Professional Documents
Culture Documents
Fig 1. (a) Traditional PFC converter with current sensor. (b) Thermal image
at full load.
Fig. 2 (a) Typical PFC scheme with digital control and a current sensor.
(b) Analog to digital conversion circuit of the input current.
I. INTRODUCTION
Manuscript received December 20, 2012; revised March 10, 2013, May 10,
2013, and August 11, 2013; accepted August 16, 2013. Date of current version
February 18, 2014. Recommended for publication by Associate Editor S. K.
Mazumdar.
V. M. Lopez and F. J. Azcondo are with the Universidad de Cantabria, Santander 39005, Spain (e-mail: lopezvm@unican.es; azcondo@ieee.org).
A. de Castro is with the Universidad Autonoma de Madrid, Madrid 28049,
Spain (e-mail: angel.decastro@uam.es).
R. Zane is with the Department of Electrical Communication Engineering, Utah State University. Logan, UT 84322-4120 USA (e-mail: regan.zane@
usu.edu).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TPEL.2013.2280077
0885-8993 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
LOPEZ et al.: UNIVERSAL DIGITAL CONTROLLER FOR BOOST CCM POWER FACTOR CORRECTION STAGES
3819
Fig. 3. Input voltage and power range of the recent works in sensorless PFC
controllers. The green area represents the goal of this study.
Fig 4. (a) Basic model of the boost converter and (b) Boost converter current
estimator.
3820
TABLE I
LIST OF CORRESPONDENCE BETWEEN THE ANALOG
AND THE DIGITAL VARIABLES
TABLE II
EXPRESSION OF THE CURRENT RIPPLE FOR THE REAL
AND REBUILT INPUT CURRENT
Fig. 5. Digital signal ire b [k], compared with the analog real input current ig ,
under ideal conditions. The on-off signal is the output of the digital device and
v L the analog inductor voltage.
qo = q (1 o )
(2)
where g and o are the per unit error of the input and output
analog-to-digital conversions, respectively.
Theoretically, the inductance L is known, but tolerances, temperature, switching frequency, and the inductor current cause a
difference between the estimated inductance (Lest ) and the real
one that depends on the core material. The inductor is modeled
as an integrator with a gain equal to the inverse of its inductance, whose output is the digitally rebuilt (estimated) input
current ireb . This signal ireb is used in the PFC current loop
instead of the real ig .
Table I shows the correspondence between the analog variables (actual variables in the converter) and their corresponding
digitally estimated variables expressed in the same units. These
variables define ig and ireb according to the expressions presented in Table II, where ig and ireb are the peak-to-peak
current ripple of the real and rebuilt input current, respectively.
If the analog and digital variables are equivalent, with vg =
vg q, vo = vo q, being q = qg = qo , L = Lest and ton = ton , then
both currents agree (ireb = ig ) when neglecting quantization
effects. The waveforms are as depicted in Fig. 5, where ireb [k]
represents the estimated current in the clock cycle k, and i
reb [j]
and i
g (jTsw ) are the estimated and real current at the end
of the switching cycle j (valley values), respectively. In this
situation, there is no current estimation error, defined as the
difference between ig and ireb , ierror = ig ireb , and therefore
ireb corresponds with an accurate quantization of ig .
At the beginning of the line voltage semicycle, ireb = ig = 0.
Since the input current is not measured, the line zero crossings
are the only points where the real current is known. Small errors
in the digital variables compared with the analog (see Table I)
cause a current estimation error in each switching period j. The
dominant current estimation errors are due to:
1) voltage data errors due to the tolerances of the voltage
dividers, noise, offset, quantization process, or nonlinearity of the ADCs (vg = vg q and/or vo = vo q and/or
qg = qo = q);
2) the difference between the estimated inductance (Lest )
and the real one (L);
3) the influence of the parasitic elements (RL , Ron , RD ,
VD );
4) delays in the drive signals (ton = ton ).
All of these errors are described separately in Section III,
where the current estimation error caused by these different
situations is modeled.
