Three-Phase Flyback Push–Pull DC–DC Converter: Analysis ... em periódicos...This paper proposes the three-phase ﬂyback push–pull con-verter, which is a three-phase version

IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 28, NO. 4, APRIL 2013 1961

Three-Phase Flyback Push–Pull DC–DC Converter:Analysis, Design, and Experimentation

Hugo Rolando Estofanero Larico, Member, IEEE, and Ivo Barbi, Fellow, IEEE

Abstract—In this paper a step-up/step-down isolated dc–dc con-verter referred to as a three-phase flyback push–pull dc–dc con-verter is presented. The power circuit is constituted by a pair ofcoupled inductors, a three-phase transformer, a capacitor, threeswitching transistors and three power diodes. The proposed con-verter offers the advantages of compact passive devices, lowconduction power losses, full duty cycle range (0–100%), and inher-ent protection against transformer saturation. Furthermore, filtersizes are minimized to duty cycles of around 1/3 and 2/3. Thesecharacteristics makes this converter suitable for many applications,especially in low-voltage high-power applications such as telecom-munications power supply, battery chargers, and renewable powersystems. The operating principle and the idealized mathematicalanalysis in continuous conduction mode are presented. Experimen-tal data were obtained from a laboratory prototype with an inputvoltage of 125 V, output voltage of 100 V, load power of 1000 W, andswitching frequency of 42 kHz. The measured prototype efficiencywas 94% for full load and 96% for 400 W.

Index Terms—DC–DC converter, flyback converter, push–pullconverter, three-phase dc–dc converter.

I. INTRODUCTION

THE voltage-fed and current-fed push–pull converters areclassical isolated dc–dc converters used widely in low volt-

age applications due to their simplicity and good transformercore utilization since with these converters two quadrants ofthe BH curve are used [1]–[7]. However, only the current-fedconfiguration is able to drive large current rate without causingtransformer saturation due to an unequal volt-second product.Also, these converters are not able to operate in the full dutycycle range, because this would lead to a short-circuit in thevoltage-fed mode and an open-circuit in the current-fed mode.

The Weinberg and flyback-current-fed push–pull convertersare also push–pull topologies and are able to operate in thefull duty cycle range (0 to 100%) [8]–[16]. This characteristicis achieved by the use of a pair of coupled inductors. However,only the flyback-current-fed push–pull converter shown in Fig. 1uses effectively the coupled inductors, since its windings drivethe energy for the whole duty cycle range, while the secondary

Manuscript received September 22, 2011; revised January 2, 2012 and March12, 2012; accepted July 13, 2012. Date of current version October 26, 2012.Recommended for publication by Associate Editor T. Sebastian.

H. R. E. Larico is with the Mobility Engineering Center (CEM),Federal University of Santa Catarina, Joinville 89218-000, Brazil (e-mail:[email protected]).

I. Barbi is with the Department of Electrical Engineering, FederalUniversity of Santa Catarina, Florianopolis 88040-970, Brazil (e-mail:[email protected]).

Digital Object Identifier 10.1109/TPEL.2012.2211037

(a)

(b)

Fig. 1. (a) Flyback-current-fed push–pull converter and (b) proposed three-phase flyback push–pull converter.

winding of other such converters drives the energy only whenthe duty cycle is less than 50%. Moreover, this converter doesnot increase the number of semiconductor devices required. Inaddition, both of these converters provide continuous constantcurrent shape at the input and output when operated with aconstant duty cycle of 1/2. Consequently, operation around thispoint provides smaller filter sizes.

This paper proposes the three-phase flyback push–pull con-verter, which is a three-phase version of the flyback-current-fedpush–pull dc–dc converter [13] as are shown in Fig. 1. The pro-posed converter is constituted by three switching transistors (S1,S2, S3), three diodes (D1, D2, D3), a capacitor C, a pair ofcoupled inductors Lf , and a high-frequency three-phase trans-former T in a wye-wye connection. The resulting characteristicsof the three-phase version are as follows:

1) operation in full duty cycle range;2) absence of the transformer saturation problem, since the

high-source impedance provided by the coupled inductorsavoids transformer saturation due to the unequal volt–second product applied to its windings;

0885-8993/$31.00 © 2012 IEEE

1962 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 28, NO. 4, APRIL 2013

v Tp1

-

+

vTp2-

+

vTp3

-

+ v Ts1

-

+

vTs2

-

+

vTs3-

+

iTp1 iTp3

iTp2

iTs1 iTs3

iTs2

NT:1T

(a)

vTp1

iTp1 iTp2 iTp3

vTp2 vTp3+

-

+

-

+

-

iTs1 iTs2 iTs3

vTs1+

-vTs2

+

-vTs3

+

-

(b)

Fig. 2. The three-phase high-frequency transformer: (a) schematic and(b) physical representations.

