Literature Review Report 2 Overview of Power Factor Correction
Literature Review Report 2 Overview of Power Factor Correction
Literature Review Report 2 Overview of Power Factor Correction
net/publication/241889542
Article
CITATIONS READS
0 2,975
1 author:
Grace Chu
Guina Research and Development
16 PUBLICATIONS 187 CITATIONS
SEE PROFILE
All content following this page was uploaded by Grace Chu on 11 August 2016.
10 JAN 2006
Abstract
In this report, we will review the fundamentals of power factor correction. We will also introduce a
power factor corrector using the most popular topology, the boost PFC pre-regulator. Then a simulation
model of a boost PFC pre-regulator by Dranga and Tse[3], together with some experimental verification
result will also be presented.
1 Introduction
The power factor correction and the reduction of the line harmonics becomes more and more important as
the amount of line-fed electronic equipment grows and the international regulation requirements are tight-
ened.
For many line-fed equipments, such as the AC/DC switching mode power supply, the input rectifier frontend
draws narrow pulsating currents form the AC line only at the peak of the supply voltage. The current pulses
contain large-amplitude harmonics and produce many undesirable effect on the AC line. They will cause
additional transformer and distribution loss, which will be dissipated in the transformer and distribution line
as heat, which is a sort of energy wastage. Moreover, the odd harmonic components result in a compensating
current flow in the neutral line of a three-phase, four-wire distribution system and produce clipping in AC
line voltage.
Besides, many countries have proposed and tightened regulations to further cut down such power line pol-
lution. The regulation IEC1000-3-2, which in 1998 was a generally accepted standard to set the limit of line
harmonics produced by different categories or classes of electronic equipments.
The above reasons make power factor correction become a ”evergreen” research topics no matter in academic
or industries over so many years. Different kinds of innovative topologies and control methods have been
introduced to further improve the performance in harmonics reductions, efficiency and stability.
This literature review report mainly gives an overview on some fundamental concepts of power factor cor-
rection. First of all, the definition of power factor will be given and we will discuss the common power factor
problem in single stage switching mode power supply. Then the passive and active power factor correction
will be introduced, in which we will review some basic topologies of PFC corrector. Following this we will go
through some common active current control schemes currently used in industry. To give more details, the
stability analysis and the bifurcation behavior in PFC are also discussed. A boost PFC simulation model
is presented following with the experimental verification to identify the stability boundaries with different
circuit parameters. In the last session, a conclusion will be given as well as the future plan of research work
to follow up the related issue.
1
2 Definition of Power Factor
In terms of strictly passive reactive load, power factor is defined as the cosine of the phase angle θ between the
voltage and current waveform. For an rms input voltage Vrms and rms input current Irms , the total power
taken from the line is Vrms Irms , which is referred as ”apparent power”. But the actual power consumed by
the load is only Vrms Irms cosθ. It means only the component of input current which is in phase with the
voltage across the load resistance Irms cosθ contributes to the load. This actual power is also referred as
”real power”. This brings an alternative way to express power factor as a ratio of real power to apparent
power.
Realpower
P owerf actor = (1)
Apparentpower
Vrms Irms cosθ
= (2)
Vrms Irms
= cosθ (3)
Phase shift
wt
Real power
(W)
(a) (b)
Figure 1: (a) Line voltage and current at the input of a reactive load. (b) Vector diagram showing the
apparent power exceed the resistive and reactive power
However, the above stated definition is only valid on condition that both the input current and input voltage
are ideal sinusoidal waveform, which is not the case in most real-world power supplies.
When the input current or input voltage is non-sinusoidal, power factor consists of two factors:
1. The displacement factor measures the phase angle, P Fdisp
2. The distortion factor measures the waveform shape, P Fdist
As mentioned before,
while
1
P Fdist = p (5)
(1 + T HD2 )
1
= p (6)
1 + (I2/I1)2 + (I3/I1)2 + (Ii/I1)2 + ......
2
Where THD stands for ”Total harmonic distortion” Ii is the rms of the ith harmonic input current.
Power factor is a useful measurement of power quality. It can vary between 0 and 1. when the current and
voltage waveforms are in phase and of ideal sinusoidal shape, the power factor is 1 (unity). However, Redl
[2] pointed out the truth that high power factor (close to 1) can still be obtained even with high current
distortion. The purpose of making the power factor equal to unity is to make the circuit look purely resistive
(apparent power equals to real power).
i
C
Load
Line AC
50Hz
Large capacitor
wt
(a) (b)
Figure 2: (a) A typical rectifier input stage with large input capacitor for a direct off-line switching mode
power supply. (b) Rectifier output voltage and current with large input filter capacitor
3
many disadvantages: They are relatively bulky. Also they are sensitive to line frequency and may produce
excessive phase shift. Moreover, they lack voltage regulation and the dynamic response is poor.
L
Vo
Line AC i
50Hz
Load
4
requirement. For example, keeping S1 ON and only operates S1 turn the corrector into boost mode, which
can be used to increase efficiency during heavy load condition. While buck mode (keeping S2 open and only
operates S1) is used intermittently during start-up. The buck-boost PFC corrector and its two-switches
version are shown in Fig. 5 and Fig 6 respectively.
D Vo
Line AC SW1
50Hz
L Load
C
Vo
SW1 L D2
Line AC D1
50Hz C
SW2 Load
Vo
L D
Line AC
50Hz C
SW1 Load
5
Vo
Line AC L D2
50Hz
C Load
SW
Vo
Line AC
50Hz
C Load
L
SW2
SW1
2L
Rin,buck−boost = (9)
d2 T
Although buck-boost corrector is perfect for power factor correction due to its low current distortion, it
exhibits a higher peak current stress in the switching components and yields a lower efficiency. Boost
converter thus become a more favorable choice based on the overall performance and we will go into detail
in later sessions.
6
6 Active Control of Line Current
Active control of line current means forcing or programming the input current to follow the envelope of the
sinusoidal input voltage waveform. We will introduce several common control methods with controller ICs
available in the industry. The general active current control scheme are shown in Fig. 10 below.
7
Vo
L D
Line AC
50Hz SW1 C
Load
Logic and
Driver
Voltage
Amplifier
Mulitiplier
Current
Amplifier/ Z= A
comparator AB/
C B
C
Ref
(a) (b)
(c) (d)
Figure 11: The inductor current waveform using various control methods
8
period doubling among switching cycles. (several hundreds KHz) While instability of the voltage loop results
in slow-scale bifurcation, which shows the phenomenon of period doubling among line cycles (50/60Hz).
Instability or bifurcation behaviors have been reported in boost PFC and quite a number of research and
publications have been focusing on the stability analysis of the boost PFC. R. Ridley [11]provided a small
signal model of boost PFC. Mazumder [12] developed a state-space model for boost PFC to analyze the
converter stability in saturated and unsaturated regions using Lyapunov method and bifurcation analysis.
Dranga and Tse [13] also demonstrated the fast-scale current loop bifurcation by simulation in boost PFC
with peak current mode control. According to their result, duty cycle and the power converter’s steady state
gain (ratio of output voltage)and circuit parameter like inductor and the load resistance, etc affects stability.
The stable/unstable regions have be identified, as well as the relationship of the critical phase angle and the
dc converter gain.
For slow-scale bifurcation, Wong, Tse and Orabi [14] developed an average boost PFC model by double
averaging and derived the slow-scale instability boundaries with different circuit parameters. Experiment
was conducted and succeeded in verifying some of the stability boundaries derived. Based on their results,
the PFC ’s stability becomes worse with higher output voltage to input voltage ratio, higher voltage loop dc
gain, smaller voltage loop time constant and/or smaller output capacitor value.
Figure 12: Slow-scale bifurcation or period doubling demonstrated by the simulation model[left] and the
enlarged waveform [right]
One of the curves is repeated with the parameters shown in Table 1. The resulting plot shows some deviation
from the initial results, especially in the high power region where almost all bifurcation seem to disappear.
Further work need be done to investigate whether such deviation is due to some deficiency in the simulation
model or some complex circuit behavior during high power stage. The resulting plots are shown in Fig 15.
9
Figure 13: Simulink model of boost PFC with average current mode control
18
2.2
UNSTABLE OPERATION STABLE OPERATION
17
2.1
τF2=1ms
boost stage feedback gain GF1
in,rms
16
slow−scale 2
15 /V
period−doubling PFC,ref
τF2=2ms
14 1.9
slow−scale τF2=3ms
voltage ratio V
13
period−doubling
1.8
12
GF2=60 G =80 1.7
F2
11
1.6
10
GF2=100 1.5
9
STABLE OPERATION UNSTABLE OPERATION
8 1.4
50 60 70 80 90 100 110 120 130 140 150 50 60 70 80 90 100 110 120 130 140 150
power (W) power (W)
Figure 14: Initial plots of stability boundary of a single-stage boost PFC and its cascaded system with a
buck converter
9 Experimental Verification
Based on the simulation model, a single stage boost PFC pre-regulator with average current mode control
is implemented with controller IC UC3854. The circuit schematics is shown in Fig. 16 and the experimental
result are shown in Fig. 17.
Based on Wong and Tse ,2002 [14], the gain in the experimental circuit is defined by:
Rvf 100
GF = · (10)
Rvi 2.6
= K · GF 1 (11)
(12)
Where K is an unknown scaling factor between the transfer function’s gain and the real circuit’s loop gain.
Similar to the simulation result, the unstable region mainly located at the low power region and the converter
becomes stable as output power increased up to 100W around.
10
Table 1: Parameter of the boost PFC pre-regulator model.
Parameters Values
Rectified line voltage Vin 70 Vrms/ 50 Hz
Boost inductance L1 1 mH
Boost capacitance C1 60 µF
Boost switching period Ts1 10µs
Reference PFC output voltage VP F C,ref 240 V
PI controller gain k 5.12
PI controller time constant Ti 0.12 ms
Voltage controller time constant TF 1 8.6 ms
16
GF1
14
12
10
8
40 60 80 100 120 140
Power /W
Figure 16: Schematics of the boost PFC pre-regulator for experimental verification
11
Stability region of single−stage boost PFC by experiment
20
Stable
Unstable
18 Stable/Unstable
16
14
GF
12
10
4
40 50 60 70 80 90 100 110 120 130 140 150
Power /W
10 Conclusion
In this report, we have reviewed some fundamental concepts of PFC. Firstly, the importance of power factor
correction in the society and the industry has been mentioned and the definition of power factor is presented.
Then we have introduced some basic PFC converter topologies such as buck, boost and buck boost, with
their advantages and disadvantages compared, as well as a few innovative topologies from the variation of
the basic topologies. Besides, common control schemes for input current shaping has been reviewed. In later
sessions, we have discussed the stability analysis and bifurcation behavior of a boost PFC. A Simulink model
has been presented and used to identity the stability boundary of a boost PFC with average current mode
control. The simulation is repeated and the resulting plot matches with the initial findings by Dranga [3]
at the low power region; while large deviations found at the high power region. A single-stage boost PFC
circuit is implemented with UC3854 controller IC for experimental verification. The experimental result has
been presented and it matches with the simulation result.
Further work need to be done to verify the stability boundaries of a single-stage boost PFC and its cascaded
system with a forward converter with transformer isolation. For the single-stage boost PFC pre-regulator,
simulations need to be repeated to re-identify the slow-scale period doubling region. At the meantime, a
new single-stage boost PFC pre-regulator needs to be built in order to verify the simulation result in high
power region. The current challenges are the co-relation of transfer function’s dc gain in the simulation
model with the dc gain of the error amplifier in the real circuit, as well as how to change the output/input
voltage ratio in the real circuit while keeping the dc gain constant. For the cascaded system, similarly,
simulation can be repeated with the original model first to verify the initial results. Then the model can
be adjusted (by adding transformer turn ratio) and compare it with the non-isolated version. Meanwhile a
transformer-isolated forward converter need to be built to verify the simulation result. With the addition
of the transformer, duty cycle can be adjusted by changing the transformer’s turn ratio, while keeping the
output voltage constant. Therefore, the effect of duty cycle on slow-scale bifurcation can be investigated
more accurately.
12
View publication stats
References
[1] C.Zhou, R. B. Ridley, F. C. Lee, ’Design and analysis of a htsteretic boost power factor correction
circuit’, in IEEE Power Electronics Specialists Conference, 1990, pp. 800-807.
[2] Richard Redl, ”Power-factor correction in single-phase switching-mode power supplies-an overview”,
Int. J. Electronics, 1994, Vol 77, No. 5, 555-582.
[3] Octavian Dranga, C. K. Tse, S. C. Wong, ”Stability analysis of complete two-stage power-factor-
correction power supplies”, ECCTD 2005.
[4] Kelley, A. W., Yadusky, W. F., ”Rectifier design for minimum line current harmonics and maximum
power factor.”, Proceeding of the Applied Power Electronics Conference, Baltimore, Maryland, USA,
pp. 13-22.
[5] Spangler, J., ”Proc. Sixth annual Appliced Power Electronics Conf., Dallas, 1991”, pp.10-15.
[9] C. K. Tse, ”Circuit theory of power factor correction in switching converters”, Int. J. Circ Theor. Appl.
2003; 31:157-198
[10] Lloyd Dixon, ”Average current mode control of switching power supplies”, Unitrode Application Note
U-140
[11] Raymond B. Ridley, ”A new-small signal model for current mode control”
[12] Sudip K. Mazumder, Ali H. Nayfeh, ” A novel approach to the stability analysis of boost power-factor-
correction circuits”
[13] Octavian Dranga, C. K. Tse, Herbert Iu, ”Bifurcation behavior of the power-factor-correction boost
converter”, International Journal of Bifurcation and Chaos, Vol. 13, No. 10 (2003) 3107(3114)
[14] M. Orabi, T. Ninomiya, S. C. Wong, C. K. Tse, ”Method of double averaging for modeling PFC switching
converter”, IEEE PESC 2004 pp.33202-3208
13