To invigorate the tappedinductor boost (TIB) topology in emerging high stepup applications for offgrid products, a lossless snubber consisting of two capacitors and three diodes is proposed. Since the switch voltage stress is minimized in the proposed circuit, it is allowed to use a device with a lower cost, higher efficiency, and higher availability. Moreover, since the leakage inductance is fully utilized, no effort to minimize it is required. This allows for a highly productive and costeffective design of the tappedinductor. The proposed circuit also shows a high stepup ratio and provides relaxation of the switching loss and diode reverserecovery. In this paper, the operation is analyzed in detail, the steadystate equation is derived, and the design considerations are discussed. Some experimental results are provided to confirm the validity of the proposed circuit.
I. INTRODUCTION
In SubSaharan African and South Asian underdeveloped countries, one can easily find the offgrid areas where people do not have an access to the power grid. In those areas, the demand for standalone offgrid power systems based on renewable energy such as sunlight and wind is rapidly increasing. This reflects the recent increase in their need for electricity. Market analysts have found that there are already more than 2 million offgrid power systems. The output voltage of these kinds of power systems is generally very low when compared to that coming from the ac mains of the power grid. They are mostly loaded by 12V automotive batteries for a continuous supply of electricity under varying weather conditions
[1]

[4]
.
Market analysts have also found that the main use of this offgrid power is for watching television and lighting as shown in
Fig. 1
. The lighting devices should be made of highly efficient LEDs since this electric power is seriously limited by the weather conditions and battery capacity. For the same reason, the televisions should be LCD televisions with an LED backlight unit since they use less power than any other type of television. These devices usually consist of 30 or more LEDs. For example, a 30Wrated medium power LED tube and a backlight unit smaller than 32inchs contain around 30 LEDs since 1WLEDs are currently in wide use. It needs to be noted that all of the LEDs are preferred to be connected in series and driven by a single circuit so that all of the LEDs carry the same current for good brightness uniformity. A challenge for the circuit engineer is that this configuration requires a 10times or higher stepup LED driver, which is practically unattainable with conventional boost topologies due to the duty ratio being very close to the unity
[5]
.
One simple and effective alternative to overcome the stepup ratio limitation in boost topologies is the tappedinductor boost (TIB) topology. By simply replacing the inductor with a tappedinductor, the TIB topology gives a much higher stepup ratio at a reduced duty ratio
[5]
. Another important advantage of the TIB topology is that the voltage stress on the switch in its turnoff state is equal to the mean of the input and output voltages weighted by the winding turn numbers, whereas a voltage equal to the output voltage appears on the switch in the conventional boost topology. However, this feature is limited to cases where the effects due to parasitic reactance are negligible. In practice, the switch of the TIB topology suffers from a severe voltage spike due to the leakage inductance of the tappedinductor. Without suppressing this voltage spike, the TIB topology is much the worse from the standpoint of the switching stress. There have been many elaborated studies on limiting the switch voltage spike of the TIB topology and improving the converter performance
[6]

