Introduction

Over the past decades, research has been directed towards EV to improve the environmental
conditions [1]. EV systems depend on electric power as its energy source rather than the fuel like
in conventional vehicle systems. The main components of the EV system are battery for energy
storage, motor for traction, AC/DC converter to charge the battery and supply energy to the motor
from the battery, and DC/DC converter to provide energy to the low-power components in the EV
system. The current use of DC/DC converters with transformers or coupled inductors causes the
following problems in the EV system; high voltage stress across the the power switch, increasing
the converter area, and increasing the EMI [2]. To design a DC/DC converter for an EV system,
there are some constraints that should be considered: high efficiency, small volume, and low EMI [3].
Several typologies of DC/DC converters are used in EV applications such as Buck converter, Boost
converter, Buck-boost converter, and SEPIC converter.
SEPIC is considered a buck-boost DC/DC converter. It differs from the regular buck-boost
converter such that the output voltage polarity in the SEPIC converter is non-inverting [4].
SEPIC converter is designed using two separate inductors [5]. However, using two coupled inductors
rather than two separated inductors reduces the required inductance for the same inductor ripple
current in half [6,7]. SEPIC converter with two coupled inductors is shown in Figure 1, where La and
Lb are the coupled inductors. Like most of the other DC/DC converters which contain a coupled

inductors or transformer in their circuit, SEPIC converter suffers from the energy stored in the leakage
inductance of the primary inductor. This energy causes a voltage stress on the switch transistor during
the turn OFF period and may damage the Metal–Oxide Field–Effect Transistor (MOSFET) switch [8].
To prevent that damage, a snubber circuit is used to create an alternative path for that energy and
protect the MOSFET switch. Besides that, the ringing created from the resonance between the leakage
inductance of the primary inductor and the parasitic capacitance of the MOSFET switch, increases the
EMI of the converter at the ringing frequency [9]. In addition, to protect the switch from the energy
stored in the leakage inductance, snubber circuit reduces the EMI as presented in [10,11].

Snubber circuits are classified as dissipative and non-dissipative. The most commonly used
dissipative snubber is the Resistor-Capacitor (RC) snubber. It consists of a capacitor and a resistor
in series connected in parallel with the transistor [12]. In this type of snubber, during the turn OFF
period, the stored energy in the leakage inductance is transferred to the snubber capacitor through
the snubber resistor. In the turn ON period, the snubber capacitor discharges that energy through
the transistor. Another dissipative snubber is the Resistor-Capacitor-Diode (RCD) snubber, which
differs from the RC snubber by containing a diode in parallel with the resistor. The operation of the
RCD snubber is the same as the RC snubber operation, except in the RCD snubber during the turn
OFF period, the energy is transferred to the capacitor through the diode not through the resistor [13].
There is a trade off between the RC and the RCD snubbers. The RCD snubber decreases the power loss
during the turn OFF period compared to the RC snubber. However, the RC snubber achieves more
suppression for the over-voltage peak during the turn OFF period [14]. Connecting the snubber circuit
in parallel with the MOSFET switch causes more current stress on it [13]. By connecting the snubber
circuit in parallel with the primary inductor rather than the MOSFET switch, the current stress on the
MOSFET switch is avoided [15]. The performance of the snubber circuit does not change by connecting
it in parallel with the primary inductor. A SEPIC converter with RCD snubber is shown in Figure 2a.
A non-dissipative LC snubber is presented for flyback converter in [16]. The snubber action
is analyzed for four different operation modes. These modes differ from each other by when the
energy stored in the snubber capacitor is returned to the supply. In mode 1, no energy is returned to
the supply after stored it in the snubber capacitor during the turn OFF period. During the turn ON
period, the snubber capacitor is discharged through the snubber inductor and the transistor. In mode 2,
the energy in the snubber capacitor is returned to the supply during the turn ON period. In mode 3,
the energy is returned to the supply during the turn ON and the turn OFF periods. In mode 4,
the energy is returned to the supply in the turn OFF period. The voltage across the snubber capacitor
determines in which mode the snubber operates.
The LC snubber has an advantage over the RC snubber, because it is able to return the stored
energy in the leakage inductance through the snubber inductor to the supply rather than dissipating it
like the RC snubber. On the other hand, an extra inductor used in this type of snubber increases the
number of components. A SEPIC converter with an LC snubber is shown in Figure 2b.

