46 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 52, NO. 1, FEBRUARY 2005
Helpful Hints to Select a Power-Factor-Correction
Solution for Low- and Medium-Power
Single-Phase Power Supplies
Arturo Fernández, Member, IEEE, Javier Sebastián, Member, IEEE, Marta María Hernando, Member, IEEE,
Pedro Villegas, Member, IEEE, and Jorge García, Member, IEEE
Abstract—This paper presents a review of power-factor-correc-
tion (PFC) circuits for low- and medium-power single-phase power
supplies. The main idea is not just to show the state of the art of this
topic but to select the most interesting topologies for each applica-
tion depending on the power level, the input voltage range, and the
output voltage. Since IEC 61000-3-2 regulations came into force,
many new topologies have been presented trying to obtain a cost-ef-
fective solution to reduce the input current harmonic content. Each
one of them has its application range due to the inherent charac-
teristics of the topology. Obviously, not every converter is useful for
the same application. This is especially perceptible in PFC circuits
due to the large amount of different solutions. Hence, this paper
tries to show the most appropriate topologies for each application,
being the input power and the IEC 61000-3-2 Class some of the
main parameters to select it. The scope of the paper is focused on
single-phase power supplies belonging to IEC 61000-3-2 Class A
and Class D with an input power level below 4 kW.
Index Terms—IEC 61000-3-2 regulations, low power,
power-factor correction (PFC).
I. INTRODUCTION
T
WENTY YEARS ago, power-factor correction (PFC) was
a synonym of sinusoidal input current and extremely low
total harmonic distortion (THD). Hence, topologies with a two
loop control system were the obvious solution. Typically, boost
or flyback converters were used. An inner control loop force the
input current to be sinusoidal and the second control loop regu-
lates the output voltage. The typical way to achieve this is using
a pulsewidth-modulation (PWM) controller with an analog mul-
tiplier. The only disadvantage of this controller is the price,
which is higher than the price of a conventional PWM controller.
From the point of view of the behavior of the converter, the main
disadvantage is the slow dynamic response of the output voltage.
As far as the objective was to get a sinusoidal input current with
a small distortion, a low pass filter with a very low corner fre-
quency should be placed in the feedback path of the voltage loop
to reduce the 100-Hz ripple and not to distort the input current.
Thus, the dynamic behavior of this converter is quite slow.
Another possibility is to take advantage of an inherent char-
acteristic of some converters to behave naturally as a resistor
emulator (RE) when the converter operates in the discontinuous
Manuscript received December 30, 2003; revised January 24, 2004. Abstract
published on the Internet November 10, 2004.
The authors are with the Grupo de Electrónica Industrial, Universidad de
Oviedo, 33204 Gijón, Spain (e-mail: arturo@ate.uniovi.es).
Digital Object Identifier 10.1109/TIE.2004.841141
conduction mode (DCM). In fact, both the flyback and the boost
converter can operate this way. This method avoids the use of
an analog multiplier and allows the use of a conventional and
nonexpensive PWM controller. The drawback of this method
is the operation in DCM. The current values are much higher
than in continuous conduction mode (CCM) and the efficiency
of the converter decreases. Hence, this system can be used in
low power applications but the stress of the semiconductors is
too high in high-power converters.
If the application needs a fast dynamic response, a second
converter should be placed in cascade. Thus, the first stage (typ-
ically a boost converter) corrects the power factor and regulates
the intermediate dc link. Then, the second stage regulates the
output voltage with a fast dynamic response and provides gal-
vanic isolation. This scheme has many advantages: the boost
converter can be very efficient and can easily regulate the in-
termediate dc link even with universal input voltage. As far as
this intermediate voltage is well regulated, the second stage can
be much optimized because it has not to deal with a large input
voltage range. Hence, it can be designed to operate in the most
optimized point for that particular topology. As can be seen,
both converters can be very efficient and, as a consequence, the
total efficiency of the system can be quite high, even though
the energy is processed twice. However, if the power level is
low (say less than 300 W), the fact of using two converters in
cascade, with two control circuits, two protection systems, etc.,
makes this solution quite expensive.
Furthermore, when IEC 61000-3-2 regulations came into
force in 2001, this problem with low power converters became
even more evident. To comply with this regulation it is not
necessary to have a unity power factor waveform. The only
requirement is to have the harmonic content of the input power
measured at nominal input voltage below a limit fixed by the
regulation. The equipment is classified in four classes: A, B,
C, and D. Class B is for portable equipment, Class C is for
lighting equipment, Class D is just for TV sets, monitors, and
computers. Finally, all the rest of equipment fits in Class A.
Each Class has different limits for harmonics ranging from the
2nd to the 39th. Hence, the shape of the current waveform can
be very different from a sinusoid. However, if the harmonic
content is below the limit, the converter complies with the
regulation.
It should be noted that most of the industry equipment fits in
Class A. The limits of this Class are absolute; this means that
0278-0046/$20.00 © 2005 IEEE