Ahmad Shukri Fazil Rahman* et al. / (IJITR) INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY AND
RESEARCH
Volume No. 1, Issue No. 3, April - May 2013, 226 – 232.
Microcontroller Based SPWM Generator: A
Conventional Design Perspective Through
Graphical Oriented Approach
AHMAD SHUKRI FAZIL RAHMAN1
4
SYAFRUDDIN HASAN
MUZAIDI OTHMAN@MARZUKI2
BAHARUDDIN ISMAIL
5
ABDUL RAHIM ABDUL RAZAK3
SYED IDRIS SYED HASSAN6
Pusat Pengajian Kejuruteraan Sistem Elektrik,d/a Kompleks Pusat Pengajian Kejuruteraan Mikroelektronik
Universiti Malaysia Perlis (UniMAP),Kampus UniMAP Pauh Putra ,
02600 Arau, Perlis, Malaysia.
Abstract- In this paper, a comprehensive design strategy was proposed and implemented for the design of a single
phase sinusoidal pulse width modulation (SPWM) waveform. Excel software was used to generate the timing
sequence, and then fed to a microcontroller unit (PIC16F877A). Through Excel, complex programming effort could
be avoided and the whole data arrays can be presented graphically. The effectiveness of the timing sequence was
validated by using PROTEUS software. The simulation result shows that the SPWM waveform achieved the desired
goals.
Keywords-SPWM, PWM, sinusoidal, triangular, modulation, carrier, excel, Proteus, inverter.
I.
INTRODUCTION
Commonly used in inverter system, the SPWM or
sinusoidal pulse width modulation is a part of
control algorithm which rises’ from the many
variety of pulse width modulation (PWM) control
technologies.
Its application has widened that include motor
control, induction heating, welding power source,
electronic converters [1, 2], wind power generation
[3],
flexible transmission system (FACTS)
controllers [4] and etc. As the name implies,
SPWM is a comparison between a reference
modulation (sinusoidal) waveform and a carrier
(triangular) waveform which resulted in PWM
(gate) signals for switching devices. The width of
each respective PWM signal is proportional to the
amplitude of a sine wave. Fig. 1 shows a typical
SPWM generation theory and its resultant PWM
signal.
There are several ways of producing SPWM signal,
the method mainly consists of modulation or with
the introduction of dedicated microcontroller [1, 3,
5-8]; which is producing greater accuracy than
analogue circuits [7]. Other known methods
emphasized the use of matlab simulink [7-9] and
Labview [10] at design stage to performed SPWM
process. With the rapid advancement of SPWM
technology [11-15], complex circuitry could be
integrated in a single chip thus reducing the
complexity and increasing its capability [16]. Even
though common method of loading complex
programming SPWM algorithm to a single chip
may compromise the chip performance [6]. Thus,
the entire process could be avoided by performing
the required task (SPWM comparison signal)
through excel, this may reduce the degree of
programming complexity and the chip performance
can be fully optimize.
It is the intentions of this paper to investigate the
graphical approach by using excel 2007 to generate
the PWM signal. These signals will be compared
(sinusoidal and triangular) and then the output from
the comparison will be used to generate the PWM
signal. Thus, the required signal will be generated
without affecting the performance of the
microcontroller. Proteus Isis version 7.7 was used
to simulate the PWM signal through PIC16F877A
microcontroller.
II.
Fig. 1. SPWM comparison Signals (top) and PWM
signal output (bottom)
ISSN 2320 –5547
SPWM DESIGN PERSPECTIVE
Presently, SPWM can be generated through
symmetric regular sampling method and
asymmetric regular sampling method [5] as per fig
2 [5] and fig 3 [17]. The excel SPWM generation
@ 2013 http://www.ijitr.com All rights Reserved.
Page | 226
Ahmad Shukri Fazil Rahman* et al. / (IJITR) INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY AND
RESEARCH
Volume No. 1, Issue No. 3, April - May 2013, 226 – 232.
principle was based on asymmetric method
whereby discrete value of sinusoidal signal was
sampled at nth steps (step size) per fc (carrier
frequency). At each step, the carrier and
modulation signal will be compared and the
intersection between the modulation waveform,
Vm(t) and the carrier signal, Vc(t) [17] will
produced a PWM signal, Tpwm with its period
proportional to the amplitude of the modulation
waveform. Larger step size or high carrier
frequency [17] would produce higher degree of the
PWM signal accuracy.
Vc(t)
Vm(t)
B
A
Tpwm
Vdc(t)
Fig. 2. Symmetric regular sampling principle
Vc(t)
t
Vm(t)
B
A
Tpwm
Vdc(t)
t
Fig. 3. Asymmetric regular sampling principle
III.
THE GRAPHICAL APPROACH- EXCEL
(2)
DESIGN
Three data columns were created to represent the
modulation, carrier and the PWM waveform plus
one column for timer or as a counter. The
modulation waveform was created first with 1001
steps or map size by using function as shown in
Equation (1), [18]. The frequency of the sinusoidal
waveform was set to 50Hz or 20ms with respective
step size of 20s.
