Power Factor Improvement of an AC-DC

Converter via Appropriate sPWM Technique

K. Georgakas, A. Safacas
University of Patras, Department of Electrical and Computer Engineering, Laboratory of Electromechanical Energy
Conversion, 26500 Rion-Patras, GREECE




Abstract The power control of a DC load can be achieved
via an AC-DC converter consisting of a rectifier bridge and
a switching element operating by sPWM technique. The use
of such a converter causes a lot of high harmonics at the AC
side, which reduce the power factor and distort the grid
voltage. Using passive filter in the converter input to avoid
the high harmonics consequences the power factor decrease.
To improve the power factor an appropriate sPWM
operation of the switching element is proposed in this paper.
The system behavior is studied through simulation and
experimental investigation. The power factor correction is
verified.
I. INTRODUCTION
It is well known that the power control of a DC load
feeding by the grid is achieved by the use of an AC-DC
converter structure operating through a sPWM technique.
In figure 1 one can see such a converter structure
consisting of a MOSFET single phase rectifier bridge in
series connected with a switching MOSFET5. In the case
of an ohmic inductive load a parallel freewheeling diode
is necessary. The rectification becomes by the parasitic
bridge MOSFET diodes, while the MOSFETs 1-4 enable
the power inversion, if an active load is considered.
AC Voltage
Source
L
f
C
f
Electrical
Load
MOSFET
1
MOSFET
2
MOSFET
4
MOSFET
3
D
o
MOSFET 5
A
B
.
.
Passive
Filter
R
L

Figure 1. An AC-DC converter structure for supplying a DC load.
The sPWM operation can be succeeded by comparison
of a sinusoidal voltage waveform (U
c
)in phase to the grid
voltage (U
g
) with a high frequency triangular waveform in
order to obtain a switching pulse waveform. The pulse
duration inside of a half sinusoidal period is not constant
and the pulse of the maximum duration is located exact at
the middle of the half period, while the pulse of the
minimum duration appears at the beginning of that, as it
appears in figure 2a. Figure 3 shows the waveforms of the
grid voltage (50Hz) and the corresponding current pulse
waveforms (switching frequency 5 kHz). In case of an
ohmic DC load the basic harmonic of the grid current
pulse waveform (fig.3a) is in phase with the grid
sinusoidal voltage waveform. If the DC load is ohmic-
inductive one, then the basic current harmonic is shifted in
relation to the voltage waveform U
g
(fig.3b). In the case
that a sinusoidal waveform U
c
is leading upon the grid
voltage U
g
by an angle a via comparison to the triangular
waveform (fig.2b), a grid current pulse waveform is
obtained of which the basic harmonic is shifted to the grid
voltage. In this way the grid current basic harmonic can
becomes in phase with the grid sinusoidal voltage, if we
have an ohmic-inductive DC load. It means that the power
factor can be corrected. In this paper an extensive
investigation of the influence of the leading or lagging
angle a to the power factor via simulation as well as
experimentally has been carried out.


2a


2b
Figure 2. Pulse waveforms obtained by sPWM when a=0
o
(2a)
and a0
o
(2b).





