Professional Documents
Culture Documents
Voltage Multiplier Cells Applied To Non-Isolated DC-DC Converters
Voltage Multiplier Cells Applied To Non-Isolated DC-DC Converters
Abstract—This paper introduces the use of the voltage multiplier be used to reduce the switching losses and the EMI generation.
technique applied to the classical non-isolated dc–dc converters in However the voltage stress is higher than in the hard-switching
order to obtain high step-up static gain, reduction of the maximum structures and the cost and circuit complexity are increased.
switch voltage, zero current switching turn-on. The diodes reverse
recovery current problem is minimized and the voltage multiplier
Thus, the weight, volume and losses of the power transformer
also operates as a regenerative clamping circuit, reducing the are limiting factors for the isolated dc–dc converters used in
problems with layout and the EMI generation. These characteris- embedded applications.
tics allows the operation with high static again and high efficiency, Non-isolated dc–dc converters as the classical boost, can
making possible to design a compact circuit for applications where provide high step-up voltage gain, but with the penalty of
the isolation is not required. The operation principle, the design high voltage and current stress, high duty-cycle operation and
procedure and practical results obtained from the implemented
prototypes are presented for the single-phase and multiphase
limited dynamic response. The diode reverse recovery current
dc–dc converters. can reduce the efficiency when operating with high current and
A boost converter was tested with the single-phase technique, voltage levels. There are some non-isolated dc–dc converters
for an application requiring an output power of 100 W, operating operating with high static gain, as the quadratic boost converter,
with 12 V input voltage and 100 V output voltage, obtaining effi- but additional inductors and filter capacitors must be used and
ciency equal to 93%. The multiphase technique was tested with a the switch voltage is high [8].
boost interleaved converter operating with an output power equal
to 400 W, 24 V input voltage and 400 V output voltage, obtaining
However, recently new non-isolated dc–dc converter topolo-
efficiency equal to 95%. gies were proposed [1]–[9], showing that it is possible to obtain
high static gain, low voltage stress and low losses, improving
Index Terms—DC–DC power conversion, dc power system,
the performance with relation the classical topologies.
switched circuits, voltage multipliers.
A new alternative for the implementation of high step-up
structures is proposed in this paper with the use of the voltage
I. INTRODUCTION multiplier cells integrated with classical non-isolated dc–dc
converters. The uses of the voltage multiplier in the classical
HE recent growth of battery powered applications and low
T voltage storage elements are increasing the demand of ef-
ficient step-up dc–dc converters. Typical applications are em-
dc–dc converters add new operation characteristics, becoming
the resultant structure well suited to implement high-static gain
step-up converters.
bedded systems, renewable energy systems, fuel cells, mobility
applications and uninterrupted power supply (UPS) [1], [2] and
II. DC–DC CONVERTERS WITH SINGLE-PHASE
[3]. These applications demand high step-up static gain, high
VOLTAGE MULTIPLIER CELLS
efficiency and reduced weight, volume and cost.
The step-up stage normally is the critical point for the de- The use of voltage multiplier in low frequency rectifiers is
sign of high efficiency converters due to the operation with high a classical solution for high dc output voltage. Some of these
input current and high output voltage, thus a careful study must structures are shown in Fig. 1. This technique is also used
be done in order to define the topology for a high step-up appli- in high-frequency isolated dc–dc converters, mainly for high
cation. output voltage (kV) applications as in Traveling Wave Tube
Some classical converters with magnetic coupling as flyback Amplifiers (TWTA), reducing the problems presented by the
or current-fed push-pull converter can easily achieve high high frequency and high-voltage power transformers [10]. The
step-up voltage gain. However, the power transformer volume charge pump technique and switched-capacitor circuit is also
is a problem for the development of a compact converter. The a classical use of the capacitor charge transference [11]. These
energy of the transformer leakage inductance can produce high structures provide an output voltage higher than the input
voltage stress, increases the switching losses and the electro- voltage without the use of magnetic elements. The operation
magnetic interference (EMI) problems, reducing the converter at high-frequency permits a reduction of the capacitor’s size,
efficiency. Active clamping soft-commutation techniques can thus enabling the design of a single integrated circuit without
external components, for low power applications [13]. Some
Manuscript received April 22, 2007; revised August 2, 2007. Recommended implementations of switched-capacitor circuits are presented
for publication by Associate Editor H. Chung. in Fig. 2 and the implementation of dc–dc converters with
M. Prudente and L. L. Pfitscher are with the UNISINOS, São Leopoldo this technique are presented in [12] and [13]. A recent use of
93022–000, Brazil.
