Modelling and simulation of power factor

corrected AC–DC converters

Ashish Pandeya, Dwarka P. Kotharia, Ashok K. Mukerjeea and Bhim Singhb
a
Center for Energy Studies, Indian Institute of Technology, Hauz Khas, New Delhi, India
b
Department of Electrical Engineering, Indian Institute of Technology, Hauz Khas, New
Delhi, India

Abstract The focus of the present work is on introducing the power electronics community to the
modelling and simulation of power factor corrected (PFC) converters. These techniques not only help
to develop a deeper understanding of these converters but also to evaluate performance and feasibility
of control strategies and topological features without fabrication of an actual system. Important PFC
converter topologies are modelled and simulated. Simulation results of a single-phase boost converter
are then compared to simulation results obtained by SIMULINK/SimPower Systems as well as
experimental results to provide an overview of the capabilities and limitations of various approaches. It
is expected that this work will be of use to students as well as researchers who are interested in
studying and researching PFC converters.

Keywords modelling; power electronics; power engineering education; power quality; simulation

Power factor corrected (PFC) converters are an important area of study and research
in power electronics.1 These AC–DC converters provide stable DC voltage at the
output with high input power factor. This capability makes PFC converters an
extremely attractive choice for offline power supplies and other AC–DC power con-
version applications because of increasing concerns about power quality and to meet
the guidelines of various power quality regulations and standards.2 Since these con-
verters cater to the unique requirements of a large number of applications, several
control strategies and topologies need to be evaluated and developed to meet the
specifications of the target application.
To this end, modelling and simulation techniques are indispensable tools. These
not only help in developing a deeper understanding of PFC converters but are also
extremely important tools for design verification and performance evaluation. These
techniques help in the evaluation of a system without risking the huge cost and effort
of developing and testing an actual converter. More importantly, the simulation study
allows checking for fault conditions and operating modes that can be difficult and
risky to test on a real system. Moreover, these techniques allow one to observe and
understand how different components interact and influence the overall performance
of the system.
As an example, input current distortion in a PFC converter (shown in Fig. 1) can
be caused by a large number of factors, such as ripple in the output voltage feed-
back and input voltage feedforward path, distortion in the input voltage, non-
linearity in the multiplier, and poorly designed current regulator. However, in an
actual system it is extremely difficult to study the impact of just one of these com-
ponents in isolation. Therefore one may prefer a simulation approach to study the

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 245

Fig. 1 Cascaded control scheme used in PFC converters.

impact of one or more components to develop and evaluate strategies to negate its
impact without worrying about the influence of other components.
Similarly, new topology ideas can also be tested by simulation studies. As an
example, excessive currents and voltages possibly caused by inadvertent shorting of
devices or other failure modes due to flaws in switching logic can be isolated and
corrected before they can manifest in an actual system. Testing directly on an actual
system can make troubleshooting prohibitive due to measurement problems as well
as continual failure of devices.
It is, however, important that the information required from the simulation study
is clearly identified before mathematical modelling is carried out. This might range
from switching-ripple-averaged information, to obtain ripple-averaged components
of various states of the system, to transitory phenomena in devices during switch-
ing transitions, such as current and voltages in semiconductor switches when the
switch turns off.
Depending upon the information required, modelling and subsequent computer
simulation could be carried out at system or component level. System-level simula-
tion essentially makes use of a simplified model that can be used to quickly verify
or develop new ideas in topology or control. Here the switching information can
be retained or ignored depending upon the modelling approach. While the overall