(1)
Where MAPSIZE is the total array size, DEGMAX is
the maximum angle value at 360, DEGMIN would
be the minimum angle value at 0, DEGRES is the
resolution for each step size and KMAP is the
mapping scaler.
The degree’s values were converted to radian
through Equation (2):
ISSN 2320 –5547
Using built in excel’s sine function (Equation 3);
the pre-calculated angle was transformed to sine
waveform.
Sinusoidal=5.4*SIN(number)
(3)
Whereby, 5.4 would be the maximum amplitude of
the waveform and (number) is the specific value in
the respective row array.
For the triangular or the carrier signal, the
frequency was set to fc = 1kHz or 1ms while the
amplitude was set to 5.4V. Since excel did not
provided any function to generate triangular
waveform; a simple linear equation was improvised
(Equation 4) as an alternative [19].
y=mx+c
@ 2013 http://www.ijitr.com All rights Reserved.
(4)
Page | 227
Ahmad Shukri Fazil Rahman* et al. / (IJITR) INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY AND
RESEARCH
Volume No. 1, Issue No. 3, April - May 2013, 226 – 232.
With m is the line slope and c would be the point of
y-axis intersection. The carrier signal was created
with 50 step size with carrier frequency, fc at 1kHz
and each respective step is at 20Hz or 50ms. These
data’s were tabulated inside excel’s data array and
the final outcome is a graph contain both the carrier
signal and the modulated signal shown in fig 4.
Fig. 4. SPWM comparison signals
The PWM signal was created by building Equation
5:
PWM=IF(((ROWCARRIER)<(ROWSINUSOIDAL)),1,0)
(5)
Equation 5 compares data, i.e. modulation and
carrier and if the condition is true the output will be
‘1’. Fig. 5 shows the PWM results for carrier
frequency, fc = 1kHz and modulation frequency, fm
= 50Hz.
Fig. 5. PWM output
The timing sequence for PWM period from fig 5
was obtained by measuring the length of each
respective PWM period in number of step sizes
multiply by 20s per step. There are 40 data’s
calculated from SPWM comparison process which
contain both positive and negative cycle. There are
20 data’s of positive and negative cycle
ISSN 2320 –5547
respectively, that is equivalent to the carrier-tomodulation frequency ratio, fc/fm = 20, [17].
The PWM period is represented as in fig 6. The
graph shows that the timing period is between
960ms and 20ms. The graph also indicates average
time occurs at the beginning, middle and end of the
modulation signal.
@ 2013 http://www.ijitr.com All rights Reserved.
Page | 228
Ahmad Shukri Fazil Rahman* et al. / (IJITR) INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY AND
RESEARCH
Volume No. 1, Issue No. 3, April - May 2013, 226 – 232.
Fig. 6. PWM period graphical representation
IV.
THE SIMULATION ASPECT
The data’s from excel transformation was
downloaded to a microcontroller unit through
Proteus simulation. The data was created based on
section 3, i.e. carrier frequency, fc = 1kHz and
modulation frequency, fm = 50Hz. The timing
program was written by using delay code under
MPLAB environment with HI-TECH C as the
compiler. The Proteus experimental set
up was based on fig 7, with PORD0 as the PWM
output and the clock frequency was set to 20MHz.
A virtual digital oscilloscope was connected to
PORTD0 to record the PWM output. The
simulation was conducted from a DELL laptop
with Intel® Core(TM)2 Duo CPU at 2.00GHz and
Microsoft Windows XP version 2002 SP2.
Fig. 7. Proteus SPWM experimental setup
V.
EXPERIMENTAL RESULTS
Fig. 8 shows the simulation outcome for the
proposed method. The result indicate a series of
pulse wave signal with fc/fm = 20 at 19.75ms which
ISSN 2320 –5547
is 1.25% lower than the calculated total modulation
period. The measured amplitude is at 2.77V. The
display outcome exhibit close resemblance with the
theoretical result.
@ 2013 http://www.ijitr.com All rights Reserved.
Page | 229
Ahmad Shukri Fazil Rahman* et al. / (IJITR) INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY AND
RESEARCH
Volume No. 1, Issue No. 3, April - May 2013, 226 – 232.
Fig. 8. Simulation result with fc=1kHz and fm=50Hz
Fig. 9 shows the simulation outcome for positive
cycle. The result indicate a series of pulse wave
signal with fc/fm half from the actual is at 10 and
the half cycle is at 9.73ms which is 2.7% lower
than the calculated total modulation period. The
measured amplitude is at 2.77V. The display
waveform also contains similar resemblance with
theoretical result.
Fig. 9. Half positive cycle simulation result
Fig. 10 shows the simulation outcome for negative
cycle. The result indicate a series of pulse wave
signal with fc/fm half from the actual is at 10 and
the half cycle is at 10.05ms which is 0.5% more
ISSN 2320 –5547
than the calculated total modulation period. The
measured amplitude is at 2.77V. The display
waveform also produce similar pattern with
theoretical result.