III. MODELING THE CURRENT ESTIMATION ERRORS
To simulate the effect of the different causes of error and
perform a first validation of the model, the system has been
modeled with the switching converter in PLECS and the behavioral control algorithm in Simulink as shown in Fig. 6.
The real input current ig and estimated input current ireb are
solved for CCM operation according to Fig. 7 and the current
estimator presented in Figs. 4 and 6. Analytical expressions
i
g (jTsw ) = ig ((j 1) Tsw ) +
i
reb
[j] =
i
reb
Tsw
[j 1] +
L
Tsw
{vg [j] vo [j] d [j]}
L
(3)
vg [j] q vo [j] q
d [j]
qg
qo
(4)
LOPEZ et al.: UNIVERSAL DIGITAL CONTROLLER FOR BOOST CCM POWER FACTOR CORRECTION STAGES
3821
Fig. 7. Digital estimated current ire b [k] compared with the analog real input
current ig under errors in data capture voltage across the inductance when
qg = qo = q. The on-off signal is the output of the digital device and v L the
analog inductance voltage.
Fig. 9. Digital estimated current ire b compared with the analog real input
current ig , when the estimated inductance value L e st is higher than the real
L. The on-off signal is the output of the digital device, and v L the inductance
voltage.
ierror [n] =
j =n
Tsw
q
q
vg [j] 1
vo [j] 1
d [j] .
L j =0
qg
qo
(5)
(6)
3822
Fig. 10. Steady-state simulated waveforms with the Simulink/PLECS model of the system with V g = 230 V rm s , V o = 400 V d c , P o = 640 W, fsw =72 kHz,
and C = 220 F, when q =qg =qo for (a) L = 1 mH and L e st = 1.8 mH, and (b) nonlinear inductance L = 1.50.147ig [mH], L e st = 1.0 mH.
i
g (jTsw ) = ig ((j 1) Tsw ) +
i
reb [j] = ireb [j 1] +
Tsw
{vg [j] vo [j] d [j]} (8)
L
Tsw
{vg [j] vo [j] d [j]} .
Lest
(9)
Comparing (8) and (9), the relation between ireb and ig yields
to expressions (10) and (11) for ig and ierror , respectively. As it
has been addressed before, ireb is the variable controlled by the
PFC control algorithm, so it has a sinusoidal shape (proportional
to the input voltage). Considering L constant over the line cycle,
ig is sinusoidal too, and no current distortion appears despite the
current estimation error, as it is presented in Fig. 10(a). Fig. 10(b)
shows the current waveforms with a nonlinear current dependent
inductance L(i) [23], as is the case of one of the inductors built
with a soft-saturation core used in the experimental results. The
model of soft-saturation inductor used in this study is presented
in [24], which considers the permeability dependence on the
magnetizing force
Lest
L
Lest
1 .
ig = ireb
ierror = ireb
Fig. 11.
(10)
(11)
(12)
d=
vg (RL + RD ) Rge VD vo
v
(Ron RD ) Rge VD vo
(14)
Defining vpar as the average voltage drop across the parasitic
elements, in each switching period
vpar = RL ig Ron ig d RD ig (1 d)
VD (1 d) = vo (1 d) vg
(15)
LOPEZ et al.: UNIVERSAL DIGITAL CONTROLLER FOR BOOST CCM POWER FACTOR CORRECTION STAGES
3823
Fig. 14. Auxiliary circuit used to adapt the drain-to-source voltage as a digital
signal. Comparison with the on-off signal and the drive signals delays to n -o
and to n -o .
j
=n
j =1
vo [j]
ton [j] .
L
(17)
Fig. 13. Digital signals estimated current ire b [k] compared with the analog
real input current ig , with on-time modification to n for the switching periods
j and j+1. The on-off signal is the output of the digital device and the inductor
voltage, v L .
