Sei sulla pagina 1di 6

This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2834418, IEEE
Transactions on Power Electronics

Comments on “A Single-Inductor Multiple-Output Switcher With


Simultaneous Buck, Boost, and Inverted Outputs”
Majid Abbasi1, Ahmad Afifi1 and Mohamad Reza Alizadeh Pahlavani1
1
Faculty of Electrical Engineering, Malek-Ashtar University of Technology (MUT) Tehran, Iran
Abstract- In a paper [1], a single-inductor simulation results to show the validity of the
multiple-output switcher was presented to obtained parameters.
produce buck, boost, and inverted outputs Owing to the nonlinear equation between
simultaneously. Unfortunately, this study the outputs and the duty cycles, deriving a
showed some unexpected mistakes, direct equation to calculate the inductor
especially in state-space equations and in the seemed impossible. Therefore, a systematic
calculation of output voltages. The objective process was presented to calculate the
of this note was to provide correct analysis inductor. The obtained parameters in this
and rectify the errors in this paper. Some work could be an effective supplement and
tables and equations have been modified that extension to the original work in paper [1].
were present in the initial draft of this paper.
The obtained parameters in this work could II. Mathematical Manipulation
be an effective supplement and extension for In this section the outputs to the input DC-
the original work in the paper [1]. DC transfer functions of the proposed
converter in paper [1] was derived with direct
Index Terms— Switching converter, buck- mathematical manipulation. The operational
boost-inverted outputs, single-inductor timing diagram of the proposed three-output
multiple-output (SIMO). SIBBI output DC–DC converter is shown in
Fig.1. The mathematical manipulations of
I. Introduction- In a the paper [1], a single-
inductor current are as follow:
inductor buck, boost and inverted (SIBBI)
Mode 1: 0 < 𝑡 < 𝐷0 𝑇
output DC–DC converter was presented. 𝑣
𝑖𝐿 (𝑡) = 𝑖𝑛 𝑡 + 𝑚 (1)
This structure could independently generate 𝐿
controlled buck, boost, and inverted outputs The peak inductor current which occurred at
simultaneously. In DC-DC power conversion 𝑛 = 𝑖𝐿 (𝐷0 𝑇) was written as:
𝑣
applications, to achieve minimization by 𝑛 = 𝑖𝑛 𝐷0 𝑇 + 𝑚 (2)
𝐿
removing the bulky transformers, cost and Mode 2: 𝐷0 𝑇 < 𝑡 < (𝐷0 + 𝐷1 )𝑇
𝑣 −𝑣
weight, the SIBBI converter can be a good 𝑖𝐿 (𝑡) = 𝑖𝑛 1 (𝑡 − 𝐷0 𝑇) + 𝑛 (3)
𝐿
solution. In Section III of the paper [1], the
authors used an inductor current ripple-based
averaging method for modeling the proposed
converter [1, Fig. 1]. By using this method,
steady-state transfer functions were obtained.
Unfortunately, certain achieved equations
were incorrect. As a consequence, the figures
of the operating range of the output, shown
in Fig. 8 of paper [1], and the voltage
parameters tabulated on Table I of paper [1]
were found to be unbelievable to the readers.
In this paper, steady-state transfer functions
between the outputs and the input, using the
inductor current ripple-based modeling
approach, have been recalculated. The
obtained results were compared with the
mathematical manipulation and the Fig. 1 Operational key waveforms for three-output
SIBBI output DC–DC converter proposed in paper [1]

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2834418, IEEE
Transactions on Power Electronics

