IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 46, NO.
3, JUNE 1999
Interaction Between EMI Filter and Power Factor
Preregulators With Average Current Control:
Analysis and Design Considerations
Giorgio Spiazzi, Member, IEEE, and José Antenor Pomilio, Member, IEEE
Abstract—The effects of a nonnegligible source impedance, due
to the presence of an input EMI filter, on the stability of power
factor preregulators (PFP’s) with average current control are
analyzed by using a state-space averaged model. Differently from
previous approaches, it allows us to derive a simple expression
for the loop gain in terms of the converter current loop gain. The
overall system stability was studied for boost, Cuk, and SEPIC
PFP topologies. Based on this model, a simple modification of the
standard current control loop is proposed which increases the
converter robustness against instabilities. Comparison between
model forecasts and experimental measurements was carried
out using two prototypes, one based on the boost topology and
the other based on the SEPIC topology, both rated at 600 W.
Finally, the model accuracy was investigated with measurements
at different current loop bandwidths.
Index Terms— Power filters, stability criteria, switched-mode
power supplies.
I. INTRODUCTION
O WING TO increased necessity for harmonic line cur-
rent reduction, high-power-factor ac–dc converters [also
called power factor preregulators (PFP’s)] are becoming an
important issue in modern electric energy power conversion
systems. In particular, limiting standards, like IEC 1000-3-
2, which have become effective since January 1996, im-
pose maximum values for current harmonics drawn from
the utility grid [1]. Thus, in the last five years, there has
been a proliferation of topologies and control techniques
which perform an input current shaping, so as to increase
the power factor [2]–[17]. Among them, the boost converter
working in continuous conduction mode (CCM) with average
current control is probably one of the most popular solutions
for single-phase power factor correction (PFC), thanks to
its simplicity, low input current ripple, and the availability
on the market of many control IC’s. Moreover, the design
of such a converter is broadly described in many papers
[2]–[5]. The same control technique can also be applied
to other topologies with an input inductor, like Cuk and
SEPIC, which, differently from the boost one, feature high-
Manuscript received October 30, 1997; revised March 3, 1999. Abstract
published on the Internet March 1, 1999.
G. Spiazzi is with the Department of Electronics and Informatics, University
of Padova, 35131 Padova, Italy (e-mail: giorgio@tania.dei.unipd.it).
J. Antenor Pomilio is with the School of Electrical and Computer Engi-
neering, University of Campinas, 13081-970 Campinas, SP, Brazil (e-mail:
antenor@dsce.fee.unicamp.br).
Publisher Item Identifier S 0278-0046(99)04138-6.
0278–0046/99$10.00  1999 IEEE
Authorized licensed use limited to: King Mongkut's University of Technology North Bangkok. Downloaded on July 12,2021 at 14:00:47 UTC from IEEE Xplore. Restrictions apply.
577
frequency isolation, step-up and step-down capability, inherent
short-circuit and overcurrent protection, input current high-
frequency ripple reduction through magnetic coupling, etc.
[11]. The power factor achievable with this structure is actually
very high and can approach unity.
However, such converters produce high-frequency noise due
to the switching action that must be filtered out in order to
comply with EMI standards, like the IEC CISPR series. In
order to do this, an external EMI input filter is generally used
between the line grid and the PFP. When the EMI filter is
added, instabilities can arise in the system due to the inter-
action between the filter and the converter. This phenomenon
is well known and many papers have already addressed it
[19]–[22]. In particular, [22] reports an analysis, for a boost
converter only, similar to that which is presented here, but the
derived loop gain does not provide easy insight into the PFC
design. Moreover, it is not able to predict instabilities that the
analysis presented here does, and this is one reason for this
paper. Another contribution of this paper is the extension of
such analysis to other PFP topologies with average current
control, like Cuk and SEPIC. Moreover, the results of the
proposed analysis suggested a simple modification of the inner
current loop of the average current control which allows us
to greatly improve the converter’s robustness against filter-
induced instabilities.
Two prototype converters were built in order to validate
the theoretical analysis: a boost and a SEPIC ac-dc converter,
both rated at 600 W. The experimental measurements done on
these prototypes show a good correspondence between model
forecasts and actual converter behavior.
II. FILTER–CONVERTER INTERACTION ANALYSIS
In order to show the nature of the problem, let us start with
a boost PFP, the simplified scheme of which, with average
current mode control, is shown in Fig. 1 together with the EMI
input filter. As we can see, the core of the input current shaping
is an inner current loop which forces the measured inductor
current to follow a suitable reference signal . The latter is
built by sensing the rectified input voltage (block in Fig. 1
is simply a scaling factor) and by multiplying it with the output
of the voltage-error amplifier of the external output voltage
loop. This signal is practically constant at frequencies above
the line frequency, since the voltage loop has a bandwidth
much lower than the line frequency in order to maintain a
 578 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 46, NO. 3, JUNE 1999