The regenerative snubber for flyback converter presented in [17,18] differs from the LC snubber,
because it is based on three coupled inductors rather than the regular two coupled inductors or
two winding transformer. The third coupled inductor or the auxiliary inductor is used as a snubber
inductor instead of using a separate inductor like LC snubber. That reduces the number of required
components and transfers the energy also to the secondary side during the turn ON period.
In this paper, a coupled inductor based LC regenerative snubber for SEPIC converter is proposed
to minimize the switching stress on the transistor, as shown in Figure 3. This method also reduces
the Electromagnetic Interference (EMI) which occurs due to the switching activity. The rest of the
paper is organized as follows. The operation of a SEPIC converter with regenerative snubber is
analyzed and design constraints for the snubber component values are determined in Section 2.
In Section 3, the advantages of regenerative snubber for SEPIC converter are verified by simulation
and experimental results. Finally, a conclusion about the results of using the regenerative snubber for
SEPIC converter is presented in Section 4.

Analysis of LC Regenerative Snubber for SEPIC Converter

For the SEPIC converter with regenerative snubber shown in Figure 3, there are four states in
each operation cycle, as shown in Figure 4. Two of these states are in ton, which is the turn ON period,
Energies 2020, 13, 5767 4 of 16
and the other two states are in to f f , which is the turn OFF period. The snubber circuit is active at the
first state of the turn ON period from t0 to t1, and at the first state of the turn OFF period from t2 to t3.
At the second state of the turn ON period from t1 to t2 and at the second state of the turn OFF period
from t3 to t4, the SEPIC converter acts like a regular SEPIC converter without any effects from the
snubber circuit. In analyses, voltage drops across diodes and transistor are neglected. The leakage
inductance at the primary side and at the auxiliary winding are taken into account.

Operating State of LC Snubber Circuit

As shown in Figure 4, the first state of the turn ON period starts at t0 when the MOSFET switch is
turned on. In this state, D1 and D2 are off, D3 is on, and the voltage across Cs is VCsmax which is the
maximum voltage. During this state, Cs discharges through the MOSFET switch and Lax, where Lax is
the auxiliary inductor. In this state, as presented in [18] for the flyback converter, the voltage across
the auxiliary inductor is equal to the reflected voltage from the primary inductor only. However, in
the SEPIC converter, the voltage across Lax is equal to the reflected voltage from the primary and the
secondary inductors, this is one of the main differences between the regenerative snubber performance
in the SEPIC and in the flyback converters. This state ends at t1, when D3 becomes off and the current
through Lax becomes zero. The equivalent circuit of the SEPIC with regenerative snubber in this state
is shown in Figure 5a. The equivalent circuit for the snubber circuit is shown in Figure 5b, where Llax
is the leakage inductance at the auxiliary side and VLax is the voltage across Lax.

Lla is the leakage inductance at the primary side, Nax is the number of turns for Lax, Na is the
number of turns for the primary side La, and Nb is the number of turns for the secondary side Lb.
Suppose that the SEPIC is designed to have Na = Nb, the voltage of the leakage inductance at the
auxiliary side is obtained as

The time constant wo for the equivalent circuit in Figure 5b is

The impedance of the leakage inductance Llax is

The current through Llax equals to

The voltage across Cs equals to

when iLlax (t) = 0, VCs reaches VCsmin . As shown in (6) and Figure 4, iLlax (t) = 0 when wot = p,
by substituting that in (7)

The power consumption in D3 during this state is

where ILlax is the average current of Llax during this state. VD3 is the voltage drop across D3.
ILlax equals to

where T is the switching time. After D3 turns off at t1, the second state in the turn ON period begins.
In this state, the snubber circuit does not affect the performance of the SEPIC and the converter acts
like regular SEPIC in the turn ON period until the turn OFF period starts at t2 when the MOSFET
switch is turned off.
At the first state of the turn OFF period, the snubber is active. In this state, D1 and D2 are on, and
D3 is off. The equivalent circuit for the SEPIC converter and for the snubber circuit in this state are
shown in Figure 6a,b, respectively. VLa is the reflected voltage from the secondary side to the primary
side and it is equal to VO. The voltage across Cs at the beginning of this state is VCsmin . This state ends
at t3 when VCs reaches VCsmax , that happens when the current through the leakage inductance iLla
= 0.
The time constant wo for the equivalent circuit in Figure 6b is

The impedance of the leakage inductance Lla is

The current through the leakage inductance Lla equals to

where I is the peak of the magnetizing current. Suppose that Cs is selected to have VCsmin = VO,
the current through the leakage inductance Lla is written as

The voltage across Cs equals to

From (15), VCs reaches VCsmax at wot = p
2 and it equals to

By taking VCsmin = VO into account, substituting (16) in (8) leads to

The power consumption in D2 during this state is

where ILla
is the average current of Lla during this state.VD2 is the voltage drop across D2. ILla
equals to

After the current through the leakage inductance Lla becomes zero at t3, D2 turns off and the
second state of the turn OFF period starts. During the steady state of this period, Cs charges to the
input voltage Vin. This leads to that the voltage across MOSFET during this period equals to the
summation of the input voltage and the output voltage. In this state, the converter acts like regular
SEPIC converter in the turn OFF period without any effects from the snubber circuit until t4 when the
MOSFET switch is turned on again. All of these states are repeated in each cycle.