Fig. 12 shows ig and ireb waveforms when the estimated volt-seconds across the inductor is lower than the actual
one due to the noncompensation of the parasitic effects with
Vg = 230 Vrm s , Vo = 400 Vdc , Po = 640 W, fsw = 72 kHz
and reactive components L = 1 mH and C = 220 F. The parasitic elements are RL = 0.3 , RD = 0.2 , VD = 0.6 V, and
Ron = 0.18 . These values are obtained from the datasheets
of the RHRP860 Fairchild Power Diode and the IRFP27N60K
International Rectifier Power MOSFET, the switching devices
used in the laboratory prototype. It can be observed that the
current estimation error, ig is not sinusoidal, with PF = 0.770
and THDi = 44%.
An additional error in the estimated volt-seconds applied to
the inductor is caused by time errors and is addressed in [19],
where it is concluded that they are mainly due to the difference
between the ON-time applied in the real converter (ton ) and
the estimated ON-time, ton . The effect of this time-error is
illustrated in Fig. 13, defining ton [j] = ton [j] ton [j] as the
) which
the power MOSFET and obtain the digital signal (vds
indicates the real ONOFF transitions in the boost converter, as
is presented in Fig. 14.
(18)
3824
Fig. 15.
vdig =
=
1 vo (Ron + RL ) + VD Re vg (Ron RD )
qg
Re (Ron + RL )
(19)
Vo (Ron + RL ) + VD Re Vg (Ron RD )
.
Re (Ron + RL )
(20)
Fig. 17. Auxiliary circuit for the actual DCM detection and waveforms.
(a) Hardware architecture. (b) Circuit waveforms.
LOPEZ et al.: UNIVERSAL DIGITAL CONTROLLER FOR BOOST CCM POWER FACTOR CORRECTION STAGES
Fig. 19.
g
reb
TDCM
.
eDCM = TDCM
in Fig. 15)
voltage signal used to estimate the input current (vo2
has 14 bits (4 LSB added). The current error for a given vdig is
defined by
ierror [n] =
Recent works [12][17], [19] avoid the input current measurement and propose a PFC digital control that includes the
measurement of the parasitic elements (RL , VD , RD , Ron ) and
applies a duty cycle command d, according to these elements, or
simply neglects their influence. But parasitic elements influence
change with the temperature, frequency, and the components
used in the PFC converter. It can be observed that in these
previously proposed solutions for sensorless PFCs use high inductance values and low switching frequencies in comparison
with the state-of-the-art of CCM PFCs that include a current
sensor.
reb
,
The estimated input current ireb , presents a DCM time, TDCM
g
while ig has a different DCM time, TDCM . A distortion in ig
g
reb
leads to TDCM
= TDCM
reducing the power factor. The compensator for current estimation error captures DCMig and DCMireb
g
reb
and measures and compares TDCM
with TDCM
to obtain the
DCM time error eDCM ,
(21)
Thus, an indirect measurement of the current estimation error is obtained by eDCM . The output voltage loop assures the
desired output voltage vo , and depending on the vdig value, two
different situations, shown in Fig. 18, can occur. If eDCM < 0
(see Fig. 18(a)), then ig < ireb, and it is necessary to increase
vdig to match DCM times. On the other hand, in Fig. 18(b), the
3825
j =n
Tsw
{(qg vdig ) d [j]}
L j =0
(22)
which is zero if vdig = /qg . Due to the finite resolution, the digital controller sets a value vdig = /qg 1 LSB. This 1 LSB
uncertainty in vdig represents an accumulated current error in the
last switching period nu of the half line cycle of, ierror [nu ] =
0.15 A, with the parameters previously presented. The new
feedback loop has a resolution of one order magnitude higher
that the feedforward compensation, resulting in ierror [nu ] =
1.40 A, limited by the digital device clock-period as presented
in detail in [22].
The output voltage loop in a PFC controller has a typical
bandwidth around fc = 510 Hz to set vo according to a reference. To not interfere with this, the DCM time compensation
loop has a lower bandwidth, using a sample period equal to Ts =
1/fc . With this assumption, a complete derivation of the model
is presented in detail in [30], where the model of the system
(plant) is given by
2
1
Mg2
K
(23)
and Mg = Vg 2 Vo and K = 2fsw L/R. Using a PI compensator, KPI /(z 1), the Jury stability criterion defines the range
of values of KPI for a stable system as
g
(s)
TDCM
1
Vo L (fu )2
= z 1 with =
vdig (s)
(24)
3826
Fig. 22. Experimental results for the DCM condition detection circuit for the
real input current.
