IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 31, NO.

12, DECEMBER 2016 8471

An Inductive and Capacitive Combined

Wireless Power Transfer System With
LC-Compensated Topology
Fei Lu, Student Member, IEEE, Hua Zhang, Student Member, IEEE, Heath Hofmann, Senior Member, IEEE,
and Chunting Chris Mi, Fellow, IEEE

Abstract—This paper proposes a combined inductive and capac- I. INTRODUCTION

itive wireless power transfer (WPT) system with LC-compensated
IRELESS power transfer technology can be divided into
topology for electric vehicle charging application. The circuit
topology is a combination of the LCC-compensated inductive
power transfer (IPT) system and the LCLC-compensated capac-
W two categories: inductive power transfer (IPT) [1], [2]
and capacitive power transfer (CPT) [3], [4]. The IPT system
itive power transfer (CPT) system. The working principle of the utilizes magnetic fields to transfer power, and the CPT system
combined circuit topology is analyzed in detail, providing the re-
lationship between the circuit parameters and the system power.
makes use of electric fields.
The design of the inductive and capacitive coupling is implemented The IPT technology has been widely studied in previous re-
by the finite-element analysis. The equivalent circuit model of the search. It consists of two planar coils with an air-gap distance
coupling plates is derived. A 3.0-kW WPT system is designed and around 150 mm for vehicle charging applications [5]. The two
implemented as an example of combined inductive and capacitive coils form a separated transformer, in which the primary side is
coupling. The inductive coupler size is 300 mm × 300 mm and
the capacitive coupler is 610 mm × 610 mm. The air-gap distance
loosely coupled with the secondary side [6]. The coupling coef-
is 150 mm for both couplers. The output power of the combined ficient between the primary and secondary coil is much smaller
system is the sum of the IPT and CPT system. The prototype has than 1.0, and usually in the range of 0.1–0.3. This means that
achieved 2.84-kW output power with 94.5% efficiency at 1-MHz the leakage inductance is significantly larger than the mutual
switching frequency, and performs better under misalignment than inductance. When it is used as a regular transformer, it can-
the IPT System. This demonstrates that the inductive–capacitive
combined WPT system is a potential solution to the electric vehicle
not transfer high power efficiently. Therefore, a corresponding
charging application. compensation circuit is necessary to resonate with the coils and
provide high voltage at both input and output to transfer high
Index Terms—Capacitive power transfer, decoupled-capacitor power.
model, electric field, electric vehicle (EV), high-frequency wire-
less power transfer, inductive and capacitive combined, inductive The IPT system can be classified according to the compen-
power transfer, magnetic field, misalignment ability, resonant cir- sation circuit topology. Conventionally, a single capacitor is
cuit, soft-switching condition. connected in series or parallel with the coil which results in
four topologies, listed as series–series, series–parallel, parallel–
series, and parallel–parallel (PP) [7]–[9]. Their advantage is
simplicity. However, they are not suitable for all load conditions
Manuscript received August 05, 2015; revised October 08, 2015 and Decem- [10]. When the load varies between light and heavy, they can-
ber 10, 2015; accepted January 11, 2016. Date of publication January 20, 2016; not maintain high efficiency. Another problem is that the coil
date of current version July 08, 2016. This work was supported in part by the
U.S. Department of Energy Graduate Automotive Technology Education Grant, position misalignment could break the resonance between the
the U.S.-China Clean Energy Research Center—Clean Vehicle Consortium, capacitor and coil, which in turn can cause power dramatically
DENSO International, San Diego State University, the University of Michigan- drop or increase.
Ann Arbor, and China Scholarship Council, and also by the U.S. Department
of Energy under Award DE-PI0000012. Recommended for publication by As- The double-sided LCC-compensated topology is a better so-
sociate Editor J. M. Rivas Davila. lution to compensate the power-transmitting coils, in which one
F. Lu is with the Electrical Engineering and Computer Science Department, external inductor and two external capacitors are required at
University of Michigan, Ann Arbor, MI 48109 USA, and also with Department
of Electrical and Computer Engineering, San Diego State University, San Diego, each side [11]. With this topology, the resonant tank acts as a
CA 92182 USA (e-mail: feilu@umich.edu). current source to both the input and output, which means the
H. Zhang is with the Department of Electrical and Computer Engineering, output current is independent of the load condition. In the bat-
San Diego State University, San Diego, CA 92182 USA, and also with the
School of Automation, Northwestern Polytechnical University, Xi’an 710072, tery charging application, the resonant circuit is, therefore, not
China (e-mail: huazhang@umich.edu). affected by the battery voltage variation. Therefore, it is suitable
H. Hofmann is with the Electrical Engineering and Computer Science Depart- to provide constant current charging condition to the batteries.
ment, University of Michigan, Ann Arbor, MI 48109 USA (e-mail: hofmann@
umich.edu). Another benefit of LCC topology is that it can provide nearly
C. C. Mi is with the Department of Electrical and Computer Engineering, San unity power factor to the input side inverter, so there is almost
Diego State University, San Diego, CA 92182 USA (e-mail: cmi@sdsu.edu). no reactive power injected into the resonant tank, which results
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. in high efficiency for this topology. In [12], a LCC-compensated
Digital Object Identifier 10.1109/TPEL.2016.2519903 system has achieved 96% efficiency at 7-kW output power, and

0885-8993 © 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
8472 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 31, NO. 12, DECEMBER 2016

