A POWER ELECTRONICS AND DIGITAL CONTROL EXPERIMENT

APPLIED TO TEACHING INTERDISCIPLINARY IN ELECTRICAL

ENGINEERING

Kleiton M. Sousa, Filipe Marangoni, Julian K. Moreno, Emerson G. Carati, Mario L. S. Martins,
Carlos M. O. Stein, Jean C. C. Silva
UTFPR Federal University of Technology - Parana, ZIP 85503-390, Pato Branco - PR
kleitonms, fi.marangoni, jk.moreno@gmail.com, emerson, mlucio, cmstein, jeanccs@utfpr.edu.br

Abstract - This paper presents a interdisciplinary versities have developed several didactic platforms for teach­
teaching experience applied to power electronics and di­ ing [3,4].
gital control. The boost converter with a discrete-time PI This paper presents a didactic experiment applied to mul­
controller implemented in a microcontroller is used in this tidisciplinary teaching to digital control and power electron­
experiment. It is presented a boost converter model and ics. This experiment is applied in the digital control course
the controller design, beyond the description of in strum en­ of electrical engineering graduation program , where the pre­
tation circuit used. Finally are presented experimental and requisites necessary for the student are knowledge in power
simulated results. This experiment is applied to UTFPR electronics, analog and digital electronics and programming
electrical engineering students. techniques microcontrolled devices. The approach is proposed
to be developed during one semester and must be carried out
Keywords - Boost converter, discrete time control, edu­
from the first classes in the course. For that, the professor must
cation, interdisciplinary.
specify this work as the final project of the course.
This paper are organized as follows. Section II, which is
I. INTRODUCTION presented to model the converter used, the instrumentation cir­
cuit used and control design implemented in the microcon­
Nowadays, most electrical engineering courses include the troller. Section III shows the results and discussions obtained
disciplines of power electronics and digital control, since the in simulation and experimentally, followed by conclusions and
demand for technologies using these areas have grown in num­ final considerations shows in Section IV .
ber of application and become indispensable for the back­
ground knowledge of graduated students motivated the train­
ing of future graduates students. Most of real application have
II. DESCRIPTION OF THE PROPOSED SYSTEM
a interdisciplinary characteristic, requiring the integration of
the concepts of both areas, power electronics and digital con­
trol. An example of such application is in power supply sys­ The diagram of the system structure is shown in Fig. 1,
tems of electronics equipments. Often, the power of the equip­ comprising a static converter feeding a load, the circuit instru­
ment is supplied by a power electronic apparatus known as mentation and the digital controller implemented in a micro­
switch-mode power supply (SMPS). The SMPS operates in controller. Initially the student should analyze the converter
order to regulate the voltage required for the system. To ac­ desired to be controlled, by finding the converter simulation
complish such task there is a control system that handles the model. After modeling the converter, the student should de­
error between the voltage reference and voltage measured. velop an instrumentation circuit to provide the adequation of
However, in the power electronics discipline on electrical the signal in the range of values appropriate for the AID con­
engineering course are normally studied the static converters version of the microcontroller. The digital control discipline
operation stages, not considering the techniques of control and is applied during the controller development and design, be­
operation of the closed-loop converters. As well as, the im­ ing the student to decide which type of controller should be
pact of the disturbances such as, circuit input variations or employed.
load transients, in the converter response. The power elec­ As an example of the use of multidisciplinary didactic ex­
tronics is a discipline of experimental and multidisciplinary periment will be shown a boost converter and a PI controller
character [1] . Moreover, the digital control discipline only in discrete time. The experiment begin with the modeling
deals with the control isolated, using theoretical simulations of the boost converter, after that, the instrumentation circuit
tools and mathematical analysis. Often in digital control dis­ is presented. Following is described a discrete time PI con­
cipline, none experimental situations is analyzed, and some troller used, obtained by discretization of a continuous-time
aspects such as actuator devices saturation effect are not ob­ controller. The controller is implemented with a microcon­
served. troller MC9S08AW60, manufactured by Freescale. Finally,
The student's education can be improved by integrating in­ the simulation and experiments results obtained are presented.
terdisciplinary areas with practical situations. It will allow the Due to the versatility of the platform, could be used with other
students capacity of measuring and analyzing real systems be­ topologies of controllers in discrete time or another static con­
havior [2] . For the undergraduate student improvement, uni- verter for quick and inexpensive way.

