Design and Implementation of A Sic Based Contactless Battery Charger For Electric Vehicles

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Design and Implementation of a SiC based

Contactless Battery Charger for Electric Vehicles


I. Villar, U. Iruretagoyena, A. Rujas, A. Garcia-Bediaga, I. Perez de Arenaza
Control Engineering and Power Electronics Area
IK4-IKERLAN Technological Research Centre
20500 Arrasate-Mondragon, Spain

Abstract—Wireless charging is gaining interest worldwide, large amounts of power over large distances [3].
mostly driven by electric vehicles (EVs) and consumer
electronics. This paper presents the design and implementation of
a contactless battery charger for electric vehicles. The power
converter is based on SiC MOSFET devices and efficient power
topologies, enabling to increase the operating frequency with the
aim of minimizing system volume and weight. The battery
charger is a 3.2kW-100kHz contactless inductive power transfer
system, designed for a domestic charger application.
Keywords—inductive power transfer, SiC, electric vehicle
charger.

I. INTRODUCTION
Induction Power Transfer (IPT) technology is commonly
put into practice in low-power converters thanks to the
availability of fast and efficient semiconductors. Nowadays, in
medium and high-power applications this fact is becoming a Fig. 1: Contactless battery charger for electric vehicles.
reality due to the development of new wide-band gap
semiconductor devices (with lower conduction resistance, Thanks to the input interleaved PFC converter, the total
higher blocking capabilities and better frequency behavior). harmonic distortion (THD) of the input current can be reduced
IPT is widely discussed as a charging technology for the complying with harmonic injection standards and controlling
Electric and Hybrid Electric Vehicles (EV, HEV) [1]-[3]. In at the same time the primary side dc-link voltage. The
this application, IPT offers an advantage compared to secondary Boost converter enables secondary side dc-link
traditional charging infrastructure as it significantly simplifies voltage regulation, and therefore the control of the transferred
the charging process, without the need of any physical power from the EV.
connection between the vehicle and the charging point.
One of the main drawbacks of the SS compensation is the
Power density is an essential factor in high-power on- voltage gain dependency with the load and the associated
board energy-conversion systems which have strong volume control difficulties. However, the transferred power depends
and weight limitations. High operation frequency along with directly on both bus voltages (1), as well as the mutual
semiconductors with very good frequency performance can inductance of the coils and the angular frequency.
provide interesting features, such as efficiency improvement,
higher power density or cooling system simplification [4]. 8 Vdc1 Vdc 2
P= · (1)
This paper analyzes the design and implementation of a π2 ωM
contactless battery charger for electric vehicles, exactly a where Vdc1 and Vdc2, are the primary and secondary bus
series-series (SS) inductive power transfer system, intended voltages respectively (see Fig. 2), Ȧ is the angular frequency
for a domestic use. The battery charger circuit is composed of and M the mutual inductance between the inductive coils.
several step converters: a passive rectifier with an interleaved
Power Factor Correction (PFC) converter, a series-series
converter with an H-bridge and a second passive rectifier, and
finally, a second interleaved Boost converter (see Fig. 1).
The series-series (SS) configuration is a common choice in
high-power contactless IPT converters, which is especially
well suited (with high-frequency features) for transferring
Fig. 2: Contactless charger topology.

978-1-4673-7151-3/15/$31.00 ©2015 IEEE 1294


With the SS configuration and with two external dc normalized current ripple in terms of the duty cycle is
voltage controllers (see Fig. 2) the contactless charger can illustrated in Fig. 3.
transfer energy with high efficiency over a wide load range 1
avoiding the previously mentioned control difficulties. 1
2
channel
channels
Moreover, the tolerance against position misalignment of the 0.8

Normalized Current Ripple


3 channels

inductive power transfer system is very good, because when


4 channels
5 channels

the coupling factor varies, i.e. the mutual inductance changes, 0.6

the voltages on both dc sides will be changed in order to 0.4


control the transferred power.
0.2
In the converter topology illustrated in Fig. 2, the primary
side voltage will vary from 300 V-450 V, while the secondary 0
0 0.25 0.5 0.75 1
side voltage will vary according to the charging requirements Duty cycle

of the on-board battery. Even with a limited voltage control in


the primary side PFC converter, with the wide voltage control Fig. 3: Current ripple in terms of duty cycle.
range of the on-board Boost converter a large number of
possible voltage and current combinations will be created Regarding the secondary side Boost converter, the current
before reaching the maximum electrical ratings of the charger. ripple expression and the influence of the number of channels
Moreover, the primary bus voltage controlled enables soft- (see Fig. 2) is also valid, but the design criterion in this case is
switching conditions for the primary H-bridge because there is different. The current ripple of inductors has to be bounded to
no need of leg phase shifting for power control and therefore minimize the inductor and also the secondary dc-link capacitor
the converter is operated in soft-switching conditions. requirements.

