Professional Documents
Culture Documents
Me2010 - Avravi Comanda Invertor
Me2010 - Avravi Comanda Invertor
Me2010 - Avravi Comanda Invertor
By
Anand Vivek Ravi
Acknowledgements
I extend a great thanks to Prof. Vinod John for giving me an opportunity to work in the
Power Electronics Group.I would like to express my admiration for him on the various interesting ideas suggested during the course of project. Work under him had always been a
great pleasure and I take this opportunity to acknowledge all the timely help rendered by
him, right from awarding me a very interesting project till the final completion of it. I would
like to acknowledge the confidence and the encouragement provide by him during the final
bottleneck situations, which was a great motivation, helping me complete the work.
I express my humble gratitude to Prof. V. Ramanarayanan, who has been a pillar of support
to all the students in the group. I,having cherished attending his lectures and seminars,
would like to thank him for teaching us to progress on becoming good power electronics
design engineers.
I would like to thank Prof. V. T. Ranganathan for his patient lectures and in making a not so
conspicuous topic into an easily comprehensive one. I would like to thank Prof. G. Narayanan
for providing insights on the various PWM techniques through his course.
I thank my close frinds Venkat, Manoj, Shan, Arun Karuppaswamy,Raju and Tarak for
being supportive and critical of me and for the interesting discussions had with them. I
would also like to thank Vishnu, Prakash and the other students of the batch for their support. I thank all PhD students of the PEG group Kamalesh Hatua, Amit Jain, Shivaprasad,
Anirban, Dipankar De, Soumitra Das and Binoj Kumar for their help. I would like to thank
Srinath in helping me out on various aspects of the project.
I thank Ms. Silvi Jose, Mr. Paul, Mr. Ravi and members of the department workshop for
their cordial demeanour. I also extend my thanks to Mr D. M. Channegowda and his team
at the department office for their good administrative activities.
I would like to thank Shankar, Anil Adapa and Mr. Krishna of M/s CHIPKRAFT TECHi
ii
Acknowledgements
NOLOGIES for helping at a very difficult situation, the result of which the project got
completed.
Finally, I would like to thank my parents who have put up with me inspite of my maverick
behaviour and for being by my side in all my failures and success. I would like to thank
my adoring grandparents and my guru shri Anandha Bharathi Ayya for providing me the
mental strength during bleak periods and the God Almighty for having been my best buddy,
for giving me the opportunity to study at this prestigious institute, to work under such an
understanding and a patient guide and finally for providing me a career through this project
along with the completion of the project.
Abstract
High power converters are used in variable speed induction motor drive applications. These
power converters consist of active front end rectifier, a DC link and an inverter feeding the
motor. The power devices employed in this converter are high voltage IGBTs.
The following figure shows a typical block diagram of the inverter fed induction motor drive
system
Contents
Acknowledgements
Abstract
iii
List of Tables
vii
List of Figures
viii
Nomenclature
1 Introduction
1.1
1.2
1.2.1
1.3
1.4
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1
Construction of Ultracapacitor . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2
2.3
2.4
2.5
2.4.1
11
2.4.2
12
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
iv
Contents
3 Hardware Design
3.1
14
Power Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
3.1.1
14
3.1.2
14
3.1.3
15
3.1.4
. . . . . . . . . . . . . . . . . . . . . . . . . .
15
3.1.5
16
3.2
Theory of operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16
3.3
Controller Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
3.4
18
3.4.1
Voltage Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
3.4.2
Current Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
4 Controller Design
4.1
4.2
4.3
4.4
21
21
4.1.1
Control Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
24
4.2.1
24
4.2.2
26
28
4.3.1
28
4.3.2
30
31
5 Experimental Results
33
5.1
Ultracapacitor Characterisation . . . . . . . . . . . . . . . . . . . . . . . . .
33
5.2
33
5.3
35
5.3.1
35
5.3.1.1
Charging at 1A . . . . . . . . . . . . . . . . . . . . . . . . .
35
5.3.1.2
Charging at 2A . . . . . . . . . . . . . . . . . . . . . . . . .
36
5.3.2
38
5.3.3
42
vi
Contents
5.3.3.1
42
5.3.3.2
42
43
5.4
45
5.5
Thermal Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
46
5.3.4
6 Conclusions
48
50
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
50
51
52
53
54
55
56
References
57
List of Tables
1.1
System Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1
2.2
10
3.1
Inductor Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15
5.1
Ultracapacitors Parameters
. . . . . . . . . . . . . . . . . . . . . . . . . . .
34
5.2
45
6.1
49
vii
List of Figures
0.1
iii
1.1
1.2
1.3
2.1
2.2
Ragones Plot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3
Electrical characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
2.4
Discharge Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
2.5
12
3.1
Power circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
3.2
17
3.3
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
3.4
19
3.5
20
4.1
22
4.2
22
4.3
d d(s)
d . . . . . . . . . . . . . . . . .