3) continuous constant current shapes at the input and theoutput for set duty cycles of 1/3 and 2/3 with consequentlysmaller filter sizes;

4) the transistors are connected to the same reference point;thus, these devices can be driven by nonisolated circuitdrivers;

5) low conduction losses since there is only a single semicon-ductor device in series with the input and output sources;

6) small filter sizes, the three-phase system increases theoperating frequency of filters to three times the switchingfrequency;

7) a good transformer copper and core utilization since thetransformer windings are placed in a single common coreas is shown in Fig. 2;

8) high-power processing, the sharing of the input and outputcurrent improves the heat transfer of the power devicesallowing high current rates.

The characteristics listed previously show that the proposedconverter merges the main push–pull converter characteristicswith those of the three-phase converters [17]–[26]. The resultis a converter with smaller filter sizes able to drive high currentrates at low voltage rates.

This study presents an idealized theoretical analysis and ex-perimental results for the proposed converter operating in con-tinuous conduction mode (CCM). Also, a comparison with theconventional and push–pull types of flyback converters high-lights the advantages provided by the proposed converter.

II. PRINCIPLE OF OPERATION

The following assumptions are made to simplify the theoret-ical description of the operating principle:

1) the converter is operating in steady state;2) the magnetic flux of the coupled inductors is continuous,

this means that the converter is operating in CCM;

TABLE IOPERATION REGIONS

3) all switching devices and the three-phase transformer areideals, so these power devices do not cause power lossesand the transformer magnetizing inductance is infinite;

4) the capacitance of C is large enough to consider a constantoutput voltage;

5) the windings of the coupled inductors are tightly coupledand have negligible resistance;

6) the magnetizing inductance Lm of the coupled inductorsLf is referred to its primary side and has a finite valuesince this device stores energy;

7) the power transistors are symmetrically driven using thethree-phase gating signals as is shown in Fig. 3;

8) the turns ratios of the coupled inductors and the three-phase transformer are considered equal (NL = NT ). Thisequality is assumed since it provides a single static gainin CCM for all duty cycle range.

The output voltage Vo is regulated employing pulse-widthmodulation (PWM) to control the duty cycle D, which is de-fined as the ratio between the transistor on time ton and theswitching period Ts . The transistors can be driven using a dutycycle range from 0% to 100%, that is, the proposed converteradmits overlapping and nonoverlapping conduction of transis-tors. Thus, the converter operation is divided into three regionslisted in Table I. In region R1 the transistors are not overlapped,while in regions R2 and R3 overlapping of the conduction oftwo and three transistors, respectively, occurs.

A. Operating Principle in R1

In this region, the energy supplied by the input source isdiscontinuous and it occurs during the transistor conductionperiods. On the other hand, the capacitor and load arrangementis continuously supplied with energy from the circuit. Therefore,the input and output currents have discontinuous and continuousshapes, respectively. The main waveforms in CCM for regionR1 are shown in Fig. 3.

The first stage starts at instant t0 when S1 is turned ON, caus-ing D1 to be reverse biased. Diodes D2 and D3 remain conduct-ing. The equivalent circuit for this stage is shown in Fig. 4(a).Ideally, the magnetomotive forces (mmf) of the three-phasetransformer are equal; thus, the currents flowing through S1,D2 and D3 have the relation NT iS1 = iD2 = iD3 . Therefore,one-third of the magnetizing current of the coupled inductorsflows through the primary winding (iLp = im /3), while two-thirds are conducted through the secondary winding (iLs =2NT im /3). The voltages across the active primary windings oftransformer and coupled inductors are functions of the input andoutput sources and are given as vT p1 = 2(Ei + NT Vo)/3 andvLp = (Ei − 2NT Vo)/3, respectively. In this stage, the coupled

