See

discussions, stats, and author profiles for this publication at:

https://www.researchgate.net/publication/318671821

STABILITY ANALYSIS OF SEPIC CONVERTER

USING MATLAB/SIMULINK

Article · June 2014

CITATIONS READS

0 243

2 authors, including:

Sidharth Sabyasachi
Visvesvaraya National Institute of Technology
15 PUBLICATIONS 31 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Inverter View project

Fault tolerance of multilevel inverters View project

All content following this page was uploaded by Sidharth Sabyasachi on 25 July 2017.

The user has requested enhancement of the downloaded file.

International Journal of Electrical, Electronics and Data Communication, ISSN: 2320-2084 Volume-2, Issue-6, June-2014

STABILITY ANALYSIS OF SEPIC CONVERTER USING

MATLAB/SIMULINK
1
SIDHARTH SABYASACHI, 2PRATYUSHA MOHANTY
1,2
Electrical & Electronics Engineering Department Centurion University of Technology and Management (CUTM) Odisha,
India
Email: sidharth.mana@gmail.com, pratyushamohanty0@gmail.com

Abstract: DC-DC converter is a converter where we are regulating the input voltage to get the desired output voltage which
is either greater than or equal to or less than the input voltage with positive or negative polarity. A lot of dc-dc converters are
available in the world out of which SEPIC (Single-Ended Primary-Inductor Converter) converter is one of them. To analyse
the converter, a lot of techniques are available. This paper presents the use of MATLAB/SIMULINK to study the stability of
the converter. The state space technique is applied to this higher order converter and converted into transfer function. The
reduced order technique is applied to the system and compared with the higher order system which gives approximate results
for all type of analysis.

Keywords: DC-DC Converter, SEPIC Converter, Transfer Function, State Space Analysis, Stability Theory, Reduced Order
Technique, MATLAB/SIMULINK.

I. INTRODUCTION converted in the form of transfer function. The

A lot of dc-dc converter has been designed for
different application purpose such as buck converter,
boost converter, buck-boost converter, flyback
converter, push-pull converter, forward converter etc.
The converters are designed based on the switching
operation which helps to transfer the energy from
load to source. The SEPIC converter is one of the dc-
dc converter which gives higher or lower output
voltage as compare to the input voltage with the help
of switching operation and energy storage elements
like inductor and capacitor. The operating mode of
the converter depends upon the flow of the load
current which can be continuous or discontinuous. So
the converter can be operated as Continuous
Conduction Mode (CCM) or Discontinuous
Conduction Mode (DCM) which depends whether the
load is heavy or light. The switching operation takes
place at a higher frequency which increases the
response of the system. While designing any
converter, we have to consider the source, switching
elements, filter elements and loads. The filtering
elements are mainly the energy storage elements
which decide the order of the system. For lower order
converters, it is easier to design the converters which
can operate in the stable mode and controller design
is also simple. For higher order converters, it is very
difficult to design the controller which can operate
the converter in the stable mode. The SEPIC
converter is a 4th order converter where two inductors
and two capacitors are used. Since the system is
nonlinear, the controller design is very difficult. So to
analyze the system behavior, the order of the system
has to be reduced using well defined methods. In this
paper, we are using small signal analysis including
the equivalent series resistance for inductor and
capacitor. The average model is designed and with
the help of state space analysis, the system is
different stability theory has been used to verify the
operating mode. Depending upon the energy state of
each state variables, the system order is reduced
which helps to analyze the system better. The higher
order as well as the reduced order system behavior
has been compared which gives approximate results.
II. STATE SPACE ANALYSIS OF SEPIC
CONVERTER
2.1. State Space Model of SEPIC Converter
The state space analysis is of the method by which
the converter can be analyzed. In this method, the
converter is converted into mathematical differential
equation. The number of equation is equal to the
number of energy storage elements present in the
converter. These equations are converted into matrix
form. The converter can operate in CCM mode as
well as DCM mode. When CCM mode is considered,
depending upon the switching operation, the
converter operation can be divided into two modes of
operation i.e. switch ON operation and switch OFF
operation. The equivalent circuit diagram for SEPIC
converter including equivalent series resistance for
inductor and capacitor is shown in fig.1. The
equivalent circuit diagrams during switch on
condition and switch OFF condition are shown in
fig.2 and fig.3 respectively.
The state space equation by taking the average of
these two circuit operation is
.
x (t )  Ax (t )  Bu (t ) (Eq.1)
y (t )  Cx  Du (t )