At 𝑡 = (𝐷0 + 𝐷1 )𝑇 the inductor current 𝑎2 𝑅3 𝐷3 2𝐿−𝑅2 𝐷22 𝑇 𝑅3 𝐷3 (𝐷2 +𝐷3 )𝑇


𝑎4 = � �+� � (21)
𝑅2 𝐷2 2𝐿+𝑅3 𝐷32 𝑇 2𝐿+𝑅3 𝐷32 𝑇
𝑖𝐿 ((𝐷0 + 𝐷1 )𝑇) = 𝑝 so that:
𝑣𝑖𝑛 −𝑣1 Constraint on voltage-second balance
𝑝= 𝐷1 𝑇 + 𝑛 (4)
𝐿 equation under the switching sequence is
Mode 3: (𝐷0 + 𝐷1 )𝑇 < 𝑡 < (𝐷0 + 𝐷1 + 𝐷2 )𝑇 illustrated in Fig.1:
𝑣𝑖𝑛 −𝑣2
𝑖𝐿 (𝑡) = (𝑡 − (𝐷0 + 𝐷1 )𝑇) + 𝑝
(5) 𝐷0 𝑣𝑖𝑛 + 𝐷1 (𝑣𝑖𝑛 − 𝑣1 ) + 𝐷2 (𝑣𝑖𝑛 − 𝑣2 ) +
𝐿
At 𝑡 = (𝐷0 + 𝐷1 + 𝐷2 )𝑇 the inductor current 𝐷3 (𝑣𝑖𝑛 − 𝑣3 ) = 0 (22)
equals to q which is: Substituting Equations (14) and (19) to
𝑣 −𝑣
𝑞 = 𝑖𝑛 2 𝐷2 𝑇 + 𝑝 (6) Equation (22) gives the DC-DC transfer
𝐿
Mode 4: (𝐷0 + 𝐷1 + 𝐷2 )𝑇 < 𝑡 < 𝑇 function V1/Vin as fallows:
𝑣 −𝑣 𝑣1 1−(𝑎2 𝐷2 +𝑎4 𝐷3 )
𝑖𝐿 (𝑡) = 𝑖𝑛 3 (𝑡 − (𝐷0 + 𝐷1 + 𝐷2 )𝑇) + 𝑞 (7) = (23)
𝐿 𝑣 𝐷 +𝑎 𝐷 +𝑎 𝐷
𝑖𝑛 1 1 2 3 3
End of this time interval, t=T, the inductor
current equals to m which is: Correction 1: Inductor current averaged
𝑣 −𝑣
𝑚 = 𝑖𝑛 3 𝐷3 𝑇 + 𝑞 (8) over the entire switching
𝐿
The outputs average current, as shown in Using Equations (1) to (7) from paper [1],
Fig.1, could be written as: the inductor current averaged over the entire
𝑝+𝑛 𝑣 switching cycle was recalculated as:
𝐼𝑎𝑣𝑒 (𝑜𝑢𝑡1) = 𝐷1 = 1 (9) 𝑣𝑖𝑛 𝑇𝑠 𝑣1 𝑇𝑠
2 𝑅 1 𝚤̅𝐿 = 𝑚 + (𝑑0 + 𝑑1 + 𝑑2 ) + 𝑑1 (𝑑0 − 𝑑2 − 1)
2𝐿 2𝐿
𝑞+𝑝 𝑣2
𝐼𝑎𝑣𝑒 (𝑜𝑢𝑡2) = 𝐷2 = (10) +
𝑣2 𝑇𝑠
(24) 𝑑2 (𝑑0 + 𝑑1 − 1)
2 𝑅2 2𝐿