(a) (b)
Fig. 2. (a) State-space average model of boost converter in CCM. (b)
Simplified model for input impedance calculation.

current perturbations is derived as follows:

Fig. 1. Basic scheme of a boost PFP with average current control plus an (2)
input EMI filter.

where it was considered that , since the external

good power factor. Thus, sets the correct current reference voltage loop bandwidth is well below the line frequency
amplitude in order to maintain the regulation of the output and, consequently, is constant at the frequency range
voltage. It is worthy to note that the same control structure we are interested in.
is used with Cuk and SEPIC preregulators, with only a small 2) From the converter small-signal model, the relation
modification of the power stage topology in order to stabilize between input current, duty cycle, and input voltage
the inner current loop [11]. The same figure also shows the perturbations is derived as follows:
circuit model which represents the interface between the filter
and the converter in which the Thevenin equivalent circuit of
the filter output was used ( is the filter attenuation). From (3)
this, we can write
where ( ) means perturbations with respect to steady-
(1) state values (note that from the arguments listed
at the end of the previous section).
The symbol was used for the first coefficient in
where . (3) because it represents the high-frequency converter input
can be interpreted as a loop gain which must satisfy the admittance, i.e., the admittance at frequency above the current
Nyquist criterion for stability. If were always lower loop crossover frequency in which is constant, while
than one, no instabilities could arise in the system, and this represents the transfer function between duty cycle and input
sufficient criterion was largely used in the past, in particular, current which is used for the current loop gain calculation.
for dc-dc converters [19], [20]. However, in the case of ac–dc In fact, from (2) and (3), the current loop gain can be
converters with high power factor, limitations exist both on derived, considering , as
the filter component values and on the converter design [18].
Thus, we will see that it is quite common to have (4)
in a frequency range above the inner current loop crossover
frequency, in particular, at low line voltage and high load A. Controller Analysis
currents and, in this case, the general approach for stability
analysis must be followed. Thus, from (1), we see that the From the analysis reported in the Appendix, which refers to
knowledge of the converter input impedance is a prerequisite a standard average current controller IC like the UC3854 or
for the stability analysis. the L4981, the expressions for coefficients and
In the following section, this input impedance will be de- are derived as follows:
rived for boost, Cuk, and SEPIC converters. For this purpose, it
is important to observe that, in this analysis, the output voltage (5a)
can be considered constant owing to the high value of the
output filter capacitor needed to filter out the low-frequency (5b)
components of the fluctuating input power.
where is the current-sensing resistance, is the
amplitude of the controller internal ramp, is the current-
III. PFP’S INPUT IMPEDANCE
error amplifier transfer function, and and are rms input
The calculation of the input impedance (or admittance) of voltage and current, respectively.
boost, Cuk, and SEPIC converters follows two steps. The current loop usually uses a proportional integral (PI)
1) From the scheme of Fig. 1, the relation imposed by regulator with a high-frequency additional pole in order to
the controller between duty cycle and input voltage and reject the high-frequency input current ripple (see Fig. 10),

