Home Energy Management System
Sponsor: Progress Energy
Spring 2011
Group Members:
Zineb Heater
Ryan Jones
Dennis Kilgore
1
Contents
Executive Summary
1
1 Definition
3
1-1 Overview
3
1-2 Motivation
4
1-3 Goals and Objectives
6
1-4 Requirements
7
1-5 Specifications
12
1-6 Design Constraints
13
1-7 Previous Projects and Commercial Products
14
1-8 Safety Concerns
17
2 Research
19
2-1 Methods
19
2-2 Components
41
3 Design
48
3-1 Power and Current Sensing Circuit
48
3-2 Power Relay
50
3-3 Wireless
53
3-4 Microprocessor
61
3-5 Main Control Board
62
3-6 Block Diagrams
65
3-7 Software
67
2
4 Prototype
77
4-1 Design Planning
77
4-2 Parts Acquisition
80
4-3 Budget
80
5 Testing
81
5-1 Wireless Modules
5-2 Microcontroller
5-3 LCD Touch Screen
81
81
83
5-4 Current/Power Sensor
83
5-5 Software
84
5-6 Power Relay
86
5-7 System Testing
86
6 Reflections
87
6-1 Features Excluded
87
6-2 Future Improvements
87
7 Conclusion
8 Product Manual
Appendices
90
92
93
3
Executive Summary
Throughout the centuries, energy as a concept has worked its way into nearly
every scientific application, through gadgets and gizmos and into the homes of
everyday Americans. According to the U.S. Department of Energy, the United
States currently consumes the most of the world’s energy, and 21% of the
country’s energy consumption comes from residential use. In an effort to curb
these facts in the future, many have thought to turn towards renewable energy
sources. However, before new energy sources are created, the idea of
responsible energy consumption should first be considered. In fact, in order to
properly integrate renewable energy sources, people will need to be aware of
their excessive energy usage on a daily basis in order to curb their dependency
on non-necessities, such as keeping lights on when out of the room or plugging
in appliances no longer in use.
In studies conducted in 2000, Vampire Draw (referencing the power used by
electronic devices when they are supposed to be dead) accounted for upwards of
13% of the average American household energy bill. This means that 13% of the
household energy bill is from wasted power – electronics that are not in use and
are not benefiting the homeowner, except they may merely save a few moments
when the device initially starts up. Much of the excess energy in these devices is
dissipated as heat, which also raises home cooling costs showing as a double
expenditure on the bill.
While renewable energy is a very altruistic motive as far as the environment and
futuristic goals are concerned, there is also a very pertinent and personal motive
to energy management – to save money by reducing household energy usage,
thus reducing the final total on the monthly bill. As the world moves forward,
technological advances will only add to the growing list of gadgets found in
American homes. Not only are there cell phones, game consoles, cable boxes,
desktop computers, laptops, and ebook readers, there are the chargers,
controllers, and monitors associated with each that are left plugged in throughout
the day.
The Home Energy Management System is a project that handles the opposite
end of the power spectrum as renewable energy, sustainability. Sustainability is
as important as finding sources of renewable energy. At the current pace of the
country, we may have sufficient energy reserves and resources to last decades,
but why can we not stretch that even further? The world needs to act today in
preparation for the future, and it begins with sustainability projects.
The goal of this project is to give users an idea of how much energy is being
used throughout the month and where their major energy drains are within their
home. The project aims to be a sustainability project. As well, the project will
provide the users with the ability to control the power going into different areas of
4
their homes. The idea is to combine a power switch and power meter into one
unit, and allow users to view the data and control the switch wirelessly.
The product will be designed so that one central unit can be used to record the
data and function as a server for the controlling application. Due to differing
layouts of houses, it would be difficult to predict whether a central unit will always
be able to connect to all of the switch/meter units, so a wireless protocol that
uses mesh networking is ideal. That way, regardless of the size or structure of
the house, all data can be communicated back to the central unit without issue.
The Xbee protocol will be used as it combines low operating power consumption
and the necessary wireless networking.
The measuring accuracy is expected to be within 10% as this will be sufficient in
providing the user with a fairly accurate predictor of their monthly power bill,
especially considering current methods (i.e. guessing) could yield up to 50%
inaccuracy.
The data will be formatted in a user-friendly way to increase desire of use and
remove any anxiety those not too computer literate might foster. As well, the
application will easily display all pertinent information attained from the device.
While the control of the house’s energy consumption from anywhere in the world
is the ultimate goal, the users will also be able to effectively budget their energy
expenditures to ensure they meet their monthly household budget.
5
1- Definition
1-1 Overview
The Home Energy Management System empowers the user to save hundreds of
dollars on electric bills. Electricity bills are rising and there is no end to it unless if
we make the right actions and implement new methods to save energy and
money. With the Home Energy Management System, the user can cut down on
costs and find out what appliances are actually worth keeping plugged in and
those that are consuming way too much energy. The U.S Department of Energy
reports that 20% of the country’s electric bills come from items that are left
plugged in when they are not in use, or items that are in standby mode. With the
Home Energy Management System, the user can monitor the energy eaters in
their homes, cut down their electric bills and protect the environment by using
fewer resources at the same time.
The user can plug whatever item they want into the device and it will tell them the
efficiency of that item by displaying the kilowatt per hour. A high- level block
diagram can be found below as figure 1-1.1. The Home Energy Management
System will help the user determine which items are costing the most to run. The
system will calculate the power, voltage, current, and power factor. The system
will allow the user to calculate the electric bill even before receiving it from the
electric company. By simply connecting various appliances to the system, it will
assess how efficient they really are. An LCD screen will display the calculated
values of current, voltage and power. The display will also provide real time
graphs of the AC current and power being used. The user could calculate their
electrical expenses by the day, week, month, or even for an entire year. But
measuring appliance consumption is just one of the various options offered by
the Home Energy Management System. The system will monitor voltage and can
also test if an outlet is working, or evaluate the quality of the electrical power
provided by the utility company. It can detect voltage drops around the house;
help to predict brownout conditions or to make sure a new home's outlets are in
working condition before escrow closes. The system will also allow the user to
access the information wirelessly. A high level block diagram of the system can
be found below in figure 1-1.1. It shows the three planned modules, and how
they interface with the touch screen and main control boards. Data readings
from each module are sent wirelessly to the main control board, which creates a
GUI for the data to be presented to a user.
6
Figure1-1.1: High level block diagram
1-2 Motivation
Motivation for this project derives from the need for more effective home energy
management system. To ensure the sustainability of the planet, we must make a
real effort to conserve resources. Our project will assist in this by solving two
main problems.
First, people often leave electronic appliances on for
unnecessary amounts of time or when the appliances should be off. Second,
wasted standby power leads to extra energy costs that should be avoided as
they waste a significant amount of energy. We should put an effort into
conserving the energy produced from nonrenewable resources because there
are limited resources. Also, the process of producing energy pollutes water, land
and atmosphere.
It obviously does cost a lot of money and time to dig resources out of the ground
and transform them into the usable form of energy that we are aware of and use
on a regular basis. If there were ways to use less energy, both time and money
could be saved. Conserving the home energy not only does lower our monthly
and yearly bills, it also benefits our health and environment as there will be less
harmful gases in the air and more resources in the ground of the coming
generations.
Whenever we save energy, not only we save money and time but we also reduce
the demand for such fossil fuels such as oil and natural gas. The less we burn
7
these fuels, the lower emission of Carbone Dioxide in the air. The average
American produces over 40,000 pounds of Carbone Dioxide emissions each and
every year. By taking few responsible actions, we can cut our annual emissions
by thousands of pounds and we can also cut our energy bills by a huge amount
of money. These resources are at the moment finite, and as such, we need to
limit their use as much as possible.
The home management energy system will be easy to use. The LCD screen will
allow the user to monitor their consumption in real time. We are considering the
display of current and power graphs in real time so the user can be aware of the
power usage. We also intend to display the power consumption in dollars and
cents. For instance, most non technical people do not understand what a kilowatt
hour would cost. So to make the system as user friendly as it could be, we will
display the kilowatts hour as well as the cost in dollars.
Our home management energy system will not make people go without. The
user can still watch his or her favorite show and charge their cell phone or laptop
as needed. The system will enable the user to reduce energy usage of
appliances that are left on a standby mode. That means that the users have a
real choice and hopefully the power to change energy use and limit waste.
Many people leave so many appliances in a standby mode. For instance, people
usually leave their cell phone chargers plugged into the wall even though they
are not charging the cell phone. If only they knew that only 5% of the power
drawn by the cell phone is actually used to charge the phone. The remaining
95% is wasted when the charger is plugged into the wall but not into the cell
phone. This sounds terrible, because we are basically wasting 95% of our energy
unnecessarily.
This group believes that many businesses could benefit from this type of
technology. Although businesses do not pay the same power rates during the
day as they do during the night, many businesses leave hundreds of computers
running at night, and entire rooms with their lights are left on. This type of power
usage is wasteful, and detracts from the overall goal of energy sustainability.
We conducted little research and found out that a large number of appliances
cannot be turned off completely without being unplugged. This kind of appliances
such as cell phone chargers, TVs and VCRs draw power 24 hours a day whether
they are being used or not. For example, a DVR and digital cable box can use up
to 43.46 W when turned off by a remote. Stand By power is ultimately the power
being consumed by these appliances while they are switched off by still plugged
into the wall. For a single appliance, the power wasted might not be a huge
concern. Certain devices such as network routers and modems, which largely do
not need to be on unless they are being used, cost anywhere from $2-$4 per
month. However, when we add up all the appliances and gadgets we have in our
homes, the wasted energy becomes significant and sometimes critical.
8
To paint a clear picture of how much we waste due to stand by appliances, we
need to picture a typical American home. The American home has typically more
than 20 appliances that are constantly costing money and energy. If we focus on
one of these appliances, let’s say the TV. We realize that the TV is actually
turned on all the time even though it looks off. It’s constantly preheating the
picture tube and powering the receiver for the remote and waiting to be turned
on. According to a study performed by Cornell University, these so called
“Vampire appliances” consume and estimated $3 billion a year or about $200 per
household. Standby power consumes over 7% of a home’s total electric bill. The
money saved could easily go toward a fun hobby or a nice vacation.
We believe that the consumer could put a stop to such a huge waste of energy.
The home management system will allow the user to be conservative when it
comes to energy use. It will help save the money and the planet. The
advantages of the home energy management system would be amplified if the
home were powered by a renewable resource such as solar or wind in situations
where the weather is overcast or the wind is not blowing.
1-3 Goals and Objectives
The goals of this project are to reduce needless power usage in homes with a
system that is effective at reducing power use and noninvasive to daily lifestyle in
a relatively inexpensive and simple way, without using more power than would be
saved. So, a main objective of the home energy system would be low power
usage. If the system is going to consume more energy than it saves, then the
system would be simply useless and ineffective. Thus, the entire system must
operate at a power usage lower than the amount of power the regular Vampire
Draw would waste.
We will create multiple modules for switching power on and off as well as
monitoring current. Also a small control unit will be built connected to a touch
screen display which will be the user interface for the system. Also, a unit which
the user can monitor the usage over the internet will be designed. The central
unit needs to be user friendly and safe to use. The main idea is that with this
system, the user will be able to see significant power bill reductions in a simple,
nearly completely passive manner.
A current flow and current measurement module will be in line with each outlet in
a system. The current used will be recorded using a current sensor that we will
build and will be sent to the main control device. This main control device might
be a microcontroller that is able to turn off the flow of power to anything beyond
one of the modules.
Another sensor will be used to sense the presence of a person in any area of the
house. The sensor might be a motion sensor or a body temperature sensor. The
9
main control device will be receiving data from both sensors and sharing data
with the LCD display and the unit that the user can monitor over the internet. A
serial the Ethernet device server will be used to allow the user to communicate
with the main control device through the internet. The user will be able to choose
to turn the power off an appliance or many appliances at the same time. This
feature can easily be accomplished with the use of the microcontroller. The
microcontroller will send out command of either a 1 or 0 to the designated
appliance chosen by the user.
With the project’s expectations in mind, the wireless communication will require a
device that can allow data to be transmitted between the units in distances of
possibly 35 to 60 feet. As houses do not always come in convenient, tiny box
shapes, the device used must reliably send the data from the measuring units to
the display units which would be set in a central location within the domicile. If
the measuring units created a peer to peer network to pass along the data to the
main unit perhaps that could solve any issues in larger facilities. As well, if a
wired network is chosen, data reliability must be taken into consideration and
preserved over the existing powerline infrastructure of the house.
The software goals of the project are to ensure that the user will want to save
energy in a simple, passive way. The GUI for the project should be clean
enough to not be distracting, but still informative enough that the user will want to
look at it often. Also, the GUI must be responsive enough that the user does not
feel frustrated while waiting for a window to load. The user must feel like they
are in control of their energy usage, by offering them many options to customize
their experience.
The demands of the project require the use of a microcontroller in each
measuring unit, and a larger processor in the display unit to control the touch
screen display. In theory, the microcontroller in the sensor units should be able to
convert the analog measurement from the metering components and send that
information to the display unit after completing and compiling any necessary
calculations. The microcontroller needs to be small enough that the size of the
sensor units will not become an eyesore to homebuyers and will still remain
applicable in an end user situation.
This project demands an environment where real time graphs of power usage
can be displayed on a touch screen with near seamless transitions. This should
be possible with a strong control unit and enough memory. To accomplish that,
an FPGA would be difficult to design. Instead, a microprocessor will be used.
Because a microprocessor is needed, a support system for that is necessary.
This includes RAM, flash memory, and communications access. Research in
the following sections will include these issues.
1-4 Requirements
10
1-4.1
Power and Current Measurements
To measure the electrical power we can use different methods. For instance, a
Hall Effect sensor can be used to measure voltage, current, and power levels by
using a current perpendicular to a magnetic field to cause an electric potential.
Another way is to use voltage response measurements where the electrical
characteristics of a circuit are determined from the amplitude and phase of a test
current flowing through the circuit. To measure power the power in this case, the
sensor will have to measure the phase difference between voltage and current.
In order to measure the energy consumed by the devices, we will needs to
measure both the voltage and the current waveforms. For this purpose, analog to
digital converters are going to be used. Due to the high Voltage and the inability
to measure current directly, voltage and current sensors have to be built. Several
sensing technologies are being considered and will be discussed.
Our application requires one measurement device to capture the voltage across
the terminals of a load, and the second device to capture the current going
through the load. However, the actual power calculation depends on the resistive
and reactive components in the circuit. In order to have an accurate power
measurement, we find it necessary to study the electric power a little bit more
and define its different components. While conducting our research, we realized
that in alternating current circuits, energy storage elements such as inductance
and capacitance may result in periodic reversals of the direction of energy flow.
The portion of power flow that averaged over a complete cycle of the AC
waveform, results in net transfer of energy in one direction is known as real
power. On the other hand, the portion of power flow due to stored energy, which
returns to the source in each cycle, is known as reactive power. AC power flow
has the three components:



Real power (P), measured in watts (W)
Reactive power (Q), measured in reactive volt-amperes (VAR).
Complex power (S), measured in volt-amperes (VA). S is usually referred
to as the apparent power and it’s the simple product of the RMS voltage
and the RMS current.
The mathematical relationship among these power components can be
represented by vectors and is typically expressed using complex numbers:
S=P+jQ.
Real power is the measure of a circuit’s dissipative elements (Resistances) and
is the one represented by P, which has a unit measure of Watts. P is the power
component we will be measuring in our application since it is the one the power
company charges us for. Typically, there are two different techniques that are
used to measure the real power P. The first way is to take the time average of
the instantaneous product of voltage and current over a period of time. Another
common method is to use the impedance angle depicted in the power triangle
11
consisting of the three components of power defined above. The cosine of the
impedance angle is directly proportional to the amount of reactance in a circuit,
which is called power factor PF. The power factor is the cosine of the phase
angle between voltage and current. A power factor of 1 represents a purely
resistive load and power factor less than 1 represents a reactive circuit. Power
factor is also defined as the ratio of true or real power to apparent power. Power
factor takes into account both the phase and wave-shape contributions to the
difference between true and apparent power. In our case, both the voltage and
the current waveforms are sinusoidal. Therefore, to measure the power factor we
simply need to read the time difference between zero crossings of the waveform.
To accurately measure the real power in our system, we need to have some
capabilities such as:



Voltage and current waveform acquisition capability
Simultaneous acquisition of both measurement waveforms
Both measurement devices must acquire simultaneously
A power factor of one is the goal of any electric utility company since if the power
factor is less than one, they have to supply more current to the user for a given
amount of power use. Low power factor is caused by inductive loads (such as
transformers, electric motors, and high-intensity discharge lighting), which are a
major portion of the power consumed in household. Unlike resistive loads that
create heat by consuming kilowatts, inductive loads require the current to create
a magnetic field, and the magnetic field produces the desired work. The total or
apparent power required by an inductive device is a composite of the real power
and the reactive power discussed above. In order for our power measurement to
be within tolerance, we have to make sure that the error caused by the power
factor is acceptable. To do this, we will install capacitors in our AC circuit to
decrease the magnitude of reactive power. Capacitors release energy opposing
the reactive energy caused by the inductor. This implies that inductance and
capacitance react 180° to each other. The presence of both in the same circuit
results in the continuous alternating transfer of energy between the capacitor and
the inductor, thereby reducing the current flow from the generator to the circuit.
When the circuit is balanced, all the energy released by the inductor is absorbed
by the capacitor. The following figure shows the power measurement error
versus the power factor error:
12
Figure1-4.1.1: Power measurement error versus power factor and phase
angle error
1-4.2
Microcontroller
The microcontroller to be used will be an essential part of the whole design. Not
only does the microcontroller have to be able to easily interface with the different
components of the design, but it also has to have enough output and inputs pins
to manage all the data being received and sent. The microcontroller has to
receive data from the current sensor, convert it to power and transmit it to the
LCD screen. This requires the microcontroller to have a large number of input
and outputs pins as there will be different tabs available to the user to program
the microcontroller. Since the microcontroller will receive data from the current
sensor, we need to make sure the voltage values measured do not exceed the
maximum input voltage to the microcontroller. This can be done by either using a
voltage divider or an attenuating circuit with an op-amp. The microcontroller must
also be able to control a power relay the power off a device if the user chooses to
do so. As microcontrollers become smaller and cheaper, the functionality of
many external components is being integrated directly into the microcontroller.
For instance, eight-bit microcontrollers come in a variety of package sizes,
random access memory and read only memory sizes, serial communication
buses, and analog inputs and outputs.
The fact is that we have a huge amount of microcontrollers to choose from. We
need to choose the one that matches our design requirements and cost
constraints. Our choice could integrate some analog and digital capabilities such
13
as an analog to digital converter, amplifiers, timers, counters, a small computer
serial interface and even an USB. This can allow us to significantly reduce the
number of components we have to use. We considered several different
microcontrollers for processing the data such as PIC 18F2455, MSP430FE4272,
CC2531 and Arduino. We considered using the PIC18F2455 because of its
popularity and easiness. Microchips microcontrollers are used widely and several
examples were found. The programming is easy as well. The MSP430FE4272
was considered because of the fact that it offers excellent support in the energy
sensing field. It also includes a DSP that manipulates current and voltage
waveforms parallel to the CPU such that the device only reports important data
every second. Its power consumption is low as well. The CC2531 was
considered because it includes RF communications with the Xbee stack protocol
and an 8-bit 8051 microcontroller. This will cut our cost as we will not need to
purchase the wireless chip separately. However, it may be very time consuming
to learn how to program it and use it.
1-4.3
Wireless communication
Our issue is to figure out the right wireless technology to use in our application.
The first step in deciding the kind of wireless components we will be using is to
decide between a one way wireless protocol and a two way wireless protocol.
The answer to this question was simple as we obviously need a two way wireless
protocol. The microcontroller will send data and receive commands. Also,
whichever wireless protocol we choose, it has to be as simple as possible to
enable an easy learning process and implementation. The group has considered
several types of wireless protocols. We will discuss all the types considered in
the research phase. We will point out their advantages and conveniences and
chose the one that serves our application the best.
One of the types wireless protocol considered is Wi-Fi, or also known as wireless
fidelity. In a Wi-Fi network, computers with the network cards connect to a
wireless router and this is connected to the internet by means of a modem. Any
user within about 60 meters of the access point can then connect to the internet.
This was more than enough if the user is using a system in a house. The choice
of Wi-Fi networks was also considered because of security reasons as the
system will be secured and a password is needed.
Another wireless protocol we are considering is the Zigbee. This last one works
in a slightly different manner than IR and RF. It works off a grid of multiple nodes
in a particular location and uses multiple communication paths. Unlike IR and RF,
this wireless meshed networking alternative could make life easier for, as it
doesn’t have to be in a direct line of sight in order to send signals.
Bluetooth technology is also considered as it is useful when transferring
information between two or more devices and also when seep is not a huge
issue. In our case, we need to transmit data between 2 devices at least.
14
Xbee modules were also considered in our design. The XBEE modules are the
core of the mesh communications protocol. They provide an easy to use
Application Interface and handle all the mesh networking by themselves. These
modules offer several advantages such as the ease of use and the possibility of
being tested independently from the microcontroller.
1-4.4
Touch Screen
The touch screen used must respond to the touch of hands which are relatively
clean, and hands that are wearing gloves. We will not consider conditions of
greasy, or dirt covered hands. In addition, the touch screen must be accurate
enough to respond to a touch in the approximate area that the user attempts to
touch it. The screen must be durable enough to last through years of use. It also
must have a function where it can sleep, but turn on if a user presses it to
activate it. Also, the screen must have minimal image burning, as many
elements of the GUI will look similar, and the system may have burn marks after
extended periods of use. The
1-4.5
Software
The software for this project must be compatible with itself. It is desired that the
software be written in an object-oriented language. It would be preferred that all
programming be done in the same language. The current programming
convention is object oriented coding. For the maximum control and compatibility,
it was found that C++ would be ideal. Java was also considered due to its user
friendliness, but the C++ libraries are not compatible.
A Graphic User Interface (GUI) will be needed to allow the user to navigate the
system. Many functions of a GUI require an advanced processor, so an FPGA
will not be ideal. In addition, GUIs require some sort of an operating system to
be run. The operating system must be chosen carefully. Because a full
operating system will not be possible in such a device, the system will need to be
run from a lightweight, minimalistic operating system.
1-5 Specifications
1-5.1
The Power and Current Sensing Circuit
This circuit will consist of a current sensor and a relay. The current measured will
be transmitted to the microcontroller and commands will be received from the
microcontroller to shut the power of an appliance. In order for the measurements
to be accurate, we have defined the following specifications for this part of the
design:


This part of the circuit should consume very low power
Be able to measure currents varying from 0 Amps to 15 Amps
15



Be able to measure voltages up to 120V
Be able to control the relay and shut the power of an appliance
Power measured must be within 10% error
1-5.2
The Microcontrollers
The uses of this project are quite advanced, so a prefabricated microcontroller or
processor will be required. An FPGA is not practical in this sense because the
architectures involved would take a long time to develop. We are considering the
uses of 3 microcontrollers for the child modules as each one of them will monitor
an appliance. The microcontroller should be able to receive data from the
current/power sensing circuit, calculate the consumed power and transmit the
data to the LCD screen.
The main control board will have its own
microcontroller. This unit will take information transmitted by the other units,
process it, and modify it in a format readable by humans.
The main
microcontroller must use less power than the unit will prevent other devices from
using.
1-5.3
Touch Screen
The touch screen used must respond to the touch of hands which are relatively
clean, and hands that are wearing gloves. We will not consider conditions of
greasy, or dirt covered hands. In addition, the touch screen must be accurate
enough to respond to a touch in the approximate area that the user attempts to
touch it. The touch screen should be large enough to be usable by an average
person, but not so large that the person should need to keep the device farther
than an arm’s length from their eyes to read all of the information. Ideally this
size is a six inch to ten inch screen. This way, there will be enough pixels on the
screen to allow the data to be displayed, but not so many as the graphics will
look distorted or pixilated.
1-5.4
The Power Relay
There are 3 installation parameters that must be planned for before we can
purchase the relays. We have to avoid over amperage. The current must be less
than the SSR rating. We also have to avoid over voltage and over temperature.
Solid state relays have an isolation feature. The control signal and power line is
completely isolated by an internal LED.
1-6 Design Constraints
We have to consider some constraints when designing the system.
constraints are derived from our assumptions and limitations:

These
Size: The PCB and the LCD must be practical for hands on use. The
physical characteristics that are constrained are that of the size and
16




weight. However, since the weight of the design components are fairly
light, weight should not be a big issue. The final product will be light
enough to be carried easily by hand.
Budget: Due to a limited budget, the team must analyze part cost
effectively. It will be important not only to spend as little as possible, but to
find the best part for the money as well. Any sponsor purchased material
would also be an ideal way to keep the budget within reason.
Durability: The system must be able to perform after several uses.
Ease of use: the LCD screen commands should be easy to understand.
The system should have a detailed manual to explain different features
and how they can be used easily
Safety: the system should be safe and reliable. There should be no risk or
injures due to electrical current.
1-7 Previous Projects and Commercial Products
Several products are available on the market for power reading. The first one that
comes in mind is the Kill a Watt. Kill a watt is a small device that tells the user
how much electricity something uses, either at a given moment or over an
extended period of time. The device needs to be plugged into the meter, then the
meter need to be plugged into the wall. This kind of devices sells for around $25.
While this device is especially useful for finding the amount of kWh used in a
month for devices that do not run constantly, it does not offer the user the option
to turn the power off an appliance. The following figure shows a typical kill a watt
device.
Figure 1-7.1: Kill a Watt, photo used with permission from Amazon
There is another class of products similar to the Cent-a-meter which show users
their energy usage on a wireless display. These work by placing a transducer on
the power company meter to measure the power consumption and send the
information to the display which converts it to cents per hour or green house gas
17
emissions as an added bonus, since the metering hardware is placed outside the
home, it also displays temperature and humidity for convenience. These products
do not directly show the power use of a specific appliance or electronic, but can
give an estimate if the user keeps a constant baseline power usage before and
during use of said appliance, noting the difference between the two. The stated
accuracy for this product is 5% which isn't quite all that accurate, but the purpose
of giving a user an idea of how much energy they are using it is close enough.
Recently there have been products introduced that are low cost metering devices
which send the monitoring data to Google PowerMeter which allows users to
view realtime power usage data from their iGoogle homepage with visualization.
This is a great example of what an easy to use UI design can be and how it could
be implemented. The fact that it is an open program that anyone can use for free
with the given hardware installed makes it a strong contender for use.
The Google PowerMeter, iGoogle UI
Permission granted from google.com
In addition to the commercial products there have been many similar senior
design projects, especially on the power metering subject. While researching
differing products, the group also came across an interesting project, called the
Tweet-a-Watt, that uses a Kill-a-Watt in conjunction with an xbee radio to send
measurement data to the a computer. The computer then runs a google app that
uses googles visualization engine to display the resulting power usage over time
for each individual outlet. This is very similar to what the the groups project and
worth noting. The difference is that the group plans to use an embedded system
18
to visualize the data on a touchscreen where the tweet-a-watt just sends it to the
users computer for collection and visualization.
UI for the Tweet-a-Watt using googles visualization engine
Permission granted from adafruit.com
A similar project with somewhat similar goals was completed by students at the
University of Illinois at Urbana-Champaign. Their project was similar in that it
metered the power usage and then wirelessly sent the information to a central
hub which was connected to a computer where they used a program to visualize
the data. The group was pleased with their project overall but didn't like some
aspects of the final product. One of their first complaints was the wireless
scheme they were using received a lot of interference and also had a limited
range of 30 ft. Because of the way they were encoding the data even minimal
interference could corrupt the data being sent over the air. Also, due to their
communication protocol, they were only able to use the 8 most significant bits of
their 10 bit Analog to digital conversion thereby losing accuracy. This is
something that they wished they could change and if doing the project in the
future they most certainly would do differently.
A few other projects were reviewed and most of them used technology discussed
in the research portion, using some combination of a voltage sensor, current
sensor, wireless or powerline communication and a more powerful processor
(usually a computer) to visualize the data collected. Many projects opted for a
graphical display of the power which is something that the group is designing for,
which isn't always the case with standalone units.
The group has found a similar senior design project. The design is the “The
power meter” from the university of Central Florida during spring 2010. The
objective of the project was to design and construct a power meter for the
household. The power meter was intended to be used to measure the power
consumption of any electrical device in the home that is connected with the use
19
of a plug. The group had set an accuracy requirement of at least 5%. This
project was implemented and test with an accuracy of 5%.
While the general idea behind the unit seems to be somewhat common, as there
are plenty of products that do similar things, and plenty of very similar senior
design projects that have been completed in the past, the group feels that each
one, including this design is different and each has its own merit, and the small
decisions that differentiate one project from the next are ever important.
1-8 Safety Concerns
The design is centered on many electrical devices. These devices can pose a
significant hazard to the team members or the used, particularly when
mishandled or not maintained. We will be measuring high AC current and
voltage. Due to the sensitivity if the design, we find it appropriate to discuss some
safety concerns and address how are we going to avoid electrical hazards. The
major hazards associated with electricity are electrical shock and fire. Electrical
shock occurs when the body becomes part of the electric circuit, either when an
individual comes in contact with both wires of an electrical circuit, one wire of an
energized circuit and the ground, or a metallic part that has become energized by
contact with an electrical conductor. The severity and effects of an electrical
shock depend on a number of factors, such as the pathway through the body, the
amount of current, the length of time of the exposure, and whether the skin is wet
or dry. Water is a great conductor of electricity, allowing current to flow more
easily in wet conditions and through wet skin. The effect of the shock may range
from a slight tingle to severe burns to cardiac arrest. The chart below shows the
general relationship between the degree of injury and amount of current for a 60cycle hand-to-foot path of one second's duration of shock.
Current value
Reaction
1 milliamp
5 milliamps
6-30 milliamps
50-150 milliamps
1-4.3 amps
5 amps++
Perception level
Shock felt, not painful
Painful shock
Extreme pain, respiratory arrest
Ventricular fibrillation
Severe burns, cardiac arrest and
possible death
While it is not likely that any safety issues will arise from our project, we find it
necessary to point some precautions to prevent electrical hazards such us:


We will inspect the wiring before each use and replace damaged electrical
cords immediately
We will use safe work practices every time the equipment is used
20

We will make sure that each team member knows the location of the
switches or circuit breaker panel in the senior design lab
21
2- Research
2-1 Methods
2-1.1
The Current/Power Meter
The main device of the design consists of the power, current and voltage sensor.
This part of the design is critical because all the other steps rely on it. If the
power or current measured are not accurate, the microcontroller will send out
incorrect values. Therefore, the displayed values on the LCD screen will be
incorrect, the measured power consumption and expenses will be inaccurate and
the whole system will be useless. The power sensing circuit need to provide
accurate calculations and has to measure currents from 0 amps up to 15 amps.
AC power is used in nearly all homes in the United States. This is because the
power companies need to send their electricity over long distances and
compared to DC power, AC power travels farther. The standard voltage of AC
power in the United States is 120V. The current sensor we need to use must be
able to detect AC current from 0 amps up to 15 amps. The current signal must be
fed into a computer in which sensors convert current into a proportional voltage
with minimal influence on the measured circuit. The oldest technique is to
measure the voltage drop across a resistor placed in the current path. To
minimize energy losses the resistor is kept very small, so the measured voltage
must be highly amplified. The amplifier’s offset voltage must be as small as
possible and its supply voltage must be at the potential of the circuit, often 110 V
to 120 V mains with high parasitic peaks from which its output must be isolated.
This requirement increases overall system cost.
In order to measure AC power, we need to keep in mind the important of the
power factor. As in the case of DC power, the instantaneous electric power in
and AC circuit is given by voltage times current. However, these two quantities
are varying continuously. Therefore, the measured AC power will actually be the
average power given by 𝑷 = 𝑽 ∗ 𝑰 ∗ 𝒄𝒐𝒔 (𝜽), where θ is the phase angle between
the current and the voltage and the current and voltage values are the RMS
values.
A microcontroller will be the center of our system as it interacts with all the parts.
The current being measured as well as the voltage across an item will be fed to
the microcontroller. The current and voltage fed to the microcontroller need to be
within the allowed range of current and voltage. A voltage divider will be used to
reduce the voltage by a given factor. The same applied to the current.
There are numerous other methods used to measure current and power. We will
try to cover a variety of methods and study each one of them separately. We will
identify the advantages and disadvantages of each method and choose the one
22
that works best for our application. The most common ways of measuring current
include the use of a resistive shunt, a transformer, a magnetic current sensor or a
Hall Effect current sensor. We will try to evaluate each one of these sensors and
decide which one works better for our application. Each current sensing method
has its advantages and disadvantages.
Resistive shunts operate by Ohm’s law giving a voltage proportional to the
current going through the shunt. It is simply a resistor in series with the load. This
offers accuracy and low offset, but it does not provide electrical isolation and has
a high thermal drift. This allows the transient spikes to eventually ruin the sensor
and potentially overload the electronics. Another issue with resistive shunts
would be heat. Like any other resistor, a resistive shunt can generate a
significant amount of heat. Heat is not a desirable effect and could cause parts
damage.
Current transformers are made up of a primary and secondary coil wrapped
around a magnetic core. The primary coil carries the current to be sensed and
induces a magnetic filed in the core. A current is generated in the secondary coil
that is proportional to the primary current scaled by the number of turns ratio. The
current transformer offers isolation but it only works for AC application. Current
transformers can also be large and bulky.
2-1.1.1Current Measuring Shunts
Current shunt resistors are low resistance precision resistors used to measure
AC or DC electrical currents by the voltage drop those currents create across the
resistance. The principle of current measurement uses Ohm’s law. The voltage
across the resistor is simply the product of the resistance and the current. For
instance, a current shunt whose resistance is 0.001 ohms having a current of 15
amps flowing through it will produce a voltage of 15 mill volts.
Figure 2-1.1.1.1 shows a typical current shunt:
Figure2-1.1.1.1: Current Sensing Shunt. Permission asked from Galco.com
23
A perfect current shut has the exact resistance claimed. Its resistance does not
change with temperature, time or current. Its inductance or resistance to AC
current change is zero. This kind of perfect shunts are usually very large and way
too expensive. However, they do have several advantages such us:



Specific current rating, for example 50 Amps or 100 Amps
Resistance accuracy, usually within plus or minus 0.25%
Resistance drift, the shunt’s resistance changes by less than 0.002% per
degree C of temperature change
Current shunts are resistors. So by nature they dissipate heat from the current
flowing through them and they get hot. Since the heat can change a shunt’s
resistance and eventually damage the shunt, current shunts are often given a
power rating or a rating factor. The heat produced is power measured in
Watts:𝑾 = 𝑰^𝟐 ∗ 𝑹. Therefore the heat produced increases with the square of
the current. So doubling the current increases the heat power dissipated by 4
times. This might create other issues and ultimately damage some parts of the
circuit if not all.
2-1.1.2 Current Transformer
Current transformers consist of two coils wrapped around a magnetic coil. The
theory behind them is very simple. When an electric current flows through a wire,
it generates a magnetic field. The magnetic field strength varies when the current
varies, so the bigger the current, the stronger the magnetic field and the stronger
the magnetic flux density and vice versa. For a time varying current, the magnetic
field generated is time varying as well. A time varying magnetic field creates an
electric current. So, if we put a second wire next to the first one where a time
varying current is flowing, a time varying current will be generated in the second
coil and will be proportional to the first current. That is how a current transformer
measures current. The reaction is called electromagnetic induction because the
measured current in the primary coil induces a current in the secondary coil.
There are also some current transformers that produce a voltage instead of a
current. The voltage is proportional to the current being measured.
The use of current transformers to measure current has many advantages. The
most important benefit of using such a method in our application would be safety.
The current transformer will be safe to use as it isolates the current being
measured from the rest of the circuit. It also controls the circuit and protects it
from the high voltage or current being measured. However, they tend to be large
and would cost more than what the budget of the system would allow. Current
transformers also have another inconvenient of saturation. Because of current
spikes, the transformer cores might get saturated and therefore the measured
current will be inaccurate. This is to avoid as accuracy is a critical point of the
system. Current transformers have many performance specifications such as the
primary current, secondary current and accuracy.
24
Figure 2-1.1.2.1 shows a typical current transformer:
Figure2-1.1.2.1: Current Transformer. Permission received from
midwestcurrent.com
The use of current transformers to measure current has many advantages. The
most important benefit of using such a method in our application would be safety.
The current transformer will be safe to use as it isolates the current being
measured from the rest of the circuit. It also controls the circuit and protects it
from the high voltage or current being measured. However, they tend to be large
and would cost more than what the budget of the system would allow. Current
transformers also have another inconvenient of saturation. Because of current
spikes, the transformer cores might get saturated and therefore the measured
current will be inaccurate. This is to avoid as accuracy is a critical point of the
system. Current transformers have many performance specifications such as the
primary current, secondary current and accuracy.
There are two kinds of current transformers, wounds and toroidal. In one hand,
Wound current transformers are made of an integral primary winding. This
primary winding is in series with the conductor carrying the current being
measured. In the other hand, the Toroidal current transformers are donut
shaped. They do not have a primary winding. They simply have the wire that
carries the current being measure threaded through a window in the transformer.
2-1.1.3 Eddy Current Sensors
Eddy current sensors are non contact sensors that are capable of measuring the
position of a conductive material. They are used to measure the position or the
displacement of a target carrying a conductive material. These kind of current
sensors operate with magnetic fields. The sensor creates an alternating current
in the sensing coil. This current produces an alternating magnetic field at the
probe tip. The magnetic field then creates create small currents in the target
material. These current are called Eddy currents and they generate a second
25
magnetic field that is opposite to the first one. The interaction between the two
magnetic fields depends on the distance between the probe and the target.
Therefore it changes as the distance changes. The sensor can measure the
change in those electric fields and produce a voltage that is proportional to the
change in distance. Figure 3 shows an Eddy Current sensor:
Eddy current sensors have several advantages such as:
 High precision and resolution
 Non contact and wear free
 Insensitivity to dirt
 Suitability for fast applications
 A reasonable price and good performance ratio
Eddy Current sensors use changes in the magnetic fields to determine the
distance to the conductive target. There also some other factors that can change
the magnetic fields intensity. The presence of conductive material can affect the
magnetic field and therefore can affect the measured current. This presents a
potential error that should be avoided in our application as accuracy is an
important part of the system.
2-1.1.4 Hall Effect Current Sensors
Transducers
The Hall Effect is considered an ideal sensing technology. The Hall element is
constructed form a thin sheet of conductive material with output connections
perpendicular to the direction of current flow. When subjected to a magnetic field,
it responds with and output voltage that is proportional to the magnetic field
strength. The heart of every Hall Effect sensing device is the integrated circuit
chip that contains the Hall element and the signal conditioning electronics. There
are many key advantages to using Hall Effect current sensors transducers.
Firstly, they can be totally isolated from another high voltage electrical system
which eliminates many safety concerns. The sensor is just measuring the
magnetic field from the wire. This is a nice advantage over current monitoring
shunt precision resistors which are less safe because they require you to put in
certain safeguards to protect your data acquisition system. Secondly, unlike the
current shunt sense resistor which can have thermal temperature heat
dissipation issues, the Hall Effect current sensor does not get hot. Even when
measuring 50 Amps. The cost, performance and availability make the Hall Effect
current sensors almost ideal for our application. The Hall Effect current sensors
have several features such us:






True solid state
Long life
High speed operation, over 100 KHz possible
Operates with stationary input
There are no moving parts
Logic compatible input and output
26