Current to Control transfer function i(s)/
25
4.4
inner current loop - Loop gain GH transfer function i(s)/ iref (s) . . . . . . .
25
4.5
27
4.6
27
4.7
d d(s)
d . . . . . .
Current to Control transfer function - ultracap charging i(s)/
29
4.8
29
viii
List of Figures
4.9
ix
d d(s)
d . . . . .
Voltage to Control transfer function - ultracap discharging v(s)/
31
32
5.1
34
5.2
35
5.3
Charging of Ultracaps - 1A . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36
5.4
Charging of Ultracaps - 2A . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36
5.5
38
5.6
39
5.7
39
5.8
Boost Converter Constant Voltage Control - 15V input, 25V output, 20ohm
load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
40
40
41
42
43
43
44
44
45
46
47
47
50
51
52
53
54
55
56
5.9
Nomenclature
Symbols
: Definitions
Vmax
Vmin
Vnominal
Vcell
Imax
Ct
Irms
RDSon
Vav
Iav
tr , tf
vb(s)
bi(s)
ev
ei
error[n]
kp [n]
Nomenclature
ki [n]
b
d(s)
xi
Chapter 1
Introduction
The power line disturbances like voltage sags and blackouts affect the continuous process
industries to a great extent, the severity being dependent on the magnitude and duration
of the sag. Faults lasting even less than 0.5s can have dentrimental effects on the production. Each industrial equipment reacts to voltage sag in different ways [2]. Voltage sags are
defined as a momentary dip in voltage - below 10 % lasting from a few cycles(10ms) to 150
cycles(0.3s) [16].
Voltage sags cause diruption of service but do not cause damage to sensitive loads. On the
Chapter 1. Introduction
loads. More than 62% of the disturbances are due to voltage sags with duration less thanhalf
a second [16]. A severe voltage is one where voltage falls below 85 %. Voltage sags are caused
by faults within the system, starting currents of motor.
In the induction motor drive, severe voltage sags on the control power supply feeding the
DSP, Gate drive cards, Current and Voltage sensing cards might cause the entire motor drive
system to shutdown. The control flow diagram is shown in figure 1.1. So a ride-through system consisting of DC-DC converter, an energy source has been designed to overcome the
effect of these voltage sags and momentary black-outs.
1.1
1.2
The project aims at developing a fault ride through system using Ultracapacitors as the
source of energy. The ride through system consists of a bank of ultracapacitor charging and
discharging through a bi-directional buck-boost converter. The system has been designed to
provide back-up over a period of 10s. The ride through system is shown in figure 1.2. The
main power to the control cards is fed through the PFC boost converter circuit, consisting
of a transformer-rectifier system in cascade with boost converter to make the input current
continuous, feeding the bi-directional converter and the sensing cards in parallel.
1.2.1
The block diagram of the ride through system consists of a power converter interfacing the
ultracapacitor and the loads. The converter is used to take control of the voltage levels
enabling the ultracapacitor to be charged to a voltage higher than the dc bus voltage or less
than that, in other words to shape the voltage profile of capacitor during charging and discharging. Without the converter, the capacitor will get charged to the dc bus voltage without
any control over the charging current. The ultracapcitor is charged through the converter
at constant current when the main power is available, and during surges/blackouts the ca-
Chapter 1. Introduction
1.3
The project project has been organised as follows elaborating the various points mentioned
in the outline
1 The characteristics of Ultracapacitors is explained in the second chapter. The constructional features of ultracapacitors along with their structural differences from the
electrolytic capacitors is discussed. With the construction, the applications of the ultracapacitors as back-up energy storage devices are put forth. Finally for the specifications
mentioned in table 1.1, the ultracapacitor bank design and cell voltage monitoring are
explained.
2 The third chapter discusses the the design of the experimental set up which consists
of:
i The bidirectional buck-boost converter which in turn involves:
- Selection of the Switching devices along with their gate drive circuit.
- Filter inductor and capacitor design.
1.4. Conclusion
Value
VA Rating
240 VA
Maximum Load
85 W
Back up Capacity
10s
30V
24V
Switching Frequency
100kHz
ii The Digital controller dsPIC30F2023 board for the generation of PWM signals
iii The Feedback signals sensing circuit for sensing the currents and voltages for the
purpose of closed loop control.
iv Power factor correction circuit which becomes the primary source of power for
the loads.
3 The fourth chapter discusses the results. First the ultracapacitor characterisation test
results are presented. Then the closed loop control test results of the converter in buck
and boost mode of operation is discussed along with the bode plots for the design of
controllers . The fifth chapter gives conclusion of the designs done and the experiments
conducted.
1.4
Conclusion
This chapter gave an outline of the project. It discussed the voltage sag effects and reason
behind the project. The next chapter describes in detail the characteristics, construction and
applications of the ultracapacitors along with their sizing for the above mentioned system
specifications.