Fig. 20. (a) Schematic diagram of the Boost PFC Converter. (b) Laboratory
prototype.
Fig. 23. Comparison of the harmonic content with the limits defined by the
Standard IEC 61000-3-2 for Classes A, C, and D.
TABLE III
POWER FACTOR AND THDI FOR DIFFERENT CONDITIONS
Fig. 21. Experimental results: ON-time (top) and the duty cycle modification
(to n ) due to the drive signal delays acquisition (bottom) over the half line
cycle.
LOPEZ et al.: UNIVERSAL DIGITAL CONTROLLER FOR BOOST CCM POWER FACTOR CORRECTION STAGES
3827
Fig. 24. Experimental results. (a) V o = 400 Vd c , L1 = 1 mH and R L 1 = 0.25 . Left: V g = 230 V rm s (50 Hz), Pg = 970 W. Right: V g = 85 Vrm s (60 Hz),
Pg = 320 W. (b) V o = 400 V d c , L2 = 1.5 mH and R L 2 = 0.35 . Left: V g = 230 Vrm s (50 Hz), Pg = 970 W. Right: Vg = 85 Vrm s (60 Hz), Pg = 320 W.
3828
The single digital signal acquired from the MOSFET drain-tosource voltage drop is used by both the feedforward and feedback compensators. The feedforward one represents a coarse
compensation of current estimation errors due to time delays.
And the new feedback loop generates a constant digital signal
to compensate current estimation errors, modifying the output
voltage measurement used to estimate the input current, and
minimizes this DCM time error. This feedback loop autotunes
the value of the digital signal when the converter operates in a
wide load or voltage range with a high resolution, resulting in
an average real input current proportional to the input voltage
in each switching period. A universal Boost PFC digital controller is achieved without current measurement, so in the point
of view of the designer the complexity of the PFC controller
decreases. With this feedback loop, parasitic element values do
not need to be measured and are compensated for automatically,
representing a step forward in comparison with the previous
works about PFC sensorless controllers. The prototype has been
tested with two different inductors (linear and nonlinear), and
the experimental results show, in both cases, high power factor with reliable performance under different load and voltage
conditions.
REFERENCES
[1] B. Mammano, Current sensing solutions for power supply designer, in
Proc. Unitrode Power Supply Seminar SEM-1200, 1997.
[2] P. Midya, P. Klein, and M. F. Greuel, Sensorless current mode controlAn observer-based technique for DCDC converters, IEEE Trans. Power
Electron., vol. 16, no. 4, pp. 522526, Jul. 2001.
[3] Y. Qiu, X. Chen, and H. Liu, Digital average current digital average
current-mode control using current estimation and capacitor charge balance principle for DC-DC converters operating in DCM, IEEE Trans.
Power Electron., vol. 25, no. 6, pp. 15371545, Jun. 2010.
[4] M. Rodriguez, J. Sebastian, and D. Maksimovic, Average inductor current
sensor for digitally-controlled switched-mode power supplies, in Proc.
Energy Convers. Cong. Expo., 2010, pp. 780787.
[5] Z. Lukic, Z. Zhenyu, S. M. Ahsanuzzaman, and A. Prodic, Self-tuning
digital current estimator for low-power switching converters, in Proc.
Appl. Power Electron. Conf., Feb. 2008, pp. 529534.
[6] B. A. Mather and D. Maksimovic, A simple digital power-factor correction rectifier controller, IEEE Trans. Power Electron., vol. 26, no. 11,
pp. 919, Jan. 2011.
[7] S. M. Park, Y. D. Lee, and S. Y. Park, Voltage sensorless feedforward
control of a dual boost PFC converter for battery charger applications, in
Proc. Energy Convers. Cong. Expo., 2011, pp. 13761380.