Authorized licensed use limited to: King Mongkut's University of Technology North Bangkok. Downloaded on July 12,2021 at 14:00:47 UTC from IEEE Xplore. Restrictions apply.
 SPIAZZI AND ANTENOR POMILIO: INTERACTION BETWEEN EMI FILTER AND PFP’S WITH AVERAGE CURRENT CONTROL 579

(a) (b)

(c)
Fig. 3. Basic PFP topologies. (a) Cuk. (b) SEPIC. (c) Small-signal model for input impedance calculation.

i.e., Consequently, the expressions for the current loop gain

and high-frequency input admittance become,
(6) respectively,

Using (2)–(5), a general expression for the PFP input (10)

admittance independent of the particular topology can be found
as (11)

(7) An important result which comes from this analysis is that

the input impedance for the boost converter does not depend
From this expression, we can see that, at frequencies be- on the instantaneous input voltage, but only on its rms value
low the current loop bandwidth , the input through .
admittance is constant and equal to , while at frequencies
above the current loop crossover frequency it C. Cuk and SEPIC
coincides with . From (5b), we see that the low-frequency The basic schemes of ac–dc converters employing Cuk and
input admittance depends on the converter operating SEPIC topologies are reported in Fig. 3. Note the damping
point, i.e., network across the energy transfer capacitor
used in order to smooth the converter transfer functions, as
(8) suggested in [11]. The input admittance calculation makes use
of the pulsewidth modulation (PWM) switch model [23]. The
where is the output power. From (8), we can see that simplified small-signal model which results, short circuiting
the low-frequency converter input impedance decreases at the output filter capacitor , is shown in Fig. 3(c).
high power and low input voltage, thus making the system From it, the expressions for and are easily
more susceptible to instabilities induced by filter–converter derived as (12) and (13), shown at the bottom of the next
interactions. page, where and
are parameters which, together with the duty cycle, depend on
B. Boost PFP the converter instantaneous operating point, i.e., on the line
The derivation of (3) for the boost converter is done starting angle . The parameter is given by
from the well-known state-space average model shown in
Fig. 2(a) for CCM operation. In this figure, is (14)
the complement of the duty cycle. Since the output voltage
Expressions without the damping network can be easily
can be considered constant, the model can be simplified,
derived, letting in (12) and (13).
as shown in Fig. 2(b). From it, we can easily derive
An example of bode plots of input impedance for
boost and Cuk or SEPIC converters are reported in Fig. 4
(9)
for different input voltage values and for in the

Authorized licensed use limited to: King Mongkut's University of Technology North Bangkok. Downloaded on July 12,2021 at 14:00:47 UTC from IEEE Xplore. Restrictions apply.
 580 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 46, NO. 3, JUNE 1999

TABLE I
BOOST CONVERTER PARAMETER VALUES

TABLE II
SEPIC CONVERTER PARAMETER VALUES

Fig. 5. Bode plots of loop gain TF (j!) for a SEPIC converter. Left-hand
0
side:  = =2; a): Ug = 127 V 20%, b): Ug = 127 V + 20%. Right-hand
0
Fig. 4. Bode plot of gain and phase of input impedance of boost (left) and
Cuk or SEPIC (right,  = =2) PFP. a): Ug = 127 V + 20%. b): Ug = 127 side: Ug = 127 V 20%; a):  = =2, b):  = =200.
V. c): Ug = 127 V 20%.0
Fig. 1. The converter and controller parameters are reported in
case of Cuk and SEPIC input impedance. The converter Table II, while the filter parameters are
parameters used are reported in Tables I and II. As we can mH, and nF. The bode plot of the resulting loop
see, all curves tend to converge at (boost) or gain is shown in the left-hand side of Fig. 5 for two
(Cuk or SEPIC) for where is the different input voltage values and rated output voltage and
current loop crossover frequency ( kHz for boost power. As we can see, at lower input voltage, the systemis
and kHz for Cuk or SEPIC, depending of the unstable, since at the crossover frequency kHz
value of line angle ). Fortunately, the dependency of Cuk and (see curve a) in Fig. 5), the phase margin is 8 , while at
SEPIC input impedance from the line angle is not so strong higher input voltage it becomes stable (at kHz, the
(see next section), so that the analysis can be performed at a phase margin is about 15 ). These curves are obtained at
fixed value of . an operating point corresponding to the peak of the input
voltage (line angle . The dependence of loop gain
IV. MODEL PREDICTIONS on the line angle is shown in the right-hand side
From (1) and (7), we are now able to predict the of Fig. 5, where the same minimum input voltage was used
high-frequency instabilities which can occur from the with two different line angles (curve a) , curve b)
filter–converter interaction. Let us consider, for example, a ). In this case, the worst condition corresponds to
SEPIC PFP with a simple single-cell EMI filter, as shown in the peak of the input voltage.

