AC 2011-2307: DESIGN PROJECTS FOR PROGRAMMABLE EMBED-

DED SYSTEM-ON-CHIP COURSE

Prawat Nagvajara, Electrical and Computer Engineering Drexel University

Prawat Nagvajara received his Ph.D. degree from Boston University, Boston, Massachusetts, in 1989, MS
and BS degrees from Northeastern University, Boston, Massachusetts, in 1979 and 1980, respectively. He
joined the Department of Electrical and Computer Engineering at Drexel University in 1990 where he is
currently an associate professor. His research focus is on application specific computing using custom
hardware and commodity high-performance accelerators and embedded systems education.

Robin Kizirian, Electrical and Computer Engineering Drexel University

Robin Kizirian completed his M.S. degree in Computer Engineering at Drexel University in Philadelphia
and his B.S. degree in Computer Engineering with a minor in Computer Science at the University of
Hartford in Connecticut. He is currently serving as a research assistant at the Engineering Technology
department at Drexel University. Robin has been involved in various projects funded by Pfizer, NASA,
NSF and Department of Education. His areas of research include Embedded Systems, Mechatronics,
Efficient Solar Energy Systems, Internet-based Quality Control and 3-D Online Education.

Page 22.437.1

c American Society for Engineering Education, 2011

 Design Projects for Programmable Embedded
System-On-Chip Course

Abstract

The integration of microcontrollers with programmable electronic devices and design

automation software tools has greatly improved today’s embedded system design. Most
recently, there has been a trend to integrate microcontrollers with programmable digital
and analog peripherals on a single chip. This technology is known as programmable
embedded system-on-chip. At the author’s institution, programmable system-on-chip
(PSoC) devices such as the Cypress PSoC boards have been adopted in embedded system
design courses. This paper presents design projects for embedded system learning
purposes. Programmable system-on-chip, with its design software, offers a hands-on
approach that aids students in understanding the concepts behind each of the building
blocks, providing them with the sufficient tools and experience to develop real-world
applications. This paper summarizes three projects that have been developed and taught
as the assignments in embedded systems course. Each of the projects involves real-time
software concepts, e.g. interrupts, interfacing with external sensors and actuators, and
digital and analog hardware theories. The project materials will be disseminated on the
authors’ institution website for future use by other institutions.

I. Introduction

The study of embedded system design with microcontrollers is relevant to today’s

automatically controlled electronic devices. A microcontroller is an integrated circuit
comprises of a microprocessor, memory, and other input and output peripherals. It offers
many features such as interrupts and General-Purpose Input/Output interfaces (GPIO).
The interrupt feature provides the chip with real-time capabilities, including the
capability of providing a proper response to a specific event. The GPIO feature allows
microcontroller to interface with other electronic devices. It is common to incorporate
microcontrollers together with programmable digital and analog peripherals on a single
chip1,2.

One new trend in embedded designs is to integrate microcontrollers with programmable

peripherals on the chip3,4. This technology is known as programmable embedded system-
on-chip (PSoC). The Cypress PSoC device is an example of a programmable system-on-
chip device. It includes a microprocessor and mixed-signal arrays of configurable
integrated analog and digital peripherals5,6 which offers a practical embedded systems
learning tool. The PSoC boards have been adopted in embedded system design courses
at the authors’ institution.

This paper presents a number of microcontroller-based design projects for embedded

system learning purposes. These projects are implemented using the Cypress PSoC
boards11. PSoC technology provides students with a hands-on approach in understanding
Page 22.437.2

the concepts behind each of the building blocks when developing real-world
applications7-9. This paper presents three projects: Pulse Width Modulation (PWM) fan
control, gravity measurement, and a wireless traffic light control system. The projects
have been taught as part of an embedded systems course and their materials will be
disseminated on the authors’ institution website for future use by other institutions. The
projects cover real-time software concepts such as interrupts, interfacing with external
sensors and actuators, and digital and analog hardware theories. These projects can better
prepare students for the industry, emphasize the impact of software on the controller’s
architecture, and introduce the use of networking as an engineering tool.

