IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 27, NO. 5,MAY 2012 2471
Analysis of Inductorless Zero-Voltage-Switching
Piezoelectric Transformer-Based Converters
Edward L. Horsley, Alfredo V. Carazo, Nam Nguyen-Quang, Martin P. Foster, and David A. Stone
Abstract—The inductorless piezoelectric transformer (PT)-
based resonant converter topology allows all the components be-
tween the half-bridge inverter and rectifier in a conventional LCC
converter to be replaced with a single ceramic component. This of-
fers potential savings in cost, size, and mass. However, zero-voltage
switching (ZVS) becomes more difficult to achieve because the
MOSFET output capacitances are augmented by the PT input ca-
pacitance. This paper presents an analytical model for the ZVS
condition in inductorless PT-based converters. Unlike previously
reported models, the proposed model is shown to offer a level
of accuracy comparable to a SPICE simulation and to correlate
well with experimental results.Using a normalization scheme and
numerical optimization techniques, the criteria for achieving in-
ductorless ZVS are found in terms of the equivalent circuit compo-
nents. Both ac-output and dc-output variants are considered, and
design charts for ensuring ZVS in five different topologies are pre-
sented. The results and design charts are applicable to any type of
PT that can be represented by the standard PT equivalent circuit.
Index Terms—Inductorless, piezoelectric transformer (PT),
resonant power conversion, zero-voltage switching (ZVS).
I. INTRODUCTION
T
HERE has been much interest over the past two decades
in the development of piezoelectric transformers (PTs) to
replace many of the reactive components in certain switch mode
power supply topologies [1], [2]. Since the reactive components
in a power electronic circuit are usually the largest and most
expensive, the use of a PT can yield savings in cost, size, and
mass.
The topologies that are most suited for use with PTs are
push-pull, half-bridge, and class-E [1]. The push-pull and class-
E topologies require one or more additional inductors to be
added to the input side of the PT [1], [3], [4]. While the
half-bridge topology is also commonly used with an induc-
tor in series or parallel with the PT input, it is possible to re-
Manuscript received August 4, 2010; revised April 10, 2011; accepted
September 6, 2011. Date of current version February 27, 2012. This paper
was presented in part at the IEEE Power Electronics and Drive Systems Con-
ference, Taiwan, November 2–5. Recommended for publication by Associate
Editor S. Y. (R.) Hui.
E. L. Horsley, M. P. Foster, and D. A. Stone are with the Department
of Electronic and Electrical Engineering, University of Sheffield, Sheffield
S1 3JD, U.K. (e-mail: ehorsley@theiet.org; m.p.foster@sheffield.ac.uk;
d.a.stone@sheffield.ac.uk).
A. V. Carazo is with Micromechatronics, Inc., State College, PA 16803 USA
(e-mail: avc@mmech.com).
N. Nguyen-Quang was with the University of Sheffield, Sheffield S1 3JD,
U.K. He is now with Ho Chi Minh City University of Technology, Vietnam
(e-mail: nqnam@hcmut.edu.vn).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TPEL.2011.2169431
Fig. 1. Inductorless PT-based ac-output converter.
move this inductor completely if the PT is designed appropri-
ately [5]–[10]. The half-bridge topology without an additional
inductor at the PT input is often referred to as the “inductorless”
configuration.
The lumped equivalent circuit model for a PT is topologically
similar to a conventional LCC series–parallel converter, but with
an additional capacitance across the PT input. When used in the
inductorless topology, such as the ac-output variant shown in
Fig. 1, the output capacitance of the MOSFETs is augmented
by the input capacitance of the PT, C
in
. Since the former is of
the order of tens to hundreds of picofarad, whereas the latter
is often several nanofarad, the amount of charge that must be
delivered by the tank current i
L 1
during the dead-time in order
for zero-voltage switching (ZVS) to be achieved is much greater
than in a conventional discrete converter. A much larger dead-
time must also be used in order to provide sufficient time for this
charge to be delivered. Therefore, inductorless converters will
typically be operated further above resonance than their con-
ventional counterparts because the maximum dead-time (the
time between one MOSFET being turned OFF and the other
MOSFET being turned ON) that can be used is dictated by the
time between the low-side MOSFET (i.e., S
2
in Fig. 1) turn-
ing OFF and the tank current crossing zero, which in turn is
partly dictated by the phase angle at which the tank is oper-
ated. Since the total capacitance across the half-bridge is con-
siderable, the voltage across the MOSFETs is heavily snubbed
during turn-off. Therefore, ZVS turn-off is easily obtained, and
the primary concern throughout this paper is the need for ZVS
turn-on.
If the PT input capacitance is too large, the tank current will
not be able to supply sufficient charge during the dead-time and
ZVS will not be achieved. Since the total capacitance across the
half-bridge is much larger compared to a discrete converter, the
power dissipation that occurs in the MOSFET switches if ZVS
is not achieved is also much greater.
While reducing the PT input capacitance will allow induc-
torless ZVS to be achieved more readily, it will also decrease
PT power density because the size of the input section must be
0885-8993/$26.00 © 2011 IEEE