3a



3b
Figure 3. Grid voltage and current in the case of ohmic load (3a) and
ohmic-inductive load (3b).
Proceedings of the 15th Mediterranean Conference on
ControI & Automation, JuIy 27 - 29, 2007, Athens - Greece
T26-024
II. POWER FACTOR INVESTIGATION USING
SIMULINK/MATLAB SIMULATION
The power factor calculation is achieved by
Simulink/Matlab simulation using the appropriate models
for the system which is shown in figure 1. First, the
simulation has been carried out without input passive
filter for two different loads: a) ohmic load and b) ohmic
inductive load (several values), by a switching frequency
f
sw
=5 kHz. The power factor as a function of the angle a
mentioned above is depicted in the figures 4, 5 for an
effective load power of P=1200W. Figure 14 shows the
power factor in the case that a DC motor is used as a
load. In figure 15 is PF=f(a) by f
sw
=10 kHz. In figures
11 and 12 the waveforms of the ac input voltage and
current are shown. The figures 6, 7, 8, 9 and 10 show the
spectrum of the input current for characteristic values of
the angle a. In the case of ohmic load (R=20) the
power factor gets its maximum value (PF
max
=0,6548) by
a=0, which is relatively a low value because of the
great current high harmonic content. If the load has
ohmic-inductive character, the maximum PF value is
obtained by a negative value of angle a. This happens
because the control voltage U
c
mentioned above is
leading by the angle a upon the grid voltage U
g
in order
to achieve that the basic current harmonic is in phase with
the grid voltage. In this point it must be remarked that the
sPWM procedure, controlled by the signal U
c
, leads to
the correction of the power factor through the moving of
the current waveform to the left in the figure 12, what can
be easy shown if the figures 11 and 12. So, the
waveforms of u
in
(t) and i
in
(t) get more similar to the
waveform of figures 13. But, by moving of the waveform
i
in
(t) increase the high harmonic content, as it can be
shown in the figures 6, 7, 8, 9 and 10, which are getting
through FFT analysis of the input current waveform i
in
(t)
using the Origin software. It is obviously that the high
harmonic content of i
in
(t), for example, by a=0 is lower
than that by a=-45
o
(fig.7 and 10). Beginning from
a=0 and gradually going on to a=-45 the harmonics
of the 5
th
, 7
th
, 9
th
and 11
th
order increase.
- 2 0 - 1 5 - 1 0 - 5 0 5 1 0
6 5 , 4 7 0
6 5 , 4 7 2
6 5 , 4 7 4
6 5 , 4 7 6
6 5 , 4 7 8
6 5 , 4 8 0
P
F
[%
]
' a ' [
o
]
R = 2 0

Figure 4. Power factor (PF) as a function of the angle a by ohmic
load and switching frequency 5 kHz without input filter
(simulation results).
-70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25
78,0
78,5
79,0
79,5
80,0
80,5
81,0
81,5
82,0
82,5
83,0
83,5
84,0
84,5
85,0
P
F

[
%
]
'a' [
o
]
L=10mH, R=20
L=30mH, R=20
L=50mH, R=20
L=100mH, R=20
L=200mH, R=20

Figure 5. Power factor (PF) as a function of the angle a by ohmic-
inductive load as a parameter and switching frequency
5 kHz without input filter (simulation results).
0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 9 0 0 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0 0
0
1
2
3
F r e q u e n c y ( H z)
I


[
A
]
Figure 6. Grid current harmonic content for angle a=18

and ohmic-
inductive load without input filter (R=20, L=30mH)
(simulation results).

0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 9 0 0 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0
0
1
2
3
F r e q u e n c y ( H z)
I


[
A
]
Figure 7. Grid current harmonic content for angle a=0

and ohmic-
inductive load without input filter (R=20, L=30mH)
(simulation results).

0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 9 0 0 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0 0
0
1
2
F r e q u e n c y ( H z)
I


[
A
]
Figure 8. Grid current harmonic content for angle a=-9

and ohmic-
inductive load without input filter (R=20, L=30mH)
(simulation results).

0 100 200 300 400 500 600 700 800 900 10000 20000 30000 40000 50000
0
1
2
3
Frequency ( Hz)
I


[
A
]
Figure 9. Grid current harmonic content for angle a=-36

and ohmic-
inductive load without input filter (R=20, L=30mH)
(simulation results).

0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 9 0 0 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0 0
0
1
2
3
F r e q u e n c y ( H z)
I


[
A
]
Figure 10. Grid current harmonic content for angle a=-45

and
ohmic-inductive load without input filter (R=20,
L=30mH) (simulation results).

