April Newsletter 048
TI India
Analog Design
Contest
Vol 3, Issue 2, April 2010
Table of Contents
Editorial- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -1
Design and Development of a Low Cost Video Bronchoscope - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2
Obstacle Detection in Flooded Roads - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -4
A Portable mobile phone charger using Pico Wind-Turbine - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 8
Student Attendance Monitoring System
using Voice Analysis & RFID - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 9
Signal Conditioning Unit for Field Deployable Microimpedance
Biosensors - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -12
MSP430 based Electronic Load Control for
Micro Hydro Power Plants - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 15
Maximum Power Point Tracking of Photovoltaic Arrays
using Adaptive Perceptive Particle Swarm Optimization Technique - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -17
Remote Home Lighting and Appliance Control System - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -20
Feedback Amplifier! - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -22
TI Tech Days on Embedded Processing at Bangalore - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 24
UniTI on Campus – Events in 1Q, 2010 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 26
Low Power RF Workshop at BITS–Goa Campus - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 27
A one-day workshop on Power Management - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -28
The Bibliophile - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -29
Editorial
The April 2010 issue of the UniTI newsletter is a special issue focusing on the Texas Instruments
India Analog Design Contest (2009). It is a little more than a year since we started preparations for
the launch of the Analog Design Contest. The intention of the contest was to encourage systemlevel design among the undergraduate student community in Indian academic institutions that offer
Engineering degrees. When I was sending out the posters for the contest last year, I was somewhat
apprehensive about the response we may receive. I was pleasantly surprised when we received an
overwhelming response from a large number of colleges. Our original goal was to shortlist 5
colleges, but the response prompted us to revise our goals to support double that number.
Running a National-level contest has been a learning experience for me. Collating the requirements of the teams, placing
orders and providing components for thirty teams from ten colleges that were geographically distributed, communicating
regularly with the teams to track their progress, getting the proposals and the final submissions reviewed, organizing the
Awards ceremony … While these were the main steps in the process, we had plenty to do behind the scenes.
I believe the first Texas Instruments India Analog Design Contest has been a great success. I congratulate all the teams
who participated and thank all the mentors who provided their time and precious guidance to the students. One of the
reviewers from our Dallas office commented – “I am very impressed with the quality of thought and creativity of virtually
all of the proposals that I have reviewed.” It was great to see the enthusiasm of the student participants and the kind of
projects they selected. They chose to work in exciting areas such as medical instrumentation, renewable energy, security,
and entertainment. It was not easy for us to shortlist the winners of the first phase – there was such intense competition
among teams!
The main improvement to the contest that has been suggested by most of the mentors as well as some of the reviewers is
that the projects need more time. The project submissions came from 3rd year and final year undergraduate students,
who were multiplexing their projects with the regular
coursework. On page 22 of this issue, you will find some
suggestions that may be useful to you if you are thinking of
making a submission to ADC 2010.
It is my pleasure to acknowledge the help I got from several
sources during the execution of ADC 2009. First and foremost,
I thank Syd Coppersmith and Mealine Calahan, who sponsored
the contest. Thanks to Gaurav Jabulee, Robert Owen, and Paul
Prazak for their help at different stages of the conference. Many
other colleagues from TI India helped us at different stages and I
am unable to list their names for paucity of space. Cranes
Software, TI India's University Partner, helped in publicizing the
contest through posters. I received help from over 30
reviewers, both from inside and outside TI. I must thank all the
colleges and the participants that sent proposals and all the
faculty mentors for their guidance. My interns Adithya
Gajulapally and Sagar Juneja have been very helpful in running
the contest.
Art by Ananya Ravikumar
I congratulate all the teams for their achievement. I look forward
to receiving the entries for ADC 2010! As always, we are open
to your comments and suggestions.
C.P. Ravikumar
“Prize 1.234 goes to Team 3.14178 ...”
1
Design and Development of a Low Cost Video Bronchoscope
“Non-invasive” is a
keyword in medical
equipment. The
Wikipedia defines
“non-invasive” as a
medical procedure
where the skin is not
broken and there is
no contact with the
mucosa. M. Aravind
Krishnan, N.
Hariprasad and S.
Ganapathy Subramanium, students of Anna University,
Chennai, were motivated by Prof. P.V. Ramakrishna to
build a low-cost medical equipment that does not
compromise on quality. The team proposed a
bronchoscope for capturing a clear image of the
respiratory system in a non-invasive manner. They say
doctors can use a bronchoscope for diagnosis of
pulmonary tumors or for surgical procedures such as
endoscopy and administering general anesthesia.
Bronchoscopes are of two types, fiber-optic
bronchoscope, which uses a fiber-optic cable for imaging
the airways, and video bronchoscope, which makes use of
CCD or CMOS image sensors for imaging. The fiber-optic
bronchoscope uses extremely fragile fiber-optic cables
and costs more. Since the students of Anna University
wanted to minimize the system cost, they elected to
design a video bronchoscope, which offers the additional
benefit of portability.
The proposed bronchoscope has two parts – the distal end
and the proximal end, both of which were built and tested
successfully. The distal end is the tip that is inserted into
the patient to capture the image of the airway.
A challenge that the team faced was to conform to size
constraints of commercially available bronchoscopes.
Providing a clear video of the innards of the human
respiratory system with very low illumination is another
technological challenge. The team chose a tiny video
sensor, about 2 mm in diameter, accompanied by a 1.5
mm LED, to form the distal tip of the bronchoscope. Since
the bronchoscope must provide a way to record and store
videos for later diagnosis, the team used MPEG4
compression. Image processing such as periodic noise
reduction were implemented to improve the quality of the
video. When they tested it on external objects, the
bronchoscope was able to provide clear images of objects
placed within a close range.
2
Distal End
The distal end comprises of a light source to illuminate the
interior respiratory system using a white SMT LED.
Imaging Sensor with lens apparatus is responsible for the
actual imaging of the respiratory system. The team used
Omnivision OV6920 sensor for this purpose. This sensor
has a very small imaging aperture (1/18th of an inch) but a
good resolution (320 x 250pixels) and viewing angle
(around 120o). The sensor is controlled using a PIC
microcontroller using an SPI interface. The output of the
sensor is in the NTSC format. Certain image processing
algorithms like periodic noise reduction are implemented
to improve the video quality.
Proximal End
The proximal end is located near the doctor and its
function is to process the images obtained from the
sensor at the distal end. A Video Decoder is used in the
proximal end to extract the video information from the
incoming video signal and convert it into digital data. The
video decoder also uses the framing and synchronization
signals from the composite video signal so as to
synchronize the chunks of video data available from its
output. The video decoder used here is Texas Instruments
TVP5146, which can be programmed easily. A video
processor (Texas Instruments TMS320DM355) was used
to encode the video for efficient storage, transfer the
video to an image processor for further analysis, and view
the image on a 7” TFT LCD Screen.
Using the NAND flash memory of the DM355, the board
was mounted on a Linux host workstation through NFS
mount. The processor was programmed to perform
MPEG4 encoding and the resulting video stream was
transmitted to the workstation using an Ethernet cable
over Internet Protocol. The video processor also performs
digital to analog conversion of the encoded video data to
drive a composite video player. The Texas Instruments
OPA360 video amplifier was used for amplifying the
analog video signal obtained from the video processor
and provide matching for driving an LCD Screen.
The LCD screen reduces the system cost, but is capable of
displaying good images of reasonable resolution. The
video obtained from the image sensor has a high amount
of periodic noise along the y-axis which is eliminated
using image processing. The periodic noise had a
frequency of 0.1 times the sampling frequency and was
eliminated using a simple image notch filter. The filtering
has been implemented in software (MATLAB) at this
stage, and will be implemented in real time using the TI
DM355 processor.
SMT 5 mm
LED
1 2
LCD Screen
3
Vdd
Clk
SCLK and Sdata
OPA360
Video
Amplifier
CVO
Ov6920 Image
Sensor
Distal Tip
Catheter
Legend and Key
1. Battery and 5V Regulator
2. 6.14 MHz Crystal Oscillator
3. PIC12C508A Microcontroller
TVP5146
Video
Decoder
TMS320DM355
Digital Video
Processor
Frontal End
- Analog/Power Signal
- Digital Video Signal
Figure 1: Block Diagram of the Proposed Bronchoscope
Challenges
UniTI got in touch with the team about the challenges
they faced during the project, and this is what they had
to say.
Aravind: The illumination sensitivity of the image sensor
was lower than what we required for this application.
Further, the video obtained from the image sensor had a
periodic noise; we need to employ image processing to
eliminate this noise.
Hari: The camera chip module obtained from the
manufacturer has to be soldered to a thin, flexible and
cylindrical surface with diameter not exceeding 3 mm. The
mechanical casing of the bronchoscope forms the most
crucial part of the system design; it is yet to be devised.
UniTI: What were the highpoints of the projects?
Gana: The electronics for the project has been fully
implemented and we believe it is a feasible design.
Although it requires some improvements, the current
system is cost-effective compared to the commercially
available bronchoscopes.
UniTI: How was the system tested?
Answer: We had tested the system using a very crude
phantom built using a cardboard box. The cardboard box
has a small hole which could be used for sending in the
distal end of the bronchoscope.
UniTI: What is the approximate cost of the project?
Gana: LCD Screen costs Rs 2000, Video processor costs
Rs 565 and encoder costs Rs 225. The total cost is
around Rs 2800.
UniTI: Congratulations & wish you all the best!
Many thanks to Prof. P.V. Ramakrishna for mentoring
the project!
3
Obstacle Detection in Flooded Roads
Each year, when the rainy season hits Indian cities, we read about the flooding of low-lying areas,
roads submerged under water, and accidents involving people who fell into potholes or manholes
when trying to cross the road. There was a particularly tragic story in Bangalore about 8 years ago,
when a family driving a car was drowned in an open gutter which was submerged under water
and acted as a deathtrap for an unsuspecting driver.
Three students of IIT Madras, Anoosh G, Raghunandan S and S Sundar Aditya, decided to do
something about preventing accidents that occur in the rainy season. Their aim was to implement
an obstacle detector which will detect abrupt changes in terrain – manholes, potholes, bumps and
drains – and warn the user. While the obstacle detector is useful to anyone who may be using a
flooded road in a rainy season, it can be used by a blind person even in dry season.
