and its symmetry in terms
of phase-shift vs. power transferred in both directions make the same control algorithm can
be used to control the converter in the whole power range and both directions. The ripple
observed in the batteries is perfectly reasonable and enables the usage of a small filter giving
the converter possibility of being a compact device. The comparison of results of the average
current output between simulation and analytical expression showed that the simulation
model correctly implements the topology.
The design procedure followed for the power stage was found to be a simplified approach,
but not compromising the final results. The option for a ready to use IGBT drives was a good
solution, considering that the process of develop new devices can be difficult and is out of
the scope of this work. The power switches have not been analyzed in depth in this design.
During operation of the converter was not observed any negative impact associated with this
approach. The main conclusions that can be taken from the power hardware project are
about the transformer. In fact during the parameterization is concluded that the project
should be even more careful because there is influence of many parameters during operation.
The capacity between turns should have been considered during this project in order to be
minimized. All other parameters that were designed revealed a correct specification, which
was confirmed by experimental results.
The selected physical structure consisting of a double-U core and two concentric single
layer windings offer good magnetic coupling, with the leakage inductance having a small
value. This way can be added an external inductor in series to perform a desired total
equivalent inductance.
The measurement system has been used at this stage to measure the main variables of
interest of the prototype and revealed an adequate bandwidth and sensitivity. The estimated
offset error of the transducers has been verified in practice. The acquisition system has not
yet been used but will be of great interest to implement a closed-loop control system.
114 Conclusions and future work
The command system for the phase-shift generation was shown to have a very good
resolution for the regulation of the converter. Although manually operated allows smooth and
precise variations of the power transferred by the converter. The simplified solution which
has been developed has proven to be a great choice to implement this type of control using
PWM generator module, which are commonly available in DSPs and MCUs.
Regard the tests utilized for the transformer, those selected was shown to be adequate
and able to collect valuable data for the parameterization of the transformer. They require
minimal additional equipment/components and allow for a parameterization with sufficient
accuracy for this stage of development. This allowed the development of a simplified
simulation model in Simulink® which validate the parameters obtained. The model to match
the reality should include parasitic capacitance with the aim of modeling the oscillations.
The experimental results obtained during the transformer load tests had the expected
results demonstrating proper operation within a wide range of power. Finally the test done
with the converter operating as dual active bridge, using the batteries as load and an inductor
in series with the transformer primary, demonstrated an operation in accordance with the
theory described and the results verified in simulation. The buck and boost operation modes
are observed with the expected changes on transformer primary current waveform.
Finally, this converter offers key advantages to the vehicle-to-grid system with a
reasonable complexity, implementing at least two of the critical requirements – insulation
between grid and vehicle, and power bidirectionality. The potential to have a very high
efficiency allows it to be compact, which is extremely important for an on-board charger.
This feature can also be used to allow very high switching frequencies (hundreds of kHz) to
reduce the ripple in the batteries, which is of great interest. The functionality of current
control for battery charging can be implemented by this converter, since the relationship
between the control variable and the average current in the batteries it is simple and deeply
studied.
7.2 Future work

Although the converter is already functional, is not yet ready to be used for the purpose
which is designed for. Some issues like evaluate the efficiency and further study of certain
phenomena should be performed.
Related to transformer, the study of its oscillations during the switching should be
considered to optimize the design. Regard the output current to the batteries, it still has a
very high oscillation component, which should lead to the study of the factors that influence
this parameter and possibly demand for new filter solutions. The relationship between the
phase-shift and the ripple should be studied, as was observed in simulation that this influence
exists.
Once solved the previous problems related to the operation of the power stage, the
converter dynamics must be studied in deep and a closed-loop control should be developed to
regulate the output according to a given reference. The interface of this control algorithm
with the converter is already implemented by the measurement and acquisition system and
the command system.
References
1. Abbott, D., Keeping the Energy Debate Clean: How Do We Supply the World's Energy
Needs? Proceedings of the IEEE, 2010. 98(1): p. 42-66.
2. Zhu, Z.Q. and C.C. Chan. Electrical machine topologies and technologies for electric,
hybrid, and fuel cell vehicles. in Vehicle Power and Propulsion Conference, 2008.
VPPC '08. IEEE. 2008.
3. Pecas Lopes, J.A., P.M. Rocha Almeida, and F.J. Soares. Using vehicle-to-grid to
maximize the integration of intermittent renewable energy resources in islanded
electric grids. in Clean Electrical Power, 2009 International Conference on. 2009.
4. Lopes, J.A.P., F.J. Soares, and P.M.R. Almeida, Integration of Electric Vehicles in the
Electric Power System. Proceedings of the IEEE, 2011. 99(1): p. 168-183.
5. Kisacikoglu, M.C., B. Ozpineci, and L.M. Tolbert. Examination of a PHEV bidirectional