LARICO AND BARBI: THREE-PHASE FLYBACK PUSH–PULL DC–DC CONVERTER: ANALYSIS, DESIGN, AND EXPERIMENTATION 1963

Fig. 3. Main waveforms of the three-phase flyback push–pull converter inregion R1 in CCM.

inductors store part of the energy supplied by the input sourceand, consequently, the input voltage is greater than 2NT Vo . Thevoltage applied across transistors S2 and S3 is 3/2 times the ac-tive primary transformer winding (VS2 = VS3 = Ei + NT Vo ).This voltage referred to the secondary side is applied acrossdiode D1 (VD1 = Ei/NT + Vo ).

The second stage starts at instant t1 when S1 is turned OFF.All diodes are forward biased; therefore, the voltages across thetransformer windings are zero. Thus, the stored energy of thecoupled inductors is directly transferred to the output throughits secondary winding (vLs = −Vo ) as shown in Fig. 4(b). Themagnetizing current referred to the secondary side is evenlyshared among the diodes (iD1 = iD2 = iD3 = NT im /3). Thevoltage across any transistor is Ei + NT Vo .

In the third and fifth stages, the input source again suppliesenergy to the circuit, part of this energy is stored by the cou-pled inductors, which is transferred to the output in the fourthand sixth stages, respectively. Fig. 3 shows that the capacitorbehavior is the opposite to that of the coupled inductors, that is,it charges in the even stages and discharges in the odd stages.

B. Operating Principle in R2

In this region, the energy flowing through the input and outputare continuous. Therefore, the input and output currents arecontinuous but both currents have pulsating shapes. The mainwaveforms in CCM for region R2 are shown in Fig. 5.

The first stage starts at instant t0 when S1 is turned ONwhile S3 is conducting. Only diode D2 is forward biased. The

(a)

(b)

Fig. 4. Equivalent circuit for (a) first and (b) second stages in R1.

equivalent circuit for this stage is shown in Fig. 6. The relationbetween the currents flowing through active power devices isNT iS1 = NT iS3 = iD2 . Thus, two-thirds of the magnetizingcurrent of coupled inductors flow through the primary wind-ing (iLp = 2im /3) and one-third is conducted by the secondarywindings (iLs = NT im /3). The voltages across the active pri-mary windings of the transformer and coupled inductors arevT p1 = (Ei + NT Vo)/3 and vLp = (2Ei − NT Vo)/3, respec-tively. In this stage, the input voltage is higher than NT Vo/2,but it subsequently becomes lower than 2NT Vo . Therefore, thecoupled inductors store part of the energy supplied by the in-put source. The voltage applied across transistor S2 is threetimes the voltage across active primary transformer windings(VS2 = Ei + NT Vo ). This voltage referred to the secondaryside is applied through diodes D1 and D3 (VD1 = VD3 = Ei/NT + Vo ).

The second, fourth, and sixth stages are similar to the evenstages of region R1 with the difference that now the coupledinductors transfer their stored energy to the output since theinput voltage is lower than 2NtVo . In the third and fifth stages,the coupled inductors store again part of the energy supplied bythe input source. Fig. 5 shows that the capacitor has the samebehavior as that of R1, that is, it charges in the even stages anddischarges in the odd stages.



Fig. 6. Equivalent circuit for the first stage in R2.

C. Operating Principle in R3

In this region, the input source supplies energy to the circuitthroughout the switching cycle, while the output receives energyfrom the circuit if only two transistors are turned ON. There-fore, the input current is continuous while the output current isdiscontinuous. The main waveforms for a switching period inR3 are shown in Fig. 7.

The first stage starts at instant t0 when S1 is turned ONwhile S2 and S3 are conducting. All diodes are reverse bi-ased. Therefore, the voltages across the transformer windingsare zero as shown in Fig. 8. Thus, the input source transfers en-ergy directly to the coupled inductors through its primary wind-


Fig. 8. Equivalent circuit for the first stage in R3.

ing (vLp = Ei). The voltage across any diode is Ei/NT + Vo .Whole magnetizing current of coupled inductors flows throughthe primary side and is evenly distributed through S1, S2, andS3 (iS1 = iS2 = iS3 = ii/3). This storage stage occurs againin the third and fifth stages.