𝐼𝑎𝑣𝑒 (𝑜𝑢𝑡3) =
𝑚+𝑞 𝑣
𝐷3 = 3 (11) Where the 𝚤̅𝐿 from (9) of paper [1] was given
2 𝑅 3 as:
Substituting n from (4) into (9), the p could 𝑣𝑖𝑛 𝑇𝑠
be derived as: 𝚤̅𝐿 = 𝑚 + (𝑑0 + 𝑑1 + 𝑑2 )(1 − 𝑑2 ) −
𝐿
1 𝐷1 𝑇 𝑣𝑖𝑛 𝑣1 𝑇𝑠
𝑝=� − � 𝑣1 + 𝐷1 𝑇 (12) 𝑑1 (1 − 𝑑0 ) (25)
𝑅1 𝐷1 2𝐿 2𝐿 𝐿
𝑣2 𝑇𝑠
Using Equations (6) and (10), the p was − (1 − 𝑑0 )(1 − 𝑑0 − 𝑑1 − 𝑑2 )
given by 𝐿
1 𝐷2 𝑇 𝑣𝑖𝑛
The step by step procedure to calculate the
𝑝=� + � 𝑣2 − 𝐷2 𝑇 (13) correct value of 𝚤̅𝐿 is given below.
𝑅2 𝐷2 2𝐿 2𝐿
Setting Equations (12) and (13) equal to each Step 1: Using (1) from paper [1] could be
other: written as:
𝑚+𝑛 𝑣 𝑑 𝑇
𝑣2 = 𝑎1 𝑣1 + 𝑎2 𝑣𝑖𝑛 (14) = 𝑚 + 𝑖𝑛 0 𝑠 (26)
2 2𝐿
Where Using (6) and (1) from paper [1] could be
𝑅2 𝐷2 2𝐿−𝑅1 𝐷12 𝑇 written as:
𝑎1 = � � (15)
𝑅1 𝐷1 2𝐿+𝑅2 𝐷22 𝑇 𝑝+𝑛 𝑣 𝑑 𝑇 (𝑣 −𝑣 )𝑑 𝑇
= 𝑚 + 𝑖𝑛 0 𝑠 + 𝑖𝑛 1 1 𝑠 (27)
𝑅2 𝐷2 (𝐷1 +𝐷2 )𝑇 2 𝐿 2𝐿
𝑎2 = � � (16)
2𝐿+𝑅2 𝐷22 𝑇 Using (6) and (7) from paper [1] could be
Substituting p from (6) into (10), the q could written as:
be derived as: 𝑝+𝑞 𝑣𝑖𝑛 𝑑0 𝑇𝑠 (𝑣𝑖𝑛 − 𝑣1 )𝑑1 𝑇𝑠
=𝑚+ +
1 𝐷2 𝑇 𝑣𝑖𝑛 2 𝐿 𝐿
𝑞=� − � 𝑣2 + 𝐷2 𝑇 (17) (𝑣 −𝑣 )𝑑 𝑇
𝑅2 𝐷2 2𝐿 2𝐿 + 𝑖𝑛 2 2 𝑠 (28)
2𝐿
Using Equations (8) and (11), the q was Using (7) from paper [1] could be written as:
given by 𝑞+𝑚 𝑣𝑖𝑛 𝑑0 𝑇𝑠 (𝑣𝑖𝑛 − 𝑣1 )𝑑1 𝑇𝑠
1 𝐷3 𝑇 𝑣𝑖𝑛 =𝑚+ +
𝑞=� + � 𝑣3 − 𝐷3 𝑇 (18) 2 2𝐿 2𝐿
𝑅3 𝐷3 2𝐿 2𝐿 (𝑣𝑖𝑛 −𝑣2 )𝑑2 𝑇𝑠
+ (29)
Similarly, setting Equations (17) and (18) 2𝐿
equal to each other and using (14) cause: Step 2: Substituting (26) - (29) to Equation
𝑣3 = 𝑎3 𝑣1 + 𝑎4 𝑣𝑖𝑛 (19) (8) of original paper gives inductor current
Where average over the entire switching cycle as:
𝑣𝑖𝑛 𝑇𝑠 𝑣1 𝑇𝑠
𝑎1 𝑅3 𝐷3 2𝐿−𝑅2 𝐷22 𝑇
𝚤̅𝐿 = 𝑚 + (𝑑0 + 𝑑1 + 𝑑2 ) + 𝑑1 (𝑑0 − 𝑑2 − 1) +
2𝐿 2𝐿
𝑎3 = � � (20) 𝑣2 𝑇𝑠
𝑑2 (𝑑0 + 𝑑1 − 1)
𝑅2 𝐷2 2𝐿+𝑅3 𝐷32 𝑇 (30)
2𝐿

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2834418, IEEE
Transactions on Power Electronics

