Design. Implementation and Experimental Analysis of Two-Stage Boost Converter for Grid Connected Photovoltaic System

Muhammad Aamir School of Electrical and Computer Science Engineering

Hanyang University, ERICA Campus Ansan, South Korea.

m _ aamir80 l@hotmail.com

Abstract-This paper proposes a two stage boost converter

circuit for the grid connected photovoltaic system. Each stage

is controlled separately by opposite pulse width modulated

signal. Due to nonlinear I-V characteristic of the photovoltaic

cells, the converter circuit is designed for the range of input

voltage 35 to 60 V, 350V output voltage and maximum output

power of 525 watt. Dynamic modeling of the converter is done

using state space averaging technique and efficient voltage

control topology is suggested. Also operation, principle and

performance characteristics of the proposed circuit are

discussed and simulated results are validated through practical

experiments.

Keywords-boost converter; photovoltaic system; Dynamic

modeling;

I. INTRODUCTION

For many years, fissile fuel has been the primary source

of energy. However, there is a limited supply of these fuels

on Earth and they are being used much more rapidly than

they are being created. Eventually, they will run out. The

world trend nowadays is to find a non-depletable and clean

source of energy. The most effective and harmless energy

source is probably solar energy. Solar energy is considered

one of the most promising energy source due to its infinite

power. Due to the high cost of the photo voltaic modules, the

focus is on the method to get the maximum energy from the

Photovoltaic (PV) system. One of the interesting applications

of photovoltaic systems is grid connected photovoltaic

978-1-4244-5540-9/10/$26.00 2010 IEEE

194

Mahmood Y ounas Shinwari School of Electrical and Computer Science Engineering

Hanyang University, ERICA Campus Ansan, South Korea

mahmood _ shinwari@hotmail.com

systems [1]. Photovoltaic systems generate electricity

especially in summer times when the sunlight is available for

quiet long time and grids are often loaded additionally by air

conditioning and other cooling systems. One of the required

features of a grid connected photovoltaic converter is the

ability to get maximum power from the photovoltaic arrays

[2]. Therefore, the maximum power point tracking is

required, as the power obtained from the photovoltaic system

depends on the environmental conditions such as the

intensity of the light [3]. Since the output voltage of the

photovoltaic array is considerably low and we require a very

high voltage level for the system to connect it to the grid, so

a single stage boost converter cannot achieve such a high

transformation ratio. More importantly the duty ratio of the

converter has inverse relation with the efficiency (TJ ) of the converter circuit [4].

1 (1) TJ

= 1+ RL (1-D)2RZoad

Where RL is the internal resistance of the inductor.

Thus considering (1) we cannot increase the duty ratio (D)

from a certain maximum limit. Now for certain high step-up

applications, coupled-inductors converters such as Flyback

and isolation Sepic converters have been proposed. But these

converters have problems of high voltage stress due to

leakage inductance of the coupled inductors and also

degraded efficiency [5]. Therefore, the bidirectional two

stage converter in cascade is proposed with high step-up

voltage ratio and maximum efficiency. The bidirectional two

stage converter has been proposed in [7]. In that proposed

circuit, switch of each stage is provided with pulse width

modulated (PWM) signal at the same time and same duty

ratio. More importantly only simulated results of the

proposed model is presented without explanation of the

circuit operation. This paper deals with the design and

operation of the two stage boost converter with opposite duty

ratio and more practical dimensioning of the components

depending on the statistics of each stage. The operation of

the circuit with this new methodology is explained; in

addition it is verified by the simulation results and practical

experiments. More simple and innovative feedback control

technique for each stage is also presented in the paper.

v.,

Stage 1 stage 2 ------ -, r - - - - - - - - -

,,L1 01 I ,,"______t_--c:-r. I II ! 1 : : '\ " T : : "j C2 T : I J J ' I ] 1 i i i I L _ _ _ _ _ _ _ _ 1 _ _ _ _ _ _ __ J