Chapter 2
Ultracapacitors - Modelling and
Sizing
This chapter describes the constructional features of ultracapacitor, applications of ultracapacitor and sizinf of ultracapacitor. Ultracapacitors store energy in electrostatic form with
the constructional features similar to that of a battery. A simple parallel plate capacitor
stores energy in the electric field between the plates, the capacitance given by
C=
A
d
(2.1)
Where A is the area of the parallel plates and d is the distance between the two parallel plates
and e is the permittivity of the medium. The electrolytic capacitor differs in construction
from the ordinary capacitor. An electrolytic capacitor has a metal electrode over which
there is a metal foil contact with the external circuit. There is a thin layer of metal oxide
formed on the electrode surface by the process of electrolysis i.e. passing current through a
electrolyte placed in contact with the electrode and the quantity of material deposited on
the electrode is given by the Faradays law of electrolysis
m=Z I t
(2.2)
where m is the mass of material deposited and z is the electrochemical equivalent and I*t is
the charge applied.
The electrolytic capacitor consists of the electrolyte which acts as the cathode and the
electrode acts as anode. There is a separator which is used to prevent contact of the anode
with the wall of the container. The breakdown strength of the oxide layer is of the order of
several MV/cm which is the reason for very high capacitance of electrolytic capacitor. The
6
typical metals used are valve metals like Aluminium, Titanium, Niobium and Tantalum,
Aluminium being more preferred because of its low cost. The electrolyte will vaporize at
high temperatures and crystallize at low temperatures making the ESR of the capacitor
requiring attention.
2.1
Construction of Ultracapacitor
2.2
The various energy storage devices like batteries, fuel cells, ultracapacitors are compared
using a log-log plot called Ragones plot. The Ragones plot is a plot of power density versus
energy density. figure 2.2 shows the Ragones plot from which the particular device can
Ultracapacitor
Batteries
Energy Storage
Electrostatic
Electrochemical
Charge/discharge cycles
Energy density
high
low
high
Shelf life
high
low
Weight
heavy
Maintenance
Maintenance free
Requires maintenance
2.3
2.4
The total power consumed by the control cards is given in table 2.2. The power requirement
of the Control cards of the inverter has been calculated to be 78W (considering a safety
factor of 2 for the total current drawn by the cards). For the above system specifications,
assuming an efficiency of 70% for the converter, the ultracapacitor is sized for 125W.
10
-15V
Annunciation card
0.052
0.005
1.71
0.238
0.033
8.13
0.736
PD card
0.122
0.066
0.504
10.64
DSP card
0.1
0.1
1.5
10
Total power
5V
11.04
45
78
2.4. Design of the Ultracapacitor Stack for the Ride Through System
2.4.1
11
Befor calculating the number of ultracapacitor cells required, it would be good to know the
discharge characteristics of ultracapacitors under constant power through the use of power
converter. The discharge characteristics of the ultracapacitor is given by figure 2.4. As per
(2.3)
dV = Vnominal Vmin
(2.4)
Imax =
Imin =
P
Vmin
P
Vmax
(2.5)
(2.6)
The maximum current is decided from the ultracapacitor datasheet, which fixes the minimum
voltage to which the ultracap can drop to.
dV =
Iavg
(t + )
Ct
(2.7)
In equation 2.7, is the timeconstant of the ultracap which is taken normally as 1.1s and t
is the time for the ultracap to provide ride throuugh, in this case being 10s. From equations
12
2.3, 2.4, 2.5, 2.6 and 2.7 the required capacitance is calculated to be 11F.
The number of cells required is calculated by
Ct = Ccell
parallelcells
seriescells
(2.8)
The number of series cells is determined by the voltage rating of the stack.
numberof cellsinseries =
Vmax
Vcell
(2.9)
The number of cells in series and parallel are calculated to be 12 and 1 respectively. Hence
the total number of ultracapacitor cells was chosen to be 12. The ultracapacitors used were
Maxwell BCAP0150 capacitors of capacitance 150F.
2.4.2
The maximum voltage withstood by each ultracapacitor is about 2.5V or 2.7V, limited by
the electrolyte decomposition voltage. The series connection of the ultracapacitor leads to
unbalance in the voltage across each cell because of the differences in self-discharge rates.
This might lead to excess voltage on one cell and a lesser voltage on some other cell. To
prevent overvoltage across a particular cell, voltage equalization is done.
The equalization method adopted here is connecting a series string of Diodes, LEDs and
2.5. Conclusion
13
is more than forward voltage drop of the diodes plus that of the LEDs, the LEDs starts
glowing thereby discharging the cell. A 1.5 resistor was chosen to be used along with one
diode (BA159) and one LED.
The ultracapacitor hardware is shown in the appendix figures A.3 and A.4
2.5
Conclusion
The chapter explained the construction of ultracaps, difference between the various energy
sources, sizing of ultracapacitors for the ride through system and the voltage management
circuit. The tests done to test the capacitance and leakage resistance of the ultracaps along
with the characterisation results is explained in later chapters.