978-1-4577-1646-1/11/$26.00 ©2011 IEEE 1037

 Converter Load

Fig. 4. : Diagram of the boost converter connected with switch

off.

, ,

rent h, are shown in (10) to the switch on and (11) to the

L __________ --.J

Microcontroller
switch off.
Fig. 1. : Block diagram of experiment. The output voltage can be obtained using the space state
equations. For the switch on the output voltage expression is
shown in (2) and to the switch off the output voltage is shown
in (3).
R
VOon = Ve (2)
re+R
R reR
VOoff = Ve +h (3)
re+R re+R
The converter equation change according to the state of the
Fig. 2. : Simplified diagram of the boost converter. switch. Therefore, it is a time-varying system. The average
transfer function of the boost converter should be determine to
design of the propose controller. According to [5] , the average
transfer function is:

va(s) _ Vo re (s+Wzn ) ( S - Wzp )

G(s) = =
d(s) (1-D)(R+re) s2+2�wos+w�
(4)
where Va (s) represents the behavior of the converter output
Fig. 3. : Diagram of the boost converter connected with switch voltage, the converter average output voltage in regime is re­
on. presented by Vo. The duty cycle is represented by D, d(s)
represents a step of duty cycle. For the simulation of the con­
verter should be calculated the converter output voltage Vo by
A. Modeling of Boost Converter (1), for a duty cycle D, and the system input d(s) will be a
The simplified diagram of a boost converter is shown in step with the same amplitude value ofD. Therefore, the trans­
Fig. 2. This converter may operate in two distinct forms, de­ fer function used will serve to represent the converter output
pending on the state of the current through the inductor L. in just one operation point. The remaining terms of (4) are:
In this paper, only the continuous current conduction mode
1
(CCM) is regarded. In CCM operation, the converter operates Wzn = -- (5)
Gre
in two stages, when the switch S is closed and when the switch
S is open.
The converter static gain depends of the switch S conduc­ Wzp = ± [(1-D)2 R+r] (6)
tion time te. Therefore, the converter output voltage is given
by: r +(1-D)2 R
(7)
E Wo =
LG (R+ro)
Vo = (1)
'
1- tc/T
[
G r (R+re)+(1-D)2 Rre ]
J
where Vo is the converter output voltage, tc is the switch S � (8)
[ ]
=
conduction time and T is the switching period.
2 LG (R+re) r +(1-D)2 R
In order to provide more accurate simulation, the boost con­
verter model considering the inductor windings resistance rL,
where r is an equivalent resistance. The resistance r depends
the capacitor equivalent series resistance re and the switch on­
of the inductor resistance, the switch on-resistance RDS and
resistance RDs should be developed. The Fig. 3 represents the
the capacitor equivalent series resistance re. The value of r is
converter with the switch on, and Fig. 4 shows the converter
given by [5] :
with the switch off.
The equations describing the circuit dynamics, choosing as D (1-D) Rre
state variables the capacitor voltage Ve and the inductor cur- r = rL+DRDS + ' (9)
R+re

1038
 •

( �� ) = ( -1/C(R+ro) ) ( �� ) + ( I�L )
o
E (10)
o -(rL+RDS )/L
•

( �� ) = ( -R/L(ro
-1/C(R+ro) R/C(R+ro)
) ( �� ) + ( I/L
o
(11)
+R) - [R(ro+rL)+rO +rL) /L(R+ro)

TABLE I
Circuit parameters
v.

Input Voltage E 12 V
Inductor L 1.7mH
Switch S IRF740
Diode MUR860
Capacitor C 44f.£F Fig. 5. : A second order lowpass circuit.
Resistor R 480
Switch resistance RDS 0. 550 10 k!l 1 k!l 1 k!l
Capacitor resistance ro 0.850
Inductor resistance rL 0. 30 5V
Switching frequency 20kHz lOill

20nF
The implemented circuit parameters are presented in the Ta­
ble I.
With the values from Table I, the average transfer function 4.7 ill

of the boost converter is given by:

-0.802482- 1.592xl048+1.479xl08
G(8)= (12)
82+10318+3.492xl04
Fig. 6. : Diagram of instrumentation circuit.
B. Instrumentation Circuit