Figure 1. Two stage boost converter basic circuit design

Voul

II. BASIC OPERATIONAL PRINCIPAL OF THE Nov AL

CIRCUIT

The basic circuit of the proposed converter is shown in

Fig 1. It is assumed that the switches are ideal. The input

voltage is instantaneous and constant and load is pure

resistive. Each stage of the converter is assumed to be

operated in a continuous conduction mode. The voltage

conversion ratio of boost converter is given by (2)

1 M(D) = -( -) 1-D

(2)

Since there are two stages of boost converter connected

together so the voltage conversion ratio for the whole circuit

will be given equation (3)

195

M(D) - (_1 ) (_1 ) 1-D1 1-Dz (3)

Now depending on the input voltage range and output

voltage for the particular application the duty ratio of both

the stages has been derived. The stages of operation have

been described in four steps. In this analysis some

consideration has been made i.e the active and passive

elements of the circuit are ideal.

First step: This step begins when the switch S 1 turns on.

During this period of operation, S2 will remain off. The

diode is reserve bias, thus isolating the right hand side of the

circuit which is now only connected to the ground as shown

in fig.2 (b). The input supplies energy to the inductor Ll.

Thus the inductor will store energy. This step finishes when

S 1 is turned off.

Second Step: As SI turns off, the output of stage

receive energy from the inductor as well as from the PV

array as shown in the fig 2( c). Thus it will give step up

output voltage. The output filter capacitor is assumed to be

very large to ensure constant output voltage. Vo (t) - Vo

Third Step: This step starts in parallel operation with

step second. As SI turns off, S2 turns on. The diode D2 is

reverse bias, thus isolating the output portion of second stage.

There is a point to be noted that the output of stage 1 is the

now input of stage second as shown in the fig 2( c). The

output from the first stage supplies energy to inductor L2.

Thus the inductor will store energy. This step finish when S2

turn off.

Fourth Step: When S2 turns off, the output of stage

second receive energy from the input of stage 1 as well as the

store energy from the inductor L2 as shown in fig 2 (d).

L1 01 11 02

(a).Two stage boost converter with 35-60V input voltage. Sl and S2 are

Mosfets

11 11 02

" ll r] 35 60V 'T CI T

e-j 9 C2 T

load

J I ! I (b) Circuit operation when switch S 1 is on, and switch S2 is off

(c) Circuit operation when switch S 1 is off, and switch S2 is on

11 L1

IAn I""'" 1-""''''''- 1 I

35 60V * ___

__

_

'--

_

_ C_l .... T _

______ C _2 +-,--j ."

(d). Circuit operation when switch S 1 again turns on, and switch S2 is off

Figure 2. Steps of operation of Two stage boost converter

III. DIMENSIONING OF THE CIRCUIT

As shown in the fig 2(a), the implemented circuit of the

converter has the following parameter: 35-60V input voltage,

350V output voltage, 450-525 watt output power and 20

KHz switching frequency. Switching frequency can be

higher depending upon the circuit designer. The load current

will vary from 1.3A to 1.5A. Now the most important part of

the circuit is the perfect design of the inductors L1 and L2

[6].

Equation (4) can be used to derive the inductor values.

L = VoD(1-D)2Ts

210min (4)

196

Ts is the switching period. 10 for the second stage is

known as shown in the specification of the circuit. So

inductance for second stage can easily be calculated using

above equation. But the 10 for the first stage is not known.

First we have to find the 10 for the first stage. Now

considering few basic equations

Pout 1]= Pin

(5)

(6)

lin for the first stage can be find using (6) and 10 can be

find using (5). In the implemented circuit the L1 is 200uH

and L2 is 550uH

Equation (7) can be used to derive the capacitor values'

c = DTsVo RLlVo

(7)

Where R is the load resistance and the V 0 is the output

ripple voltage. So the value of Cl is 630uF and C2 is 240uF.

IV. DYNAMIC MODELING

A number of AC converter modeling techniques have

appeared in the literature. In the proposed paper, state space

averaging technique is applied to model two stage boost

converter [8]. Conduction losses of the active and passive

components have been considered in the AC Modeling.

There are two basic states of operation in the proposed

circuit. Fig 3 shows the state when Sl is close and S2 is

open. While fig 2 shows the contrary state.