the total grid-to-battery efficiency of the wireless charger can

achieve 93%. It is already comparable with the traditional wire-
connected charger with LLC resonant converter [13]. Moreover,
this topology compensates using the self-inductance of the coils,
which is independent of misalignment. Therefore, the resonance
is maintained even in large misalignment cases. Although more
components are used in this compensation circuit, the two exter-
nal inductors can be integrated into the main coils to save space
and make the system more compact [14]–[16].
Compared to the IPT technology, the CPT technology has
not been well studied. Especially in the electric vehicle charg-
ing area, the CPT system has only recently been applied. Most
Fig. 1. Circuit topology of the LC-compensated IPT-CPT system.
of the present CPT systems were designed for low-power appli-
cations, including USB devices, lamps, and small robots [17]–
[20]. The transmitted distance in these applications is limited to
the millimeter range and the power is lower than tens of watts. power is 1.6 kW with 89.1% efficiency. It proves that the LCLC
When two metal plates are separated by an air-gap distance topology is a good candidate to realize capacitive charging for
of 150 mm, the capacitance is small (e.g., tens of picofarad), electric vehicles.
which requires a very large inductance value (e.g., millihenry) Comparing the circuit topologies of LCC-IPT system and
or very high switching frequency (e.g., tens of megahertz) to LCLC-CPT system, it is meaningful to reduce the external com-
provide resonant conditions to the capacitor. However, a large ponent number and simplify the circuit design. Therefore, they
inductor is difficult to make, and its parasitic resistance can can be combined together to make full use of the coupling coils
dramatically reduce the system efficiency. Also, too high of a and metal plates, which results in an LC-compensated IPT-CPT
switching frequency is also difficult to achieve in practice, and combined system. The coupling coils work as the IPT com-
the power level of a high-frequency converter is usually limited ponents, as well as the inductors resonating with the coupling
to hundreds of watts [21]. Also, the CPT approach requires metal plates. The coupling metal plates work as the capacitive
high electric fields to transfer high power, which can result in power transferring components, as well as the capacitors res-
significant safety concerns to the general public [22]. onating with the coupling coils. The circuit topology is shown
The CPT system is classified according to the compensation in Fig. 1. This system takes full advantages of the electric and
circuit topology. Similar to the IPT system, the most widely magnetic fields in the resonant circuit.
used topology in CPT systems is a single inductor connected in
series with the coupling capacitor to form a series resonance. II. CIRCUIT TOPOLOGY AND WORKING PRINCIPLE
However, this topology has only been used with large capac-
itances, where the inductance and switching frequency are A. LC Compensated IPT-CPT Topology
reasonable [23]–[26]. Another topology is the class-E converter, The corresponding IPT-CPT system topology is shown in
in which the coupling capacitor works as a regular resonant Fig. 1. The primary and secondary sides are separated by a red
component [27], [28]. However, its efficiency and power are dashed line. It is similar with the CPT system proposed in refer-
limited by the high-frequency converter. A high-frequency ence [32], except that two of the inductors L1 and L2 are coupled
PWM converter has been used to realize capacitive power to transfer power from the primary side to the secondary side
transfer [29], [30]. However, it requires the capacitance value using magnetic fields. At each side, there are only two external
to be in the tens of nanofarad range, and the distance is less components connected to the coupling coils and capacitors.
than 1 mm. Current CPT systems have achieved 1-kW power For the coupling coils, it is similar to the regular IPT system.
transfer and 90% efficiency. They have been designed for However, it is important to pay attention to the coupling
stationary charging with the charging gap in the millimeter polarity of L1 and L2 . The connection in Fig. 1 can result in an
range, and also have good field confinement [31]. addition of the inductive power with the capacitive power. If the
The double-sided LCLC compensated topology has been pro- coupling polarity is reversed, the two types of power transfer
posed in [32], in which two external inductors and two external will cancel. The coil design process will be analyzed in details
capacitors are used on each side of the coupling capacitor. Com- in Section III-A.
pared with previous work, this system has the potential to be For the coupling capacitors, the four plates are arranged as
applied to mobile charging of an electric vehicle, but generates shown in Fig. 5. The equivalent circuit model is derived in
significant leakage electric field emissions due to the large air- Section III-B, which contains two small capacitances, Cs1,2 ,
gap distance. The system has achieved 2.4-kW power transfer and two large capacitances, C1,2 . Compared to [32], no external
through an air-gap distance of 150 mm at 90.8% efficiency with capacitor is connected with the coupling capacitors, and the
two pairs of 610 mm × 610 mm aluminum plates. The system system structure is further simplified. The coupling capacitors
also has good performance with large misalignment and air-gap design process will also be presented in Section III-B.
variation. At 300-mm misalignment, the output power is 2.1 kW An H-bridge MOSFET inverter is used at the input side and
with 90.7% efficiency. At 300-mm air-gap distance, the output an H-bridge diode rectifier is connected at the output side to the
LU et al.: INDUCTIVE AND CAPACITIVE COMBINED WIRELESS POWER TRANSFER SYSTEM WITH LC-COMPENSATED TOPOLOGY 8473

The circuit in Fig. 2(b) results in the following equations:

⎧
⎪
⎪ ω = 2π · fsw
⎪
⎪
⎪
⎪ 1
⎪ Cs =
⎪
⎪
⎪ 1/Cs1 + 1/Cs2
⎪
⎪
⎨ 1
Cp1 = C1 + (1)
⎪  1/C s + 1/C2
⎪
⎪ ·
⎪
⎪ ω = 1/ L f2 Cf2
⎪
⎪ 
⎪
⎪
⎪
⎪ 1
⎪
⎩ ω = 1/ L1 · 1/C + 1/C
p1 f1

where the switching frequency is fsw , and Cp1 is the input

capacitance of the coupling capacitor from the primary side.
Equation (1) shows that Lf 2 and Cf 2 form a parallel resonance,
and the impedance is infinite. There is no current flow through
the inductor L2 . Also, L1 , Cf 1 , and Cp1 form another parallel
resonance. There is no current flow through the inductor Lf 1 .
The current on Lf 1 is also the input current to the resonant tank.
So, the input current is independent of the input voltage, which
means the circuit can work as a current source to the input.
Fig. 2(c) shows the components excited only by the output
voltage source at the desired frequency. There are two parallel
resonances highlighted in Fig. 2(c). The relationship between
components is expressed as
⎧
⎪ 1
⎪
⎪ Cp2 = C2 +
⎪
⎪ 1/C + 1/C1
Fig. 2. FHA analysis. (a) Simplified resonant circuit topology. (b) Compo- ⎪
⎨  s
nents excited by the input voltage. (c) Components excited by the output voltage.
ω = 1/ Lf 1 · Cf 1 (2)
⎪ 
⎪
⎪
⎪
⎪ 1
load. Silicon carbide (SiC) MOSFETs (C2M0080120D) from ⎪
⎩ ω = 1/ L2 · 1/C + 1/C
p2 f2
CREE are used in the inverter. The circuit in Fig. 1 can be tuned
to provide soft-switching conditions to the inverter to mitigate where Cp2 is the input capacitance of the coupling capacitor
the switching losses. In the rectifier, SiC diodes (IDW30G65C5) from the secondary side. Equation (2) shows that Lf 1 and Cf 1
from Infineon are applied. form a parallel resonance. L2 , Cf 2 , and Cp2 form another par-
allel resonance. There is no current flow through inductors L1
and Lf 2 . The current on Lf 2 is also the output current to the
B. Circuit Working Principle load. The output current is therefore independent of the output
The input-side inverter generates a square-wave voltage exci- voltage, and the resonant circuit works as a current source to the
tation, and injects power into the resonant tank. The output-side output.
rectifier converts the output ac current into dc form and supplies
the battery. Therefore, both the input and output voltage contain C. System Power Calculation
high-order harmonic components. However, the LC compensa-
tion network serves as low-pass filter; therefore, the fundamental First, the circuit in Fig. 2(b) is used to calculate the output
harmonics approximation (FHA) can be used to analyze work- current and output power. It needs to be emphasized again that
ing principle, as shown in Fig. 2. the output current only depends on the input voltage, as shown
Fig. 2(a) shows the simplified resonant circuit topology of in Fig. 2(b). The input voltage V1 is set to be the reference phasor
Fig. 1. The coupling capacitor is replaced by the equivalent in the following analysis.
circuit model in Fig. 6, which will be presented in Section III-B. The parallel resonance of Lf , Cf 1 , and Cp1 is considered, so
The input and output are replaced by two sinusoidal voltage the voltage on each component can be expressed as
sources. The superposition theorem is used to divide the two- ⎧
⎪
⎪
VCf1 = V1
source circuit into two single-source linear circuits, as shown in ⎪
⎪
⎪
⎨ jωL1 · V1 jωL1 · V1
Fig. 2(b) and (c), respectively. VL 1 = = = ω 2 L1 Cf 1 V1
Fig. 2(b) shows the components excited only by the input jωL1 +1/(jωCp1 ) −1/(jωCf 1 )
⎪
⎪
voltage source at the design frequency. There are two parallel ⎪
⎪ 1/(jωCp1 ) · V1 1/(jωCp1 ) · V1 −Cf 1
⎪
⎩ VCp1 = = = V1 .
resonances highlighted in Fig. 2(b). The components in gray jωL1 +1/(jωCp1 ) −1/(jωCf 1 ) Cp1
conduct no current, so they can be treated as an open circuit. (3)
8474 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 31, NO. 12, DECEMBER 2016