Board temperature can range from -40 to +150 degree C
Highly repeatable operation
High linearity with no saturation effects
The voltage output of the Hall Effect current sensor is proportional to the current
in the conductor. The sensor monitors the gauss level of the magnetic field
created by a current flow, not the actual current flow. The current being
measured if passed through a flux collecting core that concentrates the magnetic
field on the Hall Effect sensor. The waveform of the sensor voltage output will
trace AC or DC waveform of the measured current. The through-hole design
electrically isolates the sensor and ensures that it will not be damaged by over
current or high voltage transients. It also eliminates any DC insertion loss. Hall
Effect sensors measure the magnetic field surrounding the conductor but, unlike
current transformers, they also sense DC currents.
Figure 2-1.1.4.1 shows a picture of a typical linear Hall Effect current sensor.
Figure 2-1.1.4.1: Typical linear Hall Effect current sensor. Permission
received from HoneyWell.com
The very small output voltage of the Hall element must be highly amplified, and
the sensitivity is temperature dependent and requires adequate compensation.
Also of note is sensitivity to short current peaks in the circuit: according to the
hysteresis properties of the core material, these peaks can cause a static
magnetization in the core that result in an offset alteration of the Hall element.
The two types of Hall Effect current sensors are open loop and closed loop. In
the open loop Hell Effect current sensors, the amplified output signal of the Hall
element is directly used as the measurement value. The linearity depends on that
of the magnetic core. Offset and drift are determined by the Hall element and the
amplifier. The price of these sensors is low, but so is their sensitivity. Closed-loop
Hall sensors are much more precise. The Hall voltage is first highly amplified and
the amplifier’s output current then flows through a compensation coil on the
magnetic core. It generates a magnetization whose amplitude is the same but
whose direction is opposite to that of the primary current conductor. The result is
27
that the magnetic flux in the core is compensated to zero. The principle is similar
to that of an op amp in inverter mode, for which the input voltage is always close
to zero. The nonlinearity and temperature dependence of the Hall element are
thus compensated but the offset remains. Closed-loop current sensors work up
to frequencies of about 150 kHz. They are not cheap, though, and for high
currents they become very bulky.
2-1.1.5 Magneto Resistive Field Sensors
Magneto resistive field sensors use the change of the resistivity of a material due
to a magnetic field to measure the current. These kinds of sensors have the
benefit of small energy consumption and high sensitivity in small intensity
magnetic fields. Magneto resistive field sensors can be compared to resistors
that are magnetically controllable. The angle between the internal direction of the
magnetization and the current flow is the determining factor for the resistance.
The way the magneto resistive sensor is very simple. The sensor detects the
movement of a magnetic structure caused by changes in the magnetic flow. This
causes changes in the magnetic field resistance. The change in the magnetic
field is then converted to an electric variable. The output of the sensor can be
either a current or a voltage which reflects the changes in the magnetic field.
Figure 2-1.1.5.1 shows magneto resistive current sensors from the family
CMS20XX:
Figure 2-1.1.5.1: CMS20XX Magneto resistive current sensor family.
Permission asked from lust-tec.com
Practical magnetic field sensors based on the magnet resistive effect are easily
fabricated. To reduce the temperature dependence, they are usually configured
as a half or a full bridge. In one arm of the bridge, the barber poles are placed in
opposite directions above the two magneto resistors, so that in the presence of a
magnetic field the value of the first resistor increases and the value of the second
decreases. The linearity of magneto resistive sensors is not very high, so the
28
compensation principle used on Hall sensors is also applied here. An electrically
isolated aluminum compensation conductor is integrated on the same substrate
above the resistors. The current flowing through this conductor generates a
magnetic field that exactly compensates that of the conductor to be measured. In
this way the magneto resistive elements always work at the same operating
point; their nonlinearity therefore becomes irrelevant. The temperature
dependence is also almost completely eliminated. The current in the
compensation conductor is strictly proportional to the measured amplitude of the
field. The voltage drop across a resistor forms the electrical output signal.
Magneto resistive current sensors have many advantages such us compatibility
with integrated circuits, high sensitivity and good accuracy.
2-1.2 Power Relay Switch
A power relay is and electrically operated switch. It is designed to withstand the
current of the electrical circuit into which it’s inserted and to cut off the electrical
circuit under load. Many relays use an electromagnet to operate the switching
mechanism. A power relay is going to be used in our application as we need a
device that can cut the power of an appliance. The relay uses an electromagnet
to open and close an electrical circuit. The basic design of a relay consists of an
electromagnet coil, an armature, a spring and some contacts.
When power is applied to the power relay, the electromagnet attracts the
armature. The armature which is held in place by the spring will be pulled in the
direction of the coil until it reaches one of the contacts and therefore closes the
circuit. If the relay is closed per design, then the coil pulls the armature away
from the contact instead of trying to reach a contact. In this case, the circuit will
be open and the power will be cutoff.
We have considered several switching devices. The first device we could
consider for our design would be the TRIAC and use it as a switch. The TRIAC
switch is a small semiconductor device. It is similar to a diode or a transistor. The
TRIAC would be a very inexpensive device and it offers good switching power
consumption for non reactive loads. However, it has few cons such as the
requirement of additional circuitry if reactive loads are required to be switched on
and off and it also has a higher series resistance than conventional latching
relays. The TRIAC has two terminals, which are wired into two ends of the circuit.
There is always a voltage difference between the two terminals, but it changes
with the fluctuation of the alternating current. The TRIAC acts as a voltage driven
switch where the voltage on the gate controls the switching action and a variable
resistor controls the voltage on the gate.
Another method we could consider would be the solid state relay switch. The
solid state relay is an ON/OFF is a control device in which the load current is
conducted by one or more semiconductors. We can consider the solid state a
TRIAC with embedded additional circuitry to turn on and off active and reactive
loads. While the solid state relay requires low control circuit energy to switch the
29
output and it is also easy to control, it has some disadvantages such as high
price, high series resistance and high power consumption compared to the
latched relay switch.
Another way we could turn the power off a device will be using a mechanical
switching relay. The mechanical latching relay provides good series resistance
and is easy to control. The most important benefit of using this kind of relay is
that fact that its power consumption can be manipulated to be zero. A main
disadvantage is that it requires extra circuitry and a different algorithm to switch it
on and off. Also we need to be careful not to short the circuit with control signals
due to its bridge driving circuitry.
When we think of a power relay, we think of a switch. We considered a simplified
circuit to show how a relay would work. It is simply a circuit that turns on and off
the power. The following circuits illustrate how a relay can be represented in both
the on and off state:
Figure 2-1.2.1 shows a simplified circuit of the ON and OFF state of a power
relay:
Figure2-1.2.1: equivalent circuits for ON state (b) and OFF state (c).
Permission asked from omega.com
2-1.3 Printed Circuit Board
The design calls for the use of a printed circuit board. The printed circuit board is
generally made of one or more layers of insulating material with electrical
conductors. The insulating material could be anything of the base of ceramics,
plastic or other dielectrics materials. This part of the project consists of simply
ordering the right printed circuit board. There are many vendors and the prices
are reasonable. The vendors also provide the free software to create the circuit
board layout. Besides having the printed circuit board fabricated with minimal
flaws, the design allows us to define and design multiple layers. The use of
multiple layers in our printed circuit board will assist in the complexity of the
30
circuit. We intent to use either a 2 layers PCB or a 4 layer PCB. A 4 layer PCB
will allow us to have planes designated just for power and ground. This will help
reduce the complexity of the routing especially that none of the group members
has any soldering experience. The 4 layer PCB will also help in reducing the
inductance and impedance across the board and will prevent ground loops from
happening.
2-1.4 Communication
There are two basic forms of communication that could be considered acceptable
to be used for this type of project – powerline and wireless communication. The
biggest, and most obvious, difference between the two is the wired versus
wireless aspect. Of course, the inner workings of each vary greatly, which
creates an extensive variation in the internal categories of each option, thereby
providing several alternatives for data transmission.
The first technology that was looked into was powerline communication.
Powerline communication is merely using the existing power infrastructure in a
house as the communication path, which would require no extra wiring to be
installed for use. Within powerline communication, there are a few major
methods and protocols that are currently in use. The most important systems
include, but are not limited to: X10, Universal Powerline Bus, and the Homeplug
Standard. However, each of these technologies may be affected by noise on the
powerline which can be introduced by home appliances, such as vacuum
cleaners, blenders, televisions, and other such devices. Below is a figure
depicting how devices can be connected over the powerlines within a residence
without any wiring other than the already installed power structure.
Figure depicting powerline communication capabilities.
Permission asked from embedded-system.net.
X10 is a fairly simple protocol that has been in use for many years. However, the
X10 has a fairly low data rate at 20 bits per second, and the signal has a
31
reliability rate of 80%. Due to the possibility of a lost signal, the protocol sends
each signal twice, which contributes to the lower data rate. X10 uses a house
code, unit code, and a four bit function code to control devices. These signals are
sent using a 1 millisecond burst of 120 kilohertz at the zero crossing point of the
60 hertz power signal. Ones are represented by a burst, immediately followed by
an absence of a burst, and a Zero is signaled oppositely – an absence followed
by a burst with an expected absence. All signals are sent twice, in order to
correct for false signaling. Due to the double signaling, the data rates are very
low which makes X10 solely confined to turning devices on or off. Even such
functions as creating a “scene,” turning on lights and controlling the dimming in
one swoop, are beyond the reach of X10. Below, the table of codes displays the
actions the X10 is able to perform in one signal.
Code
Function
0 0 0 0 All units off
0 0 0 1 All lights on
0010
0011
0100
0101
On
Off
Dim
Bright
Description
Switch off all devices with the house code indicated in
the message
Switches on all lighting devices (with the ability to
control brightness)
Switches on a device
Switches off a device
Reduces the light intensity
Increases the light intensity
Shortened table of X10 Codes
The Universal Powerline Bus is another technology that is used primarily for
home control. Unlike X10, it has a much higher successful data rate. Due to the
way that the signals are sent and received, UPB’s transmission and reliability are
considerably more accurate than X10. UPB uses precisely timed pulses
superimposed on the AC sine wave. These pulses are created by charging a
capacitor to a very high voltage and discharging it at a specific time. The signal
can be picked up over a large distance due to the high voltage spikes. The
Universal Powerline Bus protocol specifies a window called the UPB Window
wherein the value can hold either 0, 1, 2, or 3. This method of communication is
not confined to the Universal Powerline Bus, but is widely used in digital
communications and is called the Pulse Position Modulation. Universal Powerline
Bus can achieve higher data rates, because each ‘bit’ of sent data actually
contains two bits. Each ‘bit’ can be transmitted on every half cycle of a 60 Hertz
power signal, allowing it to achieve 240 bits per second which makes it ideal for
home control.
32
Description of PPM implemented in UPB
Permission asked from pulseworx.com
Homeplug is a standard that is used for broadband communication over
powerlines. While the technology is worth mentioning, the specifics weren’t
considered as much as the data rates that Homeplug was capable of processing
were well over the needs of the project. This standard is included to provide a
complete look at the main powerline communication technologies, as well as the
viability of powerline communications for higher data rates.
As previously stated, the other viable way to transmit data would be to do so
wirelessly. As technology in the modern world grows, so do the vast and various
ways to communicate wirelessly. Research into the subject showed that there
were many different methods on the market today.
A company called Nordic Semiconductors produces a unit, called the RF24L01+
which was carefully analyzed for the needs of the project. Of course, the
products typically found within a communication device were also considered –
cellular, Bluetooth, Zigbee, and WiFi. All of these mechanisms are able to
communicate wirelessly, are already a part of the common knowledge of the
population, and are generally small enough to fit within the scope of the project.
The aforementioned device, Nordic’s RF24L01+, was quite promising in
research. The cost is very low, much lower than most technologies. It uses very
little power, which is why it is used in many long-term battery-powered wireless
devices, such as the Nike+ hardware which communicates with an iPod. This
wireless radio can reach accurate data rates of up to two Megabits per second,
which is much more capable than either of the powerline communications. The
33
Nordic part is extremely flexible – it is able to transfer a small amount of data
between a sensor and a computer, or even create a mesh network. The mesh
network, however, isn’t supported out of the box, and the user would have to
write their own mesh network protocol for the units. The Nordic part is a great
base to build off of and retains a very small footprint, being less than the size of a
quarter (as seen below.)
Nordic Radio, about the size of a quarter.
Permission acquired from sparkfun.com
Another wireless technology that is quite ubiquitous is cellular. Cellular is
everywhere – it reaches most places indoors, which would make it a useful
resource for this project. However the radios themselves are expensive, they
consume a lot more power than the project demands, and to move their data the
homeowner would have to pay per kilobyte sent. These disadvantages of cellular
go against the goals of the project, which includes reducing power consumption
and creating a viable, inexpensive alternative for homeowners to reduce power
usage and expenditures.
In terms of size, cellular radios are far from the smallest of the technologies
researched, having four times the footprint of the Nordic Radio. If the project
required that the power measured be in a remote location, far from the display
unit, this technology would make sense for the needs of the project. As it is, that
isn’t the case, so cellular radios aren’t really considered a contender as it doesn’t
meet the needs of the project.
Zigbee was the first technology considered by the group as a wireless solution.
Zigbee is already used in home control and monitoring, which lead the group to
further examine this interesting wireless communication device. A great
advantage to using the Zigbee technology is that it allows for the creation of
mesh networks, pictured below.
34
The ability to create a mesh network would allow for subsequent units (nodes) to
be out of the initial range of the display (parent) unit without halting
communication. The nodes would be able to send and receive data, passing
along the information received from outer nodes to the parent unit. Researching
into applications of Zigbee showed that the technology is difficult to implement
and usually leads novices into a mire of problems, which ends up causing more
trouble than it is initially worth. However, while researching Zigbee, the group
came across a product that implements the basic, attractive Zigbee advantages
in an easier to use and implement product, called Xbee.
Xbee is a wireless solution that is produced by Digi, which was based on the
Zigbee Protocol. Digi has created a radio that implements all of the best of
Zigbee without the mess. For instance, it uses an AT command set which makes
the Xbee much easier to implement directly out of the box. The latest version
supports mesh networking, though this requires much more effort and
implementation for the user. However, unlike Zigbee, the Xbee allows the group
members to take full advantage of their knowledge to create a reliable mesh
network without going beyond the capabilities expected.
As well, the range for the Xbee unit is perfect – 100 meters is described as the
short range, which allows for attenuation through the walls of a house yet would
still cover almost any normal residence’s width or length. If not, the mesh
networking would make up for any loss of signal. Compared to the Nordic radio,
the Xbee unit is much more expensive and uses more power, but it achieves
higher accurate data rates ensuring the user’s information gets from point A to
point B. While the Nordic radio has a strong battery life, it is assumed the
project’s final device will be connected to an AC power line at all points, which
makes the expected battery life of the Xbee null and void.
WiFi is a great technology for interoperability, with just about anything; it is used
in communication devices from the Xbox 360 to the cellphone, to the iPad. For
35
this project, however, this is the last and only positive advantage to using WiFi. If
WiFi were used, it would allow for an entire house network, but it would use a lot
more power than any other technology and would require a better
microprocessor in the sensor units. The new microprocessor would increase
power consumption of the whole project, thereby undermining the goal of
reduced electricity usage. The data rates that can be achieved over WiFi also
aren’t needed in this project, thus WiFi wasn’t researched further as preliminary
research showed it wasn’t a plausible alternative.
2-1.4 Microcontroller
While researching microcontrollers, a few different schemes were found that
could be used. The first one considered was from a dedicated power metering
circuit designed by Analog Devices, and two others were found while researching
possible wireless technologies. The main uses for a microcontroller in this project
are to control the flow of data to and from the sensor units and display unit,
meanwhile performing any calculations on and converting the analog signals
from the metering components. The group also sought a microcontroller that
would be programmable in the C programming language. The three
microcontrollers researched are the Microchip PIC 16C67, the Arduino family
using the ATMega 328, and the ZB1 board with support for the Xbee wireless
radio.
From a microcontroller standpoint, the Microchip PIC 16C67 was considered a
suitable scheme as it provided an entire system on a single board. The Analog
Devices circuit was designed to allow the power company to meter a residence’s
home energy usage, using a seven segment LCD display. Which the MicroChip
PIC 16C67 could be programmed in C, it was already fully designed and required
the use of the AD7755 part to be functional in the given state. While the part
excelled in all of the microprocessor aspects the group sought, the
disadvantages of the power metering had to also be taken into account. Below is
an image of the Microchip alongside the required circuit it needs to be functional.
36
PIC 16C67 and AD7755 Circuit.
Permission asked from AnalogDevices.com.
The ZB1 microcontroller board was also considered, since it was recommended
to be used alongside the Xbee unit. In fact, this board is already interfaced with
the Xbee unit so that part of the design would already be taken care of – a cause
of concern for the group since the group sought to design this part as well.
Another disadvantage of the ZB1 microcontroller was its size. The board took up
4.4” x 1.75”, which was too large for implementation in the final unit. This board is
very similar in design and use as the Arduino microcontrollers, featured later.
Unlike the Arduino family, however, there is only the one model offered of the
ZB1 which doesn’t provide the scalability that Arduino has available.
The Arduino family of microprocessor boards has everything one could possibly
want in one convenient package. The Arduino family offers several analog inputs,
twice as many digital outputs, and interoperability with various products. One
concern in using the Arduino family is the availability of a C programming
language. Conversely, the Arduino family has a development environment for
Windows, which uses a modified version of the C programming language, with
additional specific commands for controlling the pin functions of the digital pins
and the special multi-use pins.
2-1.5 Touch Screen
There are many types of touch screens available. Care must be taken to choose
the correct one for this application. The first type, called resistive touch screen
are created by applying two sheets coated with resistive material separated by
an air gap. One sheet will have vertical lines, and the other will have horizontal
lines. When the user touches the screen, the location is registered. Resistive
touch screens work well for this application. They have the advantages of being
used with styluses and pens rather than only fingers in addition to being relatively
inexpensive. Resistive screens are in general low power, and are widely
available in many sizes.
37
A second type of touch screen is surface acoustic wave. This operates by the
device sending ultrasonic waves across the surface of the device, and when the
panel is tapped, the energy of the wave created is reduced, and the device
registers a hit as shown in figure 2-3.1.1. Surface wave screens are interrupted
by outer forces such as loud sounds, and rely on the screen to be mostly clean,
otherwise the waves are dampened, and the device fails.
Figure 2-5.1.1 permission requested from www.touchuserinterface.com
Another type of touch screen is the capacitive touch screen. Although there are
many types of capacitive screens, they all operate in a similar way. A surface
capacitive screen is created by taking a thin insulator, and coating it with a
transparent conductor, as shown in figure 2-5.1.2. To use the screen, another
conductor with energy, such as the human body, can touch the screen, which
results in a change in the magnetic field of the screen. This change is
measurable as capacitance. Projected capacitance is a technology where a
conductive layer of electrodes is etched into a screen. These are extremely
small, and have a high resolution of nodes compared to other types. These
nodes are placed on the back of the glass substrate in use. This allows
progressive capacitive screens the ability to the used without direct contact.
Another type of touch screen is the mutual capacitance screen. Here, an array of
capacitors is placed in a grid of a reasonable size across the screen. Now,
whenever a part of the screen is pressed, the capacitor is discharged, and the
location of the touch can be found. Capacitive screens can only be used by
conductive materials, which is a positive since they can be cleaned with a cloth
without any interruption. If a stylus will be used, a conductive one will be needed.
Capacitive screens are in general more responsive than resistive screens, but
are more costly.
38
Figure 2-5.1.2 permission requested from www.screentekinc.com
Infrared touch screens use a grid of infrared LED and photodetector pairs around
the outer edge of a screen. These can detect interruptions in LED beams which
cross each other in perpendicular patterns. By finding the X and Y LED beam
number, the exact location can be found, as shown in figure 2-5.1.3. This system
can be activated by anything that can interrupt a beam. Thus, they are popular
as point of sale kiosks in restaurants where users do not always have a
conductor available. Unlike other types of touch screens, infrared screens do not
require any modification to the glass used, and thus increases the clarity of the
glass over the long term.
Figure 2-5.1.3 permission requested from www.planarembedded.com
39
Touch screen monitors are in reality any LCD monitor with a capacitive or
resistive overlay. Thus, the brand or quality of the monitor itself is not a large
issue. Because of the scope of the project, a small 8 to 10 inch monitor would
work very well whereas a larger monitor would be bulky and unnecessary. The
quality of the screen must be great enough to be left on for a range of 5 minutes
to an hour without burning an image. In addition, the screen must be able to go
into a sleep mode for when the device is on, but the screen is not needed.
2-1.6 Surface Mounting Methods
The next part to discuss would be how we are going to surface mount the parts
into our circuit boards. There are basically two methods to do so; the old method
is the through-hole technology. In the through-hole technology, the components
are placed on one PCB side and soldered on the other. The other method to
approach this is the surface mount technology. In the surface mount technology
the components can be assembled on both sides of the board. The components
are attached to the PCB by solder paste or non conductive glue and then
soldered.
2-1.7 Software
The software goal of the project is to create a touch screen environment which
will display graphs of time versus kilowatts per hour, with menu options to
navigate and modify the software using touch. This action requires the touch
screen itself, a processor for the operating system, and the remainder of the
computer system. Because of this, an operating system will be required.
For this project, the operating system needs to be lightweight enough to operate
in the small amount of memory that will be available, and yet flexible enough to
issue the required commands. Thus, an embedded operating system is needed.
An embedded operating system is like a standard operating system like Windows
or OSx, but it is more compact, and lacks many features commonly seen today
such as a complex user interface. These embedded operating systems fall into
the category of a Real Time Operating System, or RTOS.
Many different brands and types of RTOSes exist. The embedded system for
this project will run on a version of embedded Linux. With an embedded Linux,
we can have a stable environment to run the specified dynamic code along with
the freedom of no licensing fees or royalties. Support exists for embedded Linux
on the ARM processor. Embedded Linux only requires about 100 kilobytes of
memory to operate all basic system functions. If networking and stack utilities
are added, 500 kilobytes of memory is required.
A GUI is needed for the project. Ideally, the GUI will be minimalistic where there
exists enough relevant information on a page, but there is no clutter or
unnecessary graphics or data.
40
In a conventional GUI graphics model, mouse clicks or keyboard strokes create
events that are continuously checked by an application. Thus, it is an event
driven process. These events are sent to the focused window, and are
transferred to procedures related to the window for processing. The procedures
to do so are defined by each application or a standard procedure. The window
procedure for handling events identifies exactly one window class, where
windows with the same window procedure are considered to belong to the same
window class.
Many GUI design tools exist. One such product is GEMstudio. GEMstudio was
created by Amulet Technologies and is an industry leader for interface solutions
for embedded products. An advantage of GEMstudio is that it handles all LCD
and touch screen functions, so the microprocessor does not need to. The system
uses a proprietary LCF module with a 480x272 color TFT LCD, with backlighting
and a clear analog touch screen on a controller board with Amulet’s GEMcompliant ASIC. The system uses USB to connect to the programming
computer. The system runs on a 5V power supply. The downside to this GUI
creation tool is that the screen used must be the one provided from Amulet
Technologies. This would require us to create an embedded system and
processor to support the other functions of the device. Thus, this solution is not
ideal.
Qt software also has a branch dedicated to embedded system GUIs. Qt builds
on the standard API for devices with embedded Linux devices. It uses its own
compact window system and writes directly to the Linux framebuffer. This is the
essence of an embedded GUI.
The ideal GUI creation library found is called MiniGUI. It is open source, but
does not require a license. As taken from the introduction on minigui.org,
“MiniGUI, is one of the world famous free software projects. MiniGUI aims to
provide a lightweight graphics user interface (GUI) support system for real-time
embedded systems.” The touch screen GUI will be modeled using the MiniGUI
framework in C++. MiniGUI has a vast amount of support for projects like this
one, specifically including the ARM processor. This will assist in creating the
menus the user will navigate. Features include a multi window mechanism
where multiple windows can be open at a given time. This would be preferred
because the requirements state multiple windows of real time graphs. MiniGUI
supports widget controls such as multi line edit boxes, labels, and animations in
addition to many fonts, and natively offers support for touch screen calibration. It
can support many image file types such as GIF, JPEG, and BMP. MiniGUI can
be run in three distinct run modes. The first is the “MiniGUI-Threads” which can
allow a program to create multiple cascading windows, and allow all of the
windows to belong to a single process. “MiniGUI-Processes” causes ever task to
be its own process. This setting is the most suitable for embedded systems.
“MiniGUI-Standalone” is the single process mode where only one process is
active at a time.
41
MiniGUI uses only 700KB of flash memory and 1MB of RAM. This will allow
space for other aspects of the software. The size of the MIniGUI library can be
reduced to 500KB or less if necessary. The many functions of MiniGUI allow it to
be used in low in products with CPU frequencies as low as 30MHZ and up to
high – end products. The windows can be skinned however the creator chooses.
It’s cross-operating system features makes it simple to run MIniGUI on most
embedded opertating systems, and allow complete access to multi-window
systems. The GUI system can be configured to satisfy different requirements of
embedded systems. It has many configuration options, and it can be designated
to include or exclude certain libraries to include or exclude features the developer
may or may not be using. Thus, MiniGUI has strong scalability, which allows it to
be applied from simple devices, to complicated electronic systems. The
architecture and optimized graphics interfaces of MiniGUI allow for an extremely
fast graphics output. As mentioned before, MiniGUI was designed for real-time
systems, and basics such as compactness, high efficiency, and high
performance were in mind from the beginning of its development.
The relationship between MiniGUI and other parts of the operating system is
shown in figure (figurenumber). A MiniGUI application can use its functions by
calling APIs from the ANSI C library and the MiniGUi libraries. The portable layer
prevents the applications from taking care of input and output devices by hiding
the details of hardware and the operating system.
Figure 2-1.7.1 permission requested from MiniGUI.org
Programming in MiniGUI begins with calling at least the following five header
files for any GUI. The first is common.h. It includes the definitions of macros and
data types most commonly used in MiniGUI. The second is minigui.h. . In this
header, we include definitions of global and general interface functions, among
various other functions. The third is gdi.h. This header file includes the
definitions of interfaces for the MiniGUI graphics functions. Window.h includes
the definitions of macros, data types, and data structures, which are relative to
42
windows and declarations of function interfaces. The final header file, control.h
includes the definitions for all of the built-in controls in the MiniGUI library.
The initial user interface of any MiniGUI application is a main window. This
window can be create by calling the CreateMainWindow function. The argument
of CreateMainWindow is a pointer to the MINWINCREAT structure, and the
return value is the handle to the created main window.
Menus can be created with the CreateInfo.hmenu command. This sets up a
menu with a value of 0 indicating no menu, and 1 indicating that a menu exists.
CreateInfo.hCursor = GetSystemCursor(0) ; will return the default cursor of a
system to the location of the cursor.
Three main libraries in MiniGUI exist. Some of these may or may not be included
depending on the use of the application. The first is the core library, libminigui. It
includes the basic set of instructions needed for most GUI applications. The
other two are libvcongui and libmgext. Libmgext includes controls and graphics
for user interface routines such as “open file” dialog. Libvcongui is a virtual
console support library.
MiniGUI has three window types, main window, dialog box, and control window.
Each application creates a main window. In each main window, there are
several child windows, and the child windows are control windows or userdefined window classes. In addition, MIniGUI can be used to create a dialog
box. These boxes are actually main windows, but prompt the user to perform an
input operation within the dialog box before it will continue.
DestroyMainWindow will destroy a main window, but will not destroy the
message queue of a main window, or the window object. The function,
MainWindowCleanup will destroy the message queue and the object.
The graphs will be created using GNUPlot, a lightweight graphing utility for Linux.
This software is copyrighted, but freely distributed. GNUPlot was created for
scientists and engineers to freely view graphs and figures, and has support for
interactive screen terminals. Much documentation exists for GNUPlot on the
internet, as it has extensive documentation and support containing tutorials,
premade scripts, and other helpful documents.
GNUPlot supports most
mathematical expressions that would be available in C, FORTRAN, Pascal, or
BASIC programming languages, while ignoring white space inside expressions.
GNUPlot uses real and integer arithmetic where one can be chosen above the
other. This will be helpful in finding exact numbers. Another helpful factor in
GNUPlot is a numerical value is promoted to a integer or real value when used in
an expression. For example, “4”+”5” == 9 would be true. Graphs are created
with the “plot” function. Plot allows for the axes to be set with names as options,
and allows the user to set what data will be the source for the graphs. The
documentation warns that floating point exceptions are possible in GNUPlot.
Whether or not the system will ignore or crash the application is dependent on
43
the compiler and runtime environment.
inspected before they are used.
So, any functions will have to be
GNUPlot’s time counts wrap at 24 hours, and are accurate to 1 second. This
project will not be handling times of less than 1 minute, so there is a surplus of
breathing room in the calculations of time.
It would be preferred that all programming be done in the same language. The
current programming convention is object oriented coding. For the maximum
control and compatibility, it was found that C++ would be ideal. Java was also
considered due to its user friendliness, but the C++ libraries are not compatible.
2-2 Components
2-2.1Current Sensor and Power Relay
a Hall effect current sensor will be used to sense the current. We chose this kind
of sensors over the other technologies because of safety and reliability
considerations. As discussed earlier, the Hall Effect current sensor will be totally
isolated from the high voltage electrical system which eliminates many safety
concerns. Also, the Hall Effect current sensor will not get hot while we are
measuring the current. The cost was also one of the considerations; the Hall
Effect current sensor should not cost us more than $20 per unit. When it comes
to the power relay, we decided to use a solid state relay over an
electromechanical relay. The reasons behind our choice are the fact that the
solid state relay is smaller that electromechanical relays. This will save us space
in the PCB. Another reason is that the solid state does not have any moving
parts, which offers improved reliability. The other reason is the fact that a solid
state relay does not have contacts, so we do not have to worry about wear
issues. The relay will basically last forever as long as it’s used per the
manufacturer requirements. The cost was also considered.
2-2.2 Wireless Connectivity Module
The design of the project requires many things from the sensor units. First, the
units must be able to communicate with the display unit without additional wiring
installation. As well, they must be able to be installed anywhere in a house
without specific consideration of where the display unit is placed. Finally, the
sensor units should operate with a low power consumption since that is part of
the final goal for the project. These three constraints allowed the group to
examine many different technologies, and eliminate the differing data
transmission technologies to come to the decision of which one to use.
As the group first considered the aspect of wireless versus power line
communication, there were many advantages and disadvantages to the multiple
types which were previously described. However, the group decided that they
44
wanted to use wireless communication in the project, to provide the safety of
knowing that the signal will make it to the receiver uncorrupted by the noise on a
power line. As well, wireless communication would offer extensibility with it being
applicable without any consideration for whether the house is on a multi-phase
system or not, which interferes with power line communications. Having make
that initial decision, the group turned to the different wireless communication
technologies there are to choose from and sought the perfect type of
communication product for this project.
The group was able to initially eliminate Cellular and WiFi, as they both are
outside the scope of the project and those communication technologies are not
needed for the design. This preliminary exclusion left Bluetooth, the Nordic part,
and Xbee for further research and discussion. While Bluetooth is quite useful in
many ways, the group felt that the range provided by the lower power Bluetooth
modules wasn’t right to meet the needs of the project. While the Bluetooth can
create point to multi-point networks, and even mesh networks, the complexity to
make that happen with Bluetooth modules is beyond the scope of any of the
group members’ combined knowledge.
Without Bluetooth, the two remaining components to choose from are Xbee and
the Nordic part. Both wireless communication technologies are able to create
mesh networks, so this was not the deciding factor between these two devices.
The range and data rates put the Xbee ahead of the Nordic part, although the
Nordic device won out in the fields of cost, size, and power usage. Another point
that put Xbee ahead was the fact that the data transmission is reliable where the
Nordic part doesn't always do a good job of making sure the data is received,
which is more important to the group members over the cost.
In the end, the Xbee wireless radio will be used in the project, allowing for mesh
networking, reliable communication, more than sufficient data rates, and
interoperability with the chosen microcontroller.
Size of Xbee wireless radio
Permission asked from ____
The choice was based on several considerations such as:
 Simple 3.3v UART interface that should work with most microcontrollers
 reasonable cost ( around $17 to $25 per module)
 40 m to 90 m range (depending on specific model, we can use multiple
models depending on which room we are in)
45