The power converter developed within this paper has been Designers have two important design criteria: the inductor
build for experimental verification purposes. Even if the requirements (saturation current and losses) and the maximum
1.2kV SiC MOSFETs are not the best choice for a single grid current ripple. Regarding the worst cases, the former is
phase power converter, the semiconductor choice will give analyzed with a duty cycle equal to a half, while the latter
experimental data information and further knowledge for depends on the selected number of phases; for example, the
future high power inductive power transfer developments. critical duty cycle (i.e. the maxim grid current ripple) of a two
phase interleaved PFC is found with 0.25 and 0.75 (see Fig.
3).
II. INTERLEAVED POWER FACTOR CORRECTION
The goal of the power factor correction is to control the Regarding the dc-link capacitor, the voltage ripple has to
transferred power (or the dc voltage) and at the same time be bounded with the aim of achieving a correct operation of
absorb a sinusoidal current in phase with the grid voltage. This the following converters. The low frequency (100Hz) current
converter structure has two important design parameters: the is the most penalizing one, due to the chosen single-phase grid
number of channels and the operation frequency. Increasing connection. Thus, the capacitive value is calculated with:
the number of phases interleaved control can be reached, Po
decreasing the required inductive energy (also size of Co ≥ 2
Vo 2 ω grid Δυ o (4)
inductors) in order to respect grid current ripple requirements,
but increasing the converter complexity. High operation
frequency of the PFC converter enables the reduction of the III. INDUCTIVE POWER TRANSFER SYSTEM
input inductor size, but with the disadvantage of higher
switching losses, and thus, with a bigger cooling system. A. Compensation topology
Moreover, a bypass diode is also included to the circuit for the
dc-link pre-charging process. The maximum current ripple of Considering the equivalent circuit represented in Fig. 4, L1
a Boost inductor is calculated with: and L2 are the self inductance values while R1 and R2 are the
resistance values of the primary and the secondary coils
Vo V ·δ respectively, M is the mutual inductance, k the coupling factor,
Δi L = ·δ ·(1 − δ ) = i (2) RL is the equivalent load resistance and Ȧ the angular
f ·L f ·L
frequency:
which is equal to the grid ripple if one channel PFC is used.
Generalizing the expression in terms of the number of phases,
the grid current ripple is defined with: M = k L1 L2 (5)
Vo
Δi grid = ·K N
f · L· N (3)
where K N = (Nδ − ¬Nδ ¼)(1 − Nδ + ¬Nδ ¼)
where N is the number of channels of the Boost converter and
¬x ¼ = floor (x ) is the largest integer not greater than x. The
Fig. 4: Equivalent inductive power transfer circuit.

1295
Analyzing and simplifying the circuit (see Fig. 5), the Apart from the classical solutions, combinations of these
maximum transferred power is reached when the secondary basic compensations can also be found in the literature [5].
reflected equivalent impedance is equal to the primary The aim of these compensation methods is to compensate the
conjugate impedance, i.e. Z1*=Z2´. self inductance of the primary and secondary coils. With the
compensation of the secondary self inductance, the transferred
efficiency is already increased because only active power will
be transferred. The transferred efficiency for a series or
parallel compensated secondary turns into [6]:

η max =
(kQ )2
2 (9)
§1 + 1 + (kQ )2 ·
¨ ¸
© ¹
Fig. 5: Simplified equivalent circuit. Thus the efficiency of a secondary compensated IPT
system depends directly on the coupling and quality factors of
However, it is known that the maximum power transfer the coils. Fig. 7 shows the power transfer efficiency for
condition does not mean maximum transfer efficiency, but different quality and coupling factors.
50% efficiency (6).

R2 ' R2 ' X: 0.25

η= '
= ' '
= 0.5 Ÿ 50% (6)
1 Y: 80
Z: 0.9 X: 0.4
X: 0.8
Y: 25

R1 + R2 R2 + R 2
Y: 45 Z: 0.9049
Z: 0.9
0.8

Thus when designing an inductive power transfer system

Efficiency - η
0.6

the operation frequency is not the one complying with the 0.4
maximum transferred power, but with the maximum
efficiency point. The efficiency of a non-compensated IPT 0.2

system (7) depends on the quality factor of the coils Q, the 0

coupling factor between the coils k, and the quality factor of 100
80 1
the secondary coil Q2: 60
0.6
0.8
40
0.4