International Journal of Electrical Engineering Education 41/3

246 A. Pandey et al.

system dynamics can be studied with this technique, component-level details are not
available since the mathematical model of the power circuit essentially overlooks
the non-ideal behaviour of devices to keep the model simple.
In this work PFC converters are modelled using Kirchoff’s and Ohm’s laws to
obtain differential equations describing the circuit behaviour. These equations can
be solved using numerical methods in a programming environment such as C++ or
Matlab. The control algorithms are also embedded in the same computer program,
since the PFCs cannot operate without a feedback controller, as is evident from Fig.
1. This technique provides greater flexibility due to the choice of modelling approach
and use of a low-level programming language for simulation. Moreover it does not
require costly simulation packages and is also considerably faster. These factors
make it attractive for quick and accurate verification of new ideas and for develop-
ing an intimate understanding of the converter system.
Component-level simulation, on the other hand, essentially requires commercial
packages, such as PSpice, which make use of detailed models of various compo-
nents in a power electronics circuit.3 Since the intricacies of various components are
modelled, a component-level simulation can provide deeper insight into the opera-
tion of individual devices in a converter system. As an example, sub-switching
phenomena such as transitory current and voltage in semiconductor devices during
switching transitions can be studied with component-level simulation. This infor-
mation can be useful for designing snubbers as well as for evaluating current and
voltage stresses on the devices.
While component-level simulation can be useful in designing and testing a sub-
circuit, like a snubber network, it is generally difficult to simulate a complex circuit
such as a PFC converter. It may be argued that full circuit simulation at component
level can be as complex and time-consuming as development and testing of an actual
prototype.
A true compromise between the two widely different techniques is Simulink/
SimPower Systems, which is part of the Matlab package. Here the controller can
be designed using the mathematical capabilities of Matlab while the non-ideal
behaviour of various power circuit components such as diodes and mosfets can be
captured using the power electronics library of SimPower Systems. All this can be
accomplished in a user-friendly Graphical User Interface (GUI) based Simulink
environment. This makes the simulation faster and more like system level simula-
tion while at the same time the simulation is able to provide greater details of the
dynamics in PFC converters.

Modelling the boost family of converters

Mathematical modelling is an integral part of the computer simulation process. The
system behaviour is obtained by solving the modelling equations. Mathematical
modelling also decides the details of the system that can possibly be studied by com-
puter simulation. PFC converters can be modelled using a number of techniques
such as state-space averaging4 or singular-perturbation5 methods. However, these
techniques are more suitable for the design and analysis of feedback controllers since

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 247

the system is linearized to facilitate controller design and analysis, and are of little
use for performance simulation.
The starting point for modelling a PFC converter, however, is by application
of Kirchoff’s and Ohm’s law to the circuit, which provides first-order differential
equations describing the state of current through inductor(s) and voltage across
capacitor(s). The discrete switching function obtained from the controller is also
incorporated into the modelling equations to obtain details of switching ripple. Addi-
tionally, other components of the PFC converter system, such as the single-phase
AC supply, feedback controller and load, are also modelled and solved concurrently
to simulate the entire system. All the different components of the PFC converter are
modelled separately and then integrated to form a comprehensive model of the
system. The modelling equations are then solved using software developed in C++
for verification.
Simulations for all the converters are carried out under similar conditions with
a sinusoidal AC source, 220 VAC 50 Hz input, 500 W converter with a switching
frequency of 50 kHz. The output voltage is 400 VDC. Exceptions for topological
reasons are mentioned wherever required.

Supply system
Under normal operating conditions the supply system can be modelled as a
sinusoidal voltage source of amplitude Vm and frequency fs. The instantaneous
voltage is:
vs (t ) = Vm sinwt (1)
where w = 2pfs electrical radians/second and t is instantaneous time.
In some topologies, the input is rectified line voltage vd(t) which can be given as:
vd (t ) = vs (t ) = Vm sinwt (2)
From the sensed supply voltage, an input-voltage template u(t) is estimated for
converter topologies with AC side inductor as
u(t ) = vs (t ) Vm (3)
The input-voltage template for converter topologies with a DC side inductor is
obtained from:
u(t ) = vs (t ) Vm (4)

Feedback control
PFC converters, like most power electronics systems, cannot function without feed-
back control. Figure 1 shows a typical control scheme for PFC converters – the
current mode control. This control scheme ensures regulated DC output voltage at
high input power factor. Here the control philosophy is fairly intuitive. The output
DC voltage regulator generates a current command, which is the amount of current
required to regulate the output voltage to its reference value. The output of the
DC voltage regulator is then multiplied with a template of input voltage to generate

International Journal of Electrical Engineering Education 41/3

248 A. Pandey et al.

an input current reference. This current reference has the magnitude required to
maintain the output DC voltage close to its reference value and has the shape and
phase of the input voltage – an essential condition for high input power factor
operation.