charger system for V2G reactive power compensation. in Applied Power Electronics
Conference and Exposition (APEC), 2010 Twenty-Fifth Annual IEEE. 2010.
6. Markel, T., M. Kuss, and M. Simpson. Value of plug-in vehicle grid support operation.
in Innovative Technologies for an Efficient and Reliable Electricity Supply (CITRES),
2010 IEEE Conference on. 2010.
7. Cvetkovic, I., et al. Future home uninterruptible renewable energy system with
vehicle-to-grid technology. in Energy Conversion Congress and Exposition, 2009. ECCE
2009. IEEE. 2009.
8. Larminie, J. and J. Lowry, Electric Vehicle Technology Explained. 2003: John Wiley &
Sons.
9. Valsera-Naranjo, E., et al. Electrical vehicles: State of art and issues for their
connection to the network. in Electrical Power Quality and Utilisation, 2009. EPQU
2009. 10th International Conference on. 2009.
10. Agency, I.E., Technology Roadmap - Electric and plug-in hybrid electric vehicles.
2009.
116 References
11. Zpryme, Smart Grid Insights. 2010.
12. Interoperability, O.o.t.N.C.f.S.G., NIST Framework and Roadmap for Smart Grid
Interoperability Standards, Release 1.0. January 2010, National Institute of Standards
and Technology.
13. SAE. SAE Automotive Standards. 2011 [cited 2011 January 27]; Available from:
http://standards.sae.org/automotive/.
14. UL. Underwriters Laboratories | Standards for Safety. 2011 [cited 2011 January 27];
Available from: http://www.ul.com/global/eng/pages/corporate/standards/.
15. IEC, IEC 61851-1 Electric vehicle conductive charging system - Part 1: General
Requirements. 2010.
16. IEC, IEC 61851-21 Electric vehicle conductive charging system - Part 21: Electric
vehicle requirements for conductive connection to an a.c./d.c. supply. 2010.
17. IEC, IEC 61851-22 Electric vehicle conductive charging system - Part 22: AC electric
vehicle charging station. 2010.
18. IEC, IEC 62196-1 Plugs, socket-outlets, vehicle couplers and vehicle inlets –
Conductive charging of electric vehicles – Part 1: Charging of electric vehicles up to
250 A a.c. and 400 A d.c. 2010.
19. Comission", I.E., IEC 61851-1 Electric vehicle conductive charging system - Part 1:
General Requirements. 2010.
20. MENNEKES, MENNEKES and electric vehicle.
21. REMA, Charger Connectors for Electricle Vehicles (EV’s).
22. Yazaki. Charging connectors | Charging connectors of the Yazaki Group. 2011 [cited
2011 15-01-2011]; Available from: http://charge.yazaki-group.com/english/product/.
23. CHAdeMO. CHAdeMO Association Website. 2011 [cited 2011 February 10]; Available
from: http://www.chademo.com/.
24. GE. Electric Vehicle Equipment from GE. 2011 [cited 2011 February 10]; Available
from: http://www.geindustrial.com/products/static/ecomagination-electric-
vehicles/ge-wattstation.html.
25. GE, WattStation TM - Fact Sheet. 2010.
26. AV, AeroVironment™ EV Solutions™ Fleet Fast Charging Station, 50kW DC Secure,
Safe, Fast Charging For Your Hardworking Fleet Model EV50-FS. 2010.
References 117
27. Weise, N.D., K.K. Mohapatra, and N. Mohan. Universal utility interface for Plug-in
Hybrid electric vehicles with vehicle-to-grid functionality. in Power and Energy
Society General Meeting, 2010 IEEE. 2010.
28. Bilgin, B., A. Emadi, and M. Krishnamurthy. Design considerations for a universal
input battery charger circuit for PHEV applications. in Industrial Electronics (ISIE),
2010 IEEE International Symposium on. 2010.
29. Rei, R.J., et al. Grid interactive charging control for plug-in electric vehicles. in
Intelligent Transportation Systems (ITSC), 2010 13th International IEEE Conference
on. 2010.
30. Grenier, M., M.G. Hosseini Aghdam, and T. Thiringer. Design of on-board charger for
plug-in hybrid electric vehicle. in Power Electronics, Machines and Drives (PEMD
2010), 5th IET International Conference on. 2010.
31. Zhou, X., et al. Design and control of grid-connected converter in bi-directional
battery charger for Plug-in hybrid electric vehicle application. in Vehicle Power and
Propulsion Conference, 2009. VPPC '09. IEEE. 2009.
32. Xiaohuz, Z., et al. Multi-function bi-directional battery charger for plug-in hybrid
electric vehicle application. in Energy Conversion Congress and Exposition, 2009.
ECCE 2009. IEEE. 2009.
33. Jaganathan, S. and G. Wenzhong. Battery charging power electronics converter and
control for plug-in hybrid electric vehicle. in Vehicle Power and Propulsion
Conference, 2009. VPPC '09. IEEE. 2009.
34. Erb, D.C., O.C. Onar, and A. Khaligh. Bi-directional charging topologies for plug-in
hybrid electric vehicles. in Applied Power Electronics Conference and Exposition
(APEC), 2010 Twenty-Fifth Annual IEEE. 2010.
35. Haghbin, S., et al. Integrated chargers for EV's and PHEV's: examples and new
solutions. in Electrical Machines (ICEM), 2010 XIX International Conference on. 2010.
36. Lisheng, S., A. Meintz, and M. Ferdowsi. Single-phase bidirectional AC-DC converters
for plug-in hybrid electric vehicle applications. in Vehicle Power and Propulsion
Conference, 2008. VPPC '08. IEEE. 2008.
37. Mohan, N., T.M. Undeland, and W.P. Robbins, Power Electronics - Converters,
Applications, and Design (3rd Edition). 2003: John Wiley & Sons.
38. Rashid, M.H., Power Electronics Handbook. 2001: Academic Press.
39. Kheraluwala, M.N., et al., Performance characterization of a high-power dual active