“The objective has been to realize a robust, minimal-power system with a small footprint so that it
can be part of a cellphone, an automobile, or a walking cane used by visually impaired people,”
explained Anoosh, a final year student of Electrical Engineering in IIT Madras.
“Our intention has been to use the principles of SONAR to determine the nature of the terrain located a few meters ahead
– the look-ahead distance, as we call it. A short duration ultrasound pulse is emitted periodically and directed a few meters
in front of the person/automobile carrying the obstacle detector. The reflected signal from the terrain is analyzed to gather
information about the terrain and warn the user of any abrupt discontinuity beyond a tolerable value,” adds Sundar Aditya.
Tx
Rx
Ultrasound
transducer
Reflection from
surface
Receiver
Amplifier
ADC
Processing
(Depth measurement,
Discontinuity
Checking)
Clock
Denotes TI components
Discontinuity
No
Yes
Alarm
Figure 1: Block Diagram of Obstacle Detector
The students made use of Texas Instruments MSP430
microcontroller in their project, along with ultrasound
transducers and receivers (see block diagram in Figure 1).
The ultrasonic transducer emits pulses at regular intervals.
A receiver block listens to the echo. The echo round-trip
time is a function of the height of the system. It is assumed
the detector will be a hand-held device at the waist level,
say 1 m above the ground. The receiver detects abrupt
changes in terrain by comparing the round-trip times. The
look-ahead distance is a function of the speed at which the
user is moving, and a warning must be provided before it is
too late. For pedestrian usage, the look-ahead distance
was taken as 1.5 m, assuming a nominal walking speed
4
of 1 m/s. Assuming that acoustic waves travel at 340
m/s, the round trip time may be calculated as 10 ms.
Thus the sensor needs to be triggered at a time period
larger than 10ms. Any variation in round-trip time above
or below a threshold is detected as a discontinuity.
“We used the LV-Max Sonar EZ1 sonar range finder as a
sensor, which needs a 5V supply; this is provided using a
voltage regulator (TPS71550 or TPS71537 from Texas
Instruments). The output of the detector is a PWM
waveform, whose on-time varies according to the
distance of the obstacle,” explained Raghunandan
The pseudo-code of the control algorithm and associated sub-programs is shown in Figure 2 below.
1.
2.
3.
4.
5.
6.
7.
Initialize the microcontroller - enable global interrupts and clear CPU
registers
Initialize Port 2:
a.
Set the direction of the pin p connected to the sensor to 'input'
b.
Enable interrupts for pin p and select the interrupt edge to be '
rising edge sensitive'.
Initialize Timer A:
a.
Set comparator register r to capture mode and connect the input of
capture
compare block to pin p
b.
Select SMCLK as the clock input to the timer.
c.
Set clock divider to 1.
d.
Set the timer mode to continuous.
Sensor Interrupt Service Routine:
a.
If the pin is rising edge sensitive then,
i.
Reset the timer.
ii.
Set the Port 2 interrupt to falling edge sensitive.
b.
If the pin if falling edge sensitive then,
i.
Call Pulse_width_calculate().
ii.
Set the Port 2 interrupt select to rising edge sensitive.
c.
Reset the interrupt flag.
Timer Overflow_Interrupt Service Routing:
a.
Increment a global variable ovflow.
b.
If the input is low, set the variable ovflow to 0.
c.
Register R4 <= ovflow; /* R4 stores the number of overflows */
d.
Reset the interrupt flag.
Routine Pulse_width_calculate:
a.
Register R5 <= Value in capture/compare register of timer
b.
Call function Check_pothole().
c.
Width <= R4 + R5 /* Pulse width */
d.
Register R6 <= R4; Register R7 <= R5; * Base values for the next
iteration.
Routine Check_pothole:
a.
Delta <= number of clock ticks between two adjacent adjacent pulses
b.
If (Delta > Threshold) set warning <= true;
Figure 2: Control Algorithm for obstacle detector
TI India Analog Design Contest 2010
Log on to
http://www.ti.com/in/analog_design_contest_2010.htm
Posters are available! Write to [email protected] for a copy
5
UniTI caught up with the team from IIT Madras to get a
first-hand look at the obstacle detector project.
UniTI: What was the aim you started out with?
Anoosh: Our aim was to build a simple navigational aid for
the visually impaired. The obstacle detector can be
incorporated in their walking canes or footwear.
UniTI: Where else do you think your obstacle detector
will be useful?
Raghunandan: We think it may be useful in unmanned
military vehicles used in warfare on a hostile terrain. The
obstacle detector designed by the authors can play a very
useful role in these vehicles. They can act as the eyes for
the vehicle.
UniTI: Is the obstacle detector ready to be deployed?
Anoosh: For field deployment, many sensory cues must be
provided to the user. The algorithm for obstacle detection
needs to be much more complex. For example, the
detector must distinguish between walls and people. The
pulse width can be updated using the cumulative average
of all past measurements and the current measurement.
Cumulative average is an efficient way of storing the
history information and ensures that a gradient in the
terrain isn't mistaken for a discontinuity. The Doppler
Effect can be used to determine an approaching vehicle.
This will help the user in crossing a road or boarding a
train.
UniTI: How did you test your system?
6
Raghunandan: We are in the process of integrating the
hardware and software. Presently, we have simulated
scenarios by feeding software-generated inputs instead of
sensor-generated inputs.
UniTI: What challenges did you face in the project?
Sundar Aditya: There are many voltage levels to deal with.
The sensor needs a 5V supply and the PWM output is 5V
peak-to-peak. However, the MSP runs on a 3.6V power
supply. Therefore, we need another regulator to step
down the output of the sensor down to a range that is
compatible to the MSP.
The sensor used doesn't give robust readings for oblique
angle of incidence i.e. the sensor is not placed with
vertical incidence. The PWM reading for a change in
distance of one meter doesn't give considerable change in
Ton. The configuration of oblique incidence is necessary to
detect a pothole from a finite distance away, so that there
is sufficient time to react.
UniTI: How much will the obstacle detector cost?
Sundar Aditya: The microcontroller we used costs Rs
2950. The sensor costs Rs 750, and the voltage regular
cots Rs. 40. The electronic components cost under Rs
4000.
UniTI: Thank you and wish you all the best!
Many thanks to Prof. Nagendra Krishnapura for his
guidance!
A Portable mobile phone charger using Pico Wind-Turbine
Today, we depend on electrical power for charging cell
phones and other portable devices. With widespread use
of battery-operated mobile devices, there is an urgent
need to look for alternate ways to charge batteries in order
to conserve electrical energy which today predominantly
comes from thermal power plants. Not only will alternate
sources of energy reduce our dependence on exhaustible
resources such as coal, they will also reduce the carbon
footprint of electronic devices.
Aditya Venkataraman and Sayee Ram V, undergraduate
students of Electrical and Electronics Engineering at
National Institute of Technology, Trichy, decided to take
on the challenge of building a portable mobile phone
charger that does not require an AC power outlet.
Figure 1: Top View of the Stator
“Presently, our idea is to build a charger that is useful in
emergencies,” explained Aditya Venkataraman. “Today,
travelers on long journeys in remote areas face problems
in charging their cell phones or other mobile appliances.
We made use of a pico wind turbine, which is a miniature
of a conventional wind mill, which can provide sufficient
power for charging a cell phone.”
“The exciting challenge of harvesting ambient energy kept
us motivated during the project,” added Sayee Ram. “We
began our quest by constructing a Pico-Turbine based on a
vertical-axis Savonius generator. It is essentially a single
phase alternator designed to run purely on wind energy.
We performed a simple fan test to observe the
performance of the turbine over a range of wind speeds
and found that the idea was feasible.”
“The second stage of the project was to adapt the Picoturbine for charging a rechargeable battery bank of 6.3 V,”
said Aditya. “This required us to employ the properties of
voltage multiplication, rectification and exploit the
inherent open-circuit nature of our Pico-Turbine, at times
of no wind, which prevents battery discharging. The final
phase involved the use of this rechargeable battery bank in
a cell-phone charging circuit, which consists of a voltage
regulator and a current booster.”
Pico Wind Turbine
Figure 2: Top View of the Rotor
In order to test the Pico turbine output for different wind
velocities, the students carried out a simple test using a
table fan to provide the wind energy (see Figure 3). Of
course, in a real situation, the source of energy can be the
natural wind or manual rotation. The fan has a maximum
speed of 1350 RPM and a sweep of 400mm. We found
that when the fan was set in low speed mode, the output
voltage was in the range 0.6 to 1.4V RMS. In the medium
speed range, we obtained an output voltage in the range
1.4V to 2.1V RMS. In the high speed range, the voltage
varied from 1.9V to 2.5V. Since the highest voltage level
from the turbine is of the order of 2V, we used a tripler to
raise the voltage level to about 6V needed for the charger.
A series resistance is used to limit the current; in our
experiments, an ammeter connected in series indicated
the current flowing to the battery.
The Pico wind turbine built by NIT Trichy students consists
of a stator with six copper coils wound and connected in
series with alternate clockwise and anti-clockwise
windings as shown Figure 1. The rotor plane is mounted
on a wooden shaft with a pointed tip that rests on top of a
screw in the base. A metal plate (rotor) is attached to the
wooden shaft; six permanent ceramic magnets are
attached to the metal plate with alternate poles as shown
in Figure 2. The output of the Pico Turbine is the single
phase AC.
Figure 3: Fan Test
7
Input
Rsc
Q1
v1
Q2
R1
uA7805
30
c1
2
3
vo
Output
c2
Figure 4: Mobile charging circuit
Practical challenges
UniTI spoke to the students of NIT Trichy to
understand the challenges faced by them during their
project.
UniTI: Congratulations of completing this project, NIT
team! Do you recall any particularly satisfying
moments from the project?
Aditya: The air gap between the magnets and the coils is
of paramount importance. Initially we had to move the
plate up and down the shaft to get the right air gap. But
this made it difficult to keep the plate at a fixed distance on
the axis. So we came up with the innovative solution of
adjusting the screw in the bottom to adjust the air gap,
rather than manually shift the rotor plane along the axis.
lower than a threshold, we had to be ensure that the
battery does not discharge into the turbine. The inherent
open-circuited nature of the turbine and the blade design
helped solve this problem.