Correction 2: State Space Equations Which, difference between the (34) from
In calculation of state space equations of this work and (24) from paper [1] is the
paper [1], some matrices contained errors matrix b2 (Equation (34) of paper [1]).
which would be corrected in this section. Mode 4: In mode 4, as shown in Fig. 6 of
Mode 2: In mode 2, as shown in Fig.4 of paper [1], the differential Equation (28) of
paper [1], the differential Equation (16) of original paper for state variable v3 could be
original paper for state variable v1 could be re-written as:
𝑑𝑣3 𝑞+𝑚 𝑣 𝑖
re-written as: = − 3 − 3 (35)
𝑑𝑣1 𝑛+𝑝 𝑣 𝑖 𝑑𝑡 2𝐶 𝑅 𝐶 𝐶
1 3 3 3
= − 1 − 1 (31) 𝑞+𝑚
𝑑𝑡 2𝐶 𝑅 𝐶 1𝐶 2 1 1 Substituting the value of from (29) to
𝑛+𝑝 2
Substituting the value of from (27) to (35), the state space representation becomes:
2
𝚤̇𝐿
(31), the state space representation becomes: 𝑣̇ 1
(32) � �=
𝑣̇ 2
(36)
𝚤̇𝐿 𝑣̇ 3
𝑣̇ 1
� 1� ⎡0 ⎤
𝑣̇ 2 0 0 −
𝑣̇ 3 ⎢ 𝐿 ⎥
1
⎡0
1
− 0 0 ⎤
⎢0 − 0 0 ⎥ 𝑖𝐿
𝐿 ⎢ 𝑅1 𝐶1 ⎥ 𝑣1
⎢ ⎥
⎢1 1 𝑑1 𝑇𝑠 𝑑2 𝑇𝑠 ⎢ 1 ⎥ �𝑣 �
− − (𝑑 − 𝑑2 ) − (𝑑 + 𝑑1 − 1) 0 ⎥ 𝑖𝐿 ⎢0 0 − 0 ⎥ 𝑣2
⎢𝐶1 𝑅1 𝐶1 2𝐿𝐶1 0 2𝐿𝐶1 0 ⎥ 𝑣1 𝑅2 𝐶2
= ⎢ 1 ⎥ �𝑣 � ⎢ ⎥ 3
⎢0 0 −
𝑅2 𝐶2
0 ⎥ 𝑣2 ⎢1 −
𝑇𝑠
𝑑 (𝑑 − 𝑑2 ) −
𝑇𝑠
𝑑 (𝑑 + 𝑑1 ) −
1 ⎥
⎢ ⎥ 3 ⎣𝐶3 2𝐿𝐶3 1 0 2𝐿𝐶3 2 0 𝑅3 𝐶3 ⎦
⎢0 1 ⎥
0 0 − 1
⎣ 𝑅3 𝐶3 ⎦ ⎡ 0 0 0 ⎤

1
0 0 0 ⎤ ⎢𝐿 ⎥
⎢ 𝐿 ⎥ ⎢0 − 1 0 0 ⎥ 𝑖𝑛
𝑣
⎢− 𝑇𝑠 (𝑑2 − 𝑑0 ) − 1 0 0 ⎥ 𝑖𝑛
𝑣 ⎢ 𝐶1 ⎥ 𝑖1
⎢ 2𝐿𝐶1 𝐶1 ⎥ 𝑖1 +⎢ 1 ⎥ � 𝑖2 �
+⎢ 1 ⎥ 𝑖2 �
� ⎢0 0 − 0 ⎥ 𝑖
⎢ 0 0 − 0 ⎥ 𝑖 ⎢ 𝐶2 ⎥
3
𝐶2 3
⎢ ⎥ ⎢0 1⎥
⎢ 1 0 0 −
0 0 0 − ⎥ ⎣ 𝐶3 ⎦
⎣ 𝐶3 ⎦
Which is not equal to (19) and (33) of the Which, difference between the (36) from this
original paper. work and (29) of paper [1] is the matrix b3
Mode 3: In mode 3, as shown in Fig. 5 of (Equation (35) of paper [1]).
paper [1], the differential Equation (22) of Therefore, the matrices a1, b2 and b3 were
original paper for state variable v2 could be not same with the corresponding matrices in
re-written as: the original paper. Also, the Equation (36) of
𝑑𝑣2 𝑝+𝑞 𝑣 𝑖 paper [1] should be modified to:
= − 2 − 2 (33)
𝑑𝑡 2𝐶 𝑅 𝐶 1 𝐶 2 1 2
𝑝+𝑞
𝑋 = −𝐴−1 𝐵𝑈 (37)
Substituting the value of from (28) to
2
(33), the state space representation becomes: Correction 3: Results and Discussions
In this section the comparison was
(34) performed among mathematical
1 manipulation, state-space equations, and
⎡0 0 − 0 ⎤
𝐿 PSIM software simulation (Fig. 2) in steady-
⎢ 1 ⎥
𝚤̇𝐿 ⎢0 − 0 0 ⎥ 𝐿
𝑖
𝑣̇ ⎢
� 1� = ⎢ 1
𝑅1 𝐶1 ⎥ 𝑣1 state. The parameters values selected for this
𝑣̇ 2 𝑇𝑠 1 𝑇𝑠 ⎥ �𝑣2 �
𝑣̇ 3 ⎢𝐶 −
2𝐿𝐶2
𝑑1 (𝑑0 − 𝑑2 + 1) − −
𝑅2 𝐶2 2𝐿𝐶2
𝑑2 (𝑑0 + 𝑑1 ) 0 ⎥
𝑣 comparison were according to those
⎢ 2 ⎥ 3
⎢0 1 ⎥ tabulated in paper [1, Table I]. The
0 0 −
⎣ 𝑅3 𝐶3 ⎦
1 simulation was done based on ideal devices.
⎡ 0 0 0 ⎤
⎢ 𝐿 ⎥ Also, the switches were selected as four-
⎢ 1 𝑣
0 − 0 0 ⎥ 𝑖𝑛
⎢ 𝐶1 ⎥ 𝑖1 quadrant switch types.
+⎢ 𝑇 1 ⎥ 𝑖2 �


𝑠
2𝐿𝐶2
(𝑑0 + 𝑑1 ) 0 −
𝐶2
0 ⎥ 𝑖
3
The calculated results obtained from the
⎢ ⎥
⎢ 0 0 0 − ⎥
1 mathematical manipulation and state-space
⎣ 𝐶3 ⎦
equations in this paper worked together with
the ones achieved in paper [1] and ideal
simulation results were tabulated in a

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2834418, IEEE
Transactions on Power Electronics

Fig. 2 Simulation results of steady-state output voltages and inductor current according parameters tabulated
in paper [1, Table I]
Table I: Comparative values of output voltages
V1(V) V2(V) V3(V)
Values in [1] 5.6 -1.5 1.8
Mathematical 4.543 0.238 2.178
State-space Eq. 4.543 0.238 2.178
Simulation 4.567 0.241 2.14

Table II: Modified parameters of the converter presented in paper [1]


Vin V1 V2 V3 L C1 C2 C3 Op.Freq. R1 R2 R3 D0 D1 D3
(V) (V) (V) (V) (μH) (μF) (μF) (μF) (KHz) (Ω) (Ω) (Ω) - - -
2.3 5.6 -1.5 1.8 5.42 10 10 10 100 50 35 20 0.225 0.3657 0.2626

V1
d0=0.25
8 d0=0.2
d0=0.15

6
Normalized w.r.t Vin
Output Voltages

4 V3
d0=0.15
d0=0.2
2 d0=0.25
V2
d0=0.15
0 d0=0.2
d0=0.25

-2 0.35
0.2 0.3
0.25 0.25
0.3 0.2
0.35 0.15 d2
d1 0.4 0.1

Fig. 3 Operating range of the output voltages with variation in duty cycles (L=5.4μH)

comparative Table I. It could be observed output voltages with the variation in D0, D1,
that the deviation of the obtained values in and D2 according to the parameters tabulated
mathematical manipulation, simulation, and in Table II. As illustrated, the inverted output
state-space equation obtained in this work could not be achieved for all possible value
compared with the values in paper [1] were of output loads, duty cycles, and/or
not negligible. As a result, the parameter inductance value. Therefore, determining the
listed in Table I of the original paper could proper set of parameters could be one of the
not generate an inverted output at the V2 port. difficulties in designing of proposed
Therefore, the parameters were modified, and converter structure.
tabulated in Table II, to generate the desired Because of the nonlinear equation between
voltages value. Fig. 3 shows the level of the outputs and the duty cycles, deriving a

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2834418, IEEE
Transactions on Power Electronics

direct equation to calculate the inductor Start


seemed to be impossible. On the other hand,
obtaining the acceptable inductance
Specify the [Ro(min) Ro(max)] o=1,2,3
boundary could not be calculated using linear L=L0 (initial value of inductor)
Inx=0 , Step=LS
mathematical manipulation. Also, an optimal
inductance was needed to be able to keep Solve F(D)=0 for 23 possible values of loads
(using the Jacobian method)
output voltages constant under different
loads.
Using obtained equations in Section II, a op(L)=0
No Eq. (40) is
satisfied
function could be written as follows:
𝐹(𝐷 ) = 0 (38) Yes

