Design of A Feedback Controlled Dual Sided LCC Converter To Achieve ZVS For Electric Vehicle Wireless Charging Applications
Design of A Feedback Controlled Dual Sided LCC Converter To Achieve ZVS For Electric Vehicle Wireless Charging Applications
Design of A Feedback Controlled Dual Sided LCC Converter To Achieve ZVS For Electric Vehicle Wireless Charging Applications
Abstract—Wireless charging of electric vehicles (EVs) is However, despite many incentive programs, PEVs still suffer
gaining traction in the industry due to its convenience and from some drawbacks. Firstly, current technology is unable to
robustness. Thus, reaching a high system efficiency in wireless produce the battery that has high energy density at an affordable
power transfer (WPT) at high power loads is a critical challenge
in the EV charging market. WPT can be applied for both stationary
cost. This drawback makes EVs unacceptable to many
and dynamic charging scenarios. This paper focuses on design of an customers because the driving range of EVs is unsatisfactory.
LCC converter with frequency control. The designed converter is For instance, in commercialized EVs, the energy density of
capable of achieving zero voltage switching at different loading and lithium-ion battery is only 90-100 Wh/kg for a finished pack
input voltage conditions. Additionally, a polarized single-sided flux [3]. This number is too small compared to gasoline of which
pad topology is applied for the design of the transmitting and the energy density is 12,000 Wh/kg. The second drawback is the
receiving coils to increase the system efficiency.
long-time charging of PEVs. The average time for a single
Keywords—Compensation network, Electrical vehicles, LCC, charge of EVs is about one half-hour, which is much longer
Full bridge converter, Pure power factor, Resonant converter, soft
switching, wireless charging, zero voltage switching (ZVS) than gasoline filling process [3]. Finally yet importantly,
customers have to face safety risks when using electrical
I. INTRODUCTION components to charge EVs [4].
Conventional internal combustion engine vehicles not only Therefore, there is a need to eliminate PEVs limitations by
perform at a low efficiency, but they also are one of the main utilizing new technologies. Wireless power transfer (WPT) is
sources of greenhouse gases. Furthermore, under the pressure an advanced technology able to solve technical drawbacks of
of global energy crisis, the government in many nations is PEVs. The operation of a WPT system is similar to the
focusing on developing a new generation of electric vehicles. inductive power transfer (IPT) technology. With WPT system,
Indeed, electrification in transportation has been rapidly customers no longer have to wait a long time for charging
investigated and developed for years enabling significant process [4]. Since EVs are more frequently charged, the driving
improvements in power electronics and battery technology [1]. range is significantly increased, and the size of battery is
Consequently, plug-in electric vehicles (PEVs) become more reduced. Finally, WPT system eliminates risks of electric shock
affordable and acceptable for customers over recent years. since the system sets customers free from handling or touching
However, to challenge the 300-mile range of an internal electrical equipment such as wire or outlet [5].
combustion engine powered vehicle, a pure EV needs a large
amount of batteries which is too heavy and too expensive [2]. In a WPT system, the power is transferred through the
WPT system, as shown in Fig. 1, is an advanced technology to mutual inductance between the transmitting and receiving coils
solve these technical drawbacks of EVs through providing an [6]. The WPT for EVs is more attractive because it provides a
opportunity charging conditions. more convenient, safer way of charging the batteries. For a
stationary system, the customers can park and leave without
plugging their vehicles to an electric outlet. For dynamic system,
AC/DC
AC Battery
EVs are charging while driving, which could lead to unlimited
with PFC driving range.
II. PROPOSED SYSTEM
DC to high This research undertakes to analyze and design a WPT system
frequency AC/DC
AC
Receiving Coil to charge EVs with high efficiency. The dual-sided LCC
compensation network is shown in Fig. 2. where S1 ~ S4 are
Primary four power MOSFETs in the primary side. All the detail
Secondary
Compensation Compensation components and values are listed in Table 1. The LCC
compensation network improves the performance of WPT
Transmitting Coil system. It also enables a unity power factor for both input side
and output rectifier. As long as the coil current and voltage are
in phase, then a unity power factor is achieved. This phase
Figure 1. Wireless charging system of an electric vehicle
relation is independent of the coupling coefficient or any load
2659
• The phase relation is independent of coupling coefficients
or load variations.
