Advanced Rowing Instruments – Business Plan & Final ... - Saket Vora

Advanced Rowing Instruments
WIN BASSETT
JARED EVERETT
GREG MULHOLLAND
SAKET VORA
SPRING 2007
FINAL
DOCUMENTATION

Advanced Rowing Instruments, Final Documentation
John E. Bassett IV
Jared S. Everett
Gregory J. Mulholland
Saket R. Vora
May 1, 2007

Acknowledgments
Special thanks go to Dr. Stephen Walsh for his expertise, feedback, tireless support of the Engineering
Entrepreneurs Program, and for helping ARI reach its true potential; to Prof. Percy Hooper
for his enthusiasm in establishing a collaboration between the College of Design and the Electrical
and Computer Engineering Department; to Dr. Thomas Miller III for making the Engineering
Entrepreneurs Program possible; to Sam Dirani and Michael Caston for their tremendous effort
and patience; to Gordon Jeans for his generousity with time and resources; and to Barbara Yde
for helping to make the Silicon Valley trip a reality and for keeping this dynamic program running
on all cylinders.
Finally, special thanks go to each member of our eTeam for their tremendous effort and dedication.
Without them, the success of ARI would not be possible.
1

Contents
I Executive Summary 5
II Business Plan 8
1 Introduction to Rowing 9
2 Business 13
2.1 Business Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Marketing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.2 Market Research - Customer Survey . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.3 Market Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2.4 Brand Establishment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3 Sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.2 Direct Sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.3 Boat Distributors & Retailers . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.4 International Sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.5 Financial Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4 Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4.2 Competitor Breakdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5 Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.5.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.5.2 S.W.O.T. Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.5.3 Price & Features Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.6 Intellectual Property . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.6.1 Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.6.2 ARI’s Intellectual Assets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.6.3 Chief Competitors’ intellectual assets . . . . . . . . . . . . . . . . . . . . . . . 31
2.7 Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.8 Product Roll-Out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2

2.8.1 Phase One . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.8.2 Phase Two . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.8.3 Phase Three . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3 Corporate Structure 35
3.1 Management Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.1.1 Board of Directors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.1.2 Senior Leadership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.1.3 Virtual Employees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.1.4 Virtual Employee Rotation System . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2 Company Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.1 Employees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.2 Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2.3 New Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
III Product Development 39
4 Hardware 40
4.1 Accelerometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.1.2 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.1.3 Voltage Follower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.1.4 Orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.1.5 Underlying Mathematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.2 Power Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.2.1 Battery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.2.2 Voltage Regulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.3 Audio Amplifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.3.1 Microphone Pre–Amplifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.3.2 Power Amplifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.4 Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.4.1 Liquid Crystal Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.4.2 Driver IC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.5 Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.5.1 Control Buttons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.5.2 Audio Connectors and Volume Control . . . . . . . . . . . . . . . . . . . . . . 44
4.5.3 Other Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.6 Board Interconnects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.7 Media Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.8 Microcontroller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.8.1 General Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.8.2 Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.8.3 Processors Considered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3

5 Embedded Software 47
5.1 Variable Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.2 System Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.3 Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.4 Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.5 LCD Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.6 Stroke Rate Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.7 Secure Digital . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.8 Bluetooth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6 Non-embedded Software 52
6.1 MATLAB Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.2 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
6.3 Stroke Rate Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
6.4 Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
7 Case Design 62
7.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
7.2 Inspiration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
7.3 Modular Battery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
7.4 Packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
7.5 Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
IV Appendices 65
A Financials 66
B Project Timeline 68
C Design Day Plans 73
D Functional Specifications 79
E Virtual Employee Job Descriptions 81
F Call Logs 85
G Schematics & PCB 90
H Component Datasheets 96
I Industrial Design Documentation 112
J Business Documentation 123
K Parts List 127
4

Part I
Executive Summary
5

Executive Summary
Rowing Rowing is a simple sport — 4 or 8 athletes propel a boat down a body of water as
fast as possible. A coxswain leads the crew. He or she is the only person on the boat without
an oar, and he or she sits at the end of the boat (bow or stern) and directs the rowers. His or
her direction must be heard by all rowers. To facilitate this, a device known as a “coxbox“ has
been developed. An audio amplifier connects to speakers spread throughout the boat, allowing
rowers to hear. Additionally, a coxswain must know how his or her crew is performing. ARI uses a
combination of innovative technology and better pricing to enter this market. At the core, there is
the idea that through data-driven decisions, teams can increase performance. ARI creates products
that bring this idea to reality.
Market Rowing as a sport is increasing in popularity throughout the United States. The total
addressable market for rowing is $398M in the United States, and the segmented addressable market
for rowing electronics is $7M to $12M. These size estimates increase dramatically in the European
markets due to the greater popularity of rowing. Currently there is only one real company serving
the needs for a large number of rowing teams that field 4– and 8– man shells and it commands over
95% market share. However, their product has not seen much innovation and is priced too high for
many teams to purchase the equipment. As a result of extensive market research, ARI plans to field
two products that serve the low and high end of the rowing market. The lower-end product, the
Mercury, will be priced lower than the competition’s higher end model and thus attract a greater
number of customers. The higher-end model, the Titan, will feature significantly more capabilities
and technology than any other rowing device on the market, and will be attractive for the elite
teams.
Sales The sales cycle in the sport of rowing is seasonal due to the timing of regattas and training.
Sales can be expected to be high at the beginning of seasons and fall as teams focus on training
rather that equipment upgrades. Sales will be achieved through direct means such as phone calls,
website sales, and company representatives, as well as indirect means via boat manufacturers and
retailers. Pro-forma financial projections indicate a break-even point in just over two years with a
20% to 30% market share attained in four years.
Competition Rowing is a well-developed sport in both the US and internationally. The primary
company is Nielsen-Kellerman, established in 1978. Two small companies operate in the UK and
Australia, and in 2004 another startup in the US was formed called In2 Rowing. While technology
the In2 Rowing product is closer to what ARI intends to offer, their price point is closer to $800
to $1000. The current market share for In2 Rowing is less than 5%. ARI must work quickly to
establish brand and company recognition with limited marketing funds. Rowers maintain a social
network and word-of-mouth will be critical, as well as a demonstrated win with an established
team. ARI’s modular battery system has no comparison in the market. ARI must stay on top of
the innovation cycle, as well as become a standout leader in customer service and quality. Durability
is highly valued in the rowing community.
Intellectual Property Landscape ARI is in a position to fully develop a powerful intellectual
property portfolio of its intellectual assets. Though the Mercury and the Titan currently have no
6

patentable technologies due to prior art, issuance of a design patent is possible to secure protection
of the casing and to obtain the right to use ”Patent Pending” or ”Patent Issued” to frighten
potential infringers. Furthermore, algorithmic trade secrets are in place, and trademarks on all
company branding will be effective after public debut. Similarly, copyrights already exist and will
continue to protect ARI’s documentation as it evolves with the company. ARI has taken every
precaution with respect to competitors’ intellectual property, and the company will continue to
strengthen its intellectual property portfolio as the growth of the company allows it.
Technology ARI’s technology is its driving force. A fundamental change is the use of accelerometers
to detect metrics such as stroke rate. Combined with data logging and software tools, coaches
can analyze the performance of a boat with unprecedented granularity and with metrics previously
unavailable to them. A prototype of the Mercury model has been constructed featuring
an audio amplifier, power management circuitry, microcontroller, accelerometer, display driving
circuitry, and interface elements all built from scratch into a cohesive system. A sleek unique
case was designed and produced by two industrial designers. Future plans for the Titan model
include Bluetooth wireless support for audio and telemetry as well as Secure Digital storage for
datalogging.
7

Part II
Business Plan
8

Chapter 1
Introduction to Rowing
National Administration
Rowing in the United States is governed by USRowing. This organization governs most regattas
held within the United States. This includes all US National, Collegiate Men’s, Master’s, High
School, and Junior rowing. Interestingly the NCAA governs Collegiate Women’s rowing directly
because of its burgeoning popularity. USRowing regulates safety, judging, regatta operations,
and team behavior. A key area of regulation that affects product development is on the water
communication. According to the USRowing Rules of Rowing: “No crew shall receive any outside
assistance, coaching, or advice during a race” 1 This is important to note for the development of
any on-the-water device and is the primary legal limitation to the development of any measurement
device.
Vessels
A crew has one purpose: to move their vessel across a specified distance as quickly as possible.
These vessels are known as shells. 2, 4, or 8 rowers can work in concert to propel a shell. The
most prestigious races are performed with the 8-rower configuration. These shells are typically 60
ft in length and can range in price from $18,000 to $30,000. They can be expected to have a racing
lifetime of approximately 5 years. Some shells will in fact maintain their stiffness—the primary
metric used to determine speed—for longer, but most will have lost some considerable speed by the
end of their fifth year of racing.
1 USRowing Rules of Rowing. http://www.rci.rutgers.edu/ ronchen/rowrul1b.htm#2-410. Accessed: Dec 10, 2007.
Figure 1.1: An 8-man shell.
9

Seasons
Within the sport of rowing, there are two very distinct seasons. In the fall, teams typically compete
in s of 5000m. These races derive their history from winter races in England where many of the
rivers are too narrow for multiple boats to pass one another. Thus, they are run as time trials
on ten-second intervals These events focus on an athlete’s ability to maintain performance over
a long period of time. This includes technical ability and aerobic output. Coaches often require
long-distance and long-interval workouts of their athletes during this season. During the races
themselves, While the specific approaches to the head season vary greatly from team to team, it is
generally accepted that the season is preparation and maintenance for the spring season.
The spring season consists of a series of sprint races of 2000m. These races test the short term
aerobic and anaerobic output of the athletes. Furthermore, much higher stroke rates increase the
need for improved control by each of the athletes. These are the races that most Americans think
of when the idea of crew is suggested.
Teams
The fundamental unit of rowing is a . A boat consists of 1,2,4, or 8 people who compete in a race.
Typically, boats belong to a team. Teams consist of anywhere from 1 person to several hundred.
Beyond that, teams vary widely. Competitive categories are divided by age. While USRowing,
the governing body of rowing in the United States, has a complex system of age classification
for national events, teams can typically be broken down into three simple categories: Junior,
Collegiate, and Master. Junior teams include anyone under the age of 18 and teams associated
with high schools. Collegiate teams are those associated directly or indirectly with a University
or College. Finally, master’s teams are independent teams to whom anyone can belong. These
teams are usually comprised of adult rowers who have graduated beyond the age of collegiate and
junior competition. It is important to note that age distinctions are relatively unimportant in team
performance. Many junior teams can compete on the level of collegiate teams and master’s teams,
for example.
While age is a defining category of competition, team association also plays a role in its structure
and operation. There are three categories of association: Varsity, Club, and Independent. Varsity
teams represent a school as a fully supported entity. Athletes on Varsity teams are often lettermen
and in some cases earn scholarships. Club athletes on the other hand do not earn Varsity letters and
do not have scholarship opportunities. They are, however, associated with a school. Independent
teams allow any rower of a particular age to participate. Most Master’s clubs are independent as
are many Junior clubs. For the purposes of this document, it is important to note that Varsity
teams are often well-funded, while most other teams work under a more limited budget.
Rowing Technique
Rowing is referred to many as the ultimate team sport. Every rower must be in concert with every
other to achieve maximum speed. This means that each rower must have a reproducible, consistent
stroke.With that in mind, there are two key parts to the rowing stroke. Both are vital to the
synchronization of the boat.
Drive The drive is the part of the stroke that propels the boat forward. It begins at the catch,
where the rower drops the blade in the water. The rower then extends his legs, accelerating the
10

slide backward. It is during this part of the drive that most of the power is delivered to the boat.
Once the legs are fully extended, the rower then straightens his back. Finally, he pulls his arms
into his ribcage to complete the stroke. When the arms are in use, the rower has the finest control
of the blade of the oar. Thus, he uses this time to prepare for the extraction of the blade from the
water.
Recovery The recovery is the pause between drives, when the blade is reset at the beginning of
the stroke. At the culmination of the drive, the blade is lifted from the water. It is then rotated
90 degrees to be parallel to the water. This lowers wind resistance while the oar is out of the
water. The rower then performs the reverse of the drive. The arms are extended away from the
body, the rower leans forward, and then the legs are compressed. During this time, control of
the body is crucial. While it is extremely easy to let the body collapse forward to the catch, this
dramatically slows the boat in an effect known as . Because the rower is facing the stern of the
shell, any acceleration in that direction will need to be countered before the next stroke can begin.
If a rower accelerates too quickly down the slide, he will be forced to stop himself with a great deal
of pressure on the footplate, pushing the boat in a direction counter to that of the drive. Coaches
focus on this area of the stroke more than any other in most experienced rowers.
Coxswain While rowers are obviously a focal point of the sport,
coxswains play a vital role in the operation of any team and the success
of any boat. While rowers are responsible for providing the force
behind the boat, the coxswain performs all of the thinking tasks necessary
to optimize power over the course of the race, steer, and motivate
his rowers. He is seated at the extreme end of the shell (either bow or
stern depending on configuration) and is usually sitting or laying low in
the boat to minimize wind resistance.
A coxswain must think about many things at one time. During a
race, he must be aware of all other boats in the vicinity. He must know
what each of his rowers is doing and how it will affect the performance of
the crew as a whole. He must steer the boat using the on-board rudder.
And most importantly, he must tell his rowers what to do. This is no
small task. Coxswains are nearly always the lightest members of a crew
and therefore must assert their authority by sheer force of personality.
Coxboxes
Figure 1.2: A coxswain
encrouages his crew.
The coxswain is the primary driver of the crew. Without his leadership, a crew has difficulty
maintaining pace, steering, and concentrating. To effectively control his crew, the coxswain must
be able to communicate with them. Because rowing takes place on rivers, it is ofter very windy
during races. The rowers, sometimes over 40 feet away, need to be able to hear his commands. Thus,
a device known as a coxbox was developed. This audio amplifier feeds the coxswain’s voice—via
microphone—to a set of speakers placed along the length of the shell, enabling the rowers to hear
it. In addition, some coxbox models also give the coxswain simple data about the boat. Common
models include a timer for workouts and a counter of stroke rate(strokes per minute). The timer
is typically controlled via toggle switch. Stroke rate is determined by a magnet placed under the
11

stroke seat of the boat. A magnetic sensor is then fixed to the deck of the boat. When the magnets
pass, the box can sense the induced current and registers a stroke.
A more advanced version of the device, primarily intended for use in small boats without the
need for audio amplification, senses real time water speed. Similarly to stroke rate, a sensor impeller
is placed on the outside hull of the boat. A magnet is then placed directly over the impeller. This is
connected to the SpeedCoach near the coxswain. This speed measurement does require a secondary
device—the SpeedCoach—to be installed and an impeller to be attached to the hull of the boat.
Even though the impeller does not profoundly affect the speed of the boat, the psychological impact
of attaching any adhesive to a brand new $30,000 hull is daunting to any coach.
12

Chapter 2
Business
2.1 Business Model
Advanced Rowing Instruments sells electronic devices to rowing teams that are an alternative to
the current marketing offering as well as having rich new features which will help increase a team’s
performance—all at an affordable price. Products currently available are either not technologically
up to date, too expensive for large segments of the market, or lacking in key features. ARI focuses
on its strengths in practical innovative technology, modular design, data logging, and performance
analysis that lead to better results. ARI focuses on the 4-man and 8-man rowing teams that make
up the majority of the junior and collegiate market. These teams have specific needs in terms of
both features and affordability which are currently under-served by competitor offerings. Marketing
will take place at regatta events, holding product demonstrations, advertisements in Internet and
magazine rowing related media, and creating a web presence that taps into the online community
of rowers. Coaches of rowing teams maintain a high degree of camraderie, and coxswains are vocal
in reviewing products they use. Word of mouth marketing is an important one to maintain and
support. ARI primarily sells its products directly to consumers through direct sales and Internet
orders. Hardware devices are sold with a warranty, and periodic commercial updates will become
available for the data analysis software. It is common practice for rowing electronics to be sold
through boat manufacturers, and ARI will partner with boat manufacturers to facilitate this reseller
sales channel.
ARI will utilize the Internet to support sales to any customer in the world. A sales team will
begin by focusing on the geographic proximity of the Southeast, Atlantic, and Northeast markets.
Raleigh, NC works well as a base of operations due to its central location with respect to these
three lucrative and high density markets. In the future, sales representatives will be hired to serve
other geographic markets in the United States and the world.
ARI maintains a small research and development team which oversees product development from
both a hardware and software perspective. A sales and marketing staff canvas concentrated rowing
markets and serve as the face of the company. Manufacturing of electronic devices is contracted
through an electronic contract manufacturer. As such, ARI maintains no overhead facilities or
manufacturing resources.
ARI seeks to build upon its core technology of electronic sensors used for performance analysis.
Future plans not only include creating a solid national presence for ARI but to expand into new
13

Figure 2.1: Business Model
14

sports markets.
2.2 Marketing
2.2.1 Overview
As described above, the sport rowing is divided into several different divisions with a variety of
racing categories, from single person sculls to 8-person shells. Different racing categories have
different needs—a single scull does not have a coxswain or need a voice amplifier whereas an 8person
shell can require up to three speakers. Due to the background and motivation for entering
the market, ARI is focusing on the 4-person and 8-person competitive racing category popular in
the junior, collegiate, master, and national divisions. It is this segment’s needs that are not being
met by the current offering of rowing instruments.
2.2.2 Market Research - Customer Survey
ARI performed market research using online databases, websites, phone calls to companies in the
rowing market, interviews with coaches and coxswains, and a customer survey.
An online customer survey was created and asked questions directly related to rowing electronics.
Questions regarding the size of teams, number of equipment, replacement rates, problems with
equipment, annual amount spent on electronics, and amount teams would spend for ARI’s two
product lines were completed by the customer. The full survey and results can be found in the
appendices. The survey was sent to over 200 coaches of NCAA Division I Women’s programs,
Men’s collegiate club teams, private clubs, and high school clubs. 73 people responded the survey,
a response rate of approximately 40%. A summary of the findings can be found below:
These results indicate a good cross-section of the entire rowing population. There is a wide
variety of team size, number of coxswains, and number of rowing electronic equipment. Coaches
were asked to estimate how much money was spent on rowing related electronics. This included
purchases of coxboxes, coxvoxes, seat magnets, speakers, microphone headsets, etc.
To better meet the needs of the customer, survey takers were asked to rate causes of problems
they have experienced with their coxboxes and coxvoxes. These included concerns about the battery,
casing, controls, audio amplifier, and the system as a whole. Respondents indicated that concerns
over battery life were highest, followed by integrity of the controls and the quality of the audio
amplifier. Problems regarding the casing and the overall system were rated very low. As a note
concerning the audio amplifier: the general comments section of the survey and through message
board research, several people indicated problems involving hiss and static. These were anecdotally
caused by headset cable damage, dirty connections, and sometimes physically broken audio amplifier
circuit.
In order to determine a market acceptable price for ARI’s products, the survey takers were given
a list of features that was intended for the Mercury and asked to specify a price, in $50 increments,
they would be willing to pay for such a product. Almost 40% of respondents indicated that they
would pay $300 for the Mercury model, followed by 21% who would pay $350. The mean price for
the Mercury is $329.33.
The same question and analysis was done for the Titan product, with just over a third of
respondents indicating they would pay $500 for the product. One respondent indicated he or she
15