bridge. Industry Applications, IEEE Transactions on, 1992. 28(6): p. 1294-1301.
40. Kheraluwala, M.H., et al. Performance characterization of a high power dual active
bridge DC/DC converter. in Industry Applications Society Annual Meeting, 1990.,
Conference Record of the 1990 IEEE. 1990.
118 References
41. Alonso, A.R., et al. An overall study of a Dual Active Bridge for bidirectional DC/DC
conversion. in Energy Conversion Congress and Exposition (ECCE), 2010 IEEE. 2010.
42. Demetriades, G.D., On small-signal analysis and control of the single and the dual
active bridge topologies. 2005, KTH, Sweden: Stockholm.
43. Blake, C. and C. Bull IGBT or MOSFET: Choose Wisely.
44. SEMIKRON, SK100GB12V SEMITRANS 2. 2010.
45. Herzer, R., et al. (2010) Integrated Gate Driver Circuit Solutions.
46. SEMIKRON, SKYPER® 52 R Driver Core - Technical Explanations 2010.
47. SEMIKRON, SKYPER® 32PRO R Driver Core - Technical Explanations 2007.
48. SEMIKRON, Evaluation Board 1 SKYPER® 32PRO R - Technical Explanations 2008.
49. EPCOS, Aluminum electrolytic capacitors - General technical information. 2010.
50. EPCOS, Aluminum electrolytic capacitors with screw terminals - Series/Type: B43456,
B43458. 2010.
51. EPCOS, Film Capacitors - General technical information. 2009.
52. Bossche, A.V.d. and V.C. Valchev, Inductors and Transformers For Power Electronics.
2005: CRC Press.
53. Erickson, R.W. and D. Maksimovic, Fundamentals of Power Electronics (Second

Edition). 2001: Springer.
54. SIEMENS, Ferrites and Accessories. p. 610.
55. LEM, Voltage Transducer LV 25-P.
56. LEM, Current Transducers HY 5 to 25-P.
57. Mora, J.F., Máquinas eléctricas (Quinta Edición). 2003: McGRAW-HILL.