(12)

(13)

Authorized licensed use limited to: King Mongkut's University of Technology North Bangkok. Downloaded on July 12,2021 at 14:00:47 UTC from IEEE Xplore. Restrictions apply.
 SPIAZZI AND ANTENOR POMILIO: INTERACTION BETWEEN EMI FILTER AND PFP’S WITH AVERAGE CURRENT CONTROL
TABLE III
COMPARISON BETWEEN MODEL FORECASTS AND EXPERIMENTAL MEASUREMENTS FOR THE BOOST PFP
V. EXPERIMENTAL RESULTS
In order to test the validity of the model forecasts, two
prototypes were built and tested. The first one is based on
the boost topology, and its parameter values are reported in
Table I, while the second one employs a SEPIC converter,
the parameter values of which are reported in Table II. Both
converters are supplied from the utility grid using an isolating
transformer plus an autotransformer in order to vary the
converter input voltage. The output inductance of the supply
line, which works as a filter inductance, was measured at
different voltages so as to use it in the input filter model.
The latter is, thus, a simple single-cell – – filter, as
shown in Fig. 1.
A. Boost Prototype
With the parameter values used for the current-error ampli-
fier, the current loop bandwidth varies from 5 to 8.3 kHz in the
output voltage range V. Comparisons between
experimental measurements of the boost PFP and model pre-
dictions are reported in Table III for different operating points.
The column corresponding to the experimental measurements
reports the value of peak input voltage at which instability
arises, together with the corresponding oscillation frequency.
The column labeled MODEL I reports the same information
derived from the model and the last column (MODEL II)
reports the crossover frequency and the phase margin as given
by the model in correspondence of the measured input voltage
value in which oscillations appear. As we can see, there is
a good agreement between model forecasts and experimental
measurements.
The value used in the model for was the dc value equal
to 0.9 . However, in this case, the model is not much sensitive
to the value of this resistance. For example, using a nonlinear
value of the type , so as to better
model the skin effect in the equivalent input filter, only small
Authorized licensed use limited to: King Mongkut's University of Technology North Bangkok. Downloaded on July 12,2021 at 14:00:47 UTC from IEEE Xplore. Restrictions apply.
581
variations of the values reported in Table III were observed
(some model predictions become more accurate like number 1
for which becomes 0.02 , and others become less accurate
like number 6 for which becomes 7.2 ).
B. SEPIC Prototype
In the case of the SEPIC PFP, the current loop bandwidth
depends on the instantaneous input voltage value, i.e., on the
line angle . With the parameter values listed in Table II,
the current loop bandwidth ranges from 6.4 to 11.5 kHz
at nominal conditions. Measurements done on the SEPIC
prototype at different operating points are reported in Table IV,
together with the model forecasts. Once again, the given model
allows us to predict quite well the instability phenomenon.
The fixed value of measured oscillation frequency was due
to measurement limitations (the oscillation period ranges from
55 to 56 s).
VI. MODEL ACCURACY
A more careful reading of the data reported in Table III
for the boost converter reveals that the difference between
measurements and model predictions depends on the output
voltage value, i.e., it depends on the bandwidth of the inner
current loop (for Cuk and SEPIC converters the current loop