Chapter 3
Hardware Design
In this chapter, the hardware, which includes the power circuit, controller board, current
and voltage sensing card, has been discussed.
3.1
Power Circuit
The power circuit used is a 240VA bi-directional buck-boost converter. It comprises the
choice of switching devices, design of the filter elements, Device gate drive circuit.
3.1.1
3.1.2
The driver circuit for the MOSFET consists of an optocoupler IC HCPL3101 to isolate the
control circuit from the power circuit and IR2110 which is the driver with separate High
and Low side referenced output channels. The driver circuit is powered by the power circuit
itself with isolation through a flyback converter. The gate drive voltage obtained from the
IR2110 is in the range of 10-20V. The bootstrap capacitor is selected as 10F based on the
formula given in the IR2110 datasheet.
14
3.1.3
15
The switching frequency is chosen as 100kHz. The inductor is designed for a current ripple
of 0.2A as follows
dI
(3.1)
dt
dI is the current ripple of 0.2A. dt is the on/off time of the switch. The inductor details
V =L
are as specified in the table 3.1. The value of V is chosen such that the product V*d is
300H
Switching Frequency
100kHz
Peak Current
9.1 A
0.2A
Core
Ferrite, E65/32/27
SWG
13
Air gap
1.6 mm
maximum. This occurs at the minimum value of the ultracapacitor voltage. The required
inductance is calculated as 300H. The filter capacitor is chosen depending on the voltage
ripple which is chosen to be less than 0.1 %. The capacitor is calculated to be 5000F. Hence
five nos. of 1000F ,63V Electrolytic capacitors are chosen with a 1F ,63V high frequency
capacitor. The high frequency capacitor is used in parallel with electrolytic ones because at
high frequencies the electrolytic capacitor starts becoming resistive thereby exhibiting higher
voltage ripple at switching frequencies.
3.1.4
(3.2)
16
2. Reverse conduction loss The reverse conduction loss in the passive switch(Body diode
of complementary Mosfet) is calculated to be
Vav Iav = 3.86W
(3.3)
3. Switching loss
tr
0.5W
T
tf
ID 0.5W
T
3.1.5
(3.4)
(3.5)
The entire power circuit schematic is given in the appendix. The current in the circuit is
sensed through the wire wound current sense resistor of 0.1. LM 2576 is a 12V,3A Buck
converter IC from National Semiconductors. It eliminates the need for external power supply,
hence keeping the area of experimental set up small. It is used to power
1. Gate drive card
2. dsPIC30F2023 Controller board.
It gets power from the upstream mains when it is present and from the ultracapacitor fed
boost converter in case of outage. All the components used in the power circuit along with
their values are listed in appendix. The power circuit with the various constituents are shown
in the appendix figure A.5.
3.2
Theory of operation
The block diagram of the ride through system was shown in figure 1.2. The power circuit
used in the system is shown in figure 3.1. The switches Q1 and Q2 are MOSFETS with antiparallel diodes. The circuit represents one leg of an H-bridge circuit. The circuit operates in
buck mode to charge the ultracapacitor. The circuit diagram in buck mode is as shown in
3.2 The switch Q1 is turned on by Gate pulse and the diode D2 freewheels to conduct when
Q2 is turned off. The controllers are designed such that the current through the inductor
is constant and the Ultracapcitors are charged at constant current.When there is an outage
the ultracapacitor discharges through the active switch Q2 and the diode D1. Q2 is gated
17
18
and D1 freewheels. This is the boost mode of operation. The controllers are designed so
that Q2 switches accordingly by varying the duty ratio to the switch.
3.3
Controller Platform
The controller is implemented digitally through dsPIC30F2023. The existing controller board
[3] was used without modifications. The schematic given in appendix is also referred from
[3]. The Controller board is shown in figure appendix A.7.
3.4
To perform the controller action, all the voltages and currents should be converted into the
processors voltage range, here this being 5V. For this purpose, a LM-324 quad Op-amp
based cerrent and voltage sensing circuit was designed. The power for the LM-324 is fed
through the LM 2576 power supply IC. Hence the op-amps power supply terminals are
+12V and Ground. Negative potentials cannot be represented. The output of these cards
are fed to the ADC pins of the processor through an anti-aliasing filter circuit with a cut off
frequency less than 1kHz.
3.4.1
19
Voltage Sensing
Two voltages DC bus voltage and ultracapacitor voltage are sensed. The voltages are stepped
down by a magnitude of 10 using the non-inverting amplifier circuit. The circuit for voltage
sensing is shown in 3.4
3.4.2
Current Sensing
The current is sensed through the current sense resistor of value 0.1. A capacitor is used
in parallel with the resistor, the resultant time constant being very less than switching time
of the converter. The current sensing part is used to produce a voltage as per the equation
VOadc = 2.5 + (0.167 I)
(3.6)
Equation 3.6 is used to produce a output of 5V for a maximum current of 15A and 0V for
a maximum negative current of -15A. The circuit implementation and the hardware of the
sensing circuit are shown in fig. 3.5 and appendix figure A.6. This chapter explained the
design of entire hardware starting with power circuit to controller design and signals sensing
circuit. The next chapter explains the controller design.