The voltage on Cp1 is also the voltage on the capacitor C1 . Equation (10) indicates that the output current (−I2 ) leads
So the voltage on C2 can be expressed as the input voltage V1 by 90°. In the H-bridge diode rectifier, the
1/C2 1/C2 −Cf 1 output current (−I2 ) and the output voltage V2 are in phase.
Vc2 = · VC 1 = · · V1 . Therefore, the output voltage V2 is leading the input voltage V1
1/Cs + 1/C2 1/Cs + 1/C2 Cp1
(4) by 90°.
Considering the expression (1), the voltage on C2 can be Second, Fig. 2(c) is used to calculate the input current and
further simplified as input power. It needs to be emphasized again that the input
current only depends on the output voltage as shown in Fig. 2(c).
−Cs · Cf 1 Similarly to the previous analysis, the input current I1 can be
Vc2 = · V1 . (5)
C1 Cs + C2 Cs + C1 C2 expressed as
Since L1 and L2 are magnetically coupled and there is current −VC f 1 −jωCs · Cf 1 Cf 2 · V2 −jM12 V2
flow through L1 , there is also a voltage difference between the I1 = = + . (13)
jωLf 1 C1 Cs + C2 Cs + C1 C2 ωLf 1 Lf 2
two terminals of L2 . The dot end connection in Fig 2(b) deter-
mines the direction of this voltage. So the voltage on L2 can be This equation shows that the output voltage V2 is leading the
expressed as input current I1 by 90°. Since the previous analysis proves that
M12 the output voltage V2 is also leading the input voltage V1 by
VL 2 = · VL 1 = ω 2 M12 Cf 1 V1 . (6) 90°, the input voltage V1 and current I1 are in phase with each
L1
other. Therefore, the input power can be expressed as
M12 is the mutual inductance, which is written as
 Pin = |V1 | · |I1 | · cos(ϕV 1 − ϕI 1 ) = |V1 | · |I1 | . (14)
M12 = k L1 · L2 (7)
Expression (15) shows that unity power factor is realized
where k is the coupling coefficient. With the expression of (5)
at the input side and there is no reactive power injected into
and (6), the voltage on Cf 2 can be expressed as
the resonant tank. Therefore, there is no extra conduction loss
−Cs · Cf 1 · V1 induced by the circulating reactive current, and the efficiency of
VCf2 = VC 2 − VL 2 = − ω 2 M12 Cf 1 V1 .
C1 Cs + C2 Cs + C1 C2 the resonant circuit can be very high.
(8) Combining (13) and (14), the input power is further expressed
Equation (8) indicates that the polarity connection of L2 has as
to be the same as shown in Fig. 2(b). Otherwise, the voltage VC 2
ωCs · Cf 1 Cf 2 · |V1 | |V2 | M12 · |V1 | |V2 |
and VL 2 will be canceled, and the output power is, therefore, Pin = |V1 | · |I1 | = + .
reduced. C1 Cs + C2 Cs + C1 C2 ωLf 1 Lf 2
(15)
The voltage on Cf 2 is the same as the voltage on Lf 2 , so the
Comparing (11) and (15), it shows that the input power is the
output side current I2 can be expressed as
same as the output power, which is consistent with the previous
−VC f 2 Cs · Cf 1 · V1 /(jωLf 2 ) ω 2 M12 Cf 1 V1 assumption that the losses in all the components are neglected.
I2 = = + .
jωLf 2 C1 Cs + C2 Cs + C1 C2 jωLf 2 According to Dai and D.C. Ludois [30], the relationship between
(9) C1 , C2 , and Cs satisfies C1  Cs and C2  Cs . So (11) and
Considering the parallel resonance between Lf 2 and Cf 2 and (15) can be further simplified as
the expression (1), (9) can be simplified as √
Cf 1 Cf 2 k L1 L2
−jωCs · Cf 1 Cf 2 · V1 −jM12 · V1 Pin = Pout ≈ ωCs · · |V1 | |V2 | + · |V1 | |V2 | .
I2 = + . (10) C1 C2 ωLf 1 Lf 2
C1 Cs + C2 Cs + C1 C2 ωLf 1 Lf 2 (16)
As an H-bridge diode rectifier is used in Fig. 1 at the output
side, the fundamental output current (−I2 ) and output voltage III. INDUCTIVE AND CAPACITIVE COUPLER DESIGN
V2 are in phase with each other. Therefore, the output power can
A. Inductive Coupler Design
be expressed as
The switching frequency needs to be specified at the begin-
ωCs · Cf 1 Cf 2 |V1 | |V2 |
Pout = |V2 | · |(−I2 )| = ning of the coupler design. From (12), PCPT is proportional
C1 Cs + C2 Cs + C1 C2 to the coupling capacitor Cs , which is usually in the picofarad
M12 |V1 | |V2 | range, so the switching frequency has to be increased to the MHz
+ . (11)
ωLf 1 Lf 2 level to achieve the desired power transfer by the capacitors.
Since the inductive couplers (coils) share the same frequency as
In expression (11), the first term is determined by the capac-
the capacitive couplers (plates), the switching frequency should
itive power transfer, and the second term is determined by the
be limited to reduce the skin effect. As a result, the frequency is
IPT
⎧ set to 1 MHz.
⎪ ωCs · Cf 1 Cf 2 · |V1 | · |V2 |
⎪
⎨ PCPT = C C + C C + C C
The contribution of this paper mainly lies in the validation
1 s 2 s 1 2
. (12) of the IPT-CPT combined system. For simplicity, shielding of
⎪
⎪ M · |V | · |V | the leakage magnetic field is not considered in the coil design.
⎩ PIPT = 12 1 2
ωLf 1 Lf 2 According to a previous study, the ferrite and aluminum plate
LU et al.: INDUCTIVE AND CAPACITIVE COMBINED WIRELESS POWER TRANSFER SYSTEM WITH LC-COMPENSATED TOPOLOGY 8475