[18]
. However, they generally modify the original TIB topology, limit the switch voltage to a nonminimum value, or require an auxiliary inductor or winding to achieve proper circuit operation.
Typical offgrid power appliances.
In this paper, a lossless snubber with minimum voltage stress for a continuous current mode TIB converter is proposed
[19]
. It consists of two capacitors and three diodes. This provides more boost capability and a higher efficiency to the original TIB converter. The following features make the proposed circuit very attractive especially in commercial applications. First, a large improvement in the circuit reliability is achieved. Since the proposed circuit limits the switch voltage of the TIB converter to its minimum, it is possible to use switches with significantly lower voltage ratings which generally results in a lower cost, higher performance, and higher availability. Second, by utilizing the leakage inductance of the tappedinductor, problems related to
di
/
dt
such as the switching loss and the diode reverserecovery are alleviated. Moreover, efforts in designing magnetically coupled devices to minimize the flux leakage, such as special winding methods which in most cases sacrifice cost and productivity, are not necessary. Third, the proposed circuit can be readily incorporated into an existing system since it does not modify the original TIB topology.
The operation principle of the proposed circuit is analyzed in detail in this paper, and the steadystate stepup relationship is derived based on this analysis. The design considerations are also discussed with various boundary conditions related to the proper operation. Some experimental results from a 12V to 120V/240mA LED driver prototype for a lighting module are also provided to confirm the validity of the proposed circuit.
II. CIRCUIT ANALYSIS
 A. Operation analysis
The proposed circuit consists of a clamp capacitor
C_{c}
, a clamp diode
D_{c}
, a resonant capacitor
C_{r}
, and two path diodes
D_{a}
and
D_{b}
for charging and discharging
C_{r}
, as shown in
Fig. 2
. The tappedinductor is modeled as an ideal transformer with a turns ratio of
N
=
N
_{2}
/
N
_{1}
, a magnetizing inductance
L_{m}
, and a leakage inductance
L_{k}
. The switch
M
is modeled as an ideal switch in parallel with the output capacitance
C_{ds}
. The parasitic elements other than
L_{m}
,
L_{k}
and
C_{ds}
are basically ignored. However, they are separately considered when they produce a meaningful effect. The switching frequency is assumed to be
f
=1/
T
, and 1+
N
is denoted as
k
for convenience in deriving the equations.
TIB converter with proposed lossless snubber circuit.
Qualitative illustration of theoretical key waveforms.
Fig. 3
shows the key waveforms of the proposed converter (referring to the TIB converter with the proposed snubber). The magnetizing current
i_{m}
is shown to keep varying throughout the switching cycle. However, the operation is analyzed with the assumption that it is constant during the switching transients and snubber operation. In other words,
i_{m}
=
I
_{M1}
for
t
_{0}
~
t
_{2}
and
i_{m}
=
I
_{M2}
for
t
_{3}
~
t
_{6}
, where
I
_{M1}
and
I
_{M2}
denote the minimum and maximum values of
i_{m}
, respectively. Otherwise, the quantitative analysis becomes very complicated. The input voltage
V_{i}
, output voltage
V_{o}
, output current
I_{o}
, and clamp voltage
V_{c}
are assumed to be ripplefree and constant. It is also assumed that
C_{r}
is much larger than
C_{ds}
. The clamp voltage
V_{c}
which determines the switch voltage stress can be obtained from the input and output voltages weighted by the tappedinductor turns ratios as:
There are 7 topological states in one switching cycle of the proposed converter.
PreOnmode 1[~t_{0}, t_{6}~T]:
The switch
M
is in the offstate, and the resonant capacitor
C_{r}
is in its completely discharged state. The secondary current of
i_{m}
/
k
, which is equal to the primary current, is carried by output diode
D
_{1}
, and the magnetizing current
i_{m}
decreases with a slope of (
V_{i}

V_{c}
)/
L_{m}
. This mode is identical to the switchoff mode of the original TIB converter.
Onmode 1[t_{0}~t_{1}]:
This mode begins with the turnon of the switch
M
and represents a turnon delay due to the current slopes which are limited by the leakage inductance
L_{k}
. The equivalent circuit is shown in
Fig. 4
(a).
Topological stages in Onmodes.
Since the switch current
i_{ds}
is slanted with a slope limited by
L_{k}
, the zero current soft turnon of
M
is achieved, alleviating the switch turnon loss. This mode ends when the secondary current
i_{s}
reaches zero, providing zero current soft turnoff of output diode
D
_{1}
.
The reverserecovery current, which is not obvious in
Fig. 3
, is illustrated in detail in
Fig. 5
. Since the slope of is
i_{s}
also limited by
L_{k}
, the peak reverserecovery current of
D
_{1}
, denoted by –
I
_{rr1}
, is small, providing relaxation of the associated noise and EMI problems. This reverserecovery current slightly further discharges
L_{k}
and leaves a small negative current –
I
_{rr1}
in the secondary winding at the end of this mode.

*Key equations for Onmode1:
Onmode 2[t_{1}~t_{2}]:
The small negative current of –
I
_{rr1}
in the secondary winding, which is generated by the reverserecovery of
D
_{1}
at the end of Onmode 1, changes its path from
D
_{1}
to
D_{a}
at
t
=
t
_{1}
, and a path for the resonance of
C_{r}
and
L_{k}
to charge
C_{r}
is formed, as shown in
Fig. 4
(b). Neglecting the small jump of
I
_{rr1}
in the waveform of
i_{Da}
at
t
=
t
_{1}
, shown in
Fig. 5
, the resonance of
C_{r}
and
L_{k}
provides the zero current soft turnon of
D_{a}
by slowing down the current rise.

*Key equations for Onmode2:

whereω1=1/{N·(LkCr)1/2}, andim(t)=IM1.
After a half cycle of resonance, the voltage across
C_{r}
reaches its peak value obtained from (8) as:
and the secondary current reaches zero providing the zero current soft turnoff of
D_{a}
at
t
=
t
_{2}
. This is the end of Onmode 2.
The amount of charge released from
C_{c}
to charge
C_{r}
during this mode can be obtained from (6) as:
As illustrated in
Fig. 5
, the reverserecovery of
D_{a}
slightly further discharges
L_{k}
and leaves a small positive current
I
_{rr2}
in the secondary winding. This current turns on
D_{b}
, resulting in the equivalent circuit shown in
Fig. 4
(c). The slope of
i_{s}
at this moment can be obtained from (6) as:
If the sign of (11) is negative, the current
i_{s}
cannot increase any further beyond
I
_{rr2}
but immediately starts decreasing toward zero at
t
>
t
_{2}
, as shown in
Fig. 5
. After this current comes back to zero, the path diode
D_{b}
processes its reverserecovery and then turns back off, leaving a small negative current in the secondary winding. This small negative current brings back the equivalent circuit shown in
Fig. 4
(b) by turning on
D_{a}
again. Then the current
i_{s}
immediately restarts increasing toward zero, repeating the phase just before the first turnoff of
D_{a}
. In this case, the path diodes
D_{a}
and
D_{b}
repeatedly turn on and off, and the secondary current
i_{s}
settles to zero, eventually making both of the diodes remain in the turnoff state.
If the sign of (11) is positive, the current
i_{s}
further increases beyond
I
_{rr2}
after the turnon of
D_{b}
. In this case, the proposed converter maintains the equivalent circuit shown in
Fig. 4
(c), and
L_{k}
and
C_{r}
start another halfcycle of resonance to discharge
C_{r}
from the peak voltage given by (9). Since there is no benefit to this extra resonance, this paper suggests avoiding it by making (11) negative in pursuit of simple and efficient operation.
Reverserecovery currents of diodes.
Onmode 3[t_{2}~t_{3}]:
This mode is equivalent to the switchon mode of the original TIB converter, as shown in
Fig. 4
(d). All of the diodes are in the offstate, and the magnetizing current
i_{m}
increases with a slope of
V_{i}
/
L_{m}
and finally reaches
I
_{M2}
at the end of this mode at
t
=
t
_{3}
.
Offmode 1[t_{3}~t_{4}]:
This mode begins with the turnoff of
M
at
t
=
t
_{3}
. Neglecting the output capacitance of the diode, the path diode
D_{b}
turns on immediately, and the leakage inductance
L_{k}
starts resonating with
C_{ds}
to charge
C_{ds}
up to the clamp voltage
V_{c}
with its initial current of
I
_{M2}
. However, this resonance can also be neglected since its characteristic impedance (
L_{k}
/
C_{ds}
)
^{1/2}
, where the value of
L_{k}
is in most cases much larger than that of
C_{ds}
, can be assumed to be so much larger than 1 that there is negligible change in the resonant current from the initial value of
I
_{M2}
while the voltage across
C_{ds}
makes a large swing from zero to
V_{c}
. Therefore, this mode can be assumed to begin with the equivalent circuit shown in
Fig. 6
(a), where the capacitance
C_{ds}
is also neglected since it is in parallel with a much largervalued
C_{c}
.
During this mode, the energy stored in
L_{k}
is discharged into
C_{c}
, and the primary current is transferred to the secondary side with a slope limited by
L_{k}
. To simplify the analysis, from this point on the voltage
v_{r}
is assumed to maintain a constant value given by (9) during this mode.

*Key equations for Offmode1:

whereim(t)=IM2, andiDc(t) is the current flowing throughDc.
This mode ends when all of the energy in
L_{k}
is discharged into
C_{c}
, and (12) becomes zero providing the zero current soft turnoff of
D_{c}
.
Offmode 2[t_{4}~t_{5}]:
This mode consists of two components: one is the highfrequency ac component caused by the resonance of
C_{ds}
and
L_{k}
beginning with a fast decrease in
v_{ds}
from
V_{c}
, and the other one is the linear dc component representing a slow increase in
v_{ds}
caused by the discharge of
C_{r}
to the output. These two components are superposed so that the ac component is centered at the dc component, as shown in the interval of
t
_{4}
~
t
_{5}
in
Fig. 3
, where the ac components are shown by dotted lines.
The peak of the
v_{ds}
waveform is bounded to
V_{c}
by
D_{c}
and
C_{c}
, while the valley is bounded to the ground by the body diode of
M
. As the dc component in
v_{ds}
increases, the ac component is pulled up toward the clamp voltage
V_{c}
, which results in the repetitive clamping of the ac component of
v_{ds}
to
V_{c}
. Therefore, the ac component as well as the associated parasitic losses in the circuit decay rapidly.
Topological stages in Offmodes.
By neglecting the ac component, the equivalent circuit of this mode becomes as shown in
Fig. 6
(b). This mode ends when
C_{r}
is completely discharged, and the voltages across the switch
M
and output diode
D
_{1}
reach
V_{c}
and zero, respectively.