Figure 2.2: The Amount of Money Spent Annually on Rowing Electronics
16

Figure 2.3: Level of Concern for Components Causing Problems
would pay $900 for it, while another individual stated he or she would not buy it at any price. The
mean price for the Titan is $511.97.
The rowing electronics market has been heavily dominated by a single company for nearly two
decades. As a new company intending to break into such a market, the question was asked whether
people would even be interested in purchasing from a new company or would old loyalties dominate
their purchasing decision. The results show that nearly all the respondents are very interested or
somewhat interested in a new company, with 56% responding with the former.
In the general comments section, many areas of concern were revealed. The number one most
important factor for coaches is durability of the product. These devices are used in harsh conditions
and three years of use are expected out of them. Many responders indicated a frustration with
Nielsen-Kellerman and seemed excited that a new competitor would be in the market.
2.2.3 Market Segmentation
Rowing is the fastest growing sport among NCAA Division I Women’s Sports 1 . As the shift to the
sunbelt continues, more and more rowing clubs at the junior and collegiate levels are starting up.
According to a recent NCAA Participation study, in 2001 there were 200 men and women collegiate
teams. Many junior and collegiate teams, especially those who do not have Varsity status, face
considerably tighter budgets. Crew teams can have annual operating budgets from $20,000 to
over $100,000. Though this sounds significant, much of it is spent on regatta fees and maintenance.
Combined with the high price of shells and boathouse and lake fees, such teams often find it difficult
to afford electronic rowing devices. As a result, they cannot buy enough units to equip their entire
team, or they must sacrifice functionality (such as the ability to see a timer or a stroke rate) due
to the high prices. Our Mercury model addresses this compromise. However, teams that have
1 Rowing News Media Kit, 2006
17

Figure 2.4: What People Would Pay for Mercury
sizable budgets are free to look for products that can give them the edge in performance. Our
Titan model fits the needs of these teams by offering the technology and software to analyze their
team’s performance in order to make better training decisions.
Total Addressable Market
The Rower’s Almanac, USA based organization, conducted a national survey in 2004 and determined
a market value for the sport of rowing at $398 million. In the United States, the total
rowing population was estimated at 177,500 people. Due to the greater popularity and presence of
rowing in the Europe, this number is much higher when international markets are considered. For
instance, the size of the rowing population with respect to the total population is six times greater 2
in the United Kingdom (which has approximately 22,000 rowers) than in the United States.
Segmented Addressable Market
The total addressable market for rowing-related electronics estimated to be between $7 to $12
million market in the United States. Ms. Karen Solem-Derringer, publisher of the Rower’s Almanac,
provided the analysis as follows: Of the 177,500 strong rowing community in the US, 13.7% stated
2 Amateur Rowing Association http://www.ara-rowing.org
18

Figure 2.5: What People Would Pay for Titan
they intended to purchase rowing electronics. The averaged intended purchase amount used was
$290. ARI conducted its own survey and found an estimated average intended annual purchase
amount of approximately $485.
Share of Market
The market research described above revealed a demand for a rowing device with the features of
the ARI Mercury at a price point of $300. The features of the ARI Mercury not only matches the
core features of the Nielsen-Kellerman CoxBox, but offers improvements on it. From interviews
with coaches, coxswains, and rowers, the general consensus is that teams would like to purchase
additional electronic rowing instruments for their teams, but are unable to afford the high prices of
the current offerings. The benefit of such devices are not debated, but their benefit-to-cost ratios
are a subject a heated debate in teams. By keeping a price point below that of our competitors
and by reducing costs, ARI hopes to capitalize on moving up the demand curve and receive orders
from a greater percentage of the market. The strategy for seizing market share is two-fold: (a)
the price point will attract new customers within the market to purchase our products in greater
quantities, and (b) existing customers will switch to ARI because of our features and capabilities.
In the first four years, ARI hopes to capture 30% to 40% of the US market for rowing electronics.
19

2.2.4 Brand Establishment
Figure 2.6: Interest in Trying a New Company
Critical to successfully breaking into a dominated market will be to establish a brand recognition.
We will accomplish this in several ways:
Figure 2.7: ARI Logo
2.3 Sales
2.3.1 Overview
• Purchasing print and Internet advertisements in popular rowingrelated
media, such as Rowing News, Rower’s Almanac, and
Row2K.com.
• Creating banners with ARI logos for display at regatta
• Distributing ARI product posters at rowing related stores
• Offering freebies with ARI branding at regattas
• Community extension into message boards
Due to the seasonal nature of the sport of rowing, the sales cycle for rowing products features
a unique pattern compared to other consumer electronics. The fall season begins in September
20

and the spring season begins in March. The bulk of our sales will occur during this time frame,
with a smaller amount of additional sales occurring throughout the seasons and the summer for
replacements. For teams that have models by our competitors, there will be a lag period from
the time those products need to be replaced and our product would be a viable alternative. The
sections below describe the channels through which we intend to sell our product.
Another goal of ARI is to actively enter the international market in order to capitalize on
the greater number of rowers and the popularity that rowing enjoys overseas, particularly in the
European and Australian markets.
2.3.2 Direct Sales
Direct sales consist of sales made via phone calls or regional representatives directly to the end
customers. Phone calls could be prompted by viewing print or Internet advertisements, word of
mouth advertising, or not being comfortable with purchasing over the Internet. Regional representatives
are a common way rowing related companies interact with boat clubs by region, and ARI
will adopt a similar approach to help facilitate sales in this channel.
Regattas
A major component of marketing and advertising will be made through attending regattas. Boat
manufacturers and distributors as well as other companies who produce rowing-related products
make frequent appearances at numerous regattas throughout the year in order to develop relationships
with consumers, be on hand to answer questions or show their wares, and to obtain constant
feedback from their consumers. Regattas will also serve as an opportunity for teams to purchase
our products directly from us. These are important events, due to the relatively large number
of teams that appear at Regattas, their regularity during the rowing season, and the significant
amount of ’downtime’ between races that occurs.
E-Commerce
In order to reach consumers we cannot meet at regattas and to reach teams from across the country,
we will develop an online store where consumers can purchase their products easily and securely.
This is standard practice among our two primary competitors.
2.3.3 Boat Distributors & Retailers
Ultimately, we will make partnerships with boat manufacturers and distributors. With these partnerships,
our products will be offered to consumers when they are purchasing boats, so they can
choose to add our products to their order. The retailers will then purchase the products through
our company. This will streamline sales for consumers who wish to buy all their products at once
from the same retailer. This is currently done at some retailers with Nielsen-Kellerman products.
2.3.4 International Sales
There are new competitors that enter the picture when considering the international market. Row-
Data, which operates out of UK and CoxMate, which operates out of Australia, are the only two
manufacturers to truly speak of, and through anecdotal analysis the market share situation is same
21

there as it is here: Nielsen-Kellerman continues to dominate. This is likely due to their long-time
presence though, not through some particularly strong feature about their product.
Opportunity with Irish Venture Capitalist
On March 5, 2007, an ARI executive was contacted by Eamonn Hynes, a recent electrical engineering
graduate in Ireland who was conducting research on “electronic monitoring system for use in rowing
boats“ for a venture capital firm based out of Dublin, Ireland. ARI briefly spoke with Hynes and
indicated that after this business plan was created more discussion would follow. This chance
encounter has increased ARI’s confidence in initiating overseas sales more quickly than initially
anticipated.
ARI will look into partnering with a group of like-minded entrepreneurs for licensing and redistribution
within concentrated markets, such as in the United Kingdom and The Netherlands.
ARI will continue to create a brand and company presence while receiving a portion of each sale.
2.3.5 Financial Projections
A preliminary pro forma set of financial projections have been created. These projections extend
to Year 4, with greater detail for the first two years. For detailed information relating to cost of
goods sold (COGS) and staffing costs, please refer to the appendices.
In order to reduce initial sunk, capital, and overhead costs, the ARI executives have decided to
run the company in a very learn “startup“ mode. The executives will take a reduced salary, and
hiring will be kept to a minimum for the first year.
Figure 2.8: Breakdown of COGS by Component Type
Nearly 40% of the cost of goods can be attributed to the cost of the case material, which is due
to the unique shape of ARI products as well the specialized molds needed to produce such a case.
There is a high initial cost in the equipment needed to produce the case molds, but this capital
cost is reduced over time.
22

Figure 2.9: Financial Projections & Break Even Point
Figure 2.9 shows the pro-forma financial projections for ARI for the next four years. The full
source for these projections can be found the appendices. The chart shows a break even point just
after year 2.
23

Advanced Rowing Instruments
Projected Operating Statements & Cash Flows For Years
Ending December 31 (in $000'S)
Y1 Y2 Y3 Y4
Revenue:
Projected Annual Growth Rate
License & Royalty fees 1 3 7 9
Software Sales 1 4 7 10
Software Maintenance 0 0 1 1
Hardware Sales 50 345 552 972
Hardware Maintenance 0 18 17 41
Total Revenue
Operating Expenses:
52 370 584 1,033
Research & Development 60 56 56 56
Sales & Marketing 16 49 75 107
General Administation 113 105 105 105
Manufacturing 0 0 0 0
Total Variable Costs 25 66 105 156
Total Operating Expenses 214 275 341 424
Profit (Loss) from Operations Operations (162) (162) 94 94 244 609
Tax Provision 0 38 97 243
Net Income (Loss)
Cash Flow Analysis:
(162) 57 146 365
Contract and/or Equity Financing 250 500 1,500 800
Capital Expenditures (60) (95) (140) (165)
Change in Cash 28 462 1,506 1,000
Beginning Cash 50 78 540 2,046
Ending Cash 78 540 2,046 3,046
Variable Costs by Year: Y1 Y2 Y3 Y4
Trade Shows 8 15 20 24
Sales & Marketing Collateral 6 10 16 23
Legal Fees 2 1 3 4
Travel & Entertainment 3 8 10 15
COGS (Cost of Goods Sold) 6 32 56 90
Total Variable Costs: 25 66 105 156
Capital Purchases:
Employee Related 10 15 20 25
General 50 80 120 140
Total Capital Purchases: 60 95 140 165
Software Markup 50% 50% 60% 70%
Software Maintenance as % of Sale 5% 5% 10% 10%
Software Maintenance Increase 0% 10% 10% 10%
Hardware Markup 60% 80% 100% 120%
Hardware Maintenance as % of Sale 10% 5% 3% 4%
Hardware Maintenance Increase 0% 0% 3% 5% 5%

2.4 Competition
2.4.1 Overview
Due to the age of the sport, shell and scull manufacturers have existed for over a century. Rowing
related electronics, in contrast, have been in the market for just under 30 years. The primary
domestic competitor in the niche of rowing electronics is Pennsylvania-based Nielsen-Kellerman,
or NK. This company introduced the first coxswain amplifier in 1979, after Kellerman saw a boat
crash into a bridge because the rowers could not hear the coxswain. This device was called the
CoxVox, and added stroke rate display capability by the mid 1980s with their CoxBox. The
basic design and functionality of these two products have seen little change since their debut, save
for adding a race timer and stroke count display to the CoxBox unit. Today, Nielsen-Kellerman
features a number of additional products geared primarily to sculls. These products, such as the
StrokeCoach, SpeedCoach, and Cadence are newer in design and offer the ability to upload results
into a computer for analysis. Though crew teams could use these newer products in their 4 or 8
person shells, these products lack the much needed voice amplifier and are not compatible with
the CoxBox container commonly found on boats. Additionally, the StrokeCoach and SpeedCoach
still rely on the method of electromagnetic induction from magnets under the stoke’s sliding seat
to calculate stroke rate, requiring additional cable connections and maintenance. For boat speed
to be measured, an impeller must be attached to the shell’s exterior in a submerged position. New
shells can typically cost $20,000, so attaching an item to the hull is not the most desirable option
for teams. For these reasons, such products are ill-suited to match the needs for crew teams.
A new company called In2 Rowing out of Colorado, a subsidiary of hobby parts provider Spark
Fun Electronics, entered the scene in 2004. The founder was a novice rower in a collegiate crew
team when he had the idea to create his own coxbox, and the small In2 Rowing product lineup
directly addresses crew teams with 4 and 8 person shells. Their products feature a GPS (global
positioning system) which is used to determine boat speed. Personal experience with GPS systems
on the water have not provided satisfying accuracy. On November 6, 2006, In2 Rowing announced
their latest product, the In2 Trinity. The Trinity uses accelerometers to determine stroke rate and
boat dynamics in addition to its GPS for boat speed. In2 Rowing claims to offer a software suite
for data analysis, but no information is currently available on this. Pre-production models are
presently on sale, with production units slated to go on sale in January 2007.
Internationally, there is an Australian company called CoxMate which began in 2002 and an
English company by the name of RowData in 2003. Both these companies also target crew teams,
but the have virtually no presence within the United States.
Due to their first mover advantage, Nielsen-Kellerman overwhelmingly dominates the US market
for crew team electronics with an apparent market share of over 90%. Because of their age, they
have influenced the design of the mounting harnesses that shell manufacturers often install in the
boat. Indeed, much like the words ’Xerox’ and ’Kleenex’, Nielsen-Kellerman’s ’CoxBox’ has become
the commonly used term for this product.
2.4.2 Competitor Breakdown
Nielsen-Kellerman and In2 Rowing are our two domestic competitors, and a breakdown of each are
provided in this section.
1. Nielsen-Kellerman
25

• Year Founded: 1979
• Location: Boothwyn, Pennsylvania
• Employees: 50 to 99
• Yearly Sales: $10 to $20 million
• Market Share: Greater than 90%
• Key Products: CoxVox, CoxBox, StrokeCoach, SpeedCoach, Cadence
• Strengths: Well established, name recognition, first mover
• Weaknesses: Outdated technology for CoxBox, product suite does not target key segments
of rowing teams, expensive, non-modular design
2. In2 Rowing
• Year Founded: 2004
• Location: Denver, Colorado
• Employees: 6
• Yearly Sales: Unknown
• Market Share: Less than 5%
• Key Products: In2 Solo, In2 Trinity
• Strengths: New technology, directly addresses crew teams, data logging, non-impeller
speed determination
• Weaknesses: No name recognition, very expensive, unattractive product, non-modular
design, location
2.5 Positioning
2.5.1 Overview
ARI faces the challenge of breaking into a market with a heavily entrenched competitor. A positioning
strategy that describes how ARI will distinguish its products amidst the competition is
crucial for the company’s success. ARI has decided to focus on our four main strengths: acceleration
sensing, data analysis, modularity, and price. One of the initial aims of founding ARI was to
provide an affordable option for crew teams. Though In2 Rowing claims to have started with the
similar goal, the starter kit for their latest product sells at $799.95.
The following section describes the business and technical strengths, weaknesses, opportunities,
and threats for our company.
2.5.2 S.W.O.T. Analysis
1. Strengths
• Closely aligned with the needs of 4 and 8 person crew teams with high resource needs
• Acceleration sensing for drive and check analysis
• Detailed data and performance analysis
26

• Modular design for easy maintenance and battery charging
• Located in good geographic concentration of rowing teams
• Two models that target two major market segments in price and features
• No manufacturing facility overhead
2. Weaknesses
• No name recognition
• Beaten to market with acceleration sensing innovation
• Small company size results in smaller funds for marketing
• High initial costs must be offset early on
3. Opportunities
4. Threats
• Targeting an underserved market—teams that need basic CoxBox functionality without
the high price of existing models.
• Offering a convenience with our modular design not offered by any competitor
• Build upon our network in the rowing community
• Expand network and initiate sales in international markets
• Acceleration intellectual property might be filed by In2 Rowing
• Nielsen-Kellerman has weight to suppress entrance into the market
2.5.3 Price & Features Positioning
2.6 Intellectual Property
2.6.1 Strategy
Competitor Product Pricing
CoxVox Control Unit $309.00
Nielsen-Kellerman
CoxVox Starter Kit
CoxBox Control Unit
$410.00
$419.00
CoxBox Starter Kit $720.00
In2 Rowing
In2 Solo
In2 Trinity
$450.00
$800.00
ARI
Mercury
Titan
$330.00
$550.00
ARI’s intellectual property strategy is an adaptation of the approach Rajiv Patel used in his
article ”A Patent Portfolio Development Strategy for Start-Up Companies.” The four phase plan
is outlined below.
27

1. Phase I – Development
Identify key business goals of ARI
2. Phase II – Evaluation
Mine, analyze, and organize intellectual assets of ARI
• Patents – casing design patent
• Trade secrets – accelerometer algorithms, MATLAB algorithms
• Trademarks – Advanced Rowing Instruments, ARI, logos
• Copyrights – documentation, business plans, presentations, web site
3. Phase III – Procurement
Build portfolio by obtaining patents, trademarks, and copyrights (defense)
• File design patent
• File trademark applications
• Add copyright notices on documentation
4. Phase IV – Deployment
License own intellectual property, acquire other’s intellectual property, prosecute for infringement,
predict trends (offense)
• Crown-jewel patents – blocks competitors from entering technology market
• Fence patents – fences in competitors’ patents with all conceivable improvements so
competitors have to cross-license their patents
• Design-around patents – based on innovations created to avoid infringement on other’s
patents
2.6.2 ARI’s Intellectual Assets
Patents
A patent is the right granted by the government to exclude others from making, using, selling,
offering to sell, or importing that which is claimed in the patent. Having a comprehensive patent
portfolio can significantly increase the value of a company, but it can impose severe financial strains
as well if not executed properly. Unnecessary or old patents can quickly drain funds, and getting
a patent just to say you have one will not get a company off the ground or attract investors. The
decision ”to patent or not to patent” should not be taken lightly, and it can significantly influence
the direction of a company.
ARI’s Mercury and Titan technologies contain no patentable matter. Their inner workings
and features, such as coxswain voice amplification and using an accelerometer to measure boat
acceleration, all exist in forms of prior art. These methods have not been patented, however, so
ARI’s products are not infringing. But, since they already exist in the public market, a patent can
never be issued. Therefore, ARI is not pursuing any utility patents, or patents on functionality.
However, a possibility does exist for ARI to pursue a design patent for the casing of the Mercury
and Titan. Design patents are granted for the ornamental design of a functional item and cannot
28

include any functional claims. They prevent a substantially similar design from being made, used,
copied, or imported into the United States. It is important to know that design patents do not
protect an idea or an invention, but rather only protect ornamental design of exactly what is
pictured in the patent application.
For this reason, design patents are usually only effective in bulk. For example, getting several
design patents on similar variations will allow a company to fully harness the power of a portfolio
rather than relying on a single patent. Similarly, a design patent might be worthwhile to get
simply for the ability to print ”Patent Pending” or ”Patent Issued” on a product. This tactic can
both scare away potential infringers and make the product look more legitmate from a marketing
standpoint. ARI plans to apply for a design patent for this reason.
The design patent fees currently due to the United States Patent Office (USPTO) for a small
entity are $215. If the application is successful, there is an additional issue fee of $400 due to the
USPTO. This particular fee is for all patents — not just design patents. Again, this amount is
assuming you are a small entity. Forunately, there are no maintenance fee payments for design
patents, and once the design patent has issued, there are no other financial obligations necessary
to keep the design patent pending for the full 14 year term. This differs from utility patents that
have increasing maintenance fees due at 3.5, 7.5, and 11.5 years after issuance.
Trade Secrets
A trade secret is a practice, process, or algorithm used by a business to obtain an advantage over
competitors or customers. ARI has two key trade secrets — the algorithm used in the microcontroller
to calculate stroke rate and the algorithm used in the MATLAB code to calculate stroke
rate. Though these two algorithms are patentable due to Arrhythmia Research Technology, Inc. v.
Corazonix Corp., which ruled that the use of an algorithm to transform representations of physical
things may be patentable, there is no need to disclose their contents.
A trade secret, which costs nothing to get or maintain, better suits these two assets. They are
enforced by a ”need-to-know” information policy and non-disclosure agreements.
Trademarks
A trademark is a distinctive sign or other mark used to uniquely identify the source of its products
and/or services to consumers and to distinguish its products or services from those of other entities.
Trademarks are essential to guarantee assurance that consumers will easily recognize it a company’s
product or service. Similarly, brand loyality is also initiated by the use of trademarks.
There are two routes of obtaining a trademark. The first method is through the USPTO. Filing
a trademark application costs $250 if it is ubmitted online. If the application is successful, an
additional fee of $200 is due for a certificate of registration. A trademark filed with the USPTO
may use the R○symbol with the mark. On the other hand, no formal USPTO filing is needed for the
protection of a trademark. When a mark is used in the public market, protection is automatically
granted (if it does not infrigne on other marks). In this case, a trademark may use the TM symbol
with the mark. Though this method does not bring with it as much protection and leverage as a
registered trademark, a non-registered trademark is much less expensive with some protection still
attached.
Currently, ARI plans not to register its trademarks due to the expense. However, when funds
become available for such legal actions, ARI plans to trademark its company name, company
29

initials, company logos, and product model names. These include ”ARI TM ,” ”Advanced Rowing
Instruments TM ,” ”Mercury TM ,” ”Titan TM ,” and the logos below.
Figure 2.10: Logo with white background
Figure 2.11: Logo with black background
To best serve its purpose, the trademark notice should always accompany the trademark’s first
or most prominent appearance in a document, program, or packaging. However, the notice does
not need to be used each time the mark appears thereafter.
Copyrights
A copyright is a set of exclusive rights regulating the use of a particular expression of an idea or
information. It can protect written documents, business plans, web sites, software code, and other
ideas that are manifested on some sort of medium. Like trade secrets and trademarks, copyrights
do not need to be filed or registered and exist as soon the idea or information ”hits the paper.”
The copyright notice ( c○) may be used, but it is not required to obtain protection. It can however,
be used to deter potential infringers. Similarly, the phrase ”All rights reserved” used to be required
to obtain all rights granted under existing copyright law. However, the phrase is no longer due to
the Copyright Act enacted in 1976.
ARI’s documentation, web site material, promotional and marketing items, and software code
is copyrighted.
30