Fig. 3. Dimension of the coils. (a) 3-D view of the coils. (b) Front view of the Fig. 4. Maxwell simulation of the coupling coefficient and self-inductance.
coils. (a) Coupling coefficient at different dimensions. (b) Self-inductance at different
dimensions.

can be used together to shield the leakage field and they will The circuit model of the plate is shown in Fig. 6. To simplify
incur 1%–3% extra loss, depending on the structure [33]. the design process, the coupler is symmetric with respect to
The dimension of the coil is shown in Fig. 3(a) and (b), where the primary and secondary sides. All six mutual capacitances
d1 is the outer length of the coil, d2 is the inner length, d is the between each two plates are illustrated in Fig. 6. Because of
air gap, and w is the coil width. The width ratio is defined as the symmetry, C12 = C34 and C14 = C23 . In the equivalent
rw = w/d1 . The coil has a square shape, and the air-gap distance circuit, C1 = C2 . It needs to be emphasized that C13 and C24
is 150 mm. The coil thickness has a little effect on the coupling do not have to be equal; the same for Cs1 and Cs2 . The nodal
coefficient, and is determined by the Litz-wire diameter. In this voltage method can be used to derive the relationship between
case, the thickness is 3.2 mm. the original model and the equivalent model. The equivalent
The coupling coefficient k and self-inductance Ls of the model is used in Section II-B.
coils are simulated in Maxwell, and shown in Fig. 4(a) and For the plates, there are three parameters that need to be
(b). Fig. 4(a) indicates that the coupling coefficient k is mostly determined: l1 , l2 , and dc . In this design, the outer plate length
determined by the coil outer length d1 . When d1 is between 300 is set to be 610 mm due to the space limitation. The other two
to 450 mm, k is in the range of 0.1 to 0.25. Fig. 4(b) shows that parameters, l2 and dc , are varied to find the best dimensions.
the self-inductance Ls can be regulated by varying the width All the mutual capacitances can be obtained from Maxwell
ratio rw without affecting coupling coefficient. and converted to equivalent parameters. The plate ratio rp is
In the IPT-CPT combined system, the coils resonate with the defined as rp = l2 /l1 . The capacitances Cs = Cs1 Cs2 /(Cs1 +
metal plates. Therefore, the coil size cannot be decided with- Cs2 ) and C1,2 are shown in Fig. 7. This figure indicates that Cs
out considering the dimension of the plates. The balancing of is not highly sensitive to the plate distance dc .
the output power of the IPT and CPT system is also an im- The plate ratio is set to 0.75 to maximize the coupling capac-
portant specification. The coil dimensions will be determined in itance. The parameter dc is, therefore, the only parameter that
Section IV, where a specific 3.0-kW design example is provided. needs to be determined. It is used to determine the capacitances
C1,2 , and, hence, the resonance with the inductive coils. In Sec-
tion IV, the distance dc will be determined in a 3.0-kW design
B. Capacitive Coupler Design example.
Four metal plates are used to form capacitors to transfer power
through electric fields, and they are separated into two pairs IV. 3.0-KW EXAMPLE FOR EV CHARGING
at the primary and secondary. The structure of the capacitive
coupler is shown in Fig. 5, where l1 is the length of the outer A. Power Ratio of the IPT and CPT System
plate, l2 is the length of the inner plate, dc is the distance between This system utilizes both inductive and capacitive coupling
the inner and outer plate, and d is the air-gap distance same as to transfer power. It is important to determine the power ratio of
the inductive coupler. The plate thickness is set to be 2 mm. each part, which is the percentage of their contributions. Since
8476 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 31, NO. 12, DECEMBER 2016

Fig. 7. Capacitances at rp and dc when l1 = 610 mm. (a) Capacitance C s .

(b) Capacitance C1 , 2 .

Fig. 5. Dimension of the plates. (a) 3-D view of the plates. (b) Front view of
the Plates.

Fig. 8. Power ratio between the IPT and CPT system.

Fig. 6. Circuit model of the plates.

Since L1 , Cf 1 , and Cp1 form a parallel resonance, the inductor
L1 can be replaced by the capacitor Cf 1 and Cp1
the IPT and CPT coupler have similar physical size, they can be PIPT C1 C1 + 2Cs Cp1 + Cf 1
designed to transfer the same order of magnitude of power to rI −C = =k· · .
PCPT Cs C1 + Cs Cs /(C1 + Cs ) Cf 1
the load. According to (1), the power ratio of inductive power (20)
to capacitive power can be defined as Referring to [32], it is known that Cf 1  Cp1 , so (20) is
approximated as
PIPT M12 · (C1 Cs + C2 Cs + C1 C2 ) PIPT C1 C1 + 2Cs
rI −C = = . (17) rI −C = ≈k· · . (21)
PCPT ωLf 1 Lf 2 · ωCs · Cf 1 Cf 2 PCPT Cs C1 + Cs Cs /(C1 + Cs )
The capacitor ratio between C1 and Cs can be defined as
Considering (1), (2), and (7), (17) can be simplified as kC = Cs /C1 . Therefore, the power ratio can be written as
√ PIPT k 1 + 2 · kC
PIPT ω 2 k L1 L2 · (C1 Cs + C2 Cs + C1 C2 ) rI −C = ≈ · . (22)
rI −C = = . PCPT kC 1 + kC2 /(1 + kC )
PCPT Cs
(18) Fig. 4(a) shows that the coupling coefficient k varies from
Since L1 = L2 and C1 = C2 , (18) can be further rewritten: 0.10 to 0.20, and Fig. 7 shows that the capacitor ratio kc varies
from 0.04 to 0.10. Therefore, the power ratio between the IPT
PIPT ω 2 k · L1 C1 (2Cs + C1 ) and CPT system is shown in Fig. 8, in which the IPT system
rI −C = = (19) contributes more to the output power.
PCPT Cs
LU et al.: INDUCTIVE AND CAPACITIVE COMBINED WIRELESS POWER TRANSFER SYSTEM WITH LC-COMPENSATED TOPOLOGY 8477

Fig. 10. Normalized values of C s and k at different misalignments.

Fig. 9. Parameter relationship in a 3.0-kW Example. (a) C1 versus IPT/CPT

power percentage. (b) C1 versus L1 value.

TABLE I
SYSTEM SPECIFICATIONS AND CIRCUIT PARAMETERS

Parameter Design Value Parameter Design Value

Fig. 11. Simulated input and output voltage and current waveforms.