Components Sizing

During the first state of the turn ON period, Vcs drops to Nax
Na
2Vin, after that D3 turns off. Moreover,
during the turn OFF period to make D1 on, VCs should be greater than VO [18]. From that, and by
solving the following inequality,

Furthermore, by solving the following inequality using (16) and (17),
VCsmax

Moreover, as mentioned in Section 2 that Cs should be selected to have VCsmin = VO, by solving
the following inequality taking (20) into consideration and using (17),

where D is the duty cycle. In the first state of the turn ON period, VCs reaches VCsmin when wot = p,
which means half of the time constant. This should happen within the turn ON period, this leads to

By using (22) and (24), the upper and lower limits for Nax are determined. Moreover, by using
(28) and (30), the upper and lower limits for Cs are obtained. From the previous constraints, the value
Energies 2020, 13, 5767 9 of 16
of the leakage inductance of the primary Lla is very critical and important to determine the limits of Cs
and Nax. To calculate the leakage inductance value, the value of the parasitic capacitance of the switch
should be known. If it is not known, to calculate it, the frequency of the ringing should be measured.
It equals to

where Cpar is the parasitic capacitance. After measuring the ringing frequency, an additional capacitor
should be connected in parallel with the parasitic capacitance and then new ringing frequency should
be measured. The new frequency equals to

where Cadd is the additional capacitor. By using (31) and (32), the parasitic capacitance is calculated
and it equals to

After calculating Cpar, the leakage inductance Lla is calculated by rearranging (31)

3. Simulation and Experimental Results
   LTSpice is used to simulate the SEPIC converterwith regenerative snubber. Component parameters
   used in the simulation are shown in Table 1. The simulation setup is shown in Figure 7.

From the design constraints illustrated in the Section 2, the values of Nax and Cs are related to
each other. From (22), the lower limit for Nax is

By drawing (24) using the above constraint for the lower limit of Nax as shown in Figure 8,
the chosen pair of Cs and Nax values should be under the graph. To avoid the sensitive areas,
the optimum values of Cs and Nax are found to be in the area under the graph and bounded by
Cs = 25 nF and Nax = 6.5. The chosen pair for (Cs,Nax) is (10 nF, 5).

The simulation results of the voltage across the MOSFET switch of the SEPIC converter without
regenerative snubber and with non-coupled and coupled regenerative snubber for Vin = 24 V are
shown in Figure 9. To ensure that the regenerative snubber has the same effect on the SEPIC
converter with different values of Vin, the voltage across the MOSFET switch without snubber
and with snubber for Vin = 28 V and Vin = 22 V are depicted in Figures 10 and 11, respectively.
Without coupling of snubber inductance to SEPIC inductors, the improvement is limited, as shown in
Figures 9b, 10b, and 11b. The coupling between snubber inductance and SEPIC inductors contribute
significant reduction of ringing, as shown in Figures 9c, 10c, and 11c. Comparison between simulation
results is shown in Table 2 where Avg. P is the average power consumption in the MOSFET during
turn OFF period, D is the duty cycle, VdsOS is the voltage overshoot across the MOSFET, and VdsSS is
the steady-state voltage across the MOSFET during turn OFF period. Using the regenerative snubber,
the average power consumption in the MOSFET, the overshoot voltage during the turn OFF period
and the required time to reach the steady state are all reduced.

For the experiment, a prototype for SEPIC converter is built, as shown in Figure 12. The voltage
across the MOSFET switch in the turn OFF period without the regenerative snubber and with
the regenerative snubber experimentally for Vin = 24 V are shown in Figure 13a,b, respectively.
The regenerative snubber reduces the stress on the switch by reducing the ringing and suppressing
the voltage surge by 16 V. Moreover, the voltage across the MOSFET switch of the SEPIC during
the turn OFF period with snubber and without snubber for different input voltages are shown in
Figures 14 and 15, respectively. The voltage on the MOSFET switch without coupled inductance in the
regenerative snubber is shown in Figure 16b. Although the peak voltage is reduced compared to the
one without snubber circuit, the time for settling due to the ringing does not change significantly.