The instrumentation circuit comprises the voltage divider,

which reduces the voltage converter output tenfold, the sec­ equation of lowpass filter is shown in (15).
ond order lowpass filter which function is to filter the ripple of
boost converter output voltage and the zener diode 5 V which ( )- 1.579x108
H 8 (15)
function is protect the microcontroller if the filter output sig­ 2
- 8 +1.776xl048+1.579xl08
nal exceeds 5 V. In order to increase input impedance of the
instrumentation circuit the buffer is used between the resistive C. Microcontroller Used and PI Discrete Time Controller
divider and the low pass filter. A pull down resistor is used
in the circuit output. The filtered signal is then converted to The microcontrollers are programmable devices composed
digital by the microcontroller. of a processing unit that includes some peripherals use in in­
ternal structure. The microcontrollers have peripherals such
A second order lowpass filter can be implemented using the
as memory read and write, analog to digital converter, PWM
circuit shown in Fig. 5, which employs an opamp in an ar­
generators and other devices that vary depending on the model
rangement of a VCV S with gain G. Its transfer voltage ratio,
used. The bits of the microcontroller also varies depending on
obtained by analyzing the circuit, is shown in (13).
the model used, where as higher the number of microcontroller
bits, greater will be the numerical accuracy.
The DSP devices have a processing speed much higher com­
paring to the microcontrollers. Moreover, the DSP devices
utilize the digital representation for floating point, which pro­
where: vides greater accuracy for numerical representation. The most
R
G=I+ b (14) microcontrollers advantage is their low cost compared to DSP.
Ra The digital controllers have some advantages compared to
The complete instrumentation circuit is shown in Fig. 6. analog controllers. One of the advantages of digital controllers
The filter cutoff frequency of Fig. 6 is 2 kHz, a decade be­ is the design flexibility, where the poles and zeros allocation
low the switching frequency of the boost converter. In order is done by changing only one value in the control equation
to achieve unity gain in the circuit, is used Rb = 00. The implemented in the microcontroller.

1039
 Per = 2.2. This results to a controller given by:

7.1
C( s) = 0.0135 + - (18)
s
The discretization of the compensator C(s) is made by the
Euler method. Making s � (1 -z)/T, with a discretization
Fig. 7. : Block diagram of control loop. interval of 1 ms, is determined the discrete time compensator
C(z):
x10' 0.0071
1.5 C(z) = 0.0135 + ' (19)
l-z -1
'"
·x

�
therefore, the control action u[k] can be implemented by the
cd
0.5 difference equation:
r
0
.S k
L eli]
00
cd -0.5
,§ u[k] = 0.0135e[k] +0.0071 , (20)
-1 i=l
-1.5
_4 -3 -2 -1 o 2 the control action u[k] sets the boost converter switch conduc­
x10'
Real axis tion time, defined in (17). The adder in (20) can be imple­
mented using a variable that accumulates the error sum for
Fig. 8. : Roots locus graphic. each sample. In (21) the variable S is used to accumulate the
error sum.
S S+e[k] = (21)
This paper proposes a PI discrete-time classic controller.
This controller was chosen because it is easily implemented, Ill. RESULTS
and presents wide use in industrial control systems. The
discrete-time PI controller is obtained by discretization of a The performance verification of the discrete-time PI con­
continuous-time controller. From the discretized PI is found troller to control the boost converter was performed using the
the control action implemented in a microcontroller. The modeling presented in (10) an (11), and discretized by the Eu­
continuous-time PI controller is shown in (16): ler method with time discretization Ts 50 f-Ls. To verification
=