bandwidth depends also on the input voltage). In order to
assess the model accuracy, experimental measurements were
done on the boost PFP at different current loop bandwidths.
The result can be summarized as follows. The phase margin
given by the model in the operating conditions in which insta-
bilities occur in the prototype is plotted in Fig. 6 against the
current loop bandwidth normalized to the switching frequency.
As we can see, the model prediction becomes more accurate,
in terms of phase margin, at lower current loop bandwidths,
while the oscillation frequency prediction remains good, even
at higher current loop bandwidths. Clearly, delays in the loops
 582
COMPARISON BETWEEN MODEL FORECASTS
Fig. 6. Model accuracy as a function of the normalized current loop band-
width.
exist which are not accounted for by the simple small-signal
model employed.
VII. CURRENT LOOP MODIFICATION
The derivation of (7) suggests a simple modification of the
controller, so as to increase the system robustness against
instabilities. In particular, we can note that the second term
at the right-hand side of (7) comes from the term in
(2), i.e., from the path from to shown in Fig. 1, and
is the term which depends on the rms input voltage. If we
insert a low-pass filter into the current reference path with
a sufficiently high corner frequency so as not to appreciably
degrade the rectified sinusoidal reference, then the converter
input admittance modifies as (see Appendix)
Comparison between the resulting loop gain and
the previous one without low-pass filter is shown in Fig. 7
( kHz) which refers to the SEPIC
converter in the operating point corresponding to measure
number 7 in Table IV (a nonlinear EMI filter resistance was
used in this plot). As we can see, this simple controller
modification reduces the loop crossover frequency from 18
kHz ( in the figure) to 13.6 kHz ( in the figure) and
Authorized licensed use limited to: King Mongkut's University of Technology North Bangkok. Downloaded on July 12,2021 at 14:00:47 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 46, NO. 3, JUNE 1999
TABLE IV
AND EXPERIMENTAL MEASUREMENTS FOR THE SEPIC PFP
Fig. 7. Gain and phase plots of loop gain TF (j!) for the SEPIC converter
at the operating point corresponding to measure number 7 in Table IV. (a)
Standard controller. (b) With a low-pass filter in the current reference path
(fPB = 1:85 kHz).
increases the phase margin from 2.5 (note that this value
differs from that reported in Table IV measure number 7
because of the different model used for the filter resistance
) to 38.4 .
This low-pass filter can be inserted simply by modifying
the controller scheme with the insertion of capacitor , as
shown in Fig. 10.
In order to prove the efficacy of such provision, the mea-
sured input voltage and current waveforms of the SEPIC
(15) converter taken in the operating point corresponding to mea-
sure number 7 in Table IV are reported in Fig. 8, which reveals
the instability predicted by the loop gain . Adding
a low-pass filter in the current reference path, the system
turns out to be stable in all operating conditions, and the
corresponding input voltage and current waveforms at nominal
conditions are shown in Fig. 9, together with the instantaneous
input power. The measurements shown in the picture allow us
to estimate the converter power factor, which is practically
unity. The same low-pass filter used with the boost converter
 SPIAZZI AND ANTENOR POMILIO: INTERACTION BETWEEN EMI FILTER AND PFP’S WITH AVERAGE CURRENT CONTROL 583