• The power transferred through LCC tank depends on
k ,U AB ,U ab as shown in (21)
(21)
k L1 L2 U ab
P = U AB × I L f 1 = U AB ×
w0 L f 1 L f 2
Figure 4. Equivalent circuit referred to the primary side
As the real MOSFET is not ideal, and to ensure all switches
L2 M turn on at ZVS, some characteristics of the circuit need to be
n= (1) k= (2) Lm = kL1 (3)
L1 L1 L2 adjusted. One method is to turn on the MOSFETs at a negative
current. To ensure this negative turn-on current, the turn-off
where is the turns ratio between the two coils, is the current of MOSFETs has to be positive. The following is the
coupling coefficient, and is the magnetizing inductance design process that can ensure the negative turn-on current
referred to the primary side.
• To achieve ZVS, increase the value of Le 2
Ls1 = (1 − k ) L1 (4) Ls2 = (1 − k ) L2 (5)
1 1 Le 2 = Le 2 + ΔLe 2 (22)
L f1 C f1 = (6) L f2 C f 2 = (7)
w02 w02 • The MOSFET turn-off current:
1 1
L1 − L f1 = 2 (8) L2 − L f2 = 2 (9)
w0 C1 w0 C2 U2 ΔL 1
ab e 2 −
where & are the leakage inductances of the primary and U AB L f 2 4 + U AB
I OFF = 2 (23)
secondary coils respectively, while , , & are the w0 L f 2 4 w0 L f 1
filtering passive elements shown in Fig. 4, and is the
resonant frequency.
Also, • The minimum turn-off current is reached when:
Lm = kL1 Ls1 = (1 − k ) L1 ΔL 1 Lf 1
(10) (11)
U AB = U ab _ min 4 e 2 − (24)
L
L f2 f 2 4 Lf 2
Ls′2 = Ls2 / n 2 = (1 − k ) L2 / n 2 (12) L′f2 = (13)
n2 2 2 2
1 I OFF _ min × w0 L f 1 L f 2
ΔLe 2 = Lf 2 + (25)
C2′ = n C2 2
(14) C ′f2 = n C f2 2
(15) 4 U ab2 _ min
U ab 4CossU AB _ max
′ =
U ab (16) I OFF ≥ (26)
n td
After applying superposition on both of the sources and while where Coss is the junction capacitance of the MOSFET, and
using the derivations shown in [7], it can be concluded that: dead time between MOSFETs & or &
IV. COILS DESIGN & MATLAB/ANSYS CO-SIMULATION
U AB = U AB ∠0o (17) ′ = U ab
U ab ′ ∠ − 90o (18)
In this paper, polarized single-sided flux pad topology is
applied for the design of transmitting and receiving coils. Since
k L1 L2 U AB the coils are wound like a “D”, this topology is also called as
k L1 L2 U ab ILf 2 = ∠ − 90 o
I Lf 1 = ∠0 o (19) w0 L f 1 L f 2 (20) Double-D (DD) structure. The DD structure yields valuable
w0 L f 1 L f 2
characteristics, which are desired for wireless charging system
of EVs applications. For instance, the charge zone for a DD pad
- There are four conclusions from above equations: could be about two times larger than a conventional circular pad
with similar material cost. Moreover, DD structure is capable
• I L f 1 & U AB are in phase unity power factor of input side. of improving the coupling coefficient of transmitting and
• I L f 2 & U ab are in phase unity power factor of output receiving coils compared to traditional coil designs.
side. The structure of transmitting coil and receiving coil includes
two windings, which are connected magnetically in series and
2660
electrically in parallel. The coil width was made up of 20 turns misalignment (shown as slide_x) and the gap between the
of 4-mm diameter Litz wire. The ferrite bars are implemented transmitting and the receiving coils. According to the plot, the
in order to guide and enhance the magnetic fields. In addition, coupling coefficient decreases when the air-gap variation or
an aluminum plate is used as the shielding layer and heat sink. Slide-X misalignment increases. Furthermore, at 200 mm air-
The structure of transmitting coil and receiving coil includes gap and slide-X of 50 mm, the coupling coefficient is 0.2, that
two windings, which are connected magnetically in series and is a compromising number.