DC voltage controller
A proportional integral (PI) voltage controller is selected for zero steady-state error
in DC voltage regulation. The DC voltage vdc is sensed and compared with the set
reference voltage vdc*. The resulting voltage error ve(n) at the nth sampling instant is:
ve ( n ) = v*dc - vdc ( n ) (5)
The output of the PI voltage regulator vo(n) at the nth sampling instant of the PI
controller will be:
vo ( n ) = vo ( n -1) + K p {ve ( n ) - ve ( n -1) } + Ki ve ( n ) (6)
Here Kp and Ki are the proportional and integral gain constants, respectively. ve(n-1)
is the error at the (n - 1)th sampling instant. The output of the controller vo(n) after
limiting to a safe permissible value is taken as the amplitude of the input current
reference A (Fig. 1).

Reference supply current generation

The input voltage template u(t) obtained from the sensed supply voltage is multi-
plied by the amplitude of the input current reference A to generate a reference
current. The instantaneous value of the reference current is given as:
iL* = AB C 2 (7)
where B is the input voltage template u(t) and C is the input voltage feedforward
component obtained by low-pass filtering the sensed input voltage signal.

Active wave-shaping of input current

The input current error is the difference between the reference current and the actual
current (DiL = iL* - iL). A number of current control strategies can be employed to
contain the current error within the desired range so that the input current tracks its
reference with minimal error. The switching logic is dependent on the type of con-
verter topology used. In the present work, a rule based hysterisis current controller
is used due to the applicability of this technique to all topologies.

Modelling devices and load

The various devices and elements in PFC converters can exhibit highly complex
dynamics, however, in this work their ideal behaviour is modeled for simplicity and
appropriateness.

Passive elements
The inductor is modelled as a linear inductance Ls (or L1 and L2 in multiple in-
ductor topologies) with a DC resistance r. Similarly, a capacitor is modelled with

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 249

capacitance Cd (or C1 and C2 in topologies with multiple capacitors). The equiva-

lent series resistance (ESR) of capacitors is not considered.

Semiconductor switches
Semiconductor switches, mosfet S and diode D are modelled as pure ON–OFF
switches. No snubbers or non-idealities in the switches are modelled.

Load
The converters are modelled as resistive loads having resistance R.

Power circuit
The power circuit is modelled by first-order differential equations describing
the circuit behaviour. These modelling equations are obtained by application of
Kirchoff’s and Ohm’s laws to the power circuit.

Single-phase boost PFC converter

This is by far the most popular topology. It is used extensively for front-end high
power factor rectification in offline power supplies and electric drive applications.
The boost converter (Fig. 2a) is modelled using two differential equations for
inductor current iL and DC link capacitor voltage vdc:
piL = (vd - v p ) Ls - r(iL Ls ) (8)
pvdc = (i p - vdc R) Cd (9)
where p is the differential operator (d/dt). vp is the PWM voltage across the switch
and is defined as
v p = vdc (1 - d ) (10)
ip is the current through the boost diode and is defined as
i p = iL (1 - d ) (11)
Here d is the switching signal obtained from the current regulation loop. Its value
is 1 (ON) or 0 (OFF) depending upon the state of the switch S (Fig. 2a). As is obvious
from the circuit topology and modelling equations, the DC output voltage in this
converter has to be kept higher than the peak value of the input voltage to allow
current control. The input current and output voltage details obtained by solving the
model are shown in Fig. 2b.