UniTI: So, where do we go from here?
Aditya: While our design gives the proof of concept, many
improvements are necessary to make it practical. The
pico turbine has to be even smaller and the blade design
must be further improved to spin sufficiently fast at low
wind velocities. We believe there are ways to fix the picoturbine in a car or other automobiles to harvest the wind
energy. This would be the direction of future work!
UniTI: Thanks and best wishes to you!
Many thanks to Prof. Arul Daniel for his guidance!
UniTI: What were some design challenges you faced?
Sayee Ram: Harvesting the small amount of energy from
the pico turbine means that we must have minimum
amount of electronics which will consume power for
operation. Therefore, we have kept our design as simple as
possible. We based our design on the use of Transistors
(as amplifiers) and a voltage regulator IC 7805.
“The other challenge we faced was that the battery
management circuitry in the cell phone follows a certain
algorithm for charging to prolong the battery life; without
customization for the specific phone charger, the battery
management unit will prevent charging of the battery,”
said Aditya. “In our experiments, we used the Nokia 1100
cell phone and successfully charged it using the lead acid
battery bank. The pico-turbine will have to be customized
for different types of batteries – for example, a different
design will be needed for Lithium-Ion rechargeable battery
bank.”
Sayee Ram: When the Pico turbine is rotating at a speed
8
Art by Ananya Ravikumar
“If my fans were here, I could have used my
picoturbine to charge my mobile phone …”
Student Attendance Monitoring System
using Voice Analysis & RFID
The strength of an undergraduate class in an Indian university is often more than 50. Most universities require the student
to attend a minimum number of classes (typically 75%) to qualify for appearing in the examination. Calling out the roll
numbers of 50+ students can not only consume a considerable amount of class time, it is also an error-prone procedure
for recording attendance. The students of CMR institute of technology, Syed Azam Naushad, Shuaib Ahmed Salman, and
Shobhit Mathur, have thought about how they can help their teachers in automating the job of marking attendance. They
have designed a Student Attendance Monitoring System that combines Voice feature extraction technology with Radio
Frequency Identification (RFID) technology to give a robust solution to the problem.
The Voice feature extractor samples the voice input of students and generates a unique digitized data corresponding
to each student's unique voice features.
The digitized data is integrated with the RFID system which includes an RFID reader, a compatible RFID tag and GUI
Software running on a computer to communicate between the RFID reader and the RFID tag.
The software interfaces to a database which stores and allows comparison to the data under observation.
When a student enters the class, in an ideal case, the reader located near the door identifies the RFID tag embedded in the
student ID card and updates the attendance corresponding to the unique ID of the RFID tag. This technique is better than
swiping or card punching in terms of convenience and time savings. Such a system will permit proxy attendance, since a
student may carry the ID card of another fellow student! A second level of verification uses voice recognition feature as an
input. The students propose to store the voice information in the RFID tag itself.
Host System
(PC, PLC, etc.)
Running GUI or Process
Fixed RFID Reader
System
Interface
Clock
Processor
+V
HF
Tag
Antenna
SPI or Parallel
RFID
Reader /
Writer
Transceiver
4Ù
Matching
13.56 MHZ
AC Adapter
LED
Driver
System
Power
AC/DC
Adapter
Buzzer
Driver
Mobile RFID Reader
SPI or Parallel
Display
Processor
Keypad
Clock
+V
Antenna
RFID
Reader/
Writer
Transceiver
Matching
13.56 MHZ
AC Adapter
AC/DC
Adapter
(optional)
Battery
Management
System
Power
Battery
Figure 1: Block Diagram of RFID Reader
LEGEND
Processor
Interface
RF/IF
Amplifier
Logic
Power
ADC/DAC
Clocks
Other
9
filter, to give a slowly varying DC voltage corresponding to
the pitch of the person. The low pass filter is implemented
using the OPAmp IC UA741CN.
A sampling time window that corresponds to the
acceptable duration of the spoken word is generated by
means of a zero crossing detector. It begins when the
voice input crosses the zero threshold and remains ON for
the duration determined by the RC time constant
designed as per the time limit for input voice. The window
is generated for duration of 1 second. The analog DC
signal sampled at the Analog to Digital Converter's
(ADC0804) input pin for a maximum of 1 second
duration, gets converted into digital format and is
available on the output pins of the ADC. The conversion of
the analog input and reading the digitized output from the
ADC to the PC is implemented by interfacing the
ADC0804 to a microcontroller (8051).
RFID tag based identification
The RFID (Radio Frequency Identification-13.56MHz)
essentially consists of the following:
An RFID Reader/Writer (transceiver),
An HF tag
A processor with peripheral interfaces
Texas Instruments provides High Frequency (HF) tags
suitable for paper and plastic lamination. These tags
permit memory sizes up to 2Kb. To implement RFID
system a TRF7960 RFID Reader Evaluation Module and
RFID tag inlays of the tag-it family have been used. The
block diagram of the RFID reader is shown in Figure 1.
Voice based identification
To generate a unique voice imprint that will allow
biometric verification, we used the pitch of the voice as an
individual's characteristic. The voice input is given to a
microphone, which is connected to a preamplifier. The
amplified voice signal is filtered to get the required band of
frequencies corresponding to the pitch of the person. The
speech of a typical adult male (female) has a fundamental
frequency from 85-155 Hz, (165-255) Hz. Keeping in view
the margins due to non-idealities of circuits and to
accommodate unexpected variations in voice input, we
assume a pass band range of 70 Hz to 300 Hz . We used a
3rd order Butterworth band pass filter in order to have the
best roll-off and minimize out-of-range frequencies. The
filtered signal is passed through a Schmitt trigger to
generate rectangular pulses of non-uniform width. Since
each voice signal has a unique pitch, the frequency of
change of state in the corresponding pulse waveform will
be unique. The pulse waveform is digitized using a
monostable multivibrator (IC 7412N), which generate
uniform pulses from the non-uniform pulses. These equal
width pulses are averaged in voltage using a low pass
10
After the data is obtained on hyperterminal using serial
communication with the microcontroller 8051, it is
converted into hexadecimal form. This conversion is
required, since the data stored in each RFID tag memory
block is required to be in Hexadecimal form. The Hex data
is called the Voice Imprint of a student and is stored in the
memory block of RFID tag and the student database.
Transferring Digitized voice to RFID Tag
To store the voice imprints in the user programmable
memory of the RFID tags, we made use of the Texas
Instruments TRF7960 evaluation module and its GUI. The
GUI contains the 15693 protocol tag writing options,
which are used to write single/multiple blocks of data into
the tag. The following procedure is used to write/read the
voice sample from the tag.
The card is brought into the range of the reader
In the 15693 tab, the respective UID of the card is
entered
Then the data is entered into the data tab
Number of blocks to be written is selected
Protocol is set, and the process is executed
The coding for the project was carried out in assembly
language, and the KEIL compiler was used.
System use scenario
When the student joins the college, his/her voice imprint
will have to be stored on the RFID tag as part of his
admission process, at the time of issuing the badge to the
student. During a class session, each student is expected
to carry his/her student ID. The classroom has a computer
which stores the database. The teacher carries the RFID
reader module and connects it to the database. The
students get their respective cards in the range of the
reader, which reads the tags and extracts the voice
imprints. This data is then compared with the database
voice samples; if they match the attendance of the
respective student is updated. In order to save time, the
RFID reader can be permanently installed in the doorway
of the class and connected to the computer. The
attendance records can be updated and made available to
students on demand through a college website.
We asked the students of CMRIT what were some of the
challenges they faced during the implementation of their
project. Shobit said, “We learnt many things during the
course of our project. For example, when implementing
the 74121 monostable multivibrator, the output was not
seen due to connection of +Vcc and ground terminals to
the circuit as mentioned in the datasheet. We found out
that the +Vcc and –Vcc pins had to be connected to the IC
74121 instead of ground pin.”
Shuaib added, “Once the program was downloaded into
the Flash RAM of 8051 and we executed the program, the
output on the HyperTerminal was not obtained easily.
Many changes were done in the code again to finally get
the output data closer to what we expected. Now we
could see even the sudden change in output once
someone spoke on the mike. Similarly, after the digitized
voice data was stored in the database and the RFID card,
there was a need to read the data back into the GUI and
compare it with the data stored in the database. Here, we
faced the problem of linking the data in GUI with the
database. After evaluating various options, we decided to
use Eclipse Workspace for this task.”
UniTI asked the students what was the main achievement
in their project. “A proof of concept of the proposed
system has been implemented practically and
successfully,” said Shobit Mathur with a sense of pride.
“Of course, there are limitations in our prototype that we
need to work on. The main limitation of our RFID system
is that at present it does not take real-time voice input for
verification. Although the voice pitch of one person is
different from that of another, it is also true that the pitch
of a person's voice may vary from one time to another.
This limitation can be overcome by designing a more
sophisticated statistical voice recognition algorithm
which learns the voice pattern of a person gradually and
identifies the person in a more robust fashion. There are
some chips available today for speech recognition. The
current solution has been developed for an HF Range
antenna which has a read distance range of about 3 to 5
inches. For actual implementation, the read range needs
to be a few meters. This can be achieved by using UHF
Range antennas which have an interrogation range of
several meters.”
UniTI wishes the students all success in their
endeavors!
Many thanks to Prof. Sharmila for her guidance!
Art by Ananya Ravikumar
“Sunil,your voice does not tally with the sample
on the database. Please use the same voice you
had used at the time of enrollment.”
Shuaib, Azam, Mrs. Sharmila (Mentor) and Shobit with
the project
11
Signal Conditioning Unit for Field Deployable Microimpedance
Biosensors
Early detection of cancer can save human lives.
Undergraduate students of Electronics and
Telecommunication Engineering, Abhra Bagchi and
Gaurav Konar from Bengal Engineering and Science
University, selected this important field for their project
submission to the Texas Instruments India Analog Design
Contest (2009).