Stability Analysis Of SEPIC Converter Using Matlab/Simulink

59
International Journal of Electrical, Electronics and Data Communication, ISSN: 2320-2084 Volume-2, Issue-6, June-2014
rl1L1 rc1C1 D L2 100µH
rl2 1mΩ
C1 800µF
rc1 3mΩ
Vs MOS rc2C2 R
rl2L2 C2 3000µF
rc1 1mΩ

Output Voltage Waveform for Average Model

15

O u tpu t V oltage(V o) in V olt

Figure 1: SEPIC Converter
rl1L1 rc1C1
10

5
Vs rc2C2 R
rl2L2
0

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Time(sec)
Figure 4: Output Voltage with duty cycle 0.5

Figure 2: SEPIC Converter with Switch ON

rl1L1 rc1C1
D For ideal converter, when no equivalent series is
Vs
rl2L2
rc2C2
Figure 3: SEPIC Converter with Switch OFF
In this converter, there is no direct connection
between input and output. So Du(t) is absent.
A  A1d  A2 (1  d )
Where (Eq.2)
B  B1d  B2 (1  d )
A=4×4 System Matrix, B=4×1 Control Matrix,
d=Duty Cycle, C=1×4 Output Matrix. A1, B1 for
switch ON and A2, B2 for switch OFF condition. x is
the state vector and y is the output vector.
The average model is created for this converter in the
MATLAB by using the linear differential equation. In
average model, ripple is neglected and it is used for
steady state analysis. The model is shown in fig. 6.
The output voltage waveform is shown in fig. 4 with
duty cycle 0.5. It is shown that the converter is giving
10 V and the steady state is achieved approximately
at 0.03 sec. The circuit parameter is given in table 1.
2.2. Variation of Voltage gain with Duty Cycle
considered, the output voltage is related to the input
voltage by duty cycle.
Vin d
R Vo  (Eq.3)
1 d
By varying the duty cycle from 0 to 1, the output
voltage can be varied. The equation 3 is rearranged to
convert into voltage gain (Vo/Vin). The graph is
taken between voltage gain and duty cycle which is
shown in fig.7. When the duty cycle is 0, the voltage
gain is 0. By increasing the duty cycle, the voltage
gain increases, and at duty cycle of 0.5, the voltage
gain is 1. That means the duty cycle below 0.5, the
converter steps down the input voltage. At 0.5, the
output voltage is equal to the input voltage. Duty
cycle above 0.5, the voltage gain increases above 1
and since the derivation is for ideal converter, when
duty cycle is 1, the voltage gain is infinity.
The equation (3) is not valid for practical converter.
That’s why the equation is modified by including the
internal resistances for inductor and capacitor. In
ideal converter, the output voltage does not depend
upon the load resistance whereas in practical
converter, the output voltage depends on both load
resistance and equivalent series resistance.
10
8
V oltag e G a in ( V o /V in )

Table 1. Converter Parameter

6
Circuit Parameters Values
Input Voltage (Vs) 10V 4
Duty Cycle 0.5
Output Voltage 10V 2
Switching Frequency 100kHz
Load R 1Ω 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
L1 100µH duty cycle
rl1 1mΩ Figure 7: Voltage gain Vs Duty Cycle with variable load

Stability Analysis Of SEPIC Converter Using Matlab/Simulink

60
International Journal of Electrical, Electronics and Data Communication, ISSN: 2320-2084 Volume-2, Issue-6, June-2014

The output voltage for practical converter is III. STABILITY ANALYSIS

Vin
Vo  (Eq.4) 3.1. Simulation of the Converter
1d d rl1 1d rl2  2d 1 rc1 The converter is derived by converting into
  
d 1d R dR R mathematical equation. The parameters are defined in
the program. The averaging state space model in
By rearranging the equation, the voltage gain is matrix form is defined in the program. With the help
calculated and it is drawn against the duty cycle. The of programming language, the state space model is
figure 5 shows the variation of voltage gain with converted into transfer function which is a 4th order
respect to duty cycle for different load. It is shown function.
that when the load resistance increases, the maximum
voltage gain is also increases. In this figure, R is 3 62 7 13
varied from 1 to 10 Ω and it is observed that upto H  4.995s 1.66510 s 7.28410 s1.04110
duty cycle 0.8, there is no change of voltage gain, but s4393s37.933106s22.319109s1.0451013
when it is more than 0.8, the voltage gain is different
for different value of R which is increasing for Now the converter is reduced into 2nd order by using
increasing of R. model order reduction technique which is decided
from state energy graph. The 2nd order transfer
50 function is
R=10 ohm
6
40
R=9 ohm R=8 ohm H 2  3.515s  1.664 10
R=7 ohm s  362.8s  1.671106
2
R=6 ohm
R=5 ohm Here, the duty cycle is taken 0.5. The output voltage
30
V o /V in

R=4 ohm is Vo=9.994 V which is calculated for practical

R=3 ohm converter and we have observed form the voltage
20 gain graph that at 0.5, the change in voltage gain is
R=2 ohm
R=1 ohm
very small. That’s why the output voltage nearly to
10 10 V.