Vin 310 V Vo u t 320 V

PIP T 2160 W (72%) PC P T 840 W (28%)
d1 300 mm l1 610 mm
With the dimensions in Table I, the FEA analysis of the two
d2 90 mm l2 457 mm couplers are conducted at different misalignment conditions.
d 150 mm dc 25 mm For the capacitive coupler, the variations of C1,2 are within 3%
rw 0.35 rp 0.75
k 0.130 kc 0.052
when the misalignemnt increases to 30 cm, so only the variation
fs w 1 MHz Cs 5.0 pF of Cs is considered. For the inductive coupler, the variation of
Lf 1 14.2 μH Lf 2 14.2 μH L1,2 are within 2%, so only the variation of coupling coeffcient k
Cf 1 1.78 nF Cf 2 1.78 nF
L1 256.2 μH L2 264.1 μH
is considered. The normalized values of coupling capacitor Cs
C1 96.1 pF C2 96.1 pF and inductive coupling coeffcient k at different misalignment
conditions are shown in Fig. 10. It shows that, in this system,
the capacitive coupler has better misalignment ability than the
inductive coupler.
B. System Parameter Design LTspice is used to simulate the system input and output power
In order to get significant power transfer through CPT cou- and waveform, as shown in Fig. 11. The voltage and currrent
pling, the maximum power ratio is set to be 3.0. The coil length are almost in phase with each other at both the input and output,
is set to be 300 mm and the coupling coefficient is 0.130. Fig. 7 which is consistent with the analysis in Section II. Also, the
shows that Cs is around 5.0 pF. In a 3.0-kW system, the input cut off current at the switching transient is 3 A to provide soft
voltage is 310Vdc and output voltage is 320 Vdc . Substitution of switching.
Vin , Vout , k, Cs , and Pin into (1), (2), (15), and (20), shows that LTspice simulation also provides the voltage stress between
C1 determines the percentage contribution of IPT and CPT, and the metal plates. The RMS voltage between P1 and P2 is 4.9 kV,
also the inductance value L1 , as shown in Fig. 9. Considering and the RMS voltage between P1 and P3 is 3.4 kV. The break-
the space limitation for the coil, its inductance L1 cannot be too down voltage of dry air is 3.0 kV/mm and the distance between
large, so the minimum power ratio is set at 2.0. Therefore, the P1 and P2 is 25 mm, so there is no concern with arcing.
acceptable power ratio is between 2.0 and 3.0. The magnetic field distribution of the inductive coupler is
The plate distance dc is set to be 25 mm and C1 = 96.1 pF. shown in Fig. 12(a), and the electric field distribution of the
Fig. 9(b) indicates that L1 = 264.1 μH, which corresponds to capacitive coupler is shown in Fig. 12(b).
rw = 0.35. Fig. 9(a) shows that the IPT system provides 72% of According to the ICNIRP 2010, the safety exposure limit for
the total power and the CPT system provides 28%. The designed the general public to magnetic fields is 6.25 μT [34]. In our
system specifications and parameters are shown in Table I. L1 is setups, an aluminum shield was not used. The simulation result
7% smaller than L2 to provide soft-switching conditions [11]. in Fig. 12(a) shows that the safe range is 600 mm away from
8478 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 31, NO. 12, DECEMBER 2016

High-frequency thin-film capacitors from KEMET are used

to realize the compensation capacitors Cf 1 and Cf 2 . At 1 MHz,
the RMS voltage rating of a single capacitor is about 300 V,
and the RMS current rating of a single capacitor is about 1.0 A.
Therefore, multiple capacitors are connected in series and par-
allel to satisfy the voltage and current requirements. According
to the datasheet, the capacitance values have ±5.0% tolerance
from the nominal value. To maintain the accuracy of the ca-
pacitance values, they are measured and tuned by the Agilent
Network Analyze E5072B. The dissipation factor of the capac-
itors is 18 × 10−4 at 1 MHz, which is used to calculate the
equivalent series resistance of the capacitors.
Fig. 12. Fields distribution of the coupler. (a) Magnetic field distribution. Aluminum plates are used to construct the capacitive cou-
(b) Electric field distribution.
plers. The outer plate length is 610 mm, and the inner plate
length is 457 mm. The distance between the outer and inner
plates is 25 mm, and the thickness of the plates is 2 mm. PVC
tubes are used to hold the plates. The plates are clamped into
slots on the tubes, and it is convenient to move the plates in
the slots to adjust the misalignment. Since there is high voltage
stress between the plates, ceramic spacers are used to sepa-
rate the outer and inner plate at the same side. In this design,
the other circuit components are connected to the edge of the
plates, which influences the current distribution in the plates. In
future research, the connection of the plates and its influence on
efficiency will be studied.
Fig. 13. Experiment prototype of a 3.0-kW LC-compensated IPT-CPT SiC MOSFETs (C2M0080120D) are used in the input in-
System.
verter, and SiC diodes (IDW30G65C5) are used in the output
rectifier, as mentioned in Section II-A. The digital signal pro-
the magnetic coupler. In the future, if aluminum shielding is cessor TMS320F28335 is used to generate the PWM signals for
utilized, the magnetic radiation can be significantly reduced. the inverter. It is convenient to adjust the switching frequency
According to the IEEE C95.1 standard, the safety limit for and dead-time of the signals in the programming codes. The
electric fields is 614 V/m [35]. Fig. 12(b) shows that the safe datasheet shows that the parasitic capacitor between the drain
range is 700 mm away from the coupler. In the future, more and source of the MOSFET is only 100 pF at 310 V, so it is
research will be conducted to either reduce or shield the electric easy to charge and discharge this capacitor during the switching
field radiation. transient. The MOSFETs in the inverter can therefore achieve
soft-switching conditions.
V. EXPERIMENTAL DETAILS In Fig. 13, the distance between the inductive and capacitive
coupler is 300 mm to eliminate the influence between them. If
A. Experimental Setup the two couplers are placed close to each other, it can induce
A 3.0-kW experiment prototype is constructed to verify the some drawbacks. There are capacitive couplings between the
proposed system, which is shown in Fig. 13. All the components coils and plates, and so the equivalent capacitance between the
are implemented using the values in Table I. The inductors and plates can be changed by the proximity of the coils. As a result,
coils are made by 3000-strand AWG 46 Litz-wire with a single the resonances in the circuit can be changed. Also, the magnetic
wire diameter of 39.8 μm. Since the skin depth of copper is fields of the coils can generate eddy current losses in the metal
65.2 μm at 1 MHz, there is no significant increase of ac resis- plates, which can reduce the coil inductance and hence system
tance induced by the skin effect. However, according to Lu et al. efficiency. However, the long distance between the two couplers
[36], the proximity effect between bundles in the Litz-wire can- will decrease the power transfer density. In future research, the
not be neglected, and the measurement by Agilent LCR meter two couplers will be integrated as a single coupler to increase
E4980A shows that the ac resistance is 3.4 times the dc resis- the power density. In this case, the interaction between the coils
tance at 1 MHz. The inductors have an air core, and are wound and plates will be presented in detail.
on PVC tubes to eliminate the magnetic losses. The coils are
wound on plastic board and there is no magnetic material, as
B. Experimental Results
presented in Section III-A. The insulation between the adjacent
turns in the coils and inductors should be considered to avoid The experiment is conducted with the designed prototype.
voltage breakdown. In future designs, the high-frequency mag- The input dc source voltage is 310 V, and the electronic dc load
netic material 51 from Fair-Rite will be used in the inductors and works at constant 320 V to simulate the battery pack on the
coils to reduce their volume as well as magnetic fields radiation. vehicle. The experimental waveforms are shown in Fig. 14.
LU et al.: INDUCTIVE AND CAPACITIVE COMBINED WIRELESS POWER TRANSFER SYSTEM WITH LC-COMPENSATED TOPOLOGY 8479

Fig. 14. Experimental input and output waveforms.