Micro impedance biosensors for low-level pathogen
detection and target protein markers for cancer detection
are of growing importance in health industry for early
diagnosis and treatment. Microimpedance biosensors
offer the advantage of rapid detection without requiring
any labeling. The sensing mechanisms of impedimetric
biosensors normally involve measurement of impedance
and phase. However, a majority of measurements at
present are carried out with the help of desktop LCR
meters. The commercially available portable LCR meters
suffer from limitations for this application:
The maximum frequency is usually limited to 10 kHz
or 100 kHz
There is no direct display of the phase value.
To overcome these problems, we have developed an
improved phase detection technique using readily
available discrete components which can operate in the
frequency range from 200Hz to 300 kHz and measure a
low phase shift.
Impedance change in a micro-impedance biosensor is
primarily caused due to the modulation of the electrical
double layer impedance Rd and/or the solution resistance
Rs. In the low frequency region, Rd dominates; in mid
frequency range, the contributions of Rd and Rs are
similar; in high frequency range, Rs dominates. Rd is a
function of the applied voltage and it is desirable to keep
the applied voltage within 500mV to maintain a stable
double layer. The sensitive frequency of micoimpedance
biosensors varies depending on the electrode design and
is usually in the range 500Hz to 500 kHz. For bacteria
detection, the preferable is around 1 kHz. For online
monitoring of cell growth, the preferred frequency range is
above 10 kHz. For cell culture detection, the background
medium is made conductive to keep the cells alive. Thus
the output phase of such a sensor is usually low. In view of
these constraints, the circuit level specifications are as
follows:
12
Frequency of applied signal: 100Hz to 100 KHz
Amplitude of the applied signal - within 500mV
Impedance detection range -100? to 100k? .
Phase detection from 0.1° to 90° over the entire
frequency range
Power supply: 5V or less
The circuit level specifications are realized by a combined
impedance phase detector circuit which will require a
sinusoidal excitation. The principle of operation of the
circuit is shown in Figure 1. The sensor circuit is
highlighted in Figure 2.
Sine Wave
Generator
Amplifier1
Comparator1
LPF
XOR Gate
SENSOR
Amplifier2
Microcontroller
MSP430FG4618
Comparator2
Digital Storage
Oscilloscope
Digital
Voltmeter
Test & Measurement
Figure 1: Block Level Representation of the Signal
conditioning Unit
To Amplifier1
R1
To Amplifier2
R
C
Sine Wave
Frequency: 1KHz
Amplitude: 500mV
Figure 2: Sensor circuit scheme
Sensor
The amplitude of the signal at the output of Amplifier 2 is
a measure of the impedance. Both the DC signal and the
output of Amplifier2 are interfaced to the MSP430
microcontroller. The on-chip 12 bit ADC in the
microcontroller is used to read both the values with high
accuracy and sensitivity which are essential for correct
computation. These values are then used to compute the
impedance and phase difference using a look-up table
method.
The following are the formulae used for the computations:
Impedance of sensor: R/√(1+ù2R2C2)
Phase of sensor: -tan-1(ùRC)
Selection of Components
Figure 3: Complete System which shows the MSP430
board interfaced to the PCB that was built for
implementing the signal conditioning circuit
If the applied signal is A sin(wt), then the output of sensor
is G.A sin(wt+è) where G is the gain of the sensor (G<=1)
and è is the phase difference between the input signal and
sensor output. Input signal is also directly applied to
amplifier 1 and output of sensor is applied to amplifier 2
These two signals are amplified and passed through
comparators which result in two square waves with a
phase shift between them. The EXOR of the square waves
produces a pulse train with a frequency that is double that
of the input frequency and ON time equal to the time delay
between the two signals. The output of the EXOR gate,
when passed through an active 2nd order low-pass RC
filter, gives a DC value which is proportional to the ON
time of the pulse train and hence to the phase difference.
Assuming that the è is expressed in degrees, the DC value
is given by
DC Value =
è
180
Comparators: A phase shift of around 0.1O in 100kHz
frequency is equivalent to a time difference of around 3ns.
In order to measure such a small time difference, we
require a high speed comparator with rise time and fall
time preferably within 1.5 times the requirement. It would
be desirable that the comparator has a low offset voltage
for maintaining stability in the rising and falling edge of the
output pulse. The comparator should be able to drive a
CMOS XOR gate. We selected Texas Instruments
comparator TLV3501AIDG4 based on the above
considerations. This IC has a speed of 4.5ns, rise time and
fall time of 1.5ns, an offset voltage of 6.5mV and operates
with a supply voltage of 3.6V.
Amplifiers: The amplifiers are primarily required for
increasing the voltage level to reduce the noise in the
output of the comparator. As the output of the sensor is
less than 500mV, the offset voltage and the noise voltage
of the amplifier should be low. As the phase shift at the
output of the sensor is low, it is required that the input
capacitance of the amplifier is significantly lower than the
sensor capacitance to maintain accuracy. These
requirements justify the choice of OPA4376 amplifier
from Texas Instruments, which has a low offset voltage of
5µV, a low noise of 7.5nV/√Hz in 1 kHz frequency, and
an input capacitance of 6.5pF.
Vdd
13
XOR gate: The primary concern in the selection of the
XOR gate is the propagation delay which should be of the
order of 5ns. Also the input capacitance should be low so
that no additional phase shift occurs at the output of the
comparator. The maximum load capacitance tolerance
needs to be high, so that the output does not get distorted
after interfacing with the low pass filter amplifier. To meet
these requirements, SN74LVC1G386DBVR XOR gate has
been used.
Sinusoidal waveform generator: The sinusoidal
waveform must be generated over a frequency range of
100 Hz to 100 kHz with sufficient stability and should be
operated with a single supply voltage. The frequency drift
with supply voltage should be low. The output impedance
should be low in order to drive the sensor circuit. Based on
these considerations, ICL8038 was chosen; the frequency
sweep of this generator can be adjusted from 100Hz to
100 kHz.
Amplifier 3 for low pass filter: Amplifier 3 is used in
building a low pass filter. The primary requirement is that a
minimum DC value of 2mV will have to be filtered. Thus
the noise and offset voltage should be low. Also, the input
capacitance should be less than 15pF to keep the output
of the XOR gate undistorted. The OPAMP should be
operated from a single supply voltage of 3.6V. The
component OPA4376 from Texas Instruments satisfies all
these requirements.
Power management chip: The power management chip,
TPS63002 is required to down convert the voltage from
5V to 3.6V. ICL8038 requires 5V supply and the other
components require a 3.6V supply.
Microcontroller: A microcontroller with a built-in 12-bit
ADC that gives a resolution of 0.8mV in 3.6V is ideal for
our application. Such a resolution helps in measuring a
small change in impedance and phase. The
microcontroller is useful in realizing computational
functions like square, square root, and tangent, as well as
providing an interface to an LCD display. For the purpose
of our instrumentation, we decided to use the Texas
Instruments MSP-EXP430FG4618 experimenter board,
since it is easy to operate and meets the above
requirements. The MSP430FETPIF programmer was used
to program the microcontroller.
14
Practical Challenges in Implementation
Even after performing the simulation and testing the
components individually, deviations from the expected
characteristics were observed after making the PCB.
Despite proper biasing, the amplifiers showed
distortion at the DC bias point crossing, which is
similar to the zero point crossing distortion. This is
more prominent for input amplitudes less than 1V
and single supply voltage. This may be attributed to
some mismatch in biasing circuit with single supply
voltage.
The outputs of the amplifiers are expected to be in
the same phase when no external capacitance is
present in the sensor circuit. But this does not
happen in practice. There are mismatches in the
amplifiers introduced from the fabrication process
tolerances. Another significant contribution for the
mismatch comes from the output node capacitances
introduced from the PCB board layout. The signal
lines at the outputs of the two amplifiers may be
coupled with different parasitic capacitances and
this creates an additional phase shift.
The output of the comparators also deviated from
the expected behavior. There is a baseline phase
shift in the absence of any external capacitance in
the sensor circuit; this is a reflection of the distortion
at the output of the amplifiers and the process
mismatch in the comparators. The low-to-high and
high-to-low transition times have increased to 100ns
probably due to the parasitic capacitances
introduced from the interconnect lines in the PCB.
This limits the detection of very low phase values at
high frequency. The output of the comparator is also
distorted at the DC bias point crossing, which is
probably due to similar distortion at the output of the
amplifier.
Achievements
UniTI caught up with Abhra Bagchi and Gaurav Konar to
understand their main achievements.
“We successfully built the analog front end for
microimpedance detection which works at the entire
frequency range of interest,” said Abhra Bagchi. “We can
operate our circuit from 100 Hz to 100 kHz.”
UniTI: Congratulations to you and best wishes!
Many thanks to Prof. Chirashree Roy Chaudhuri for
her guidance!
MSP430 based Electronic Load Control for
Micro Hydro Power Plants
The increasing rate at which conventional sources of
energy are getting depleted has laid an emphasis on
renewable energy sources such as wind energy and
pico/mini/micro-hydro power plants. Undergraduate
students of Electrical Engineering from IIT Delhi, Ankit
Agarwal, Anubhav Gupta and Shailesh Kumar Dubey,
decided to work on a project that would help towards this
important cause.
Traditionally, synchronous generators have been used for
power generation, but the use of induction generators has
been on the rise. For renewable energy applications of low
and medium power (up to 100 kW), the induction generator
offers considerable advantages:
manufacturing, maintenance and operational simplicity
brushless and rugged construction
lower unit cost
good dynamic response
self-protection against faults
ability to generate power at varying speed and troublefree operation for many years
The self-excited induction generator (SEIG) does not need
an external power supply or other related equipment such
as field breaker, rheostat, and automatic voltage regulator
to produce the excitation magnetic field. Thus SEIG is a
good candidate for wind, biogas, and hydro powered
electricity, especially in remote and isolated areas. These
advantages facilitate induction generator operation in
stand-alone/isolated mode for supplying power for local
load and grid.
Principle of Operation
A three-phase squirrel-cage induction motor is connected
in delta configuration and excitation capacitors are
connected in C-2C configuration. The excitation
capacitance must be sufficient to provide the required
voltage for the load on the SEIG at the operating speed.