produces the same benefits, thus eliminating the instabilities

in all operating conditions, even at the higher current loop
bandwidth (17 kHz).

VIII. CONCLUSION
In this paper, the interactions between the input EMI filter,
and PFP’s with average current control have been analyzed. A
simple expression for the loop gain in terms of the converter
current loop gain was derived for boost Cuk and SEPIC pre-
regulators. The derived loop gain allows useful insight into the
converter controller design. In particular, based on this model,
a simple modification of the standard converter current control
loop is proposed which greatly increase the system robustness
against instabilities induced by filter–converter interaction.
Measurements done on two prototypes demonstrated the
model validity and its limitations.

Fig. 8. Experimental results of the SEPIC PFP in the operating point

corresponding to measure number 7 in Table IV. From top to bottom: zoom APPENDIX
of ig (t); zoom of ug (t); ug (t) 50 V/div; ig (t) 2 A/div. From the control scheme shown in Fig. 10, which represents
the standard implementation of the average current mode
control (see [3]) we can derive the expression for the duty
cycle as

(A1)

From Fig. 10, the multiplier produces an output current

which is given by (note that signal in Fig. 10, which
represents a feedforward path, is constant during a line period
and, thus, it can be considered constant at the much higher
frequencies we are interested in)

(A2)

where . At steady state, the average (in

a switching period) input current is equal to its reference, i.e.,

(A3)

where uppercase means steady-state conditions. Considering

Fig. 9. Input voltage and current waveforms and their product relative to a perturbation around an istantaneous (during the line period)
the SEPIC converter at nominal conditions (ug (t) 100 V/div; ig (t) 5 A/div;
pg (t) = ig (t) 1 ug (t) 1 kW/div.
working point, from (A2) we can obtain

(A4)

Now, substituting (A4) into (A1) and using (A3), we can

write (remember that we assumed for this calculation that
)

(A5)

from which the coefficients of (2) can be easily derived as

(A6)

(A7)
Fig. 10. Average current-mode controller scheme.

Authorized licensed use limited to: King Mongkut's University of Technology North Bangkok. Downloaded on July 12,2021 at 14:00:47 UTC from IEEE Xplore. Restrictions apply.
 584 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 46, NO. 3, JUNE 1999
where a sinusoidal input current was assumed. In this deriva-
tion, we neglected capacitor , which is the control modi-
fication proposed in the paper. Taking it into account, (A2)
modifies as
where . Consequently, (7) becomes (15).
REFERENCES
[1] Electromagnetic Compatibility, Part 3: Limits—Sect. 2: Limits for Har-