2.6.3 Chief Competitors’ intellectual assets
Patents
A patent search was performed on ARI’s chief competitors below. Only United States patents were
considered since international intellectual property law is beyond the scope of this documentation.
Nielsen-Kellerman
6 US patents
• 7,059,170 – Method and apparatus for measuring relative humidity of a mixture
• 6,257,074 – Vane anemometer with thermally isolated sensors
• 5,939,645 – Vane anemometer having a modular impeller assembly
• 5,783,753 – Vane anemometer having a modular impeller assembly
• 5,357,794 – Faraday effect small boat speed transducer and waterproof connection for same
• 5,099,689 – Apparatus for determining the effective force applied by an oarsman
In2Rowing, subsidiary of Spark Fun Electronics
0 US patents
RowData (United Kingdom)
0 US patents
Coxmate (Australia)
0 US patents
Davies Row Tech (United Kingdom)
0 US patents
As seen above, no US patents by ARI’s chief competitors are related to ARI’s Mercury or Titan
products. Furthermore, ARI does not infringe any other patents. Complete patent searching
results can be seen in Appendix X.
Trademarks
A trademark search was performed on ARI’s chief competitors below. Only United States trademarks
were considered since international intellectual property law is beyond the scope of this
documentation.
Nielsen-Kellerman
20 US trademarks
In2Rowing, subsidiary of Spark Fun Electronics
0 US trademarks
RowData (United Kingdom)
0 US trademarks
31

# Serial Number Reg. Number Trademark Status
1 78506116 3108038 CADENCE LIVE
2 78580831 3060395 KESTREL LIVE
3 76053082 2459706 WATCHWARE LIVE
4 76053296 2471487 SPEEDCOACH LIVE
5 76053250 2451024 STROKECOACH LIVE
6 76595853 2966128 POCKET WIND LIVE
7 76557341 2888771 NIELSEN KELLERMAN LIVE
8 76421535 2763731 KESTREL TRACKER LIVE
9 976179164 TRACKER DEAD
10 76053298 2549188 COX-BOX LIVE
11 76053295 2728273 POCKET WIND LIVE
12 76053256 2909386 POCKET WEATHER LIVE
13 75881525 2414021 COX-VOX LIVE
14 75881545 2429691 LIVE
15 75881531 2426744 INTERVAL LIVE
16 75881547 2831077 NK LIVE
17 75881546 2831076 NK LIVE
18 75570343 2346672 INTERVAL LIVE
19 75196901 2116023 KESTREL DEAD
20 74248416 1773499 PACECOACH DEAD
Coxmate (Australia)
0 US trademarks
Davies Row Tech (United Kingdom)
0 US trademarks
Table 2.1: Nielsen-Kellerman trademarks
As seen above, ARI does not infringe any of its chief competitor’s trademarks. Furthermore,
ARI does not infringe any other trademarks. Complete trademark searching results can be seen in
Appendix X.
2.7 Manufacturing
ARI develops both hardware and software products for the end consumers. Because the hardware
products have a core in electronics, the equipment required to manufacture such products is very
expensive. Each Mercury or Titan model will have a printed circuit board(s), microcontrollers,
LCD displays, batteries, etc. Additionally, the cylindrical casing for the CX and VX will have
to be uniquely created, requiring a specialized and expensive molding machine for production in
mass quantities. As a result, we have decided that the best route for manufacturing is to enlist the
services of an electronic contract manufacturer.
An electronic contract manufacturer (ECM), is a good fit for our company because they deal
32

with relatively low volume, high technology products. Our primary experience with an ECM
is with Steve Yauch, President of Carolina Electronic Assemblers located in nearby in Smithfield,
NC. Carolina Electronic Assemblers (CEA) can help serve companies with prototyping and product
development, whether a simple printed circuit board is needed or the complete product from PCBs
to casings. In a tour of CEA’s facilities, the ARI management team was confident that CEA was a
good match for our needs. Furthermore, they can assist companies with developing ideas from just
quick drawings on paper or fully detailed schematics. Mr. Yauch explained that they have staff
who can coordinate with over a hundred different electronic component suppliers, so the ARI team
will not be directly responsible for the intricacies of global supply chains.
This is an optimal arrangement for the company still in its start-up stage. By contracting CEA
to manufacturer our products, the employees at ARI can devote more of their efforts toward product
research and development, sales, and marketing. Additionally, the volumes that ARI anticipates is
too small to justify the cost of setting up an overseas production facility.
A preliminary bill of materials is currently under development as prototyping for the VX progresses.
A substantial expense, apart from high cost components such as large LCD screens or the
accelerometers, will be from the case. Due to the aquatic environments our products are required
to operate in, the case must be waterproof and rugged. The modularity of our case design poses
additional challenges. Sam Dirani indicated that high impact ABS plastic would be used for the
shell, with rubber co-molded into it. Apart from the high cost of the initial dies, such materials
should be had a much lower per-unit cost when in full production. When our staff increases, we
will have the manpower to further explore the total pricing for a limited run of the VX and CX
units. Our goal is to obtain a sample contract with Mr. Yauch about the total bill of materials for
our products and their fees.
2.8 Product Roll-Out
Introducing your product is a key challenge when trying to break into a heavily dominated niche
market. ARI executives Vora and Mulholland spoke with Dr. Mitzi Montoya-Weiss, professor of
strategic marketing at NC State University, about this issue and gained useful insights. We have
developed a three phase product roll out plan that charts a local-to-national path for ARI’s presence
in the market.
2.8.1 Phase One
The first step is to achieve a demonstrated win with the products. ARI will select a premier collegiate
crew team in the local area (such as the University of North Carolina’s Women’s Varsity
Crew team) or a crew team that the management team has an existing relationship with (such as
Norfolk Academy’s crew team in Virginia) and ask them to pilot ARI’s products. The company
will work closely with them, obtaining constant evaluations and immediately making possible adjustments.
This will be an opportunity to get extensive user feedback and showcase the company’s
commitment in customer service. This phase will last approximately a year, as ARI tracks progress
through the fall and spring collegiate crew seasons. A case study for marketing will be created
from the success of this pilot program. The case study, along with positive word-of-mouth from a
premier crew team, will be used for publicity and gives ARI’s products credibility.
33

2.8.2 Phase Two
Phase two will involve building upon the demonstrated win achieved in phase one. The company
website will include an e-commerce engine and online store for those teams willing to directly
purchase from us. Primary efforts, however, will be in developing personal relationships with crew
teams throughout the southeast United States and up the eastern seaboard toward New England.
During the fall and spring seasons, there are weekly regattas where shell and scull manufacturers
and distributors (such as WinTech, Resolute, Vespoli) and other crew product related companies
have regional sales representatives. These representatives are on hand to field questions from crew
teams, carry spare parts for on-site repair, discuss future products or sales, and to cultivate goodwill
among crew teams. ARI will build a presence at these regattas, having products on display for teams
to try out and to develop relationships with crew teams. Where appropriate, ARI representatives
will also visit clubs to do personal product demonstrations. When it is time to make a purchase
involving rowing electronics, the goal is to make every crew team think “let me look at that new
company, Advanced Rowing Instruments.“ Two regattas will be of particularly importance for ARI.
The Head of the Charles in Boston, MA is one of the most famous and prestigious regattas in the
world. Over 1,600 teams compete during the three day event. The Dad Vail regatta in Philadelphia,
PA is the largest collegiate regatta in the United States and over 100 teams are present. Substantial
marketing resources will be directed at these two regattas due to the opportunity to reach a very
large number of teams from around the country. Phase Two of the product roll-out plan is expected
to last approximately eight months to 1 year.
2.8.3 Phase Three
Toward the end of phase two, sales at regatta events and through the online store to grow. After
building up a strong regional presence, it will be time to start expanding nationally and internationally.
ARI will achieve this in two ways. First, using networking resources regional sales
representatives will be hired for large markets in the the western United States. These representatives
will utilize the model of attending regattas and clubs to spread awareness and to facilitate
greater sales activity. The second part will involve negotiating with national shell manufacturers
and distributors to sell ARI’s products on their websites. Currently, companies offer such thirdparty
products as a way of providing a one-stop place for a team’s crew product needs. For example,
if a crew team orders a NK CoxBox with an 8-man shell from Resolute Racing, Resolute Racing
will order the unit from Nielsen-Kellerman at cost and install it into the shell with an installation
fee. Similarly, ARI will work to have our products listed as third-party accessories that boat
manufacturer and distributors will sell. This will give ARI more national exposure and increased
sales.
34

Chapter 3
Corporate Structure
3.1 Management Team
3.1.1 Board of Directors
Figure 3.1: Organizational Chart for Prototype Development
The Board of Directors advise and provide guidance over the direction of the company. Gregory
Mulholland and Mark Adams are the two current members of the board of directors.
35

The Board of Directors as an entity are not responsible for day-to-day decisions. They receive
at the minimum quarterly reports on the health, growth, and direction of the company. If at any
time the direction of the company diverges from the interests of the board, a meeting will be held
with the members of the board and the companys day-to-day leadership (the chief executive officer
and president) to resolve such issues.
3.1.2 Senior Leadership
Chief Executive Officer – Saket R. Vora
The Chief Executive Officer is one of the two members in charge of the day-to-day operations of
the company. The CEO of ARI oversees the business related elements of the company such as
financials, customer satisfaction, and sales. In addition to these areas, the CEO also monitors
product development and provides input, support, and guidance where needed.
For the prototype development, Vora actively worked on the audio amplifier, schematics, LCD
driving, accelerometer configuration, case design and packaging.
President – Gregory J. Mulholland
The President is one of the two members in charge of the day-to-day operations of the company.
The president of ARI is primarily responsible for the technical elements of the company, namely
product development and services. The president will manage the hardware and software divisions
of the company, ensuring they are operating in sync with each other. Product development includes
primary research, design, prototyping, testing, and support.
Mulholland brings to the company extensive experience in the sport of rowing, having rowed
competitively for over six years and serving as president of the NC State Crew club team for two
years. His knowledge of the sport and contact network will be an invaluable asset to all facets
of the company. For the prototype development, Mulholland worked on the embedded software,
interfacing with accelerometer, and system testing.
Vice-President Business Strategy – John Edwin “Win“ Bassett IV
The Vice-President of Business Strategy will work closely with the CEO with the business elements
of ARI but have greater responsibility in seeing these parts executed successfully. In particular, the
VP Business Strategy will perform exhaustive intellectual property investigations, oversee market
research and sales strategy, and help create and manage all critical business documents such as
claims, business plan, and documentation.
Bassett brings technical writing expertise to ARI, along with a long held interest in patent law.
For the prototype development, Bassett completed the computer-based software and participated
in case design discussions. Bassett also directed graphic design skills for marketing materials.
Vice-President Hardware – Jared S. Everett
The Vice-President of Hardwarea is responsible for the product development, testing, and debugging
of the prototype. This person must be knowledgeable about system design, embedded systems,
analog circuits, and have extensive testing experience.
36

For the prototype development, Everett served as the driving force for the system development,
printed circuit board design, case design and packaging, product assembly, testing, and display
driving.
3.1.3 Virtual Employees
ARI is targeting seven to eight virtual employees (VE) join our team. Refer to the Appendices
for detailed descriptions of each VEs job responsibilities. It is important to realize that being a
start-up company, not a single member of the team will only do one particular task. The VPs
will not just manage but tangibly contribute to product development, the VEs will be performing
tasks in both technical and business related areas. However, it is important for each employee to
be responsible for a particular aspect of the business or product. This not only gives an employee
a sense of ownership, but also allows a direct way of performance evaluation and accountability.
3.1.4 Virtual Employee Rotation System
In order for the virtual employees to better understand the different business and technical areas
of the company, the ARI executives executed a four week rotational system that moved groups
of virtual employees through the various fields. At the end of the four week period, the virtual
employees were asked to choose which area they would like to work most in. While some virtual
employees ended up working primarily in those areas, there was still some fluidity between technical
units as the semester progressed. The virtual employees helped out where assistance was needed.
3.2 Company Growth
ARI started with two people who had a vision of creating a better product and putting that product
into more hands than what was currently offered. To see that vision become a reality, it is going
to take more than just two people.
3.2.1 Employees
In the start-up phase, ARI will have an employee base of around 5 to 6 people; four representing
leadership positions and two more assisting with sales, marketing, product development, or support.
Because of our product roll out plan, ARI intends to slowly hire additional sales representatives to
work regional markets. This will keep staffing and facilities costs low as we break into the market.
When the initial product design stage is complete, ARI will transition hardware division positions
into e-commerce positions, as building up the website will gain importance.
As ARI begins to expand nationally, additional sales staff will be added accordingly.
3.2.2 Facilities
ARI will maintain an office of approximately 2,000 sq.ft for sales, management, and laboratory
work.
37

3.2.3 New Markets
The technology core of ARI, using accelerometers to obtain useful information about sports training,
can be expanded to other sports beyond rowing. Part our company roadmap is to create a parent
company tentatively called , which will be responsible for developing products using our technology
for other sport markets. Advancements can be made in sports such as cycling, tennis, track & field
events, skiing, bobsledding, etc.
The reason for using the company name Advanced Rowing Instruments at this point in time is
to establish credibility in the rowing industry, ARI’s primary market.
38

Part III
Product Development
39

Chapter 4
Hardware
4.1 Accelerometer
4.1.1 Overview
Over the last ten years, a number of advancements have been made in MEMS technology. Common
now are acceleration sensing devices known as accelerometers. These devices are fabricated using
the same Silicon lithography techniques as microprocessors and use electrical measurements to
determine the real time, 1-dimensional force being placed upon them. 2– and 3– axis models
are available and are fundamentally several 1-d devices in one package. Accelerometers are well
suited to the sport of rowing because their packaging is extremely small, allowing them to be easily
contained within a watertight container.
All data in the sport of rowing comes from acceleration in one way or another. Because rowing
is a series of strokes and corresponding recoveries, the boat is constantly in a state of acceleration
and deceleration. At the catch, the boat will decelerate briefly when the oars enter the water.
When the drive begins, the boat will experience a growing forward acceleration which levels, then
declines. During the time, the boat is propelled forward at its fastest rate. During the recovery, the
oars are no longer in the water, so a quasi-constant deceleration due to the water will be seen. Each
of these phases can be seen clearly in the voltage output of the accelerometer. Through careful
analysis of the accelerometer output, boat speed, stroke rate, and transient accelerations can be
quantified and reported.
4.1.2 Implementation
ARI’s Mercury leverages this powerful new technology by using it to calculate the stroke rate.
An accelerometer board from Freescale Semiconductor is used to collect acceleration data in 3dimensions.
The accelerometer development board has three output pins, one for each axis of
motion. Each pin outputs an analog voltage proportional to the acceleration along its respective
axis of motion. The three analog output signals are then routed through an LM324N quad op-amp
IC before connecting to the PIC microcontroller. The analog acceleration data is then converted
through the analog-to-digital converter port into digital data to be processed by the PIC. The
microcontroller finally uses its stroke rate algorithm to convert the acceleration data into stroke
rate before pushing the information to the two digit LCD.
40

To reduce noise, the Freescale accelerometer has a sensitivity select parameter with four selectable
sensitivity settings. For the Mercury, this parameter has been set to the highest sensitivity
setting of 800 mV/g, which is for a g-range of 1.5g. This sensitivity setting allows for optimum
resolution for calculation of stroke rate, however it results in some signal clipping. The stroke rate
algorithm requires relative values, as it uses a modified zero-crossing approach to determine time
between strokes. Because the values are relative, noise created by clipping is not a problem. However,
the acceleration and speed calculation algorithms of the Titan will require a lower sensitivity
setting to prevent noise due to clipping.
4.1.3 Voltage Follower
The ADC port of the PIC18F4520 requires a source impedance of no greater than 2.5kω. The
output impedance of the accelerometer’s analog output was measured to be a high impedance
output, in the range of 100kω to 150kω. In order to turn this high impedance output to a low
impedance source for the ADC port, an operational amplifier configured as a voltage follower (unity
feedback) was used in between the accelerometer and the PIC18 microcontroller.
4.1.4 Orientation
The accelerometer datasheet in the appendices illustrates the orientation of the axes with respect
to the packaging of the development board. The accelerometer development board was mounted on
the B1 board so that the z–axis was oriented perpendicular to the base of the Mercury. Because of
the way these devices are most often mounted in shells, this precise orientation results in the z–axis
dominating the direction of forward movement of the shell. The x–axis is best suited to measure
the set of the shell, while the y–axis is best suited to measure the longitudinal balance of the shell.
4.1.5 Underlying Mathematics
Vector mathematics are used to calculate the stroke rate from the electrical data collected by the
accelerometer. If one assumes for mathematical purposes that the box has 1 measurement axis
normal to the direction of travel, a 2-dimensional Pythagorean Theorem can be used to determine
the magnitude of the acceleration vector:
aD =
�
a 2 DX + a2 DZ
Similarly, a three dimensional magnitude can be calculated:
aD =
�
a 2 DX + a2 DY + a2 DZ
It is important to note that these equations give only magnitude as an output. Because the stroke
rate is based on the frequency of the largest magnitude acceleration, this is acceptable. For the
Titan’s velocity calculation, however, a more in-depth analysis must be completed:
v =
� ��
�2 ��
axdt +
41
�2 ��
aydt +
�2 azdt
(4.1)
(4.2)
(4.3)