II. Programmable System-on-Chip Technology Overview

Programmable System-on-Chip (PSoC) is an integrated circuit whose architecture

consists of a CPU core, configurable analog and digital blocks, and programmable
routing and interconnects. The core includes multiple core options, choice of internal or
external oscillator, flash memory, SRAM, sleep and watchdog timers, and multiple clock
sources. The number of digital and analog blocks varies from one device to another, but
is a major factor in differentiating the PSoC from a regular microcontroller. Having the
ability to configure analog and digital circuitry on the same chip provides the engineer
with the flexibility of integrating both logic resources with complex analog flows. At the
heart of the analog blocks is a collection of switch capacitors, op-amps, comparators,
ADC, DAC, and digital filter blocks. Re-routing between the I/O pins also offers the user
more logic freedom and simplification as the global buses allow for signal multiplexing
and logic operations6.

The projects in this paper are based on the CY8C2xxxx family of PSoC1 devices. These devices
can be programmed by using the PSoC Designer software that can be downloaded for free from
the Cypress website11. The software simplifies the coding process by allowing the addition or
removal of components by dragging and dropping the desired digital or analog peripherals. Each
of the building blocks is fully explained and documented in separate PDF documents. After
creating the necessary blocks, the software automatically builds the API libraries to access and
control these blocks. The processor is programmed by using the API libraries and C Language.
The boards that are used in this paper are shown in Figure 1. Figure 1(a) displays the PSoC1
Evaluation board while Figure 1(b) shows the wireless PSoC1 Low Power RF node. Both boards
are Cypress Semiconductor products.

Page 22.437.3

Figure 1(a): PSoC1 Evaluation Kit Figure 1(b): PSoC CYFI Low Power RF Node
III. Pulse Width Modulation Fan Control Project

An instructional application of embedded systems is designing an efficient fan-based

temperature controller. Van Ess et. al presented this design project in their laboratory manual5.
The relevance of the project can be justified by the sensitivity of electronic devices to
temperature. Most of these devices contain fans which have to be controlled to keep the
electronic environment at a stable temperature. This represents a practical application that can
familiarize students with different engineering theories and concepts such as PWM and
frequency measurement techniques.

The system design is shown in Figure 2(a). The fan is equipped with an integral tachometer
which allows the monitoring and controlling of the fan’s speed. The controller utilizes a simple
pulse width modulator to change the width of the pulse provided to the fan and thus, allowing to
change its speed. Although the fan’s speed is calculated by measuring the frequency from the
tachometer, this measurement is only possible when the fan is fully driven. For this purpose, a
tachometer overriding state machine has been programmed into the PWM interrupt handler to
generate an override signal and calculate the speed appropriately. In addition to this setup, an
integrating speed controller has been added to capture the fan’s speed and adjust the PWM duty
cycle to drive the fan with a constant speed. This concept teaches students about closed feedback
loop systems and shows them their significance in adapting to sensitive variations in a system.

Figure 2(a): Schematic for Fan Project5 Figure 2(b): The Implemented Project

The temperature measurement is performed by using a thermistor. Thermistors provide

resistance values as a function of the measured temperature. By calculating their resistance, an
accurate measurement of the temperature can be obtained. This relationship between the
resistance and the temperature is described by Equation 1: The Steinhart-Hart equation5.

where T is the temperature in Kelvin

Page 22.437.4

R is the resistance in ohms

A, B, and C are the constants
By using the Steinhart-Hart equation, the necessary look-up thermistor table is created. As for
the resistance calculation, it can be performed in Figure 2(a) by Equation 2.

Comparing the measured resistance to the look-up table, it is possible to find out the temperature.
Once the temperature is determined, the controller can use the information to modify the fan’s
speed accordingly to keep the system cool. The project is implemented on the PSoC1 Evaluation
Board and is displayed in Figure 2(b).

IV. Gravity Measurement Project

Real-time computing is an essential phenomenon in the study of embedded systems. A common

technique to address this issue is by using interrupts. An interrupt-driven system allows an
asynchronous signal to pause the processor, save its execution state, and then execute the code in
the interrupt handler. Once the interrupt has been handled, the saved state of execution is
recovered and the program pointer returns to where the processor had been stalled. The gravity
measurement project serves as a useful tool to teach about the concept of requesting interrupts
and handling them. In PSoC1, an interrupt has to be declared and coded in C language by using
the #pragma interrupt_handler directive. This statement will order the compiler to save and
recover the value of the registers, and return from the interrupt subroutine while the execution
has ended. In the boot file, the Interrupt Vector Table has to also be edited so that it jumps to the
required handler once the interrupt has been requested.