L1 RL1

Vgl- J

L1 RL1 R02 r",vv''----:n

RCI f RC2 Rload l CI c2 l I I T

Figure 3. Circuit diagram when S 1 is ON and S2 is OFF

Now writing state equation for the circuit as shown by

the fig. 3 diLl 1 ( ( )) --:it = L1 V 9 -RLl + RS1

diL2 1 ( . ( RC2 )) - = - VC1 + lL2Ra + VC2 -1 dt L2 RLoad+RC2 dVCl _ 2.. (-i ) dt - C1 L2

dVC2 _ 1 (. (1 _ RC2 ) _ v. ( 1 ) ---- lL2 C2 dt C2 Rload+RC2 RLoad+RC2

Where

11 RLI RDI L2 R12

I-"'''''''

f :"' RC2 n Vg RS2 1 c. c2 1 I I I T Rload

Figure 4. Circuit diagram when Sl is OFF and S2 is ON

(S.a)

(S.b)

(S.c)

(S.d)

And the state equation for the circuit shown by the figA diLl = 2.. (V 9 - iLl (RLl + RD2 + RCl) + iL2Rc1 -VCl) dt L1

dV C2 _ 1 ( V. ( 1 )) ---- -C2 dt C2 RLoad+RC2

(9.a)

(9.b)

(9.c)

(9.d)

The next step is to evaluate the state space averaged

equilibrium equation. So weight (S.a) and (S.c) by D1, (S.b)

and (S.d) by (1-D1). Similarly weight (9.a) and (9.c) by D2,

(9.b) and (9.d) by (I -D2)

197

ILl \ All AI2 0 ILl BI 3 d IL2 A21 An A23 A 2 IL2 0 = + vlNI dt vel A31 A32 0 0 vel 0

ve2 0 A 2 0 A44 ve2 0

1 All = L1 (( D

1 -1)(RLl + RD2 + RC1) - D1(RLl + RS1))

A = RCl (1 - D1) 12 L2 A -(1-D1) 13 - ----u-A = RCl D2 21 L2

1 A22 = L2 (-D2(Rc1 + RL2 + RS2) + (1 - D1)Ra) 1 A23 = -L2

A24 = (1-D2) ( RC2 -1) L2 RLoad+RC2

(1 - D1) A31= C1

A -_1_ (1_ RC2 ) 42 - (1-D2) Rload+RC2 1 ( 1 ) A44 = --C2 RLoad+RC2

1 B1 = L1 Thus we can assess the behavior of the converter using

above stat-space model.

V. EXPERIMENTAL RESULTS

The experimental results of the proposed circuit are

shown. In Fig.9 the prototype design has been shown. The

switches used are IPW60R04SCS, and FFH60UP40S. Fig.S

shows the waveforms of the inductor current in relation with

the duty cycle. You will notice a high inductor current in the

first stage and low inductor current in the second stage. Also

the duty cycle is opposite for both the switches. Fig.6 shows

that inductor current for each stage gets its maximum value

alternately depending on the duty cycle.

Figure 5. Waveform of the PWM signal for the switches and

Inductor Current Waveform of both the stages (lOv/div, IOA/div,

IOV/div, SA/div respectively)

Figure 6. Waveform of the inductor current of both the stages

(IOA/div, SA/div respectively)

198

Figure 7. Waveform of the PWM signal provided to switches

SlandS2, with inductor current waveforms ofL! and L2

Fig. 7 shows the experimental waveform of the inductor

currents, when the same duty cycle is provided to both the

switches. Now a discontinuous conduction mode has been

examined in the waveform of the inductor current. But the

converter circuit should be operated in the continuous

conduction mode. Under this situation either the load

resistance should increase or inductance of the second stage

should be increased. But the load resistance is fixed. So the

second option of increasing inductance can be used. Now if

the inductance is increased, than inductor L2 goes into

saturation as can be seen by the peaks of the inductor current

waveforms. So this situation can only be solved by providing

PWM signal alternatively to the switches of the both the

stages.

