Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

A Comparison of Isolated DC-DC Converters for

Microinverter Applications

Luciano Andres Garcia Rodriguez Juan Carlos Balda


Department of Electrical Engineering Department of Electrical Engineering
University of Arkansas University of Arkansas
Fayetteville AR, USA Fayetteville AR, USA
lgarciar@uark.edu jbalda@uark.edu

Abstract – Developing a cost-effective and high-efficiency isolation between the PV panel and the grid established by
microinverter to interface a single photovoltaic (PV) the article NEC 690. Different flyback topologies are
module with residential ac voltages is not an easy task analyzed in [2]–[12], in particular, the interleaved topologies,
taking into account the cost target set by the Department which mean there are normally two flyback transformers
of Energy (DOE) of 18 cents per Watt and the instead of one. An important benefit of using two
requirements imposed by the National Electric Code transformers is the reduction of the input and output ripple
(NEC) which lead to galvanic isolation. This paper currents, which reduce the requirements on the input and
presents a detailed analysis of three isolated topologies output capacitances, and also increase the lifetime of the
which are the top candidates for the dc-dc stage of a two- capacitors. Another benefit is the possibility of providing
stage microinverter. The considered topologies are the more power to the grid. Push-pull topologies are presented in
flyback, interleaved flyback and push-pull converters. A [13], [14], and [15].
complete set of design equations for each topology is The utilization of transformers has some adverse effects.
presented and used for designing the dc-dc stage of a For example, the leakage inductances of the transformer
microinverter system with the goal of comparing the generate losses and produce voltage spikes during the
ratings of the active and passive components since these commutation process which can damage the semiconductor
affect system cost. Validation of the design equations are devices. Due to these effects, passive snubbers are applied
done through PSpice simulations. through the terminals of the transformer to provide a path for
the currents when switches are turned off. Snubbers provide
I. INTRODUCTION a solution for the spikes but they dissipate the energy
The DOE performance targets for PV power electronic associated with the transformer leakage inductances. Another
interfaces determine that the inverter efficiency must be more solution is the implementation of an energy recovery circuit
than 95% and the reliability of the system must be at least 30 that improves system efficiency by recycling the energy of
years [1]. There is also a DOE cost target of 18 ¢/W [1]. the leakage inductances [5], [6], [7], and [16]. These circuits
Therefore, reducing the part number and eliminating add other switches and gate drivers increasing the system
components with high failure rates are extremely important. cost. Moreover, core and copper losses are the highest loss
The weakest component of a microinverter is the electrolytic component in a microinverter system [16]. Hence, the design
capacitor because of its reduced lifetime dependent on of the transformers is an important task.
operating temperatures and currents. Unfortunately, the use Different operation modes are defined depending on the
of a capacitor cannot be avoided, but another technology with shape of the current flowing though the transformer.
a better lifetime could be used if the required capacitances are Conditions for working at continuous conduction mode
reduced. Hence, design equations for sizing the input and (CCM), discontinuous conduction mode (DCM), and
output capacitors are used to analyze the capacitance required boundary conduction mode (BCM) are derived. The benefits
for the same specifications and operation conditions for the of being in one mode or in other are analyzed in this paper.
converter topologies addressed below. A microinverter could be realized using a single power
The isolated flyback and push-pull topologies are the two stage or a dual power stage as shown in Fig. 1 [18]. The
main ones utilized in microinverters. These topologies are single power stage (Fig. 1(a)) uses a dc-dc converter which
isolated topologies in order to meet the safety requirement of produces a boosted output voltage similar to a rectifier 60-Hz