electrically in parallel. The coil width was made up of 20 turns
V. FEEDBACK FREQUENCY CONTROL
of 4-mm diameter Litz wire. The ferrite bars are implemented
in order to guide and enhance the magnetic fields. In addition, In charging applications, there is a need to regulate the
an aluminum plate is used as the shielding layer and heat sink. amount of energy which charging the EVs. There are three
popular control methods known as primary side control,
The next step is to build a simulation model of transmitting
secondary side control, and dual-side control [1]. The name of
and receiving coils in Ansys/Maxwell as shown in Fig 5. The
each method presents where the control is applied at. In this
purpose of this step is to investigate fully the operation of two
paper, the controller is applied at the primary side.
coils in WPT applications. Finally, the co-simulation between
Matla/Simulink, Ansys/Simplorer and Ansys/Maxwell will be The primary side control could be accomplished by adjusting
done. In particular, the controller is simulated in the switching frequency or the phase-shifted between legs.
Matlab/Simulink, the transmitting and receiving coil are Because the compensation network of the WPT system is a
simulated in Ansys/Maxwell, and the power converter is resonant converter, phase shift on the primary side causes
simulated in Ansys/Simplorer. circulation of energy and ZVS is not achieved.
In this paper, the primary side frequency control is applied in
order to regulate the transferred power. With double sided LCC
topology, there are multiple neighbor resonance frequencies.
For this project, the components are calculated and selected
based on 85kHz resonant frequency. And control algorithm
becomes more complicated with wider frequency range.
Therefore, this paper focuses on the gain and frequency
relationship from the neighbor switching frequencies between
50kHz and 120kHz as shown in Fig 6.
Gain vs frequency (k=0.25)
8
load=60
load=100
7 load=150
(a) load=200
4
fs(ii)
3
2
fs(i)
1
fs(iii)
(b) 0
Fig. 5. Single-Sided Double D Coils (a) result of coupling 50 60 70 80 90 100 110 120
Switching frequency [kHz]
coefficient (b) variation on air gap and slide-X misalignment
Fig. 6. Characteristic of double sided LCC converter.
The air-gap 200 mm is selected since this distance is also the
normal gap between electric vehicles floor and the ground. The Generally, for the operation frequency beyond resonant
preferred Slide-X between two coils is 50 mm because for frequency, it enables ZVS. In this double sided LCC
drivers with normal driving skills, they are able to park cars in compensation network, depending on the coupling coefficient
that range of misalignment between the transmitting and of the transmitting and receiving coil and load resistance,
receiving coils. A 3D plot is shown in Fig. 5.b) to represent the resonance frequency can be different and multiples. As a
change of the mutual coupling with the variation of the lithium battery for electric vehicle, the load resistance roughly
from 10 to 200 Ω is used for this project analysis and
2661
simulation. some small VDS changes across the MOSFET can be seen
From the DC characteristic of Fig.6, with coupling through simulation. The power level will change depending on
coefficient of 0.25, it can be seen that the there are three regions the coupling coefficient. The weaker the coupling coefficient,
of ZVS marked in light pink. The switching frequency beyond the lower the energy level.
the resonance enables ZVS. And the switching frequencies
below the resonance cannot achieve ZVS. Therefore, the
switching frequency should stay in these three ZVS zones
depending on the input voltage. In other words, the change of
frequency is required depending on the input voltage, but it is
nonlinear function of the input. For high gain, the switching
frequency needs to decrease and jump to the ZVS (i) region as
input voltage decreases. And for lower gain, switching
frequency needs to increase to the ZVS (iii) region as the input
voltage increases. The detail explanation is showing in Fig 7.