The project has been setup as in Figure 3(a). A number of Infrared digital sensors, as in Figure
3(b), have been placed in a vertical fashion on a wall. The sensors have been fixed equidistantly
at one foot from each other. They are connected to the general purpose ports on the PSoC board.
A round object is dropped at free fall from the top. Once the ball is dropped, each of the sensors
detects the ball and signals the controller as a general purpose input/output interrupt request to
keep track of the elapsed times at which the ball crossed the sensor. The elapsed times are
gathered within the interrupt handler and later used for calculating the gravitational acceleration
of the ball. The calculations are conducted by utilizing linear regression methodologies and
finding the best linear fit of a Velocity v.s. Time graph.

Page 22.437.5
 Figure 3(a): Gravity Measurement Project Setup Figure 3(b): IR Digital Sensor

Figure 4 represents the pseudo code for the controller algorithm. The elapsed times are captured
and calculated in the interrupt handling section. The elapsed times are stored in an array whose
pointer is incremented at every signal from a sensor. Once all the sensors have triggered the
interrupt, the Boolean that signifies the completion of the collection process becomes true. At
this point, the post-data-collection processing takes place. The velocities corresponding to the
elapsed times are calculated and the linear regression technique is applied. The calculated slope
of the regression line is the value of the gravitation acceleration. The value for gravity is finally
displayed on the LCD screen that is provided by the PSoC board.

//Initiation
Distance = Distance between sensors
Collection_Done = False
Start Timer

//GPIO Interrupt Handling and Processing

Read Timer Value
Store Timer Value in Array
Increment Pointer

If Pointer = # of Sensors
Collection_Done = True

//Post-Data-Collection Processing
If Collection_Done = True
Calculate Elapsed times from Array
Calculate Velocities corresponding to Elapsed Times
Apply Linear Regression Technique
Gravity = Slope

Display Gravity Measurement on LCD Screen

Page 22.437.6

Figure 4: Controller Algorithm to Calculate the Average Gravity

The linear regression methodology10 used is described in Equations 3, 4, and 5.

where x and y are the variables

n is the number of values
m is the slope of the regression line
b is the intercept point of the regression line
r is the correlation coefficient

Table 1 displays sample gravity measurements data and the corresponding best linear-fit graph.
As can be seen, the slope of the line is 31.659 ft/s2 which is approximately equal to the average
gravitational acceleration of 32.2 ft/s2.

Table 1: Sample Gravity Calculation Data

Elapsed Time Velocity

Best Linear Fit for Gravity Measurement
(seconds) (ft/s)
20
0.134 4.3148
15
Velocity (ft/s)

0.257 7.8345
10
0.366 11.9256 y = 31.659x + 0.0085
5 R² = 0.9955
0.458 14.2693
0
0.535 17.4754 0 0.2 0.4 0.6 0.8
Elapsed Time (Seconds)
0.597 18.5344

V. Wireless Traffic-Light Control Project

The main project in the embedded systems course involves designing a wireless traffic-light
controller. This project presents the concepts of network protocols and topologies and their usage
as engineering tools. It combines these networking tools with microcontroller technologies to
develop an integrated wireless embedded design. Using PSoC devices for wireless embedded
systems provides the advantage of fewer components and shorter design cycles. PSoC utilizes
Page 22.437.7

the CyFi Star Network Protocol11 as the basis for its wireless system. The protocol is portrayed
in Figure 5. In the star network topology, packets are transmitted back and forth between the
nodes and the centralized hub. These communications are bidirectional and take place as from
node to hub or hub to node. Up to 250 nodes can be connected to one hub. The nodes and the
hub are implemented using wireless PSoC1 devices and are powered using two AAA batteries.

Figure 5: Wireless Star Network Topology

In each of the wireless components, the PSoC device executes the network protocol, runs the
programmed application, and drives the CyFi transceiver. The CyFi transceiver features a 2.4
GHz RF (Radio Frequency) solution and is controlled by an active power management system
for low power communications among the wireless components. Figure 6 displays the proposed
system design for the wireless traffic light controller. The system consists of two nodes and a
hub. On one side, the sensor node interfaces with analog proximity sensors, collects traffic
information, and then transmits it to the hub. The hub correspondingly forwards the information
to the traffic light controller node. The traffic light controller node sends a signal to a PSoC1
traffic light controller board to either maintain its state or change it accordingly.