The second, fourth, and sixth stages are similar to the evenstages of region R3 with the difference that now the coupledinductors transfer their stored energy since the voltage NT V o/2becomes higher than the input voltage. Fig. 7 shows that in theeven stages, the energy consumed by the load is supplied by thecapacitor that is recharged in the odd stages.


III. MATHEMATICAL ANALYSIS

In this section, the main equations for CCM that allow thecurrent and voltage stresses of the proposed three-phase flybackpush–pull dc–dc converter to be computed are developed. Thequantitative information contained in Figs. 3, 5, and 7 is usedfor this purpose.

The input and output currents of the proposed converter suchas the flyback converter have pulsating shapes. Consequently,the rms currents in the power devices as well as the output volt-age ripple in CCM are largely dependent on the average value ofthe magnetizing current. Thus, the expressions to compute theseelectrical values are developed assuming that the magnetizinginductance of coupled inductors is large enough for its currentripple to be neglected.

A. Static Gain

If the proposed converter is operating in steady state then theaverage voltage across primary winding of coupled inductorsshould be zero for a switching cycle. Therefore, consideringthe symmetrical operation, this is also true for one-third of aswitching period

VLp =1Ts

∫ t0 +Ts

t0

vLpdt =3Ts

∫ t2

t0

vLpdt = 0. (1)

The substitution of voltages and times that correspond toregions R1, R2, and R3 yields

(Ei − 2NT Vo) ton − NT Vo

(Ts

3− ton

)= 0 (2)

(2Ei − NT Vo)(

ton − Ts

3

)

+ (Ei − 2NT Vo)(

2Ts

3− ton

)= 0 (3)

Ei

(ton − 2Ts

3

)+ (2Ei − NT Vo) (Ts − ton) = 0. (4)

Solving these equations, (5) is obtained which represents thestatic gain of the proposed converter in CCM. The result showsthat for the converter, there is a single expression for the wholeduty cycle range and that it is the same as that of the conventionalflyback converter

Vo

Ei=

1NT

D

1 − D. (5)

B. Magnetizing Current Ripple of Coupled Inductors

The magnetizing current ripple can be determined in thestored stage or transferred stage. Considering the former, wehave the following:

ΔIm =1

Lm

∫ t1

to

vLpdt. (6)

Substituting the voltages and times for each region results in(7), (8), and (9), which correspond to regions R1, R2, and R3,

respectively

ΔIm =(Ei − 2NT Vo)ton

3Lm(7)

ΔIm =(2Ei − NT Vo)

3Lm

(ton − Ts

3

)(8)

ΔIm =Ei

3Lm

(ton − 2Ts

3

). (9)

Using (5) in the aforementioned equations results in the per-unit magnetizing current ripple given by

ΔIm =ΔIm Lm

VoTs=

⎧⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎩

N1 − 3D

3for R1

N(2 − 3D)(3D − 1)

9Dfor R2

N(1 − D)(3D − 2)

3Dfor R3.

(10)

C. RMS Currents of Coupled Inductors

The primary and secondary rms currents of coupled inductorsare computed through

ILprms =(

3Ts

∫ t2

t0

(iLp)2dt

)1/2

(11)

ILsrms =(

3Ts

∫ t2

t0

(iLs)2dt

)1/2

. (12)

The average values for the magnetizing current and the outputcurrent of any region are related as follows:

Im =Io

NT (1 − D). (13)

Neglecting the magnetizing current ripple of the currentshapes shown in Figs. 3, 5, and 7, and using the relationshipgiven by (13) the per-unit primary and secondary rms currentsof the coupled inductors are obtained as (14) and (15)

ILprms =ILprms

Io=

⎧⎪⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎪⎩

1(1 − D)NT

√D

3for R1

13(1 − D)NT

√9D − 2 for R2

1(1 − D)NT

√5D − 2

3for R3

(14)

ILsrms =ILsrms

Io=

⎧⎪⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎪⎩

11 − D

√3 − 5D

3for R1

13(1 − D)

√7 − 9D for R2

11 − D

√1 − D

3for R3.