978-1-4673-4355-8/13/$31.00 ©2013 IEEE 2084


T1 D2
+ ● +
T1 D1
+ + LP LS CO

CO ●
LP LS -
T2
● - ●
CIN CIN D1
LP LS

Q1

(a) - Q1 Q2

-
(a) (b)
Q1 T1 D2 LO
+
● ●
LP LS CO
+ -
● ●
(b) CIN LP LS
Fig. 1 (a) Single power stage, and (b) dual power stage D1
-
Q2
microinverter topologies.
(c)
Fig. 2 (a) Flyback converter, (b) Interleaved Flyback converter,
sinusoidal waveform. Then, this voltage is converted into an and (c) Push-pull converter.
ac waveform using an unfolding bridge [18]. Advantages of
using a single power stage are the reduction of the component directly when a switch is on; therefore, the magnetizing
number and power switching losses at the inverter stage inductance is designed to be as large as possible.
because the unfolding switches commutate at line frequency. The input of the microinverter system is a single
The dual power stage, illustrated in Fig. 1(b), has a dc-dc photovoltaic (PV) panel; for example, Kyocera KD205GX-
stage which steps up the voltage from the PV panel to an LP which produces an output power of 205W and a voltage
adequate dc value. That voltage is the input of the inverter of 26.6V at its maximum power point.
stage. One main benefit of the two stage microinverter is the The output voltage of the dc-dc converter is designed to
capability to provide reactive power which can be used in be the input of the inverter stage which is connected to the
distribution systems for power factor correction. grid. Therefore, the input and output currents are defined by
The organization of the paper is as follows: section II the input and output voltages and the power provided by the
presents the design specifications of the dc-dc converters, PV panel. The choice of the switching frequency fsw is a
section III provides the active and passive current and voltage compromise between losses in the transformer, switches and
ratings, section IV addresses design equations of each diodes, and the sizes of the transformers and capacitors.
analyzed converter, section V presents equations for the Table I summarizes the converter design specifications.
power losses in the semiconductor devices, and section VI
provides a comparison analysis of the three topologies. III. ACTIVE AND PASSIVE COMPONENT RATINGS
For the three topologies, the switch current from an initial
II. DESIGN SPECIFICATIONS AND SELECTED
value and rises to at the on time . The value of is
CONVERTER TOPOLOGIES zero under DCM or BCM operation. The rising slope of the
Fig. 2(a) presents a flyback converter (FC) whose main flyback and interleaved flyback input current is a function of
components are switch , (output) diode and a flyback the magnetizing inductance of the transformer (primary
transformer. Energy is stored in the transformer when the inductance ), the input voltage and on time [17]:
switch is turned on, the output diode is reversed biased and
the output capacitor supplies the load. The energy stored in TABLE I. DC-DC CONVERTER DESIGN SPECIFICATIONS
the transformer is transferred to the load when is turned
off, and conducts. When the energy stored in the Parameter Value
transformer is completely transferred to the load before a Output power, 205 W
switching period is completed, the converter is working under Rated output voltage, 200 V
DCM; otherwise, it is working under CCM.
Output voltage ripple, 5%
Fig. 2(b) shows an interleaved flyback converter (IFC)
which is formed by two flyback converters in parallel and Output current, 1.025 A
working in antiphase, so twice the number of components are Input voltage ripple, 5%
required. Fig. 2(c) presents a push-pull converter (PP) that is
realized by a center tapped transformer, two switches and two Input voltage, 26.6 V
diodes. A difference with the interleaved flyback converter is Input current, 7.7 A
that the transformer used in the push-pull converter is not
Critical switching frequency, 82.5 kHz
designed to store energy, since the energy is transferred

2085
(1) B. Interleaved Flyback converter

The magnetizing inductance of the transformer establishes


the boundary condition among the different modes of (4)
operation. In the case of the push-pull converter, the rising ! 1 2
,
slope of the input current is determined by the amount of 2 2
output inductance which also establishes the condition for
CCM, BCM and DCM modes of operation [17]: 1
,
1
2