*Key equations for Offmode2:
The duration of this mode, denoted by
t
_{45}
, can be obtained from (14) as:
Offmode 3[t_{5}~t_{6}]:
The secondary current that was flowing through
D_{b}
is switched to
D
_{1}
at
t
=
t
_{5}
, providing the zero voltage soft turnon of
D
_{1}
. The equivalent circuit is shown in
Fig. 6
(c), where
C_{ds}
in parallel with
C_{c}
is neglected. In this mode, the clamp capacitor
C_{c}
resonates with
L_{k}
, and the charge lost by
C_{c}
in Onmode 2 is replenished. This mode ends when the diode
D_{c}
turns off with zero current after a halfcycle of resonance, resulting in the soft reset of the proposed snubber.
Since the assumption of a constant voltage on
C_{c}
makes it inappropriate to characterize this mode by setting up and solving differential equations, the charge balance in
C_{c}
is considered instead. Assume a sine current whose amplitude is unknown but whose angular frequency is equal to
k
/{
N
·(
L_{k}C_{c}
)
^{1/2}
}, a halfcycle integration of this current flowing into
C_{c}
can be balanced with (10) to give the following key equations.

*Key equations for Offmode3:

whereω2=k/{N·(LkCc)1/2}, andim(t)=IM2.
It should be noted that the charge gained by
C_{c}
through the clamping of
v_{ds}
in Offmodes 1 and 2, which is not considered in (17)(19), in practice, has the effect of reducing the amount of charge that should be replenished in this mode for the snubber reset, consequently reducing the amplitudes of the sine terms in (17)(19). The larger the charge gained in Offmodes 1 and 2, the smaller the peak of the resonant currents in Offmode 3. This makes the operation of the circuit in this mode less affected by the snubber. Therefore, it can be said that the neglect of the charge gained by
C_{c}
through the clamping of
v_{ds}
in Offmodes 1 and 2 leads to a worst case analysis without depreciating (17)(19).
Offmode 4[t_{6}~T]:
This mode is identical to PreOnmode 1. The equivalent circuit is shown in
Fig. 6
(d). This mode ends with the turnon of
M
at
t
=
T
. This completes one switching cycle of the proposed converter.
 B. Steadystate analysis
The finite current slopes due to
L_{k}
in the proposed converter generate Onmode 1 and Offmode 1. These modes occupy relatively small portions of one switching cycle. In addition, their contributions to the stepup ratio partially cancel each other out since they reduce both the onduty and the offduty. This will be further investigated in part of the next chapter. Therefore, in deriving the steadystate stepup ratio of the proposed converter, these two modes are neglected. Otherwise, the procedure and result become very complicated and do not give very useful insights.
Waveform of primary winding voltage.
VALUES FOR EACH INTERVAL
In order to apply the voltsecond balance condition on
L_{m}
for one switching cycle, the voltage waveform across it, denoted by
v_{p}
, is analyzed in
Fig. 7
, where Onmode 1 and Offmode 1 are excluded as mentioned above. The duration of each interval and the corresponding average value of
v_{p}
are shown in
Table I
. Summing all of the products of the expressions in each row of
Table I
and equating the result with zero result in:
Since the average voltage across
L_{k}
in Onmode 2, which can be obtained from (5), is zero, the magnetizing current can be considered to be linearly increasing all the way from
I
_{M1}
to
I
_{M2}
during Onmodes 2 and 3, that is,
I
_{M2}
=
I
_{M1}
+
V_{i}DT
/
L_{m}
. Then, the output current
I_{o}
can be obtained by:

i) integrating the secondary current assumed to be linearly decreasing fromIM2/k toIM1/kduring Offmodes 2~4,

ii) integrating the sine term in (19) representing the droop due to the charging current ofCcduring Offmode 3,