G. Emmendoerfer, E. F. Romaneli, and R. Gules are with the Federal Univer-
the capacitor charge transference in a new class of conversion
sity of Technology CPGEI-UTFPR, Curitiba 80230-901, Brazil. power cells for dc–dc voltage step-up has been presented in [5]
Digital Object Identifier 10.1109/TPEL.2007.915762 and [6]. The voltage lift technique is utilized to implement a
0885-8993/$25.00 © 2008 IEEE
872 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 23, NO. 2, MARCH 2008
Fig. 3. Voltage Multiplier cell integrated with classical dc–dc converters. (a)
Buck. (b) Boost. (c) Buck-boost.
integration of the voltage multiplier cell are the same for all
Fig. 2. Switched-capacitor circuits. basic converters. However, the use of the voltage multiplier
integrated with the buck converter does not introduce practical
advantages because the output voltage must be lower than
series of high voltage and wide conversion range applications the input voltage. As the main application of the structures
converters. studied is high static gain converters, only the boost topology
Another alternative to overcome the limitations of the clas- is analyzed in detail in this paper.
sical dc–dc converters for high performance and large conver- The voltage multiplier cell also operates without the reso-
sion ratio applications is proposed in this paper with the integra- nant inductor . However, the inclusion of this small induc-
tion of a voltage multiplier cells with non-isolated dc–dc con- tance (typically 1 H to 4 H) allows the power switch to op-
verters. The utilization of the voltage multiplier is also presented erate with zero-current-switching (ZCS) turn-on and the nega-
for the multiphase dc–dc converters, for better performance in tive effects of the reverse recovery current of all diodes are mini-
high power applications. mized. These characteristics reduce the converter commutation
losses, allowing the operation with high switching frequency,
maintaining high efficiency.
A. Circuit Description of the Single-Phase Converter
It is possible to add more multiplier cells in order to achieve
The basic structure of the single-phase voltage multiplier higher step-up ratios, as shown in Fig. 4. The reduction of the
cell is composed by the diodes , the capacitors reverse recovery current of all diodes is obtained with only one
and the resonant inductor . This voltage resonant inductor in the first voltage multiplier cell. The voltage
multiplier cell can be integrated with the classical converters multiplier cell increases the static gain of the classical boost by
as buck, boost and buck-boost, composed by the switch (S), a factor , where M is the number of multiplier cells.
inductor (L), output diode and filter capacitor , However, the maximum switch voltage is lower than the
as presented in Fig. 3. The new features obtained with the output voltage. In the simplest case shown in Fig. 3(b),
PRUDENTE et al.: VOLTAGE MULTIPLIER CELLS 873
Fig. 12. Integration of the voltage multiplier capacitor with the output capac-
itor.
defined by (10), (13) and (14). When the energy stored in the
capacitor is transferred to the capacitor, the diode
is blocked (instant ). The average voltage stored in the
output capacitor is equal to the output voltage of the classical
boost converter plus the voltage. The average voltage of
the capacitors and are equals to the output voltage of
the classical boost converter, and this is the maximum voltage
applied in all diodes and power switch
(10)
(11)
Fig. 11. Main theoretical waveforms of the single-phase converter.
(12)
(13)
(3)
3) Third Stage ( Fig. 8): At the instant , the switch As can be observed in Fig. 11, the switch turn on is ZCS.
S is turned-on with ZCS commutation and the current in the The resonant inductor limits the current variation (di/dt) in
resonant inductor and in the output diode reduce linearly all diodes, reducing the diodes reverse recovery current. The
to zero as defined by (7), at the instant . Thus the output voltage in all semiconductors is half of the output voltage, con-
diode also is blocked with low reverse recovery current. The sidering a low voltage ripple in the multiplier capacitors.
capacitor voltage can be considered constant due to the The basic structure also can be modified as shown in Fig. 12.
short duration of the third stage The voltage multiplier capacitor can compose the output
filter capacitor, reducing the output capacitor voltage level. A
(7) symmetrical output voltage is obtained with this
configuration, because the capacitor is charged with half
(8) of the output voltage even for unbalanced loads .
(9) Therefore, the symmetry between and is obtained
without the use of a control circuit. The configuration proposed
4) Fourth Stage ( Fig. 9): When the output diode is in Fig. 12 can be interesting for an integration of the step-up
blocked, the diode conducts transferring the energy stored dc–dc converter with a half-bridge inverter, as in a class D
in the capacitor to the capacitor , in a resonant way, power amplifier powered by a battery.