monic Current Emission (Equipment Input Current 16 A per Phase),
IEC 1000-3-2, 1995.
[2] M. J. Zhou, “Design trade-offs in continuous current-mode controlled
boost power-factor correction circuits,” in Proc. HFPC Conf., 1992, pp.
209–220.
[3] C. Silva, “Power factor correction with the UC3854,” Unitrode Corp.,
Merrimack, NH, Application Note U-125, 1986.
[4] E. X. Yang, Y. M. Jaing, G. C. Hua, and F. C. Lee, “Isolated boost
circuit for power factor correction,” in Proc. VPEC Seminar, 1992, pp.
97–104.
[5] N. Fröhleke, R. Mende, H. Grotstollen, B. Margaritis, and L. Vollmer,
“Isolated boost fullbridge topology suitable for high power and power
factor correction,” in Proc. IEEE IECON’95, 1995, pp. 405–409.
[6] L. Balogh and R. Redl, “Power-factor correction with interleaved
boost converters in continuous-inductor-current mode,” in Proc. IEEE
APEC’93, 1993, pp. 168–174.
[7] C. Zhou, R. B. Ridley, and F. C. Lee, “Design and analysis of
a hysteretic boost power factor correction circuit,” in Proc. IEEE
PESC’90, 1990, pp. 800–807.
[8] C. A. Canesin and I. Barbi, “A unity power factor multiple isolated
outputs switching mode power supply using a single switch,” in Proc.
IEEE APEC’91, 1991, pp. 430–436.
[9] J. Lo Cascio and M. Nalbant, “Active power factor correction using a
flyback topology,” in Proc. PCIM Conf., 1990, pp. 10–17.
[10] C. M. Seixas and I. Barbi, “Analysis of a power factor correction system
employing the multiphase boost converter operating in dicontinuous
conduction at constant frequency,” in Proc. COBEP Conf., 1993, pp.
207–212.
[11] G. Spiazzi and P. Mattavelli, “Design criteria for power factor prereg-
ulators based on SEPIC and Cuk converters in continuous conduction
mode,” in Conf. Rec. IEEE-IAS Annu. Meeting, 1994, pp. 1084–1089.
[12] D. S. L. Simonetti, J. Sebastian, F. S. dos Reis, and J. Uceda, “Design
criteria for Sepic and Cuk converters as power factor preregulators in
discontinuous conduction mode,” in Proc. IEEE IECON’92, 1992, pp.
283–288.
[13] G. Spiazzi and L. Rossetto, “High-quality rectifier based on coupled-
inductor sepic topology,” in Proc. IEEE PESC’94, 1994, pp. 336–341.
[14] R. Redl, L. Balogh, and N. O. Sokal, “A new family of single-
stage isolated power-factor correctors with fast regulation of the output
voltage,” in Proc. IEEE PESC’94, 1994, pp. 1137–1144.
[15] A. F. de Souza and I. Barbi, “A new ZVS-PWM unity power factor
rectifier with reduced conduction losses,” IEEE Trans. Power Electron.,
vol. 10, pp. 746–752, Nov. 1995.
Authorized licensed use limited to: King Mongkut's University of Technology North Bangkok. Downloaded on July 12,2021 at 14:00:47 UTC from IEEE Xplore. Restrictions apply.
[16] D. Maksimovic, Y. Jang, and R. W. Erickson, “Nonlinear-carrier control
for high-power-factor boost rectifiers,” IEEE Trans. Power Electron.,
vol. 11, pp. 578–584, July 1996.
[17] Z. Lai, K. M. Smedley, and Y. Ma, “Time quantity one-cycle control
for power-factor correctors,” IEEE Trans. Power Electron., vol. 12, pp.
369–375, Mar. 1997.
[18] V. Vlatkovic, D. Borojevic, and F. C. Lee, “Input filter design for power
factor correction circuits,” IEEE Trans. Power Electron., vol. 11, pp.
199–205, Jan. 1996.
[19] R. D. Middlebrook, “Input filter considerations in design and application
of switching regulators,” in Conf. Rec. IEEE-IAS Annu. Meeting, 1976,
pp. 366–382.
[20] R. D. Middlebrook, “Design techniques for preventing input-filter oscil-
lations in switched-mode regulators,” in Proc. Power Conversion Conf.,
May 4–6, 1978, pp. A3-1–A3-16.
[21] S. Y. Erich and W. M. Polivka, “Input filter design for current-
programmed regulators,”in Proc. IEEE APEC’90, 1990, pp. 781–791.
[22] R. Redl and A. S. Kislovsky, “Source impedance and current-control
loop interaction in high-frequency power-factor correctors,” in Proc.
IEEE PESC’92, 1992, pp. 483–488.
[23] V. Volperian, “Simplified analysis of PWM converters using the model
of pwm switch: Part I and II,” IEEE Trans. Aerosp. Electron. Syst., vol.
26, pp. 490–505, May 1990.
Giorgio Spiazzi (S’91–M’93) was born in Legnago,
Verona, Italy, in 1962. He graduated with honors in
electronic engineering and received the Ph.D. degree
in industrial electronics and informatics from the
University of Padova, Padova, Italy, in 1988 and
1993, respectively.
Since 1993, he has been a Permanent Researcher
with the Department of Electronics and Informatics,
University of Padova. His main research interests
are advanced control techniques for dc/dc con-
verters, power-factor controllers, and soft-switching
techniques.
José Antenor Pomilio (M’93) was born in Jundiaı́,
Brazil, in 1960. He received the B.Sc., M.Sc., and
D.Sc. degrees in electrical engineering from the
University of Campinas, Campinas, Brazil, in 1983,
1986, and 1991, respectively.
From 1988 to 1991, he was Head of the Power
Electronics Group at the Brazilian Synchrotron
Laboratory. Since 1991, he has been an Assistant
Professor in the School of Electrical and Computer
Engineering, University of Campinas. During
1993–1994, he held a post-doctoral position in
the Electrical Engineering Department, University of Padova, Padova, Italy.
His main research interests are switching-mode power supllies, electrical
drives, and active power filters. He is Vice-President of the Brazilian Power
Electronics Society (SOBRAEP).
Dr. Pomilio is currently the IEEE Power Electronics Society Liaison to
Region 9.