Single-phase semi-boost converter

These converters have been proposed to eliminate the blocking diode D (Fig. 2a) in
the single-phase boost converter. This enhances converter efficiency and reliability.
Two variants of semi-boost converter topologies are considered, namely sym-
metrical (Fig. 3a) and asymmetrical (Fig. 3b). Both types have a similar model and
consequently identical characteristics. The variation leads to simplified current
regulation in the symmetrical variant. Here again the converter is described by two

International Journal of Electrical Engineering Education 41/3

 250 A. Pandey et al.

(a)

(b)

Fig. 2 (a) Power circuit of boost converter. (b) Simulation results for boost converter.

differential equations for inductor current iL and DC link voltage across capacitor
vdc:
piL = (vs - v p - r ◊ iL ) Ls (12)
pvdc = (i p - vdc R) Cd (13)
where PWM voltage and current are given by
v p = vdc (d1 - d2 ) (14)
i p = iL (d1 - d2 ) (15)

International Journal of Electrical Engineering Education 41/3

 (a)

(b)

(c)

Fig. 3 (a) Power circuit of symmetrical semi-converter. (b) Power circuit of asymmetrical
semi-converter. (c) Simulation results for semi-converter.

International Journal of Electrical Engineering Education 41/3

 252 A. Pandey et al.

respectively. Here d1 and d2 are switching states of switches S1 and S2 respectively.

The modelling equations show that the dynamics of this converter is similar to the
single-phase boost converter. Here too, the output voltage has to be maintained at a
value higher than the peak value of input voltage for proper current regulation. The
validity of the model is demonstrated in the simulation results shown in Fig. 3c.

Single-phase half-bridge converter

The half-bridge converter shown in Fig. 4a allows bi-directional power flow and
eliminates the use of blocking diode D (Fig. 2a). The half-bridge configuration can
be operated in both modes, as rectifier and inverter. There are three modelling equa-
tions describing the state of input current and voltages across the capacitor:

(a)

(b)

Fig. 4 (a) Power circuit of half-bridge converter. (b) Simulation results for half-bridge
converter

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 253

piL = (vs + v p1 - v p 2 - riL ) Ls (16)

pvC1 = - {i p1 + (vdc R)} C1 (17)
pvC 2 = - {i p 2 - (vdc R)} C 2 (18)
DC link voltage is given as
vdc = (vC1 + vC 2 ) (19)
and the PWM voltages and current are given by
i p1 = d1iL (20)
i p 2 = d2 iL (21)
v p1 = d1vC1 (22)
v p 2 = d 2 vC 2 (23)
Here d1 and d2 are switching states of switches S1 and S2.
It becomes apparent from the model that the voltage across each capacitor has to
be above the peak value of the input voltage. Therefore the DC link has to be main-
tained at about twice the input voltage. Simulation results in Fig. 4b show the input
current, voltage across capacitor C1 and output voltage.

Single-phase voltage source converter

Voltage source converters (Fig. 5a) provide excellent control over power flow in
both directions. They can be operated as AC–DC converters to generate regulated
DC voltage at high input power factor. The power flow can easily be reversed to
operate the converter as a DC–AC converter. This capability makes the system
ideally suited to electric drives and line interactive UPS applications.
The converter is described by two differential equations for inductor current iL
and voltage across capacitor vdc:
piL = (vs - v p - r ◊ iL ) Ls (24)
pvdc = (i p - vdc R) Cd (25)
where PWM voltage and current are
v p = vdc (da - db ) (26)
i p = i L (d a - d b ) (27)
respectively. Here
da = 1 if switch S1, otherwise da = 0
db = 1 if switch S2, otherwise db = 0
The simulation results shown in Fig. 5b validate the modelling equations.