20
Chapter 4
Controller Design
This chapter explains the design of controllers and their transformation to the z-domain and
finally implementation in dsPIC30F2023 through C programming.
4.1
The closed loop control is done to prevent variations in the output voltage or output current,
due to any disturbances, by introducing a control variable in the form of duty ratio. This
control variable d, is used to turn on the active devices in the switching circuits for a duration
corresponding to its value. Closed loop control varies the duty ratio, in case of any deviation
in output parameter, from its present steady state value to a new steady state value thereby
correcting for the variation in the output.
4.1.1
Control Objective
22
23
of inner current loop as unity. Then the controller is designed for the above objectives and
bandwidth of the outer voltage loop is kept atleast one-tenth the bandwidth of inner current
loop. This is because when the outer voltage loop is active, the actual current in the circuit
should have reached the set current reference. The outer voltage loop sets the inner current
reference. The references, output capacitor voltage and inductor current, are the dynamic
variables in the system.
24
4.2
The buck converter was first tested with a resistive load, before charging the ultracaps. The
control involved providing an inner current loop and an outer voltage loop. The bandwidth
of the inner current loop is kept as one-tenth the switching frequency. In the following
section, the controller equations, bode plot and their implementation in z-domain is given.
4.2.1
The figure 4.1 shows the control structure adopted. The bandwidth of the voltage loop is
kept as one-tenth the bandwidth of the current loop. The small signal transfer function of
current to control is given by
bi(s)
Vg
1 + sRC
gb(s) = b =
L
R LCs2 + s R
+1
d(s)
The bode plot of the transfer function with the experimental values of
Vg-19V,R-10, C=1000F, L=300 is shown in figure 4.3
bi(s)
gb(s) = b = 1.9
1+
d(s)
1+
s
100
s
33333
s 2
1825
(4.1)
s
b
1 + 1825
d(s)
=
s
ebi (s)
250
(4.2)
The bode plot of the loop gain GH is given in figure 4.4 The digital implementation of the
above PI controller is done using bilinear transformation. The controller implemented with
a sampling frequency of 20kHz is
kp [n] = (0.144 error[n])
(4.3)
(4.4)
The duty ratio is generated by the addition of equations 4.3 and 4.4
d d(s)
d
Figure 4.3: Current to Control transfer function i(s)/
Figure 4.4: inner current loop - Loop gain GH transfer function i(s)/ iref (s)
25
26
4.2.2
(4.5)
The Voltage loop PI controller is chosen for a bandwidth of 1000 rad/sec. The transfer
function of the PI controller used is
b
h(s)
=
s
1 + 100
id
ref (s)
=
s
ecv (s)
100
(4.6)
(4.7)
(4.8)
(4.9)
The reference current for the inner loop is set by 4.9. The bode plots for the outer voltage
loop are shown in figures 4.5 and ??
27
50
40
Magnitude (dB)
30
20
10
0
-10
-20
Phase (deg)
-45
-90
0
10
10
10
Frequency (rad/sec)
10
10
Figure 4.6: Outer Voltage Loop - Loop gain GH transfer function V(s)/Vref (s)
28
4.3
The control structure adopted for charging ultracapacitors is the same as shown in figure
4.1. The controller gains are different.
4.3.1
s 2
16
(4.10)
s
b
1 + 1000
d(s)
=
=
s
ebi (s)
253
(4.11)
(4.12)
(4.13)
(4.14)
The duty ratio is given by equation 4.14. During charging of the ultracapacitors, the duty
ratio varies from zero to one, with the ultracapacitor stack charging slightly over the set
reference and finally the duty ratio settles at zero at this overcharged value. The bode plots
of the current to control transfer function and current loop gain are shown in figures 4.7 and
4.8
d d(s)
d
Figure 4.7: Current to Control transfer function - ultracap charging i(s)/
29
30
4.3.2
In this case the outer voltage loop controller can be just a proportional controller. The small
signal transfer function of voltage to current is given by
vb(s)
1
gb(s) = b
=
Cs
i(s)
vb(s)
1
gb(s) = b
=
12.5s
i(s)
(4.15)
The Voltage loop PI controller is chosen for a bandwidth of 1000 rad/sec. The transfer
function of the PI controller used is
b
h(s)
=
id
ref (s)
= 12500
ecv (s)
(4.16)
(4.17)
(4.18)
4.4
31
The boost converter control structure implemented is a voltage loop control. Here, in addition to a PI controller, a lead compensator is used to improve phase margin at the gain
cross over frequency. The voltage to control transfer function is given by
L
1 s R((1D)
2
vb(s)
Vg
gb(s) = b =
LC
L
2
2
R(1 D) 1 + s R((1D)2 + (1D)
d(s)
2s
In the experiment done, the values used were C=1000F, Vg=10V, L=300H, R=20,
D=0.375 The transfer function turns out to be
s
1 26042
vb(s)
gb(s) = b = 38.4
s 2
s
+ 510
1 + 26042
d(s)
b
h(s)
=
(4.19)
s
s
b
1 + 510
1 + 510
d(s)
=
s
s
ecv (s)
1 + 10000
208
(4.20)
error(s) in equation 4.20 is the voltage error which is given by the difference in the voltage
between the set reference value through the processor and the sensed value from the system.