Reverse protection diode and SEPIC diode are selected as PN junction diode. Although PN
junction diode has higher forward voltage drop than a schottky diode, it has lower reverse current
leakage at high temperature. The voltage drop on these diodes is around 0.9 V. Therefore, for a 4.4 W
load these diodes contribute 11 W of power loss. The ferrite core is based on N87 material and has
4 W of power loss. The power loss due to the series resistance of inductors is 1 W. The conduction
and switching losses of the MOSFET switches are 1Wand 3W, respectively. The gate driver circuit
including its low-dropout regulator contributes 1.5 W of power loss. The rest of the circuit contributes
less than 0.5 W. Therefore, for 4.4 W load the output power is 131 W and total power loss is 22 W. This
results in 85 % efficiency. Since snubber circuit improves switching losses which is only 14 % of the
total loss, the effect of snubber circuit on the efficiency is limited.
To measure EMI for the SEPIC converter, Rohde & Schwarz FSH8 spectrum analyzer is used.
The SEPIC converter without regenerative snubber has a peak at the ringing frequency which is
20 MHz, as shown in Figure 17a. After using the regenerative snubber as shown in Figure 17b, the peak
at the ringing frequency is suppressed by 8 dB. On the other hand, another peak appears at 29.75 MHz,
but still less than the peak without regenerative snubber by 2.8 dB.

The EMI for the SEPIC without regenerative snubber and with regenerative snubber for input
voltage Vin = 28 V and 22 V are shown in Figures 18 and 19, respectively. For Vin = 28 V as shown
in Figure 18a, the converter has a peak of 48.6 dB at the ringing frequency. In Figure 18b, using the
regenerative snubber reduces that peak to 40 dB. Moreover, at high frequency, the EMI is reduced by
10 dB. The same thing happens for Vin = 22 V, as shown in Figure 19, where the converter has a peak
of 50.4 dB at the ringing frequency. Using the regenerative snubber reduces the peak at the ringing
frequency to 42.3 dB. Moreover, EMI at high frequency is reduced by 10 dB. EMI for SEPIC with
non-coupled regenerative snubber is shown in Figure 20 which does not reduce the EMI compared
to the coupled regenerative snubber. Performance measurement results with regenerative snubber

and without regenerative snubber for different input voltages Vin = 24 V, 28 V, and 22 V and with
non-coupled regenerative snubber for Vin = 24 V are summarized in Table 3.

4. Conclusions
   An LC regenerative snubber for SEPIC converter for EV applications is proposed. The operation
   of the SEPIC converter with regenerative snubber is analyzed. Determination and sizing for the
   components of the regenerative snubber are presented. The effects of using the proposed snubber
   on the SEPIC are verified by simulation and experimental results. The peak of the voltage stress on
   MOSFET switch of the SEPIC converter is reduced by 16 V. An improvement on the EMI performance
   of the SEPIC converter is achieved by 8 dB reduction at the ringing frequency. Moreover, EMI is
   reduced at high frequencies by 10 dB. Using the proposed snubber does not affect the efficiency of the
   SEPIC. The values of the snubber capacitor and inductor are sensitive, they should be selected to have
   enough margin not to be affected by the parasitic capacitance and the leakage inductance values.

References and Authors

Department of Electrical and Electronics Engineering, Istanbul Okan University, Istanbul 34959, Turkey;
abd.m.ksasbeh@gmail.com (A.K.); kellecib@hotmail.com (B.K.)
2 Department of Electrical Engineering, Istanbul Technical University, Istanbul 34469, Turkey
3 MOBI Research Group, ETEC Department, Vrije Universiteit Brussel (VUB), 1050 Brussels, Belgium;
Ahmet.Aksoz@vub.be (A.A.); Omar.Hegazy@vub.be (O.H.)
4 MOBI Core-Lab, Flanders Make, 3001 Heverlee, Belgium