(kQ )2
20
0.2
0 0

η max = (7)
Quality Factor - Q Coupling - k

(kQ ) 2
+ 2 1 + Q2 + (kQ ) + 2 2
Fig. 7: Power-transfer efficiency with secondary compensation.
where Q is defined as:
IV. DESIGN AND DEVELOPMENT OF THE CONTACTLESS
ωL1 ωL2 BATTERY CHARGER
Q = Q1 Q2 = · (8)
R1 R2 In order to validate the operation of a contactless battery
charger, the primary side converter (PFC + H-Bridge) along
Analyzing this expression, in a really optimistic scenario with the coils have been designed and build (see Fig. 8). The
(k=0.5 and Q=100), the efficiency is not very high on-board Boost converter has not been developed and the final
(Ș=91.72%), and so, a compensation circuit is required in complete solution will be presented in a future paper.
order to increase this value. Four classical compensation
methods can be found in the literature: series-series (SS),
series-parallel (SP), parallel-series (PS) and parallel-parallel
(PP) (see Fig. 6).

Fig. 8: Implemented converter topology.


a) SS b) SP
TABLE I illustrates the specifications of the contactless
battery-charger, while TABLE II summarizes the main power
electronic design parameters.

TABLE I: SYSTEM SPECIFICATIONS

c) PS d) PP
Description Param. Value Unit
Fig. 6: Classical compensation topologies. Maximum output power Pomax 3.2 kW

1296
Battery voltage Vbat 350-400 V
Primary bus voltage Vdc1 300-450 V
Grid RMS voltage Vgrid 230 V
Maximum input current ripple ǻiL 20 %
Maximum dc voltage ripple ǻVdc 5 %

TABLE II: MAIN POWER ELECTRONICS DESIGN PARAMETERS

Description Param. Value Unit


PFC operation frequency fswPFC 100 kHz
IPT operation frequency fswIPT 100 kHz
PFC number of phases nPFC 2
Fig. 9: Pareto front of the optimal inductors.
A. Power Factor Correction
For the primary side PFC converter, a two channel The chosen solution from the Pareto front has an
interleaved converter has been chosen. As an initial step for inductance value of 191 μH, with an E42/21/15 core with a
the design of the optimum interleaved PFC converter, various 3.5 mm gap and with 36 turns of a 0.071x420 Litz wire.
discrete SiC semiconductors from Cree [7] have been The PFC output capacitor, with a maximum ripple voltage
compared, considering the worst operating point of the of 5%, is calculated with the maximum output power, the
converter. The results are summarized in TABLE III, where minimum output voltage, and applying the following equation:
conduction and commutation losses of three SiC
semiconductors are compared. The C2M0080120D has been Po 3200
the chosen semiconductor for the PFC converter. Co ≥ = = 1.1 mF
2 ripple (%) 300 2 2π 100 0.05 (11)
Vo
TABLE III: SIC SEMICONDUCTOR CHARACTERISTICS FOR IN A TWO CHANNEL
100
INTERLEAVED PFC CONVERTER.
B. High Frequency Primary Inverter
For the high frequency primary inverter a H-Bridge has
C2M0160120D C2M0080120D C2M0040120D been chosen. Once more a comparative evaluation of various
discrete SiC semiconductors has been made (see TABLE IV),
RdsON(Tmax) 290 mŸ 128 mŸ 84 mŸ and the C2M0080120D has been the chosen semiconductor
solution for the high frequency H-Bridge inverter.
Pcond 12.42 W 5.48 W 3.06 W
TABLE IV: SIC SEMICONDUCTOR CHARACTERISTICS IN AN H-BRIDGE
PON100kHz 6.34 W 6.00 W 5.49 W CONVERTER INTENDED FOR SS COMPENSATION.
POFF100kHz 5.36 W 8.29 W 11.19 W
C2M0160120D C2M0080120D C2M0040120D
PTOT100kHz 24.14 W 19.77 W 20.29 W
RdsON(Tmax) 290 mŸ 128 mŸ 84 mŸ
Tj100kHz 140 ºC 122 ºC 118 ºC
Pcond 11.87 W 5.24 W 3.44 W