4.2 Power Management
The devices used in the Mercury required a range of source voltages to run. The audio amplifier
accepts an input voltage between 8V and 18V. Each of the digital devices (i.e. microcontroller,
quad op-amp voltage follower, LCD driver) accept a 5V input voltage. The LCDs and accelerometer
accept a 3.3V input voltage. To allow for full system integration from a single battery source, a
power management circuit was required.
4.2.1 Battery
The power for the system comes from a modular battery pack comprised of eight 1.2V rechargeable
NiMH batteries. Each battery provides 2500 mAh. Based on the estimated power consumption
of our product, this will result in a maximum single-charge lifetime of approximately 10 hours. A
major decision had to be made between using lithium ion versus nickel metal hydride as the battery
technology. Li–Ion has higher charge density, and can support a greater number of recharge cycles.
However, the cells are more expensive and not proven to be as safe as NiMH, which is a paramount
concern for a modular battery system. Due primarily to cost and ease of prototyping, NiMH
batteries were chosen.
4.2.2 Voltage Regulators
The battery pack provides nominal 9.6V directly to the bottom board (B2) of the system. This
voltage is of an appropriate value to be used unregulated to power the audio amplifier. To obtain
appropriate supply voltage for the other system components, two voltage regulators were needed
to provide 5V and 3.3V regulated source voltages. The 5V regulator received the unregulated 9.6V
power supply from the battery. The 5V output was then cascaded with the 3.3V regulator. Taps
were then connected from the regulated voltage output to to the appropriate devices through the
board interconnects. The power switch was wired in series with the battery pack. Thus, when the
switch is open it cuts power to the entire system. Resistors configured in a voltage divider network
were used in various places to fine-tune voltages where necessary.
4.3 Audio Amplifier
A core functionality of the Mercury and Titan devices is to amplify the coxswain’s voice so all the
rowers in a boat can hear him or her clearly. There are either one, two, or three speakers in a
shell. The standard speaker commonly found in shells is an 8ω load. The audio amplifier for ARI’s
products include two components: the microphone pre–amplifier and the power amplifier.
4.3.1 Microphone Pre–Amplifier
The type of microphone used in the widely available Nielsen-Kellerman headset is an Electrec
condensor type. This type of microphone requires a bias voltage supplied on the positive input
wire. This type of biasing is known as “phantom power“. Through detailed systematic testing, it
was determined that the input signal from the headset microphone needed an initial amplification
of 4x before being sent to the power amplifier. A NPN transistor model 2N 3904 with a resistor
network was used and constructed in the prototyping area on the printed circuit board. A 5kω audio
42

taper trim potentiometer was used between the pre-amplifier and the volume control potentiometer
to adjust the levels based on post-assembly variations.
4.3.2 Power Amplifier
A TDA2003 10 W Mono Channel power amplifier is used as the primary audio amplifier for the
device. This IC can take input voltages between 8V and 18V, and the battery pack supplied
a nominal 9.6V to it. Volume control was accomplished using a 5kω audio taper potentiometer
configured as a voltage divider into the input pin of the TDA2003 IC. The feedback and coupling
capacitors and resistors were selected according to the application test circuit provided in the
TDA2003 datasheet. The output of the power amplifier was sent directly to the speaker out.
Repeated and detailed testing was needed to properly configure the output of the audio amplifier
so that the microphone input would sound adequately loud but avoid hissing, garbled sounds, or
clipping. When the coxswain yells into the Mercury unit, current consumption of the audio amplifier
peaks as high as 400 mA at 9.6V supply voltage.
4.4 Display
The Mercury utilizes two 7-segment based LCDs to display two types of information. First, a six
digit timer shows the time elapsed of the current workout or race. Second, the current stroke rate
of the boat (in terms of the number of strokes per minute) will be displayed in a two digit display.
The production model will include a backlight to ensure visibility.
The Titan will display acceleration and velocity in addition to timer and stroke rate information.
It will also provide the option of displaying graphs of stroke rate, acceleration, and speed in realtime.
A pixel-based LCD will be used instead of character based display to allow greater flexibility
for the display of this additional information. The display must necessarily be larger to display
information in an easy to read, and uncluttered format.
4.4.1 Liquid Crystal Display
The Mercury utilizes Veritronix transflexive liquid crystal displays. These displays provide a number
of advantages. First, they provide large, easy to read 0.5” characters. Second, they utilize a positive
transflexive display type. This means that the images can be easily viewed in a variety of lighting
environments, including direct sunlight and very low light, making the display ideal for outdoor
use. Third, the LCDs utilize static bias which allows simple high and low signals to be sent to
each individual segment without generating complicated waveforms. This approach, in conjuncture
with an LCD driver IC, minimizes embedded code overhead and simplifies the circuitry between the
LCD and the microcontroller. Fourth, they are statically driven which provides the best contrast
ratio over the broadest temperature range, allowing for excellent visibility in even the most adverse
weather conditions.
4.4.2 Driver IC
The interface between the Veritronix transflexive liquid crystal displays and the microcontroller
consists of two Microchip AY0438 32-Segment CMOS LCD Driver ICs. These chips are designed to
work well with the PIC18 microcontroller. They both reduce circuit complexity and reduce overhead
43

in the embedded code. The AY0438 chips accept a serial data bitstream from the microcontroller
and activate the corresponding segments of each digit. Each AY0438 chip can drive 4 digits. The
Mercury has 8 digits to drive: six for HH:MM:SS and two for the stroke rate. The AY0438 chips
can be cascaded, so two chips are used in the Mercury model. For these reasons, just three control
lines are needed for the microcontroller to drive the entire display.
4.5 Interface
4.5.1 Control Buttons
Primary control of the device is performed using four control buttons mounted to the faceplate:
power, run, pause, and reset. The power button is used to start and stop operation of all systems
within the device. When the device is powered up, the ARI logo will automatically flash across
the LCD before initializing the timer and stroke rate displays. The run button is used to start
the timer and the stroke rate calculation. The pause and run buttons are then used to pause and
continue the timer. The reset button is used to reset the timer to 0:00:00.0.
4.5.2 Audio Connectors and Volume Control
The coxswain microphone is connected using a standard BNC connector. The speaker output
consists of an Amphenol Series 44 5-pin connector. Both of these match the current standard
connector types for coxswain microphones and speaker systems. This allows the Mercury and
Titan to interface with the customer’s existing infrastructure. Eventually, these connectors will
interface with the ARI coxswain microphone and ARI speaker system as well. The volume of the
audio amplifier is controlled by a vertically mounted scroll wheel on the faceplate. The wheel is
recessed into the faceplate for added durability.
4.5.3 Other Indicators
In addition to the liquid crystal display, the Mercury and Titan both include a red LED to indicate
power. This LED is also used to indicate battery low when flashing. The Titan will include
an additional button to activate/deactivate Bluetooth telemetry and a blue LED to indicate active
Bluetooth connection. The Titan will also include an additional red LED to indicate proper
installation of an SD card.
4.6 Board Interconnects
The inner structure of the Mercury is compromised of 3 circuit boards. The top board (A1) contains
the LCD, LCD driver ICs, and all button and LED connections to interface with the faceplate. The
middle board (B1) houses the the PIC18 microcontroller and the accelerometer board, including
the voltage follower used to interface the sensor with the PIC. Additionally, the prototype of the
Mercury includes an In-Circuit Debugger connector that can be used to flash the microcontroller
and test the embedded code without removing any components. The bottom board (B2) contains
the audio amplifier, pre-amplifier circuit for the microphone, and the power management circuitry.
In order to improve communication between circuit boards and facilitate this 3 board approach,
a system of board interconnects were used. The primary interface between the boards consists of
44

a series of single line pins that function as a single bus for all interboard communication. Custom
connector cables are used to connect the pin headers of each board, with one connection running
between each board (i.e. A1-B1, A1-B2, and B1-B2). The structure of each of these connections
places power rails at the first pin and ground rails at the last pin, to keep power lines clean.
Communications signals are then added in between. In addition to the primary commmunication
interface, the bottom board (B2) was fabricated as a PCB (printed circuit board) and utilizes PCBmount
screw terminals for external wires running to the battery, microphone BNC jack input, and
speaker output. This allows for both a solid electrical connection, and a secure mount to the board
that is easy to disassemble as necessary.
4.7 Media Storage
A key distinguishing feature of our CX model is that it allows for the acceleration data (sampled
at up to 100 times a second) and the coxswain’s voice to be recorded for multiple races. This data
will be stored inside the CX on a SecureDigital card, a widely available digital storage solution.
SecureDigital’s advantages over its CompactFlash is that it is smaller and consumes less power.
Utilizing a SecureDigital card also enables the end user to adjust how much storage is available,
from 16 MB to 1 GB.
The SecureDigital card slot, mounting, and interface is available through electronic parts distributors
such as DigiKey. This will be incorporated into our device in an easy to reach location
that is available when the CX is twisted and opened using the modular case design. Data is stored
on the SecureDigital card through the microprocessor which handles the real-time audio encoding.
Additionally, a computer data port can be included on the front plate of the CX behind a
waterproof cover that will allow the data to be transferred directly off the CX onto a computer.
4.8 Microcontroller
4.8.1 General Considerations
Many factors were taken into account regarding the microcontroller. In addition to the discrete
hardware requirements, discussed in the next section, form factor, development kit availability,
and power consumption. In the general sense, power consumption scales with chip complexity and
speed. Thus, a processor that can complete the specified instuctions at the lowest frequency will
consume the least power and is more desirable than a higher power model.
In addition to power, form factor is a major concern. Becuase early in the development process,
ARI will not have access to surface mount equipment. Because of this, processors that are packaged
in both easy-to-use PDIP packages and compact QFP packages.
Finally, to develop a product effectively, an easy-to-use development kit is vital. This must
include a programmer and minimal support circuitry.
4.8.2 Requirements
Multiple microcontrollers were examined for use in both the Mercury and Titan models. A number
of requirements were seen for each of the product lines, as can be seen in table 4.8.2. For control
purposes, digital input and output pins are needed for each model. In addition, each model requires
45

an 8-bit or higher analog-to-digital converter. Microcontrollers by their nature have digital I/O
pins. These are the only two features that the Mercury model requires.
Feature Mercury Titan
Digital I/O • •
A/D • •
MAC •
SPI •
USART •
Table 4.1: Required processor features.
Because of its increased feature set, the Titan model has a more stringent set of requirements.
Most notably are its off-chip peripheral communication interfaces. This microprocessor must have
a number of standards built in for ease and efficiency of use. Both Serial Peripheral Interface
and Universal Synchronous Asynchronous Reciever Transmitter (USART) must be implemented
for Secure Digital and Bluetooth, respectively. In addition for the more advanced analysis, such
as real-time velocity calculation, a Multiply Accumulate Unit must be included in the processor.
This will allow for the integration and filtering processes to be completed.
4.8.3 Processors Considered
A number of processors were considered for the ARI devices. These included product offerings
from TI and Microchip. Ultimately, Microchip was selected. Datasheet samples for considered
processors can be found in the appendices.
Texas Instruments Two processor lines from Texas Instruments were considered. The C54xx
and C55xx processor types were both examined for possible use. A development kit for the C54xx
had been previously ordered. Unfortunately, the C54xx did not have an A/D module built it.
Because of power concerns, this model was discarded. In addition a large amount of supporting
circuitry was necessary and as such these processors were overlooked. In future development, the
C55 should be closely examined, however, because it has native SD support.
Microchip Microchip was selected for the Mercury model because its development kit was relatively
inexpensive and readily available. The PIC18F4520 was selected as the primary processor
for the Mercury. It had the necessary features for both Mercury and Titan with the exception of a
MAC . In addition, Microchip offers DSP products that are mostly compatible with code written
for the PIC18. Therefore, when the Titan comes under development, existing Mercury code can be
easily ported to the new processor for new functionality on a shorter software development cycle.
Products offered by other compaines were not as easy to integrate in a short time, nor were they
able to meet the develoment cost profile necessary. Most other processor offerings also required
external crystal oscillation, increasing layout complexity. Becuase of the short development cycle,
this was undesirable.
46

Chapter 5
Embedded Software
5.1 Variable Use
The libraries included with the Micropchip MCC18 Compiler did not include memory management.
Therefore, all variables will declared as externs. The variable declarations can be seen in Listing 5.1.
The variables were then called within functions as shown in Listing 5.2. Using this method prevented
Listing 5.1: Variable Declarations
int hours =0;
int minutes =0;
int seconds =0;
int tempTmr1 ;
int magnitude ;
signed int register x a x i s =0, y a x i s =0, z a x i s =0;
short s t a t e =0;
int nowSecondsA=0, nowSecondsB=0, nowExactA=0, nowExactB=0;
float period =0;
int r a t e =0;
struct LCDseg seg ;
int tempPeriod ;
int tempRate ;
short on=0;
int outSeconds =0;
short resetB =0;
memory conflict. Although any major increase in stored data would prevent this methodology from
succeeding.
47

extern int hours ;
5.2 System Initialization
Listing 5.2: Extern Variable Declarations
The system is initialized using a set of three functions and register assignments. The code can be
seen in Listing 5.3. It first sets the oscillator frequency to the maximum of the chip. Port B is
then set as the digital input port for button use. Finally, the analog to digital, LCD, and timer
initialization functions are called.
OSCCON=0xFF ;
TRISB=0b00100110 ;
LATB|=0 x01 ;
INTCON3=0b11011000 ;
initAD ( ) ;
initLCD ( ) ;
i n i t C l o c k ( ) ;
Timer
Listing 5.3: Main Initialization Code
Timer initialization code can be seen in Listing 5.4. The timer is first initialized to 0xC000 for
the boot-up sequences. The interrupt for the timer is enabled, and the prescaler is set to 1:256.
Finally, the timer itself is started.
Listing 5.4: Timer Initialization Code
void i n i t C l o c k ( ) {
T0CON=0b00000111 ; // p r e s c l e 1:256
INTCON=0b11100000 ; // e n a b l e Timer0 I n t e r r u p t
INTCON2&=0b11111011 ;
RCONbits . IPEN=0b1 ;
TMR0H=0xC0 ;
TMR0L=0x00 ;
T0CON|=0 b10000000 ; // s t a r t
}
LCD
The LCD initialization code is very simple. It can be seen in Listing 5.5. It sets port D as an
output and calls functions to load ARIonto the LCD screen while the device boots. This is held for
48

approximately two seconds until the timer takes over. This is not absolutely necessary but serves
as an eye-catcher for the product.
Listing 5.5: LCD Initialization Code
int initLCD ( ) {
PORTD=0x0 ;
TRISD=0x0 ;
fillLCD ( 0 , 0 , 0 , 0 , 0x77 , 0x50 , 0x10 , 0 ) ;
return 1 ;
}
Analog to Digital Converter
The Analog to Digital Converter initialization code can be seen in Listing 5.6. It sets the wait
time and hold time as short as possible to increase sampling frequency. In addition, it disables
the interrupt. Therefore, the converter will need to be polled. After initialization, the converter is
polled continuously.
void initAD ( ) {
ADCON0=0b00000001 ;
ADCON1=0b00011100 ;
ADCON2=0b00000000 ;
PIR1bits . ADIF=0;
}
5.3 Interface
Listing 5.6: A-to-D Initialization Code
There are digital I/O buttons on the interface. Two of them (start and pause) are enabled as
interrupts. When one is pressed, it enters its interrupt routine. The pause button sets the onbit
to 0; the start button sets the onbit to 1. This enables the output LCD to continue to increment
when the on but is 1, but but only allows the internal counter to increment when it is 0. Thus, the
internal timer is constantly counting, but the external one only updates when the device is in run
mode.
The reset button is not enabled as an interrupt. If enabled as an interrupt, it would be active
upon first press. Instead, each time the timer is incremented, it polls the button. If the reset button
is held for more than 2 seconds, the timer is reset to zero. The code to do this is shown in listing
5.7.
49

5.4 Timer
Listing 5.7: Reset Button Polling
i f ( ! resetB )
resetB=PORTB&0b00100000 ;
i f ( resetB && PORTB&0b00100000 ){
resetB =0; outSeconds =0; on=0;
} else
resetB =0;
The timer system of the pic18f4520 is a relatively simple one. The processor itself has four timers
included within it. Only timer zero is being used at present. The timer counts from 0xE5E0 to
0xFFFF, then interrupts. When this interrupt is thrown, the interruptTimer function is called.
The code for interruptTimer can be see in Listing 5.8. In addition, the function calls the LCD
update functions. If the variable onis set to 1, then the visible timer is incremented. Otherwise,
the internal timer is incremented for stroke rate calculation purposes only. This will be discussed
later.
void incrementTime ( ) {
seconds++;
i f ( on ) outSeconds++;
i f ( outSeconds==60){
}
}
5.5 LCD Access
Listing 5.8: incrementTimer()
outSeconds =0;
minutes++;
i f ( minutes==60){
hours++;
minutes =0;
}
The LCD is accessed via two cascaded driver chips. They accept three signals: load, data, and
clock. Load is simply a latch. Therefore, if load is held high, the driver constantly accepts data.
The two cascaded drivers accept 64 bits serially on the rising edge of the clock. These bits each
correspond to a segment of the display. convertDig()is a function written to convert the numbers
0-9 to their corresponding 7-seg outputs. These values are then shifted sequentially bitwise to the
driver chips using port D.
50

5.6 Stroke Rate Calculation
Figure 5.1: Rate Calculation Finite State Machine
After the A-D conveter is polled three times (once for each axis), calcStrokeRate()is called. This
function is a software finite state machine. An illustration can be seen in figure 5.1. When a rising
endge is seen, the timer value is saved. When two rising edges are seen in a row, the values are
subtracted and divided by 60. This returns the number of stroke periods in one minute. This
is known as the stroke rate. An improvement in accuracy could be achieved if the device used a
peak-finding algorithm instead of a zero-crossing one. Zero crossing algorithms are susceptible to
noise. As a prototype, however, the zero-crossing is easy to implement and fairly reliable.
5.7 Secure Digital
Long term recording of data for later analysis will be performed on the Titan. This will require
outputting data via the SPI port to the SD card itself. There are two options in this case. The
first is to do raw data writes. It will require memory management but be implemented easy on the
microcontroller. The other option it to implement a FAT filesystem. This would allow native reads
and writes on a standard windows computer. From a processor perspective, it would require more
overhead. The FAT apporach will be implemented in the Titan.
5.8 Bluetooth
In addition to a FAT filesystem, Bluetooth would be implemented. Bluetooth modules can be easily
connected to the USART port of the device. It can be connected as a virtual serial port to a host
computer and the data downloaded to the analysis software. This will also be included in the Titan
model.
51

Chapter 6
Non-embedded Software
6.1 MATLAB Platform
MATLAB by TheMathWorks is a numerical computing environment and programming language. It
can plot functions, create graphical user interfaces, and interface with programs in other languages.
MATLAB can also perform math functions numerically or symbolically.
MATLAB’s open graphical user interface (GUI) design enviroment (GUIDE) allows users to
develop GUIs that interface with their code and provide a more user-friendly working environment.
Similary, GUIs are useful because they remove end users from the command line interface of MAT-
LAB and provide an easy way to share code between nonprogrammers. ARI’s software that comes
packaged with the Titan is built on the MATLAB platform using GUIDE.
6.2 Features
Users of ARI’s software are able to import any ASCII character-delimitted text file using the File
menu (this file is populated with voltage values from the accelerometer in the Titan). The software
then plots an abstract acceleration curve and resultant stroke rate graph. The hold function can
be used to plot multiple boats on the same axes using color-coded lines, so the user can compare
performance across the fleet. Similarly, the grid can be turned on and off, and the user can insert
and save comments in the left side of the window. An initial interface was created in Macromedia
Fireworks as seen below in Figure 6.1. A screenshot of the actual GUI can be seen in Figure 6.2.
The code in ARI’s software to read in any ASCII character-delimitted text file can be seen below.
It was obtained from MATLAB Central’s File Exchange.
function RESULT = readbar(filename,msg)
%READBAR Read ASCII delimited file.
% RESULT= READBAR(FILENAME,MESSAGE) reads numeric data from the ASCII
% delimited file FILENAME showing the progress through a progress bar. The
% delimiter in the ASCII file is guessed by READBAR. The MESSAGE will
% be shown in the progress bar, together with the estimated time left (in
% seconds). The result is returned in RESULT.
%
52