Bode plots of the voltage to control transfer function and the loop gain transfer function are
shown in figures 4.9 and 4.10 respectively.
100
80
60
Magnitude (dB)
40
20
0
-20
-40
-60
-80
-100
Phase (deg)
-90
-180
-270
2
10
10
10
Frequency (rad/sec)
10
10
d d(s)
d
Figure 4.9: Voltage to Control transfer function - ultracap discharging v(s)/
32
80
60
Magnitude (dB)
40
20
0
-20
-40
-60
0
Phase (deg)
-45
-90
-135
-180
-225
-270
1
10
10
10
10
Frequency (rad/sec)
10
10
Chapter 5
Experimental Results
In this Chapter, the results obtained as a result of experiments conducted are presented.
5.1
Ultracapacitor Characterisation
The first in the series of experiments was to test the Ultracapcitors individually for their
Capacitance and leakage resistance. For this, the ultracaps were charged at constant current
to a voltage slightly below the rated voltage using a 30V,2A power supply and discharged
through a 5 load. The data was logged using YOKOGAWA MX100 Datalogger. To
monitor the leakage resistance, the cells were charged to 2.5V and disconnected from external
circuits. The voltages across the cells were monitored over a weeks period at 24 hour interval.
Table 5.1 gives the values of the 12 ultracapacitors being used. The values logged were used
in excel and line of best fit was constructed and the values found. The values for the above
12 capacitors were plotted using excel and the plot is shown in figure 5.1
5.2
The voltage monitoring circuit was tested with a comination of resistor, LEDs and diodes
of different ratings. The results which were consequential in choosing the components used
in the ultracapacitor board are summarised in figure 5.2 The characteristic required should
be such that at over 2.5V, the ultracap should exhibit a significant leakage current to bring
the voltage down to 2.1 or 2V, at which the leakage current should be almost zero. As per
the figure 5.2, the characteristic is satisfied by a series combination of 1 LED, 1.5 resistor
and 1 diode, which is hence chosen.
33
34
123.43
9.528
123.77
10.26
131.33
9.838
146.11
9.094
148.14
9.646
150.69
10.24
152.72
10.63
153.37
11.42
158.05
10.11
10
158.78
11.4
11
159.62
12.6
12
160.43
16.97
35
5.3
5.3.1
Charging at 1A
The input voltage was kept fixed at 24V. The charging was done with a charging current
reference of 1A. The scope waveform shows the capacitor getting charged from 7V to 23V in
almost 175secs. This is validated below. For an ultracapacitor current of 1A, output from
sensing circuit should be 2.67V which can be seen on scope. Figure 5.3 shows the charging
at 1A.
dv
dt
Iadc = 2.5 + (0.167 1) = 2.67V
C
12.5
t = dv = (23 7)
= 200s
I
1
I=C
(5.1)
(5.2)
(5.3)
36
Charging at 2A
The input voltage was kept fixed at 24V. The charging was done with a charging current
reference of 2A. The scope waveform shows the capacitor getting charged from 4V to 23V
in almost 90secs. This is validated below. For an ultracapacitor current of 2A, output from
sensing circuit should be 2.83V which can be seen on scope. Figure 5.4 shows the charging
at 2A.
(5.4)
(5.5)
37
C
12.5
= (23 4)
= 125s
I
2
(5.6)
38
5.3.2
First the boost converter was tested with the controller programmed to take a 4V input and
generate a 6V output with the latter as the reference voltage. The following describe the
waveforms captured on scope
1. Green Waveform : Input Voltage,
2. Blue waveform : Output Voltage.
The above waveforms are each 2V/div for figures 5.5,5.6,5.7 and at 10V/div for 5.8,5.9 and
5.10.
1. Figure 5.5 shows output voltage of the boost converter whiich was at slightly less than
4V reaching 6V in nearly 1sec. The voltage spikes on the waveforms are switching
spikes and power supply noise.
Figure 5.5: Boost Converter Constant Voltage Control - 4.7V input, 6V output, 20ohm load
2. The figure 5.6 shows a step change in input voltage from 4V to 5.7V. The output
voltage increases momentarily and settles down at 6V in about 1sec.
3. Figure 5.7 shows a step change in voltage from 5.6V to 4.5V, the output decreases
alongwith the input and finally settles down at the set reference of 6V.