Annex A – Workplan
120 Annex A – Workplan
Annex A – Workplan 121

Veículo Elétrico Com Interface para A Rede Elétrica Grid Interface For An Electric Vehicle

Uploaded by

Copyright:

Available Formats

Veículo Elétrico Com Interface para A Rede Elétrica Grid Interface For An Electric Vehicle

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Veículo Elétrico Com Interface para A Rede Elétrica Grid Interface For An Electric Vehicle

Uploaded by

Copyright:

Available Formats

Faculdade de Engenharia da Universidade do Porto

Veículo elétrico com interface para a rede

Luís Miguel Faria Miranda

Dissertation carried out under the

Supervisor: Prof. Doutor Rui Manuel Esteves Araújo

O campo dos conversores DC-DC isolados é uma area importante da electrónica de

Aos meus colegas de curso, em particular aos finalistas da especialização em Electronica e

À Joana, cujo o amor, apoio e compreensão estiveram sempre presentes, principalmente

Agradecimentos ................................................................................ vii

List of Figures ................................................................................. xiii

List of Tables ................................................................................. xvii

Acronyms ....................................................................................... xxi

Chapter 7 ...................................................................................... 113

References .................................................................................... 115

Annex A – Workplan ......................................................................... 119

Figure 1.2 Basic diagram of Battery Electric Vehicle (BEV) ............................................2

Figure 1.3 EV interface evolution ...........................................................................4

Figure 1.4 V2G based system ................................................................................5

Figure 2.1 Conceptual architecture of a battery electric vehicle [8] .............................. 10

Figure 2.4 Battery costs through 2020 [10] ............................................................. 14

Figure 2.5 US V2G Unit Forecast [11] .................................................................... 14

Figure 2.7 Mennekes® connector drawing and pinout [20] ........................................... 18

Figure 2.8 Commercial charging station [24] ........................................................... 21

Figure 2.9 Residential EVSE [24] .......................................................................... 21

Figure 3.1 Full Bridge PWM rectifier/inverter .......................................................... 34

Figure 3.2 Single-phase multilevel rectifier/inverter ................................................. 35

Figure 3.3 Half-Bridge converter (Buck/Boost) ......................................................... 36

Figure 3.4 Dual active bridge converter [39] ........................................................... 37

Figure 3.5 Chart of alternatives for the converter architecture .................................... 38

Figure 3.6 System architecture ............................................................................ 42

Figure 4.2 Fundamental model [41] ...................................................................... 44

Figure 4.3 Equivalent circuit 1 ............................................................................ 45

Figure 4.5 Equivalent circuit 3 ............................................................................ 46

Figure 4.6 Equivalent circuit 4 ............................................................................ 46

Figure 4.7 Fundamental waveforms [41] ................................................................ 47

Figure 4.10 Soft switching region vs. Phase-shift and M .............................................. 50

Figure 4.11 Simulation model ............................................................................. 51

Figure 4.14 Output current vs. phase-shift ............................................................. 55

Figure 4.19 Output voltage ripple ........................................................................ 58

Figure 5.2 Considered equivalent circuit ................................................................ 68

Figure 5.3 5 to 15V circuit ................................................................................. 71

Figure 5.4 Internal equivalent circuit of HALT signal ................................................. 71

Figure 5.5 HALT signal interface circuit ................................................................. 72

Figure 5.7 Transformer simplified equivalent model (primary referred) .......................... 76

Figure 5.8 Transformer design flowchart ................................................................ 77

Figure 5.9 Voltage waveform applied to the transformer ............................................ 80

Figure 5.10 Coil former drawing .......................................................................... 83

Figure 5.11 Mean turn length identification ............................................................ 84

Figure 6.2 Multiplexed analog inputs ..................................................................... 91

Figure 6.3 Parallel analog inputs .......................................................................... 91

Figure 6.4 Phase-shift generator .......................................................................... 95

Figure 6.5 Practical transformer equivalent circuit (adapted from [57]).......................... 96

Figure 6.8 Equivalent circuit for DC resistance measurement ....................................... 98

Figure 6.9 Equivalent circuit for shorted secondary ................................................... 99

Figure 6.11 Step response with shorted secondary. .................................................. 100

Figure 6.12 Equivalent circuit for open secondary ................................................... 101

Figure 6.14 Magnetization current identification (dashed) ......................................... 102

Figure 6.15 Slope estimation of magnetization current ............................................. 102

Figure 6.16 Transformer model with parameters ..................................................... 104