In Fig. 14, the red curve is the square wave input voltage V1
provided by the inverter, and the pink curve is the sinusoidal
current I1 injected into the resonant tank. The green curve is
the output square wave voltage V2 to the rectifier, and the blue
curve is the driver signal to one of the MOSFETs. It shows
that the input voltage and current are almost in phase, so the
system achieves unity power factor. There is little reactive power
injected into the circuit, and the efficiency is maintained. The
output voltage V2 is 90° leading the input V1 , which agrees
with the analysis in Section II-C. Although there is noise in the
driver signal, its magnitude is within 3 V, which is lower than
the threshold voltage of MOSFETs. Therefore, it is safe to use
Fig. 15. Measured output power and efficiency for IPT and CPT system.
this signal as driver. (a) IPT + CPT system. (b) Only IPT system. (c). Only CPT system.
Fig. 15(a) shows the relationship between the output power
and efficiency of the IPT and CPT combined system. In the
increases to 20 cm, the capacitive coupler can maintain 80.2%
well-aligned case, the output power reaches a maximum of
of the well-aligned power, and the inductive coupler can only
2.84 kW with an efficiency of 94.45%. When both the inductive
maintain 38.5% of the original power. Therefore, the import
and capacitive couplers have 20-cm misalignment, the output
of the capacitive coupler to the IPT system can improve the
power drops to 1.35 kW, and the efficiency drops to 91.49%.
misalignment ability of the system.
Fig. 15(b) shows the output power and efficiency of the IPT
system. In this case, the two pairs of plates are separated far
C. Comparison of Modeled and Experimental Performance
away to eliminate the capacitive coupling. Only the coils are
used to transfer power. When the coils are well aligned, the max- The experimental performance of the IPT and CPT combined
imum output power is 1.95 kW with an efficiency of 94.89%. system is compared to the modeled and simulated system in
When the coils have 20-cm misalignment, the output power is Section IV. The comparisons include the input voltage and cur-
0.75 kW, and the efficiency drops to 91.68%. rent waveforms and the system power and power distribution at
Fig. 15(c) shows the output power and efficiency of the CPT different misalignment conditions.
system. In this case, the two coils are separated to eliminate the Fig. 16 shows the comparison of the input voltage and current
inductive coupling. Only the metal plates are used to transfer waveforms between the simulated and experimental results, and
power. When the plates are well aligned, the maximum output they agree well with each other. The dashed blue curve is the
power is 0.86 kW with the efficiency of 93.04%. When the plates simulated input voltage, which is a square wave to accelerate
have 20-cm misalignment, the output power is 0.69 kW, and the calculation speed. The solid red curve is the experimental input
efficiency drops to 91.18%. voltage, which has a pulse at the switching transient. This is
The output power in Fig. 15(a) is the sum of the output power because there is voltage stress on the MOSFETs at the turn-on
in Fig. 15(b) and (c), which shows that both the inductive and transient, which can induce some switching losses. However,
capacitive coupler can transfer power in the combined system. the magnitude of the pulse is lower than 50 V, so the induced
The capacitive coupler has better misalignment ability when power loss can be neglected. The dashed green curve and solid
compared to the inductive coupler. When the misalignment pink curve are the simulated and experimental input current,
8480 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 31, NO. 12, DECEMBER 2016

than the experiment result. This is because the calculation is

based on FHA, and all the high-order harmonics currents are ne-
glected. However, with the increase of misalignment, the high-
order currents are becoming relatively larger when compared to
the fundamental current, so the experiment result is larger than
the calculation. It also needs to be emphasized that the differ-
ence of power between the calculated and experiment results
also increase with misalignment. For example, Fig. 10 shows
that, at 20-cm misalignment, the IPT power drops to 25.2%, and
the CPT power drops to 62.1%. However, Fig. 15 shows that,
in experiments, the IPT power drops to 38.5%, and the CPT
power drops to 80.2%. This difference also comes from the ap-
proximation of FHA approach. However, the absolute value of
the power difference is still in an acceptable range. In future re-
search, the calculation expressions will be modified to consider
Fig. 16. Comparison of waveforms in simulated and experimental results.
all the high-order components.
Fig. 17(b) shows the calculated and experimental power ratios
between the IPT and CPT system. The maximum ratio differ-
ence is within 0.30. When the misalignment increases, both the
inductive and capacitive power decreases. The power ratio is
also decreasing, which means the capacitive power becomes
comparable to the inductive power.

VI. CONCLUSION
This paper proposes a wireless power transfer system that
combines inductive and capacitive power transfer. The innova-
tion is that the inductive coupler and capacitive coupler resonate
together, using the compensation components to transfer power.
An LC compensation topology is required at both the primary
and secondary sides. The working principle of the IPT + CPT
system is analyzed to derive the expression of the output power
and the system design process. The inductive and capacitive
couplers are designed using 3-D simulations in Maxwell. A
3.0-kW prototype is constructed to verify the proposed con-
cept. The output power of the combined system is the sum of
the IPT and CPT systems. The high voltage between the metal
plates is an important safety concern in application. In the fu-
Fig. 17. Comparison of experiment and calculation results. (a) Output power ture, the system will be optimized to reduce the voltage and the
of IPT and CPT combined system. (b) Power ratio between IPT and CPT system.
corresponding radiated EMI. Also, the inductive and capacitive
coupler can be constructed together to make the system more
respectively. The cut off currents of the switching transient are compact.
the same.
When there is misalignment, Cs and k decrease as shown REFERENCES
in Fig. 10. The FEA analysis shows that the variations of L1,2
and C1,2 can be neglected. Therefore, (12) and (15) and the [1] J. Kim, B. Lee, J. Lee, and S. Lee, “Development of 1 MW inductive
power transfer system for a high speed train,” IEEE Trans. Ind. Electron.,
parameters in Table I and Fig. 10 are used to calculate the system vol. 62, no. 10, pp. 6242–6250, Oct. 2015.
power at different misalignment conditions. The comparisons of [2] C. Zheng, J. Lai, and L. Zhang, “Design consideration to reduce gap vari-
the calculation and experiment results of the system total power ation and misalignment effects for the inductive power transfer system,”
IEEE Trans. Power Electron., vol. 30, no. 11, pp. 6108–6119, Nov. 2015.
and power ratio are shown in Fig. 17. [3] D. C. Ludois, M. J. Erickson, and J. K. Reed, “Aerodynamic fluid bearing
Fig. 17(a) shows that the calculated and experimental results for translational and rotating capacitors in noncontact capacitive power
of the system total power are close at different misalignment transfer systems,” IEEE Trans. Ind. Appl., vol. 50, no. 2, pp. 1025–1033,
Mar./Apr. 2014.
conditions. The maximum power difference is within 250 W. [4] J. Dai and D. Ludois, “A survey of wireless power transfer and a critical
When the misalignment is small, the calculation is larger than comparison of inductive and capacitive coupling for small gap applica-
the experiment result. This is because the calculation does not tions,” IEEE Trans. Power Electron., vol. 30, no. 11, pp. 6017–6029,
Nov. 2015.
consider the power losses in the circuit components. When the [5] C. Park, S. Lee, S. Jeong, G. Cho, and C. T. Rim, “Uniform power I-
misalignment increases, the calculation decreases to be lower type inductive power transfer system with DQ-power supply rails for
LU et al.: INDUCTIVE AND CAPACITIVE COMBINED WIRELESS POWER TRANSFER SYSTEM WITH LC-COMPENSATED TOPOLOGY 8481