Wireless Node Wireless Node

Traffic Controller
Wireless Hub

Distance Infrared
Sensors
Page 22.437.8

Figure 6: The Proposed System for Controlling Traffic Lights Wirelessly

The idea behind using a traffic controller board is to have enough power to drive a relay system
to light up the traffic lights. The PSoC1 Evaluation board can satisfy the necessary 5V operating
voltage for the relays whose coils are connected to a 12-volt power supply. For safety purposes,
the installed traffic light bulbs operate at 12 volts and utilize a low current of approximately
500mA. Inside the controller, a state machine as in Figure 7 has been programmed to change the
state of the lights according to the signal from controller node. At any of the states except for
Green/Red, the state is maintained until an expired timer signal has been received. When the
state is Green/Red, the state is not changed until an expired timer signal is received and traffic
information indicates that a car is waiting.

Figure 7: Traffic Light Controller State Machine

Cypress has included the Application Programming Interface (API) for its CyFi topology to
simplify the control of its Star Network Protocol. Before packets can be transmitted among the
nodes, each of the nodes has to be bound with the hub. The binding process is started by the hub
when it initiates its listening process to any requesting nodes. The nodes then connect to the hub
by requesting to be bound and registering their device IDs. This means that the hub stores all the
nodes’ device IDs; however, the nodes do not know each other’s IDs. After the nodes are bound,
the network has been established and data can be transferred. Figure 8 demonstrates the different
stages through which a packet traverses when it is transmitted from the sensor node (Node 1) to
the control node (Node 2). At the sensor node, the packet is programmed to include the control
node’s device ID in its payload. The packet also contains the traffic information data that the
sensor node is planning to send to the control node. Since CyFi utilizes a star network protocol,
the packet has to be sent to the hub initially before it is forwarded to its destination. Thus, the
header information indicates that the sender is the sensor node and the receiver is the hub. When
the packet arrives to the hub, the hub determines the packet’s destination from its payload. It
creates a new packet which includes the sender’s device ID and a copy of the data in its payload.
At this stage, the header information indicates that the sender is the hub and the receiver is the
Page 22.437.9
control node. Once the packet arrives at the control node, the node identifies that the sender is
the sensor node and then analyzes the data that it has received.

Figure 8: Packet Transmission from Node 1 to Node 2

The overall system has been setup in a laboratory environment at the author’s institution. Figure
9 displays the implemented wireless traffic light controller and highlights the different sections
of the system.

Traffic
Lights

Control Sensor
Node Node
Relay
System
Analog IR Sensor

Hub
Traffic
Controller
Page 22.437.10

Figure 9: The Implemented Wireless Traffic Light Controller

VI. Evaluations

The students in the embedded systems course were presented with pre-course evaluations
to determine their strengths and weaknesses prior to performing the laboratory
experiments that accompany the course. After being taught the programmable system-on-
chip technology and getting involved in the projects, the students were asked to complete
the same course evaluation questionnaire.

The questionnaire targets the students’ understanding of the course materials and
measures the enhancement in the learning of embedded systems concepts and
applications. The evaluation questionnaire is the same for pre- and post-evaluations and
is displayed in Figure 10.

The pre- and post-course evaluation results are presented in Figures 11 and 12. The final
results demonstrate a marked increase in the students’ understanding of microcontroller-
based technology and embedded systems. By using programmable system-on-chip
devices, the students gained experience in designing systems composed of both hardware
and software components. Their understanding about sensors and actuators in addition to
their knowledge in computer architecture, interrupts, and mixed signal systems drastically
improved.