• Correspondence: ozturksb@itu.edu.tr

1. Naunin, D. Electric Vehicles. In Proceedings of the IEEE International Symposium on Industrial Electronics,
   Warsaw, Poland, 17 June 1996; pp. 11–24. [CrossRef]
2. Bellur, D.M.; Kazimierczuk, M.K. DC-DC Converters for Electric Vehicle Applications. In Proceedings of the
   2007 Electrical Insulation Conference and Electrical Manufacturing Expo, Nashville, TN, USA, 22–24 October
   2007; pp. 286–293. [CrossRef]
3. Sakka, M.A.; Mierlo, J.V.; Gualous, H. Electric Vehicles—Modelling and Simulations; Soylu, S., Ed.; IntechOpen
   Limited: London, UK, 2011; Chapter 13, ISBN 978-953-307-477-1. [CrossRef]
4. Ridley, R. Analyzing the SEPIC Converter; Power Systems Design: St. Annapolis, MD, USA, 2006; pp. 14–18.
5. de Melo, P.F.; Gules, R.; Romaneli, E.F.R.; Annunziato, R.C. A Modified SEPIC Converter for
   High-Power-Factor Rectifier and Universal Input Voltage Applications. IEEE Trans. Power Electron. 2010, 25,
   310–321. [CrossRef]
6. Falin, J. Designing DC/DC converters based on SEPIC topology. Analog Appl. J. 2008, 4Q, 18–23.
7. Betten, J. Benefits of a coupled-inductor SEPIC converter. Analog Appl. J. 2011, 2Q, 14–17.
   Energies 2020, 13, 5767 16 of 16
8. Erickson, R.W.; Maksimovic, D. Fundamentals of Power Electronics, 2nd ed.; Springer: New York, NY, USA,
   2001; ISBN 978-1-4757-0559-1. [CrossRef]
9. Dutta, A.; Ang, S.S. Effects of parasitic parameters on electromagnetic interference of power electronic
   modules. In Proceedings of the 2017 IEEE Applied Power Electronics Conference and Exposition (APEC),
   Tampa, FL, USA, 26–30 March 2017; pp. 2706–2710. [CrossRef]
10. Yuvarajan, S.; Nelson, R.; Quek, D. A study of the effects of snubber on switching loss and EMI in an MCT
    converter. In Proceedings of the 1994 IEEE Industry Applications Society Annual Meeting, Denver, CO, USA,
    2–6 October 1994; pp. 1344–1349.
11. U-Yaisom, C.; Khanngern, W.; Nitta, S. The study and analysis of the conducted EMI suppression on
    power MOSFET using passive snubber circuits. In Proceedings of the 3rd International Symposium on
    Electromagnetic Compatibility, Beijing, China, 21–24 May 2002; pp. 561–564. [CrossRef]
12. Yatsugi, K.; Nomura, K.; Hattori, Y. Analytical Technique for Designing an RC Snubber Circuit for Ringing
    Suppression in a Phase-Leg Configuration. IEEE Trans. Power Electron. 2018, 33, 4736–4745. [CrossRef]
13. Liu, S.; Lin, H.; Wang, T. Comparative Study of Three Different Passive Snubber Circuits for SiC Power
    MOSFETs. In Proceedings of the 2019 IEEE Applied Power Electronics Conference and Exposition (APEC),
    Anaheim, CA, USA, 17–21 March 2019; pp. 354–358. [CrossRef]
14. Wu, Y.; Yi, Q.; Wu, Y.; Yang, F.; Zhang, Z.; Wen, J. Research on Snubber Circuits for Power Electronic Switch
    in DC Current Breaking. In Proceedings of the 2019 14th IEEE Conference on Industrial Electronics and
    Applications (ICIEA), Xi’an, China, 19–21 June 2019; pp. 2082–2086. [CrossRef]
15. Hren, A.; Korelic, J.; Milanovic, M. RC-RCD clamp circuit for ringing losses reduction in a flyback converter.
    IEEE Trans. Circuits Syst. II Express Briefs 2006, 53, 369–373. [CrossRef]
16. Domb, M.; Redl, R.; Sokal, N.O. Nondissipative turn-off snubber alleviates switching power dissipation,
    second-breakdown stress and VCE overshoot: Analysis, design procedure and experimental verification.
    In Proceedings of the 1982 IEEE Power Electronics Specialists Conference, Cambridge, MA, USA, 14–17 June
    1982; pp. 445–454. [CrossRef]
17. Ai, T.-H. A Novel Integrated Nondissipative Snubber for Flyback Converter. In Proceedings of the IEEE
    ICSS 2005 International Conference on Systems & Signals, Kaohsiung, Taiwan, 28–29 April 2005; pp. 66–71.
18. Liao, C.-S.; Smedley, K.M. Design of High Efficiency Flyback Converter with Energy Regenerative Snubber.
    In Proceedings of the 2008 Twenty-Third Annual IEEE Applied Power Electronics Conference and Exposition,
    Austin, TX, USA, 24–28 February 2008; pp. 796–800. [CrossRef]

This article is published under CC BY 4.0