Figure 6.1: ARI software mockup
% RESULT= READBAR(FILENAME) shows a standard MESSAGE field where the
% FILENAME and the estimated time left will appear.
%
% The function is suitable for huge ASCII files. A CANCEL button is
% available to interrupt the loading process. In that case an empty
% RESULT is returned
%
% Example:
% X = rand(390625,32); % 100M variable !!!
% save example.txt X -ASCII
% Y = readbar(’example.txt’);
%
% The function READBAR is (estimated) 15% slower than load, but
% provide a method to interrupt the execution and to note the progress of
% loading. It is useless for small files.
% Copyright 2004 Stefano Gianoli, ETH Zurich
% gianoli@chem.ethz.ch
% $Revision: 5.0 $ $Date: 2004/06/15 12:06:46 $
%
53

Figure 6.2: ARI software
% v4: improved compatibility: read files generated with
% save(’filename’,’var1’,...,’-ASCII’)
% v5: bugfix for NxM data where N

% create waitbar
h = waitbar(0,’Calculating the size of the file’,’CreateCancelBtn’,@readbar_cancel_Callback,’In
% open the file
fid = fopen(filename,’r’);
% Make sure it can be opened, otherwise close the waitbar and stop
if fid == -1, close(h), error(’Unable to open file’), end
% get the number of iteration necessary to read all the file: WARNING this
% line must be before the determination of the size of RESULT, since the
% number of iteration is needed
SetIterNum(filename);
% determine the size of the RESULT matrix and initialize it to zero
SizeRESULT = GetSizeResult(fid,h);
if length(SizeRESULT) == 2
RESULT = zeros(SizeRESULT);
nIter = GetIterNum;
else
close(h)
error(’Invalid size for RESULT: check BUFFERSIZE or End Of Line delimiters’)
end
close(h)
h = waitbar(0,msg,’CreateCancelBtn’,@readbar_cancel_Callback,’Interruptible’,’on’,’BusyAction’,
hAxes = findobj(h,’type’,’axes’);
hxLabel = get(hAxes,’xlabel’);
% save the cpu time at the beginning reading iterations
a = cputime;
tmp =[];
waitbar(0,h,msg);
for iIter=1:nIter
% update the waitbar
try
waitbar(iIter/nIter,h);
set(hxLabel,’string’,[int2str((nIter/iIter-1)*(cputime-a)) ’ seconds left’]);
catch
% if the wait bar have been cancelled, interrupt the iterations
if ~ishandle(h), RESULT = []; return, end
end
% retrive the buffer size for the current iteration
s = iGetBufSize(iIter);
% read the file in a char vector
55

B = fread(fid,s,’*uint8’);
% retrive the range indexes for the rows of the RESULT matrice to be
% used to store the converted buffer of char read
r = iGetIndexRow([iIter iIter+1]);
% try to read convert the char in the buffer and store them in RESULT
B = strtrim(B);
try
if any(r) && r(1) ~= r(end)
if length(tmp)>0
tmp = [tmp dataread(’string’,char(B))];
RESULT(1+r(1):r(2),:) = tmp;
tmp =[];
else
RESULT(1+r(1):r(2),:) = dataread(’string’,char(B));
end
else
tmp = [tmp dataread(’string’,char(B))];
end
catch
delete(h);
disp([’On file’ mfilename ’ ==> using dataread’])
error(lasterr)
end
end
delete(h);
%//////////////////////////////////////////////////////////////////////////
% end of main
%//////////////////////////////////////////////////////////////////////////
function SetIterNum (f)
global GBUFFERSIZE GITERATIONNUM % protected variable
% protected max buffer size -> GBUFFERSIZE
% protected number of iteration -> GITERATIONNUM
% get the file size
info = dir(f);
fSize = info.bytes;
% check that the buffer size is in the range 2^12 to 2^20
GBUFFERSIZE = min([max([4096 floor(fSize/50)]) 2^18]);
% get the binary value
binbuf = dec2binvec(GBUFFERSIZE);
% convert to the maximum value multiple of 2
binbuf(1:end-1) = 0;
% return to the decimal value of the maximum buffer size
GBUFFERSIZE = binvec2dec(binbuf);
% get the number of iteration, rounded to next integer
56

GITERATIONNUM = ceil(fSize/GBUFFERSIZE);
function GITERATIONNUM = GetIterNum
global GITERATIONNUM % protected variable
%//////////////////////////////////////////////////////////////////////////
function S = GetSizeResult(f,h)
global GBUFFERSIZE GITERATIONNUM ITEMROWS ITEMBUFFER % protected variable
% max buffer size -> GBUFFERSIZE
% number of iteration -> GITERATIONNUM
% index of the first rows of the RESULT matrix to be filled at the i-th iteration -> ITEMROWS
% buffer size to be read in order to fill the irow of the RESULT matrix at the i-th iteration -
% save the current position
currpos = ftell(f);
% go at the beginning of the file
fseek(f,0,’bof’);
c = 0; SUBITER = 0;
while ~feof(f)
%----------------------------
% AIM: >>find the number of column c1 || any(A==10), break, end;
%----------------------------
% AIM: >>find the index of the first row (ITEMROWS) and the respective
% buffer size (ITEMBUFFER) for the i-th iteration

hAxes = findobj(h,’type’,’axes’);
hxLabel = get(hAxes,’xlabel’);
% for each iteration determine ITEMROWS and ITEMBUFFER
a = cputime;
for i = 1:GITERATIONNUM
try
waitbar(i/GITERATIONNUM,h);
set(hxLabel,’string’,[int2str((GITERATIONNUM/i-1)*(cputime-a)) ’ seconds left’]);
catch
% if the wait bar have been cancelled, interrupt the iterations
if ~ishandle(h), RESULT = []; return, end
end
end
% read in the buffer char vector A the file f
[A,count] = fread(f,GBUFFERSIZE,’*uint8’);
% find the number of Carriage return character into the buffer A
tempCR = (A == 10);
% the number of ITEMROWS is equal to the sum of the CR
ITEMROWS(i) = sum(tempCR);
% the number of the bytes from the end of the char buffer, excluding
% the CR, carriage return, i.e. +1, as a negative number, to be added
% to the buffer at the i-th iteration
ITEMBUFFER(i) = FindLastOneNeg(tempCR)+1;
% set the file position to the end of the just after the last CR, i.e.
% subtract the value already read after the last CR.
fseek(f,ITEMBUFFER(i),’cof’);
% set the length char buffer ITEMBUFFER to the number of bytes to read in the i-th
% iteration, to stop just after the last CR, i.e. position of last CR
% from the end of the file, as a negative number, plus bytes actually
% read
ITEMBUFFER(i) = ITEMBUFFER(i) + count;
ITEMROWS = [0; cumsum(ITEMROWS)];
% restore the current position
fseek(f,currpos,’bof’);
% return the size the matrix RESULT in S
S = [ITEMROWS(end) c];
%//////////////////////////////////////////////////////////////////////////
function v = iGetIndexRow(i)
global ITEMROWS
v = ITEMROWS(i);
58

%//////////////////////////////////////////////////////////////////////////
function v = iGetBufSize(i)
global ITEMBUFFER
v = ITEMBUFFER(i);
%//////////////////////////////////////////////////////////////////////////
function p = FindLastOneNeg(A)
global GBUFFERSIZE;
% find the index of last value in A that is equal to 1, starting from the
% bottom of the file, and returning a negative value, that is the number
% that must be subtracted to the length of the file in order get the last
% CR. that correspond to find the max index, but since this could be time
% consuming for a huge vector, it is possible to find this value in the
% latest 2048 characters (hopefully thre are no line bigger than 2048 char
% in a ACSII file). A check is done in order to avoid an error message for
% files smaller than 2048 bytes
p = max(find(A(end-min(end,GBUFFERSIZE-1)+1:end)))-min(length(A),GBUFFERSIZE);
if isempty(p)
p = -1;
end
%//////////////////////////////////////////////////////////////////////////
function m = GetStdMsg(f)
% get the standard message (containing the file name)
[v,w,z] = fileparts(f);
w = strrep(w,’\’,’\\’);
w = strrep(w,’_’,’\_’);
m = [’Loading ’’’ w z ’’’’];
%//////////////////////////////////////////////////////////////////////////
function X = strtrim(X)
tmp = 0;
X = uint8(strrep(char(X’),sprintf(’\t\n ’),sprintf(’\n’)))’;
while length(X) ~= tmp
tmp = length(X);
X = uint8(strrep(char(X’),sprintf(’ ’),sprintf(’ ’)))’;
end
%//////////////////////////////////////////////////////////////////////////
function readbar_cancel_Callback(obj, eventdata)
% Callback function that delete the waitbar once the user press the cancel
% button
handles = guihandles(obj);
hfields = fields(handles);
59

delete(handles.(hfields{1}))
6.3 Stroke Rate Calculation
The code for ARI’s stroke rate calculation can be seen below.
clear all
% Read input text file and put into array Y
Y = readbar(’example.txt’);
% Plot abstract acceleration
[m,n] = size(Y);
fs = 100;
ts = (1/fs);
t = 0:ts:((m-1)*ts);
subplot(2,1,1)
plot(t,Y)
title(’Acceleration’)
xlabel(’(s)’)
grid on
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Calculate stroke rate
[zero] = 0; % vector of time values
i = 1;
for i=1:(m-1)
a = Y(i);
b = Y(i+1);
mul_test = a*b;
if (Y(i) == 0)
zero = [zero i];
elseif (mul_test < 0)
zero = [zero i];
end
end
zero = zero(2:length(zero));
%[zero]
[rate] = 0; % vector of stroke rate values
k = 1;
for k=1:2:(length(zero)-2)
SR = (1/(zero(k+2) - zero(k)));
indices = (zero(k+2) - zero(k));
rate = [rate SR*(ones(1,indices))];
60

end
while (length(rate) ~= m)
rate = [rate 0];
end
%[rate]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Plot stroke rate
subplot(2,1,2)
plot(t,rate)
title(’Stroke Rate’)
xlabel(’(s)’)
grid on
6.4 Future
In the future, ARI’s software could receive its input data file via Secure Digital or Bluetooth
interface from the Titan. Ideally, a coach could sit the Titan next to his or her laptop after a
regatta and transmit the acceleration data file wirelessly via Bluetooth. He or she could then open
ARI’s software and import the data file to analyze his or her team’s performance.
61

Chapter 7
Case Design
7.1 Overview
ARI’s unique case design is the product of a multidisciplinary collaboration with the NCSU College
of Design. This collaboration was made possible through the cooperation of Mr. Percy Hooper,
Associate Professor of Industrial Design. ARI was fortunate to recruit not one, but two gifted
industrial designers, Sam Dirani and Michael Caston to approach the challenge of designing a
waterproof case with a modular battery compartment and distinctive appearance. Mr. Dirani and
Mr. Caston worked closely with the technical design team to develop a case design that would
interface effectively with the internal electrical components. The result is an impressive balance of
form and function.
7.2 Inspiration
The industrial design team sought to create a unique shape that would stand apart from the
traditional cylindrical shape of ARI’s competitors. They looked to nature for inspiration and found
the “water boatman“, an insect with a sleak appearance that seems to have tiny “oars“ for legs.
The shape of this insect was used as inspiration for the sweeping curves and elongated top of the
prototype’s case. Additionally, the sweeping lines accentuated the traits of speed, a teardrop of
water, and the shape of the ARI logo.
The industrial designers also strove to integrate a handle into the case design, instead of using
a plastic or fabric strap that is seen in the competition’s offerings. The result is a subtle handle
recessed into the underside of the faceplate. The handle recess was contoured to fit the shape of a
hand, with depth based on the mean finger size of a coxswain.
With the interface design, rubber molded buttons were chosen for simplicity of operation, even
with gloves. Instead of using discrete “up“ and “down“ buttons for volume control, a vertically
mounted scroll wheel was used with a groove surrounding it to allow one’s finger to comfortably
adjust the volume wheel. Additionally, the groove also allows the wheel to sit below the surface of
the curve, preventing the wheel from being sheared off in case of a fall.
62

7.3 Modular Battery
A repeated complaint of the current offerings by ARI’s competitors stem from the battery. The
battery in these units are not user removable. They provide approximately 3 to 5 hours of charge,
and after two or three years are not able to hold charge, despite a recent recharge. At this point, the
coxbox must either be replaced or sent back to the manufacturer for a costly battery replacement.
ARI decided to design a modular battery pack which would allow teams to easily recharge and
purchase additional batteries for their units. By choosing the AA battery size form factor, the
Mercury and Titan units are unique in the feature that if a coxswain forgets to charge the unit
before a regatta, he or she has the option of using standard alkaline AA batteries found in any
gas station or convenience store in the vicinity as a temporary backup. Though such low capacity
batteries may not last an entire regatta, it is better than having a dead coxbox unit.
The modular battery unit is found in the bottom portion of the cylindrical case, and connects
to the rest of the unit via a bayonet-style mount. Production models include a rubber o–ring and
compression seal to waterproof this connection.
7.4 Packaging
A significant challenge in the development of the Mercury prototype was packaging the electronics
into the case. The case had to fit existing mounting cups which are 4 inches in diameter. At
the same time, the electronics that allow for manageable prototyping and assembly are larger than
what production units would have. Frequent meetings between the industrial designers and the ARI
engineering team were had to ensure both parties were on the same page regarding the packaging
of the electronics into the case.
The A1 board was to be sandwiched between the top faceplate and the case’s middle section.
The B1 and B2 boards were connected together using metal standoffs, and the B1 board was
attached to the case’s middle section through the use of two plastic blocks. The B1 board was
carefully placed to allow the microphone BNC and Amphenol 44 speaker connectors to protrude
into the middle section of the case. The metal standoffs were chosen to keep the bottom face of the
B2 board approximately 1/8“ inch from the bottom edge of the case’s middle section. Three small
columns, with a diameter of approximately 3/8“ inch, were glued along the inside diameter of the
middle section. Once all three boards were in place, long screws were inserted through the columns,
the A1 board, and the corresponding pillars in the faceplate to hold the entire case together.
7.5 Production
The industrial designers had at their disposal a state-of-the-art ABS plastic printer that can print
3-dimensional models. Based upon the initial sketches and meetings between the designers and
the engineers, Dirani created a three dimensional model of the Mercury using the Alias computer
program. This model was sent to the ABS plastic printer. This first run encountered some problems,
but were instructive for initial packaging and solving future assembly concerns. The computer model
was refined further and a second run was printed. The model was printed in three separate pieces:
the modular battery portion, the middle portion of the case, and the top faceplate. The faceplate
had a unique shape cut out to allow the displays to be seen. A thin piece of acrylic was cut, heat
shaped, and formed to match the graceful curve of the faceplate.
63

After assembly holes were drilled and cut, the ABS plastic was sanded and primed, and a
glossy coat of paint was applied. The modular battery compartment and the bottom portion of the
middle case had a rubber-like paint applied to it to provide better grip and feel. This is important
for ensuring the Mercury and Titan sit firmly in the mounting cup commonly found in shells.
For production models, Dirani stated that high impact ABS plastic would be used, with rubber
co-molded into the case design.
64

Part IV
Appendices
65

Appendix A
Financials
66

Advanced Rowing Instruments
Projected Operating Statements & Cash Flows For Years
Ending December 31 (in $000'S)
Y1 Y2 Y3 Y4
Revenue:
Projected Annual Growth Rate
License & Royalty fees 1 3 7 9
Software Sales 1 4 7 10
Software Maintenance 0 0 1 1
Hardware Sales 50 345 552 972
Hardware Maintenance 0 18 17 41
Total Revenue
Operating Expenses:
52 370 584 1,033
Research & Development 60 56 56 56
Sales & Marketing 16 49 75 107
General Administation 113 105 105 105
Manufacturing 0 0 0 0
Total Variable Costs 25 66 105 156
Total Operating Expenses 214 275 341 424
Profit (Loss) from Operations Operations (162) (162) 94 94 244 609
Tax Provision 0 38 97 243
Net Income (Loss)
Cash Flow Analysis:
(162) 57 146 365
Contract and/or Equity Financing 250 500 1,500 800
Capital Expenditures (60) (95) (140) (165)
Change in Cash 28 462 1,506 1,000
Beginning Cash 50 78 540 2,046
Ending Cash 78 540 2,046 3,046
Variable Costs by Year: Y1 Y2 Y3 Y4
Trade Shows 8 15 20 24
Sales & Marketing Collateral 6 10 16 23
Legal Fees 2 1 3 4
Travel & Entertainment 3 8 10 15
COGS (Cost of Goods Sold) 6 32 56 90
Total Variable Costs: 25 66 105 156
Capital Purchases:
Employee Related 10 15 20 25
General 50 80 120 140
Total Capital Purchases: 60 95 140 165
Software Markup 50% 50% 60% 70%
Software Maintenance as % of Sale 5% 5% 10% 10%
Software Maintenance Increase 0% 10% 10% 10%
Hardware Markup 60% 80% 100% 120%
Hardware Maintenance as % of Sale 10% 5% 3% 4%
Hardware Maintenance Increase 0% 0% 3% 5% 5%

Appendix B
Project Timeline
68

March 2007
Sunday Monday Tuesday Wednesday Thursday Friday Saturday
1 2 3
Get Patent the searching, Provisional 18 patent daysAp
4 5 6 7 8 9 10
Get the Provisional Patent searching, patent Silicon Application, 18 ValleyTrip, days 10 days? 5 days?
Photography slideshow, 40 days?
11 12 13 14 15 16 17
Get the Provisional patent Application, 10 days?
Patent searching, Permanent Flash 18 presentation, Positions, days 136.13 33 days? days?
Video presentation, 34 days?
Photography slideshow, 40 days?
18 19 20 21 22 23 24
Permanent Flash Patent presentation, Positions, searching, 136.13 18 33 days days? days? Document "Fence" patents, 26 days?
Video presentation, 34 days? Licensing, 26 days?
Photography slideshow, 40 days?Plans to acquire other intellectual property, 26 days?
"Design around patents" to avoid our infringement on other patents, 26 days?
25 26 27 28 29 30 31
Patent searching, 18 days Complete Document Permanent Flash non-PPAs presentation, "Fence" Positions, (trademark, patents, 136.13 33 days? 26 days? patent (1-3 technologies)), 11 days?
Video Licensing, presentation, 26 days? 34 days?
Plans to acquire Photography other intellectual slideshow, property, 40 days? 26 days?
"Design around patents" to avoid our infringement on other patents, 26 days?
Getting Boat, 17 days?