Figure 5.6: Boost Converter Constant Voltage Control - Step Increase in Input
Figure 5.7: Boost Converter Constant Voltage Control - Step Decrease in Input
39
40
The boost converter was then tested to simulate the voltage decrement across the ultracapacitor from 24V to 14V. This was done by varying a 60V,5A power supply from 24 to 14V,
although the load in the intended application is 6. Here the test was done with a 20,50W
resistor. The various test results are as follows
All the tests were done with a reference voltage of 25V.
1. Figure 5.8 shows the converter reaching steady state voltage of 25V from initial voltage
of slightly over 18V when the controller is turned on. The steady state voltage is
reached in almost 1sec.
Figure 5.8: Boost Converter Constant Voltage Control - 15V input, 25V output, 20ohm load
2. Figure 5.9 shows a step increase in the input from 15V to 22V. The output voltage
increases and finally settles down to 25V in about a sec.
Figure 5.9: Boost Converter Constant Voltage Control - Step Increase in Voltage
41
3. Figure 5.10 shows a step decrease in the input from 22V to 15V. The output voltage
decreases and finally settles down to 25V in about a sec. This is the condition that
happens when the ultracap is conducting, where the voltage decreases from 24 to 14V.
Figure 5.10: Boost Converter Constant Voltage Control - Step Decrease in Voltage
42
5.3.3
The figure 5.12 shows the discharge profile of ultracapacitors under a 50W (23V, 10) load.
It can be seen that the discharge takes place for almost 25secs, after which the PWM is
turned off because the ultracapacitor has dischared to its minimum voltage level which has
been set as 15V.
43
5.3.4
The integrated opeartion is about creating the effect of voltage sag at the terminals of the
ultracapacitor and the converter and letting the ultracapacitor discharge with a 75W load
and the power supply would be turned on to charge the capacitors once again. Figures 5.13
44
45
5.4
The Power circuit was tested in the buck mode with a 10 load for efficiency and with duty
ratios varying from 0.5 to 1. The results are tabulated as in table 5.2
0.4
7.75
0.78
0.4
79
19
0.266
6.23
0.63
0.33
76.8
19
0.536
9.26
0.93
0.487
84.2
19.1
0.57
90
19.1
0.85
12.25 1.22
0.64
92.2
19.1
0.72
94
19.1
0.79
94
19.1
1.548
0.87
94.3
16.7
1.67
46
5.5
Thermal Test
A temperature rise test was done on the entire system by running the converter in buck
mode. The results were monitored using FLUKE TI20 THERMAL IMAGER. The
pictures taken using the thermal imager are shown below. It can be seen that the processor
dsPIC30F2023 is the hottest in the entire system. The current levels in the power circuit
were not too high to cause any significant copper loss in the inductor or conduction loss in
the conducting devices. The figure 5.17 shows the temperature rise of the half bridge and
47
Chapter 6
Conclusions
The project on providing ride through for control power supply could be completed with the
following points noted during course of the project
1. The boost converter control was succcessfully tested and the results captured with
both power supply and the ultracapacitor input.
2. The charging of ultracapcitors in buck mode was tested at two current levels - 1A and
2A and the charging profile was found to be satisfactory. Voltage spikes of very high
magnitudes were experienced both at lower and higher voltage levels, which might have
a detrimental effect on the life of ultracapacitors.
3. The integration of buck and boost converter operations and hence testing of the whole
system could be completed successfully with the tests carried out by turning off the
power supply and allowing the ultracapacitor to feed the load and turning on the power
supply, whence it started charging back to the set reference.
The back up offered for various loads is shown in table 6.1.
The following issues might also be interesting.
1. Gate Drive Card : When charging was tried, with the capacitor already charged to
slightly more than half the reference voltage, the controller immediately saturated, but
the high side switch of the half bridge remained switched off. This finally turned out
to be an issue with the bootstrap capacitor on the Gate driver IR2110. If the low side
48
49
50s
50W
25s
75W
9s
switch never turns on, there is no other way to charge the bootstrap capacitor. So an
isolated 12V power supply was used for charging this bootstrap capacitor.
2. Issues with dsPIC30F2023 target board : There was a plenitude of processor failures.
The reason is attributed to momentary overvoltages above its rated voltage of 5V
through the Analog input channels.So a 4.7V zener was used to avoid the processor
failures. The processor failures can also be reasoned due to the temperature rise. So
forced cooling (using fans) has to be adopted to avoid the processor failures. It might
also be because the target processor board layout is not satisfactory on the thermal side.
Sufficient space has to provided around the dsPIC30F2023 processor before connecting
it to the other components.