As mentioned before the minimum PFC inductance value PON100kHz 2.10 W 3.80 W 7.00 W
is determined with a duty cycle of 0.75, which is the worst
case scenario of a two channel Boost converter. The minimum POFF100kHz 0W 0W 0W
operation frequency for the PFC is 100 kHz, and in order to
comply with the maximum grid current ripple the minimum PTOT100kHz 13.97 W 9.04 W 10.44 W
inductance value will be:
Tj100kHz 114 ºC 99 ºC 99 ºC
Vo max (2· Dworst − 1) · (1 − Dworst )
Lmin = = 190.6 ȝH (10) C. Primary side converter
f · ΔI L
With the previous design considerations a primary side
In order to design an optimal inductor with standard ferrite converter has been designed, where the PFC and the primary
cores and Litz wires, an optimization routine has been high frequency inverter are sharing the same cooling system
developed. The Pareto front of the various solutions is (see Fig. 10). For the bus capacitance, six aluminum
illustrated in Fig. 9. Each inductor has to support a rms current capacitors have been used, with a capacitance value of 820 μF.
of 8.2 A and a rms current ripple of 2.56 A, which corresponds
to a ripple current of 20%.

1297
1 1
ω= = (12)
L1 C1 L2 C2
where L1 and L2 are the self inductance values of the primary
and secondary, respectively, and C1 and C2 are the required
capacitance values for a specific resonant frequency.
Apart from the minimum coupling requirements for the
minimum transfer distance, the self inductance values of the
inductive coils will be maximized in order to reduce the
requirements for the resonant capacitors of the resonant tank.
Fig. 10: Layout of the primary side converter.
In order to reach this optimum combination several 3D
A picture of the final primary side converter solution is finite element simulations have been carried out with
illustrated in Fig. 11. FLUX3D (see Fig. 12). Circular coils where chosen for the
laboratory prototype, due to the fact that only the covered
inductance surface affects on the value of the coupling factor
[7].
The diameter of each inductive pad is limited to 300mm
and 210mm, respectively, due to space limitations. Moreover,
it is demonstrated that a higher coupling value is reached if the
secondary or received pad is smaller than the primary pad [7].

Fig. 11: Picture of the primary side converter.

D. SS compensation
As mentioned before, the power transfer of the series- Fig. 12: FEM simulation to find optimal coils.
series compensation depends on the values of both bus
voltages, the operation frequency and the mutual inductance For an operation frequency of 100 kHz, and an output
between the two inductive pads. In order to keep commutation power value of 3.2 kW, with a maximum primary bus voltage
losses to a minimum, the frequency will be kept constant to of 450 V and a maximum secondary bus voltage of 400 V, the
the resonant frequency values, and it will not be used as a expected mutual inductance value is 70 μH for the closet
control parameter. transfer distance. The characteristics of the designed and
The maximum value of the bus voltages are set to 450V implemented inductive coils are summarized in TABLE V. In
and 400V, respectively. The maximum power will be set for order to increase the coupling factor ferrites are introduced in
the maximum bus voltage values, and the maximum mutual the bottom of each pad, as well as a 1 mm thick aluminum
inductance value. When the two inductive pads are centered sheet for shielding purposes.
and as close as required, the maximum coupling will be
reached. Thus, when the two pads are separated one from the TABLE V: INDUCTIVE PAD CHARACTERISTICS AT 100 KHZ.
other, or misaligned, the bus voltage will reduce their value in
order to keep the charging power constant. The maximum Parameter Value
distance and misalignment characteristics of the inductive
Primary Self Inductance L1 115 μH
power transfer system will be limited by the current
limitations of the H-Bridge inverter. Thus an iterative design Secondary Self Inductance L2 78 μH
process has been carried out in order to determine a
competitive and feasible solution. Primary Resistance Rac1 1ɏ

1) Inductive pads Secondary Resistance Rac2 0.6 ɏ


In the series-series compensation, the resonant frequency Mutual Inductance M 70 μH
depends only on the self inductance values of each inductive Primary Turns T1 25
pad:

1298
Secondary Turns T2 19
Primary Litz Wire d1 0.071x420
Secondary Litz Wire d2 0.071x1260

For a transfer distance of 100 mm the mutual inductance is


reduced reaching a maximum coupling factor of 0.4. In Fig. 13
for a transfer distance of 100 mm, the dependency with the x
and y misalignments is illustrated.
Coupling factor

0.5

0.4 Fig. 14: Voltage derating characteristics of a R74 capacitor from KEMET [8].
k - Coupling factor

0.3
The capacitor tanks are build with several small
0.2 capacitance values in order to have a high capacitance
0.1
resolution.
0 In Fig. 15 the 3D CAD of the resonant tank, i.e. the
100
inductive pads and the resonant capacitos, is illustrated, along
50 100
0 50 with the final build solution. The shielding of the resonant
-50 -50
0 capacitors is made with the aluminum sheet that is located
dy - Y axis misalignment -100 -100
dx - X axis misalignment
under the ferrites.