Extendable range, we can purchase longer range units that will cost a bit
more and seamlessly interface with the lower range units
Low power (Less than 40 mA at 3V), they also come with adjustable
power so the unit may be put in low power mode
Availability of the Development kit available (for around $180 to $230).
The kit will include 2 development boards (1 usb, 1 serial) – 2 Xbee-Pro
modules and other accessories
Availability of the manuals on Digi’s website
2-2.3 Microcontroller
The group’s initial idea in the component selection was to pick a microcontroller
which would allow them to design the power metering side of the circuit, and
would preferably be able to work with a C-language programming.
This idea allowed the group to very easily decide that the PIC16C67 was not the
microcontroller they wanted, because they did not want to rely on the AD7755 to
measure the power. The group did not want to leave the power metering side in
the hands of the microprocessor manufacturer. This leaves the ZB1 board and
the Arduino family of microprocessor boards.
The ZB1 board was carefully considered due to it’s similarity in structure to the
Arduino family of boards, and it has a built in socket for an Xbee radio. However,
because of the built in socket, the board is quite large and doesn’t allow for the
group to design much in the way of space savings. In the end, while the board
would work it would end up being too large for any practical application of the
sensor units.
The Arduino family of boards, being all that is left, is the final choice of the group.
There is a board for almost any size application, which makes it convenient in
respect to the group’s desire to design and control the size and structure of the
final device.
At one point, the group considered the idea of having the display simply be an
Android application. This would have allowed the information from the device to
be seen from anywhere in the world, and this would have allowed the user to
control their home energy usage even if on vacation. To do this, the project
would have required a smaller microcontroller than what is being used for the
display unit and would have been one of the larger Arduino boards. However,
this idea was scratched in favor of a touch-screen embedded solution which has
a processor of its own, to allow for better visualization and control of the screen.
46
For the sensor units, small is the key. The Arduino Pro mini manufactured by
Sparkfun electronics was just the right size for the project, as seen below. The
Arduino has the appropriate input and output pins to handle the analog
measurements and the serial communication of the Xbee module. The ability to
program the microcontroller in a modified C-programming language was an
enormous boon and definitely a major factor in the decision to use it.
Small size of the Arduino Pro Mini
Permission asked from robot.com.au
The programming of the Arduino is done in a modified version of C/C++. The
modified part of the development language is the special functions that have
been added in to give users direct access to helpful tools and pin modes. Some
of the more useful functions that will be used in the project are the following:
pinMode() – Configures the specified pin to behave either as an input or
an output
digitalWrite() – Write a HIGH or a LOW value to a digital pin. If the pin
has been configured as an OUTPUT with pinMode(), its voltage will be set
to the corresponding value: 3.3V for HIGH, 0V (ground) for LOW. If the pin
is configured as an INPUT, writing a HIGH value with digitalWrite() will
enable an internal 20K pullup resistor, LOW disables it.
analogRead() – Reads the value from the specified analog pin. The
Arduino board contains a 6 channel (8 channels on the Mini and Nano, 16
on the Mega), 10-bit analog to digital converter. This means that it will
map input voltages between 0 and 3.3 volts into integer values between 0
and 1023.
analogReference() – Configures the reference voltage used for analog
input (i.e. the value used as the top of the input range). Can be the default
3.3V or any of the internal references available (1.1V or 2.56V), or
external (voltage supplied to the aref pin).
millis() – Returns the number of milliseconds since the Arduino board
began running the current program. This number will overflow (go back to
zero), after approximately 50 days.
delay() – Pauses the program for the amount of time (in miliseconds)
specified as parameter. (There are 1000 milliseconds in a second.)
47
attachInterrupt() – Specifies a function to call when an external interrupt
occurs. Replaces any previous function that was attached to the interrupt.
Most Arduino boards have two external interrupts: numbers 0 (on digital
pin 2) and 1 (on digital pin 3). The Arduino Mega has an additional four:
numbers 2 (pin 21), 3 (pin 20), 4 (pin 19), and 5 (pin 18).
There is also a very handy serial library that allows serial communication using
the RX and TX pins, which also means that pins 0 and 1 cannot be used as
digital I/O pins. The serial library will be used to communicate with the Xbee
module. Some of the special functions that the group will be using are:
Begin() – Sets the data rate in bits per second (baud) for serial data
transmission. For communicating with the computer, use one of these
rates: 300, 1200, 2400, 4800, 9600, 14400, 19200, 28800, 38400, 57600,
or 115200. You can, however, specify other rates - for example, to
communicate over pins 0 and 1 with a component that requires a
particular baud rate.
End() – Disables serial communication, allowing the RX and TX pins to be
used for general input and output. To re-enable serial communication,
call Serial.begin().
Read() – reads incoming serial data.
Print() and Println – Prints data to the serial port as human-readable
ASCII text. This command can take many forms. Numbers are printed
using an ASCII character for each digit. Floats are similarly printed as
ASCII digits, defaulting to two decimal places. Bytes are sent as a single
character. Characters and strings are sent as is.
Another great upside to the Arduino Pro Mini is that it allows for an unregulated
power supply to act as the power supply, up to 12 volts through the RAW pin.
This allows for leeway in the power supply design on the sensor units and
flexibility of the microcontroller board itself. It is also possible to provide the unit
with a regulated power supply at 3.35V through the VCC pin but this would be
unnecessary and allow for the design to incorporate a single power supply for all
components. Below is the built in power regulator which connects to the VCC pin.
48
The Regulator built into the Arduino Pro Mini
Permission granted by the Creative Commons License
Microcontroller:
Operating Voltage:
Input Voltage:
Digital I/O Pins:
Analog Pins:
DC Current per I/O Pin:
Flash Memory:
SRAM:
EEPROM:
Clock Speed:
Atmega168
3.3V
3.35 – 12 V
14
6
40mA
16KB(2KB used by the bootloader)
1KB
512 bytes
8MHz
2-2.4 Motion Detector
A motion detector may be added to the unit to give additional features. A motion
detector is a device that can quantify motion so that it can be integrated into a
device system to alert the system if an object comes within its field of view.
These are most used in home security systems. If a motion sensor were to be
added, it would most likely be prebuilt, rather that built from scratch to save time.
If added, this piece would allow the system to be set to turn all electronic devices
off when a person is not detected in that particular room. The idea behind this is
the user would have an even more passive role in his or her energy reduction.
There are three different types of motion detection sensors. The first type of
sensor is the passive infrared sensor. These are the most widely available type
49
of motion sensor. This is a device that measures infrared light radiating from
nearby objects. The term “passive” indicates that the device itself does not emit
any beams or radiation, it merely detects them. The sensor is made of a thin
pyroelectric material film. It can be made from gallium nitride, cesium nitrate, or
polyvinyl fluorides. To focus the infrared radiation, most detectors use fresnel
lenses. This is generally put into a case as shown below.
Figure 2-2.4.1 infrared motion detector
After research, it was found that this type of sensor can detect a person through
body heat emitted as infrared radiation. It will sense anything with a weight
greater than approximately thirty pounds.
The other two types of motion sensors are active sensors. Ultrasonic sensors
send out pulses and measure the reflection angles. If anything within the field
moves, the reflection angles will change, and the unit detects the movement.
The final type of sensor are microwave sensors which send out microwave light
pulses and measure the reflection off of a moving object. This type is most
similar to a police radar gun.
An ideal motion detector for this project would be one that could be set to on or
off, while still not resulting in Vampire Draw. The amount of power used is the
largest drawback for this addition.
2-2.5 Other Components
50
Other than the main components discussed above, the design will require the
use of some basic components such as resistors, capacitors, amplifiers, filters
and regulators. The need of adjusting voltages to suit the required microcontroller
input voltage for instance will require the use of an attenuator or a regulator. The
same rule applied to the current sensor and relay. We will use any of these
components as needed. The cost will virtually be zero as we plan on using parts
for the senior design lab.
51
3- Design
3-1 Power and Current Sensing Circuit
We have decided to use a Hall Effect current sensor in our design. As explained
earlier, this kind of sensors has so many benefits and is appropriate in our
design. After extended research, we decided on using the Hall Effect sensor from
Honeywell.com. The sensor’s description is below:






















Current Measuring Range DC:0 to 72A
Current Measuring Range AC:0 to 72A
Supply Voltage Range DC:6V to 12V
Sensor Output: Open Loop
Response Time:3ms
Current Consumption:20mA
Output Type: Linear
Supply Voltage DC Max:13.2V DC
Supply Voltage Range DC:12V
Connector Type:3 Pin
Device Type: Open Loop Linear
External Depth:35.6mm
External Length / Height:10.4mm
External Width:34.3mm
Hole Diameter:10.9mm
Operating Temperature Max:85°C
Series: CS
Supply Current:20mA
Supply Voltage DC Min:5.4V DC
Temperature Operating Min:-25°C
Transducer Function: Linear Current Sensor
Reasonable price of around $15 per unit
To implement this part of the design, we will wire the power source to the sensor.
The sensor will have the positive and common input terminals clearly marked. A
typical sensor runs on between 4.5 and 8 volts at 5 milliamps DC. Also, we will
need to test the sensor by measuring its output with the voltmeter. The Hall
Effect sensor output will be connected to the microcontroller. The following figure
shows a possible configuration of the part of the design.
Figure 3-1.1 shows how the Hall Effect sensor will be connected:
52
Figure 3-1.1: Hall Effect sensor circuit
The circuit is an inverting amplifier circuit. The 100k variable resistance shifts the
input signal. It can be set so that the output lies within the Arduino's 0 to 5V input
range. The 1k variable resistance sets the amplification factor. By varying both
resistances it is possible to set the current measurement range of the circuit.
Since this is an inverting amplifier circuit, the output is inverted so that when the
current increases the signal input to the Arduino decreases from 5V to 0V.
3-2
Power Relay
Power relays are typically discussed in terms of several things such as the power
the high power side can handle in Amps, the voltage and power type (AC or DC)
the coil needs to operate and the number and type of contacts the relay has. The
first thing we need to discuss is the power rating. We can say that a relay is rated
for its capacity to handle power. The relays we have been searching were
described as 20A, 30A, and several different values of current. We realize that
the relay rating must be higher that the maximum rating of the appliances we will
test our system on. The second thing is the coil voltage and type. This
characteristic is typically omitted when working in a known environment. The
third thing is the number and type of contacts. This characteristic is used to
control various things at once and control them by turning them on or by turning
them off. The number of contacts or poles is the number of things that the relay
can control at once. The relay is just an electromagnetically controlled switch.
Those contacts or so called poles are described as "normally open" (NO) or
"normally closed" (NC). This simply describes what the rest state means. For a
power relay, that means if no power is applied to the coil wire. In the typical case
where a user wants to turn something on, we would use a normally open relay
because when we apply power to the relay, the contacts close, and power is sent
to the desired device. In our case however, we want the relay to be able to turn
things off. Therefore we will choose a normally closed relay. So when we apply
the power to the relay, the contacts open and the power is no longer sent to the
specified appliance or device.
53
As discussed earlier, a solid state relay SSR is an electronic device we can use
to switch the electrical current, rather than an electromechanical device. An
electromechanical relay uses a magnetic coil and mechanical contacts. When
current flows through the coil, it pulls down a piece of iron called an armature,
causing the mechanical contacts to touch and thus close an electrical circuit. We
had the choice of either a solid state relay or a mechanical relay. The solid state
relay has no mechanical moving parts, but instead uses a three terminal device
such as a TRIAC, a back to back thyristors, or a field effect transistor FET to
conduct the electrical current. When the third terminal is energized by the control
input, the device conducts. Essentially the solid state relay controls a larger
electrical current by accepting a small control signal. A main benefit of using it
would be the fact that here are no moving parts. Also solid state relays have no
internal arcing or contacts to wear out, so they can last virtually forever. They
also have extremely low control input requirements, and are immune to vibration.
Other considerations that have led to our choice were the fact that the solid state
relays are typically smaller than the electromechanical relays. This will help us
conserve valuable space in printed-circuit board applications. Also, the solid state
relays offer improved system reliability because they have no moving parts or
contacts to degrade. Adding to that, solid state relays provide high performance,
including no requirement for driver electronics and bounce free switching. They
provide improved system life cycle costs, including simplified designs with
reduced power supply and heat dissipation requirements. Another benefit would
be the possibility of using a solid state relay as surface mount technology parts,
which means lower cost and easier surface mounting technology printed circuit
board manufacture.
Another issue with the electromechanical relay would be the bounce time. The
maximum bounce time of an electromechanical is the period from the first to the
last closing or opening of a relay contact during the changeover to the other
switching position. Bouncing causes short term contact interruptions. In our
case, bounce can easily lead to false pulse counting as contacts continue to
make and break the circuit during bounce. Contact bounce does not occur in
semiconductor based solid state relays. So when the user decides that the power
should be turned off a device, it is turned off.
Figure 3-2.1 shows the relationship between operate time and contact bounce
time in electromechanical relays:
54
Figure 3-2.1 : relationship between operate time and contact bounce time.
Permission asked from Clare.com
Since solid state relays do not have contacts, wear issues are not of our concern.
The absence of contacts and moving parts means that the solid state relays are
not subject to arcing and do not wear out. Contacts on the electromechanical
relay on the other hand can be replaced on some larger relays but contact
replacement is not practical. Another issue with the electromechanical relay
would be the shorted coils. Shorted coils can occur if excessive heat melts the
coil insulation. Open coils can be caused by over voltage or over current
conditions applied to the coil. The circuitry used to drive the electromechanical
relay can cause open coil failures if the drive circuit itself fails or is subjected to
transients. Solid state relays can be driven directly from logic circuits, so an
intermediate drive circuit is not required. AC load solid state relay have the
benefit of zero crossing switching which reduces noise in the circuit by restricting
the switching operations to the point where the voltage crosses zero.
Since a solid state relay will be used. A possible configuration is considered. The
relay will be connected to the load and to a control voltage. The sensitivity the
minimum control voltage and current at the solid state relay turns on depends on
the characteristics of the isolating component or circuit and will be documented in
the relay data sheet. The following figure shows a possible configuration of the
relay circuit.
Figure 3-2.2 shows a photo coupled solid state relay:
55
Figure 3-2.2: photo coupled solid state relay. Permission asked from
omega.com
3-3 Wireless
The Xbee wireless modules support AT commands and parameters can be set
by appending AT to any of the commands. The commands are fairly simple, used
to identify, and address modules. Initially the modules will need to be configured
to be on the same network and channel so that they can communicate with one
another, this can be done on start-up of the arduino or, be stored in the nonvolitile memory onboard the Xbee module. The display units address must then
be set as the destination address in each of the sensor units, so that the display
unit will be the one receiving the data. Alternatively, the addresses can be set so
that the data is received by all the xbee units in the network and program the
Arduinos to ignore any incoming information that doesn't include the relay on/off
command.
Command Description
Valid Values
Default Value
ID
Network ID of the module
0-0xFFFF
3332
CH
Channel of the module
0x0B-0x1A
0x0C
SH, SL
Serial number of the
module
0-0x0FFFFFFFF
Differs
MY
16 bit address
0-0xFFFF
0
DH, DL
Destination address
0-0x0FFFFFFFF
0 for both
BD
Baud rate for serial comm
0-7
3
Table 3-3.1 of the Xbee module commands
The Xbee module has different modes that it can be in at any given point. The
idle mode is important as it is the connecting mode between all the other modes.
The module can quickly switch between idle and any other mode, and to change
modes the module must first enter idle from one mode before transitioning to the
other mode. The other modes are Transmit, receive, sleep and command mode.
56
Figure 3-3.1 Mode state diagram
Permission asked from digi.com
The Transmit and receive modes are very similar in action and are closely
related as when one module is in transmit mode, another module is going to be
in receiving mode. The Xbee modules can be set to either Direct or indirect
transmission modes. The indirect transmission mode is very important for the
project as it will allow for a node to sleep and the message to be held by the
sender until the sleeping node polls the sender to let it know that it is ready to
receive data. In a strictly peer to peer network, only Direct transmission is
allowed, where messages are immediately sent, but if a network has a
coordinator the coordinator may hold a message and allow for indirect
transmission. The transmit mode is even more complex in a mesh network, as
the modules also need to discover a route for the data, whether this is from
sender to receiver or by way of one or more intermediate nodes.
57
Figure 3-3.2 Transmit Mode Sequence
Permission asked from Digi.com
Another important tool which increases the reliability that data is received is the
ACK, an acknowledgment that the data is received. If the sender of a message
doesn't get an acknowledgment the message is resent up to three times before
triggering an ACK failure and losing the message.
The sleep mode is a powerful energy saving tool that can be utilized to conserve
power. There are two main sleep modes that can be implemented, either a pin
sleep, or a cyclic sleep. When the module is in pin sleep mode, it will stay so until
the sleep pin is brought back down to a zero. When it exits, it will not poll for
messages, unless commanded to do so. The cyclic sleep is where the real power
is, it allows for a timed sleep where upon exiting it will automatically poll for
messages which is something that is perfect for the parameters of the project.
58
Figure 3-3.3 Polling allows sleeping nodes to receive data when they wake
Permission asked from Digi.com
Command Mode is used to access module data and program certain
characteristics of the module. Command mode is entered by sending the three
character command sequence “+++” when there has not been any other
characters sent within the guard time, which is one second by default, but can be
set while in command mode. If done correctly the module will respond with
“OK<CR>” meaning the module is now in command mode. The most common
way for a user to fail to enter command mode is due to a Baud rate mismatch,
which by default is set to 9600 bps. To send an AT command, the user can then
use the format seen below to send commands, using AT followed by an two
character ASCII command and optional space, the parameter in HEX all followed
by a Carriage Return. The unit will then send an “OK<CR>” response if the
command is successfully executed, otherwise sending “ERROR” indicating that
the command was not successfully executed. To make any changes permanent
the user must then send the WR code to write the changes to long term memory.
Figure 3-3.4 AT command Structure
Permission asked from Digi.com
While in command mode there are a plethora of commands that can be issued to
perform many kinds of functions. There are diagnostic commands that will let the
user see any failures and check version information. These are as follows:

VR (firmware version): Reads which firmware version is stored in the
module. There are either three or four hexadecimal characters and if there
are only three characters, the last character is assumed to be a zero.
59






VL (firmware version-verbose): Shows information on specifics of the
RF module like the date the application was built, the bootloader version
and built date.
HV (hardware version): Indicates which version of the hardware is being
used.
DB (received signal strength): Used for read the signal strength
received in dBm of the last RF signal received. There is an inherent range
in the module from -40 to -96 dBm. If there is no signal sent in, “0” will be
displayed.
EC (CCA failures): Reads the amount of failures through each of the
modules. A failure in this case is when the RF module does not transmit a
packet due to the detection of energy that is above the CCA threshold
level (set with CA command). When the user wants to reset the count, the
user must set the parameter to zero.
EA (ACK failures): Reads the amount of failures through each of the
modules. A failure is when the module expires its transmission retries
without receiving an acknowledgement on a packet transmission. When
the user wants to reset the count, the user must set the parameter to zero.
ED (energy scan): Finds the max energy in (-)dBm in each channel
followed by a carriage return with an extra carriage return to signify the
end of the command.