Single-phase interleaved converter

Interleaved converters have been proposed for high current applications to reduce
device current rating and inductor size by paralleling the switching devices and

International Journal of Electrical Engineering Education 41/3

 254 A. Pandey et al.

(a)

(b)

Fig. 5 (a) Power circuit of voltage source converter. (b) Simulation results for voltage
source converter.

inductors (Fig. 6a). The inductor size is reduced by a quarter for the same current
ripple. This is a significant advantage as inductor size can severely constrain the
design of high current PFC converters. The interleaved converter can be modelled
by three differential equations describing the states of two inductor currents and the
voltage across the capacitor:
piL1 = (- vd + v p1 + vd ) L1 (28)
piL 2 = (- vdc + v p 2 + vd ) L2 (29)
pvdc = {iL1 + iL 2 - i p1 - i p 2 - (vdc R)} Cd (30)

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 255

(a)

(b)

Fig. 6 (a) Power circuit of interleaved converter. (b) Simulation results for interleaved
converter.

where iL1 and iL2 are currents through inductors L1 and L2 respectively. Here the
input current is is given as
is = iL1 + iL 2 (31)
The PWM voltages and currents in the system are given as
v p1 = vdc d1 (32)
v p 2 = vdc d2 (33)
i p1 = iL1d1 (34)

International Journal of Electrical Engineering Education 41/3

 256 A. Pandey et al.

i p 2 = iL 2 d1 (35)
Here d1 and d2 are the states of switches S1 and S2 respectively.
Figure 6b shows the input current, current through L1 and the output voltage of
the converter.

Single-phase multilevel converter

The multilevel converter (Fig. 7a) is proposed for high voltage applications and more
specifically where the voltage rating of switching elements and output tank capa-
citors constrains the design. The multilevel converter is also modelled by three

(a)

(b)

Fig. 7 (a) Power circuit of multilevel converter. (b) Simulation results for multilevel
converter.

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 257

differential equations. These represent states of voltages across two capacitors and
current flowing in the inductor:
piin = {[(d1 - 1) Ls ]vC1} + {[(d2 - 1) Ls ]vC 2 } + vd Ls (36)
where the output voltage
vdc = (vC1 + vC 2 ) (37)
vC1 = {[1 - d1 ]iL - (vC1 R)} C1 (38)
vC 2 = {[1 - d2 ]iL - (vC 2 R)} C 2 (39)
Figure 7b shows input current, voltage across capacitor C1 and output voltage
obtained by solving the modelling equations.

Discussion of key results

The modelling and simulation of the single-phase boost converter is analysed to
highlight the capabilities and limitations of the method. Simulations are carried out
using software developed in C++ as well as using the Matlab based Simulink and
SimPower Systems package. Key results obtained by simulation are compared to
experimental results.

Simulation in C++
Figure 8 shows a flowchart for solving the modelling equations for computer
simulation using C++ programming language. Performance simulation by develop-
ing programs provides an intimate understanding of the converter system, as it
requires extensive interaction with the modelling equations and controller algorithm.
It also gives greater flexibility in controller design and verification. The differential
equations are solved using a fourth-order Runga–Kutta method. The controller input
d in the differential equation is obtained by solving the controller algorithms con-
currently. Essentially, this technique is as accurate as the mathematical model of the
system.

Simulation in MATLAB®, SIMULINK® and SimPower Systems

A more detailed simulation of the converter system is possible using Matlab based
packages. The SimPower Systems library in Simulink offers detailed models of
semiconductor devices used in the construction of power circuits. This power elec-
tronics circuit can be interfaced with other SIMULINK libraries, uniquely suited to
implementing the various control schemes that are required for the PFC converters.
This approach has several advantages since it offers a detailed model of the power
circuit while the controller can be tested and designed using the mathematical capa-
bility and flexibility offered by Simulink and Matlab.
The speed of simulation in the Simulink environment can be increased signifi-
cantly by discretizing the model using the PowerGUI tool and fixed-frequency
solvers. Here the different components of the PFC system, such as power circuit,
voltage and current controller and sensors etc. can also be incorporated for perfor-

International Journal of Electrical Engineering Education 41/3

258 A. Pandey et al.

Fig. 8 Flowchart for simulation in C++.

mance simulation of PFC converters (Fig. 9a). The power circuit can be developed
using readymade components from the SimPower library, which includes AC power
source, semiconductor devices, inductors and capacitors as shown in Fig. 9b. For
semiconductor devices, a universal bridge is used to allow discretization of the
model.