iii) subtracting ii) from i),

iv) and dividing iii) by one switching periodTas:
Solving (21) for
I
_{M2}
and substituting the result into (20) gives:
Eliminating
V_{c}
in (22) by using (1) yields the stepup ratio of the proposed converter as:
where α is as given in (22).
It can be seen from (23) that the parameter α plays a role in increasing the effective duty ratio of the proposed converter. This makes the proposed converter show a higher stepup ratio than the original TIB converter. It can be also seen that a smaller value of
C_{r}
or a larger value of
I_{o}
/
V_{i}
results in a smaller value of α. In an extreme case where α=0, (23) becomes the stepup ratio of the original TIB converter.
III. DESIGN CONSIDERATIONS
Various conditions validating the analysis in the previous chapter are presented in this chapter. Some characteristics of the proposed converter are also discussed in detail. To achieve a seamless flow of the analysis, the validity of (23) will be assumed for the time being. However, in a later part of this chapter, it will be verified by simulation results.
 A. Nodischarge of Crin Onmode
In order to make
C_{r}
not discharge through an extra resonance mode after being charged to its peak voltage in Onmode 2, the sign of (11) is required to be negative as explained in Onmode 2. This condition can be simplified in terms of
V_{o}
/
V_{i}
as:
This defines the lower boundary of the stepup ratio. The righthand side of (24) has a minimum value of 9 at
N
=3 and it becomes larger as
N
increases or decreases.
 B. Continuity of is
Due to the resonant charging current of
C_{c}
in Offmode 3, a current droop appears in the secondary current
i_{s}
as shown in the
t
=
t
_{5}
~
t
_{6}
interval of
Fig. 3
. This droop is represented by the sine term in (19). If this droop is too large to maintain the secondary current continuous, the proposed converter enters a mode similar to the discontinuous current mode, and the output voltage becomes significantly higher than that predicted by (23). However, this is beyond the scope of this paper, and it should be avoided by the following condition:
Eliminating
I
_{M2}
in (25) using (20) gives a relationship which makes it more convenient to determine the continuity of
i_{s}
as:
This defines the upper boundary condition for the stepup ratio.
 C. Soft Reset of the Snubber
If the duty ratio of the proposed converter is too large, the switch can turn on prematurely before
L_{k}
and
C_{c}
complete their halfcycle of resonance to make the resonant current in
D_{c}
zero. In this case, the peak resonant current automatically increases to maintain the charge balance in
C_{c}
as the switching cycle continues. This does not significantly change the average behavior of the proposed converter over one switching cycle. However, this hard reset of the proposed snubber needs to be avoided since the clamp diode
D_{c}
in this case fails in achieving soft turnoff which results in an increase in associated side effects such as switching loss and electric noise.
To make sure that the proposed snubber can achieve a soft reset in Offmode 3, the sum of the time required for
C_{r}
to completely discharge in Offmode 2 and the time required for
L_{k}
and
C_{c}
to complete a halfcycle of resonance in Offmode 3 should be shorter than (1–
D
)
T
. This condition can be obtained from (16) and
ω
_{2}
in Offmode 3 as:
By eliminating
I
_{M2}
in (27) using (20), another upper boundary condition for the stepup ratio of the proposed converter to ensure the soft reset of the snubber can be obtained as:
 D. Characteristic Analysis
In order to make the analysis of the proposed converter more convenient and illustrative, the practical parameters are assumed to be
V_{i}
=12V,
f
=100kHz,
N
=3,
L_{m}
=80
μ
H,
C_{c}
=330nF, and
C_{r}
=4.7nF.
Fig. 8
shows the stepup ratio of the proposed converter in comparison with that of the original TIB converter. The solid lines represent the analytic stepup ratios obtained from (23) for different values of
I_{o}
. The shortdashed lines represent the simulated stepup ratios obtained from SPICE simulations for the same values of
I_{o}
, where
L_{k}
=2
μ
H is assumed. The dashdot and dotted lines represent the upper boundaries given by (26) and (28) for the secondary current continuity and soft reset of the snubber, respectively. The longdashed line is the lower boundary given by (24) with no discharge of
C_{r}
during the onmode.
Fig. 8
can be explained from the following standpoints.
Stepup ratios and boundary conditions.
1) Effect of L_{k}
:
Fig. 8
shows that there is a large difference between the analytic and simulated stepup ratios of the conventional TIB converter. This can be explained by using
Fig. 9
, where the solid and dotted lines represent the secondary current
i_{s}
and the drain voltage
v_{ds}
of each converter from the moment the switch turns off. The turnoff delay, denoted by
t_{offdly}
, is defined for each converter by the time required for
i_{s}
to rise from zero to the steadystate value of
I
_{M2}
/
k
.
Turnoff transient waveforms of i_{s} and v_{ds}.
As already explained in Offmode 1, the drain voltage
v_{ds}
of the proposed converter is assumed to be clamped at