Overflow Tasks
ID Name Start Finish
18 Design & order t-shirts Fri 2/23/07 Fri 3/9/07
19 Rotational VE work period Mon 2/12/07 Sun 3/11/07
23 Identify Key Business Goals Fri 3/2/07 Fri 3/2/07
25 Mine, analyze, and organize intellectual assets (patents, trademark, trade secrets, copyrights) Fri 3/2/07 Fri 3/2/07
31 Draft and Get Non-Disclosure Agreement Wed 2/21/07 Fri 3/2/07
39 Interface Design Fri 2/23/07 Wed 3/14/07
44 Display interfacing Wed 2/21/07 Wed 3/21/07
52 Test Audio Fri 3/2/07 Fri 3/2/07
62 Discuss connections with Case Design Fri 2/23/07 Mon 3/19/07
9 Get TV and Mounting Mon 3/5/07 Mon 4/23/07
53 Interface with existing microphone & speakers Fri 3/9/07 Tue 3/13/07
65 Package accelerometer, laptop, and multimeter in waterproof housing Fri 3/9/07 Mon 3/12/07
70 Final Documentation Sat 3/10/07 Fri 4/27/07
15 Powerpoint Mon 3/12/07 Mon 4/2/07
47 Signal Processing, Display driving code Mon 3/12/07 Wed 4/4/07
66 Row on the water and collect real-world data Mon 3/12/07 Sat 3/17/07
40 Prototype 1 Fri 3/23/07 Fri 3/23/07
45 ADC Interfacing, testing with function generator Mon 3/19/07 Fri 3/23/07
55 Final Layout for the Audio Amp Circuit Fri 3/23/07 Fri 3/23/07
54 Connect to interface Mon 3/19/07 Mon 3/19/07
41 Prototype 2 Mon 3/26/07 Wed 4/4/07
46 Accelerometer Interfacing Fri 3/30/07 Fri 3/30/07
49 Memory Card Interfacing Mon 3/26/07 Wed 4/11/07
50 Bluetooth Interfacing Mon 3/26/07 Wed 4/11/07

April 2007
Sunday Monday Tuesday Wednesday Thursday Friday Saturday
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
Find and meet with a patent lawyer for 3rd party review, 6 days?
Getting Boat, 17 days?
22 23 24 25 26 27 28
Document Flash presentation, "Fence" patents, 33 days? 26 Permanent days? Positions, 136.13 days?
Licensing, Video 26 days? presentation, 34 days?
Plans to acquire other intellectual property, Photography 26 days? slideshow, 40 days?
"Design around patents" Technical to avoid presentation, our infringement 7 days? on other patents, 26 days?
29 30
Complete non-PPAs Document Permanent Flash (trademark, presentation, "Fence" Positions, patent patents, 136.13 33 (1-3 days? 26 technologies)), days? 11 days?
Video Licensing, presentation, 26 days? 34 days?
Plans to acquire Photography other intellectual slideshow, property, 40 days? 26 days?
"Design around patents" to avoid our infringement on other patents, 26 days?
Getting Boat, 17 days?
Document Permanent Flash presentation, "Fence" Positions, Find patents, and 136.13 33 meet days? 26 with days? a patent lawyer for 3rd party review, 6 days?
Video Licensing, presentation, 26 days? 34 days?
Plans to acquire Photography other intellectual slideshow, property, 40 days? 26 days?
"Design around patents" to avoid our infringement on other patents, 26 days?
Getting Boat, 17 days?
Document Permanent Flash presentation, "Fence" Positions, patents, 136.13 33 days? 26 days?
Video Licensing, presentation, 26 days? 34 days?
Plans to acquire Photography other intellectual slideshow, property, 40 days? 26 days?
"Design around patents" to avoid our Technical infringement presentation, on other patents, 7 days? 26 days?
System Architecture Poster, 5 days?

Overflow Tasks
ID Name Start Finish
8 Mount for Boat Mon 4/2/07 Mon 4/23/07
9 Get TV and Mounting Mon 3/5/07 Mon 4/23/07
11 ARI & EEP Banners Mon 4/2/07 Mon 4/16/07
42 Final Case Manufacturing Thu 4/5/07 Thu 4/12/07
15 Powerpoint Mon 3/12/07 Mon 4/2/07
41 Prototype 2 Mon 3/26/07 Wed 4/4/07
47 Signal Processing, Display driving code Mon 3/12/07 Wed 4/4/07
48 Final Microprocessor Chip layout for Vx Wed 4/4/07 Sat 4/7/07
49 Memory Card Interfacing Mon 3/26/07 Wed 4/11/07
50 Bluetooth Interfacing Mon 3/26/07 Wed 4/11/07
70 Final Documentation Sat 3/10/07 Fri 4/27/07
10 Get Monitors Mon 4/9/07 Mon 4/23/07
12 Trifold Brochure Tue 4/17/07 Tue 4/24/07
13 One sheet brochure Tue 4/17/07 Tue 4/24/07
68 Package all circuit boards Sun 4/22/07 Tue 4/24/07
69 Test all components Tue 4/24/07 Wed 4/25/07

Appendix C
Design Day Plans
73

Design Day
Overview
Electrical and Computer Engineering (ECE) Senior Design Day is held every semester on the
campus of North Carolina State University, showcasing that semester’s senior design projects. This
semester, Senior Design Day was held on April 27, 2007 in the McKimmon Center for Extension
and Continuing Education in Raleigh, NC. ARI, along with the other Engineering Entrepreneur
Program (EEP) participants, set up a booth in the back left corner of Room 1.
Booth Setup
ARI decided to go big this year with the booth at Senior Design Day. A 4-man rowing shell was
borrowed from the NC State Club Rowing Team, and a wooden boat mount was constructed to
hang the shell behind the two tables of the booth. The shell was approximately 40 feet, and it was
hung 5 feet above the ground on its side by bright orange straps, so Design Day attendees could
see inside the boat. Inside both ends of the shell sat two promotional posters of the Mercury in
poster frames.
Two 6 foot by 3 foot tables were lined up in front of the boat mount, and white table clothes
covered their tops. A black banner with the ARI logo also hung over the front of the center of the
two tables. Two 20 inch widescreen monitors flanked a 42 inch LCD rear-projection television in
the center of the two tables. The Mercury prototype, along with an exploded view of the prototype
and foam model of the casing, were layed out across the left side of the booth. A Nielsen-Kellerman
Cox-Box was also shown beside the Mercury prototype to show the original product that ARI set
out to improve. A picture of the front view of the booth at Design Day is show in Figure C.1. A
picture of the side view of the booth at Design Day is show in Figure C.2.
Figure C.1: ARI Design Day Booth — Front View
74

Marketing
Figure C.2: ARI Design Day Booth — Side View
Two promotional posters of the Mercury and a one page handout on ARI and its products were
used at Senior Design Day. The two posters are shown below in Figures C.3 and C.4. The one
page handout is show in Figure C.5.
Demo
As Senior Design Day attendees came up to the booth, they were greeted by picture slideshows on
the two widescreen monitors at each end of the booth. These pictures consisted of the ARI team
working in the lab, the ARI team at Lake Wheeler checking out a rowing shell, and 3-dimensional
Alias models of the Mercury prototype. The center LCD television ran an animated screensaver of a
3-dimensional ARI logo. When attendees noted their interest in ARI’s products, a click of the mouse
turned the screensaver into a technical prototyping presentation with music in the background. As
soon as the presentation finished, the animated screensaver reappeared.
Attendees were also able test the timing and audio amplification features of the Mercury prototype.
The Mercury was plugged into the existing speakers on the shell hanging behind the booth,
and attendees were able to speak into a coxswain microphone and hear their amplified voice. Similarly,
attendees were able to start, stop, and reset the race timer on the prototype.
75

ROWING
REVOLUTIONIZED
MERCURY
76
Figure C.3: ARI Design Day Poster 1

MERCURY
DATA-DRIVEN
DECISIONS
77
Figure C.4: ARI Design Day Poster 2

Figure C.5: ARI Design Day Handout
78

Appendix D
Functional Specifications
79

Advanced Rowing Instruments
Editor: Saket Vora, CEO
Author(s): Saket Vora, Greg Mulholland
Version: 2.0
Last Edited: May 2, 2007
Created on: October 16, 2006
Mercury and Titan
Functional Specifications
Page 1 of 9

Appendix E
Virtual Employee Job Descriptions
81

Advanced Rowing Instruments
Internal Document -- Confidential
Job Descriptions for Virtual Employees
We plan to target having six to eight virtual employees (VE) join our team. The following will provide a
description of each VE’s job responsibilities.
It is important to realize that being a start-up company, not a single member of the team will only do one
particular task. The VPs will not just manage but tangibly contribute to product development, the VEs will
be performing tasks in both technical and business related areas.
We are implementing four week rotation scheme so that VEs can see all aspects of our company. In the
beginning of March, they will fall into a defined role. In the final two weeks, nearly all focus will be shifted
to product development and finalization.
However, it is important for each employee to be responsible for a particular aspect of the business or
product. This not only gives an employee a sense of ownership, but also allows a direct way of
performance evaluation and accountabilities.
Business Related
1) Public Relations
One VE will be assigned to public relations media presence and will be helped by the VP Business
Strategy and the CEO. This VE will be responsible for the following:
• Maintaining the website
• Aiding in the creation of promotional materials such as example advertisements, design day video
& photos.
• Assisting with customer and business surveys
2) Intellectual Property
One VE will be assigned to helping with intellectual property and licensing issues. This VE will also assist
in customer surveys, financial analysis, and market studies. Working closely with the VP of Business
Strategy and CEO, a provisional patent application will be created and submitted, and licensing deals will
be formulated.
• Perform patent and copyright searches
• Help complete provisional patent application
• Aid in analyzing the survey and market data with the VP Business Strategy and the CEO.
3

Advanced Rowing Instruments
Internal Document -- Confidential
Product Development Related
The bulk of the project will be devoted to product development. Again, the goal of these task designations
is to assign areas of primary responsibility for each VE.
3) Audio and Sensors
One VE will be assigned to the audio amplifier and accelerometer portion of the circuit design. This VE
will work closely with the CEO and VP Hardware. This VE will be responsible for the following:
• Testing and documenting of the accelerometers and audio amplifier.
• Aid in designing volume control of audio amplifier and interface design
• Work with the microprocessor groups to ensure functionality with the ADCs.
• Research other sensors for use in our products.
4) Case & Interface
One VE will be assigned to the case design & interface component of the product. This person will
interact with industrial design students and materials. The VE will work closely with the VP Hardware and
the President. This VE will be responsible for the following:
• Designing and developing the interface for the case
• Construction of the casing
• Aiding in the assembling of the device
• Aiding in designing and implementing the user interface components
5) Power & Test
One VE will be responsibility for the power supply (battery) portion of the product as well as to monitor
and document the overall testing of the product through all phases of development. This VE will work
closely with the VP Hardware, President, and the Case & Interface members. This VE will be responsible
for the following:
• Researching, and developing the battery and power management of the product
• Testing of the battery and monitoring overall testing
• Work with the case & interface VE to design and implement swivel-lock system for an
interchangeable battery.
6) Microprocessor
One VE will be assigned to the processors portion of the design. This is a large portion of the design and
will involve working closely with the VP Hardware and President. Embedded system code will be written
for signal processing and data functions. Work will also be done involving Bluetooth, storage, and display
functionality.
• Researching, designing, and testing of the microprocessors
• Testing of the processors
• Working with other members on Bluetooth, storage, and display functionality
7) Bluetooth & Storage
One VE will be assigned to handle the Bluetooth and storage related components of the design.
Bluetooth will be used to wirelessly transfer the recorded accelerometer data to a PC and, time
4

Advanced Rowing Instruments
Internal Document -- Confidential
permitting, to allow the use of a Bluetooth headset for the coxswain. The storage part involves storing the
accelerometer to an on-board or removable memory device, possibly a Secure Digital, Compact Flash, or
USB stick.
• Developing and testing of the Bluetooth module
• Developing and testing of the memory storage unit
• Aiding in the implementation of the processors in the circuit
8) Display
One VE will be assigned to primarily aid in the development of the embedded software that will run on the
processors contained inside the instrument. This person will be assigned certain functions to implement in
code. This VE will work closely with the VP Software and VP Hardware. This VE will be responsible for
the following:
• Developing and testing of the display unit
• Aid in the user interface design of the case
• Work with the microprocessor members to ensure functionality
5

Appendix F
Call Logs
85

Advanced Rowing Instruments
Internal Document -- Confidential
Company Calls for Market Research
NIELSEN-KELLERMAN
Amy Winner
Advertising and Public Relations
800-784-4221
awinner@nkhome.com
Nov 03
Amy Winner is out of the office until November 15 th . Left a message on voicemail, sent email to follow up
with her.
Got a call back from her. She said she couldn’t reveal any sort of company data. Gave me the contact for
US Rowing.
VESPOLI
Julio Viyella
Southeast Sales Representative
203-843-1491
southeast@vespoli.com
Amanda Kukla
MidAtlantic region
203-675-4409
midatlantic@vespoli.com
Teresa Strickland
Sales & Service
203-773-0311
Nov 3
Kukla – left a message.
Strickland – was forwarded to voicemail of Mike Vespoli (yeah, the CEO, Owner), and left a message with
him.
Julio Viyella – was in Chattanooga at a regatta, will call back Monday mid-morning.
- Got a call back from Kukla – says Vespoli doesn’t sell NK equipment.
WINTECH RACING
Brendan M. Crotty
Mid-Atlantic Sales Representative
1200 Belleview Blvd, #B2
Alexandria, VA 22307
Phone: 203-868-5779
bcrotty@wintechracing.com
Management team – Marketing

Advanced Rowing Instruments
Internal Document -- Confidential
Lynda Confessore
Nov 3
Brendan – left a message
HIDDEN CREEK SPORTS
Rowing retailer, Pennsylvania
General Inquiries: info@hiddencreekspots.com
866-901-0901
Spoke with Ann Williams
They have been in business only since April 2006
So far, they have sold 10 CoxBox and 10 CoxVox Box
Expects another demand in February
Estimates to sell 30 of each in a year.
CHESAPEAKE ROWING
Rock Hall, MD
rowing@dmv.com
410-639-7172
Tom McGlenn – the guy to talk to, left name and number will call back
LAKE UNION CREW
Competitive Crew team in Seattle Washington, over 300 athletes
206-860-4199
HARVARD CREW TEAM
harvcrew@hcs.harvard.edu
DURHAM BOAT COMPANY
Durham, NH
603-659-7575
HUDSON BOAT WORKS
Stephen Ross, Sales and Service
519-473-9864
Jeff McIntyre, Technical Advisor & Sales and Service
Main Office
Craig McAllister, Marketing Manager
519-473-9864

Advanced Rowing Instruments
Internal Document -- Confidential
Nov 03
Spoke to someone there, they don’t have records (without having to go through every single file) on how
many of the units are sold. They said that NK sends them the boxes, then they install them in the boat if
the customer wants it.
REGATTA SPORT
www.regattasport.com
Head Office
Phone: (North America) 1.800-567-2739
Canadian hQ
Left a message
RESOLUTE RACING SHELLS
www.resoluteracing.com
Phone: 401-253-7384
Sue Hinkley,
Probably 90% have the wiring installed in the boats, (120 boats a year)
What we usually do is install the wiring and speakers and seat magnet, (buy these three from Nielsen-
Kellerman..use the retail prices, but then include an installation price).
Usually teams have the CoxBox or CoxVox
SpeedCoach 10-15% of the boats will also order the speedcoach (include the impeller).
25 maybe at the highest
150 boats a year, Increasing 30% a year
especially high school, and girls in particular (because of scholarships) (because of title IX)
regions – areas around NJ and Philadelphia, new clubs all the time
also down in florida, and S.C. and Virginia, “booming”
West coast already has a lot, not sure how much there are moving
www.row2k.com.
Rowers Almanac – better idea of numbers. They made something called a bible – has a listing of all the
crews in the world and contacts for most of them. Telephone – 301-520-8066
info@rowersalamanac.com www.rowersalamanac.com
Vespoli makes 500 boats a year
Seattle company (foupe? Didn’t catch name) , then Durham boats in NH, really small
Frank Biller @ nkhome.com -� sales manager for rowing products.
Rowing News – tripp davis.
Nov 08

Advanced Rowing Instruments
Internal Document -- Confidential
Talked
again with
Ms. Hinkley about the Rowers
Almanac. She said that they have an extra copy around and would mail it to me. However, the National
Survey is a special thing that the Rowers Almanac did last year.
The person to talk to at Rowers Almanac:
Karen (maiden) Solem (now) Derringer
301-520-8066
ROWERS ALMANAC
301-520-8066
- try to talk to Karen Derringer
- mention Sue Hinkley’s name
- ask about the rowers almanac
- ask about national survey
-- got the answering machine to Karen Derringer, left a message.
Nov 14
Called Rower’s Almanac
- got the answering machine to Ms. Derringer and left a message.
Nov 20
Called Rower’s Almanac
- got the answering machine to Ms. Derringer and left a message.

Appendix G
Schematics & PCB
90

Appendix H
Component Datasheets
(Selected Pages)
96

� Advanced Multibus Architecture With Three
Separate 16-Bit Data Memory Buses and
One Program Memory Bus
� 40-Bit Arithmetic Logic Unit (ALU)
Including a 40-Bit Barrel Shifter and Two
Independent 40-Bit Accumulators
� 17- × 17-Bit Parallel Multiplier Coupled to a
40-Bit Dedicated Adder for Non-Pipelined
Single-Cycle Multiply/Accumulate (MAC)
Operation
� Compare, Select, and Store Unit (CSSU) for
the Add/Compare Selection of the Viterbi
Operator
� Exponent Encoder to Compute an
Exponent Value of a 40-Bit Accumulator
Value in a Single Cycle
� Two Address Generators With Eight
Auxiliary Registers and Two Auxiliary
Register Arithmetic Units (ARAUs)
� Data Bus With a Bus Holder Feature
� Address Bus With a Bus Holder Feature
� Extended Addressing Mode for 8M × 16-Bit
Maximum Addressable External Program
Space
� 192K × 16-Bit Maximum Addressable
Memory Space (64K Words Program,
64K Words Data, and 64K Words I/O)
� On-Chip ROM with Some Configurable to
Program/Data Memory
� Dual-Access On-Chip RAM
� Single-Access On-Chip RAM
� Single-Instruction Repeat and
Block-Repeat Operations for Program Code
� Block-Memory-Move Instructions for Better
Program and Data Management
� Instructions With a 32-Bit Long Word
Operand
� Instructions With Two- or Three-Operand
Reads
POST OFFICE BOX 1443 • HOUSTON, TEXAS 77251−1443
�����������
����������� ������� ������ ���������
SPRS078G − SEPTEMBER 1998 − REVISED OCTOBER 2004
� Arithmetic Instructions With Parallel Store
and Parallel Load
� Conditional Store Instructions
� Fast Return From Interrupt
� On-Chip Peripherals
− Software-Programmable Wait-State
Generator and Programmable Bank
Switching
− On-Chip Phase-Locked Loop (PLL) Clock
Generator With Internal Oscillator or
External Clock Source
− Time-Division Multiplexed (TDM) Serial
Port
− Buffered Serial Port (BSP)
− 8-Bit Parallel Host Port Interface (HPI)
− One 16-Bit Timer
− External-Input/Output (XIO) Off Control
to Disable the External Data Bus,
Address Bus and Control Signals
� Power Consumption Control With IDLE1,
IDLE2, and IDLE3 Instructions With
Power-Down Modes
� CLKOUT Off Control to Disable CLKOUT
� On-Chip Scan-Based Emulation Logic,
IEEE Std 1149.1 † (JTAG) Boundary Scan
Logic
� 12.5-ns Single-Cycle Fixed-Point
Instruction Execution Time (80 MIPS) for
3.3-V Power Supply)
� 10-ns Single-Cycle Fixed-Point Instruction
Execution Time (100 MIPS) for 3.3-V Power
Supply (2.5-V Core)
� 8.3-ns Single-Cycle Fixed-Point Instruction
Execution Time (120 MIPS) for 3.3-V Power
Supply (2.5-V Core)
� Available in a 144-Pin Plastic Thin Quad
Flatpack (TQFP) (PGE Suffix) and a 144-Pin
Ball Grid Array (BGA) (GGU Suffix)
Please be aware that an important notice concerning availability, standard warranty, and use in critical applications of
Texas Instruments semiconductor s and disclaimers thereto appears at the end of this data sheet.
† IEEE Standard 1149.1-1990 Standard-Test-Access Port and Boundary Scan Architecture.
All trademarks are the property of their respective owners.
���������� ���� ����������� �� ������� �� �� ����������� �����
�������� ������� �� �������������� ��� ��� ����� �� ����� �����������
�������� ��������� ���������� ���������� ���� ��� ����������� �������
������� �� ��� �����������
Copyright © 2004, Texas Instruments Incorporated
1