(2) 1
2
Table II presents the rated currents and voltages of the
diodes and switches of the converters expressed as a function
of the converter design specifications, and Tables III to V
C. Push-Pull converter
show the duty cycle of each topology as well as the average
and root mean square RMS currents for the components.
(5)
IV. CONVERTER DESIGN EQUATIONS 2 1
,
8 2
This section presents the design equations for the
magnetizing inductance of the transformers and for the input
and output capacitors of the converters which are shown in
16
(3), (4) and (5) below. These equations allow the designer to
choose between continuous, discontinuous and boundary
conduction modes of operation. The inductance LPcritic and
Locritic are the minimum inductances to establish CCM. 2

A. Flyback converter

V. CONVERTER SWITCHING LOSSES


(3)
This section presents equations for the conduction and
2 1 switching losses [9] in the diodes and switches for all modes
,
8 2 of operation of each topology. The conduction losses are
function of the on-resistance and turn-on voltage of the
diodes and switches. In the case of the diodes, they are noted
1 1
8
,
2
as and and for the switches, and . The
switching losses are function of the turn-on and off times as
well as the switching frequency.
1 Having a good model of the converter losses is important
2 2
in order to select the adequate switching frequency.
A. Flyback converter
1) Continuous conduction mode
a) Switch losses:
TABLE II. VOLTAGE AND CURRENT RATINGS (6)
Interleaved
Ratings Flyback Push-Pull 2
Flyback
Switch
2
voltage,
Switch current,
2
Diode
2
voltage,
I1 2
Diode current,
2 3

2086
TABLE III. FLYBACK AVERAGE AND RMS ELEMENT CURRENTS.
Flyback Converter
DCM Mode CCM Mode
2 /
Duty cycle, D

Input average I I I
D D
current, 2 2

Output average 1 NP 1 NP
current, 2 NS 2 NS

Diode RMS 1 NP NP 1
current, 3 NS NS 3

Switch average
current, 2 2

Switch RMS
current, 3 3

Input capacitor 1
RMS current, 3 4 3 4

Output capacitor NP NP
1
RMS current, NS 4 NS

TABLE IV. INTERLEAVED FLYBACK AVERAGE AND RMS ELEMENT CURRENTS.


Interleaved Flyback Converter
DCM Mode CCM Mode
/
Duty cycle, D

Input average I D I I D
current,
Output average 1
NP
1
NP
current, NS NS

Diode RMS NP 1 NP 1
current, NS 3 NS 3

Switch average
current, 2 2

Switch RMS
current, 3 3

Input capacitor 2 2
RMS current, 3 3

Output capacitor NP 1 2
1
RMS current, NS 3 3

2087
TABLE V. PUSH-PULL AVERAGE AND RMS ELEMENT CURRENTS.
Push-Pull Converter
DCM Mode CCM Mode
1
Duty cycle, D 2

Input average
I D I I D
current,
Output average NP NP I I
I D D
NS NS 2
current,
Diode RMS 1 NP NP 1
I I I I I
current, 6 NS NS 6

Switch average
current, 2 2
Switch RMS
current, 3 3

Input capacitor 2 2
RMS current, 3 3

Output capacitor NP 1 NP
RMS current, NS 3 2 NS √12

b) Diode losses:
(7)
2
1 1 I1 2
3
2 3

b) Diode losses:
2) Discontinuos conduction mode
a) Switch losses: 1 (11)

0 (8) 2 1
3
2) Discontinuos conduction mode
2 a) Switch losses:
2 (12)
0
2 3
b) Diode losses:
(9) 2
2 3 I2
3
b) Diode losses:
B. Interleaved Flyback converter (13)
2
1) Continuos conduction mode 3
a) Switch losses:
C. Push-Pull converter
(10)
1) Continuos conduction mode
a) Switch losses:

2088
2VIN (14) TABLE VI. COMPONENT CHARACTERISTICS.
Component Characteristic
IFC transformers Lp=10µH,Ns/Np= 8
2VIN
FC transformer Lp=4.91µH, Ns/Np= 8