Proceedings of the 15th Mediterranean Conference on
ControI & Automation, JuIy 27 - 29, 2007, Athens - Greece
T26-024
0.20 0.21 0.22
-200
0
200
i
in
u
in
u
(
t
)

[
V
]
,




i
(
t
)
*
1
0

[
A
]
t [s]

Figure 11. Grid voltage and current waveforms for a=0
o
and ohmic-
inductive load (R=20, L=30mH) without input filter
(simulation results).

0.06 0.07 0.08
-400
-300
-200
-100
0
100
200
300
400
i
in
u
in
u
(
t
)

[
V
]
,




i
(
t
)
*
1
0

[
A
]
t [s]

Figure 12. Grid voltage and current waveforms for a=-45
o
and ohmic-
inductive load (R=20, L=30mH) without input filter
(simulation results).

0.20 0.22
-200
0
200
I
in
U
in
u
(
t
)

[
V
]
,




i
(
t
)
*
1
0

[
A
]
t [s]

Figure 13. Grid voltage and current waveforms for a=0
o
and ohmic
load (R=20) without input filter (simulation results).

-70 -65 -60 -55 -50 -45 -40 -35 -30 -25 -20
72
73
74
75
76
77




































P
F

[
%
]
a [
o
]
5 kHz
10 kHz

Figure 14. Power factor (PF) as a function of the angle a by DC
motor load and switching frequency 5 kHz and 10 kHz
without input filter (simulation results).
As mentioned above, the phase angle of the basic
current harmonic
1
can be defined through the angle a.
Using the values of the calculated power factor and of the
spectrum analysis the value of
1
can be calculated as
follow:
As the power factor of the basic harmonic is
2
k 2
1
1
I
cos =PF 1+THD THD=
I
, with



(1)
where I
1
= rms value of the basic current harmonic and

2 2 2 2
k 3 5 7
2
500
I = .... I I I I + + +

(2)

the rms value of the high harmonics, using the known
values of PF, I
1
and I
k
, the values of cos
1
and also
1
can
be calculated. The results of such a calculation are shown
in table I.
For example, for a=18, ohmic-inductive load
(R=20, L=30mH), f
sw
=5 kHz, U
g
=220 V, f
g
=50 Hz and
P=1200W, the following values have been obtain:
I
1
=3,56 A, I
3
=0,8918 A, I
5
=0,16 A, I
100
=1,112 A,
I
200
=0,534 A, I
300
=0,3 A, I
400
=0,2 A, I
500
=0,4 A,
PF=78,27 %,
2 2 2
k 3 5
2
2
500
1
I = .... 1,622531121 A,
THD=0, 455767168, cos 7827 1 0, 455767 =0,857 168 4 0,
I I I

+ + =
+ =

TABLE I.
CALCULATION OF THE COS1 USING THE SIMULATION RESULTS
(WITHOUT INPUT FILTER)
Angle a PF cos1 1 [

]
18

(lag) 0,7827 0,8574 30,9

9

(lag) 0,7928 0,8792 28,45

0

0,8009 0,8928 26,77

-9
0
(lead) 0,8069 0,8929 26,76
-18
0
(lead) 0,8108 0,8986 26,02
-36
0
(lead) 0,8128 0,9019 25,59
-45
0
(lead) 0,8113 0,9025 25,51
-54
0
(lead) 0,8085 0,9141 23,92
-63
0
(lead) 0,8042 0,8986 26,02