(15)

D. RMS Currents of Three-Phase Transformer

The sharing of the input and output currents by the proposedconverter provides constant current amplitudes through the pri-mary and secondary transformer windings for whole duty cyclerange as shown in Figs. 3, 5, and 7, which have instantaneous


TABLE IICOMPARISON OF CONVERTERS

values of im /3 and NT im /3, respectively. Therefore, the rmsvalues of these currents, neglecting the magnetizing current rip-ple, are a function of the conduction time of the semiconductordevices in series with these windings. Thus, the primary andsecondary rms currents of the three-phase transformer are com-puted through

IT prms =Im

3

√D (16)

IT srms =Im NT

3

√1 − D. (17)

Substituting (13) in the aforementioned equations yields (18)and (19), which represent the per-unit rms currents through the

primary and secondary windings, respectively

IT prms =ILprms

Io=

√D

3(1 − D)NT(18)

IT srms =IT srms

Io=

√1 − D

3(1 − D)(19)

E. Output Voltage Ripple

The output voltage ripple neglecting the magnetizing currentripple is given by

ΔVo =1C

∫ t2

t1

(iLs − Io)dt. (20)


f

D0.2 0.4 0.6 0.8

0.1

0.2

0.3

0.4

ConventionalPush-pullProposed

(a)

D0.2 0.4 0.6 0.8

0.1

0.2

0.3

0.4


(b)

Fig. 9. (a) Per-unit magnetizing current ripple of Lf , and (b) per-unit outputvoltage ripple in CCM for NT =1.

From Figs. 3, 5, and 7, the voltage ripple for regions R1, R2,and R3 yields

ΔVo =1C

(Im NT − Io)(

Ts

3− ton

)(21)

ΔVo =1C

(2Im NT

3− Io

)(2Ts

3− ton

)(22)

ΔVo =1C

(Im NT

3− Io

)(Ts − ton) . (23)

The substitution of (5) into (21), (22), and (23) yields (24),which is the per-unit output voltage ripple in CCM

ΔVo =ΔVoC

IoTs=

⎧⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎩

D(1 − 3D)3(1 − D)

for R1

(3D − 1)(2 − 3D)9(1 − D)

for R2

(3D − 2)3

for R3.

(24)

D0.2 0.4 0.6 0.8

0.5

1.0

1.5

2.0


(a)

Lf

D0.2 0.4 0.6 0.8

1.2

1.4

1.6

1.8


1.0

(b)

Fig. 10. (a) Per-unit rms capacitor current, and (b) per-unit rms secondaryinductor currents in CCM.

TABLE IIISPECIFICATIONS OF PROTOTYPE

F. RMS Capacitor Current

If negligible output voltage ripple is assumed, then the rmscurrent of the capacitor can be calculated as follows:

IC rms2 = ILsrms

2 − Io2 . (25)

Substituting (15) in (25), the per-unit capacitor rms current isgiven by the following:

IC rms =IC rms

Io=

⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩

√D(1 − 3D)3(1 − D)2 for R1

√(3D − 1)(2 − 3D)

9(1 − D)2 for R2

√3D − 2

3(1 − D)for R3.

(26)


LT

GND2

C21000 F250V

C11000 F250V

Plug 2

D212A600V

D312A600V

D112A600V

S120A/600V

R812k

D1215V

R5/6.8

T3BD330

T6BD329

D6MUR880E8A/800V

S120A/600V

R712k

D1115V

R4/6.8

T2BD330

T5BD329

D5MUR880E8A/800V

S120A/600V

R612k

D1015V

R3/6.8

T1BD330

T4BD329

C3

630V0.33 F

GND1

Plug 1 R112k

DSPTMS320LF2407

12345678

91011121314

+5VDC

GND1

+15VDC

Plug 3

R11820

R10820

R9820

SN74LS07N

D4MUR880E8A/800V

D7

C4

630V0.33 F

R233k

D8

D9

D7-D9MUR880E8A/800V

Fig. 11. Schematic circuit of the implemented laboratory prototype.