0 3.2. Step Response

0 0.2 0.4
Duty Cycle
Figure 5: Voltage gain Vs Duty Cycle with variable load
2.3 State Energy
The fig. 8 shows the state energy graph for the state
variables. The converter has four state variables. The
contribution of energy for each state variable is
calculated by using Hankel Singular values and it is
shown that the energy contribution by two state
variables are larger as compare two other two state
variables and the system is operating in the stable
mode. So the converter can be easily converted from
4th order to 2nd order system and other two state
variables can be neglected. The responses of both the
systems are compared and it is observed that both the
0.6 0.8 1 The step response of the both the model i.e. original
model and reduced order model are drawn with the
help of simulation. It is observed that the peak
amplitude for both the model is 1.63 and overshoot is
64%. The rise time and the settling time for both the
model are almost equal. So we can conclude that the
system can be analyzed by reducing the order of the
system which is easier than higher order system. The
figure 8 shows the step response of both the model.
The figure 9 shows the step response of the original
model. The peak amplitude, rise time, settling time
are shown in the figure itself. From figure 8, both the
step response are overlapping with each other, which
shows equal response.
Step Response

responses are approximately equal. Original

1.5
System: 2nd Order 2nd Order
Peak amplitude: 1.63
Hankel Singular Values (State Contributions) Overshoot (%): 64
2.5 At time (seconds): 0.00243
Stable modes
System: 2nd Order
System: 2nd Order Settling time (seconds): 0.0203
A m p litu d e

2
1 Rise time (seconds): 0.000902
S ta te E n e r g y

1.5

1 0.5

0.5

0
0
0 1 2 3 4 5
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035
State Time (seconds)
Figure 8: Step Response of both the model
Figure 8: State energy

Stability Analysis Of SEPIC Converter Using Matlab/Simulink

61
International Journal of Electrical, Electronics and Data Communication, ISSN: 2320-2084 Volume-2, Issue-6, June-2014
Step Response
Pole-Zero Map
4000
H 1 1 1 0.999 0.998 0.99 sys
System: H
1.5 Peak amplitude: 1.63
3000 1
Overshoot (%): 64

I m a g i n a r y A x i s ( s e c o n d s-1 )
At time (seconds): 0.00243 2000
System: H System: H
1
System: H Settling time (seconds): 0.0202 Final value: 0.996 1000
A m p lit u d e

1 Rise time (seconds): 0.000903 3.5e+05 3e+05 2.5e+05 2e+05 1.5e+05 1e+05 5e+04
0

-1000
1
0.5 -2000

-3000 1
1 1 1 0.999 0.998 0.99
-4000
0 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 Real Axis (seconds-1) x 10
5

Time (seconds) Figure 12: Pole Zero map of both the Model
Figure 9: Step Response of the original model
The figure 12 shows the pole zero map of both the
3.3. Bode Plot of the Converter
model. It is observed that the locations of zeros are in
The bode plot of the converter is compared for both
the left half of the s-plane in both the model. So the
the model which is shown in figure 10. There is a
system is stable.
slight change in the 2nd order which can be neglected
3.5. Root Locus Plot of the System
for stability analysis. The bode plot for the original
model with phase margin and gain margin is shown 5
x 10 Root Locus
in figure 11. It is observed that gain margin is infinity 5 0.96 0.92 0.86 0.76 0.58 0.35

and phase margin is 23.10. So the system is inherently

stable. In reduced order model, gain margin is infinity 0.984
I m a g i n a r y A x is (s e c o n d s- 1 )

and phase margin is 23.20. So this system is also 0.996

System: H2 System: H
Gain: 1.09e+08 Gain: In f

showing the stability of the system. Pole : -4.74e+05

Damping: 1
Overshoot (%): 0
Pole : -3.33e+05
Dampin g: 1
Overshoot (%): 0
Frequency (rad/s): 4.74e+05 Frequency (rad/s): 3.33e+05
Bode Diagram 1.8e+06
0 1.6e+06 1.4e+06 1.2e+06 1e+06 8e+05 6e+05 4e+05 2e+05

200
M a g n itu d e (d B )

0.996

0 0.984

0.96 0.92 0.86 0.76 0.58 0.35

-5
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2
-2000 Real Axis (seconds )
-1 5
x 10
Original
Figure 13: Root Locus of both the Model
P h a s e (d e g )