Verification of simulation techniques

The simulation results obtained by C++ and Matlab based packages require careful
study in light of experimental results to gain confidence in the modelling and
simulation strategy and to appreciate the capabilities and limitations of the various
techniques.
For comparison between the different techniques, identical values are taken for
all components of the power circuit. The component details are provided in Table I.
Additionally, experimental verification is carried out by developing a boost converter
prototype using a UC3854A commercial PFC controller ASIC.
Comparison of output voltage and input current waveforms under steady-state
operation, obtained for the single-phase boost converter by simulation in C++ (Fig.
2b) and the Matlab package (Fig. 10a) show insignificant differences. The minor
differences are primarily due to the use of different solvers.

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 259

(a)

(b)

Fig. 9 (a) Complete circuit model in SIMULINK. (b) Power circuit using SimPower Systems
and SIMULINK.

TABLE 1 Details of system components

Parameter C++ Matlab package Experimental

Input voltage 220 VAC, 50 Hz 220 VAC, 50 Hz 220 VAC, 50 Hz

Inductance 1100 mH 1100 mH 1100 mH
Capacitance 470 mF 470 mF 470 mF ± 20%
Input EMI filter No No Yes
Voltage controller Discrete PI Discrete PI Linear compensator
Current controller PWM PWM PWM
Switching frequency 50 kHz 50 kHz 50 kHz

International Journal of Electrical Engineering Education 41/3

 260 A. Pandey et al.

(a)

(b)

Fig. 10 (a) Simulation results for boost converter using SimPower System and SIMULINK.
(b) Experimental results for boost converter (upper trace: output voltage, lower trace:
input current).

The experimental results obtained from the single-phase boost converter shown
in Fig. 10b match the simulation results. Furthermore, evaluation of converter
dynamics under load variation (180 W to 500 W) shows similar results for the experi-
mental (Fig. 11a) and simulation (Fig. 11a and 11b) techniques. Some differences

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 261

are due to the choice of voltage regulator. In general, a discrete PI controller

provides slightly better dynamic response than a linear feedback controller, which
was used in the experimental prototype. Additionally, the dynamic response of a dis-
crete PI controller is also dependent on the instant of disturbance as controller action
is taken only at zero crossings of the input voltage.

(a)

(b)

Fig. 11 (a) Voltage dynamics during loading (experimental results). (b) Voltage dynamics
during loading (Simulation C++).

International Journal of Electrical Engineering Education 41/3

 262 A. Pandey et al.

(c)

Fig. 11 (c) Voltage dynamics during loading (simulation SimPower System and
SIMULINK).

An interesting feature of all PFC converters is that the output voltage dynamics
is dependent on low-frequency ripple on the DC bus voltage. This ripple, present in
all PFC converters, is generated by the input power pulsating at twice-the-line fre-
quency whereas the output power is essentially constant.6 Ripple generation can be
explained mathematically using the model of the single-phase boost converter.
In the converter the input power Pin can be given as
Pin (t ) = Vm sin wt Im sin wt = (VI - VI cos wt ) (40)
Here Vm and Im are the peak values of rectified input voltage and current respec-
tively. V and I are their respective RMS values.
Neglecting the effect of Ls, the input power can be obtained in terms of output
voltage and diode current as
Pin (t ) = Vdc ◊ i p (41)
Applying Kirchoff’s current law to the circuit (equation (9))
i p (t ) = Idc + ic (t ) = (VI - VI coswt ) Vdc (42)
Essentially, the time-varying component of diode current ip(t) flows into the capa-
citor and the DC component flows directly into the load. Therefore the capacitor
current can be given as
ic (t ) = VI coswt Vdc (43)