V_{c}
as soon as the switch turns off. The voltage across
L_{k}
is also clamped since all of the voltages at the tappedinductor terminals are clamped as shown in
Fig. 6
(a). Therefore, all of the winding currents vary linearly with slopes determined by the value of
L_{k}
and the voltage across it. The slope of
i_{s}
is also shown in
Fig. 9
(a). Assuming the voltage across
C_{r}
is kept constant during the turnoff transient caused by
L_{k}
, the turnoff delay of the proposed converter, denoted by
t
_{offdly(prop.)}
, can be obtained as:
On the other hand, in the original TIB converter,
L_{k}
and
C_{ds}
freely resonate after the switch turns off, as shown in
Fig. 9
(b), since the drain voltage
v_{ds}
is not limited. The time required for the current
i_{s}
to reach the steadystate value of
I
_{M2}
/
k
, denoted by
t
_{offdly(conv.)}
is equal to 0.5π(
L_{k}C_{ds}
)
^{1/2}
. This corresponds to a quartercycle of resonance. This is much shorter than
t
_{offdly(prop.)}
given in (29).
Meanwhile, the turnon delay, defined by the time required for
i_{s}
to reach zero after the switch turns on, can be obtained for the proposed converter from (3) as:
Since no snubber elements are associated with (30), the turnon delay of the original TIB converter can be expressed by (30) and it can be considered equal to that of the proposed converter neglecting a small difference in
I
_{M1}
.
The turnoff delay reduces the offduty of the converter and adds a small positive voltsecond on the primary winding, resulting in a slight increase in the stepup ratio. On the other hand, the turnon delay reduces the onduty of the converter and adds a small negative voltsecond on the primary winding, resulting in a slight decrease in the stepup ratio. In the proposed converter, the effects of the turnon delay and the turnoff delay partially cancel each other out since the turnoff delay is relatively large. However, in the original TIB converter, this barely happens since the turnoff delay is very small. This makes the stepup ratio of the proposed converter less affected by
L_{k}
. Therefore, the original TIB converter shows a large difference between the analytic stepup ratio, where
L_{k}
is neglected, and its simulated counterpart, where
L_{k}
is taken into account.
2) No discharge of C_{r} in the onmode
: Assuming
N
=3, (24) gives a minimum stepup ratio of 9, which is shown by the longdashed line in
Fig. 8
. Outside this boundary, which corresponds to a stepup ratio lower than 9, an extra resonance takes place to partially discharge
C_{r}
, resulting in a reduced voltage across
C_{r}
at the end of the onmodes. This also results in a reduction in the time required for the complete discharge of
C_{r}
given by (16) in Offmode 2. As can be seen from
Fig. 7
, a decrease in
t
_{45}
adds a small negative voltsecond on the primary winding, resulting in a slight decrease in the stepup ratio. This is not considered when deriving the analytic stepup ratio of (23). Therefore, a simulated stepup ratio outside this boundary decreases faster with a decreasing
D
than its analytic counterpart, as shown in
Fig. 8
.
3) Soft reset of the snubber:
Fig. 10
compares the soft reset and hard reset of the proposed snubber. In the soft reset case, the primary current
i_{p}
starts increasing from
I
_{M1}
/
k
at
t
=
t
_{0}
and reaches
I
_{M1}
at
t
=
t
_{1}
. In the hard reset case, the switch prematurely turns on at
t
=
t
_{0}
'
before
L_{k}
and
C_{c}
complete their halfcycle of resonance. Therefore, the primary current starts increasing from a value higher than
I
_{M1}
/
k
and reaches
I
_{M1}
at
t
=
t
_{1}
'
. Since the slope of
i_{p}
after the switch turnon is the same in both cases, the time required for
i_{p}
to reach
I
_{M1}
from the switch turnon is shorter in case of the hard reset. This in effect reduces the turnon delay defined by Onmode 1 and adds a small positive voltsecond on the primary winding, resulting in a slight increase in the stepup ratio. This is shown in
Fig. 8
, where the boundary condition given by (28) for the soft reset of the snubber is shown by a dotted line. Outside this boundary, the proposed converter performs the hard reset. Therefore, the simulated stepup ratio increases slightly faster with an increasing
D
than its analytic counterpart in which the snubber hard reset is not considered.
Soft reset and hard reset of proposed snubber.
4) Continuity of i_{s:}
By comparing the stepup ratio given by (23) and the boundary condition for the continuity of
i_{s}
given by (26), it can be easily seen that they have very similar forms, and that their traces would also have very similar shapes if they came close to each other. Due to this similarity between (23) and (26), there is no meaningful intersection between them, and the range of the stepup ratio retaining the continuity of
i_{s}
cannot be well defined. Therefore, this paper suggests designing the proposed converter so that the boundary given by (26) is placed far above the operating condition as exemplified in
Fig. 8
.
5) Trends of boundary conditions:
The analytic stepup ratios of the proposed converter for different values of
N
are illustrated in
Fig. 11