The value of capacitor excitation for no load and rated
load conditions are determined iteratively. The equivalent
circuit parameters are obtained from an open-circuit and
blocked-rotor test. The performance of the SEIG depends
on its magnetizing characteristics. The synchronousspeed test is conducted to obtain the magnetizing
characteristics. The SEIG feeds two loads in parallel, such
that the total power is constant:
Pout = Pc + Pd
where Pout is the generated power, Pc is the consumer
load power, and Pd is the dump-load power.
The dump-load power Pd may be used for non-priority
loads such as heating, battery charging, cooking, etc. The
amount of dump-load power is controlled by the IGBT
(Insulated Gate Bipolar Transistor) chopper. The duty
cycle of the gate pulse of the IGBT gives the average
conduction period of the chopper and, hence, the amount
of power in the dump loads. A variable mark-space ratio
chopping approach has been adopted for the IGBT
chopper because it produces a variable unity powerfactor load with just a single ballast or dump load. The
output power of the SEIG is kept constant by the ELC.
The limitation of SEIG is its poor voltage regulation, since
the generated voltage depends on the speed, capacitance,
load current, and power factor of the load. While the input
power remains constant in an unregulated micro-hydro
turbine, the output power varies due to changing consumer
loads. To maintain SEIG output power at a constant level, a
dump load is connected in parallel with the consumer load
(Figure 1) and the power supplied to the dump load is
controlled using an Electronic Load Controller (ELC).
Main
Load
SEIG
Prime
Mover
Excitation
Electronic
Load
Controller
Dump
Load
(consists of all the
power electronic circuitry)
Figure 1: SEIG Block Diagram
15
Voltage Sensing: An Innovative approach
Ankit Agarwal explained an innovation used by the team
for reducing the cost. “Traditionally in such applications
Hall-effect voltage sensors (e.g. Lem LV100-250) have
been used. Alhough such sensors provide excellent
accuracy and dynamic response, their cost is high – a
sensor may cost around $75. Further, such sensors
measure instantaneous voltage, whereas what the
controller requires is the root mean square (RMS) value. A
somewhat expensive DSP solution is often used to do a
real-time computation of RMS value.
“We thought of a completely new approach to sense the
voltage. We made use an off-the-shelf multimeter based
on ICL7106/7109, which costs around only $2.5. This is
a win-win solution, since the IC provides the RMS value
and eliminates the need for RMS computation. A low-cost
16-bit microcontroller, such as the Texas Instruments
MSP430, which is much cheaper, is more than sufficient
for our application. We made use of the MSP430F1611
Microcontroller IC. The MSP430 has many other benefits,
such as the on-chip ADC which permitted us to eliminate a
board-level ADC.”
“There is one down-side in using the ICL 7106/109 ICs,”
added Shailesh “They do not provide isolation. We had to
make use of optocouplers (IC 6N136) to take care of that.”
Use of Analog Multiplexers:
“ICL7106 gives the output in a form that is ready to drive
an LCD,” said Anubhav. “We had to decode the voltage
reading from ICL7106 by decoding the 7-segment display
code that it generates. In order to keep the number of
inputs to the microcontroller minimum, we used an 8:1
analog multiplexer (MPC508) for each digit in the voltage
reading. The microcontroller sends the control signal (a 3
digit BCD) to each of the multiplexers in order to get the
required segment outputs. Therefore the number of
connections to/from microcontroller was reduced from
21 to 6 for the purpose of sensing.”
The block diagram of the load controller is shown in Figure
2. The Voltage Sensor used here gives an output in the
form of square waves between 4V and 9V. This output is
sent to an analog multiplexer which is controlled through
a microcontroller. The output is XORed with another
square wave and we get a digital output of 5 volts. But
since our microcontroller works at 3.3V, the XOR gate
output is given to the voltage regulator and then sent to
16
microcontroller. The sensed voltage is compared with a
reference voltage value, which is hardcoded in the
microcontroller; this value is proportional to the rated
terminal voltage of the SEIG and may be altered when
required. Depending on the error value, the pulse width of
the PWM waveform is increased or decreased. The error is
compared with a very high frequency triangular wave and
the comparator output is read by the PI Controller section
of the code in microcontroller. The PI controller is turned
on only after an initial voltage of around 230V has been
achieved after the start, so that the initial errors when the
voltage is rising do not interfere with the integral part of
the controller. The output of the PI controller section is the
required PWM signal for switching the IGBT to regulate
the dump load. We used the IGBT from Toshiba,
GT60M303. The PWM output is then given to the IGBT
chopper through opto-isolation and pulse driver circuit.
OPAMP is used for amplification as the PWM output is at
3.3V and the IGBT switch functions at around 15V.
Control Signals
Voltage
Sensing
Unit
Analog MUX
IGBT Switch
XOR Gates
Operational
Amplifier
Voltage
Regulator
Microcontroller
Figure 2: Block Diagram of Electronic load controller
Accomplishments
UniTI spoke to the students from IIT Delhi to understand
their main accomplishments.
“We initially did a simulation of the load controller in
Simulink,” said Anubhav. “Then we built sub-systems of
the controller using breadboarding. A PCB design for the
controller is now complete and we are awaiting its
fabrication.”
“We expect that our load controller will result in
substantial cost reduction, since we use a low-cost
voltage sensor and a low-cost microcontroller in place of
more expensive voltage sensors and DSP. Such load
controllers can be deployed in pico/micro hydropower
plants, where cost minimization is an important
consideration,” said Shailesh. “Such power plants make a
lot of sense in rural India.”
UniTI: Congratulations and wishes you a good luck!
Many thanks to Prof. S.S. Murthy for his guidance!
Maximum Power Point Tracking of Photovoltaic Arrays
using Adaptive Perceptive Particle Swarm Optimization Technique
In the Jan 2010 issue of UniTI newsletter, we covered
the concept of Maximum Power Point Tracking (MPPT)
in Photovoltaic Arrays. Since the conversion efficiency
of PV arrays is low, we must extract the maximum
power from them through the use of MPPT. The
undergraduate students of Electrical Engineering from
Jadavpur University, Uddipta Maity, Bishnu Dubey, &
Aritra Chowdhury, decided to use an optimization
technique called Swarm Optimization to get the most
power out of PV arrays.
The Adaptive Perceptive Particle Swarm Optimization
(APPSO) algorithm used in their work is a variation of the
conventional Particle Swarm Optimization (PSO) which is
described in the literature. PSO is attributed to Kennedy
and Eberhart who published their work in IEEE
International Conference on Neural Networks, 1995. They
later wrote a book on Swarm Intelligence (Morgan
Kauffman Publishers, 2001). PSO is inspired by the way a
flock of birds fly or the way a school of fish swim.
Compared to PSO, the APPSO offers flexibility in the
motion dynamics of the particle in the search space
through variation in perception radius, the number of
sampling directions, and the number of sampling points
per direction. Compared to PSO, which takes around 3 to
5 seconds to track to the maximum power point under
partial shading conditions and reaches the optimum with
96.41% accuracy, the APPSO can track to the MPP with
98.21% accuracy in 2 to 4 seconds. APPSO-based MPPT
requires a single pair of sensors to control multiple PV
arrays, lowering the system cost.
A major challenge in using a PV source is to tackle its
nonlinear output characteristics that get more
complicated when the entire array does not receive
uniform insolation. Nonuniform insolation conditions
prevail when clouds pass by or when neighboring
buildings cast their shadow on the PV array, resulting in
multiple peaks in the P-V characteristics. Many MPPT are
unable to effectively discriminate between local and
global maxima in the P-V characteristics. This project uses
Adaptive Perceptive Particle Swarm Optimization in
tracking the global Maximum Power Point of a solar PV
array more closely than possible using PSO. A two-stage
power electronic system architecture based on APPSO
was developed.
Ig
PV
Array
Grid
DC-DC
converter
IL
ADC
Grid
Connected
loads
MPPT
controller
Figure 1: System configuration for PV-based system
In conventional PSO, for an n-dimensional optimization problem, an n-dimensional search space is considered. However,
in Perceptive PSO and the proposed APPSO algorithms, an (n+1) dimensional search space is considered. The added
dimension represents the underlying performance of particles at their positions in n-dimensional space. Figure 2
describes the APPSO.
Algorithm APPSO
Input: Randomly initialized positions and velocities of particles in an n+1
dimensional search space;
Initialize the following variables for each particle p:
perception radius,
the maximum and minimum number of observing directions,
maximum and minimum number of sample points along a direction
maximum velocity;
Output: Particle positions that provide near-optimal solution;
begin
for each particle p do
17
set personal best position(p) = initial position(p);
while not terminating condition do begin
for p := 1 to number of particles
Select random positions for the neighboring particles of p;
Compute the local best position for particle p;
Compute the best velocity for particle p;
Evaluate the fitness function;
if the present solution is better than personal best position then
Update personal best position of the particle p;
Minimize spacing between p and neighbors along any direction within
limits;
Increase the number of sampling directions within limits;
else if present solution is worse than personal best position then
Increase spacing between p and neighbors along any direction within
limits;
Reduce the number of sampling directions within limits;
else
Retain spacing between p and neighbors and the number of sampling
directions without change;
end for;
end while;
end;
Figure 2: Adaptive Perception Particle Swarm Optimization Algorithm
Texas Instruments MSP430FG4618 is used to implement
MPPT algorithm. A square wave of varying duty cycle is
generated using the microcontroller. This square wave
drives the dc-dc buck converter. At regular intervals, MPPT
algorithms is executed and the duty cycle of the output is
varied. In our case, the required interval was 10 seconds.
To generate this interval, we used the interrupt facility of
Basic Timer 1 After specified intervals, the CPU is
interrupted by the timer and the MPPT algorithm is
invoked as the interrupt service routine. If a change in
duty cycle is necessary as per the MPPT algorithm, a new
duty cycle is established and the program to drive the
PWM generator continues with the modified duty cycle.
All the voltages and currents which are sensed from the
7-V to 40-V
Unregulated
Dc Input
+VIN
1
Feedback
4
TL2676-06
3 OND
+
CIN
100 µF
PV array are analog variables and must be digitized
before they are processed by the microcontroller. The
ADC ADS1204-1 is used for this purpose. The DC/DC
converter is used to feed the power generated by the PV
array to the grid. We used the TLV2575HV-05 buck
converter from Texas Instruments. Further the output
voltage of the microcontroller which acts as a PWM and
changes the average output voltage of the DC/DC
converter is connected to pin 5 of the buck converter.