G. Voltage Stresses Across Semiconductor Devices

The voltages across turned-off transistor and reverse-biaseddiode for any region such as in the case of the conventionalflyback converter is the sum of the output and input voltagesreferred to the primary or secondary side, respectively

VS = Ei + NT Vo (27)

VD =Ei

NT+ Vo. (28)

In CCM, these voltages can be expressed as

VS = EiD

1 − D(29)

VD =Ei

NT

D

1 − D. (30)

IV. COMPARISON OF CONVERTERS

In this section, the proposed converter is compared with theflyback-current-fed push–pull and conventional flyback convert-ers, as shown in Table II. It can be seen that the static gain andsemiconductor voltage stresses are the same for all converters,while the current and voltage ripples, and the rms currents differfor each converter. The comparison reveals that the current andvoltage ripples of the proposed converter are 1/3 lower than thoseof the conventional converter, and 3/2 lower than the values re-quired for the push–pull type as shown in Fig. 9. Moreover, therms currents through the capacitor and the secondary inductorwindings of the proposed converter are lower than those valuesfor the other ones as can be seen in Fig. 10. Consequently, theproposed converter allows a greater reduction in filter sizes thanthe other converters.

The comparison of the rms transformer currents, which are thesame rms currents across the switching transistors and diodes,shows that the rms currents of the proposed converter are 1/3lower than those of the conventional converter and 3/2 lower than

TABLE IVSPECIFICATIONS OF POWER DEVICES

those of the push–pull type. Therefore, the proposed converteris able to drive high-current rates with lower current stressesacross semiconductor devices.

In addition, the proposed converter presents two particularpoints, one-third and two-thirds of the duty cycle, where thecurrent ripple as well as the voltage ripple and rms current ofthe capacitor fall to zero. Therefore, for these points, the coupledinductors and output capacitor sizes are minimized.

V. EXPERIMENTAL RESULT

The validation of theoretical analysis was carried out throughan experimental prototype designed with the specificationsshown in Table III. The schematic circuit of the laboratory pro-totype is shown in Fig. 11, the power components of which arelisted in Table IV. The gate drive signals were generated byKit DSP TMS320LF2407. Nonisolated drive circuits were em-ployed to drive the switching transistors. The photo in Fig. 12shows the laboratory prototype, where the passive power de-vices and main parts of the circuit can be identified. It can beobserved that the windings of the three-phase high-frequencytransformer were placed in a single common core.

The experimental current waveforms obtained from the ex-perimental prototype operating in regions R1, R2, and R3 are


Fig. 12. Photograph of the implemented laboratory prototype.

1

4

2

3

V: [10A/div]H: [4 s div

1-Output current

3-Input current

4-Switch current S1

2-Diode current D1

Fig. 13. Experimental current waveforms for D = 0.289, Ei = 150 V,Vo = 100 V, and Po = 980 W.

shown in Figs. 13, 14, and 15, respectively. The results showthat the input and output currents have pulsating shapes in any ofthe regions. The currents at the input and output sides differ foreach region, in region R1 the input current is discontinuous andthe output current is continuous, in region R2 both of currentsare continuous, while in region R3 the input current becomescontinuous and output current assumes a discontinuous flow.In all regions, the input and output currents are evenly sharedthrough the transistors and diodes, and their current amplitudesacross these devices have constant values.

The converter efficiency curves shown in Fig. 16 were ob-tained for regions R1 and R2 with input voltages of 125 and

1

4

2

3

V: [10A/div]H: [4 s div

1-Output current

3-Input current

4-Switch current S1

2-Diode current D1


1

4

2

3

Time [4 s div

1-Output current [10A/div]

3-Input current [5A/div]

4-Switch current S1 [5A/div]

2-Diode current D1 [10A/div]


0 200 400 600 80090

92

94

96

98

η

Po [W]

[%] R1: =125-VDCEi

R2: =85-VDCEi

Fig. 16. Measured efficiency of the laboratory prototype in regions R1 andR2, both for output voltage of Vo = 100 V.

85 V, respectively, and in both cases the output voltage wasmaintained at a constant value of 100 V. The efficiency curvefor R3 is not shown since the laboratory prototype is not de-signed to operate in this region, so the excessive power lossesat low input voltage result in a poor efficiency. The efficiencyfor an output power of 980 W was 94% in region R1 and 93%in region R2. The maximum efficiency in both regions was ap-proximately 96% which occurred at an output power of 400 W.


VI. CONCLUSION

In this paper, a three-phase dc–dc isolated converter, referredto as a three-phase flyback push–pull converter, is introduced.Theoretical analysis of the converter is presented and experi-mental results are used for its validation. The proposed convertermerges the main characteristics of push–pull converters withthose of three-phase dc–dc converters, resulting in a converterwith low power losses, small filter sizes, compact transformer,and high-power processing.