In the table I one can see that using the proposed
operation mode of the sPWM method, an improvement of
the power factor can be achieved. We remark that by
a=0 is PF=0,8009, while by a=-36 is PF=0,8128, it
means that an improvement of 1,48% has been succeeded.
Figure 15 shows the function PF=f(a) by switching
frequency f
sw
=10 kHz and P
1,2
=1000W, 1600W. In the
case of this switching frequency value the results are very
similar to the case of f
sw
=5 kHz.
If a passive filter with appropriate L-C values
(L=100mH, C=1F) at the converter input is used, the
simulation results of the system are those which are
depicted in the figures 16, 17, 18, 19, 20, 21 and 22. The
PF values are increased in all cases, so that an
improvement of about 10% is achieved (in comparison to
PF
max
in the two cases: with and without filter). The
spectrum analysis show that the high order harmonics
(f
k
>800 Hz) doesnt appear (fig.18,19 and 20). The current
waveforms are different compared to those in case that no
filter is used, as the figures 21 and 22 show.
Table II is obtained in similar way as table I. One can
remark that by leading angle a=-54 the current basic
harmonic is in phase with the grid voltage. The maximum
Proceedings of the 15th Mediterranean Conference on
ControI & Automation, JuIy 27 - 29, 2007, Athens - Greece
T26-024
value of PF is at a=-36, by f
sw
=5 kHz and P=1200W.
Comparing the PF values for a=0 and a=-36 one can
see that an improvement of 4,47% has been succeeded.

-50 -40 -30 -20 -10 0
80
81
88
89
90
1000 [W]
1500 [W]
P
F

[
%
]
degrees [
0
]

Figure 15. Power Factor (PF) as a function of the angle a for
switching frequency 10 kHz without input filter by the
output power as a parameter (simulation results).


-60 -50 -40 -30 -20 -10 0
90
91
92
93
94
95
96
97
98
P
F

[
%
]
'a' [
o
]
L=10mH , R=20
L=30mH , R=20
L=0mH , R=20
L=100mH, R=20
L=200mH, R=20
Figure 16. Power factor (PF) as a function of the angle a by
five values of ohmic-inductive load with input
filter and switching frequency 5 kHz
(simulation results).

- 9 0 - 80 - 7 0 - 60 - 5 0 - 40 - 3 0 - 20 - 1 0
9 1
9 2
9 3
9 4
9 5
9 6
P
F

[
%
]
' a ' [
o
]
1 2 0 0 W
5 0 0 W

Figure 17. Power Factor (PF) as a function of the angle a=0
o
for
switching frequency 10 kHz with input filter by output
power as a parameter (simulation results).


0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 4 8 0 0 5 0 0 0 5 2 0 0
0
1
2
3
F r e q u e n c y ( H z )
I


[
A
]

Figure 18. Grid current harmonic content for angle a=0

with input filter

and ohmic-inductive load (R=20, L=10mH) (simulation
results).

0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0 1 4 0 0 5 0 0 0
0
1
2
3
4
F r e q u e n c y ( H z )
I

[
A
]

Figure 19. Grid current harmonic content for angle a=-27

with input
filter and ohmic-inductive load (R=20, L=10mH)
(simulation results).

0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0 4 8 0 0 5 0 0 0 5 2 0 0
0
1
2
3
F r e q u e n c y ( H z )
I


[
A
]

Figure 20. Grid current harmonic content for angle a=-40,5

with input
filter and ohmic-inductive load (R=20, L=10mH
(simulation results).

1 . 2 0 1 . 2 1 1 . 2 2
- 2 0 0
0
2 0 0
i
i n
u
i n
u
(
t
)

[
V
]
,





i(
t
)
*
1
0

[
A
]
t [ s ]

Figure 21. Grid voltage and current waveforms for a=0
o
and ohmic-
inductive load (R=20, L=30mH) with input filter
(simulation results).

0 . 1 0 0 . 1 1 0 . 1 2
- 2 0 0
0
2 0 0
i
i n
u
i n
u
(
t
)

[
V
]
,




i(
t
)
*
1
0

[
A
]
t [ s ]

Figure 22. Grid voltage and current waveforms for a= -36
o
and ohmic-
inductive load (R=20, L=30mH) with input filter
(simulation results).