on-line electric vehicles,” IEEE Trans. Power Electron., vol. 30, no. 11, [28] L. Huang, A. P. Hu, A. Swwain, and X. Dai, “Comparison of two high
pp. 6446–6455, Nov. 2015. frequency converters for capacitive power transfer,” in Proc. IEEE Energy
[6] S. Moon, B. Kim, S. Cho, C. Ahn, and G. Moon, “Analysis and design Convers. Congr. Expo., Pittsburgh, PA, USA, Sep. 2014, pp. 5437–5443.
of a wireless power transfer system with an intermediate coil for high [29] T. Kitabayashi, H. Funato, H. Kobayashi, and K. Yamaichi, “Experimental
efficiency,” IEEE Trans. Ind. Electron., vol. 61, no. 11, pp. 5861–5870, verification of capacitive power transfer using one pulse switching active
Nov. 2014. capacitor for practical use,” in Proc. IEEE Int. Power Electron. Conf.,
[7] J. Kim, D. Kim, and Y. Park, “Analysis of capacitive impedance matching Hiroshima, Japan, May 2014, pp. 2517–2522.
network for simultaneous wireless power transfer to multiple devices,” [30] J. Dai and D. C. Ludois, “Single active switch power electronics for kilo-
IEEE Trans. Ind. Electron., vol. 62, no. 5, pp. 2807–2813, May 2015. watt scale capacitive power transfer,” IEEE J. Emerg. Sel. Power Electron.,
[8] D. Ahn and S. Hong, “Wireless power transmission with self-regulated vol. 3, no. 1, pp. 315–323, Mar. 2015.
output voltage for biomedical implant,” IEEE Trans. Ind. Electron., [31] J. Dai and D. C. Ludois, “Capacitive power transfer through a confor-
vol. 61, no. 5, pp. 2225–2235, May 2014. mal bumper for electric vehicle charging,” IEEE J. Emerg. Sel. Power
[9] S. Huang, E. Lee, and T. Huang, “Inductive power transfer systems for PT- Electron., to be published, doi: 10.1109/JESTPE.2015.2505622
based ozone-driven circuit with flexible capacity operation and frequency- [32] F. Lu, H. Zhang, H. Hofmann, and C. Mi, “A double-sided LCLC-
tracking mechanism,” IEEE Trans. Ind. Electron., vol. 61, no. 12, compensated capacitive power transfer system for electric vehicle charg-
pp. 6691–6699, Dec. 2014. ing,” IEEE Trans. Power Electron., vol. 30, no. 11, pp. 6011–4014,
[10] M. Fu, H. Yin, X. Zhu, and C. Ma, “Analysis and tracking of optimal load Nov. 2015.
in wireless power transfer systems,” IEEE Trans. Power Electron., vol. 30, [33] F. Lu, H. Hofmann, J. Deng, and C. Mi, “Output power and efficiency sen-
no. 7, pp. 3952–3963, Jul. 2015. sitivity to circuit parameter variations in double-sided LCC-compensated
[11] S. Li, W. Li, J. Deng, T. D. Nguyen, and C. Mi, “A double-sided LCC wireless power transfer system,” in Proc. IEEE Appl. Power Electr. Conf.,
compensation network and its tuning method for wireless power transfer,” Charlotte, VA, USA, Mar. 2015, pp. 507–601.
IEEE Trans. Veh. Technol., vol. 64, no. 6, pp. 2261–2273, Jun. 2015. [34] “ICNIRP guidelines for limiting exposure to time-varying electric, mag-
[12] J. Deng, F. Lu, S. Li, T. Nguyen, and C. Mi, “Development of a high netic and electromagnetic fields (1 Hz to 100 kHz),” Health Phys.,
efficiency primary side controlled 7 kW wireless power charger,” in Proc. vol. 99, pp. 818–836, 2010.
IEEE Electr. Veh. Conf., Florence, Italy, Dec. 2014, pp. 1–6. [35] IEEE Standard for Safety Levels with Respect to Human Exposure to Ra-
[13] J. Deng, S. Li, S. Hu, C. Mi, and R. Ma, “Design methodology of LLC dio Frequency Electromagnetic Fields, 3kHz to 300 GHz, IEEE Standard
resonant converters for electric vehicle battery chargers,” IEEE Trans. Veh. C95.1, 2005.
Technol., vol. 63, no. 4, pp. 1581–1592, May 2014. [36] F. Lu, H. Zhang, H. Hofmann, and C. Mi, “A high efficiency 3.3 kW
[14] J. Deng, W. Li, S. Li, and C. Mi, “Magnetic integration of LCC com- loosely-coupled wireless power transfer system without magnetic mate-
pensated resonant converter for inductive power transfer applications,” rial,” in Proc. IEEE Energy Convers. Congr. Expo., Montreal, Canada,
in Proc. IEEE Energy Convers. Congr. Expo., Pittsburgh, PA, USA, Sep. 2015, pp. 2282–2286.
Sep. 2014, pp. 660–667.
[15] J. Deng, W. Li, T. Nguyen, S. Li, and C. Mi, “Compact and efficient
bipolar pads for wireless power chargers: Design and analysis,” IEEE
Trans. Power Electron., vol. 30, no. 11, pp. 6130–6140, Nov. 2015.
[16] W. Li, C. Mi, S. Li, J. Deng, T. Kan, and H. Zhao, “Integrated LCC
compensation topology for wireless charger in electric and plug-in elec-
tric vehicle,” IEEE Trans. Ind. Electron., vol. 62, no. 7, pp. 4215–4225,
Jul. 2015.
[17] D. Shmilovitz, A. Abramovitz, and I. Reichman, “Quasi resonant LED
driver with capacitive isolation and high PF,” IEEE J. Emerg. Sel. Topics Fei Lu (S’12) received the B.S. and M.S. degrees
Power Electron., vol. 3, no. 3, pp. 633–641, Sep. 2015. in electrical engineering from the Harbin Institute of
[18] C. Liu, A. P. Hu, and N. C. Nair, “Coupling study of a rotary capacitive Technology, Harbin, China, in 2010 and 2012, re-
power transfer system,” in Proc. IEEE Int. Ind. Technol. Conf., Gippsland, spectively. He is currently working toward the Ph.D.
Australia, Feb. 2009, pp. 1–6. degree in electrical engineering from the University
[19] A. P. Hu, C. Liu, and H. Li, “A novel contactless battery charging system of Michigan, Ann Arbor, MI, USA.
for soccer playing robot,” IEEE Int. Mechatron. Mach. Vis. Pract. Conf., He is working on the high-power and high-
Auckland, New Zealand, Dec. 2008, pp. 646–650. efficiency capacitive power transfer through an air-
[20] H. Funato, H. Kobayashi, and T. Kitabayashi, “Analysis of transfer power gap distance up to 100’s of millimeters. He is also
of capacitive power transfer system,” in Proc. IEEE Power Electr. Drive working on the application of widebandgap devices
Syst., Kitakyushu, Japan, Apr. 2013, pp. 1015–1020. on the WPT system. His research interest includes
[21] P. Srimuang, N. Puangngernmak, and S. Chalermwisutkul, “13.56 MHz wireless power transfer for the application of electric vehicle charging.
class E power amplifier with 94.6% efficiency and 31 watts output power
for RF heating applications,” in Proc. IEEE Electr. Eng. Comp. Telecom-
mun. Inf. Tech. Conf., Nakhon Ratchasima, Thailand, May 2014, pp. 1–5.
[22] A. Kumar, S. Pervaiz, C. K. Chang, S. Korhummel, Z. Popovic, and
K. K. Afridi, “Investigation of power transfer density enhancement in
large air-gap capacitive wireless power transfer system,” in Proc. IEEE
Wireless Power Transfer Conf., Boulder, CO, USA, May 2015, pp. 1–4.
[23] C. Liu and A. P. Hu, “Power flow control of a capacitively coupled con-
tactless power transfer system,” in Proc. IEEE Ind. Electr. Conf., Porto,
Portugal, Nov. 2009, pp. 743–747.
[24] M. Kline, I. Izyumin, B. Boser, and S. Sanders, “Capacitive power transfer
for contactless charging,” in Proc. IEEE Appl. Power Electr. Conf., Fort Hua Zhang (S’14) received the B.S. and M.S. de-
Worth, TX, USA, Mar. 2011, pp. 1398–1404. grees in electrical engineering from Northwestern
[25] C. Xia, C. Li, and J. Zhang, “Analysis of power transfer characteristic of Polytechnical University, Xi’an, China, in 2011 and
capacitive power transfer system and inductively coupled power transfer 2014, respectively, where she is currently working
system,” in Proc. IEEE Mech. Sci. Electr. Eng. Comput., Jilin, China, toward the Ph.D. degree in electrical engineering.
Aug. 2011, pp. 1281–1285. From September 2014 to August 2015, she was a
[26] M. P. Theodoridis, “Effective capacitive power transfer,” IEEE Trans. joint Ph.D. Student founded by the China Scholarship
Power Electron., vol. 27, no. 12, pp. 4906–4913, Dec. 2012. Council with the University of Michigan, Dearborn,
[27] B. H. Choi, D. T. Dguyen, S. J. Yoo, J. H. Kim, and C. T. Rim, “A novel MI, USA.
source-side monitored capacitive power transfer system for contactless Since September 2015, she has been with San
mobile charger using class-E converter,” in Proc. IEEE Veh. Technol. Diego State University, San Diego, CA, USA. Her
Conf., Seoul, Korea, May 2014, pp. 1–5. research interest includes coupler design of high-power IPT and CPT system.
8482 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 31, NO. 12, DECEMBER 2016