2 Push-pull output inductor Lo=35.62µH


I1 2
3 Diode VF=1.1V, RON =25 mΩ
VGS(TH) =0.9V, RDS(ON)=32mΩ,
Transistor
b) Diode losses: (15) TON=11nsec, TOFF =21nsec

1
2 3
(per (3)). For the case of the push-pull converter, an inductor
2) Discontinuos conduction mode of 35.62 µH has to be connected at the output to obtain the
a) Switch losses: same critical switching frequency (per (5)).
(16) Figure 3(a) shows the peak input current I2 as a function
0
of the switching frequency. From Table II, this peak current
2VIN is also the maximum value of the current though the switches.
2 DCM operation produces for every converter current peaks
I2 greater that in the case of CCM. The flyback converter is the
3
topology with the highest peak input current while the IFC is
b) Diode losses:
(17) the one with the smallest one.
1 2
Figures 3(b) and (c) present the RMS input and output
3 6 capacitor currents for the three converters. The smallest input
RMS capacitor current is obtained in the case of the IFC and
the smallest output RMS capacitor current in the case of the
push-pull converter. The ripple current is considerable due to
VI. CONVERTER COMPARISON STUDY
the absence of an input inductor, even for the IFC and push-
The three considered topologies are designed to operate in pull converters; so it is necessary to place capacitors
the three possible modes of operation under the requirements connected in parallel with the PV panel.
of Table I, which are typical specifications for a The RMS current through the semiconductor devices is
microinverter design. The conditions of the different modes shown in Figs. 3(d) and (e). A reduction in the current
of operation are determined using (1) and (2) and the design through the semiconductor devices is produced under CCM
equations and tables, applying the specifications of Table I. mode. The amount of current is considerably reduced in the
As mentioned above, the primary transformer inductance in push-pull converter and the IFC which have the minimum
flyback converters determines the critical condition for the amounts of current through the devices.
conduction modes while the output inductance does it in the
Figure 3(f) shows the total losses in the semiconductor
push-pull converter. The magnetizing inductance in the push-
devices as percentage of the total power. The characteristics
pull transformer is designed as large as possible to carry a
of the semiconductor devices used can be found in Table VI.
small part of the input current. The goal of the comparison is
The flyback converter has the highest losses even having
to determine the topology having the best performance under
fewer components than the others. Basically, the
the given conditions.
semiconductor devices in the FC must conduct larger currents
In order to make a realistic comparison, the converters are and switch at higher currents that in the case of the IFC and
designed to have the same critical frequency. Then, the most push-pull converter. For each converter the total
important equations are plotted as a function of the switching semiconductor losses are smaller when operating under BCM
frequency making possible the analysis of the characteristics because the turn-on switching losses are zero.
of each converter operating at the different operation modes.
From the design specifications in Table II, the peak
The IFC is designed using the transformers of Table VI voltages of the switches are similar in every topology
which have a magnetizing inductance of 10 µH and a turn working at all conditions (about ~53 ) while the peak
ratio of 1/8. Using the specifications of Table I, the design diode voltages are about 10% higher in the push-pull than in
equations (4), and considering the primary flyback the flyback converters (about 444 for the push-
transformer inductance as the critical inductance, the critical pull and 399 for the others).
frequency is 82.5 kHz. Therefore, the converter is working
under CCM for frequencies greater than 82.5 kHz and under On one hand, the required input capacitance is smaller in
DCM for frequencies smaller than that. In order to design a the IFC than in the other topologies. On the other hand, the
flyback converter with the same critical frequency, it is required output capacitance is lesser in the push-pull than in
necessary to reduce the magnetizing inductance to 4.91 µH the other ones. The capacitance needed is reduced when the
converters are working under CCM of operation.