[8] Z. Yang and C. Paresh, A novel technique to achieve unity power factor
and fact transient response in AC-to-DC converters, IEEE Trans. Power
Electron., vol. 16, no. 6, pp. 764775, Nov. 2001.
[9] M. Rodriguez, M. V. M. Lopez, F. J. Azcondo, J. Sebastian, and
D. Maksimovic, Average inductor current sensor for digitally controlled
switched-mode power supplies, IEEE Trans. Power Electron., vol. 27,
no. 8, pp. 37953806, Aug. 2012.
[10] J. W. Shin, B. H. Cho, and J. H. Lee, Average current mode control
in digitally controlled discontinuous-conduction-mode PFC rectifier for
improved line current distortion, in Proc. Appl. Power Electron. Conf.
Expo., 2011, pp. 7177.
[11] Y. S. Roh, Y. J. Moon, J. C. Gong, and C. Yoo, Active power factor
correction (PFC) circuit with resistor-free zero-current detection, IEEE
Trans. Power Electron., vol. 26, no. 2, pp. 630637, Feb. 2011.
[12] K. I. Hwu, H.W. Chen, and Y. T. Yau, Fully digitalized implementation
of PFC rectifier in CCM without ADC, IEEE Trans. Power Electron.,
vol. 27, no. 9, pp. 40214029, Sep. 2012.
[13] W. Zhang, G. Feng, Yan-Fei Liu, and B. Wu, A digital power factor
correction (PFC) control strategy optimized for DSP, IEEE Trans. Power
Electron., vol. 19, no. 9, pp. 14741485, Nov. 2004.
[14] W. Zhang, G. Feng, Yan-Fei Liu, and B. Wu, A digital power factor
correction (PFC) control strategy for power factor correction and FPGA
LOPEZ et al.: UNIVERSAL DIGITAL CONTROLLER FOR BOOST CCM POWER FACTOR CORRECTION STAGES
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
implementation, IEEE Trans. Ind. Electron., vol. 21, no. 6, pp. 1745
1753, Nov. 2006.
H.-C. Chen, Duty phase control for single-phase boost-type SMR, IEEE
Trans. Power Electron., vol. 23, no. 4, pp. 19271934, Jul. 2008.
H-C Chen, Single-loop current sensorless control for single-phase boosttype SMR, IEEE Trans. Power Electron., vol. 24, no. 1, pp. 163171,
Jan. 2009.
H-C Chen, C-C Lin, and J-Y Liao, Modified single-loop current sensorless control for single-phase boost-type SMR with distorted input voltage,
IEEE Trans. Power Electron., vol. 26, no. 5, pp. 13221328, May 2011.
Fairchild Semiconductor. (2012, Dec. 5). Power factor correction. [Online]. Available: Internet: http://www.fairchildsemi.com/products/pfc/
F. J. Azcondo, A. de Castro, V. M. Lopez, and O. Garcia, Power factor
correction without current sensor based on digital current rebuilding,
IEEE Trans. Power Electron., vol. 25, no. 6, pp. 15271536, Jun. 2010.
V. M. Lopez, F. J. Azcondo, F. J. Diaz, and A. de Castro, Autotuning
digital controller for current sensorless power factor corrector stage in
continuous conduction mode, in Proc. Control and Modeling for Power
Electron., 2010, pp. 18.
V. M. Lopez-Martin, F. J. Azcondo, and A. de Castro, Current error compensation for current-sensorless power factor corrector stage in continuous
conduction mode, in Proc. Control and Modeling for Power Electron.,
2012, pp. 18.
V. M. Lopez-Martin, F. J. Azcondo, and A. de Castro, High-resolution error compensation in continuous conduction mode power factor correction
stage without current sensor, in Proc. Power Electron. Motion Control
Conf., 2012, pp. 18.