Where, D is a vector of duty cycles [D1 D2


op(L)=1
D3] and the F is a function that returns a
vector value as:
(39) No
𝐹(𝐷 ) = [𝑉1 − 𝑉𝑜1 (𝐷), 𝑉2 − 𝑉𝑜2 (𝐷), 𝑉3 − 𝑉𝑜3 (𝐷) ] Inx=0 op(L)=op(L-LS)

Where, V1, V2 and V3 are desired output Yes

voltages and 𝑉𝑜𝑖 (𝐷), 𝑖 = 1, 2, 3 are the output Inx=Inx+1


voltages calculated in Equations (23), (14),
and (19). For a specified range of output
Yes
loads, determined by the user, the inductor Inx=100 Stop

value was increased from L0, the lowest No


bound of inductance, in Ls steps and used the
MATLAB numerical non-linear equation L=L+LS

solvers [2]. For each value of inductance, the


solver found a vector of duty cycles D. Only
values of inductance were acceptable which Fig. 4 Flowchart of the inductor design process
the corresponding D vector satisfied the could be kept constant at present of loads
following condition: variations.
𝐷𝑖 > 0 𝑓𝑜𝑟 𝑖 = 1,2,3
In addition, other parameters like the
� 𝑎𝑛𝑑 (40)
𝐷0 + 𝐷1 + 𝐷2 + 𝐷3 = 1
maximum inductor current and input current
ripple were affected by the inductor value.
The aim was to find the inductor values so The optimal inductive value helped the
that exists a vector of D that keeps the output designer to consider more operational
voltages constant under all possible value of parameters by changing the inductor around
output loads. A function was defined to show the optimal value without losing the proper
the acceptable values of inductor as follows: converter operation. Fig. 4 shows the
(41) proposed flowchart of the inductor design
1 𝑓𝑜𝑟 𝐿 𝑡ℎ𝑎𝑡 (38) 𝑎𝑛𝑑 (40) 𝑎𝑟𝑒 𝑠𝑎𝑡𝑖𝑠𝑓𝑖𝑒𝑑
𝑜𝑝(𝐿) = � process. This algorithm was run for output
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Finally, a good point for the inductor could loads range specified as:
𝑅1 ∈ [20 50]
be the value of L with the highest acceptable
𝑅2 ∈ [25 45] (42)
boundary. With this value and in close loop
𝑅3 ∈ [10 50]
structure, the controller could determine a
The obtained op(L) function is illustrated in
valid D vector so that the output voltage
Fig. 5. As shown in this figure, the best value
for the inductor could be selected as

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2834418, IEEE
Transactions on Power Electronics

III. CONCLUSION
An attempt was made to modify some
equations presented in paper [1]. For this
purpose, by using mathematical manipulation
of converter parameters, the outputs to inputs
steady state DC transfer functions have been
obtained. The state- space matrices were
recalculated and compared by mathematical
and simulation results.
A systematic process was presented to
calculate the inductor. In summary, the
Fig. 5 The op(L) function versus inductor values
obtained parameters in this work could be an
L=3.8μH which had the maximum boundary effective supplement and extension to the
at which the Equation (40) could be solved original work in paper [1].
for all possible value of output loads.
Fig. 6 shows the operating range of output IV. References
[1] P. Patra, A. Patra, and N. Misra, “A Single-
voltage and corresponding duty cycles, Inductor Multiple-Output Switcher With
keeping constant voltage, with the variation Simultaneous Buck, Boost, and Inverted
of the load currents in V1, V2, and V3. Outputs,” IEEE Transactions on Power
Electronics, vol. 27, no. 4, pp. 1936-1951, 2012.
As depicted, the constant output voltages
[2] W. Gander, and J. Hrebicek, Solving problems
could be achieved based on inductance in scientific computing using Maple and
L=3.8μH and calculated duty cycles under Matlab®: Springer Science & Business Media.
different loads.

Load current in V2 (R1=50, R3=20, L=3.8μH) Load current in V1 (R2=35, R3=20, L=3.8μH)

Load current in V3 (R1=50, R2=35, L=3.8μH)


Fig. 6 Operating range of the output voltages with variation in load currents

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

Potrebbero piacerti anche