together with all of the boundary conditions presented in this paper. Since the proposed converter is for use in highstep applications, the smallest value of
N
is chosen to be 3. The output current
i_{s}
assumed to be 0.24A, and all of the other parameters are the same to those used in
Fig. 8
.
As the turns ratio
N
increases, the minimum duty ratio for no discharge of
C_{r}
in the onmodes shows a decreasing tendency as denoted by
P
_{1}
,
P
_{2}
, and
P
_{3}
, and so does the maximum duty ratio for the soft reset of the proposed snubber as denoted by
P
_{4}
,
P
_{5}
, and
P
_{6}
. The boundary for the continuity of
i_{s}
, which is already high enough, becomes even higher as
N
increases.
Trends of boundaries according to N (I_{o}=0.24A).
IV. EXPERIMENTAL RESULTS
A prototype of the proposed converter has been built for a 120V/240mA LED lighting module. The input voltage ranges from 9V to 12V depending on the offgrid power system loaded by a 12V battery, and the required stepup ratio ranges from 10 to 13.3. In light of
Fig. 11
, the tappedinductor turns ratio is designed as
N
=3, and the operating duty ratio ranges from 0.66 to 0.72. The switching frequency is chosen to be 100kHz. The tappedinductor is designed to have a magnetizing inductance of 80
μ
H and it is made with highly productive separate windings to have good insulation and low capacitance between windings. The leakage inductance is measured to be 2.4
μ
H. Since the switch voltage is minimized in the proposed converter, a 60Vrated AOD442 with
R
_{DS(ON)}
=20mΩ is chosen as the switch. A 400V/1Arated ES1G is chosen as the boost diode, and 100V/1Arated MBRS1100s are chosen as the clamp and path diodes. The snubber capacitors are designed as
C_{c}
=330nF and
C_{r}
=4.7nF. The key parameters are summarized in
Table II
.
KEY PARAMETERS
The operating waveforms at the rated input voltage of 12V are shown in
Figs. 12
and
13
. The operating duty ratio is 0.68, and the measured efficiency is 97%. The continuous switch drain voltage is measured to be 39.2V as expected by (1), and the peak switch drain voltage, which is caused by parasitic inductances between the components, is measured to be 50.3V. All of the waveforms match well with the theoretical waveforms shown in
Fig. 3
. In addition, the operation and features of the proposed converter such as the minimum switch voltage stress, softswitching of the switching devices, relaxation of the diode reverserecoveries, and the soft reset of the snubber are all verified.
Key experimental waveforms.
Additional experimental waveforms.
For a comparison with the original TIB converter without a snubber, the switch of the proposed converter is replaced by a 200Vrated AOD2210 with
R
_{DS(ON)}
=105mΩ. The efficiency is 94.5%, which is lower than the previous result of 97%. This is due to increased conduction loss since the AOD2210 has a higher
R
_{DS(ON)}
than the AOD442. The proposed converter is then converted into the original TIB converter by removing all of the snubber elements and adjusting the duty cycle to be 0.72 to maintain the output voltage. An extra 470pF capacitor is added between the switch drain and source to keep the voltage spike from becoming too high by exceeding 600V. The efficiency of this original TIB converter is measured to be 91.5%, which is lower than the previous result of 94.5%. This is due to the increased switching loss. The results are summarized in
Table III
. The key loss factors affecting the results are also shown in the table.
MEASURED EFFICIENCY
V. CONCLUSIONS
For emerging high stepup applications in offgrid areas, a passive lossless snubber for the continuous current mode TIB converter is proposed. In the proposed converter, the switch voltage is minimized so that a more inexpensive and efficient device can be selected. Softswitching is also provided for the switching devices, resulting in a further improvement in the efficiency. Since the leakage inductance is fully utilized, there is no need to minimize it, which allows for a costeffective and highly productive design of the tappedinductor. In addition, the proposed converter has a high stepup ratio and provides relaxation of the diode reverserecovery.
In this paper, the operation principle and characteristics of the proposed converter are analyzed in detail. In addition, various boundary conditions that need to be considered in the design are presented. Some experimental results are also presented to confirm the operation and features of the proposed converter.
BIO
Jeongil Kang received his B.S., M.S., and Ph.D. degrees in Electrical Engineering from the Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea, in 1995, 1997, and 2002, respectively. In 2002, he joined Samsung Electronics Co., Ltd., Suwon, Korea, and is currently working on the design and control of semiconductor light source drivers for various display and lighting devices as a Principal Engineer in the Visual Display Research & Development Office. His current research interests include the development, modeling, and control of power converter topologies. He is a Member of the Korean Institute of Power Electronics (KIPE).