Magnetics powder core catalog. (2011). [Online]. Available http://www.
mag-inc.com/File%20Library/Product%20Literature/Powder%20Core%
20Literature/2011-Magnetics-Powder-Core-Catalog-2012-Update-.pdf
V. M. Lopez, A. Navarro-Crespin, R-Schnell, C. Branas, F. J. Azcondo,
and R. Zane, Current phase surveillance in resonant converters for electric discharge applications to assure operation in zero-voltage-switching
mode, IEEE Trans. Power Electron., vol. 27, no. 6, pp. 29252935, Jun.
2012.
R. W. Erickson and D. Maksimovic, Fundamentals of Power Electronics,
2nd ed. Berlin, Germany: Springer Science + Business Media Inc, 2001.
J. Morroni, L. Corradini, R. Zane, and D. Maksimovic, Adaptive tuning
of switched-mode power supplies operating in discontinuous and continuous conduction modes, IEEE Trans. Power Electron., vol. 24, no. 11,
pp. 26032611, Nov. 2009.
Fu-Zen Chen and D. Maksimovic, Digital control for improved efficiency
and reduced harmonic distortion over wide load range in boost PFC rectifiers, IEEE Trans. Power Electron., vol. 25, no. 10, pp. 26832692, Oct.
2010.
K. De Gusseme, D. M. Van de Sype, K. De Gusseme, D. M. Van de
Sype, A. P. M. Van den Bossche, and J. A. Melkebeek, Input-current
distortion of CCM boost PFC converters operated in DCM, IEEE Trans.
Ind. Electron., vol. 54, no. 2, pp. 858865, Apr. 2007.
D. Maksimovic, Y. Jang, and R. Erickson, Nonlinear-carrier control for
high-power boost rectifiers, IEEE Trans. Power Electron., vol. 11, no. 4,
pp. 578584, Jul. 1996.
V. M. Lopez-Martn and F. J. Azcondo, Modeling of a high resolution
DCM times feedback loop for sensorless boost PFC stages, in Proc.
Control and Modeling for Power Electron., 2013, pp. 16.
Victor Manuel Lopez (S10) was born in Torrelavega, Cantabria, Spain, in 1985. He received the
Electronics and Control Engineering degree from
the University of Cantabria, Santander, Spain, in
2009, where since then he has been working toward
the Ph.D. degree in the Department of Electronics
Technology, Systems and Automation Engineering
(TEISA).
From June to September 2012, he was a Visiting Scholar in the Colorado Power Electronics Center (CoPEC), University of Colorado at Boulder, and
from September to December 2012 in the Utah State University Power Laboratory, at Logan, Utah. Both periods with the Prof. Regan Zane as advisor. His
research interests include design, modeling, and digital control of topologies for
acdc converters and for resonant inverters.
Mr. Lopez received the IEEE/IEL Electronic Library Award in 2009.
3829
Francisco J. Azcondo (S90M92SM00) received the electrical engineering degree from the Universidad Politecnica de Madrid, Madrid, Spain, in
1989, and the Ph.D. from the University of Cantabria,
Spain, in 1993.
From 1995 to 2012, he has been an Associate
Professor in the TEISA Department, ETS II y T, University of Cantabria, Cantabria, Spain, where he has
been a Professor since 2012. He has been a Visiting
Researcher at the Department of Electrical, Computer, and Energy Engineering, University of Colorado, Boulder, CO, USA, from Feberuary to August 2004 and 2010, the Department of Electrical Communication Engineering, the University of Toronto,
ON, Canada, during the summer 2006, and the Power Electronics Lab (UPEL)
at Deptartment Electrical Communication Engineering, Utah State University,
Logan, UT, USA, from June to September 2013. Since Feberuary 2004, he is
Special Member (appointed as graduate faculty) at the Department of Electrical,
Computer, and Energy Engineering, University of Colorado Boulder, renovated
until spring 2017. He has been the Chair of the IEEE IES PELS Spanish
Joint Chapter from September 2008 to June 2011. His research interests include
modeling and control of switch-mode power converters and resonant converters,
digital control capabilities for SMPS, current sensorless control for power factor
correction stages and applications such as outdoor lighting, electrical discharge
machining, and arc welding.