Heath Hofmann (M’89–SM’15) received the B.S. Chunting Chris Mi (S’00–M’01–SM’03–F’12) re-
degree in electrical engineering from The University ceived the B.S.E.E. and M.S.E.E. degrees in electrical
of Texas at Austin, Austin, TX, USA, in 1992, and the engineering from Northwestern Polytechnical Uni-
M.S. and Ph.D. degrees in electrical engineering and versity, Xi’an, China, and the Ph.D. degree in elec-
computer science from the University of California, trical engineering from the University of Toronto,
Berkeley, CA, USA, in 1997 and 1998, respectively. Toronto, ON, Canada, in 1985, 1988, and 2001,
He is currently an Associate Professor with the respectively.
University of Michigan, Ann Arbor, MI, USA. His He was with University of Michigan, Dearborn,
research interests include the design, analysis, and MI, USA, from 2001 to 2015. He was the President
control of electromechanical systems, and power and Chief Technical Officer of 1 Power Solutions,
electronics. Inc., from 2008 to 2011. He is the Cofounder of Gan-
non Motors and Controls LLC and Mia Motors, Inc. He is currently a Professor
and Chair of electrical and computer engineering and the Director of the Depart-
ment of Energy—funded Graduate Automotive Technology Education Center
for Electric Drive Transportation, San Diego State University, San Diego, CA,
USA.
He has conducted extensive research and has published more than 100 jour-
nal papers. He has taught tutorials and seminars on the subject of HEVs/PHEVs
for the Society of Automotive Engineers, the IEEE, workshops sponsored by
the National Science Foundation, and the National Society of Professional En-
gineers. He has delivered courses to major automotive OEMs and suppliers,
including GM, Ford, Chrysler, Honda, Hyundai, Tyco Electronics, A&D Tech-
nology, Johnson Controls, Quantum Technology, Delphi, and the European
Ph.D. School. He has offered tutorials in many countries, including the U.S.,
China, Korea, Singapore, Italy, France, and Mexico. He has published more than
100 articles and delivered 30 invited talks and keynote speeches. He has also
served as a Panelist in major IEEE and SAE conferences. His research interests
include electric drives, power electronics, electric machines, renewable-energy
systems, and electrical and hybrid vehicles.
Dr. Mi received the “Distinguished Teaching Award” and the “Distinguished
Research Award” of the University of Michigan Dearborn. He also received the
2007 IEEE Region 4 “Outstanding Engineer Award,” the “IEEE Southeastern
Michigan Section Outstanding Professional Award,” and the “SAE Environ-
mental Excellence in Transportation.”