In addition, the operation of the proposed converter in re-gion R2 provides continuous power flow from the input sourceto the ouput load that allows filters sizes to be reduced. Theprototype efficiencies obtained for regions R1 and R2 showthat the proposed converter has good efficiency, considering thepower losses in dissipative overvoltage clamp circuits. There-fore, the proposed converter is suitable for many applications,particularly for low-voltage high-power applications such astelecommunications power supply, battery chargers, and renew-able power systems.

REFERENCES

[1] M. Shoyana, and K. Harada, “Steady-state characteristics of the push-pullDC-to-DC converter,” IEEE Trans. Aerosp. Electron. Syst., vol. AES-20,no. 1, pp. 50–56, Jan. 1984, 2013.

[2] E. R. Hnatek, Design of Solid State Power Supply, 3rd ed. New York: VanNostrand Reinhold, 1989, pp. 116–117.

[3] H. W. Whittington, B. W. Flynn, and D. E. Macpherson, Switched ModePower Supplies, 2nd ed. New York: Wiley, 1992, pp. 32–33.

[4] Q. Li, and P. Wolfs, “A review of the single phase photovoltaic module in-tegrated converter topologies with three different DC link configurations,”IEEE Transactions on Power Electronics, vol. 23, no. 3, pp. 1320–1333,May 2008.

[5] J.-M. Kwon, E.-H. Kim, B.-H. Kwon, and K.-H. Nam, “High-efficiencyfuel cell power conditioning system with input current ripple reduction,”IEEE Trans. Ind. Electron., vol. 56, no. 3, pp. 826–834, Mar. 2009.

[6] M. Mohr, W. T. Franke, B. Wittig, and F. W. Fuchs, “Converter systemsfor fuel cells in the medium power range—A comparative study,” IEEETrans. Ind. Electron., vol. 57, no. 6, pp. 2024–2032, Jun. 2010.

[7] G. Ortiz, D. Bortiz, J. Biela, and J. W. kolar, “Optimal design of a 3.5 kV/11 kW DC-DC converter for charging capacitor banks of power mod-ulators,” IEEE Trans. Plasma Sci., vol. 38, no. 10, pp. 2565–2573,Oct. 2010.

[8] A. H. Weinberg, “A boost regulator with a new energy-transfer principle,”in Proc. Spacecraft Power Condition. Electron. Semin., Eur. Space Res.Org., Sep. 1974, pp. 115–122.

[9] P. W. Clarke, “Converter regulation by controlled conduction overlap,”U.S. Patent 3 938 204, Feb. 10, 1976 (assigned to Bell Telephone Labo-ratories, Inc., Murray Hill, NJ 07974, USA).

[10] V. J. Thottuvelil, T. G. Wilson, and H. A. Owen, Jr., “Analysis and designof a push-pull current-fed converter,” in Proc. Power Electron. Spec. Conf.1981, pp. 192–203.

[11] R. Redl, and N. O. Sokal, “Push-pull current-fed multiple-output regulatedwide-input-range DC/DC power converter with only one inductor and with0 to 100% switch duty ratio: operation at duty ratio below 50%,” in Proc.Power Electron. Spec. Conf. 1981, pp. 204–212.

[12] S. K. Mazumder, K. Acharya, and C. M. Tan, “Design of an all-SiC parallelDC/DC Weinberg converter unit using RF control,” IEEE Trans. PowerElectron., vol. 23, no. 6, pp. 2893–2904, Nov. 2008.

[13] D. A. Ruiz-Caballero, and I. Barbi, “A new flyback-current-fed push-pull DC-DC converter,” IEEE Trans. Power Electron., vol. 14, no. 6,pp. 1056–1064, Nov. 1999.

[14] J. Zhang, X. Huang, X. Wu, and Z. Qian, “A high efficiency flybackconverter with new active clamp technique,” IEEE Trans. Power Electron.,vol. 25, no. 7, pp. 1775–1785, Jul. 2010.

[15] Y. Chuan, R. Xinbo, W. Xuehua, and C. K. Tse, “Isolated buck-boostDC/DC converters suitable for wide input-voltage range,” IEEE Trans.Power Electron., vol. 26, no. 9, pp. 2599–2613, Sep. 2011.