SangKyoo Han received his M.S. and Ph.D. degrees in Electrical Engineering and Computer Science from the Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea, in 2001 and 2005, respectively. For the next six months, he was a PostDoctoral Fellow at KAIST where he developed digital display power circuits and preformed several research activities. Since 2005, he has been with the Department of Electrical Engineering, Kookmin University, Seoul, Korea, as an Associate Professor. He has also worked for the Samsung Power Electronics Center (SPEC) and the Samsung Network Power Center (SNPC) as a Research Fellow. His primary areas of research interests include power converter topologies, LED drivers, renewable energy systems, and battery chargers for electric vehicles. Dr. Han is a Member of the Korean Institute of Power Electronics (KIPE).
Jonghee Han is an Executive Vice President of the Visual Display Business for Samsung Electronics. In this role, he leads the Visual Display Research & Development Office. Since joining Samsung in 1988, Mr. Han has played a significant role in the success of Samsung’s Visual Display business, helping it solidify its top position within the consumer electronics industry. One of his many successes includes the launch of the awardwinning ‘Bordeaux’ TV series, which became one of the world’s bestselling LCD TVs in 2006. During his tenure, Mr. Han has accomplished a number of “firsts” in the TV industry, serving at the forefront of some of the most noteworthy TV advancements. For example, he was an integral force in the development of the world’s first 3D LED TV in 2010 and first Smart TV in 2011, among numerous other developments. More recently, he played a critical role in the creation of both the world’s first Curved OLED TV and Curved UHD TV. In particular, his research specialties include image processing, audio signal processing, analog/digital mixed systems, display devices, and power electronics for both innovative displays and audio products. Along with contributing his vast knowledge to Samsung’s Visual Display Business, he is also a Member of the Korean Institute of Power Electronics (KIPE). Mr. Han holds a bachelor’s degree in electrical engineering from Inha University, Incheon, Korea.
Misak S.
,
Prokop L.
2010
“Offgrid power systems,”
in Proc. EEEIC
14 
17
Frame D.
,
Tembo K.
,
Dolan M. J.
,
Strachan S. M.
,
Ault G. W.
2011
“A community based approach for sustainable offgrid PV systems in developing countries,”
in Proc. IEEE Power and Energy Society General Meeting
1 
7
Freitas A. A.
,
Daher S.
,
Antunes F.
,
Ximenes S.
,
Viana F.
,
Sá 1, E.
,
Silva F. S.
,
Soares E. A.
2010
“Offgrid PV system to supply a rural school on DC network,”
in Proc. ICREPQ
paper number 535
2013
Clean Energy Ministerial Global LEAP awards
http://www.globalleapawards.org
Vazquez N.
,
Estrada L.
,
Hernandez C.
,
Rodriguez E.
2007
“The tappedinductor boost converter,”
in Proc. ISIE
538 
543
Jang D.H.
,
Kang J.I.
,
Han S.K.
2013
“High efficiency lossless snubber for photovoltaic maximum power point tracker,”
Trans. KIPE
18
(5)
485 
491
Nakamura M.
,
Ogura K.
,
Nakaoka M.
2004
“Softswitching PWM boost chopperfed dcdc power converter with load side auxiliary passive resonant snubber,”
Journal of Power Electronics
4
(3)
161 
168
Ahmed T.
,
Nagai S.
,
Hiraki E.
,
Nakaoka M.
2005
“A new three winding coupled inductorassited high frequency boost chopper type dcdc power converter with a high voltage conversion ratio,”
Journal of Power Electronics
5
(2)
99 
103
Wai R.J.
,
Duan R.Y.
2005
“High stepup converter with coupledinductor,”
IEEE Trans. Power Electron.
20
(5)
1025 
1035
DOI : 10.1109/TPEL.2005.854023
Dong W.
,
Zhao Q.
,
Liu J.
,
Lee F. C.
2002
“A boost converter with lossless snubber under minimum voltage stress,”
in Proc. APEC
509 
515
Tseng C.J.
,
Chen C.L.
1998
“A passive lossless snubber cell for nonisolated PWM DC/DC converters,”
IEEE Trans. Ind. Electron.
45
(4)
593 
601
DOI : 10.1109/41.704887
Tseng C.J.
,
Chen C.L.
1998
“Passive lossless snubbers for DC/DC converters,”
IEE Proc.Circuits, Devices and Systems
145
(6)
396 
401
DOI : 10.1049/ipcds:19981877
Kang J.M.
,
Lee S.H.
,
Hong S.S.
,
Han S.K.
2012
“Voltage clamped tappedinductor boost converter with high voltage conversion ratio,”
Trans. KIPE
17
(1)
34 
40
Li R. T. H.
,
Chung H. S.H.
2010
“A passive lossless snubber cell with minimum stress and wide softswitching range,”
IEEE Trans. Power Electron.
25
(7)
1725 
1783
DOI : 10.1109/TPEL.2010.2042074
Smith K. M.
,
Smedley K. M.
2001
“Engineering design of lossless passive soft switching methods for PWM converters. I. with minimum voltage stress circuit cells,”
IEEE Trans. Power Electron.
16
(3)
336 
344
DOI : 10.1109/63.923765
Levy H.
,
Zafrany I.
,
Ivensky G.
,
BenYaakov S.
1997
“Analysis and evaluation of a lossless turnon snubber,”
in Proc. APEC
757 
763
Kim D.H.
,
Jang J.H.
,
Park J.H.
,
Kim J.W.
2013
“Singleended highefficiency stepup converter using the isolated switchedcapacitor cell,”
Journal of Power Electronics
13
(5)
766 
778
DOI : 10.6113/JPE.2013.13.5.766
Kang J.i.
,
Han S.K.
,
Han J.
2014
“Lossless snubber for tappedinductor boost converter for high stepup application,”
in Proc. ICIT
253 
260