2089
25

Input Capacitor RMS Current (A)


90 FC CCM FC CCM
FB DCM FC DCM
20
Input Peak Current (A)

IFC CCM IFC CCM


70
IFC DCM IFC DCM
PP CCM 15 PP CCM
50 PP DCM PP DCM
10
30
5
10
0
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Frequency (kHz) Frequency (kHz)
(a) (b)

3 100
IFC CCM FC CCM
Output Capacitor RMS Current (A)

90
2.5 IFC DCM FC DCM
PP CCM 80 IFC CCM

Switch RMS Current (A)


PP DCM IFC DCM
2 FC CCM 70 PP CCM
FC DCM PP DCM
60
1.5
50
1 40

30
0.5
20
0 10
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Frequency (kHz) Frequency (kHz)
( c) (d)

15
3
14
Total Semiconductor Losses (%)

FC CCM
FC DCM 13
Diode RMS Current (A)

2.5
IFC CCM
IFC DCM 12
2 PP CCM
11
PP DCM
10 FC CCM
1.5 FC DCM
9
IFC CCM
8 IFC DCM
1 PP CCM
7 PP DCM
0.5 6
0 25 50 75 100 125 150 0 100 200 300 400 500 600 700 800 900 1000
Frequency (kHz) Frequency (kHz)
(e) (f)

Fig. 3: (a) Input peak current (I2) vs. fsw, (b) input capacitor RMS current vs. fsw, (c) output capacitor RMS current vs. fsw, (d) switch RMS
current vs. fsw, (e) diode RMS current vs. fsw, and (f) total semiconductor losses vs. fsw.

VII. CONCLUSIONS converter has the advantage over the interleaved flyback
This paper presented a complete set of design equations converter in the required amount of output capacitance
which allows the designer to determine the sizes of the because of the influence of the output inductance which
components required for a specific microinverter application. smooths the output current resulting in a smaller charge to be
The push-pull converter and the interleaved flyback converter stored in the capacitor. The interleaved flyback converter
seem to be the most adequate topologies. The push-pull requires the smaller input capacitance because, differently