1 TMS320VC5509A Features
� High-Performance, Low-Power, Fixed-Point
TMS320C55x Digital Signal Processor
− 9.26-, 6.95-, 5-ns Instruction Cycle Time
− 108-, 144-, 200-MHz Clock Rate
− One/Two Instruction(s) Executed per
Cycle
− Dual Multipliers [Up to 400 Million
Multiply-Accumulates per Second
(MMACS)]
− Two Arithmetic/Logic Units (ALUs)
− Three Internal Data/Operand Read Buses
and Two Internal Data/Operand Write
Buses
� 128K x 16-Bit On-Chip RAM, Composed of:
− 64K Bytes of Dual-Access RAM (DARAM)
8 Blocks of 4K × 16-Bit
− 192K Bytes of Single-Access RAM
(SARAM) 24 Blocks of 4K × 16-Bit
� 64K Bytes of One-Wait-State On-Chip ROM
(32K × 16-Bit)
� 8M × 16-Bit Maximum Addressable External
Memory Space (Synchronous DRAM)
� 16-Bit External Parallel Bus Memory
Supporting Either:
− External Memory Interface (EMIF) With
GPIO Capabilities and Glueless Interface
to:
− Asynchronous Static RAM (SRAM)
− Asynchronous EPROM
− Synchronous DRAM (SDRAM)
− 16-Bit Parallel Enhanced Host-Port
Interface (EHPI) With GPIO Capabilities
� Programmable Low-Power Control of Six
Device Functional Domains
� On-Chip Scan-Based Emulation Logic
All trademarks are the property of their respective owners.
TMS320C55x and MicroStar BGA are trademarks of Texas Instruments.
† IEEE Standard 1149.1-1990 Standard-Test-Access Port and Boundary Scan Architecture.
November 2002 − Revised February 2007 SPRS205I
Features
� On-Chip Peripherals
− Two 20-Bit Timers
− Watchdog Timer
− Six-Channel Direct Memory Access
(DMA) Controller
− Three Serial Ports Supporting a
Combination of:
− Up to 3 Multichannel Buffered Serial
Ports (McBSPs)
− Up to 2 MultiMedia/Secure Digital Card
Interfaces
− Programmable Phase-Locked Loop
Clock Generator
− Seven (LQFP) or Eight (BGA) General-
Purpose I/O (GPIO) Pins and a General-
Purpose Output Pin (XF)
− USB Full-Speed (12 Mbps) Slave Port
Supporting Bulk, Interrupt and
Isochronous Transfers
− Inter-Integrated Circuit (I 2 C) Multi-Master
and Slave Interface
− Real-Time Clock (RTC) With Crystal
Input, Separate Clock Domain, Separate
Power Supply
− 4-Channel (BGA) or 2-Channel (LQFP)
10-Bit Successive Approximation A/D
� IEEE Std 1149.1 † (JTAG) Boundary Scan
Logic
� Packages:
− 144-Terminal Low-Profile Quad Flatpack
(LQFP) (PGE Suffix)
− 179-Terminal MicroStar BGA (Ball Grid
Array) (GHH Suffix)
− 179-Terminal Lead-Free MicroStar BGA
(Ball Grid Array) (ZHH Suffix)
� 1.2-V Core (108 MHz), 2.7-V – 3.6-V I/Os
� 1.35-V Core (144 MHz), 2.7-V – 3.6-V I/Os
� 1.6-V Core (200 MHz), 2.7-V – 3.6-V I/Os
13

PIC18F2420/2520/4420/4520
28/40/44-Pin Enhanced Flash Microcontrollers with
10-Bit A/D and nanoWatt Technology
Power Managed Modes:
Run: CPU on, peripherals on
Idle: CPU off, peripherals on
Sleep: CPU off, peripherals off
Idle mode currents down to 5.8 μA typical
Sleep mode current down to 0.1 μA typical
Timer1 Oscillator: 1.8 μA, 32 kHz, 2V
Watchdog Timer: 2.1 μA
Two-Speed Oscillator Start-up
Peripheral Highlights:
High-current sink/source 25 mA/25 mA
Three programmable external interrupts
Four input change interrupts
Up to 2 Capture/Compare/PWM (CCP) modules,
one with Auto-Shutdown (28-pin devices)
Enhanced Capture/Compare/PWM (ECCP)
module (40/44-pin devices only):
- One, two or four PWM outputs
- Selectable polarity
- Programmable dead time
- Auto-Shutdown and Auto-Restart
Master Synchronous Serial Port (MSSP) module
supporting 3-wire SPI (all 4 modes) and I 2 C
Master and Slave Modes
Enhanced Addressable USART module:
- Supports RS-485, RS-232 and LIN 1.2
- RS-232 operation using internal oscillator
block (no external crystal required)
- Auto-Wake-up on Start bit
- Auto-Baud Detect
10-bit, up to 13-channel Analog-to-Digital
Converter module (A/D):
- Auto-acquisition capability
- Conversion available during Sleep
Dual analog comparators with input multiplexing)
Flexible Oscillator Structure:
Four Crystal modes, up to 40 MHz
4X Phase Lock Loop (available for crystal and
internal oscillators)
Two External RC modes, up to 4 MHz
Two External Clock modes, up to 40 MHz
Internal oscillator block:
- 8 user selectable frequencies, from 31 kHz to 8 MHz
- Provides a complete range of clock speeds
from 31 kHz to 32 MHz when used with PLL
- User tunable to compensate for frequency drift
Secondary oscillator using Timer1 @ 32 kHz
Fail-Safe Clock Monitor:
- Allows for safe shutdown if peripheral clock stops
Special Microcontroller Features:
C compiler optimized architecture:
- Optional extended instruction set designed to
optimize re-entrant code
100,000 erase/write cycle Enhanced Flash
program memory typical
1,000,000 erase/write cycle Data EEPROM
memory typical
Flash/Data EEPROM Retention: 100 years typical
Self-programmable under software control
Priority levels for interrupts
8 x 8 Single-Cycle Hardware Multiplier
Extended Watchdog Timer (WDT):
- Programmable period from 4 ms to 131s
Single-supply 5V In-Circuit Serial
Programming (ICSP) via two pins
In-Circuit Debug (ICD) via two pins
Wide operating voltage range: 2.0V to 5.5V
Programmable 16-level High/Low-Voltage
Detection (HLVD) module:
- Supports interrupt on High/Low-Voltage
Detection
Programmable Brown-out Reset (BOR
- With software enable option
© 2007 Microchip Technology Inc. Preliminary DS39631B-page 1

High-Performance Modified RISC CPU:
Modified Harvard architecture
C compiler optimized instruction set architecture
Flexible addressing modes
83 base instructions
24-bit wide instructions, 16-bit wide data path
Up to 24 Kbytes on-chip Flash program space
Up to 2 Kbytes of on-chip data RAM
Up to 1 Kbytes of nonvolatile data EEPROM
16 x 16-bit working register array
Up to 30 MIPS operation:
- DC to 40 MHz external clock input
- 4 MHz - 10 MHz oscillator input with
PLL active (4x, 8x, 16x)
Up to 21 interrupt sources:
- 8 user-selectable priority levels
- 3 external interrupt sources
- 4 processor trap sources
DSP Features:
Dual data fetch
Modulo and Bit-Reversed modes
Two 40-bit wide accumulators with optional
saturation logic
17-bit x 17-bit single-cycle hardware fractional/
integer multiplier
All DSP instructions are single cycle
- Multiply-Accumulate (MAC) operation
single-cycle ±16 shift
Peripheral Features:
dsPIC30F2011/2012/3012/3013
dsPIC30F2011/2012/3012/3013 High-Performance
Digital Signal Controllers
Note: This data sheet summarizes features of this group
of dsPIC30F devices and is not intended to be a complete
reference source. For more information on the CPU,
peripherals, register descriptions and general device
functionality, refer to the “dsPIC30F Family Reference
Manual” (DS70046). For more information on the device
instruction set and programming, refer to the “dsPIC30F/
33F Programmer’s Reference Manual” (DS70157).
High-current sink/source I/O pins: 25 mA/25 mA
Three 16-bit timers/counters; optionally pair up
16-bit timers into 32-bit timer modules
16-bit Capture input functions
16-bit Compare/PWM output functions
3-wire SPI modules (supports four Frame modes)
I 2 C module supports Multi-Master/Slave mode
and 7-bit/10-bit addressing
Up to two addressable UART modules with FIFO
buffers
Analog Features:
12-bit Analog-to-Digital Converter (ADC) with:
- 200 ksps conversion rate
- Up to 10 input channels
- Conversion available during Sleep and Idle
Programmable Low-Voltage Detection (PLVD)
Programmable Brown-out Reset
Special Microcontroller Features:
Enhanced Flash program memory:
- 10,000 erase/write cycle (min.) for
industrial temperature range, 100K (typical)
Data EEPROM memory:
- 100,000 erase/write cycle (min.) for
industrial temperature range, 1M (typical)
Self-reprogrammable under software control
Power-on Reset (POR), Power-up Timer (PWRT)
and Oscillator Start-up Timer (OST)
Flexible Watchdog Timer (WDT) with on-chip lowpower
RC oscillator for reliable operation
Fail-Safe Clock Monitor operation:
- Detects clock failure and switches to on-chip
low-power RC oscillator
Programmable code protection
In-Circuit Serial Programming (ICSP)
Selectable Power Management modes:
- Sleep, Idle and Alternate Clock modes
CMOS Technology:
Low-power, high-speed Flash technology
Wide operating voltage range (2.5V to 5.5V)
Industrial and Extended temperature ranges
Low-power consumption
© 2006 Microchip Technology Inc. DS70139E-page 1

Freescale Semiconductor
Technical Data
±2.5g - 10g Three Axis Low-g
Micromachined Accelerometer
The MMA7261QT low cost capacitive micromachined accelerometer
features signal conditioning, a 1-pole low pass filter, temperature
compensation and g-Select which allows for the selection among 4
sensitivities. Zero-g offset full scale span and filter cut-off are factory set and
require no external devices. Includes a Sleep Mode that makes it ideal for
handheld battery powered electronics.
Features
• Selectable Sensitivity (2.5g/3.3g/6.7g/10g)
• Low Current Consumption: 500 µA
• Sleep Mode: 3 µA
• Low Voltage Operation: 2.2 V – 3.6 V
• 6mm x 6mm x 1.45mm QFN
• Fast Turn On Time
• High Sensitivity (2.5 g)
• Integral Signal Conditioning with Low Pass Filter
• Robust Design, High Shocks Survivability
• Environmentally Preferred Package
• Low Cost
Typical Applications
• HDD MP3 Player: Freefall Detection
• Laptop PC: Freefall Detection, Anti-Theft
• Cell Phone: Image Stability, Text Scroll, Motion Dialing, E-Compass
• Pedometer: Motion Sensing
• PDA: Text Scroll
• Navigation and Dead Reckoning: E-Compass Tilt Compensation
• Gaming: Tilt and Motion Sensing, Event Recorder
• Robotics: Motion Sensing
Device Name
ORDERING INFORMATION
Temperature
Range
Package
Drawing
© Freescale Semiconductor, Inc., 2006. All rights reserved.
Package
MMA7261QT – 20 to +85°C 1622-02 QFN-16, Tray
MMA7261QR2 – 20 to +85°C 1622-02 QFN-16,Tape & Reel
Document Number: MMA7261QT
Rev 0, 10/2006
g-Select1
MMA7261QT
MMA7261QT: XYZ AXIS
ACCELEROMETER
±2.5g/3.3g/6.7g/10g
1
g-Select2 2
VDD 3
VSS 4
Bottom View
16-LEAD
QFN
CASE 1622-02
N/C
Top View
16 15 14 13
5 6 7 8
N/C
X OUT
N/C
Y OUT
N/C
Z OUT
N/C
12
Sleep
Mode
11 N/C
10 N/C
Figure 1. Pin Connections
9
N/C

The Freescale accelerometer is a surface-micromachined
integrated-circuit accelerometer.
The device consists of two surface micromachined
capacitive sensing cells (g-cell) and a signal conditioning
ASIC contained in a single integrated circuit package. The
sensing elements are sealed hermetically at the wafer level
using a bulk micromachined cap wafer.
The g-cell is a mechanical structure formed from
semiconductor materials (postillion) using semiconductor
processes (masking and etching). It can be modeled as a set
of beams attached to a movable central mass that move
between fixed beams. The movable beams can be deflected
from their rest position by subjecting the system to an
acceleration (Figure 3).
As the beams attached to the central mass move, the
distance from them to the fixed beams on one side will
increase by the same amount that the distance to the fixed
beams on the other side decreases. The change in distance
is a measure of acceleration.
The g-cell beams form two back-to-back capacitors
(Figure 3). As the center beam moves with acceleration, the
distance between the beams changes and each capacitor's
value will change, (C = Aε/D). Where A is the area of the
beam, ε is the dielectric constant, and D is the distance
between the beams.
The ASIC uses switched capacitor techniques to measure
the g-cell capacitors and extract the acceleration data from
the difference between the two capacitors. The ASIC also
signal conditions and filters (switched capacitor) the signal,
providing a high level output voltage that is ratiometric and
proportional to acceleration.
MMA7261QT
Acceleration
Figure 3. Simplified Transducer Physical Model
PRINCIPLE OF OPERATION
SPECIAL FEATURES
g-Select
The g-Select feature allows for the selection among 4
sensitivities present in the device. Depending on the logic
input placed on pins 1 and 2, the device internal gain will be
changed allowing it to function with a 2.5g, 3.3g, 6.7g, or 10g
sensitivity (Table 3). This feature is ideal when a product has
applications requiring different sensitivities for optimum
performance. The sensitivity can be changed at anytime
during the operation of the product. The g-Select1 and g-
Select2 pins can be left unconnected for applications
requiring only a 2.5g sensitivity as the device has an internal
pull-down to keep it at that sensitivity (480mV/g).
Table 3. g-Select pin Descriptions
g-Select2 g-Select1 g-Range Sensitivity
0 0 2.5g 480mV/g
0 1 3.3g 360mV/g
1 0 6.7g 180mV/g
1 1 10g 120mV/g
Sleep Mode
The 3 axis accelerometer provides a Sleep Mode that is
ideal for battery operated products. When Sleep Mode is
active, the device outputs are turned off, providing significant
reduction of operating current. A low input signal on pin 12
(Sleep Mode) will place the device in this mode and reduce
the current to 3uA typ. For lower power consumption, it is
recommended to set g-Select1 and g-Select2 to 2.5g mode.
By placing a high input signal on pin 12, the device will
resume to normal mode of operation.
Filtering
The 3 axis accelerometer contains onboard single-pole
switched capacitor filters. Because the filter is realized using
switched capacitor techniques, there is no requirement for
external passive components (resistors and capacitors) to set
the cut-off frequency.
Ratiometricity
Ratiometricity simply means the output offset voltage and
sensitivity will scale linearly with applied supply voltage. That
is, as supply voltage is increased, the sensitivity and offset
increase linearly; as supply voltage decreases, offset and
sensitivity decrease linearly. This is a key feature when
interfacing to a microcontroller or an A/D converter because
it provides system level cancellation of supply induced errors
in the analog to digital conversion process.
Sensors
4 Freescale Semiconductor

Pin Descriptions
g-Select1
Table 4. Pin Descriptions
1
2
3
4
Top View
16 15 14 13
5 6 7 8
Figure 4. Pinout Description
Pin No. Pin Name Description
1 g-Select1 Logic input pin to select g level.
2 g-Select2 Logic input pin to select g level.
3 VDD Power Supply Input
4 VSS Power Supply Ground
5 - 7 N/C No internal connection.
Leave unconnected.
8 - 11 N/C Unused for factory trim.
Leave unconnected.
12 Sleep Mode Logic input pin to enable product or
Sleep Mode.
13 ZOUT Z direction output voltage.
14 YOUT Y direction output voltage.
15 XOUT X direction output voltage.
16 N/C No internal connection.
Leave unconnected.
V DD
g-Select2
VDD VSS 0.1 µF
Logic
Inputs
Logic
Input
N/C
N/C
X OUT
N/C
1
g-Select1
2 g-Select2
3
4
V DD
V SS
Y OUT
N/C
12
Sleep Mode
Z OUT
N/C
MMA7261QT
Z OUT
Y OUT
X OUT
Figure 5. Accelerometer with Recommended
Connection Diagram
12
11
10
9
13
14
15
Sleep Mode
N/C
N/C
N/C
1 kΩ
0.1 µF
1 kΩ
0.1 µF
1 kΩ
0.1 µF
BASIC CONNECTIONS
PCB Layout
POWER SUPPLY
Figure 6. Recommended PCB Layout for Interfacing
Accelerometer to Microcontroller
NOTES:
1. Use 0.1 µF capacitor on VDD to decouple the power
source. Do not exceed capacitor values of 2.2 or 3.3
µF on VDD-GND. 2. Physical coupling distance of the accelerometer to
the microcontroller should be minimal.
3. The flag underneath the package is internally
connected to ground. It is not recommended for the
flag to be soldered down.
4. Place a ground plane beneath the accelerometer to
reduce noise, the ground plane should be attached to
all of the open ended terminals shown in Figure 6.
5. Use an RC filter with 1.0 kΩ and 0.1 µF on the
outputs of the accelerometer to minimize clock noise
(from the switched capacitor filter circuit).
6. PCB layout of power and ground should not couple
power supply noise.
7. Accelerometer and microcontroller should not be a
high current path.
8. A/D sampling rate and any external power supply
switching frequency should be selected such that
they do not interfere with the internal accelerometer
sampling frequency (11 kHz for the sampling
frequency). This will prevent aliasing errors.
9. PCB layout should not run traces or vias under the
QFN part. This could lead to ground shorting to the
accelerometer flag.
MMA7261QT
Sensors
Freescale Semiconductor 5
Accelerometer
V DD
V SS
Sleep Mode
g-Select1
g-Select2
XOUT Y OUT
Z OUT
C
C
R
R
R
C
C
C
V RH
P0
P1
P2
A/DIN A/D IN
A/D IN
Microcontroller
V DD
V SS
C

+X
X OUT @ +1g = 2.13 V
Y OUT @ 0g = 1.65 V
Z OUT@ 0g = 1.65 V
MMA7261QT
1
2
3
4
Top View
+Y
16 15 14 13
5 6 7 8
-Y
16-Pin QFN Package
Top View
12
11
10
X OUT @ 0g = 1.65 V
Y OUT @ -1g = 1.17 V
Z OUT @ 0g = 1.65 V
9
X OUT @ 0g = 1.65 V
Y OUT @ +1g = 2.13 V
Z OUT@ 0g = 1.65 V
DYNAMIC ACCELERATION
-X
STATIC ACCELERATION
X OUT @ -1g = 1.17 V
Y OUT @ 0g = 1.65 V
Z OUT @ 0g = 1.65 V
Side View
-Z +Z
* When positioned as shown, the Earth’s gravity will result in a positive 1g output.
Direction of Earth’s gravity field.*
Sensors
6 Freescale Semiconductor
Top
Bottom
: Arrow indicates direction of mass movement.
Side View
X OUT @ 0g = 1.65 V
Y OUT @ 0g = 1.65 V
Z OUT @ +1g = 2.13 V
X OUT @ 0g = 1.65 V
Y OUT @ 0g = 1.65 V
Z OUT@ -1g = 1.17 V