The output voltage control scheme is implemented on the
primary side in Fig. 7 based on the gain and ZVS switching
frequency relationship. In this control block, sensed output
voltage is compared with reference voltage at the input
of error amplifier, then go through PI controller to reduce error,
settling time and overshoot. The result is translated to sawtooth
signal and compared with another reference voltage to generate
PWM modulation to the primary MOSFETs gate driver.
The simulation result based on this control structure is shown
in Fig. 8. A constant output voltage is obtained while input
voltage changes for a wide range. In addition, Fig. 9 shows that
ZVS is always achieved with different gain level, wide
variation of input voltage and switching frequencies.
Vo
+ error fs Control scheme Gate Fig. 8. Simulation result of closed loop of control algorithm. (a) Vout. (b) Iout.
PI controller PWM
- of 3 ZVS region driver (c) Frequency (d) Vin.
Vref
start
Yes
Is this the right gain?
Stay at this frequency
No zone (ii)
Is the gain too small?
No
Yes
Is the gain too big?
Yes
No Is PI controller asking Is PI controller asking No
for bigger than (ii) for smaller than (ii)?
Yes Yes
Jump to the higher Jump to the lower
switching frequency ZVS switching frequency ZVS
f(iii) to get lower gain f (i) to get higher gain
Fig. 7. The control flow of proposed method for the primary side
Fig. 9. Simulation result of ZVS on S1 with varying input voltage. (a)
VI. VALIDATION RESULTS MOSFET current and voltage with 500V input. (b) MOSFET current and
voltage with 440V input. (c) MOSFET current and voltage with 360V input.
In this paper, the MOSFET is chosen to be IPW65R019C7 (d) MOSFET current and voltage with 200V input.
from Infineon Technologies. The deadtime is chosen to be 150
ns, and efficiency does not change much once ZVS is achieved VII. CONCLUSION
as the simulation data shows from LTspice. After deadtime
A dual sided resonant LCC converter is designed in this
longer than 350 ns, MOSFET body diode would conduct and paper for EV charging applications. The designed converter can
2662
achieve a high system efficiency due to its soft switching [4] S. Li, C.C. Mi, “Wireless Power Transfer for Electric Vehicle
features. A feedback frequency control is applied on the Applications,” IEEE Journal Of Emerging And Selected Topics In Power
Electronics, vol. 3, no. 1, March. 2015.
converter to increase the system efficiency while tracking the
[5] R. Bosshard and J. W. Kolar, “Inductive Power Transfer for Electric
output power. The simulation model is fully developed and Vehicle Charging”, IEEE Power Electronics Magazine, Sept 2016.
analyzed using LTspice, MATLAB and Ansys Simplorer / [6] A. Khaligh, S. Dusmez, “Comprehensive Topological Analysis of
Maxwell. The simulation system efficiency from dc power Conductive and Inductive Charging Solutions for Plug-In Electric
source to power electronics load is 97.48 % for maximum power Vehicles,” IEEE Transactions On Vehicular Technology, vol. 61, no. 8,
with coupling coefficient of 0.25. October. 2012.
[7] S. Li, W. Li, J. Deng, T. Nguyen, C. Mi, “A double-sided LCC
REFERENCES compensation network and its tuning method for wireless power transfer”,
IEEE Transactions on Vehicular Technology, vol. 64, pp. 2261-2273, Jun.
[1] “Electric Fuel Pumps for Wireless Power Transfer, Enabling rapid growth
2015.
in the electric vehicle market”, IEEE Power Electronics Magazine, Vol.
4, No. 2, June 2017. [8] S. Lu, X. Deng, W. Shu, X, Wei, S. Li, “A New ZVS Tuning Method for
Double-Sided LCC Compensated Wireless Power Transfer System,”
[2] S. Li, C.C. Mi, “Wireless power transfer for electric vehicle applications”,
Energies, vol. 11, no. 2, pp. 307, Feb. 201
IEEE J. Emerg. Sel. Top. Power Electron., vol.3, no.1, pp. 4-17, 2015
[3] V. Etacheri, R. Marom, R. Elazari, G. Salitra, and D. Aurbach,
“Challenges in the development of advanced Li-ion batteries: A review,”
Energy Environ. Sci., vol. 4, no. 9, pp. 3243–3262, 2011.
2663