The output of the microcontroller cannot be directly
connected to pin 5 because the microcontroller output
pin has a maximum value of 0.6-0.7V for Logic 1, and the
dc-dc converter turns off only at 2.4V. Thus, we need to
amplify the microprocessor output by a factor of 4-5V.
This can be done with a low noise amplifier.
Output
2
L1
L2
330 µF
20 µH
5 ON/OFF
01
1N6818
+
COUT
330 µF
C1
100 µF
Figure 3: Circuit Implementation of DC-DC converter
18
5-V
Regulated
Output
1-A Load
+
We simulated a scenario where there are two PV modules, one of which is fully illuminated and the other is partially
illuminated. Hence, in the context of the present problem n=2. The particle velocities are initialized randomly as shown in
Table 1, where Vop refers to the open circuit voltage of
either array. We tried several variations of the APPSO
Table 2: Different types of Adaptive Perceptive
algorithm, as illustrated in Table 3. The simulation results
Particle Swarm Optimization Algorithm
are summarized in Table 3. An analysis of Table 3 reveals
that APPSO3 yields the Maximum Power Point that is
Algorithm Perception No. of
No.of
closest to the global optimal. While PSO yields 96.41%
accuracy in reaching the global MPPT, the APPSO yields
Radius
directions sample points
98.21% accuracy.
APPSO1
Fixed
Fixed
Fixed
Table 1: Initialization of Particle Velocities
Agent
V1(v)
V2(v)
1
0.2Vop
0.3Vop
2
0.5Vop
0.4Vop
3
0.6Vop
0.1Vop
4
0.8Vop
0.7Vop
5
0.9Vop
0.4Vop
APPSO2
Fixed
Fixed
Variable
APPSO3
Fixed
Variable
Fixed
APPSO4
Fixed
Variable
Variable
APPSO5
Variable
Fixed
Fixed
APPSO6
Variable
Fixed
Variable
APPSO7
Variable
Variable
Fixed
APPSO8
Variable
Variable
Variable
Table 3: Results of Maximum Power Point Tracking obtained using different Optimization Algorithms
Algorithm
PV1 Voltage
PV2 Voltage
Power
Optimal MPPT
Optimal MPPT
Optimal MPPT
APPSO1
45.8V
46.7V
381W
APPSO2
45.9V
46.8V
378W
APPSO3
45.4V
46.2V
46.1V
384W
APPSO4
45V
45.5V
APPSO5
45.7V
APPSO6
45V
391W
383W
379W
45.8V
46.3V
46.8V
APPSO7
45.7V
46.3V
379W
APPSO8
45.9V
46.7V
380W
PSO
46V
47V
377W
Conclusions
UniTI spoke to the students of Jadavpur University to
further understand their project.
UniTI: How would you summarize the achievements of
your project?
378W
Aritra: We were interested in building a PV based system
for energy conversion. At the same time, we were
interested in optimizing the parameters of the PV arrays
so as to operate the combination at maximum power
point.
Bishnu: For achieving the maximum power point, we
implemented both the Particle Swarm Optimization
algorithm as well as a modified Adaptive Perceptive PSO
algorithm that we have proposed. In the two-module test
case that we tested, the APPSO algorithm performed
better than PSO.
Uddipta: Since a single of voltage and current sensors are
used in our system, the cost of data acquisition is
reduced. The cost of our system is under $200.
Art by Ananya Ravikumar
“Bee Number 42, increase your velocity immediately –
I repeat, increase your velocity immediately!”
UniTI: Congratulations & wish you all the best!
We thank Prof. Shubhajit Roy Chowdhury and
Prof. H. Saha for their mentoring!
19
Remote Home Lighting and Appliance Control System
Do you switch off your office lights and fans before you
leave? What about the lights and fans in a lab or a row of
cubicles? Imagine having a remote control that can help
you turn off the electrical appliances before you leave
your work place or home. Two undergraduate students of
Electrical Engineering at IIT Kharagpur, Aamod Shanker
and Praveen Kumar Sharma, decided to work on this
project for their entry into the Analog Design Contest
(2009).
1
A wireless remote control unit based on TI eZ430
RF2480 Zigbee transmitter. The remote control has
push buttons for specifying the control action. The
remote controller uses the Zigbee low power RF
protocol to transmit data, and as a result the life of
the batteries in the remote controller can be
lengthened.
In the remote controller, four micro switches are provided,
corresponding to four profiles (appliances that need to be
controlled); The GPIO pins (P2.3, P2.4, P4.4, P4.5) of the
MSP430F2274 in the remote controller were connected
to micro-switches. The switches send interrupts to the
eZ430 transmitter, which in turn sends the encoded value
of the profile to the receiver. The custom switch boards
are provided with up-down switches corresponding to
each profile, and a switch board can be made a part of the
profile by activating the up-down switch. Hence a switch
(and thus the associated appliance) can be made a part of
multiple profiles at the same time. Pressing the button
corresponding to a particular profile on the user console
turns on all the switches in that profile and switches off
the others. Thus, there are not too many controls at a
user’s disposal, but the user can still exercise a sufficient
degree of control.
2
A control station contains a TI eZ430RF2480 Zigbee
receiver, which receives the user commands from
the remote control unit. The control station uses an
MSP430F2274 microcontroller and sends serial
data over power lines to custom designed switches
for the purpose of power management. The power
line communication modulator was designed to
communicate over pre-existing wiring using On Off
Keying
The data about the requested profile stays stored in the
message buffer of the MSP430F2274 on the Zigbee
module at the control station. The serial communication
using UART standard protocol is required between
MSP430F2274 on the Zigbee module and
MSP430F2618 on the switch board to turn appliances
on/off. This serial communication takes place over the
power lines with the help of the power line
communication modulator.
3
Custom designed switch, which is intended to
replace an ordinary switch used in controlling
appliances. The custom switch comprises of a power
line communication demodulator and a TI
MSP430F2618 developer board. The latter accepts
the demodulated serial input and drives a relay to
switch to turn the supply on or off. There is also a
provision for manually turn the power on/off; this is
achieved using a push button switch which can
interrupt the microcontroller.
There are three components in their project
Figure 1: (a) Remote Controller (b) Control Station
(c) Custom-designed Switch
20
Live Wire
MSP 430
MCU
Zigbee
transmitter
Hand Held Module
Zigbee
receiver
MSP 430
MCU
Receiver/Control Station
Frequency
Controlled
Switch
Frequency
Controlled
Switch
Switches
Figure 2: System block diagram
The modulator has a 120 KHz clock generator which is
much greater than the baud rate of the serial
communication (with the baud rate set at 300bps). The
serial output of the 2274 is ANDed with the clock to get
the OOK modulated signal. The demodulator circuit
consists of two filter stages – an RLC series low-pass filter
followed by a selective amplifier that amplifies only the
higher frequency components. Hence the net effect is an
amplified signal a(t) in the band of interest, and this signal
is the modulated waveform transmitted at the control
station. The original signal O(t) is recovered by passing a(t)
through a LM358 comparator. O(t) is sent to the serial
input of the MSP430F2618. Serial communication is
hence implemented over the power line. See Figure 3.
+5V
2k2
RELAY
To MSP 430
IN4007
+
1mH
1k
910k
-
358
MAINS
From Interfacing
Circuit
thershold
2u2
2
o/p from
MSP 430
BC368
8K Ohm
2k2
Figure 3: Demodulator
Figure 5: Relay Driver Circuit
Discharge Resistor
Cdtx
Cp1-2
Ls
Cs
L
Tx
Rx
Cdrx
220V Mains
Lp
N
6.8V
6.8V
MOV 275V
Figure 4: Isolation Circuit
An isolation circuit is needed to separate the modulator
and demodulator circuits from the AC supply. Our isolation
circuit consists of a series capacitor with a high voltage
rating, which drops the AC voltage across itself. There is a
varistor to protect against voltage surges, and a regulatory
Zener at the input of the DC circuit. The isolation circuit is
designed to pass only frequencies around 120kHz, and
hence blocks out the 50Hz AC coupling noise as well. See
Figure 4.
Powerline communication occurs between the Live and
Earth wires because the neutral wire does not reach those
switches connected to appliances. The microcontroller at
the switchboard (MSP430F2618) needs to turn on/off
the power between the live and neutral. This was achieved
by using a relay with the control input from the
microcontroller (Figure 5). Two of the three output
terminals were connected to live and neutral, and the
control pins received inputs from the microcontroller via
the driver. The driver was designed for a coil current of
40mA for the relay with <1ma input current.
Conclusions
UniTI caught up with Aamod and Praveen to
understand more about their project.
UniTI: What is the main achievement in your project?
Aamod: In this project, we managed to create a robust
remote switching system based on the low-power
wireless networking protocol - Zigbee. A very minimal
circuit is used to achieve powerline communication. We
used only simple ICs such as comparators and AND gates,
which leads to a very cost effective solution.
Praveen: Centring the design around the Ultra Low Power
MSP430 processor has its own big advantages. Since all
the modules will be powered on at all times, it is important
that the electronics itself does not consume too much
power. And today, no processor provides a better feature
set than the MSP430 series for the power it consumes.
Aamod: Unlike the contemporary remote control systems
which use simple RF transmission, we have gone for the
Zigbee to achieve better power efficiency and range.
UniTI – Congratulations on completing the project and
best wishes to you!
We thank Prof. Goutam Saha for his guidance to the
students!
21
Feedback Amplifier!
Feedback on Analog Design Contest 2009
The Contest gave real insights to our students on what is involved in a hardware project. They understood how to select
components by looking at various parameters and circuit characteristics such as input impedence, output impedence and
gain. Since most of the students who worked on the projects were from second or third year undergraduate program,
they got good exposure into areas such as microcontrollers and analog circuits. The PCB design part was a problem that
the students struggled with.
I feel that it will help if TI can specify some application areas in which students can submit projects. It will also be good to
encourage cross-departmental participation by giving them higher chances of selection.
Dr.Chirasree RoyChaudhuri
Mentor, Bengal Engineering and Science University Shibpur; Kolkata
The contest was helpful in giving our students exposure in building real time systems. It gave them a lot of practical
knowledge. In our college, the students were all from third year or final year of engineering; the students worked under a
time crunch. In future, it would be desirable to provide them more time.