FEATURES
• Drives up to 32 LCD segments of arbitrary configuration
• CMOS process for: wide supply voltage range,
low- power operation, high-noise immunity, wide
temperature range
• CMOS and TTL-compatible inputs
• Electrostatic discharge protection on all pins
• Cascadable
• On-chip oscillator
• Requires only three control lines
APPLICATIONS
• Industrial displays
• Consumer product displays
• Telecom product displays
• Automotive dashboard displays
DESCRIPTION
The AY0438 is a CMOS integrated device that drives a
liquid crystal display, usually under microprocessor
control. The part acts as a smart peripheral that drives
up to 32 LCD segments. It needs only three control
lines due to its serial input construction. It latches the
data to be displayed and relieves the microprocessor
from the task of generating the required waveforms.
The AY0438 can drive any standard or custom parallel
drive LCD display, whether it be field effect or dynamic
scattering; 7-, 9-, 14- or 16-segment characters; decimals;
leading + or -; or special symbols. Several
AY0438 devices can be cascaded. The AC frequency
of the LCD waveforms can either be supplied by the
user or generated by attaching a capacitor to the LCD
input, which controls the frequency of an internal oscillator.
The AY0438 is available in 40-lead dual in-line plastic
and 44-lead PLCC packages. Unpackaged dice are
also available.
32-Segment CMOS LCD Driver
PIN CONFIGURATION
VDD
LOAD
SEG 32
SEG 31
SEG 30
SEG 29
SEG 28
SEG 27
SEG 26
SEG 25
SEG 24
SEG 23
SEG 22
SEG 21
SEG 20
SEG 19
SEG 18
SEG 17
SEG 16
SEG 15
SEG 29
SEG 28
SEG 27
SEG 26
SEG 25
SEG 24
SEG 23
SEG 22
SEG 21
SEG 20
SEG 19
40-Lead Dual In-line
44 PLCC
CLOCK
SEG 1
SEG 2
SEG 3
VSS
DATA OUT
DATA IN
SEG 4
SEG 5
LCDΦ
BP
SEG 6
SEG 7
SEG 8
SEG 9
SEG 10
SEG 11
SEG 12
SEG 13
SEG 14
© 1995 Microchip Technology Inc. DS70010I-page 1
7
8
9
10
11
12
13
14
15
16
17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
AY0438
AY0438
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
NC
SEG 30
SEG 31
SEG 32
LOAD
VDD
CLOCK
SEG 1
SEG 2
SEG 3
VSS
SEG 18 18
6
SEG 17 19
5
SEG 16 20
4
SEG 15 21
3
SEG 14 22
2
SEG 13 23
1
SEG 12 24
44
SEG 11 25
43
SEG 10 26
42
SEG 9 27
41
NC 28
40
AY0438
39
38
37
36
35
34
33
32
31
30
29
NC
DATA OUT
DATA IN
SEG 4
SEG 5
LCDΦ
BP
SEG 6
SEG 7
SEG 8
NC

®
DESCRIPTION
The TDA 2003 has improved performance with the
same pin configuration as the TDA 2002.
The additional features of TDA 2002, very low
number of external components, ease of assembly,
space and cost saving, are maintained.
The device provides a high output current capability
(up to 3.5A) very low harmonic and cross-over
distortion.
Completely safe operation is guaranteed due to
protection against DC and AC short circuit between
all pins and ground, thermal over-range, load dump
voltage surge up to 40V and fortuitous open
ground.
ABSOLUTE MAXIMUM RATINGS
October 1998
TDA2003
10W CAR RADIO AUDIO AMPLIFIER
Symbol Parameter Value Unit
VS Peak supply voltage (50ms) 40 V
VS DC supply voltage 28 V
VS Operating supply voltage 18 V
IO Output peak current (repetitive) 3.5 A
IO Output peak current (non repetitive) 4.5 A
Ptot Power dissipation at Tcase = 90°C 20 W
Tstg, Tj Storage and junction temeperature -40 to 150 °C
TEST CIRCUIT
PENTAWATT
ORDERING NUMBERS : TDA 2003H
TDA 2003V
1/10

®
DESCRIPTION
The TDA 2003 has improved performance with the
same pin configuration as the TDA 2002.
The additional features of TDA 2002, very low
number of external components, ease of assembly,
space and cost saving, are maintained.
The device provides a high output current capability
(up to 3.5A) very low harmonic and cross-over
distortion.
Completely safe operation is guaranteed due to
protection against DC and AC short circuit between
all pins and ground, thermal over-range, load dump
voltage surge up to 40V and fortuitous open
ground.
ABSOLUTE MAXIMUM RATINGS
October 1998
TDA2003
10W CAR RADIO AUDIO AMPLIFIER
Symbol Parameter Value Unit
VS Peak supply voltage (50ms) 40 V
VS DC supply voltage 28 V
VS Operating supply voltage 18 V
IO Output peak current (repetitive) 3.5 A
IO Output peak current (non repetitive) 4.5 A
Ptot Power dissipation at Tcase = 90°C 20 W
Tstg, Tj Storage and junction temeperature -40 to 150 °C
TEST CIRCUIT
PENTAWATT
ORDERING NUMBERS : TDA 2003H
TDA 2003V
1/10

Appendix I
Industrial Design Documentation
112

Advanced Rowing Instruments
Industrial Design – Elements to Incorporate in Design
This document describes the elements to consider in designing the coxbox product.
Vx – the Low End Model
Information the Coxswain Needs to See
• Stroke rate – two digit display.
o 16 mm tall and 21 mm wide the measurements for our current display.
o This is updated approximately every 3-5 seconds.
• Timer for showing the time elapsed for a race or practice – 5 or 6 digit display.
o 18mm and 62mm wide are the measurements for our current display.
• Stroke Sensing Status
o Two states are “Ready” and “Not Ready”
o Needed because displaying the stroke rate requires more than 1 full
stroke to measure.
o Can be 1 element or two (meaning can be 1 lamp/LED or 2.)
• Low Battery or Battery Indicator
o If implemented as a low battery indicator, one element (ie, lamp or LED) is
needed.
o Not sure if circuitry exists that can find the battery life.
Controls or Interactivity
• Power On and Off.
o Can be separate control or integrated with another control.
• Volume level control.
• Timer Control
o Start, Pause, and Reset functionality
Connection Jacks
• Male BNC for headset microphone
o Does not have to be on the front panel
• Proprietary five pin connector for speaker
o Does not have to be on the front panel
Case Body
• Modular battery system
• Carrying handle (does not have to be rigid)
1

Figure I.1: Refined Final Design
114

Figure I.2: Final Design Direction
115

Figure I.3: Final Design Direction
116

Figure I.4: Interface Ideas
117

Figure I.5: Case Top View
118

Figure I.6: Case Top View
119

Figure I.7: Case Bottom View
120

Figure I.8: Case Perspective View
121

Figure I.9: Case Side View
122

Appendix J
Business Documentation
123

RI Coaches Survey file:///C:/Documents%20and%20Settings/Saket/Desktop/ARI_ID/ari_co...
Survey for Coaches
This survey is to learn about the purchasing habits, the features, and the problems coaches have with coxswain
boxes. It is for a senior design project in electrical engineering through the Engineering Entrepreneurs Program at
North Carolina State University in Raleigh, NC.
Team/School Name
Your Name (optional)
1. How many of the following does your team currently have or own?
NK's Vox-Box: NK's Cox-Box: Shells: Rowers: Coxswains:
2. How many times over the course of a term do you spend money on repairing or replacing the following equipment:
(Please circle one choice for each product type)
a) Vox-Boxes: nmlkj 0 - 1 times nmlkj 2 - 3 nmlkj 3 - 4 nmlkj 6 or more nmlkj N/A
b) Cox-Boxes: nmlkj 0 - 1 Times nmlkj 2 - 3 nmlkj 3 - 4 nmlkj 6 or more nmlkj N/A
c) Speakers: nmlkj 0 - 1 Times nmlkj 2 - 3 nmlkj 3 - 4 nmlkj 6 or more nmlkj N/A
d) Microphones: nmlkj 0 - 1 Times nmlkj 2 - 3 nmlkj 3 - 4 nmlkj 6 or more nmlkj N/A
e) Wiring Harnesses: nmlkj 0 - 1 Times nmlkj 2 - 3 nmlkj 3 - 4 nmlkj 6 or more nmlkj N/A
f)
Seat
Magnets/Sensors: nmlkj 0 - 1 Times nmlkj 2 - 3 nmlkj 3 - 4 nmlkj 6 or more nmlkj N/A
3. How much money do you spend annually to repair, replace, or purchase the equipment listed above?
nmlkj $0 to $50 nmlkj $51 to $200 nmlkj $201 to $400 nmlkj $401 to $600 nmlkj $601 to $1000 nmlkj $1000+
4. On average, how many years of service do you get out of your Vox-Boxes and Cox-Boxes before they need
replacement? (Please select one choice for each product type)
Vox-Boxes nmlkj
Cox-Boxes nmlkj
Less than 1
year nmlkj 1 to 2 years nmlkj 3 to 4 years nmlkj
Less than 1
year nmlkj 1 to 2 years nmlkj 3 to 4 years nmlkj
Which months do you generally purchase these devices?
5 or more
years nmlkj N/A
5 or more
years nmlkj N/A
5. Please rate the following factors why Vox-Boxes and Cox-Boxes must be repaired or replaced. (If you have had no
of 3 5/2/2007 1:09 AM

General Comments to the Customer Survey
I'd like to say that I don't have many repairs on NK equipment because it is SO expensive and
impossible to repair!!!!! And I would love to buy more but they are way way way too expensive.
so we deal with what we have, and hope for the best. many times they don't work
1 year warranty isn't much, Service support, durability, ease of use. NK equipment is not
durable, but they've got the market, and thus we're stuck with them. Their headsets are really
bad--don't last at all.
Customer Service: Delivery of equipment and spare parts in a timely and direct fashion.
Accessibility of representatives. Knowledge about the rowing community and needs of a
rowing program. Will this company exist in 10 years when their equipment needs
maintenance?
No idea about the longevity of the product or the service of the company. It shouldn't be
hard to be better than NK, but thats not good enough. I'd probably wait to see how the product
worked for a year or two before deciding to purchase them for my program. I don't like being
the test monkey for products unless they're free.
Customer service, creditability, and longivity
None. I would consider us willing to experiment with something else because, in my opinion,
NK products are unreliable relative to what you spend. Their plugs and wiring
harnesses are poorly designed and ANY improvement someone else could offer would
be reason enough to switch.
None, if the company offered a 60 day money back guarantee or trial period. I would love to
try something new more modern and lighter weight.
Unproven Durability
possibility of problems in software and hardware. issues with the casing not being durable
enough.
reliability and functionality. I have done buisness with NK for almost 20 years
We have done this with other companies. I would want a 100% return warranty if anything
doesn't work, or if the product does not perform to spec.
Reliability and confidence that the company will be there in 5-6 years.
support and repair, at which NK sucks
Reliablity and customer service when there are problems with the unit.
This type of investment is a large amount for something unproven
Glitches
If it could withstand the rigors of daily rowing life. Materials used, quality control in
processing, and price
Software becomes complex and requires a willingness to undergo a learning curve. Easy is
good! If it works seamlessly, it would be worth a try, though.
It would need to be tested by reputable teams and coxswains. NK practically dominates the
market, so I would have to be very convinced that the new competing product is better. That
being said, I don't think you'd have to try very hard to make a better product. Would you
consider puting in a walkie-talkie device so the coach can speak into a device from the launch
and have it feed directly into the sound system in the boat? That would be a great and I would

ARI – Email Correspondence with Eamonn Hynes, Dublin, Ireland.
Greetings,
I came across your ARI pdf on the internet. This is probably all a bit random for you and I'm sure
you've got exams to worry about...
I was wondering if I could ask you a question about section 2.4.2?
In this section you state that NK's turnover is $10 to $20 million. Is this an estimate, or is it
based on some real report/document?
I tried looking for NK company information on various US government websites, but to no avail -
all I could find was your document.
I should be most grateful if you could back-up this figure for me.
Many thanks,
Eamonn
Eamonn,
Greetings and thanks for the email. It's interesting to find someone who found and read our
report. I don't know if you could infer it from the report, but it was written as part of an electrical
engineering senior design project at NC State University in Raleigh, North Carolina, USA. My
friend Greg Mulholland (and longtime rower) and I created a virtual company and did market
research, wrote a business plan, etc. We are actually building our prototype coxbox this
semester.
We had a very very difficult time finding out anything about Nielsen Kellerman. They are a small
private company and so they aren't required to publish or release anything. I even tried calling
them asking for the information but they wouldn't reveal it.
We found the $10 to $20 million estimated sales on a database called ReferenceUSA. You may
be able to find it through your university library, provided that you are a student as well.
It was nice to hear from you. Let me know if you'd like to get updates on our progress in building
a new coxbox. Are you also a rower?
Sincerely,
Saket Vora
Greetings,
Thank you for your prompt reply. The information you provided was most useful. I am working
for a venture capitalist who had a proposal from a team of guys here in Dublin about an

Appendix K
Parts List
127

Advanced Rowing Instruments
Near Complete Project Parts List
Part Description
VI-602-DP-FC-S Varitronix LCD (6 digit)
VI-201-DP-FC-S Varitronix LCD (2 digit)
MMA7260QT 3-axis Freescale Accelerometer w/board
PIC18F4550 Microchip PIC controller
PIC18F4550 (Production Model) Microchip PIC controller
CD4543B 7-Segment Display Driver
TDA2003V 10W IC Car Radio Audio Amplifier
AY0438 4-digit Serial LCD Driver
AY0438 (Production Model) 4-digit Serial LCD Driver
HR845CT SecureDigital Card Connector
497-1868-5 IC Mux/Dexmux SGL 8 Channel Analog
P3K1502 POT 5K Ohm 9mm Vertical Pla Bushing
647-1002-1 Module Bluetooth SPP W/Ant Class 2
296-19842-1 IC RS232 3V 5.5V Driver 20-SOIC
DS1818-10+CT IC Econoreset 3.3V
425-1732-5 IC Regulator LDO 3.3V
497-1441-5-ND IC Precision 5V
497-1453-5 IC Precision 15V
AE10009ND IC Socket Straight 40 Pins
V2012-ND PC Board FR4 1-Side PPH
DV164006-ND ICD2 PICDEM2 Dev Board Kit
044 103 10305 Amphenol Series 44 - 5 Pin Plug Housing
044 104 10305 Amphenol Series 44 - 5 Pin Receptacle Housing
044 100 1414P Amphenol Series 44 - Pin Contacts
044 102 1414S Amphenol Series 44 - Socket Contacts
Industrial Design Decal Paper, 10 sheets of clear laser paper
VI-602-DP-FC-S Varitronix LCD (6 digit)
VI-201-DP-FC-S Varitronix LCD (2 digit)
TDA2003V 10 W IC Car Radio Audio Amplifier
425-1622-5-ND IC Regulator LDO 9V with On/Off Control
311-1202F-10K 10k Audio Thumbwheel Pot
311-1202F-5K 5k Audio Thumbwheel Pot
311-1204F-10K 10k Audio Thumbwheel Pot, w/SW
571-4356687 8 position DIP switch
A413-ND 50-pin Solder Tail DIP Socket
A411-ND 40-pin Solder Tail DIP Socket
A403AE-ND 18-pin Solder Tail DIP Socket
A402-AE-ND 16-pin Solder Tail DIP Socket
A405-ND 24-pin Solder Tail DIP 0.300 CTR
A209-ND WW SIP SOC W/Screw Machine Pin
A460-ND Conn Socket SIP 40POS TIN
AE7108-ND IC Socket Mach Pin WW 8 POS Gold
AE7140-ND IC Socket Mach PIN WW 40 POS Gold
571-4356687 TV5 SP SLDR 3.2 HOLE - RoHS:

Index
A/D converter, 46
ABS
3D printer facility, 63
accelerometer, 52
implementation, 40
orientation, 41
overview, 40
sensitivity selection, 41
Advanced Sporting Instruments
expansion into, 38
advertising
Internet, 21
word of mouth, 21
Alias, 75
amplification, 75
ASCII, 52
attendees, 75
audio amplifier
volume control, 44
audio connectors, 44
banner, 74
Bluetooth, 46, 61
user interface, 44
Board of Directors, 35
boat, 10
boat mount, 74
booth, 74, 75
Carolina Electronic Assemblers, 33
case study
for marketing, 33
catch, 40
check, 11
Chief Executive Officer
129
description of, 36
Club Rowing Team, 74
control buttons, 44
copyrights, 30
Cox-Box, 74
CoxBox
history of, 25
coxbox
definition, 11
CoxMate, 25
Coxmate, 31, 32
coxswain, 11, 75
with regard to amplification of voice, 42
Dad Vail
product roll out, 34
data logging
in business model, 13
Davies Row Tech, 31, 32
demonstrated win, 33
deployment, 28
Design Day, 74
digital I/O, 46
Dirani, Sam
with respect to manufacturing, 33
distributors
partnerships with, 34
drive, 40
Eamonn Hynes, 22
EEP, 74
electronic contract manufacturer, 32
role in business model, 13
European sales, 21
evaluation, 28

FAT, 51
features, 52
foam model, 74
GPS, 25
GUI, 52
GUIDE, 52
handout, 75
Head of the Charles
product roll out, 34
head race, 10
impeller, 25
In2 Rowing, 25
breakdown of, 26
with regards to positioning, 26, 27
In2Rowing, 31
industrial design
inspiration, 62
overview, 62
intellectual assets, 28
intellectual property, 28, 31
strategy, 27
LCD, 48
Liquid Crystal Display, 43
advantages of, 43
driver IC, 43
logo, 74
MAC, see Multiply Accumulate Unit
MATLAB, 52
MCC18, 47
McKimmon Center, 74
Mercury, 74, 75
Microchip, 46
microphone, 75
modular
in business, 13
modular battery, 42, 63
Montoya-Weiss, Mitzi, 33
Multiply Accumulate Unit, 46
music, 75
NCSU College of Design, 62
Nielsen-Kellerman, 25, 31, 74
130
breakdown of, 25
distributors, 34
market share, 25
relation to sales, 21
with regards to headset microphone, 42
with regards to positioning, 27
Norfolk Academy, 33
online store, 21, 34
patents, 28, 29, 31
performance analysis
in business model, 13
phantom power, 42
PIC18F4520, 46
with regards to ADC port, accelerometer, 41
poster, 74, 75
power management circuit, 42
voltage regulators, 42
prescaler, 48
President
description of, 36
printed circuit board, 45
with respect to manufacturing, 33
procurement, 28
product demonstrations, 34
prototype, 74, 75
recovery, 40
regattas
product roll out, 34
related to sales, 21
regional representatives
sales, 21
regional sales representatives, 34
Resolute, 34
Row2K.com
advertising, 20
RowData, 25, 31
Rower’s Almanac
with respect to advertising, 20
with respect to market research, 18
Rowing News
with respect to advertising, 20
sales channels, 21
sales cycle, 20

screensaver, 75
SD, see Secure Digital
Secure Digital, 46, 51, 61
user interface, 44
Serial Peripheral Interface, 46
shell, 9, 74, 75
slideshow, 75
software, 52, 61
Solem-Derringer, Karen, 18
SpeedCoach, 12
SPI, see Serial Peripheral Interface
sprint race, 10
stroke rate, 11, 52
supply chain, 33
survey
market research, 15
Texas Instruments, 46
TheMathWorks, 52
TI, see Texas Instruments
timer, 48
Titan, 52, 61
trade secrets, 29
trademark, 31
trademarks, 29
Trinity, In2 product, 25
University of North Carolina, 33
USART, 46, 51
USPTO, 29
USRowing
overview, 9
regulations, 9
Vespoli, 34
Vice-President, Business Strategy, 36
Vice-President, Hardware, 36
virtual employees
description of, 37
voltage follower
for accelerometer, 41
voltage regulator, 42
WinTech, 34
Yauch, Steve, 33
131