3. Another issue with the converter was that, the controller programs wouldnt run in
debug mode with the boost converter mode of operation and hence had to done in programmer mode. The reason is, in buck mode of operation, the main 12V power supply
powered both the converter and the processor with the MPLAB ICD2 experiencing
no undervoltages due to the switching. But in boost mode of operation, main power
supply is absent and the switching spikes got propogated through the current sense
resistor, which is connected to ground, whereby the debugger experience undervoltages
and shutdown in the middle of execution/running the target. This was solved by executing the processor in a stand alone mode without the debugger and connecting the
debugger only when it was necessary to program.
4. The final issue being the switching spikes which need to be eliminated. This is because
of stray inductances due to the usage of long wires in various paths.
Appendix A
Schematics and Experimental Setup
Power Circuit Schematics
5
JP5
Vbus
D
1.2k
1
2
Vproc
1
2
GND3
Vp
R2
1
2
PWM1
GND3
4.7k
Vcc1
Q1
2N2222
2
2
JP3
8
7
6
5
C3
10u
Isense
1
2
C4
0.1u
GND1
GND3
Vcc
JP4
1.2k
1
2
3
PWM2
GND1
PWM2
1
2
1
2
3
4
R5
JP3
U2
R4
1
Vp
GND1
Vp
HO
VB
VS
NC
VCC
COM
LO
7
6
5
4
3
2
1
Q2
2N2222
8
7
6
5
GND2
C2
0.1u
C6
0.1u
C8
0.1u
D1
BYW29-150
Vcc
N
R8
1
L0
12
GND2
R6
4.7k
C7
10u
HPCL3101
4.7k
2
H0
2
C5
10u
IR2110
PWM1
NC
VDD
HIN
SD
LIN
Vss
NC
R7 12
2
U3
8
9
10
11
12
13
14
GND2
JP7
GND2
C1
0.1u
HPCL3101
R3
4.7k
1
2
1
2
3
4
JP6
U1
R1
GND3
JP2
Vcc2
1
2
GND
Vcap
Vcc1
A.1
GND1
POWER CIRCUIT
M
H0
TP2
Vcap
2
300uH
FUSE
TP1
R9
1
10k
L1
F1
TP5
TP6
D2 15 V
1
2
IRF540N
D
Vbus
M2
TP3
TP10
1
1
B
C12
100u
1
LM2576
C11
C10
2
L0
R10
10k
1
D3
N
1000u,63V
1000u,63V
U4
1u,63V
1
2
3
4
5
6
2
680uH
D11
C13
1000u,35V
Vcc2
1N5814
GND3
C9
IRF540N
M2
2
1
2
3
4
5
6
L2
NC
Vproc
1
15 V
GND3
TP8
TP9
1
TP7
TP4
R11
GND
2
0.1 ohms, 10W
GND3
Isense
5
50
A.2
51
Controller Schematics
52
A.3
A.4
53
54
A.5
A.6
55
56
A.7
References
[1] R. Srinath , Digital control of solar photovoltaic converters, M.Sc. thesis, EE Dept,
IISc., Bangalore, December 2009.
[2] Kelly et al, Voltage regulator for contactor ride through ,IEEE transactions on Industry
Applications, Volume 36, NO.2, March/April 2000, Pages 697-704.
[3] C. Karrupuswamy , Bi-directional inverter with high frequency ac-link,, M.Sc. thesis,
EE Dept, IISc., Bangalore, March 2009.
[4] S. Giridharan, A novel transformer-less uninterruptible power supply, M.Sc. thesis, EE
Dept, IISc., Bangalore, September 1996.
[5] V. Ramanarayanan ,Course material on switched mode power conversion, IISc., Bangalore.
[6] G.S. Ramana Murthy, Design of inductors and transformers at power frequecies - A
modified Area-Product approach , M.Sc. thesis, EE Dept, IISc., Bangalore, March 1999.
[7] Robert W. Erickson and Dragan Maksimovic, Fundamentals of Power Electronics , II
edition, Springer India Academic Publishers, 2001.
[8] IRF540N datasheet,International Rectifier,http://www.irf.com.
[9] HCPL-3101 datasheet, Avago Technologies,http://www.avagotech.com.
[10] dsPIC30F Family Reference manual,http://www.microchip.com.
[11] IR2110 datasheet,International Rectifier http://www.irf.com.
57
58
References
[12] Ultracapacitor Application notes, datasheets and case studies, Maxwell Ultracapacitors,
http://www.maxwell.com.
[13] Ultracapacitor brochure, LS ultracapacitor.
[14] Camara et al, Design and New Control of DC/DC converters to Share Energy Between
Supercapacitors and Batteries in Hybrid Vehicles, IEEE transactions in Vehicular Technology, Vol.57, no. 5, September 2008, Pages 2721-2735.
[15] Jasvinder Singh Khoral,Power Factor Correction of Switching Power Supplies using
UC3845, M.E. project report, EE dept. IISc., Bangalore, June 1994.
[16] Sarmiento and Estrada,A Voltage Sag study in an Industry with Adjustable Speed
Drives, IEEE industry applications magazine, January/February 1996.
[17] Ragones plot,http://www.wikipedia.org.