I feel that there is no specific need to place any rules about cross-departmental participation in projects. Let the students
decide if the project demands cross-departmental participation.
Prof. Indumathi
Mentor, CMR Institute of Technology; Bangalore
Through this project, the students got exposed to many things, such as which are the companies that make analog
circuits, how they should understand a wide range of issues before embarking on the project, and how you need to
innovate in order to make a difference. I believe the contest is most relevant to final year undergraduate students. Since
the students of Power Electronics in IIT Delhi have a natural orientation for hardware projects, it worked out well. Students
did not have sound knowledge in application domain, which was a problem – they needed time to understand the domain
first. I feel the students need more time to complete the projects. I also feel that the announcement must stress on the
need for innovative project ideas. I think it would be desirable to give priority to cross departmental participation.
Dr. G. Bhuvaneswari
Mentor, Indian Institute of Technology Delhi; New Delhi
It is good that TI provided the students an opportunity to make hardware projects – today, many students prefer to do
software projects. This contest helped attract students to hardware design and I believe our students learned a lot in the
process. They also learnt a lot in the application domain they selected. For example, the students didn't know anything
about Maximum Power point tracking when they started, but now they have enough knowledge. Since students have the
courses in the regular semester, they are unable to devote as much time as they would like to – we should provide some
more time to them.
TI should specify some broad areas in which students can work; this way, the students will come up with some
marketable products. It would be good to promote cross departmental participation in the projects.
Prof. H. Saha
Mentor, Jadavpur University; Kolkata
22
Our students got the opportunity to work on practical problems and use analog ICs. The design experience and hands-on
experimentation were great opportunities for them. The contest definitely helped in building interest in hardware design
among students, particularly in analog domain.
TI has done a very good job in organizing the event, they made the students work by instructing them through regular
mails and monitoring their work and cooperate them in all ways.
It will be nice if TI promotes cross departmental participation but for that they should provide the idea or field in which
students can work and make their projects.
Dr. S. Arul Daniel
Mentor, NIT, Trichy ; Tamil Nadu
The contest gave students an orientation towards specific application; they got interested in hardware design, particularly
in the analog domain. They also got a feel for how to benchmark your work against the best. The contest helped ignite an
innovative spirit in them.
I feel M.Tech and M.S. (Research) students should be given chance to prove themselves. I feel cross-departmental
participation should be encouraged, since it may allow the sharing of knowledge.
Prof. Goutam Saha
Indian Institute of Technology, Kharagpur; West Bengal
Suggestions for ADC 2010 Contestants
Please get an early start and use the summer holidays to work on your project. Although the proposals are due in
2
July, you need not wait until the reopening of the semester to get started! Your faculty advisors will also have more
time on their hands during the summer. Even if your project does not get shortlisted, you will still find the exercise
rewarding.
Plan your project well – there is a design phase, an implementation phase, a testing phase. Each of these may take
2
several weeks. Do not take an existing design and try to build it using TI components – you will find it more
rewarding if you select a problem that is new and apply your imagination in solving the problem. Discussions
among your team and with faculty mentors will help during this phase.
Selection of the components requires care. There is a vast selection of ICs available and you will have to spend
2
some time looking at alternatives.
Writing the report will itself take a couple of weeks. You must show the draft of your report at least a week in
2
advance to your mentors so that they get a chance to suggest improvements.
PCB design for the final project will require dedicated work.
2
When you carry out the design and implementation, investigate several possible approaches to the problem and
2
consider their relative merits and demerits. Be innovative in your design and see how you can cut down the total
cost without sacrificing the performance.
23
TI Tech Days on Embedded Processing at Bangalore
On the 8th and 9th of December, 2009, the Grand Ashok
Hotel in Bangalore played host to TI Tech Days on
Embedded Processing, a two-day conference during
which TI India, along with its partners, showcased some of
the best technologies that they have to offer. The
conference included talks, seminars, workshops and
exhibitions on the technology that is currently driving the
industry towards new dimensions.
The session on the 8th of Dec. started with a welcome
address by Praveen Ganapathy, Director, Business
Development, TI India. He spoke about how TI Developers'
Conference has been revamped and has been replaced
with TI Tech Days in order to provide more convenient
training and networking opportunities for TI customers
closer to their location. TI held over 30 Technology Days
worldwide in 2009. TI Tech Days are usually 1-2 day
training/networking events that bring integrated training
on embedded processors, analog and a variety of
applications, tailored by location and featuring
opportunities for discussion and interaction.
Keynote Demonstrations
TI Tech Days on Embedded Processing featured “keynote
demonstrations” by several TI customers – Windows
Embedded, Ittiam, Tandberg, Displaytronics, Mindtree,
TES and QNX. They displayed new products that they have
built using TI semiconductor. The products ranged from
Video Surveillance Solutions to Handheld ebook readers.
Prem Anand from Ittiam Systems (Bangalore) showcased
a 720P H.264 main profile video decoder solution based
on Texas Instruments' OMAP 3x platform. Their solution
uses the 'split architecture,' and balances the
computational load on ARM Cortex-A8 processor and the
DSP 6x processor which are present on the OMAP 3x.
Pradeep Bordia from TANDBERG demonstrated their
PrecisionHD™ USB Camera. As the name indicates, this
USB camera delivers HD quality video (720p at 30
frames/s). The camera is a low-cost solution for H.264
video encoding, as compared to dedicated solutions.
Vishal Borker from NextBit Computing displayed a
versatile 'universal' set-top box product that was designed
and developed in their Chennai office. The set-top box is
compatible with Internet Protocol and allows a mixing of
several applications such as on-demand video, Internet
radio, and Internet browsers.
Sharmila and Vijayendra of Mindtree demonstrated the
telemedicine solution from Mindtree, which is based on TI
DaVinci platform. Their solution supports several sensors,
USB and Internet connectivity, and audio/video
conferencing. The DaVinci SoC supports two processor
cores – an ARM Core which is used for running the Linux
operating system software and a DSP core which is used
for signal processing.
24
Product Presentations
The presentations at TI Tech Days were held in three
parallel tracks. The intention behind these presentations is
to expose the attendees to the “latest and greatest”
technologies from Texas Instruments. Track 1 and Track
3 featured talks and short workshops on a variety of TI
products related to embedded processing - Sitara
microprocessors, Stellaris Microcontrollers, OMAP,
MSP430, C2000/Piccolo, ARM, Hawk Board, and
Beagleboard. Track 2 featured a workshop on the
'DM6467/DM365 Architecture- Video Codec
Development and Optimization'. This workshop was
spread over two days.
A product exhibition had been organized, where stalls
were put up by TI customers, such as Wipro, PathPartners,
Cranes Software, Mindtree, Hawkboard.org, and
Windows Embedded. Participants could go up to any of
the stalls and get a live demo of the products on display as
well as get any information that they required on the
products by these companies.
Tutorial Programs
The conference was followed by a three-day tutorial
program at Texas Instruments Bangalore campus. The
following tutorials were held.
1
Introduction to CCS V4 – This tutorial was conducted
by E. Manivannan. It covered an overview of CCS v4
and its benefits, hands-on exposure to CCS V4,
debugging, and server-side scripting
2
A workshop on “Embedded Systems using MSP430
and TI Analog” – This three-day workshop was
inaugurated by Prof. C.R. Venugopal (SJCE, Mysore).
A keynote talk on “Low Power Embedded Systems –
Yesterday, Today, Tomorrow" was given by Ramesh
Ramamritham. A panel discussion on “MCU in the
curriculum” was held, where Bhooshan Iyer from
Cranes Software, C.P. Ravikumar from Texas
Instruments, and C.R. Venugopal from SJCE, Mysore,
were the panelists. On Day 2, Vasanth from Cranes
Software conducted a full-day hands-on training on
MSP430 microcontroller using MSP430
development tool from Cranes Software.
Demonstrations of several other MSP430
development tools were demonstrated by Sultan of
Cranes Software. This was followed by a hands-on
training on “Analog Lab-in-a-box” which is a tool that
can be used to teach embedded system design and
analog interfacing concepts to students.
3
A full-day training on Embedded QT was conducted by
Prabindh Sundareson of Texas Instruments. The
following topics were covered – Overview of QT/E,
Pre-requisites and Configuring QT/E, Drawing with
widgets - Events and Properties, Drawing with
widgets - Events and Properties, Using QT/E as a
middleware. He also conducted a hands-on training
on using QT/E for creating a user interface.
4
A two-day workshop on TI C6000 DSP Programming
and Optimization was conducted by experts from
Texas Instruments. Speakers included Venugopala
Krishna, Anirban Sengupta, Girish Murthy, Y
Madhusudan, and Sathish Kumar. The architecture of
C6000, compiler-driven and programming
optimizations were discussed.
Vishal Borker of NextBit Computing demonstrating a
Thin Client solution based on TI technology
Sharmila and Vijayendra of Mindtree demonstrating a
Telemedicine Camera solution
A keynote demonstration at TI Tech Days 2009 in
Bangalore. Praveen Ganapathy (TI) introducing Prem
Anand from Ittiam Systems, who demonstrated their
Thin Client solution based on TI semiconductors.
Anirban Sengupta, a speaker in the TI C6000 DSP
Programming Tutorial conducted as part of TI Tech
Days on Embedded Processing
Pradeep Bordia of Tandberg demonstrating the
TANDBERG PrecisionHD USB Camera based on TI
embedded processing solutions.
Audience in the C6000 DSP Programming Workshop
25
UniTI on Campus – Events in 1Q, 2010
On January 7, K.R.K. Rao of TI India delivered a talk on
“Analog Electronics” at the Education Forum held as part of
the International Conference on VLSI Design (Nimhans
Convention Center, Bangalore). The talk was attended by
more than 100 participants.
MSP430 microcontroller. The workshop also included
demos on MSP430. The workshop was attended by 60
student participants and 10 faculty members. We thank
Prof. Maheshwaran, Faculty Advisor, ECE Department for
the hospitalities.