@ 2013 http://www.ijitr.com All rights Reserved.
Page | 230
Ahmad Shukri Fazil Rahman* et al. / (IJITR) INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY AND
RESEARCH
Volume No. 1, Issue No. 3, April - May 2013, 226 – 232.
Fig. 10. Half negative cycle simulation result
A graphical comparison was conducted between
simulation values and calculated values. Graph in
fig 11 and 12 show the comparative presentation
for positive and negative cycles respectively. Graph
in fig 11 shows the percentage error output for
positive cycle and it can be seen that significant
error occurs at the beginning of the modulation
process and high error rate occurs during the
beginning of transition period from positive to
negative cycle.
Fig. 11. Percentage of error for positive cycle
Graph in fig 12 shows the percentage error output
for negative cycle, it can be seen that significant
error occurs at the beginning of the modulation
ISSN 2320 –5547
process and maintain a stable errorless value
circulating around peak modulation before
plunging at the end.
@ 2013 http://www.ijitr.com All rights Reserved.
Page | 231
Ahmad Shukri Fazil Rahman* et al. / (IJITR) INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY AND
RESEARCH
Volume No. 1, Issue No. 3, April - May 2013, 226 – 232.
Fig. 12. Percentage of error for negative cycle
VI.
CONCLUSION
[8]
The proposed method presented in this paper shows
that the graphical method could provide an
alternative solution toward SPWM generation.
PWM signal generated by excel method was
simulated through Proteus environment. Results
from calculation and simulation obtained validate
the process effectiveness with acceptable error
values.
VII. REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
L. H. Walker, “10-MW GTO converter for battery
peaking service,” Industry Applications, IEEE
Transactions on, vol. 26, no. 1, pp. 63-72, 1990.
M. H. Ohsato, G. Kimura, and M. Shioya, “Fivestepped PWM inverter used in photovoltaic
systems,” Industrial Electronics, IEEE Transactions
on, vol. 38, no. 5, pp. 393-397, 1991.
Y. Lu, P.-p. Jiao, and B. Zhang, "The principle and
realization of single-phase SPWM wave based on the
counter method." pp. 10-13.
J. Segundo-Ramirez, and A. Medina, “Modeling of
FACTS Devices Based on SPWM VSCs,” Power
Delivery, IEEE Transactions on, vol. 24, no. 4, pp.
1815-1823, 2009.
Y. Zhou, Q. Yang, Q. Zhang et al., "The Generating
Method of SPWM with Double-Interruption Based
on DSP." pp. 617-620.
B. K. Chaudhari, S. S. Wekhande, S. S. Dhamse et
al., "A low cost high performance SPWM inverter."
pp. 1033-1036 vol.2.
G. Shuangxi, C. Shufu, and Z. Ying, "Sinusoidal
Pulse Width Modulation Design Based DDS." pp. 14.
ISSN 2320 –5547
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
Y. Wei, W. Kai, L. Zhengyu et al., "Digital SPWM
inverter design and implementation." pp. 5847-5850.
Z. Chongqing, and C. Jianli, "A new method of
solving SPWM switch point based on natural
sampling." pp. 325-329.
L. Fan, L. Kun, and L. Yang, "A design and
implementation of edge controller for SPWM
waves." pp. 764-767.
B. Mwinyiwiwa, Z. Wolanski, and O. Boon-Teck,
“Current equalization in SPWM FACTS controllers
at lowest switching rates,” Power Electronics, IEEE
Transactions on, vol. 14, no. 5, pp. 900-905, 1999.
Y. Zhaoyang, Z. Kun, L. Jianxia et al., “A Novel
Absolute Value Logic SPWM Control Strategy
Based on De-Re-Coupling Idea for High Frequency
Link Matrix Rectifier,” Industrial Informatics, IEEE
Transactions on, vol. 9, no. 2, pp. 1188-1198, 2013.
L. Gang, Z. Hao, C. Hui et al., "A novel gang SPWM
control method for six-phase induction motor." pp. 14.
H. Patangia, and D. Gregory, "A Novel Multilevel
Strategy in SPWM Design." pp. 515-520.
G. Milan, M. Mohamadian, S. M. Dehghan et al., "A
novel SPWM strategy for single- to three-phase
matrix converter." pp. 495-500.
Y. Yang, Y. Gao, and L. Chen, "Design and Test of
Novel Programmable Digital Three Phases SPWM
Chip." pp. 1-3.
B. Mwinyiwiwa, Z. Wolanski, and O. Boon-Teck,
“Microprocessor-implemented
SPWM
for
multiconverters with phase-shifted triangle carriers,”
Industry Applications, IEEE Transactions on, vol.
34, no. 3, pp. 487-494, 1998.
J. Yan, M. Ryan, and J. Power, Using fuzzy logic:
towards intelligent systems: Prentice Hall, 1994.
"Linear equation," http://en.wikipedia.org/wiki/
Linear_equation.
@ 2013 http://www.ijitr.com All rights Reserved.
Page | 232