the behavior of the output voltage during a switching interval,

the system was again discretized using a simulation step of
(16) 0.5 f-Ls.
The proposed controller is implemented in a MC9S08AW60
where K represents the proportional gain and I/Ti the integral microcontroller. This micro controller has 8-bit and an 20 MHz
gain [6]. clock. The main peripherals used for the project implementa­
The block diagram of the control loop is shown in Fig. 6. tion are a PWM generator and an AID converter. Due to the
The discrete-time controller is represented by C(z), the blocks calculations required for the control action, the acquisition rate
G(s) and H(s) are defined by (12) and (15), respectively. The was limited to 1 kHz. The prototype is implemented using a
control action u[k] determines the boost converter switch con­ microcontrolled board developed during the Microcontrollers
duction time ton , calculated by [7]: Systems course. This board includes AID and PWM separated
connectors, which becomes easier to interface with instrumen­
tation circuit and PWM drive. Moreover, power and program­
(17)
ing connectors are included in this board to allow control flex­
ibility through reprogramming. The Fig. 9 shows a picture
where E is the converter input voltage and T the switching detailing the microcontrolled board.
period. To calculate the conduction time should be make the
A. Open Loop Convener
acquisition of the boost converter input voltage.
The Ziegler-Nichols method is used to tune the continuous­ The Fig. 10 shows the simulated output with a 0. 5 duty cy­
time controller. In this kind of tuning is necessary be found the cle using the state space equations. Still in Fig. 10 is shown
value of the gain Ker which makes the system critically stable the graph relating to the converter average transfer function
using an integral gain equal to zero (Ti 00 ) . With the value
= response. The software used for simulation is MATLAB.
of Ken the parameters in (16) are calculated as K 0.45Ker = In Fig. II(a) is shown the detail of simulated output volt­
and Ti = Per/1.2, where Per is the system critical oscillation age ripple and in the Fig. 11(b) is shown the same detailing
period with gain Ker [6] . obtained experimentally. The output voltage oscillations sim­
The critical gain can be determined by many ways, this pa­ ulated and experimental behave similarly.
per used the roots locus graphic, using MATLAB as a tool for Finally, the Fig. 12(a) shows the simulated PWM and the
finding the critical gain. In Fig. 8 is shown the roots locus. output voltage, the Fig. 12(b) shows the same waveforms for
The critical gain is Ker 0.031. The critical period is
= the circuit implemented.

1040
 ;, 1I ---; ----:- ;,
� .::::L .. _·
rI --�- -1I
····r .. ..
--- -- - --- ---- --- ---- ---- --- -�-----

. ..
.. ....·· ··· ..r-· :....
-----i- -
..

- -t -

---i---�-i-
---1'
---- ,- -----
r ----
---I-

...:
�
, , , ---,- ---....- ----�---
__ --'------..--- --_0. ------1-
-

E
� � ----� ----�- �-�---
�
...� .. . . .. '. .....�. ..�.....'....
I
f - --�- ---�- -----T ----r-- -- 1--
--
I
---I � � -----+ -- -�--
,..... ----- -

� ---- ------

... �.� ...�.... ...+ ---,-- �.... -;---

I
�,

I , , , ,

Time (20 Ils/div)

(a) Simulated
--=. .-

Fig. 9. : Microcontroller board and its main connections.

I

- State space equations

_. _ .. Transfer function
- ....J.. -J.. -t -- -

30 :--"1 r---
'-

lJl
�

�
r
20

- t-----.;----
-- · !--- �
� --t--------_:_- ---�
----:--�---t--!---r--­ •
I
i'"'"'"-'
I I
;--
I
.-- L
---l--------� -------+ ----� --
10 ------ ---�- -------

. ,___ -' ________ '- _______1... _________....________

. ..: _________
'M" ,�,
1',"0000"'
20.�
m:ir.·��" U
• J 2.I0v

0 ...i�-----�------ -- ------ �
(b) Experimental
o 2 4 6 8 10 12
Time (ms) Fig. 12. : Detail of PWM and output voltage: (a) simulated (b)
experimental
Fig. 10. : Simulated system response with a 0. 5 duty cycle.

L D

24
>'

E
� 23

- 22
�
21

Time (20 Ils/di v)

(a) Simulated

Fig. 13. : Block diagram of the implemented circuit.

B. Closed Loop Converter

The diagram of the implemented system is shown in Fig. 13.

The converter input voltage was measured to calculate the
switch conduction time. The two signals (output and input
voltage) goes to microcontroller, which makes the AID con­
version, calculating the switch conduction time. The PWM
signal generated by the microcontroller is sent to a driver.
'0111.2010
" ':0'
The system simulation using MATLAB is shown in
Fig. 14(a). The reference voltage is 24 V. The circuit values
Time (20 Ils/div)
and other parameters are the same presented in Table I. The
(b) Experimental Fig. 14(b) shows the system experimental response. When
connected to the input source, the converter output stands at
Fig. 11. : Detail of the output voltage: (a) simulated (b) exper­
12 V for 150 ms until the control begins to operate, nullifying
imental.
the error at 700 ms. The initial time interval is due to u[k]

1041
 I
45 .-' -�C"'"'II -
.

40

35 t!
-E

J
30

l
>'
'-'
25
- PWM

I 1 �
Q)

�
01) 20
o:s I
�

I
-
15
I I
t
�r lM'9o(t<!OfGI I�HM"'"
� �
I
I'--i

,
10

5 '"'
I I
.IIIIY N......
.... ..
e J " .,V

0 �
n.n:2'

Time (20 �s/div)

o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Fig. 15. : Output voltage, input voltage and PWM for E = 8

(a) Simulated V.
-h
-
I -

,I l I 'I r
..I.....
I -E

/ t -PWM

.J I ,
�1 n1
)
n
f
�ilL I•

_J
'-----i __ ..