PRUDENTE et al.: VOLTAGE MULTIPLIER CELLS 875
C. Design Considerations of the Single-Phase Converter power considered in this example is equal to 150 W for a nom-
The main equations to design the single-phase converter are inal output power equal to 100 W
presented with an example, considering the following specifica-
tions.
Output power: 100 W. Switching frequency
Input Voltage: 12 V.
Output Voltage: 100 V. Voltage of the multiplier capacitor
Switching Frequency: 50 kHz. Maximum output power (25)
Number of multiplier stages: .
1) Static Gain: The multiplier capacitor is charged with The maximum output power is limited by the energy stored
the output voltage of the classical boost converter (18) at the in the multiplier capacitor. If the load power is increased above
fourth operation stage. As this capacitor is connected in series of the value, the output voltage will be reduced, limiting
with the converter output at the transference of the energy stored the output power at the value of . Therefore, the proposed
in the input inductance (first and second stages), the output ca- converter will operate with constant output power in an overload
pacitor is charged with the boost output voltage multiplied by condition until the output voltage to reach the value of the output
two voltage of the classical boost, calculated by (18). Thus, for a too
small multiplier capacitance, the proposed structure will operate
(18) as a classical boost converter and the voltage multiplier will op-
erate only as a non-dissipative snubber. The intrinsic power lim-
(19) itation of the circuit can increase the converter reliability in the
overload operation, but a current protection circuit is necessary
for an effective short-circuit protection.
Therefore, for a circuit composed by M series stages, as pre-
Considering a low value of the multiplier capacitance (1 F
sented in Fig. 4, the output voltage will be multiplied by the
and 3.3 F) as in the implemented prototypes, a polypropylene
factor . Thus, the static gain of the proposed converter,
capacitor can be used and the equivalent series resistance (ESR)
operating in continuous conduction mode is presented in
can be not considered ( at 100 kHz). For higher
capacitance values the electrolytic capacitor can be used and the
ESR losses must be considered in the capacitor definition.
6) Resonant Inductor : The resonant inductor can be
defined by the maximum current variation (di/dt) at the turn-on
(20) commutation, in order to minimize the commutation losses. In
the third operation stage presented in Figs. 8 and in 11 ,
2) Switch Duty-Cycle: The nominal duty-cycle is defined by occurs the reduction of the resonant inductor current at the
switch turn-on. The current variation is limited by the presence
of the resonant inductor, defined by
(21)
(26)
3) Switch Voltage: The maximum voltage in all diodes and
power switch is equal to the voltage, that is equal to the Considering the maximum di/dt at the turn-on commutation
output voltage of the classical boost, calculated by (22). The equal to 25 A/ s, the resonant inductance is defined by
voltage in all components is half of the output voltage
(27)
(30)
(31)
(32)
10) Theoretical Efficiency: The expected converter effi- Fig. 15. Power switch voltage and current (10 V/5 s/div).
ciency can be determined by (33) based on the losses calculated.
The total losses of the filter inductor implemented in the
prototype is equal to W
%
(33)
Fig. 18. Input inductance current (5 V/5 s/div). Fig. 21. Multiplier diode voltage (50 V/10 s/div).
TABLE I
SIMULATED CONVERTERS
Fig. 29. Main theoretical waveforms of the multiphase converter (D > 0:5).
Fig. 34. Fifth stage D < 0:5(t ;t ). Fig. 37. Turn-on snubber.
(34)
where
Therefore, the increment of the number of diodes does not in-
M number of multiplier stages; crease the total diode conduction losses because the average cur-
D switch duty-cycle. rent of each diode is proportionally reduced
(35)
(40)
2) Switch Voltage: The maximum voltage applied across the
power switches ( and ) and the output diodes ( and
) are equal to the multiplier capacitor voltage ). The 5) Passive Components: The design of the input inductance
maximum voltage in these components is is the same of the classical boost converter
(36)
(41)
The maximum voltage applied across the multiplier diodes
is two times the multiplier capacitor voltage, even if —Input current ripple
structure presents more than one multiplier stage The multiplier capacitor can be calculated by (42), where
is the capacitor voltage ripple
(37)
Fig. 41. Output voltage and the power switch voltage (50 V/5 s/div).
Fig. 47. Uninterrupted power supply implemented with the multiphase boost
voltage multiplier.
Fig. 44. Input current and the switch voltage (50 V/10 s/div).