TABLE I. (EXPERIMENTAL DATA)

Vin lin Vout lout Pout D1 D2 Efficiency (V) (A) (V) (A) (Watt) % 35 5.3 350 0.5 177 59 53 95 37 6.11 350 0.61 214 59 53 94 39 6.70 350 0.70 246 61 52 94 41 7.30 350 0.80 280 61 52 94 43 8.0 350 0.91 323 62 51 94 45 8.2 350 1 347 60 51 94 47 8.67 350 1.1 381.8 61 50 94 49 9.16 350 1.2 419.8 60 50 94 50 9.63 350 1.3 452 60 52 94

Table.1 shows the step by step increase in input voltage

provided by the photovoltaic system. The output voltage of

the system remains 350 V but the load keeps on increasing

form 0.5A to 1.5A, thus giving maximum output power of

525 watt.

VI. CONTROL STRATEGY

Fig 8 shows the control circuit of two stage converter.

For the control of the circuit, a simple PI controller has been

employed [9]. Each stage is controlled by its isolated PI

controller. Output voltage of each stage is compared with a

reference voltage using error amplifier and the error signal is

generated. This error signal is provided to the modulator

which compares it with the ramp signal and thus provides the

pulse width modulated signal according to the requirement.

For sensing output voltage of each stage a suitable voltage

divider is used. So the above topology for the control of the

converter can easily be done using any efficient controller

(TL494). Fig. 9 shows the control part of the circuit in

practical design.

11 L2 02

'"" ..............

o.-j J $2 9 1 C2 Rload

ToS1 To5J

Figure 8. Close loop control of the converter circuit

VII. CONCLUSION

A high efficiency two stage boost converter circuit for

continuous input current operation is proposed for a wide

input photovoltaic module application. Key feature of the

proposed circuit are reduction of transformer, high voltage

transmission ratio, close loop control system and more

practical dimensioning of the circuit elements. The circuit

has been designed, simulated and implemented with 35-60v

input, 50V output and 50W output power. Experimental

results verify the validity of the novel circuit.

ACKNOWLEDGMENT

This research work is sponsored by 'Higher Education

Commission (HEC), Govt. of Pakistan' under the scholarship

program titled: MS level training III Korean

UniversitieslIndustry.

199

REFERENCES

[I] Boeke U., van der Broeck H., "Transformer-Less Converter Concept for a Grid-Connection ofThin-Film photovoltaic Modules" IEEE International Conference on Industrial Application Society Annual Meeting, 2008.

[2] Hoharn D. and Ropp M., " Comparative study of Maximum Power Point tracking Algorithms Using an Expeimental, programmable, Maximum power point Tracking Test Bed", IEEE Photovoltaic Specialists Conference, September 2000, pp. 1699-1702

[3] Hussein k.h.,Mutta I., hoshino T., and Oskakada M., " Maximum photovoltaic power point tracking: An algorithm for rapidly changing atmospheric conditions", IEEE procedings, Generation, Transmission, and Distribution, Vol. 142, No. I, january 1995

[4] Robert W. Erickson and Dragon Maksimovic, " Fundamentals of power electronics" Second Edition

[5] Qun Zhao, Lee F.C., " High-efficiency, high Step-Up DC-DC converters" IEEE Transactions on Power Electronics, Vol. 18, No. I, january 2003.

[6] Ned Mohan, tore M. Underland , and William P. robbins "Power Electronics : Converters, Apllications, and Design" Third Edition.

[7] Gragger J.V., himmelstoss, and Pirker F. "Analysis and Control of a Bidirectional two-stage Boost Converter" International symposium on Power Electronics, Electrical Drives, Automation and Motion, SPEEDAM 2008.

[8] Galotto L., Canesin C. A., Cordero R., Quevudo, c.A., Gazineu, R. "Non-Linear Controller Appled to Boost DC-DC Converters Using The State Space Average Model" IEEE international Power Electronics Conference, 2009.

[9] Jin Nan., Tang Hou-jun, Liu Wei, Ye Peng-sheng, "Analysis and Control of Buck-Boost Chopper type AC voltage regulator" IEEE 6th International Power Electronics and Motion Control Conference, 2009.

Figure 9. Hardware of the two-stage boost converter