TABLE II.
CALCULATION OF THE COS1 USING THE SIMULATION RESULTS
(WITH INPUT FILTER)
Angle a PF cos1 1 [
o
]
0

0,8986 0,9334 21
-27
o
(lead) 0,934 0,985 9,93
-31,5
o
(lead) 0,9381 0,9866 9,39
-36
o
(lead) 0,9388 0,9871 9,21
-45
o
(lead) 0,9349 0,9973 4,21
-54
o
(lead) 0,9179
1 0


III. EXPERIMENTAL RESULTS
A MOSFET converter has been designed and
constructed in the laboratory and its operation was
Proceedings of the 15th Mediterranean Conference on
ControI & Automation, JuIy 27 - 29, 2007, Athens - Greece
T26-024
controlled by a microprocessor 80C196MC. This system
including an input filter (L3mH, C3F) has been used
for the experimental investigation. The experimental
results are depicted in the figures 23, 24, 25, 26, 27, 28
and 29.
-100 -80 -60 -40 -20 0 20 40 60
0,70
0,75
0,80
0,85
0,90
0,95
P
F

[
%
]
'a' [
o
]

Figure 23. Power factor (PF) as a function of the angle a for
ohmic-inductive load (R=180, L=30mH) by swi-
tching frequency 5 kHz with input filter
(experimental results).

-60 -40 -20 0 20 40 60
0,65
0,70
0,75
0,80
0,85
P
F

[
%
]
'a' [
o
]

Figure 24. Power factor (PF) as a function of the angle a for
ohmic-inductive load (R=180, L=30mH) and
switching frequency 2,5 kHz with input filter
(experimental results).
-60 -40 -20 0 20 40 60
0,70
0,75
0,80
0,85
0,90
0,95
1,00
P
F

[
%
]
'a' [
o
]

Figure 25. Power factor (PF) as a function of the angle a for
ohmic-inductive load (R=180, L=100mH) by swi-
tching frequency 5 kHz with input filter
(experimental results).
-60 -40 -20 0 20 40 60
0,75
0,80
0,85
0,90
0,95
1,00
P
F

[
%
]
'a' [
o
]

Figure 26. Power factor (PF) as a function of the angle a for
ohmic-inductive load (R=180, L=100mH) by swi-
tching frequency 10 kHz with input filter
(experimental results).

-40 -30 -20 -10 0 10 20
0,94
0,95
0,96
0,97
0,98
0,99
1,00
1,01
P
F

[
%
]
'a' [
o
]
10 kHz
5 kHz

Figure 27. Power factor (PF) as a function of the angle a for
ohmic load R=30 and by switching frequency
5 kHz with input filter (experimental results).


Figure 28. Grid voltage and current waveforms for a=0
o
, ohmic
load (R=180) and switching frequency 10 kHz
(experimental results).


Figure 29. Grid current waveform for a=0
o
, ohmic-inductive
load (R=180, L=30mH) and switching freque-
ncy 2,5 kHz (experimental results).