1 2 3 4 5
Survey Question Least like Not Neutral Like Most
me like me me Like me
1. I have experience working with software O O O O O
program interfacing with hardware.
2. I understand how embedded systems and O O O O O
microcontrollers are applied to engineering
problems.
3. I have experience designing systems that O O O O O
have hardware and software components.
4. I have used design automation tool that O O O O O
integrates programmable hardware and
software development.
5. I understand how software programs O O O O O
interface with sensors and actuators.
6. I have knowledge in computer architecture O O O O O
and organization, and assembly language.
7. I have used interrupts and interrupt service O O O O O
routines in real-time software.
8. I have worked with mixed-signal systems O O O O O
(digital and analog).
9. I design embedded systems with design O O O O O
stages and debugging plan.
10. I am prepared for career in embedded O O O O O
systems design.
Page 22.437.11

Figure 10: Pre and Post Course Evaluation Questionnaires

 Pre-Course Evaluations
70

Percentage of Students
60
50 Most Like Me
40
Like Me
30
Neutral
20
Not Like Me
10
0 Least Like Me
1 2 3 4 5 6 7 8 9 10
Question

Figure 11: Results from the Pre-Course Evaluations

Post-Course Evaluations
70
Percentage of Students

60
50 Most Like Me
40
Like Me
30
Neutral
20
Not Like Me
10
0 Least Like Me
1 2 3 4 5 6 7 8 9 10
Question

Figure 12: Results from the Post-Course Evaluations

VII. Conclusion

Enhancing microcontroller teaching techniques is important for insuring quality and up-to-date
education for future generations of electrical and computer engineers. The paper proposes the
integration of programmable system-on-chip technologies into microcontroller learning.
Programmable system-on-chip devices, such as the Cypress PSoC boards, have been utilized in
the authors’ institution for the laboratory experiments of an embedded systems course. These
devices offer programmable digital and analog blocks that introduce the students to the different
digital and analog design criteria and provide them with a better hands-on industry-like
experience. Three main projects were incorporated into the course, each of which targeted a
specific field of embedded systems learning. Based on an evaluation questionnaire completed by
Page 22.437.12

each student at the beginning and end of the course, significant improvement in student learning
(and understanding of microcontroller-based concepts and their integration into embedded
systems) was noted. Integrating PSoC devices into microcontroller coursework appears to be of
great benefit in electrical and computer engineering education.

References
[1] Mar, M., Sullam, B., and Blom, E., “An Architecture for a Configurable Mixed-Signal Device”, IEEE Journal of
Solid-State Circuits, Vol. 38, pp. 565-568, March 2003.

[2] Fang, W., Kedar, S., Owen, S., Wei, G., and Brooks, D., “System-on-Chip Architecture for Intelligent Sensor
Networks,” Proceedings of the 2006 International Conference on Intelligent Information Hiding and Multimedia
Signal Processing(IIH-MSP’06), Pasadena, CA, December 18-20, 2006.

[3] Faura, J., Horton, C., Krah, B., Cabestany, J., Aguirre, M.A., and Insenser, J.M., “A New Field Programmable
System-on-a-Chip for Mixed Signal Integration”, Proceedings of the European Design and Test Conference
(ED&TC’97), Paris, France, March 17-20, 1997.

[4] Wallner, S., “Design Methodology of a Configurable System-on-Chip Architecture”, Proceedings of the 12th
Annual IEEE Symposium on Field-Programmable Custom Computing Machines (FCCM’04), Napa, CA, April 20-
23, 2004.

[5] D. Van Ess, E. H. Currie, and A. Doboli, “Laboratory Manual for Introduction to Mixed-Signal, Embedded
Design”, ISBN 978-0-9814679-1-7

[6] Seguine, D., “Just Add Sensor – Integrating Analog and Digital Signal Conditioning in a Programmable System
on Chip”, Proceedings of IEEE Sensors, Vol. 1, pp. 665-668, November 2002.

[7] Hall, T., and Hamblen, J., “System-on-a-Programmable-Chip Development Platforms in the Classroom”, IEEE
Transactions on Education, VOL. 47, No. 4, November 2004.

[8] Nurmi, J., Madsen, J., Ofner, E., Isoaho, J., and Tenhunen, H., “The SoC-Mobinet Model in System-on-Chip
Education”, Proceedings of the 2005 IEEE International Conference on Microelectronic Systems Education
(MSE’05), Anaheim, Ca, June 12-14, 2005.

[9] De Man, H., “System-on-Chip Design: Impact On Education and Research”, IEEE Desgin & Test of Computers,
Vol. 16, Issue 3, 1993.

[10] D. Montgomery, E. Peck, and G. Vining, “Introduction to Linear Regression Analysis”, ISBN: 978-
0471315650

[11] Cypress Semiconductor Website: http://www.cypress.com

Page 22.437.13