Having a module join an established can lead to some
there are many AT commands to help with that, including
commands help users make sure that all of their
communicating on the same channels and are able to
network:

CH (channel): Sets the channel that the XBees‟ communicate to each
other. The channel value can range from 0x0B to 0x1A with a default
parameter value of 0x0C (12). To calculate the center frequency, the
equation is: center frequency = 2.405 + (channel – 11(d))*5MHz. (Where d
is the decimal value of the parameter.
ID (pan ID): Reads the personal area network (PAN) ID of the module. If,
and only if, all the PAN values match then the modules could
communicate with each other. It has a default hex parameter of 0x03332
(13106) and a range of 0-0x0FFF.
DH (destination address high): Will set and read the upper 32 bits of the
64- address. Has a parameter range of 0x0FFFFFFFF.
DL (destination address low): Will set and read the lower 32 bits of the
64- address. Has a parameter range of 0x0FFFFFFFF. When DH and DL
are put together, one can see the 64-bit combination that is the address.
This address needs to be the same for all other modules to communicate
to each other.



difficulties, but
failures. These
modules are
join the same
60













MY (16-bit source address): Sets and reads the 16-bit source address of
the RF module. When the MY is set to 0x0FFFF, the 16-bit packet
received is disabled and the 64-bit packet is enabled.
SH (serial number high): Reads the 32 high bits of the 64-bit address.
This value cannot be changed and is a read-only value.
SL (serial number low): Reads the 32 low bits of the 64-bit address. This
value cannot be changed and is a read-only value.
RR (XBee retries): Accounts for the amount of tries past the standard
three retries inherent to the module. The parameter range for this
command is from zero to six.
RN (random delay slots): Accounts for the exponential value of the
carrier sense multiple access – collision avoidance algorithm.
MM (MAC address): Displays the MAC mode value. With a default
parameter at zero, there are four parameters from zero to three. Zero
meaning the use of the Digi mode that includes the 802.15.4 RF packet,
one is just the 802.15.4 packet without acknowledgment, two is the
802.15.4 packet with the acknowledgment and three being solely enabling
Digi mode without acknowledgment.
NI (node identifier): Identifies particular nodes using solely printable
ASCII data that doesn‟t start with a space the ending has a carriage return
command and once the maximum bits have been entered, the command
will automatically end. This command can accept anything up to a 20character string of ASCII values.
ND (node discover): This parameter value is the same as the 20character value given in the NI command. This command goes into each
module and demands an ND command packet that is 64 bits long.
NT (node discover time): The NT command will, once the parameter is
set, give a specified time for whichever RF module to respond back with
its 64-bit address. The parameter range for this command is 0x01-0x0FC
(1-100 milliseconds)
NO (node discover options): Will decide, based on its own parameter,
whether or not to have its own ND response. The parameter values are
either zero or one. Zero being that there will not be an ND response and
with a one, there will be an ND response.
DN (destination node): This command will convert any NI string to a
physical string as long as the DL and DH are set to the NI and the RF
module cannot be changed at the time.
CE (coordinator enable): Reads the behavior between the end device
and the coordinator of the module. One being the coordinator and zero
being the end device, which happens to the default parameter.
SD (scan duration): reads or sets the exponential value of the scanning
time through multiple devices. The scan time can be calculated as with the
following equation: scan time = ((# of channels to scan)*(2^SD)*15.36
milliseconds). SD has a parameter range of 0 – 0x0F with a default value
of four.
61








ED (energy scan): Displays the max energy is each channel. The value
that is displayed is the energy level in –dBm. The scan time can be
calculated as with the following equation: scan time = ((2^ED)*15.36
milliseconds).
SC (scan channel): This command is very important for the two
commands above, SD and ED. This command is used to display and or
set which channels the energy scan and scan duration will be used in.
A1 (end device association): Uses a combination of bits in a certain
order to set certain behaviors for the end device. The parameter range is
from zero to 0x0F with a default parameter value of zero. Bit zero deals
with whether or not the coordinator will associate itself on PAN ID with
others based on the NI command. Bit one deals whether or not the
coordinator will associate itself with others based on the channel settings.
Bit two deals with whether or not the device will attempt association and
deals with only non-beacon systems. Bit three deals with whether or not
the pin wake will pole the coordinator for any pending data. Bits four
through seven are reserved and pre set.
A2 (coordinator association): Uses a combination of bits in a certain
order to set certain behaviors for the coordinator. The parameter range is
from zero to 0x7 with a default parameter value of zero. Bit zero deals with
whether or not the coordinator will perform an active scan to locate
available PAN ID. Bit one deals whether or not the coordinator will perform
an energy scan. Bit two deals with whether or not the device will attempt
association with any other devices. Bits three through seven are reserved
and will not be changed.
AI (association indication): Checks to see if there are any errors while
the RF module was trying to associate. The parameter range for this
command is from zero to 0x013. This value is a read-only value.
DA (force disassociation): Disassociates the coordinator and the enddevice and then tries to re-associate them.
FP (force poll): Requests indirect messages from the Coordinator.
AS (active scan): Sends out a Beacon request to each channel. Indicates
the time set for each Beacon with each channel.
The Zigbee units can be one of three profiles, a Coordinator, Router, or End
Device. The coordinator is responsible for choosing the channel and PAN ID.
Routers and End Devices connect to the network that the Coordinator starts.
Routers must join a PAN that was created by a coordinator and can route data
through the network. Once it has joined a network it can allow End Devices to
join as children. End Devices can only join the Pan and cannot assist in data
routing, the End Device profile was designed with battery operated standalone
devices in mind and this profile most likely will not be used as End Devices
cannot participate in a mesh network.
The Zigbee PAN, Personal Area Network, is formed by a Coordinator. To start
the PAN the Coordinator performs scans to determine which channel it should
62
use. It first performs an energy scan to determine the channels that have the
least energy levels, which are the best to use to reduce any interference. Once
that is completed, the coordinator performs a PAN Scan to determine if there are
any other PAN networks already operating on a channel. If it gets a response on
a PAN ID it will skip to the next PAN ID and try that one. Once it finds a channel
that has the least energy profile and has a PAN ID that isn’t already being used
by another PAN the coordinator will use that channel and ID to begin its own
PAN. In the groups design, the Display unit will always be the coordinator.
Figure 3-3.5 PAN scan looking for unused PAN ID
Permission requested from digi.com
The newest models from Digi, the Xbee 2.5 models, allow mesh networking.
Mesh networking was important to the design of the project as it allows for a
node to be out of range of another node yet still communicate with it as long as
there is an intermediate node to pass the information along. To allow for this, the
Xbee units have an automatic route discovery which was touched on earlier in
the discussion of the Transmit Mode. To properly use mesh networking the
Transmissions must be done using the so called Unicast Transmission. Unicast
transmissions are always addressed to the 16 id address, but this isn’t
necessarily always static, so the network can automatically detect the 16 bit
address that belongs to which 64 bit address. If the 16 bit address isn’t known
the node that wishes to transmit data will initiate a network address discovery
using the 64 bit address. During this process, each node will check its 64 bit
address against the one sent out and if there is a match it will reply with its 16 bit
network address. To communicate in a mesh network the Xbee units use a route
discovery method. The routing is accomplished by tables in each node that store
the next hop. If the next hop is not known, route discovery must take place first, if
the network is large enough, all possible paths will not be able to be stored and
route discovery will happen more often.
Node
Router 3
Coordinator
Router 5
Destination
Next Hop
Router 6
Coordinator
Router 6
Router 5
Router 6
Router 6
Table3-3.2 of Mesh Network Pathing
63
The Xbee module will also be using the Arduino shield for communication with
the arduino. This shield makes communicating with the arduino only requiring the
use of the RX and TX pins and powers the Xbee module from the Arduinos
power supply a snap. These can be found pre-made or, can the parts and board
can be bought separately and self soldered. The pinout of the Xbee shield can be
seen below. This particular pinout also allows for the Xbee module to be
connected to and programmed directly from a computer through an FDTI to USB
cable.
Figure 3-3.5 The pinout of the Xbee adapter
Permission granted from adafruit.com
3-4 Microprocessor
Each of the sensor units will have an Arduino board and an Xbee radio to allow
for measurements and control of the relay allowing power to flow to a device. The
microprocessor with control the flow of measurement data from the analog signal
will perform any necessary calculations and send that information through the
Xbee radio to the display unit for storage and display.
64
Figure 3-4.1
The Arduino can also be programmed to perform certain actions based on the
state of the switch. While the switch is off, there will be no power flowing through
the socket and no measuring data is needed so the arduino can be put into sleep
mode and can be woken by a wake signal on an interrupt pin which can come
from the display unit by way of the Xbee radio when the relay is allowing the
socket to be used. It could also be programmed to sleep between periods where
the relay is allowing current to pass but is not being used (current measurement
= 0), waking up after a predetermined period of time, or on an interrupt when
current starts flowing.
3-5 Main Control Board
For the main unit, embedded system was selected for this project. It will be
purchased from Future Designs Inc. The model is the DK-57VTS-LPC3250. It
was chosen because it is a premade embedded system, which will save time on
development. It runs at 5 VDC, and 2.3 A. The system has three components:
the LCD touch screen, the CPU Module, and the I/O controller. It uses a base
Carrier Board, a CPU module with SOMDIMM memory, and an LCD Carrier
board. The boards come with expansion connectors, so the project can be
expanded when necessary. In addition, there is a Micro SD memory socket,
which can be used if the embedded Linux and the rest of the software begin to
use too much memory on the board. The screen on the DK-57VTS-LPC3250 is
a 5.7 inch screen with a VGA connection as shown in figure 3-3.1. Its resolution
is 640 x 480 pixels.
65
Figure 3-5.1 permission requested from www.teamfdi.com
The Carrier board specifications include:

10/100 Ethernet Port, USB Host and Device ports

One CAN port (Male DB9), One RS-232 port (Male DB9), External I²C
interface

3-axis Digital Accelerometer & Temperature Sensor

Real-time Clock with SuperCap backup

TFT interface for Graphics LCD displays up to 1024x768 resolution, 18-bit
color

Flexible Power Supply input can be wall supply or 5V USB
This will allow us to have accurate time measurements, and reliable input/output
while leaving a low power usage footprint. The components of the Carrier board
can be seen in figure 3-3.2. The main components that will be necessary are the
USB inputs, and the flash memory units. This is where all of the data and
program structures will be placed so the ARM processor can process them. An
additional flash memory card may be necessary, which can easily be afforded by
purchasing one. The maximum additional flash memory can be upgraded to 4
gigabytes. 4 gigabytes will cost approximately $15.
66
Figure 3-5.2 permission requested from www.teamfdi.com
CPU module specifications:

Based on SODIMM form factor (Dual Inline Memory Module)

LPC3250 266MHz ARM926EJ-S core

Vector Floating Point Co-processor

256KB of Internal SRAM

64MB of External DDR SDRAM

512MB of External NAND FLASH

1KB of EXTERNAL Secure EEPROM

10/100 Ethernet PHY

Micro SD Memory Socket

Mini-JTAG Debug Connector

PCB Dimensions 2.66” x 1.89”
The CPU used is the ARM92EJ-S. It supports 32-bit, 16-bit, and 8-bit instruction
sets. It features Jazelle Direct Bytecode eXecution, which allows the ARM
processor to execute Java bytecode. The ARM9 incorporates Enhanced DSP
instructions to support greater efficient digital signal processing algorithms. The
processor uses multiply-accumulate to support this. Also, the ARM9 has a nonunified cache, so instruction fetches do not release data, and separate data and
address bus signals. An example of an ARM9 can be seen in figure 3-3.3.
67
Figure 3-5.3 permission requested from digi-key.com
3-6 Block diagrams
3-6.1
Current/Power Sensor Circuit
The block diagram below shows a basic functionality of the current sensing
circuit. The Hall Effect current sensor will measure the current and provide a
voltage output that is proportional to the current being measured. This value will
be sent to the microcontroller. The microcontroller will then calculate the power
and transmit the data wirelessly to the LCD screen when the user can monitor
and track their power usage. The microcontroller will also send a 1/0 command to
the power relay in order to turn the power off a device:
68
Sensor: Current
Wall outlet
Measurement
Power relay
Microcontroller
Transceiver
Figure 3-6.1.1: current sensing diagram
3-6.2
Power Relay Diagram
As stated earlier, the power relay will receive command from the microcontroller
to turn the power off an appliance. The user will choose to turn the power off a
device. The wireless receiver will receive the command and send t to the
microcontroller. A voltage change will then cause the relay to turn the power off
the device. The diagram below shows the basic functionality of this part of the
design:
69
Transceiver
Microcontroller
Power relay
Power relay
Figure 3-6.2.1: Power Relay Diagram
3-6.3
Microcontroller and Touch Screen
Diagram
This is a high level diagram how the system will operate. The modules will
transmit data to the main control board, which will process the data and send it to
the touch screen where the user can modify the options for the system, and view
his power usage.
3-7 Software Design
70
3-7.1
Microprocessor Programming
The Arduino microprocessor board is programmed using a modified version of
the C programming language. It has commands built into it to handle its Arduino
specific needs and to also control the pins specifically. The Arduino Pro Mini has
14 digital input/output pins which can be programmed to be one or the other,
using special commands, pinMode(), digitalWrite(), and digitalRead(). Some of
the pins have special functions, such as the 0 and 1 pins which are the RX and
TX, respectively, for serial communication, which will be used to communicate
with the Xbee module. Pins 2 and 3 can be used as external interrupts, which
means that they can be programmed to wake the microcontroller from a sleep
mode.
Pin
Primary Use
Secondary Use
0, 1
Digital IO
Serial TX and RX
2,3
Digital IO
External Interrupts
3,5,6,9,10,11
Digital IO
Pulse Width Modulation
10,11,12,13
Digital IO
SPI communication
One of the main benefits of using the Arduino boards is the compatibility with
Standard C libraries which the group is intimately familiar with. This will aid the
group in properly programming the microprocessor and implementing special
functions, and will allow the group the flexibility to design and program the boards
as they desire. The standard libraries that the Arduino uses are:





stdbool.h: This library is used for Boolean type of variables. This library
will be useful in declaring and determining if certain conditions are met
before sleeping, transmitting data or turning off the power going through
the unit.
string.h: This library is used to manipulate strings. This library will be
used to help send the data to and from the Xbee modules and to be
accessed by the display unit for display.
stdio.h: This is the standard library for input and outputs. This will be used
throughout the entire designed program. The stdio.h library also contains
allows usage of the printf statement. This is important as it will help control
the flow of data in and out of the unit.
stdlib.h: Among other functions, the stdlib.h library allows the usage of
the absolute value function. This will allow the microprocessor to handle
any negative values it may come across use them appropriately.
math.h: This is the standard math library. This library allows functions
such as round. The round function will mostly be used to make the raw
71
data given by the main logic steps, converting the inputted power
measured into kilowatt hours, into a more manageable value.
The Microcontroller programming will consist of three major methods: a
getmeasurement method, a transmitdata method, and a switch method. Each of
these methods performs a specific task related to data flow of the sensor units.
The getmeasurement() method will read the value on the appropriate analog
input pin and get the measurements and perform the needed calculations on
them. This is essential as this is the data gathering method used by the
microcontroller. This method would look similar to
int cur = 0;
cur = analogRead(0); //returns value between 0 and 1023
//perform calcualtions and store value to be sent
The transmitData method will send out the last recorded measurement to the
display unit through the serial communication pin 1, TX to the Xbee module,
addressing it to the display unit and appending the originating unit name.
Int data = getmeasurement();
Serial.println(data, 4);
The switch method will command the relay to turn on or off which will be
connected to one or more of the digital pins. This should be a fairly simple
method, only setting the digital out pin to either a 1 or 0 and then terminating.
The microprocessor programming scheme can be thought of as a state machine
where certain conditions, or no conditions, send it to different states. The starting
point would be the getMeasurement() method, where the microcontroller would
get a measurement. From there, it would automatically go into the transmitData()
method where the data would be sent to the display unit. After transmitting the
data, it would go into one of two states, if the relay is on, it would enter a timed
sleep after which it would go to the getMeasurement() method. If the relay were
set to be off, the microprocessor would go into an untimed sleep and would
awaken when the relay is turned on. The state diagram is given below to aid in
visualization of the states. After implementing these state, more states could
easily be added with any additional parameter that are introduced, possibly going
to sleep when there is no current draw on the socket and waking when there is a
current draw.
72
3-7.2
GUI
3-7.2.1 General
The GUI needs to have a main window that holds all of the subwindows with live
graphs inside of them as shown in figure 3-5.1.1. The GUI must be detailed
enough to be intuitive for a user to change or modify the energy usage options,
but clean enough to not be distracting. In each window, there must be sufficient
information as to allow the user to identify where they are, and view any
necessary data, while not cluttering up the screen with unimportant information or
graphics. Each window needs to be a clickable object that when clicked, leads to
the proper subwindow. The color selection will be mute colors, so to not be
distracting, or difficult to view. The fonts for buttons will be Times New Roman in
size 20, and the titles will be Times New Roman in size 32. All buttons in the
GUI need to be large enough for the user to be able to press without much effort.
Buttons that are too small could cause users with larger fingers to unintentionally
change functionality, or go to the wrong subwindow. In addition, the buttons
need to be far enough apart as to not allow the touch screen to have small errors
and register a hit in the area for an incorrect button.
3-7.1.2
Main Menu
The goal of the main menu is to have a clean user friendly appearance at first
glance, but still have relevant information that the user can use at hand. It needs
to be easy to understand, but helpful enough that the user will want to look at it
73
often. The main screen will feature three live line graphs of each of the attached
modules. Each of these objects will connect to the corresponding subwindow for
that graph. The user can simply press on the graph they wish to view more
details about, and they will be sent to the corresponding subwindow. The
Options button will send the user to a subwindow where basic user settings can
be accessed. The ON/OFF button will cut off all power to the connected
modules. This button will change color to RED when all modules are set to ON,
and the text field inside it will display TURN ALL OFF. The button will be GREEN
when all modules are set to OFF, and the text field will display TURN ALL ON.
The budget remaining will be a text field displaying how much of the user’s
remaining budget in KWH or US Dollars is remaining. Which is displayed can be
changed in the options menu.
Figure 3-7.1.1
3-7.1.2 Subwindows
Each subwindow must have the title of what the subwindow is at the top. Every
submenu will have an option to go back to the main screen. The subwindows for
each module will contain the live graph of energy usage along with several fields.
The first two fields will be text fields with a calculation of average usage of energy
in KWH and US Dollars per day and per month, respectively. The third text field
will show the amount of KWH saved in the last 30 days by using the Home
Energy Management System. It will indicate a value for that particular module for
the time it was not operational. The fifth text field will show the user the
remaining budget portion the device expects the Module to use for the remainder
of the payment period. The final field will be a button to shut off all power to that
unit. This button will change color to RED when the module is set to ON, and the
74
text field inside it will display TURN OFF. The button will be GREEN when the
module is set to OFF, and the text field will display TURN ON. The information
needs to be presented in a way that is easy to understand, yet informative
enough that it needs to further explanation.
Figure 3-7.1.2
3-7.1.3 Options
The options screen will hold information relevant to the overall operation or
presentation of the device. The screen shown in figure 3-5.1.3 indicates what the
user will see on the options subwindow. The first option will be a button that
allows the user to change their budget in KWH. This integrates with other
calculations in the project. The second option is a simple button that changes
the metrics of KHW or Dollars. When the individual modules are inspected, they
show the average usage in month or day. This option will allow the user to see
their power use in Dollars or KWH. The third button will allow the user to change
their personal costs of KWH per Dollar. This includes the initial cost per KWH,
the cap of KWH before the homeowner is charged more money per KWH, and
the cost of power above the cap. This will function by a keypad appearing next
to the option buttons with an enter button below them. The number will
correspond to a value that will be put into the price per KWH variable. The user
will be able to input the cost of KWH in cents, and even input a partial cent value.
After the amount is input, the ENT button will be pressed, and the number will be
stored.
75
Figure 3-7.1.3
3-7.2 Operation
The embedded system should work together to give a seamless user experience.
The block diagram in figure 3-5.2.1 shows how this will be done. The touch
screen should accurately find clicks that correspond to the screen output, and
send a click message to the Mouse Click Listener. The Mouse Click Listener will
then notify the operating system that an object was clicked, and send that
message to the C++ assistant for processing. Each screen should close the
previous screen, and open only the new, previously selected screen. At no time
should more than one screen exist at a time. Thus, the MiniGUI must close each
window every time the user wishes to navigate to a new screen. The C++
assistant will tell the MiniGUI to update to whatever screen or action that was
clicked by the user. GNUPlot will become active whenever a window containing
one or more graphs becomes what is output on the screen. This can be seen in
the Sequence Diagram in figure 3-5.2.2. Data for the graphs themselves will
come from the wireless receiver sent from the three modules. This data is stored
in a database on the system, which is accessed by GNUPlot whenever it needs
to plot a graph. The database for the system will hold all data for the previous
month, and a double floating point variable containing how much energy had
been saved all time in KWH. In addition, the database will contain variables for
what budget the user wishes to set, in US dollars or KWH.
76
Figure 3-7.2.1
Figure 3-7.2.2
77
3-7.3GNUPlot
The data for the graphs will need to be put into two dimensional arrays so that
GNUPlot can read the data. To update the data, a separate array with the same
two dimensional arrays will be created as temporary storage. The new data will
be put into index 0 of the newly created array. Each value of the original array
will be moved one unit forward in the X index, with the values at the last X index
to be purged. The data will be refreshed every 10 minutes, so this process will
occur at least this often. The program to ensure this will happen is called Cron.
Cron is a UNIX based job scheduler that can be configured to run scheduled jobs
at certain times or over certain time intervals. Cron has a low memory footprint,
and can be used to access the internet, or update tables. GNUPlot will be called
to create a line graph. So, Cron will be called to update the database at regular
10 second intervals. Now, GNUPlot can be called any time from the GUI with the
plot function. The plot function will be configured to plot a line graph with the
appropriate data in the X and Y axes. A new graph will be redrawn on every
window and subwindow as each window will be closed when navigating to
another window. So, GNUPlot needs to be opened and closed each time, and a
fresh graph drawn every time a window is opened. The following figure 3-5.3.1 is
an example of how the GNUPlot output should appear. As the device is turned
on and off, the graph records the data, and passes it along the data arrays. The
graph is then redrawn every time Cron makes an update.
Figure 3-5.3.1
3-7.4 Software Parameters
All of the programming and software requirements will be written in C or C++
which will be written in Microsoft Visual Studio 2010. Visual Studio was chosen
for its ease of debugging and large array of support and features like intellisense.
The majority of the libraries that were researched include libraries that can be
accessed in Visual Studio. The software will be written on an AMD based
processing computer using Microsoft Visual Studio 2010, and the Intel based
78
computers available in Engineering building 2, room 274 using Microsoft Visual
Studio 2008.
C++ is the language of choice for many embedded systems. C++ is a robust
object-oriented programming language with the ability to use large libraries of
functions that it can import. Once a designed object is implemented, it must be
instantiated somewhere in the code. First, the memory will be allocated, then the
object will be initialized. After these steps are taken, the object can be used.
The ARM will support its use and functions of libraries. The embedded system
will be loaded with a Linux based embedded operating system called embedded
Linux. The operating system is needed because of the complex procedures that
will take place such as opening and closing libraries for GNUPlot and MIniGUI.
C++ is highly compatible with Linux and its derivatives, so the programming will
be homogenous throughout the project.
The use of the software to the user will be very simple. As shown in figure 35.4.1, a user will open a window, and the window will change to the selected
window. At this time, MiniGUI will draw the window, and if necessary, GNUPlot
will plot any needed graphs on the page after accessing the database for the
relevant information.
Figure3-7.4.1
3-8 Embedded Linux
79
The operating system for the main control board will run embedded Linux.
Linux is in general a lightweight operating system currently in use for a
broad variety of embedded type systems like media players, mobile
phones, networking equipment, and automation. Embedded Linux has the
advantage of not requiring and royalties or licensing fees to use and
develop, as it is open source and free. It has a stable, regularly updated
kernel that is well supported and documented.
Embedded Linux has a version ported to the ARM series of
microprocessors. The company ARM, actually contributes to the
development for the operating system on their processors. ARM works
with many different types of Linux distributions including Canonical,
Debian, Fedora, Linaro, Maemo, Movial, and Thundersoft. Information
about how to develop on the platform can be found at
http://infocenter.arm.com .
80
4- Prototype
4-1 Design Planning
Each member of the group will be in charge of a portion of the project. The
project could have easily been implemented using a single microcontroller.
However, we decided to use three of them so each one of us can learn the skills
needed for this project. The use of three microcontrollers will require the use of
three printed circuit boards. This will allow each of us to learn how to solder parts
on the board. Also, each part of the design will be built and tested separately.
Each one of the circuit boards will have a microcontroller and wireless
transceiver. The other unit will have a current sensor and a power relay.
To continue with safety in mind, we will design our current sensor and solid state
relay circuit to fit into an AC wall box enclosure. The wall box enclosure will be a
PVC duplex socket box type. The circuit will be designed on a circuit board and
installed inside the box enclosure. A hole will be drilled in the side of the PVC
housing box to accommodate the bulkhead style stereo phone jack to support the
interconnect to control the solid state relay and sample the current utilized by the
appliance plugged into the power socket. And finally, a standard PVC cover plate
will be fastened to the duplex socket to prevent accidental exposure of high
voltages. The following figure is an illustration of the box type we will be using.
Figure 4-1.1: PVC electrical box
In prototyping the system for the solid state relay control and the sensor signal,
we have planned to use commonly available stereo phone cables for its three
conductors wiring. The cable that we will be using is a three foot length male to
81
male stereo audio cable. Since all the voltages on this cable are low voltages, it
will be sufficient for the application to carry the signal to switch the relay on and
off as well as sample the signal from the current sensor. The shield will be used
as a common or ground between the microcontroller, solid state relay and current
sensor. The following figure illustrates a standard stereo male to male audio
cable:
Figure4-1.2 : Male to male audio cable
As shown above, part 1 will be used as a signal path to the solid state relay
positive input side. The number 3 will serve as a ground and will connect to the
negative input side of the solid state relay as well as a common ground path for
the sensor signal which will be carried on the number 2 of the audio cable. To
provide us with a modular approach, we will use a 3.5 mm stereo audio phone
jack to accommodate the male to male audio cable described above. The stereo
audio phone jack will be on the power control and monitor side and on the
microcontroller module side.
In keeping with safety standards, a standard household AC power plug wall box
will be used for the power control and monitor module. It should contain the solid
state relay, the current sensor and a stereo phone jack for interconnect to the
stereo phone cable connected to the microcontroller.
The proof of concept model will be created by making a model home. The model
home will be constructed from wood, and painted to look similar to a house. The
model will be approximately three feet high, and four feet wide. The home will
have two of the four sides open, so the model can be rotated. This way, the
model can be seen from either side. Inside of the model will be our system.
Each module will be connected to its own power outlet at different locations in the
model. The touch screen unit will be secured to a wall where it can be accessed
at eye height.
The model must be powered by a separate power source, so an extension cord
will be required to get power from a nearby wall. In the model, the power will be
split to the three modules, and the main control unit. Each module will be
connected to a wall outlet.
82
The system will be demonstrated by using the touch screen to turn the light bulbs
on and off, while showing the graphs of their power consumption, and the ability
of the system to perform specified tasks.
The technical demonstration will consist of testing similar to the test plan of the
system. First, the touch function will be demonstrated. In this step, the
demonstration will consist of moving from window to window, showing the
accuracy of each touch, and the fact that the system responds to a touch. Next,
the demonstration will show the real time graphs of the unit measuring energy
usage. The main screen will be shown, following each of the subwindows to
demonstrate that they all function properly.
4-2 Parts Acquisition
We will purchase all of the components needed online. We are still considering
several suppliers and looking for the best prices we can get on the parts. For the
current sensor and the power relay, we are considering Honeywell Inc as they
carry a large amount of these parts to choose from. They also have a great
customer support. We are considering several suppliers when it comes to the
other parts.
4-3 Budget
The UCF Office of Sustainability will be sponsoring the project. We anticipate
building most of our parts in the UCF lab using parts that we already have on
hand. The cost may vary depending on the parts choices. The prices listed below
are estimates based on web searches.
The budget for the Home Energy Management System is a work in progress.
Parts have not been ordered, so exact costs are not yet known. There is an
emphasis on making the device as inexpensive as possible, but given the time
constraints, this may not be achieved. Prices for all components will be
compared to similar items for function first, then price. The total budget is $2018.
Each part will be ordered, and the unit price recorded. Then, the data will be
input to the table, and the overall cost of the project will be calculated. This value
will be compared to the budgeted value. The table below shows the components
that are planned to be purchased.
83
Quantity
Unit Price
Total Price
Budget
Over/Under
Multi-touch embedded
solution
DK-57VTS-LPC3250
1
Module Control Boards
PCB Printing
PCB soldering
Microcontroller
Zigbee Chip
1
3
3
3
3
$670.00
$30
$60.00
$600
$120.00
Concept Model
Drywall
Power outlets
Screws
Wood base
Light Bulbs
Extension Cord
Power Strip
1
3
20
1
3
1
1
$8
$5
$5
$10
$5
$10
$20
Misc parts
Documentation
1
2
$50
$25
$515.00
Total
$515.00
$500.00
$2,018
Table 4-3.1 Budget
84
5- Testing
Our testing approach will be broken into two phases. In the first phase each of
the components will be tested individually to ensure they meet their own
individual requirements. This phase will also assist in developing our
understanding of each component. After this phase our integration plan may
need to be altered as we understand each component’s functionality better. Then
in the second phase the entire integrated system will be tested including
functional and non functional tests.
5-1 Wireless Modules
There are a number of tests that must be performed. The unit must be tested to
ensure basic functionality, range, and reliability. The most basic test would be to
test basic functionality to ensure the unit can properly transfer data bytes. Then
the transceiver’s range must be tested. This must be done in the product’s
intended environment. A variety of environmental factors must be tested,
including having a transceiver that needs to reliably communicate up to 3 rooms
away. Incorporated into the range tests will be a test of reliability, with reliability
being measured as the percentage of the transmitted bytes that were received by
each transceiver as a function of the range.
The wireless module chosen is Xbee. The Xbee includes built-in security
features. These security features include the addressing method used by the
transceiver. There are over 65,000 unique addresses that can be set for each
transceiver. The source and destination of each message is addressed. The
system also has the ability to encrypt the sent data using 128-bit AES (Advanced
Encryption Standard). This will help secure information sent over the wireless
serial link between the user and the microcontroller unit. Even though the Xbee
does an excellent job of securing data when configured properly, the system is
designed so that the data being transmitted wirelessly is not very sensitive to
intrusion. It is just a matter of turning on and off appliances in one’s home, so
security should not be a huge issue. The two essential pieces of data that will be
sent though the transmitter are the power measured and the commands on
turning the power on and off an appliance or more.
5-2 Microcontroller
There are a number of tests that must be performed. The microcontroller must
be tested to ensure basic functionality and reliability. The most basic test would
be to test the pins functionality to ensure that the unit performs the appropriate
functions and have good signal integrity. The microcontroller’s operation needs
to be tested at a wide range of currents sensed in order to assure a reliable
product. The signal communication lines need to be tested for reliability, with
85
reliability being measured as the percentage of the transmitted bytes that were
received correctly. We have very few safety considerations when it comes to the
microcontroller. We only need to make sure that the right voltage is applied to it.
We also need to avoid overheating it. If the microprocessor becomes too hot, it
may burn individuals, cause damage to the unit, and even cause damage to the
whole system. This can be avoided by smart design. We will pay great attention
to detail and follow through with the manufacturer recommendations.
5-3 LCD touch screen
There are several tests that must be performed in order to make sure this part of
the project is functioning properly. For the touch screen, we will first test if the
screen responds to touch with clean hands. Then, each test will be done with a
mild quantity of dirt on the finger used. Last, each test will be done when
touching the system with the back end of a pen. The first tests to be performed
should ensure basic functionality and reliability. The most basic test would be to
test basic functionality to ensure the unit can properly display data and graphics.
As we plan on displaying real time graphs, we need to make sure that the power
consumption graph is accurate as well as the cost and Kilo watt hour. Then, the
LCD backlight should be tested in various conditions and temperature ranges.
This testing will allow us to set the appropriate contrast and brightness and allow
for minimum power consumption. Pixel reliability needs to be measured in order
to make sure that the LCD is displaying the transmitted data properly. We also
need to keep in mind that the LCD consists of a glass plate structure. It needs to
be properly insulated from the casing of the user interface unit in order to assure
minimal stress and decrease the possibility of shatter. On the two sides of the
LCD, protective plastic coverings will be placed in order to protect against
damage during shipping. These must be removed prior to design integration in
order to reduce circuit malfunction and the chance of overheating. While this part
of the project is straight forward, we need to keep in mind various risks that may
arise. The group is aware of the difficulty of graphic LCD programming. We are
in contact with some manufacturers to determine whether there exists support
with graphic LCD programming and researching current graphic LCD
implementation. Another risk that may arise is the difficulty integrating the LCD
with the rest of the product design. In order to decrease this risk, our group has
researched LCD communication protocols, driver circuits, compatible
microprocessors, and required features needed for LCD functionality.
The
screen will be tested to ensure proper brightness in dark environments or
ambient light situations. What will be tested is pixel accuracy and brightness. In
addition, the screen will be tested to respond to touches in the proper locations,
and send back accurate touch locations.
5-4 Power/Current Sensor
There several test that will be performed on this part of the design to ensure
basic functionality and reliability. The power/current sensor will undergo a
86
calibration procedure. We will first take a reading from Arduino monitor at 0 Amps
current. Then we will take a reading from Arduino monitor and at a current value
close to the maximum current we intend to measure. The equation 𝐲 = 𝐦 ∗ 𝐱 + 𝐜
will then be used to find the calibration. After the values of m and c have been
found, they can be inserted into the Arduino code. The current sensor we will be
using requires a supply voltage of between 6V and 12V. Its output is centered on
Vcc/2. The output voltage increases about 50mV per Amp when the supply
voltage is 12V.
5-5 Software
Software will be tested from low level to high level, with the lowest level including
the functions that facilitate data manipulation and touch recognition, middle level
functions representing GUI creation and graph plotting, and high level functions
representing graphing data in the GUI.
5-5.1 Function Tests and Touch Recognition
Low level functions
Low level functions are defined as functions which facilitate data manipulation,
and touch recognition. Tests will be performed one each function. This portion
of the testing will take more time than others, but it crucial to ensure the success
of further portions of the project. A success will be defined as the functions
return the proper data type with an accurate value.
Middle level functions
Middle level functions are defined as functions that create GUI windows or plot
the line graphs. A success will be defined as the GUI correctly generated, or the
plots generated with accurate, correctly oriented data.
High level functions
High level functions are defined as functions that graph data within the GUI. A
success will be defined as the graph correctly placed within the GUI, and the
graph plotting accurate data.
5-5.2 GUI Test Cases
The GUI test cases will test the functionality of the GUI changing screens and
modifying attributes. After the GUI tests are passed, the GUI elements will be
integrated in the project.
Case 1- Window changes
Case 1 will test the basic functionality of moving throughout the GUI. This will be
performed after the screen has been tested for returning proper touch locations,
87
which will be necessary for the results of this test to be valid. Special cases will
include multiple touches, and having wet hands and wearing gloves.
Case 1 will validate the following:
 The GUI will load to the main menu
 The GUI will load module A when module A is touched
 The GUI will load module B when module B is touched
 The GUI will load module C when module C is touched
 The GUI will change the TURN ALL OFF to TURN ALL ON
 The GUI will change the TURN ALL ON to TURN ALL OFF
 The GUI will open the options menu
 The GUI closes previous windows when a new one is opened
 The GUI on screen recognizes key presses of the number pad
 The GUI on screen stores the value of the number pad after enter is
pressed
A success in each category will be the function operates as describes.
Case 2- Graph plotting
Case 2 will test the plotting functionality of GNUplot. The functions to create a
line plot, and to fetch data from an outside source will be tested. This is simply a
test that the plot function is working, and a line plot can be successfully
generated. A success will be a line plot created from a data set of random
values.
Case 3- Graph taking in data from the modules
Case 3 will test the GNUplot’s ability to access the data in the database of the
flash memory. This is different from Case 2 in that the data will be input from the
modules A-C. The plots must represent actual data from the database which can
change in real time. A success will be a relevant graph generated from the data.
Case 4- Plots automatically update
Case 4 will test the ability of the system to automatically update the graphs. The
plots must be line plots without any graphical errors or faults. The plots must
represent actual data from the database which can change in real time. A
success will be a plot which updates after 10 minutes.
Case 5- Plotting inside the GUI
Case 5 will test the plotting functionality while inside the GUI. This will test if the
plot windows will appear and update. The plots must represent actual data from
the database which can change in real time. The plots must be line plots without
88
any graphical errors or faults. A success will be the plots updating inside the GUI
every 10 minutes.
Case 6 – Module Testing
Case 6 will test each module’s ability to be turned on and off, while the real time
graph displaying information. This test is a test of most of the system functioning
properly. A success will be shown if the device is turned off, and the graph dips
to zero energy use until the device is turned on again.
5-6 Power relay
The power relay we will be using is very reliable, but the constant work and heat
the relay is subject to might break its coil or make its contact points stick or burn
in time. This will prevent current from going through the circuit and ultimately
prevent the relay from functioning properly. To avoid such issues, we will test the
relay each time before using the system. The relay we will be using is a four
terminal relay. There are two tests that need to be considered when dealing with
a relay test or problem. The tests we need to perform will actually depend on the
problem itself. We will decide whether the problem is with the actual relay or the
power, ground or trigger circuits that activate the relay, in our case this part is the
microcontroller that sends the command 0 or 1 to the power relay. We will first
test the relay contacts, operation and activation circuit. Our power relay is more
prone to failure when it is warm. To test this part of the design, our power rely
can be considered into two separate halves. The first or primary half of the relay
utilizes an electromagnet to close the secondary electrical circuit inside the relay.
This electromagnet is activated by a simple power (positive side) and ground
(negative side). The second half of the relay is the switch that controls the power
to a particular accessory like a light bulb or a laptop or what ever the user is
using to test this system. The following steps will be used to test the power relay
and make sure it is functioning properly before connect it to the rest of the circuit
components:



We will test the relay incoming voltage using a voltmeter or test light, to
make sure it is receiving power.
We will then use a multipurpose meter to check for continuity between the
two power relay terminals. There are two control terminals on the relay.
The two thicker wires that connect to the relay hook up to the power
terminals and the other two go to the control circuit terminals. There
should be no continuity between the terminals.
We will the control circuit terminals on the relay by connecting a wire from
one of the terminals to the positive terminal of the control circuit and the
other relay terminal to ground.
5-7 Entire System Testing
89
The project will be tested in what will as closely as possible approximate a real
end user application. A scenario will be set up with electronics plugged in through
the measuring units. In this scenario the group hopes to demonstrate the vampire
draw that almost all electronics have inherent in their design, which one of the
project goals is to reduce. The scenario will then also demonstrate the power
control features based on differing constraints.
In the scenario, three household electronics will be plugged in and operated in
different modes to demonstrate the power metering capabilities of the sensors
and the graphing of the units on the display, showing the total power usage over
time as well as the power usage of each individual unit over time.
Then, the ability to use the touch screen to turn on or off specific units will be
demonstrated by allowing one of the evaluators to perform the task also
demonstrating the ease of use of the system to control remote (though not
remote in the scenario's case) outlets.
Once manual control has been demonstrated, automatic control will be
demonstrated, using pre-programmed settings to demonstrate the capabilities in
a timely manner, say an accelerated 24 hour period condensed into 5 minutes,
showing that depending on the 'time of day' the system will handle things
differently.
On top of all the automation demonstration, the system will also be tested against
commercial power metering products. Light bulbs will be used to do the testing
as their power usage can easily be calculated 'on paper' and the group can
compare the expected values with the actual values of both the commercial
products and the project.
Many different devices will be tested with loads ranging from large to minute, say
a stovetop versus the draw of the transformer in a cell phone charger, both
charging a phone and doing nothing but sitting there plugged in.
90
6- Reflections
6-1 Features excluded
When the group originally discussed this project, there were quite a few ideas
that were thrown out, as part of the design process. Originally, the group
considered a touch screen interface, and after several other versions were
discussed it came back to the touch screen display. The touch screen idea was
originally scrapped in favor of the project not having a display screen at all.
Instead, the group considered developing an application for the Android platform
that would be able to view the data and control the switches on the final device.
Another feature that the group would have liked to implement was a web
interface that would, as with the Android application, allow the user to control and
view the system from anywhere.
These two ideas were married together under the desire for user accessibility. If
the homeowner had, for instance, gone to their parent’s for the holidays and
accidently left an iron plugged into one of the units, they would be able to use
either the Android application or web interface to double check the iron, and if left
on, shut off power to the device. This would give the user a sense of confidence
that their house was safe. As well, being able to pull from both the Android
application and the web interface, the user would be given several options of
convenience to ensure that the minor investment of this product would be
rewarded with major payoffs in the future.
Another feature that could have been implemented is a smaller unit that does not
measure power. This unit would only control a relay switch connected to an
outlet. This would have allowed for more extensible use of the product as these
units would be cheaper and allow for users to control outlets all around their
house. If this feature were adapted to the online or Android application, the user
would find themselves with complete control over all power outlets within their
domicile and unquestionable ability to monitor power usage. When the user that
is a parent tells their child ‘Lights out!’ or ‘No more computer or television!’ there
would be no way getting around the truth of whether they used the electricity in
their rooms to continue to play their entertainment device on mute.
6-2 Future Improvements
Due to the lack of time and resources, not all desirable features or extensions
can be included in the project. Given more time, and proper resources, the
system would surely be more advanced and user friendly.
One future improvement is the idea that the accuracy of the system could be
improved to the level where it would be acceptable for any possible use – even to
91
the point where power companies would approve the use of the meter. In order
to do this, higher tolerances would need to be implemented, such as more bits in
the analog to digital conversion to give a more accurate reading as well as more
accurate analog parts. While accuracy of data transmitted was one of the major
components the group sought to take care of, the accuracy of the data collected
was equally important. However, due to the ability of the group and costs of the
bits, and high standards of the power companies, it would be difficult to improve
upon the original idea to the point where it would be acceptable for this type of
use at this time.
The device could also have been made smaller, and possibly cheaper, with more
development time and a greater in-depth knowledge of the different parts of the
system. If the group could have been expanded, or the time of the group’s efforts
could’ve been equivalent to small research and development division of a
corporation, it’s possible the group could have produced a smaller product.
However, with the means and ability of the group as it were, there was
unfortunately no way to create this desired tiny wall-fly gadget.
Another advanced feature would be the ability for the system to know if a person
is inside the house. This way, a user would never need to use the console to
dictate commands to turn modules on and off; the device itself would know. This
could be accomplished by interfacing a motion detector or an infrared camera to
detect people moving around in the house. This way, the system could turn any
rooms off that a person is not in.
In addition it would be nice to have a way to interface the product with anything
that doesn’t use a power outlet. This however, would involve either some minor
electrician work or the use of transducers over the lines leading into a unit. It
would be difficult to manage this as part of the current project, since it would
change with each and every household and would again, require the use of an
electrician or the homeowner to purchase and install a transducer. Further, if a
transducer were used, the homeowner would have to decide on an accurate
device so that the readings of the group’s final product were still reliable.
Another feature that could be implemented is a smaller unit that does not
measure power. This unit would only control a relay switch connected to an
outlet. This would have allowed for more extensible use of the product as these
units would be cheaper and allow for users to control outlets all around their
house. This would be perfect for the user if they wished to control an outlet while
away – for instance, at night the user would charge their electronic device (cell
phone, e-book, iPad), but during the day the charger remains unused yet
dissipates power. If a cost-effective alternative device could be created to ensure
the Vampire draw of this outlet is negated, the unit could be very valuable in
reducing the overall household energy costs.
92
Summing some of the previously mentioned improvements, scalability is a large
issue. If the system were implemented in a large scale commercial building as it
were being constructed, energy cost savings would be very immense. Imagine if
you will, a work environment that causes rooms with no occupants to use
absolutely no power. We believe that this is the essence of the project, and the
overall goal. We wish to provide businesses with this product, and allow them to
begin saving money and reach higher profits.
Another feature that the group would have liked to implement was a web
interface that would, as with the Android application, the user would be able to
control and view the system from anywhere. This would have been a great
feature as someone could shut off power to something when they are away from
the house, say an iron plugged into one of the units.
Unfortunately, the system itself must run on power. It is not yet known if this
device, when implemented in a home will save the owner energy. Therefore, it
would be beneficial if the device had its own power source such as a solar panel
or small wind turbine which could be attached outside the home.
More software features can be added such as the ability to control your unit from
the internet or a mobile phone, or text messages sent to you about your power
usage from your home. These features would add to the overall value and utility
of the system. Either of these features would need to include WiFi or Ethernet
support to the system, and a server to be included with the software package.
This may not be possible with the current processor and embedded solution.
Some old personal computer parts may be able to facilitate this function.
93
7- Conclusion
In recent years, the general population has become more aware of their power
usage. As the supply of power doesn’t always match the growth of need for
power and the resources to create power dwindle, energy costs will continue to
rise. It has happened in the past with gasoline, and the past scare and current
(and projected) costs of energy has sparked an interest in all kinds of people to
reduce their energy usage or eliminate it. The carbon footprint fear has pushed
people to create green houses made entirely out of recyclable waste, and green
buildings that rely on solar panels and natural lighting to provide enough light for
everyone within to see throughout the business hours.
It may not be feasible to claim that the world will eliminate their need for
electricity, but at some point the citizenry will have to closely watch their
electricity usage. Products like this project are going to become more ubiquitous
as a way to monitor power usage and eradicate waste.
This project was designed to help eliminate wasteful electricity usage, in the form
of vampire draw. The designed product will most certainly accomplish this, and
will even give users the ability to monitor their power usage on items that may be
old, are used often, or just electronics that they’re curious about the product’s
power consumption. Many people already do the basics – they turn off lights
when they’re not in the room, they turn off the television once they’ve finished
their favorite program, they close doors and windows when the air conditioner is
on, but they really have no idea how much they’re saving.
Unlike many others, this project gives users the option of viewing their power use
in terms of actual money instead of an abstract value that many people cannot
understand. It provides an easy to read and, more importantly, an easy to
understand view of the power usage of whichever outlets the user is currently
monitoring.
The user can also not only see how much the unit is currently costing, but can
see the trends in power use by the outlet, which can have a profound effect on
the user – especially when they see the outlet consuming power when the object
isn’t even being used. Since the user will be able to see the amount they are
saving – in actual dollars and cents – the user can completely predict their daily
energy usage if they are under a tight budget. With the downward economy as it
currently is, the group feels that this feature alone would make the product stand
out if a user was searching for a similar device.
While the device could always be improved, the group feels that the project is
strong enough to stand on its own in a growing sea of competitors. The product
should perform admirably well in any and all situations, and has value to anyone
that uses it. Even though the device could be made smaller and more accurate,
94
that is something that would come from later iterations, alongside the Android
application and web-based interface. The product will meet all of the standards
set forth by the group, allowing the user to reduce their home energy
consumption while becoming aware of – and reducing the effect of – vampire
draw. As this is the very first version of the product, and as a first iteration of a
product, it is quite amazing.
8- User Manual
To begin using our product, you need to download and install the software.
Before plugging in any of the measuring units, you should start the software.
Then the end user can plug the power cord and the transformer into the outlet
where they want to measure.
After plugging into the wall the end user may plug in what he or she wishes to be
measured.
All that is left is for the end user to use the software.
The software will run all on its own and the user can read the power being used,
the power they save by having the system turn the outlet off, and how much
power is used by the end of the month if their current power usage habits
continue.
The software interface allows the user to turn all the modules off, or single
modules individually.
95
Appendix
Appendix A: Image Permissions

RE: Permission
11/01/10
Sc, Info
info.sc@honeywell.com
To zinachuck@knights.ucf.edu, Sc, Info
From: Sc, Info (info.sc@honeywell.com)
Sent: Mon 11/01/10 7:08 PM
To: zinachuck@knights.ucf.edu
Cc:
Sc, Info (info.sc@honeywell.com)
Thank you for your interest in Honeywell Sensing and Control Products.
The information from our website, i.e. pictures, are public domian and can be used in
your information.
If you have any further questions, or need any assistance, do not hesitate to contact
us.
Honeywell International Inc.
Sensing and Control Division
Customer Response Center
Phone: 1-800-537-6945
International: 815-235-6847
Fax: 815-235-6545
E-Mail: info.sc@honeywell.com
Website: www.honeywell.com/sensing/
From: zinachuck@knights.ucf.edu [mailto:zinachuck@knights.ucf.edu]
Sent: Sunday, October 31, 2010 10:36 PM
To: Sc, Info
Subject: Permission
Hello,
96
My name is Zineb Heater. I am an electrical engineering student at the University of
Central Florida. My team members and I are working on our senior design project. We
are contacting you to ask for your permission to use a photo of a Hall Effect current
sensor from your website in our senior design document. We would greatly appreciate
the permission.
Thank you
Zineb Heater and the Team

RE: sales@andeng.net - Permission
11/10/10
Beth Anderson
banderson@andeng.net
To zinachuck@knights.ucf.edu
From: Beth Anderson (banderson@andeng.net)
Sent: Wed 11/10/10 2:59 PM
To: zinachuck@knights.ucf.edu
Attachments, pictures and links in this message have been blocked for your safety.
Show content | Always show content from banderson@andeng.net
Hi Zineb:
You can use the photograph for your senior design project.
Good luck.
Beth Anderson, P.E.
Anderson Engineering
20526 330th Street
New Prague, MN 56071
phone: 507-364-7373
fax: 507-364-7374
cell: 952-292-6415
email: banderson@andeng.net
1-800-893-4047
www.andeng.net
From: Paula Mills On Behalf Of Test Account
Sent: Wednesday, November 10, 2010 8:28 AM
97
To: Beth Anderson
Subject: FW: sales@andeng.net - Sending mail server found on relays.ordb.org - Permission
From: zinachuck@knights.ucf.edu [mailto:zinachuck@knights.ucf.edu]
Sent: Tuesday, November 09, 2010 10:32 PM
To: sales
Subject: sales@andeng.net - Sending mail server found on relays.ordb.org - Permission
My name is Zineb Heater. I am an electrical engineering student at the University of Central Florida. My
team members and I are working on our senior design project. We are contacting you to ask for your
permission to use a photo of a current Transformer from your website in our senior design document. We
would greatly appreciate the permission. Please find below a link to the photo we would like to use.
http://www.midwestcurrent.com/ctseries.php?submitted=1&voltAmps=15&ratio=&accuracy=
Thank you
Zineb Heater and the Team

RE: Permission
To see messages related to this one, group messages by conversation.
11/15/10
Reply ▼
Bill Bratly Add to contacts
To zinachuck@knights.ucf.edu
From: Bill Bratly (pickerwest@sbcglobal.net)
Sent: Mon 11/15/10 7:03 PM
To: zinachuck@knights.ucf.edu
Hotmail Active View
Hello Zineb,
98
Can you use the attached photo? This is a relatively new series of relay and a true power relay.
Best Regards,
Bill Bratly
Picker Components Corp.
(818) 997-3933
(888) 997-3933 toll free
Fax (818) 997-8903
pickerwest@sbcglobal.net
From: zinachuck@knights.ucf.edu [mailto:zinachuck@knights.ucf.edu]
Sent: Sunday, November 14, 2010 8:34 PM
To: sales@pickercomponents.com
Subject: Permission
My name is Zineb Heater. I am an electrical engineering student at the University of Central Florida. My
team members and I are working on our senior design project. We are contacting you to ask for your
permission to use a photo of a power relay from your website in our senior design document. We would
greatly appreciate the permission.
Thank you
Zineb Heater and The Team

RE: Permission
To see messages related to this one, group messages by conversation.
11/25/10
99
Fiona Davis
fiona.davis@jennic.com
To zinachuck@knights.ucf.edu
From: Fiona Davis (fiona.davis@jennic.com)
Sent: Thu 11/25/10 2:47 PM
To: zinachuck@knights.ucf.edu
Hotmail Active View
Hello Zineb,
This is ok - l have attached the picture for you. Good luck with your design project.
Best regards,
Fiona
From: zinachuck@knights.ucf.edu [mailto:zinachuck@knights.ucf.edu]
Sent: 24 November 2010 19:40
To: fiona.davis@jennic.com
Subject: RE: Permission
Hi Fiona,
We would like to use the Example application Diagram in this PDF file:
http://www.jennic.com/files/product_briefs/JN5148_PB_1v3_3.pdf
The photo of interest in page 2.
Thank you
Zineb Heater and the Team
From: fiona.davis@jennic.com
To: zinachuck@knights.ucf.edu
Subject: RE: Permission
Date: Wed, 24 Nov 2010 09:19:00 +0000
Hello Zineb,
Can l ask which photo you would like to use?
Best regards,
Fiona
From: zinachuck@knights.ucf.edu [mailto:zinachuck@knights.ucf.edu]
Sent: 24 November 2010 06:20
100
To: info@us.jennic.com
Subject: Permission
My name is Zineb Heater. I am an electrical engineering student at the University of Central Florida. My
team members and I are working on our senior design project. We are contacting you to ask for your
permission to use a photo of a Wireless Microcontroller from your website in our senior design document.
We would greatly appreciate the permission.
Thank you
Zineb Heater and The Team
Figure depicting power line communication
101
PPM figure
Sparkfun image
AD7755 and PIC Figure
102
Arduino pro mini size picture
Digi figures, all xbee related figures
Tweet-a-watt and xbee adapter pinout images
103
Google permission
from Dennis
<dennyroxsox@gmail.com>
to consult@minigui.com
Kilgore
hide details 6:02 PM
(21 hours ago)
date Sat, Nov 27, 2010 at 6:02 PM
subject Photograph permissions.
mailed- gmail.com
by
Hello,
My name is Dennis and I am working on a senior design project for the University of Central
Florida. I was wondering if I could have permission to use the attached image in my work.
from Dennis
<dennyroxsox@gmail.com>
to desktopmonitors@planar.com
Kilgore
hide details 8:34 PM
(18 hours ago)
date Mon, Nov 29, 2010 at 8:34 PM
subject Photograph permissions
mailed- gmail.com
by
Hello,
104
My name is Dennis and I am working on a senior design project for the University of Central
Florida. I was wondering if I could have permission to use the attached image in my work.
from Dennis
<dennyroxsox@gmail.com>
to sales@screentekinc.com
Kilgore
hide details 6:06 PM
(21 hours ago)
date Sat, Dec 4, 2010 at 6:06 PM
subject Photograph permissions
mailed- gmail.com
by
Hello,
My name is Dennis and I am working on a senior design project for the University of Central
Florida. I was wondering if I could have permission to use the attached image in my work.
from Dennis
<dennyroxsox@gmail.com>
to support@fdi.com
Kilgore
hide details 6:08 PM
(21 hours ago)
date Sat, Dec 4, 2010 at 6:08 PM
subject Photograph permissions
mailed- gmail.com
by
Hello,
My name is Dennis and I am working on a senior design project for the University of Central
Florida. I was wondering if I could have permission to use the attached image in my work.
from Dennis
<dennyroxsox@gmail.com>
to webmaster@digikey.com
Kilgore
hide details 6:09 PM
(21 hours ago)
date Sat, Dec 4, 2010 at 6:09 PM
subject Photograph permissions
mailed- gmail.com
by
Hello,
My name is Dennis and I am working on a senior design project for the University of Central
Florida. I was wondering if I could have permission to use the attached image in my work.
105
106