In general, the differences between simulation and
experimental results are small. In all cases the power
factor has high values (0,850,99) depended on the
switching frequency, the load R-L values, the output
power and the L-C values of the input filter.
IV. CONCLUSIONS
The simulation and experimental results show that there
is a leading angle a by which the power factor becomes
maximum. The value of this angle depends on the nature
of the load, the output power, the input filter and the
switching frequency. A sinusoidal signal (voltage U
c
)
created by microprocessor and leading upon the sinusoidal
grid voltage determines the sPWM converter operation
and so the appropriate value a can be achieved. The
target is to shift the grid current waveform relatively to the
Proceedings of the 15th Mediterranean Conference on
ControI & Automation, JuIy 27 - 29, 2007, Athens - Greece
T26-024
grid voltage in order to be the basic current harmonic in
phase with the grid voltage.
REFERENCES
[1]. K.Georgakas, A. Safacas, Power Factor Correction and
Efficiency Investigation of AC-DC Converters Using Forced
Commutation Techniques, ISIE, Congress, Dubrovnik, June 20-
23, 2005.
[2]. K.Georgakas, A. Safacas, Efficiency and Power Factor
Investigation of Characteristic Converter Topologies via
Simulation, ICEMS, Conference Nanjing, China, September 27-
29, 2005, Proceedings, pp. 1422-1427.
[3]. K.Georgakas, A. Safacas, I. Georgakopoulos, Determination of
the Optimum Power Factor and Efficiency Values of a Single
Phase Converter Supplying a DC Drive via Simulation, ICEM,
Conference Chania, Greece, September 2-5. 2006, Proceedings,
paper No.307.
[4]. K. Georgakas, A. Safacas, I. Tsoumas, Current high harmonics
investigation of an AC-DC-AC converter consisting of high
frequency semiconductor elements supplying a DC machine,
Electromotion 2005 , 27 29 September, 2005, Lausanne,
Switzerland, proceedings, paper DS2/6.
[5]. Nabil A. Ahmed, Kenji Amei, Masaaki Sakui, AC chopper
voltage controller-fed single-phase induction motor employing
symmetrical PWM control technique, Elsevier Sciences, Electric
Power System Research, 55(2000), pp. 16-20.
[6]. Nabil A. Ahmed, Kenji Amei, and Masaaki Sakui, A New
Configuration of Single-Phase Symmetrical PWM AC Choper
Voltage Controller, IEEE TRANSACTION ON INDUSTRIAL
ELECTRONICS, VOL.46, NO.5, October, 1999, pp.942-952.
[7]. BOR-REN, High Power Factor AC/DC/AC Converter With
Random PWM, IEEE TRANSACTIONS ON AEROSPACE
AND ELECTRONIC SYSTEMS, VOL.35, NO.3 July 1999, pp.
935-943.
[8]. Jose R. Rodriguez, Juan W. Dixon, Jose R. Espinoza, Jorge Pontt,
Pablo Lezana, PWM Regenerative Rectifiers: State of the Art,
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS,
VOL.52, NO.1 FEBRUARY 2005.
[9]. Bhim Singh, Brij N. Singh, Ambrish Chandra, Kamal Al-Haddad,
Ashish Pandey, Dwarka P. Kothari, A Review of Single-Phase
Improved Power Quality AC-DC Converters, IEEE
TRANSACTIONS ON POWER ELECTRONICS, VOL.50 NO.5,
OCTOBER 2003.
[10]. Ramesh Srinivasan, Ramesh Oruganti, A Unity Power Factor
Converter Using Half-Bridge Boost Topology, IEEE
TRANSACTIONS ON POWER ELECTRONICS, VOL.13,
NO.3, MAY 1998.
[11]. Hideaki Fujita, Hirofumi Akagi, Shin-ichi Shinohara, A 2 - MHz
6-kVA Voltage-Sourse Inverter Using Low-Profile MOSFET
Modules For Low-Temperature Plasma Generators, IEEE
TRANSACTIONS ON POWER ELECTRONICS, VOL.14,
NO.6, NOVEMBER 1999.
[12]. K.Viswanathan, R. Oruganti, and D. Srinivasan, Dual-Mode
Control of Cascade Buck-Boost PFC Converter, 2004, 35
th

Annual IEEE Power Electronics Specialists Conference, Aachen,
Germany, 2004, pp. 2178-2184.
[13]. Alfio Consoli, Mario Cacciato, Antonio Testa, Francesco
Gennaro, Single Chip Integration for Motor Drive Converters
With Power Factor Capability, IEEE TRANSACTION ON
POWER ELECTRONICS, VOL.19, NO.6, November, 2000, pp.
1372-1379.
[14]. Jean-Christophe Crebier, Bertrand Revol, and Jean Paul Ferrieux
Boost-Chopper-Derived PFC Rectifiers: Interest and Reality,
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS,
VOL.52, NO.1, FEBRUARY, 2005.
[15]. Mohan, Undeland, Robbins, POWER ELECTRONICS
Converters Applications and Design, John Wiley and Sons, 1995.

Proceedings of the 15th Mediterranean Conference on
ControI & Automation, JuIy 27 - 29, 2007, Athens - Greece
T26-024