2nd order
-90
The root locus for both the system is shown in figure
13. For both the system, the damping is 1 and
-180 1 2 3 4 5 6 7
overshoot is 0%. So the system is critically damped.
10 10 10 10 10 10 10
Frequency (rad/s)
3.6. Nyquist Diagram
Figure 10: Bode Plot of both the Model
Bode Diagram Nyquist Diagram
Gm = Inf , Pm = 23.1 deg (at 1.79e+03 rad/s) 4
50 0 dB

System: H2
Real: 0.201
M a g n itu d e (d B )

0 Imag: 3.56
Frequency (rad/s): -1.28e+03
-2 dB
2 dB
2
-50
4 dB -4 dB System: H2
Im a g in a r y A x is

Phase Margin (deg): 179

System: H26 dB -6 dB
-100 Phase Margin (deg): 23.2
10 dB
Delay Margin (sec): 0.000226
-10 dB
Dela y Margin (sec): 0.0358
At frequency (rad/s): 87.3
Closed lo op stable? Yes
At frequency (rad/s): 1.79e+03
0 Closed lo op stable? Yes

-1500
-45
P h a s e (d e g )

-2
-90 System: H2
Real: 0.278 System: HH2
Imag: -3.56 Peak gai
gainn (dB): 11.1
-135 Frequency (rad/s): 1.28e+03 Frequency (rad/s): 1.27e+03

-180 2 -4
10 10
3
10
4
10
5
10
6
10
7 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
Frequency (rad/s) Real Axis
Figure 11: Bode Plot of the original Model Figure 14: Nyquist Diagram for both the Model
3.4. Pole Zero Map of the System
Stability Analysis Of SEPIC Converter Using Matlab/Simulink

62
 International Journal of Electrical, Electronics and Data Communication, ISSN: 2320-2084 Volume-2, Issue-6, June-2014

The Nyquist diagram for both the model is shown in The stability can also be checked by using Nichols
the figure 14 which shows the closed loop system is chart which is shown in figure 15. It is observed from
stable for both the model. the figure that the closed loop system is stable. And
3.7. Nichols Chart both the diagram is approximately same.
Nichols Chart
40
0.25 dB System: H2
0 dB CONCLUSIONS
System: H2
0.5 dB Peak gain (dB): 11.1 Phase Margin (deg): 179
20 1 dB Frequency (rad/s): 1.27e+03 Dela y Margin (sec):-10.0358
dB
3 dB At frequency (rad/s): 87.3

0
6 dB Closed loop stable? Yes -3 dB
-6 dB
It is observed that the system can be analyzed with
System: H
-12 dB
Phase Margin (deg): 179
the reduced order model. The stability of the system
O p e n -L o o p G a in (d B )

Delay Margin (sec): 0.0375

-20 -20 dB
At frequency (rad/s): 83.3
Closed loop stable ? Yes
can be checked by using different method with the
-40 -40 dB
help of MATLAB. All the plots are draw by writing
-60 -60 dB
the code in the m.file. So it will be easier for
-80 -80 dB designing the controller of the system. In this paper,
-100 -100 dB the converter is reduced to the 2nd order model. The
-120 -120 dB
energy of state variables are observed and only 2 out
-140 dB
of 4 variables are contributing more energy. So we
-140
-360 -315 -270 -225 -180 -135 -90 -45 0 are neglecting the energy of other two variables. We
Open-Loop Phase (deg)
are comparing the both the models by different plots
Figure 15: Nichols Chart for both the Model
and it is concluded that both the system giving same
responses.

Figure 6: Average Model of the SEPIC Converter

REFERENCES [3] Vimal Singh, Dinesh Chandra, and Haranath Kar,

“Improved Routh–Padé Approximants: A Computer-Aided
Approach”, IEEE Transactions On Automatic Control, Vol.
[1] Binod Kumar Padhi, Anirudha Narain, “CONTROLLER
49, No. 2, February 2004.
DESIGN FOR SEPIC CONVERTER USING MODEL
ORDER REDUCTION”, ASAR International Conference,
[4] Vorperian, “Analysis of the Sepic converter by Dr. Vatché
Bangalore Chapter-2013, ISBN: 978-81-927147-0-7.
Vorpérian,”Ridley Engineering Inc, www. Switching
powermagazine. com, 2006.
[2] Robert W. Erickson, Dragan Maksimovic, “Fundamental of
Power Eletronics”, Springer International Edition, 2nd
Edition.



Stability Analysis Of SEPIC Converter Using Matlab/Simulink

63

View publication stats