[16] Y.-C. Hsieh, M.-R. Chen, and H.-L. Cheng, “An interleaved flyback con-verter featured with zero-voltage transition,” IEEE Trans. Power Electron.,vol. 26, no. 1, pp. 79–84, Jan. 2011.

[17] R. W. De Doncker, D. M. Divan, and M. H. Kheraluwala, “A three-phase soft-switched high-power-density dc/dc converter for high-powerapplications,” IEEE Trans. Ind. Appl., vol. 27, no. 1, pp. 63–73, Jan./Feb.1991.

[18] P. D. Ziogas, A. R. Prasad, and S. Manias, “Analysis and design of a threephase off-line dc/dc converter with high frequency isolation,” IEEE Trans.Ind. Appl., vol. 28, no. 4, pp. 824–832, Jul./Aug. 1992.

[19] D. S. Oliveira, Jr. and I. Barbi, “A three-phase ZVS PWM DC/DC con-verter with asymmetrical duty cycle for high power applications,” IEEETrans. Power Electron., vol. 20, no. 2, pp. 370–377, Mar. 2005.

[20] C. Liu, A. Johnson, and J.-S. Lai, “A novel three-phase high-power soft-switched dc/dc converter for low-voltage fuell cell applications,” IEEETrans. Ind. Appl., vol. 41, no. 6, pp. 1691–1697, Nov./Dec. 2005.

[21] G. Franceschini, E. Lorenzani, M. Cavatorta, and A. Bellini, “3boost:A high-power three-phase step-up full-bridge converter for automotiveapplications,” IEEE Trans. Ind. Electron., vol. 55, no. 1, pp. 173–183, Jan.2008.

[22] H. Cha, J. Choi, and P. N. Enjeti, “A three-phase current-fed DC/DCconverter with active clamp for low-DC renewable energy sources,” IEEETrans. Power Electron., vol. 23, no. 6, pp. 2784–2793, Nov. 2008.

[23] R. L. Andersen and I. Barbi, “A three-phase current-fed push-pull dc-dcconverter,” IEEE Trans. Power Electron., vol. 24, no. 2, pp. 358–368, Feb.2009.

[24] E. Agostini and I. Barbi, “Three-phase three-level pwm dc-dc converter,”IEEE Trans. Power Electron., vol. 26, no. 7, pp. 1847–1856, Jul. 2011.

[25] S. Lee, F. Park, and S. Choi, “A three-phase current-fed pushpull dc-dcconverter with active clamp for fuel cell applications,” IEEE Trans. PowerElectron., vol. 26, no. 8, pp. 2266–2277, Aug. 2011.

[26] S. V. G. Oliveira and I. Barbi, “A three-phase step-up dc-dc converter witha three-phase high-frequency transformer for dc renewable power sourceapplications,” IEEE Trans. Ind. Electron., vol. 58, no. 8, pp. 3567–3580,Aug. 2011.

Hugo Rolando Estofanero Larico (M’12) wasborn in Arequipa, Arequipa, Peru, in 1980. He re-ceived the B.S. degree in electrical engineering fromthe National University of San Agustin (UNSA),Arequipa, Peru, in 2004, and the M.S. and Ph.D.degrees in electrical engineering from the FederalUniversity of Santa Catarina (UFSC), Florianopolis,Brazil, in 2007 and 2011, respectively.

He is currently a Professor in the Mobility Engi-neering Center, UFSC. His research interests includedesign, modeling, control, and simulation of static

power conversion systems.

Ivo Barbi (M’76–SM’90–F’11) was born inGaspar, Santa Catarina, Brazil, in 1949. He receivedthe B.S. and M.S. degrees in electrical engineeringfrom Federal University of Santa Catarina (UFSC),Florianopolis, Brazil, in 1973 and 1976, respectively,and the Ph.D. degree from the Institut National Poly-technique de Toulouse, France, in 1979.

He is currently a Professor at the Power Electron-ics Institute, UFSC. He founded the Brazilian PowerElectronics Society and the Power Electronics Insti-tute of the Federal University of Santa Catarina.

Three-Phase Flyback Push–Pull DC–DC Converter: Analysis ... em periódicos...This paper proposes the three-phase ﬂyback push–pull con-verter, which is a three-phase version

Documents