2090
from the push-pull converter, it can operate with a duty cycle Micro-inverter,” IEEE Applied Power Electronics
of 50% reducing the input ripple current through the Conference, pp. 555-562, 2012.
capacitor. The RMS ripple current through the input [7] R. Dong-Kyun, K. Young-Ho, K. Jun-Gu, W. Chung-
capacitors is critical because the current on the low-voltage Yuen, and J. Young-Chae, “Interleaved Active Clamp
side (PV panel side) is higher than that on the output side. Flyback Inverter using a Synchronous Rectifier for a
For this reason, commercial products have a larger parallel Photovoltaic AC Module System,” 8th International
set of capacitors in the input side to reduce the input current Conference on Power Electronics-ECCE Asia, pp. 2631-
handled by each capacitor, and thus, to satisfy lifetime 2636, May 30-June 3, 2011.
[8] Y. Hsieh, M. Chen, and H. Cheng, “An Interleaved
warranties.
Flyback Converter Featured with Zero Voltage
Increasing the switching frequency of the converter Transition,” IEEE Transactions on Power Electronics,
reduces the current thought the components, the ripple vol. 26, no. 1, pp. 79-84, January, 2011.
thought the capacitors and the amount of capacitance needed. [9] A. C. Nanakos, E. C. Tatakis, and N. P. Papanikolau, “A
However, the switching losses are proportional to the Weighted-Efficiency-Oriented Design Methodology of
switching frequency. If the converter operates under BCM, Flyback Inverter for AC Photovoltaic Modules,” IEEE
the turn-on switching losses are zero. Therefore, operating Transactions on Power Electronics, vol. 27, no. 7, pp.
the interleaved flyback converter under BCM seems to be the 3221-3233, July 2012.
best choice because it enables switching at very high [10] C. Liao, and K. M. Smedley, “Design of High Efficiency
frequencies without degrading significantly the system Flyback Converter with Energy Regenerative Snubber,”
efficiency while making it possible to require an input IEEE The Applied Power Electronics Conference and
capacitance sufficient small that permits using a more reliable Exposition, pp. 796-800, 2008.
capacitor technology. [11] T. Shimizu, K. Wada, and N. Nakamura, “A Flyback-
type Single Phase Utility Interactive Inverter with Low-
Frequency Ripple Current Reduction on the DC Input for
ACKNOWLEDGEMENTS an AC Photovoltaic Module System, ” IEEE Power
The authors are grateful through the support provided for Electronics Specialists Conference, vol. 3, pp. 1483-
this research work by the National Science Foundation 1488, 2002.
Experimental Program to Stimulate Competitive Research [12] A. Kyristsis, E. Tatakis, and N. Papanikolaou, “Optimum
(NSF EPSCoR) and the Arkansas Science & Technology Design of the Current-Source Flyback Inverter for
Authority (ASTA) through the Vertically-Integrated Center decentralized Grid Connected Photovoltaic Systems,”
IEEE Transactions on Energy Conversion, vol.23, no. 1,
for Transformative Energy Research - VICTER
pp. 281-293, March 2008.
(www.victercenter.com).
[13] F. Shinjo, K. Wada, and T. Shimizu, “A Single-Phase
Grid-Connected Inverter with a Power decoupling
REFERENCES Function,” IEEE Power Electronics Specialists
[1] U.S Department of Energy, “$1/W Photovoltaic Systems: Conference, pp. 1245-1249, 2007.
white paper to explore a grand challenge for electricity [14] T. Shimizu, and S Suzuki, “Control of a High-Efficiency
from solar”, August 2010. PV Inverter with Power Decoupling Function,” IEEE
[2] T. Shimizu, K. Wada, and N. Nakamura, “Flyback-Type Energy Conversion Conference and Exposition (ECCE
Single-Phase Utility interactive Inverter With Power 2011), pp. 1533–1539, June 3, 2011.
Pulsation Decoupling on the DC Input for an AC [15] J. Wang, J. Li, W. Zhang, “Interleaved Push-pull
Photovoltaic Module System,” IEEE Transactions on Converter with Very Low Input and high Output,”2nd
International Conference on Power Electronics and
Power Electronics, vol. 21, no. 5, pp. 1264–1272,
Intelligent Transportation System, 2009.
September 2006.
[16] Z. Zhang, M. Gao, Q. Mo, M. Chen, Z. Qian, “Loss-
[3] N. Kasa, T. Iida, and L. Chen, “Flyback Inverter
Model Based Interleave Technique to Improve the
Controlled by Sensorless Current MPPT for Photovoltaic
Efficiency of Micro-inverter”, IEEE Industrial
Power System,” IEEE Transactions on Industrial
Electronics Conference (IECON 2011), pp. 3888 - 3893,
Electronics, vol. 52, no. 4, pp. 1145-1152, August 2005.
November 2011.
[4] J. Zhao, H. Zhao, F. Dai, “A novel ZVS PWM
[17] Simon Ang, Alejandro Oliva, “Power-Switching
Interleaved Flyback Converter,” IEEE Conference on
Converters”, Third Edition, CRC Press, ISBN 978-1-
Industrial Electronics and Applications, pp. 337-341,
4398-1533-5, 2011.
May 2007.
[18] Perry Tsao, “SolarMagicTM ICs in Microinverter
[5] G.Jun-yin, W. Hong-fei, C. Guo-cheng, and X. Yan,
Applications”, Application Note 2116, National
“Research on Photovoltaic Grid-Connected Inverter
Semiconductor, May 4, 2011.
Based On Soft-Switching Interleaved Flyback
Converter,” IEEE Conference on Industrial Electronics
and Applications, pp. 1209-1214, 2010.
[6] Q. Mo, M. Chen, Z. Zhang, Y. Zhang, Z. Qian, “Digitally
Controlled Active Clamp Interleaved Flyback Converters
for Improving Efficiency in Photovoltaic Grid-connected

2091

You might also like