It::t�' I
'"' - u r '.1ev

�..M...",
U . .... . B
'Ii 011 V :'OIlY N...... _ J " .tV

�.
.··GOf�'H
":)001:15

Time (100 ms/div) �

n:ll: oH

Time (20 �s/div)

- Vo - E

(b) Experimental Fig. 16. : Output voltage, input voltage and PWM for E = 16
V.
Fig. 14. : System response with PI control: (a) simulated (b)
experimental.
log and digital electronics and programming techniques of mi­
crocontrollers. The student could observe, from the experi­
have initial value equal to zero, causing a negative te, accord­ mental results, the validity of theoretical models. In this ex­
ing to (17). A negative te causes the switch remains open, and periment we used a boost converter and a discrete-time PI
the output voltage of the circuit will be equal to input voltage. controller, but other converters, e.g., buck and buck-boost,
The controller will act only when the value of te is greater than and other controllers can be used without the need for major
zero, which happens some time after the error have been ac­ changes to the proposed experiment. It is suggest that when
cumulated, increasing the value of u[k]. This initial interval applied to a class of undergraduate students, more kind of con­
can be avoided using u[k] 12 as initial value. The behav­
= verters can be analyzed. The class can be divide in groups,
iors of the experimental and simulated system shows similar when each group should investigate a different converter or
responses. digital controller.
We performed variations in converter input voltage. The re­
sult for an input voltage of 8 V is shown in Fig. 15 and Fig. 16
shows the result when the input voltage is 16 V. The Fig. 15 REFERENCES
and Fig. 16 show the command pulse and the input and output [1] A. Ferreira, E. Vendrusculo, and J. Pomilio, "Integrated
voltage of converter. In the presented cases, control acted on platform for power electronics applications fast evalua­
the switch conduction time to maintain output voltage constant tion and teaching purposes," Power Electronics Educa­
at 24 V. tion, 2005. IEEE Workshop, pp. 81 - 86, June 2005.
[2] E. McShane, M. Trivedi, and K. Shenai, "An improved
IV. CONCLUSION
approach to application-specific power electronics educa­
With the proposed experiment, the student will be able to tion. curriculum development;' Education, IEEE Trans­
analyze, design and implement static converters using digi­ actions on, vol. 44, no. 3, pp. 282 -288, Aug. 2001.
tal control systems. To perform this experiment, the students [3] L. Rolim, R. Stephan, W. Suemitsu, and J. Silva Neto,
should have background knowledge on power electronics, ana- "Ufrj power electronics teaching lab: ten years," Power

1042
 Electronics Education, 2005. IEEE Workshop, pp. 69 -
73, June 2005.
[4] F. S. Garcia, A. A. Ferreira, and J. A. Pomilio,
"Plataforma de ensino de eletr6nica e potencia versatil
e de baixo custo," Eletronica de Potencia, vol. 13, no. 8,
pp. 85-90, May 2008.
[5] M. Kazimierceuk, R. Cravens, and J. Harrington,
"Closed-loop input impedance of a voltage-mode­
controlled pwm boost dc-dc converter for ccm," Circuits
and Systems, 1994. , Proceedings of the 37th Midwest
Symposium on, vol. 2, pp. 1253 -1256 vol.2, Aug. 1994.
[6] P. Cominos and N. Munro, "Pid controllers: recent tuning
methods and design to specification," Control Theory and
Applications, lEE Proceedings , vol. 149, no. 1, pp. 46
-

-53, Jan. 2002.

[7] L. Roggia, F. Beltrame, J. Baggio, H. Hey, and J. Pin­
heiro, "Implementation issues of a digital control system
applied to a pfc boost converter," Industrial Electronics,
2009. IECON '09. 35th Annual Conference of IEEE, pp.
1519 -1524, nov. 2009.

1043