Fig. 13: Coupling factor in terms of the misalignment.

2) Resonant capacitors
Once the inductance values are set, the capacitor
requirements for 100 kHz are established. Those requirements
are summarized in TABLE VI. The capacitors are
dimmensioned according to their specified maximum rms
current and voltage values at 100 kHz, along with the required
capacitance value.

TABLE VI: RESONANT CAPACITORS REQUIREMENTS. Fig. 15: Primary and secondary inductive pads with compensation capacitors
(3D CAD and final prototype picture).
Irms Vrms
Value
(@100kHz) (@100kHz)
V. EXPERIMENTAL IPT SYSTEM VALIDATION
C1 22 nF 11 Arms 800 Vrms
A picture of the contactless inductive power transfer
C2 32 nF 11 Arms 550 Vrms charger prototype is illustrated in Fig. 16.

In the case of the resonant capacitors several


considerations have been taken into account:
ƒ For a operation frequency of 100 kHz MKP
capacitors have been chosen.
ƒ In order to use the same capacitor series for both
sides 700 V capacitors have been chosen,
intended for a series connection of capacitors to
support the expected voltage values.
ƒ In the chosen capacitor series the largest
capacitance value is 22 nF, and the voltage
derating starts influencing from this frequency on Fig. 16: Picture of the build prototype. Primary side inverter and both
(see Fig. 14). inductive pads.

1299
In a first step the correct operation of the power factor Similarly, with a 50% of misalignment, the primary bus
correction converter has been validated (see Fig. 17). The voltage is reduced in order to keep the output power value (see
primary current and voltage waveforms are in phase, with a in Fig. 19)
low harmonic distortion and reaching a high power factor
value.

Fig. 19: Primary side inverter voltage and current values for a transfer
Fig. 17: Input voltage and current waveforms. distance of 40mm and a misalignment of 50%.

In a second step, the validation of the high frequency


inverter along with the inductive pads has been carried out. VI. CONCLUSIONS
For the validation of the charger a variable resistor has been In this paper the design and implementation of a
used as a load. A simple rectifier has been build for validation contactless inductive power transfer battery charger for
purposes. The voltage and current waveforms of the primary electric vehicles is presented. The design and the development
side high frequency inverter are illustrated in Fig. 18 for of this battery charger is carried out using the latest
various power transfer distances. semiconductor technology. The main steps and drawbacks of
the design process are discussed, underlining the design keys
in order to achieve an optimal IPT power electronics
converter.

REFERENCES
[1] C.S. Wang, O.H. Stielau, and G.A. Covic, “Design considerations for a
contactless electric vehicle battery charger,” IEEE Transaction on
Industrial Electronics, vol. 52, no. 5, pp. 1308-1314, Oct. 2005.
[2] H. Takanashi, Y. Sato, Y. Kaneko, S. Abe, and T. Yasuda, “A large air
gap 3 kW wireless power transfer system for electric vehicles,” in
Energy Conversion Congress and Exposition (ECCE), 2012.
[3] B. Goeldi, S. Reichert, and J. Tritschler, “Design and Dimensioning of a
Highly Efficient 22 kW Bidirectional Inductive Charger for E-Mobility,”
in International Exhibition and Conference for Power Electronics,
a) Intelligent Motion and Power Quality (PCIM), 2013.
[4] A. Rujas, I. Etxeberria, J. San Sebastian, U. Larrañaga, and T. Nieva,
“Design and Experimental Validation of a Silicon Carbide 100kW
Battery Charger operating at 60kHz,” in 15th IEEE Workshop on Control
And Modeling for Power Electronics (COMPEL), 2014.
[5] J. Villa, J. Sallan, J. Sanz Osorio, and A. Llombart, “High-Misalignment
Tolerant Compensation Topology For ICPT Systems,” IEEE
Transaction on Industrial Electronics, vol. 59, no. 2, pp. 945-951, Feb.
2012.
[6] K. Van Schuylenbergh and R. Puers, “Inductive Powering: Basic Theory
and Application to Biomedical Systems”, 1st ed. Springer Science, 2009.
[7] R. Bosshard, J. W. Kolar, and B. Wunsch, “Accurate Finite-Element
Modeling and Experimental Verification of Inductive Power Transfer
Coil,” in 29th Annual Applied Power Electronics Conference and
Exposition (APEC), 2014.
[8] http://www.cree.com/
b)
[9] http://www.kemet.com/Lists/ProductCatalog/Attachments/347/KEM_F3
Fig. 18: Primary side inverter voltage and current values for a transfer 029_R74_125C.pdf
distance of a) 40mm and b) 80mm.

1300

You might also like