International Journal of Electrical Engineering Education 41/3

Power factor corrected AC-DC converters 263

The capacitor current ic(t) causes ripple on the output voltage vrip (t) which can now
be given as
vrip (t ) ª (1 Cd )Ú ic (t ) dt = -(Idc 2wCd ) sin (2wt ) (44)
Here Idc/(2wCd) is the magnitude of the ripple and -sin(2wt) is the shape of the ripple
and can also be clearly seen in simulation and experimental results.
The most significant deviation from the model is caused by non-conduction of
the diode during zero-crossover of the input voltage. This can be evaluated by
rearranging equation (8) from the boost converter model as
Ls piL = {vd - vdc (1 - d )} - r(iL ) (45)
Here the left-hand side term is the inductor voltage. At zero crossing of the input
voltage (i.e. vd = 0) the inductor voltage can be given as
LpiL = - vdc (1 - d ) (46)
The resistive voltage drop (r iL) can be neglected since the current is negligible at
that moment and r is extremely small. Evaluation of equation (46) shows that
the inductor voltage should swing from -vdc to 0 as the switch turns ON and OFF
respectively.
However, the actual system deviates from the mathematical model due to non-
conduction of the diode around the zero crossing. This leads to a significant dip in
the inductor voltage while the model predicts the inductor voltage to be equal to the
output voltage at that instant.
While this aberration, shown in Fig. 12, is not critical for most studies it can effect
some important developments in the area of sensor reduction techniques7,8 where the
theoretical design of a voltage estimation technique based on the model can

Fig. 12 Experimental waveform of inductor voltage in single-phase boost converter.

International Journal of Electrical Engineering Education 41/3

264 A. Pandey et al.

completely fail due to the inability of the model to accurately predict system
behaviour.
However, for all design and development purposes, the model accurately predicts
the circuit behaviour and provides switching details. Several unmodelled phenom-
ena, such as diode reverse recovery, AC losses in switching devices as well as losses
due to skin and proximity effects and saturation of magnetic material in inductors,
ripple current stress on capacitors and switching-frequency ripples on the output
voltage etc., are more efficiently handled in the design process rather than attempt-
ing to incorporate them in the model.

Conclusions
Key PFC converter topologies have been modelled and simulated. Important results
from modelling and simulation of PFC converters are compared with results
obtained from an experimental prototype as well as from a standard Matlab
package to highlight the capabilities and limitations of the approach. It has been
observed that the modelling and simulation approach is quite adequate for design,
development and academic research. This method accurately captures the various
intricacies of PFC converters, while at the same time has unsurpassed performance
in terms of speed. Additionally, this method is uniquely suited to the academic com-
munity since it also helps in developing an intimate understanding of the converter
system.

References
1 B. Singh, B. N. Singh, A. Chandra, K. Al-Haddad, A. Pandey and D. P. Kothari, ‘A review of single-
phase improved power quality AC–DC converters,’ IEEE Trans. Ind. Electron., 50 (2003).
2 Electromagnetic compatibility (EMC) – Part 3: Limits- Section 2: Limits for harmonic current emis-
sions (equipment input current <16 A per phase), IEC61000-3-2 Document, first edition, 2002.
3 M. H. Rashid (ed.), Power Electronics Handbook (Prentice Hall, Englewood Cliffs, NJ, 1993).
4 R. W. Erickson, Fundamentals of Power Electronics (Chapman & Hall, New York, 1997).
5 S. Wall and R. Jackson, ‘Fast controller design for single-phase power-factor correction systems,’
IEEE Trans. Ind. Electron., 44 (1997), 654–660.
6 N. Mohan, T. Undeland and W. Robbins, Power Electronics: Converters, Applications and Design,
second edition (John Wiley & Sons, New York, 1995).
7 A. Pandey, B. Singh and D. P. Kothari, ‘A Novel DC bus voltage sensorless PFC rectifier with
improved voltage dynamics,’ in Proc. 28th IEEE IECON’02 Sevilla, Spain, 5–8 Nov. 2002, pp.
226–229.
8 N.-C. Choi and M.-H. Seo, ‘Power factor correction circuit having indirect input voltage sensing,’
US Patent 5,764,039, 9 June 1998.

International Journal of Electrical Engineering Education 41/3