On January 22, C.P. Ravikumar visited Muthyammal
College of Engineering, Tamil Nadu to inaugurate MINDSS
2010 and deliver a keynote talk on “From Semiconductor
Devices to VLSI Systems-on-Chip - An unparalleled story of
innovation.” The event was attended by more than 300
students and faculty from several colleges. Thanks are due
to Prof. Madheswaran, Principal of the college, for his
invitation.
On February 23, C.P. Ravikumar visited Mookamabigai
College of Engineering, Kalamavur, to inaugurate an AICTEsponsored staff development program on VLSI Design. He
also delivered an expert talk on the topic of VLSI Testing.
We thank Prof. Susan Christina, Head of the Department,
ECE, for the invitation.
On January 30, Prohor Chowdhury from Texas Instruments
visited IIT Kharagpur to take part in the seminar on
Microcontrollers from Texas Instruments and deliver a talk
on “C28x and Stellaris family of devices”. The seminar was
attended by about 100 students. We thank Mr. Praveen
Agarwal for the hospitalities.
On February 5, C.P. Ravikumar visited Goa University to
take part in the VLSI Design Symposium. He spoke on the
topic “From Semiconductor Devices to VLSI Systems-onChip - An unparalleled story of innovation.” There were
about 120 participants, including students and faculty from
Engineering institutions in Goa. Thanks are due to the
organizers of the conference and to the executive
committee members of the VLSI Society of India (Goa
chapter).
On February 20, G. Anand Kumar from Texas Instruments
visited VLB Janakiammal Engineering College, Coimbatore,
and conducted a half-day workshop on Texas Instruments
On February 24, C.P. Ravikumar visited IIT Chennai to
deliver two invited talks on the topics “Multicore SoC
Design for next level Power/Performance” and “Memory
System Optimization” in the instructional enhancement
program organized by SMDP program of the Ministry of
Information Technology.
On March 1, Rajagopal Karthik from Texas Instruments
visited Coimbatore Institute of Technology and conducted
a seminar for the students of Electrical and Electronics
Engineering on the topic High speed I/O design and Signal
integrity. We thank Prof. S. Vasantharathna, Assistant
Professor, Electrical & Electronics Department for the
invitation and the hospitalities.
During March 29-31, a three-day faculty development
program on "Applications of Analog Electronics" is being
organized jointly by the RNS Institute of Technology
(RNSIT) and the Board of Studies, VTU through UniTI. K.
Radhakrishna Rao (TI India) and Krishnamurthy Bhat (BEC,
Bagalkot) will be speakers in the program. We thank the
management of RNSIT for lending support to conduct this
event.
Cranes Software, TI India's University Program partner, conducted the following events in Indian campuses in 1Q,
2010.
On Jan 7, 2010, Vinod Thomas delivered a full-day tutorial
on “DSP and Embedded Processing WITH TI Platforms” at
K.L. College of Engineering, Vijayawada. The objective of
the tutorial was to provide a tutorial overview of the TI's
digital signal processing and embedded processing
solutions. The tutorial was part of the conference “RF and
Signal Processing Systems (RSPS-2010)” organized by the
college. Many thanks to Prof. Siddaiah of KL College of
Engineering for local arrangements. The tutorial was
attended by 45 participants from academia and industry.
A full-day workshop on “TI DSP and MCU” was conducted
at Siddaganga Institute of Technology (SIT, Tumkur) on
25th Jan'2010. Prashant Sonar and Hussain Saheb of
Cranes Software were the speakers at the workshop. The
workshop was attended by 30 participants from SIT.
Thanks are due to Prof. Stephen and Prof. Kiranmayi for
local arrangements.
26
Jayaram of Cranes Software conducted a half-day tutorial
on “Beagleboard Development Platform based on OMAP
3530 processor” at IIT, Chennai, on 24th February, 2010.
The workshop was attended by 20 participants from
across engineering colleges. Thanks are due to Prof.
Srinivasan and Prof. Nitin Chandrachoodan for local
arrangements.
Shabarinathan of Cranes Software conducted a half-day
tutorial on “Multiprocessor Technology (DSP+FPGA)” at IIT,
Chennai, on 24th February, 2010. The workshop was
attended by 20 participants from various Institutes.
Thanks are due to Prof. Srinivasan and Prof. Nitin
Chandrachoodan for making the local arrangements.
Requests for half-day, full-day, and longer workshops at
College Campuses may be directed to Imran Sayeed of
Cranes Software ([email protected])
Low Power RF Workshop at BITS–Goa Campus
Atul Lele and Ramakrishna Reddy of TI India conducted a
day-long workshop on Low Power RF at BITS Goa Campus
on Jan 23, 2010. The workshop was part of the Student
Technical Festival “Quark” being conducted at the BITS
campus. About 40 participants took part in the hands-on
workshop, where several low-power RF tools from Texas
Instruments were used to drive home the concepts of
wireless sensor networks and their applications.
In wireless sensor networks, the nodes of the sensor
network are often separated by short distances which can
be covered using low power RF communication. Since the
nodes are powered by batteries, it is important that the
power required for the communication is as low as
possible. Special low-power protocols such as SimpliciTI
and Zigbee have been invented for this purpose. In this
workshop, the intention is to expose student community to
the concepts of embedded systems, sensor networks, and
low-power RF communication. The eZ430-RF2500 tool
from Texas Instruments was used to illustrate these
concepts. This tool is based on MSP430 microcontroller
and the CC2500 radio communication chip capable of RF
communication in the frequency band centered at 2.4GHz.
In addition, the applications of the Chronos wireless watch
from TI was also demonstrated.
Ramakrishna Reddy lecturing at the event
The workshop included both lectures and hands-on
components. In the lectures, Atul and Rama covered
aspects of MSP430 and Low Power RF communication.
The Students carried out hands-on experiments on the
eZ430-RF2500 in the post-lunch lab session.
We acknowledge the help from Swetansu Mahapatra,
student coordinator, for his help in coordinating the event.
Participants at the event, mostly undergraduate
students from BITS Goa and other colleges
Hands-on session on eZ430-RF2500
Atul Lele lecturing at the event
27
A one-day workshop on Power Management
A one-day workshop on Power
Management was held at Manipal
University in Karnataka on Feb 26, 2010.
Ramprasad Ananthaswamy, Venky
Srinivasan, and Sridhar Ramaswamy from
Texas Instruments, India were the
speakers at this workshop.
Ram Ananth delivered the opening talk in
the workshop, starting with what power
management is and its significance in
modern-day electronic systems. He
covered some of the key developments in
the area of power management over the
last few decades. He then spoke abut
basic power converter topologies and
their applications. For the benefit of
students, he spoke about opportunities in
power management industry for
graduating engineers. He ended his talk
with a coverage of new frontierssolar/wind/fuel cells/renewable energy.
Venky Srinivasan speaking at the Power Management Workshop
Sridhar Ramaswamy spoke about system
design and provided an overview of how
one must start at the top to understand
the system-level requirements and
translate them into IC design
requirements. He also spoke about
system optimization to reduce cost and
maintain the desired level of performance
and power.
Venky Srinivasan talked about Power
Management ICs and the building blocks
of such integrated circuits. He went into
the details of various cells in circuits and
covered the aspects of cell-level and toplevel design.
The workshop was coordinated by Prof.
Harish Padiyar of Manipal University.
UniTI thanks Prof. Padiyar and his
colleagues for the local arrangements.
Audience at the Power Management Workshop (Manipal University)
UniTI Website
http://www.uniti.in/
28
The Bibliophile
Op Amps for Everyone and Everyone for Op Amp!
Book Review
Title - Op Amps for Everyone
Authors - Bruce Carter and Ron Mancini
Publishers - Newnes Publishers (An imprint of Elsevier Science Publishers)
Indian Reprint available from Elsevier Science
Cost – Rs 450/Reviewed by Dr. K. Radhakrishna Rao, Texas Instruments, India
I was introduced to the Op Amp during my student days, when I was taught about the Analog Computer. I was
impressed by the way the Op Amp was presented in text books as well as in the class room. Later, when I started teaching
in 1970s, I came across the Application Notes on Op Amps by Burr-Brown Engineers. I still have a copy of the book
Applications of Operational Amplifier (Third generation techniques) by J.G. Graeme of Burr-Brown (1973, McGraw Hill). I
and my colleagues at IIT Madras readily adopted some of the system-level applications such as AGC (Automatic Gain
Control) and amplitude stabilization of oscillators using Op Amp and multipliers.
With Burr and Brown's acquisition by Texas Instruments, it is quite appropriate that
an authoritative book on Op Amps has been published by Bruce Carter and Ron Mancini of Texas Instruments. This book
will influence everyone (students, faculty and practicing engineers) through its comprehensive and thorough treatment.
The all-pervading nature of the Op Amp in the present day design of systems is well brought out. The authors discuss the
characteristics of Op Amp, its specifications, and design methodology. The authors point out the common pitfalls
encountered by designers, and make an effort to correct the wrong concepts and beliefs retained by the designers.
I urge the teachers of analog courses to adopt the building blocks, namely, Op Amp, Comparator and the Analog
Multiplier, to convey to the students almost all analog concepts of importance today. Without doubt, this book will be an
invaluable resource in such a course. I hope “Op Amp for Everyone” will lead to “Everyone for the Op Amp!”
29
TI has the largest portfolio of Embedded Processors for Academia
TI Embedded Processors
Microcontrollers (MCUs)
Digital Signal Processors
(DSPs)
ARM®-Based Processors
16-bit ultralow power
MCUs
32-bit realtime C2000
MCUs
32-bit ARM
Cortex™-M3
MCUs
ARM
Cortex-A8
MPUs
Highperformance
DSPs
MSP430™
Delfino™
Piccolo™
Stellaris®
Sitara™
& ARM9
C6000
DaVinci &
OMAP
Low-power
DSPs
C5000™
Code Composer Studio – One Install for All
Intro to
Engg
Intro to
Program
Digital
Control
Digital
Power
Mech.
Engg
Intro to Linux
Signals & Systems
Computer Science
Communications
MCU / Embedded Systems / MPU
Bio Engr
SensorNets
Lighting &
Sensing
Robotics
Real-Time DSP
Image/Video
Processing
Undergraduate, Graduate and Research Projects
Low Cost Software & Dev. Tools
Moving exercises from the lab to the hostel
BioMed
DSP Apps
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement