1 | P a g e

Body Temperature Regulation Vest for Quadriplegic Wearer – Phase II

December 11

2009

ECE 480 Design Team One:

Jarred Davis

Merav Nahoom

Christa Pline

Jawad Zaheer

Rachel Bouserhal

Facilitator: Ramakrishna Mukkamala

2 | P a g e

Executive Summary

One of the potential obstacles of a spinal injury is a reduced ability to regulate

one‟s body temperature. This inability of the body to regulate its own temperature in

extreme weather can impair basic body functions. This disorder typically accompanies

paralysis of the legs and arms, leaving a person wheelchair-bound. In an effort to provide

greater independence and increased safety for people with this disability, a jacket was

developed to regulate body temperature. This jacket controls the temperature of an

individual by heating or cooling water and sending it through a network of tubes in the

lining of the jacket. By expanding on this, a more convenient and versatile design was

made. An automated system that detects the ambient temperature and reacts accordingly

was added as well as a more convenient user interface and LCD screen.

Acknowledgments

Team 1 would like to thank:

Mr. Gregg Mulder

Dr. Charles Owen

Mr. Steven Blosser

Mr. Neal Langford

Dr. Rama Mukkamala

3 | P a g e

Table of Contents

EXECUTIVE SUMMARY ...................................................................................... 2 INTRODUCTION .................................................................................................. 4 BACKGROUND .................................................................................................... 5 DESIGN APPROACH ........................................................................................... 6 TECHNICAL DESCRIPTION .............................................................................. 11

Hardware ..............................................................................................................................................11 Peltier Junction ............................................................................................................................................11 PIC microcontroller ....................................................................................................................................12 Transistors ...................................................................................................................................................12 Relays ............................................................................................................................................................14 Vest ...............................................................................................................................................................15 Keypad ..........................................................................................................................................................16 Liquid crystal display(LCD) .......................................................................................................................20 Temperature Sensor ....................................................................................................................................21

Software..................................................................................................................................................23 User Interface ..............................................................................................................................................23 Pulse Width Modulation .............................................................................................................................26 TESTING ............................................................................................................ 29

Peltier Junction Testing ..............................................................................................................29

Duty Cycle Testing .........................................................................................................................31 Heating .........................................................................................................................................................31 Cooling ..........................................................................................................................................................33 Automatic .....................................................................................................................................................33

Testing with the User ....................................................................................................................35

FINAL COST ...................................................................................................... 37 CONCLUSION .................................................................................................... 37 APPENDIX 1 – TECHNICAL ROLES, RESPONSIBILITIES, AND WORK ACCOMPLISHED. .............................................................................................. 40 Merav Nahoom ............................................................................................................................................41 Christa Pline ................................................................................................................................................43 Jarred Davis .................................................................................................................................................44 Rachel Bouserhal .........................................................................................................................................46 Jawad Zaheer ...............................................................................................................................................47 APPENDIX 2 – LITERATURE AND WEBSITE REFERENCES ........................ 49

4 | P a g e

Introduction

A healthy human is able to maintain a constant body temperature of 98.6°F

(37°C) despite the temperature of the environment. In a hot environment, the body sends

a signal to the brain via the spinal cord with the information that the body is overheating;

the brain then sends a signal back down the spinal cord and tells the body to cool itself by

perspiration. Many people with spinal cord injuries, however, are unable to complete this

operation. With the loss of the ability to sweat, a person needs a careful control of his or

her environmental conditions or their body temperature will begin to rise excessively.

Similarly in a cold environment, the body may not be able to get the messages through to

the brain that the body is cooling down and if left untreated, the person will soon become

hypothermic.

There are a variety of clothing products designed to heat or cool individuals in

order to regulate body temperature. Some heating garments use a heat source connected

to stationary equipment, with a heating system attached to the clothing. Due to the high

power requirements, the battery power needed for this type of system is either very bulky

(lead acid) or prohibitively expensive (lithium-ion), resulting in very limited mobility

because of the electrical connections. Cooling vests use ice water that is pumped from a

reservoir and is circulated throughout the vest.

These designs, aimed at those working or living in harsh environments, often are

not intended for those with medical conditions, such as quadriplegia. Some quadriplegics,

as well as having trouble regulating temperature, have trouble sensing when something is

too hot or cold. Many of the heating products on the market use heating coils or similar

technology which can easily overheat and are unsafe for those with limited sensation.

5 | P a g e

Background

The goal of this project was to modify a heating/cooling system for the use of

quadriplegic wearers. Last semester, in combined effort with the MSU Resource Center

for Persons with Disabilities (RCPD), a jacket was designed to control the temperature of

an individual by heating or cooling water and sending it through a network of tubes lining

the jacket. The system attaches directly to, and is powered by, the users‟ wheelchair.

When activated, it pumped heated (or cooled) water though medical grade tubing in the

jacket, this in turn influenced the temperature of the wearer. There were limitations in the

design that are addressed to complete the project.

One of the limitations in the first design was power consumption. Since the

system is powered by the wheelchair‟s battery, power conservation was a necessity. The

peripherals draws large amount of current that drained the battery quickly. Pulse width

modulation was attempted to achieve energy conservation last semester. However, high

currents kept burning out the control system. Consequently, no pulse width modulation

was successfully implemented. A high current is also a limiting factor in implementing

any automated setting. In this design, a different approach was applied to achieve power

conservation. This approach steps down the voltage to an appropriate level; it will be

discussed further in the following sections.

For a design to be truly effective, it must be easy for the customer to use. In phase

I of the project there was limited user control. The only settings were off, heating and

cooling. The switches for these settings were not accessible by the user. Adding an

extendable keypad allowed not only additional heating and cooling options but also a

more accessible and convenient user interface.

6 | P a g e

With such a specialized product, a complete working design would be significant

as no product like this is currently available.

Design Approach

In addition to the limitations addressed in the previous design a number of new

features were added. Two separate approaches were taken in the design process. First, the

previous system needed to be addressed; this was done by researching and testing the

previously implemented components. The new components also needed to be researched

and analyzed. The design decisions were based on a number of design specification

including safety, power consumption, accessibility, comfort, portability and cost. These

specifications were rated (as seen in Table 1) and their weights in the component matrix

(Table 2) were used to aid in component decisions.

Table 1 Design Factor Matrix

Factor Relative Order of

Significance

Relative Weight

Safety 1 10

Power Consumption 2 9

Automatic Control 2 9

Accessibility 4 8

Comfort 5 8

Cost 6 6

Robustness 7 5

Portability 9 4

7 | P a g e

Table 2 Component Based Design Matrix

Design Component Microcontroller Thermoelectric

Device Garment Style Material

PIC PSoC Peltier Heater Jacket Vest Poly-Mesh Nylon

Factor Weight

Safety 10 5 5 9 6 5 5 5 5

Power Consumption 9 5 5 7 6 5 5 5 5

Automatic Control 9 6 8 8 6 5 5 5 5

Accessibility 8 5 5 5 5 4 7 5 5

Comfort 8 5 5 5 4 4 7 9 5

Cost 6 7 6 8 4 5 5 7 5

Robustness 5 5 5 5 5 6 5 9 5

Portability 4 5 5 9 3 5 5 5 5

Total 351 363 463 336 312 383 394 330

The basic functionality of the system was mapped out on FAST diagram and is

shown in Figure 1. Each block in the diagram was then broken down into the hardware

and software requirements as well as how they would be measured, applied, and

connected.

Figure 1: FAST Diagram

8 | P a g e

Once the components required to implement the new design were selected, an

estimate for the budget was compiled. Many components were already included in the

design from the previous semester. Each component was checked for usability, and

several were incorporated in the new design including the heatsink, fans, water pump,

tubing, and connectors. This aided in keeping the cost within the budget. Table 3 shows a

breakdown of the original cost estimate from the initial design phase.

Total Budget: $500

Table 3 Initial project budget estimate

Parts Cost Range

Vest $100 - $120

Temperature Sensors $20-50

Circuit Components $10 -$20

LCD Display $10-$30

Relays $10-$20

Microcontrollers $10 - $20

Thermoelectric Device $20-$30

Total $180 - $290

Once these decisions were made, a timeline was drawn up in a Gantt chart which

can be seen in Appendix 5. Although the timetable was a convenient gauge of where we

should stand in the design process, we were not able to adhere to it completely. We were

not grossly off track but the time required for troubleshooting was more than expected

due to which the time for testing and modification suffered. Despite these discrepancies,

the end goal was reached with time to finalize and refine.

9 | P a g e

One of the major changes that were made from the original design was to

implement the automated system without measuring the body temperature. In the

original design a body temperature reading was going to be taken into account for the

automatic control of the system. This reading was to be taken under the arm of the wearer

using a thermistor. Because the vest was designed to be worn over a shirt, an accurate

reading was very hard to obtain. If the wearer lifted their arm at all, the temperature

dropped drastically. To resolve this issue, we tried to take an average of readings but the

sensor took at least a minute to stabilize once it was disturbed.

Internal body temperature needs to be at a constant value of 98.6 degrees

Fahrenheit, an increase of one degree can be quite significant. These small changes are

hard to pinpoint without an accurate reading. The temperature under the arm is already

generally about one degree less than that of the internal temperature of the body and any

movement can change the reading.

A study was conducted, Experimental Hyperthermia in Traumatic Quadriplegia,

that showed the change in internal temperature in quadriplegics over time when in an

extreme environment. The results of this study were included in the decision process of

the duty cycles. The complete study can be seen in Appendix 3.

Another change made to the design was adding low and medium manual options

to the heating and cooling settings rather than the full on or full off options in the original

design. As a microcontroller was needed for an automatic setting this was a natural

extension of the system options. The medium and low setting could easily be obtained by

changing the operating duty cycle of the system controlled by the microcontroller. This

also aided in controlling the power issue, power is conserved when the system is in the

10 | P a g e

off portion of the duty cycle. Different duty cycles were tested and decided on according

to the test responses. These results can be seen in the testing section.

The same kind of thermoelectric device was kept from the previous design

because of its capabilities to heat and cool rather than just heat or just cool. Different

models of this component were tested (the results can again be seen in the testing section)

and a decision was made. These particular thermoelectric devices need a reversed current

polarity to change from heating to cooling and vice versa. A few options were discussed

for reversing the polarity of the current but in the end a relay system was decided upon to

do this task.

Another important decision in the design process was the choice of a

microcontroller. After some deliberation the decision was between a PIC (Programmable

Interface Controller) and a PSoC (Programmable System on Chip). Many factors were

compared including the cost, complexity, and limitations of each part. In the end, the PIC

was chosen because it was readily available and all team members were familiar with it.

The PIC had all the capabilities required to complete the design specifications of this

project and will be discussed further in the following technical description of the

components.

Implementing and programming the microcontroller was a major addition to the

system that permitted additional options desired for the overall system. These options

include extra manual options as well as the automatic setting. Using the design approach

of mixing digital and analog components was the difference between the new and old

system design and made all the options of the desired system attainable.

11 | P a g e

Technical Description

Our system required a great deal of both software and hardware design. The entire

system is monitored and controlled by a microprocessor. Using multiple relays, the

output signal from the microprocessor drives the pump and peltier junction of the device.

The operation and individual components of the project will be explained in more detail

in the following sections.

Hardware

Peltier Junction – This device is used to heat and cool the water that is

circulated through the jacket that is worn by the user using the thermoelectric effect. The

thermoelectric effect is an exchange in voltage for a change in temperature.

Thermoelectric cooling devices are typically used only for cooling, though some can be

used for both heating and cooling by reversing the polarity of the current flow; a peltier

junction is one such component.

There is a calorific effect of an electrical current at the junction of two different

metals. This effect is called the Peltier effect. When current flows through a circuit, heat

is exuded from the upper junction and absorbed at the lower junction. The heat absorbed

can be quantified by

Where Q is the heat absorbed by the lower junction per unit time, I is the current

through the circuit. is the Peltier coefficient of the entire thermocouple, and Πa and

Πb are the peltier coefficients of each material. Peltier cooler/heaters are solid-state active

12 | P a g e

heat pumps that transfer heat from one side of the device to the other. The peltier junction

is activated using a 13.8 volt source. When a positive voltage is sent to the peltier

junction, it outputs a cold temperature, inversely a negative voltage results in the warm

temperature output.

PIC microcontroller – The microcontroller is the driving force of this entire

project. PIC is a 40 pin microprocessor that has AC/DC converter capabilities as well as a

built in pulse width modulator. Both of these features were ideal in the context of this

design. To automatically adjust the temperature of the jacket the PIC receives a voltage

reading from a temperature sensor. Using the Analog-to-Digital Convertor (ADC) of the

microchip, an input voltage value is converted to a hexadecimal value. This value is then

used to determine the appropriate pulse width modulated output of the microcontroller.

Since the system is able to heat and cool, the microprocessor has two separate outputs for

each state. Temperatures below 70°F cause an output to occur at pin 16 of the PIC.

Similarly, temperatures above 80°F cause an output at pin 17. The pulse width

modulation is based on the severity of the temperature. For example, a temperature of

30°F will result in a pulse width modulated signal that is completely high, where as a

temperature of 50°F will produce an output signal that is high only 50 percent of the time.

The internal Pulse Width Modulation of the PIC did not satisfy the specifications of the

system‟s design. However, this was implemented with software.

Transistors – Transistors are used to boost the current of the output signal from

the PIC microcontroller. The relay used to control the cooling of the Peltier junction

13 | P a g e

requires a 5 V and 50 mA signal to switch states. The output signal of the microprocessor

is 5 volts, but only 7 mA. Thus, the current of the output signal has to be increased.

Originally, a line driver IC was implemented to amplify the current; however, this design

had the drawback of causing the output voltage to drop to 3.6 volts. This voltage level

then became too low to switch the relay. By implementing a transistor stage to the output

signal of PIC, current is allowed in increase without causing a voltage drop. The output

of the cooling relay will be going to the double pull double throw (DPDT) relay that will

be switching the Peltier junction. Another transistor stage was needed to boost the current

going into the DPDT from the heating pin of the PIC. Current boosting stages using

transistors are simple and require little calculations.

Figure 2: Current Boosting Circuit

The basic circuit for such a stage is shown in Figure 2. First, the correct transistor must

be chosen. To do so, the correct current gain and maximum collector current must be

figured out. To find the maximum collector current needed the equation below is needed.

IC (max) Vs

RL

To find minimum required gain the equation below could be used. Where

IL is the

current through the load, and

IIC is the maximum current sourced by the microcontroller.

14 | P a g e

HFE f 5 IL

IIC (max)

For our purposes, the 2N222A transistor satisfied all the requirements. Finally, the base

resistance is based on the equation below where,

VIC is the voltage coming out of the

microcontroller.

L

FEICB

I

HVR

5

Figure 3: Current Boosting Switching and Transistor Stage

1

2

3

4

5

1

3

56

4

2

8 7

24V

13.8V

13.8V

5V

8.2 K

4.7 K

HEATING

PIN 16

COOLING

Pin 17

PIC

18

45

20

DP

DT

RE

LA

Y

SP

DT

RE

LA

Y

Relays – Two different relays are used in this design, a single pull double throw

(SPDT) relay and a double pull double throw relay (DPDT). The DPDT is a power relay

with a control voltage of 24V. Figure 4 is a pin diagram of the DPDT relay. Pins 8 and 7

are the control voltage pins while pins 5 and 6 are the output pins. When no voltage is

applied to pins 7 and 8, the output receives the voltage seen at pins 1 and 2. When the

relay is switched on, the output receives the voltage seen at pins 3 and 4. The Peltier

15 | P a g e

junction needs +13.8V to cool and -13.8V to heat. Pin 8 is connected to the output of the

transistor heating stage and pin 7 is connected to 24V. Pin 5 and 6 are connected to

ground and input of the Peltier junction respectively. Pin 4 and 3 are connected to 13.8V

and ground respectively. Pins 2 and 1 are connected to ground and the output of the

cooling SPDT relay respectively. This is done to create an “OFF” state when using Pulse

width modulation. Therefore when the microprocessor sends a signal from the cooling

pin the SPDT relay will turn on and the Peltier receives 13.8V. During a cooling cycle

when the cooling pin goes low, for pulse width modulation, the relay sees no voltage

from pins 1 and 2 and the Peltier receives 0V. Similarly when the microprocessor sends a

signal from the heating pin the relay turns on and sends a -13.8V to the Peltier, which is

the voltage at 4 and 3. Again when the heating pin is low in a heating pulse width

modulation cycle the Peltier receives 0V which is the voltage seen by pins 1 and 2. A

schematic of this circuit is shown in figure 3.

Figure 4: Power Relay

Vest – The idea of changing the original jacket concept to a vest came from the

sponsor of this project. This change, along with being easier to manufacture, makes the

garment easier to use for the consumer. To make the vest as versatile as possible,

channels were sown in the lining for the tubes, rather than attaching them directly. There

are more channels than tubes to permit the tubes to be repositioned for maximum

comfort. This also allows the tubes to be completely removed for cleaning purposes.

16 | P a g e

Keypad – The 4 rows x 3 columns keypad matrix (12 keys), 96AB2-152-R made

by Grayhill Inc, is used to allow the user to choose the system‟s mode of operation. Even

though the keypad has 12 keys, only seven keys are required for the options of this

project. The first three buttons on the first row determine the manual heating level while

keys four through six on the second row control the manual cooling level. The seventh

key activates automatic mode. The initial idea was to connect all the matrix pins directly

to the PIC, but to save pins on the microcontroller it was decided to implement the

system using one pin. This implementation requires the use of the ADC on the PIC. Each

button on the keypad is represented with a different voltage value; this was done using

multiple voltage dividers. The output voltage is send to the pin 34 of the PIC. The

schematic of the keypad interface required to give a unique value is shown in the figure 5

while the different voltage value for each key is shown in the Table 4 below.

17 | P a g e

Figure 5: Keypad Schematic

Table 4: Voltage values of keypad buttons

Button Operation Value (V)

Range of

Operation (V)

1 Hi Heat 0.1412 0.1 – 0.2

2 Medium Heat 0.3057 0.25 – 0.35

3 Lo Heat 0.4876 0.4 – 0.5

4 High Cool 0.5904 0.55 – 0.6

5 Medium Cool 1.1549 1.1 – 1.2

6 Lo Cool 1.6636 1.6 – 1.7

7 Automatic 1.9349 1.8 – 2.0

5V

15KΩ 3.3KΩ 680Ω 180Ω

560Ω 680Ω 820Ω

PINS 1 2 3 4 5 6 7

1nF

Row 2 Row 3 Row 4 Col 2 Col 3 Col 1 Row 1

18 | P a g e

Throughout the assembly process there were many unexpected problems that

occurred in the system. When the automatic and manual modes of the system were first

tested, the system was wired almost completely on a proto board with the ambient

temperature simulated with a voltage source. This was possible since the system reads the

temperature as a voltage value. For example, 0.78 volts seen at the pin is equivalent to

78°F. The microprocessor worked perfectly using this method of testing. In the next

phase of testing, the entire system was transferred to a breadboard. This transferal caused

some of the input values going to the microcontroller to change. As a result, the system

did not work as anticipated. The programming of PIC uses voltage ranges to determine

the appropriate output. Each button of the keypad has an input voltage value that fits into

a specified voltage range as indicated in Table 4. Because of the input voltage values

changing, the specified voltage ranges needed to be changed as well. To ensure that the

voltage ranges were accurate each button voltage was tested using a digital multimeter.

Once the new ranges were programmed into PIC, the system worked properly.

Another source of error developed when the LCD was programmed to display

ambient temperature. The programming is written to display the input temperature value

whenever the user presses a button on the keypad during the manual operation of the

system. In automatic mode, the ambient temperature is recorded and displayed every

three minutes. When the LCD was first programmed, the low cooling mode and the

automatic mode were both displaying the wrong temperature value. The rest of the

operation modes did not share this issue. The reason for this problem was attributed to the

fact that the buttons triggering the automatic and low cooling modes had voltages that

were very close to the reference voltage that is used in the ADC conversion of the PIC.

19 | P a g e

For the purpose of this project, PIC uses a 2.5 volts reference to convert decimal values to

hexadecimal values, thus any input value close to 2.5 would cause an inaccurate

conversion. To remedy this issue the voltage ranges for each button were lowered by

choosing different resistor values so that the output voltage from the keypad was never

greater than 2 volts. The current circuit diagram of the keypad with modified resistor

values is shown above in Figure 5.

When the automatic function of the code was initially tested it did not behave as

expected. In the correct execution of automatic mode the system is programmed to read

ambient temperature and output a signal based on that value. If automatic mode is being

initialized for the first time, it will output a constant value for one minute based on the

input temperature to allow the peltier junction to establish an initial value. After the

minute has expired the code determines the appropriate pulse width modulation output.

However, the original coding caused the peltier to output cooling for one minute

whenever automatic mode was initialized no matter what the ambient temperature was.

After extensive debugging, it was determined that the initial cooling was caused by a

voltage spike that occurred when the entire system was turned on. The programming

would output a signal for cooling when it received an input voltage over .80 volts. When

the system would turn on the voltage spike had a high enough voltage to trigger the

cooling condition in the program. To fix this problem a delay has been added (Figure 6)

to the beginning of the main function. This allows the system to wait approximately ten

seconds before reading the input voltage from the temperature sensor.

Figure 6: Delay Loop

while(1)

{

DelaySec();

if(timeCnt >= 10)

{

timeCnt = 0;

break;

}

}

20 | P a g e

Figure 7 - User Interface

Liquid crystal display(LCD) – LCDs are electronically controlled light valves.

An LCD with a small number of segments, such as those used in digital watches and

pocket calculators is used in this project. The LCD will display the current ambient

temperature and the mode that the user chose which makes it easier for the wearer to

know about the surroundings. The wiring diagram for connecting LCD to the

microcontroller is not complicated and it uses the parallel data port to send the data. We

have utilized 10kΩ potentiometer in order to control the intensity of the screen. The

schematic of LCD connected to the PIC is shown in figure 8 below. Additional

information on those pins could be found in the software section on the LCD.

21 | P a g e

Figure 8: Wiring Diagram of LCD connected to PIC

PIC

18

f45

20

GND

VDD

Contrast

Reg Select

R/W

Enable

Data0

Data1

Data 2

Data 3

Data 4

Data 5

Data 6

Data 7

LC

D M

OD

UL

E

5 V

RD0

RD1

RD 2

RD 3

RD4

RD5

RD6

RD7

RC5

RC7

RC6

Temperature Sensor

The temperature Sensor is used to measure the ambient temperature on which the

system does heating and cooling. The National Semiconductor designed the precision

integrated-circuit temperature sensors, LM34 whose output voltage is linearly

proportional to the Fahrenheit temperature and does not require any external calibration.

It outputs approximately +10mV/°F over its wide operating temperature range of -50°F to

+300°F. As it is calibrated in Fahrenheit temperature compared to other linear sensors

calibrated in degrees Celsius, the user is not required to do extra calculations to obtain

convenient Fahrenheit scaling.

The LM34 only has three terminals. Pin 1 is connected to ground while pin 3 is

connected to a 5 volt source. The output of the sensor is connected to Pin 2 of the PIC

microprocessor. This reading is sent to a built-in Analog to Digital converter on the PIC

22 | P a g e

which stores the value in hex. Figure 9 shows the wiring diagram of how it is connected

to PIC.

Figure 9: Wiring diagram of Temperature sensor with Microprocessor

LM 34

PIC

18

f45

202

The output of the sensor is used to determine which mode of Peltier junction (heating or

cooling) with PWM, should start. The temperature is checked every minute to account for

any changes. The actual ambient temperature will also be sent to the LCD screen for the

user to know the surrounding temperature.

The sensor is attached outside the system box beside the switch so it can detect

the ambient temperature as shown in figure 9.

According to initial plans, we connected the output of the sensor with 100kΩ and

connected to -5V to operate the sensor over its full temperature range of −50°F to

+300°F. As it was hard to get -5V from the system and the ADC of the PIC cannot take

in negative voltages. Thus, we decided to directly connect the sensor to pin 2 and operate

the sensor over limited range of positive temperatures.

23 | P a g e

Software

Since the entire system is monitored and operated through the PIC

microcontroller, the programming was very extensive. Two separate functions, labeled

„automatic‟ and „manual‟, are used to execute the system‟s mode of operation. A larger

function labeled „main‟ contains both of the previously described functions and allows

the system to run continuously when going between automatic control and manual

control operations. The upper level design of the software in figure 10 is broken down to

the two main parts mentioned above.

Figure 10: upper level design

Within the automatic and the manual functions, the code was broken down into smaller

functions. Each function has its own single purpose and will be discussed further in the

following section.

User Interface

Keypad - In order to determine which button was pressed in the keypad, as

discussed in the previous section, each key press generated a different voltage. One of the

libraries supported by the PIC allows reading voltage and then converts it to a

hexadecimal value. The library that is used for that conversion is ADC.h, it converts an

Is Automatic?

start

Automatic Manual

yes No

Check button status Check button status

24 | P a g e

analog value to a digital value to be used in software. After reading the value from the

key pressed, demonstrated in figure 11, the value that was read is compared to a range of

values to determine which key was pressed. The hexadecimal value is calculated with a

reference voltage of 2.5 volts. The hexadecimal value is a 10 bit voltage value. This

means we have 1024 bits. Therefore, the calculation is as follows:

Figure 11 – Function to determine a key press

LCD Screen - The LCD connects to eleven pins on the PIC. There are eight I/O

data bus line pins on the LCD that is connecting to the PIC. Those pins control the

displayed message on the LCD. The Register Select (RS) pin of the LCD is set to zero

when giving instruction input to the LCD, and set to one when giving data input to the

LCD. The Read Write pin (R/W) is set to zero in order to write to the LCD and set to one

when reading from the LCD. Enable (E) pin enables the signal to allow reading and

writing to and from the LCD.

void KeyPressed()

{

keyValue = Read ADC result // reference voltage 2.5

if(keyValue >= 0x0014 && keyValue <= 0x0028) // 0.1-0.2 V

buttonNum = '1';

else if(keyValue >= 0x0033 && keyValue <= 0x0047) // 0.25-0.35 V

buttonNum = '2';

else if(keyValue >= 0x0051 && keyValue <= 0x0066) // 0.4-0.5 V

buttonNum = '3';

else if(keyValue >= 0x0070 && keyValue <= 0x007A) // 0.55-0.6 V

buttonNum = '4';

else if(keyValue >= 0x00E1 && keyValue <= 0x00F5) // 1.1-1.2 V

buttonNum = '5';

else if(keyValue >= 0x0147 && keyValue <= 0x015C) // 1.6-1.7 V

buttonNum = '6';

else if(keyValue >= 0x0170 && keyValue <= 0x0199) // 1.8-2 V

buttonNum = '7';

}

25 | P a g e

The main two functions to control the LCD are presented in Figure 12 in pseudo

code.

InitLCD() – clears the screen and set the cursor to the home position of the screen. The

W/R pin is set to zero to allow writing messages to the screen. Before each instruction is

transmitted RS is set to zero and E is set to one after transmission RS is set to zero.

DATA includes the eight I/O pins. After each instruction, a delay is implemented. The

delay is to allow time for the LCD to react before implementing the next instruction.

InputChar(char x) – outputs a character to the screen. In the case of this function, a data

input is sent to the LCD; therefore, the RS pin is set to one.

Figure 12 – LCD functions

void InitLCD()

{

int i;

RW = 0;

RS = 0;

E = 1;

DATA = Clear Screen

E = 0;

Delay();

RS = 0;

E = 1;

DATA = Set cursor to home position

E = 0;

Delay();

}

void InputChar(char x)

{

RS = 1;

E = 1;

DATA = x;

E = 0;

Delay();

}

26 | P a g e

Pulse Width Modulation

As mentioned earlier, Pulse Width Modulation (PWM) is the preferred method to

regulate motor speed because no additional heat is generated. In addition, it is energy

efficient when compared to linear regulation. The PWM duty cycle is defined as

Ton/Toff (%) in one period and the range is 0% - 100%. Controlling the pulse width

modulation of the system is an important aspect of this project. By controlling the duty

cycle of the Peltier Junction, the heating or cooling rate of the water can be controlled.

Moreover, implementing PWM will help in controlling and reducing the amount of

current that is drawn from the battery, hence reducing power consumption.

The period achieved by using the PIC‟s PWM is about 400 µs. The formula used to

calculate it is as follows:

PR2 – is an 8-bit PWM register

TOSC – is a timer oscillator

TMR2 Prescale Value – Range of prescale value for timer 2.

For a 40MHz crystal oscillator used in the system, TOSC is equal to , PR2 is 255 and

max TMR2 prescale is 16. Therefore, the maximum period will be

= 256 • 4 • • 16 = 409.6 µs

For more details on how to calculate the period for PWM refer to the PIC data sheet on

page 149.

A higher period was required for this system (one minute period). Therefore, a

different approach was taken in order to accommodate the necessary period.

27 | P a g e

Timers are a PIC peripheral function that can run in the background without interfering

with the rest of the program. On the 18F4520 there are four timers that can be used,

which run via hardware in the background. These timers are numbered Timer0, Timer1,

Timer2, Timer3. A timer can be set up using the timer.h library functions. In this example

code, Timer0 is setup, so the internal clock can be used:

// Setup the timer with a 1:256 prescaler with 16 bits resolution

OpenTimer0( TIMER_INT_OFF & T0_16BIT & T0_SOURCE_INT & T0_PS_1_256 )

Figure 13 – PWM using the internal clock of the PIC

The frequency of the internal clock is based on the frequency of the external clock

divided by four. In this case, the external clock is 40MHz; therefore, the frequency of the

internal clock would be 10MHz. In order to create a delay of one second, the prescale

value needs to be considered. By dividing the frequency of the internal clock by the

prescale value of 256 a result of 39,062 is obtained. In the DelaySec() function in Figure

13 the clock gets incremented until it reaches the desired value (39,062). To create a

period of one minute with a Duty Cycle of 50%, the function will count thirty seconds

before it will change the state of the pin.

unsigned int timeCnt = 0;

void DelaySec()

{

while(ReadTimer0() < 39062)

;

WriteTimer0(0);

timeCnt++;

}

OpenTimer0(TIMER_INT_OFF & T0_16BIT & T0_SOURCE_INT & T0_PS_1_256);

while(1)

{

if(timeCnt > 30)

{

timeCnt = 0;

PORTDbits.RD3 = !PORTDbits.RD3; }

DelaySec();

} CloseTimer0();

28 | P a g e

Figure 14 - Counter Values and PWM Output

Figure 14 shows how the counter works with the PWM that is implemented in Figure 13.

The explanation for the calculation of the timing and the frequency of the internal clock

can be obtained from the PIC data sheet pages 123-125.

29 | P a g e

Testing

Peltier Junction Testing

As it was one of the more important components, extensive testing was performed

to decide which peltier junction had the best performance. Figure 15 and 16 shows the

graphs of the performance of three different peltier junctions for heating and cooling.

Figure 17 shows the current drawn by each device

The first model, Peltier 1, was the original from last semester; the device was

supposed to be optimized, but it was drawing more current than expected. Because of this

fact a new eltier, Peltier 2, of the same model was obtained and tested as well. This drew

less current but had a similar time response. The third, Peltier 3, was another model rated

at a lower current. Despite being rated at a lesser current, the device drew the same

amount as Peltier 2 and had a slower response time. Peltier 2 was chosen as the final

thermoelectric device.

30 | P a g e

Figure 15: Heating Response of Peltiers

Heating

0

10

20

30

40

50

60

0 10 20 30 40 50

Time (minutes)

Wate

r

Tem

pera

ture

(C

) Peltier 1

Peltier 2

Peltier 3

Figure 16: Cooling Response of Peltiers

Cooling

0

5

10

15

20

25

0 10 20 30 40 50

Time (minutes)

Wate

r

Tem

pera

ture

(C

)

Peltier 1

Peltier 2

Peltier 3

Figure 17: Current Response of Peltiers

Current Draw

0

1

2

3

4

5

6

7

8

9

0 2 4 6 8 10 12

Time (minutes)

Cu

rren

t (A

)

Peltier 1

Peltier 2

Peltier 3

31 | P a g e

Duty Cycle Testing

Heating

Another area where a large amount of testing was required was the duty cycles for

the automatic and high/medium/low settings of the manual option. As many different

types of tests were performed as time allowed. The first round of testing was done before

the microcontroller was implemented. Figure 18 shows a comparison of water

temperatures reached by the system for a 40% duty cycle both inside (at an ambient

temperature of 25 oC) and outside (ambient temperature of 10

oC). The system at this

time was running without a person wearing the jacket, so the body temperature was not

affecting it.

Figure 18: Duty Cycle Testing

Duty Cycle 40%

-10

0

10

20

30

40

50

0 10 20 30 40 50

Time (Minutes)

De

gre

es

Ce

lsiu

s

Inside (25 C)

Outside (10 C)

Duty Cycle

32 | P a g e

After the microcontroller was implemented and the system was operational in all

aspects, more testing was done to decide which duty cycles would be the most beneficial

to the system. For high heating, a 100% duty cycle was decided on. Figure 19 shows

three different options for medium heating. Medium heat 1 is the original estimate of a

60% duty cycle with a period of one minute. Medium heat 2 is the water temperature

with 60% heating and 40% cooling (period of one minute), and medium heat 3 is a 40%

heating duty cycle with a period of five minutes.

Figure 19: Medium Heat Options

Medium Heat Duty Cycles

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

45.00

50.00

0 2 4 6 8 10 12 14 16

Time (Minutes)

Wate

r T

em

pera

ture

(C

)

Medium Heat 1

Medium Heat 2

Medium Heat 3

Figure 20 shows the low heating options.

33 | P a g e

Figure 20: Low Heating Duty Cycle Options

Low Heating Options

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

45.00

0 10 20 30 40 50

Time (minutes)

Wate

r T

em

pera

ture

(C

)

Low Heat 1

Low Heat 2

Testing for the final prototype was different from previous testing. With the first

tests, water temperature was tested in an open reservoir. When bringing the system out

into a very cold environment a lot of heat was lost through this open reservoir method. To

conserve heat as much as possible the reservoir was (done away with), the casing was

closed as tightly as possible, and as much insulation was added as possible to the tubes

that were exposed to the air. In these tests, a thermometer was placed inside the vest and

the temperature readings were taken in this manner.

Cooling

The high cooling was also set at 100%. The cooling response of the Peltier is a

little more limited. A medium duty cycle was chosen at 60% and a low at 30%.

Automatic

The automatic setting were based on the previous duty cycle testing, expected

values of body temperature readings (see Appendix 3 for study results), and input from

the sponsor. Table 5 shows the ranges of temperatures and corresponding duty cycle

settings.

34 | P a g e

Table 5: Initial Duty Cycle Estimates

Temperature Range Setting Duty Cycle

Below 40o Heating 100%

Between 40 o

and 50 o

Heating 80%

Between 50 o

and 60 o

Heating 60%

Between 60 o

and 70 o

Heating 40%

Between 70 o

and 80 o

Off 0

Between 80 o

and 90 o

Cooling 50%

Above 90 o

Cooling 100%

After testing the Duty Cycle with the above values, some changes were made.

The results for 80% and 100% Duty Cycle were very close. Also, the 60% and 40% Duty

Cycle were too high for the temperature rage. The new values for the new Duty Cycle are

indicated in Table 6 below.

Table 6: Updated Duty Cycles

Temperature Range Setting Duty Cycle

Below 40o Heating 100%

Between 40 o

and 50 o

Heating 60%

Between 50 o

and 60 o

Heating 40%

Between 60 o

and 70 o

Heating 20%

Between 70 o

and 80 o

Off 0

Between 80 o

and 90 o

Cooling 50%

Above 90 o

Cooling 100%

35 | P a g e

Testing with the User

Now that the duty cycles were chosen and tested, it is necessary that we test the

system on the user. The vest needs to maintain the user‟s body temperature in extreme

weather. First, the user was taken outside where the ambient temperature was about

32oF. His initial body temperature reading was 98

oF. His body temperature was

monitored for 20 minutes. Readings were taken and recorded every ten minutes. After 20

minutes of being outside his body temperature dropped to a 97 oF with the system off. We

then turned on the system. Given the ambient temperature, the system was on the highest

heating setting. After ten minutes his body temperature was raised to a 97.5 o

F and then

maintained at that magnitude for the rest of the test. Figure 21 below represents the

results of the test. The blue portion of the graph represents the time period with the

system off, and the red region represents the time period the system is on.

This test insured proper functioning of the system in a worst-case scenario. The

system was outside with the user for 20 minutes and not turned on; all components were

cold before the system was turned on. The pump sees more resistance in cold weather,

and thus becomes slower. In normal use the system will turn on the moment the ambient

temperature was detected. It will take it a shorter time period to heat the vest since the

components would not be cold already.

36 | P a g e

Figure 21 – results of user test

37 | P a g e

Final Cost

Table 7 Final Cost

Parts Cost Range

DC/DC Converter $12.63

Temperature Sensors $1.25

Keypad $12.26

Fabric $15

Tubing Connectors $27.48

Relays $4.69

Zipper $3

Total $76.31

Conclusion

Our system is designed to be used with quadriplegics who have some mobility. It

has an automatic setting. This is done by an ambient temperature sensor which translates

the read temperature into a voltage. This voltage is then converted to hexadecimal and is

read by the microprocessor. Depending on the value read by the temperature sensor, the

microprocessor will drive the system to go in one of six modes: four ranges of heating

and two ranges for cooling. Different ranges of temperature were achieved by using pulse

38 | P a g e

width modulation. For example, the highest heating cycle would have a 100% duty cycle

and the lowest would have a 20% duty cycle.

The output signals of the microprocessor will go into transistor current boosting

stages. This is done so that the relays could be driven. The outputs of the transistor stages

will go into the switching circuitry. By manipulating this circuitry +13.8V sent to a

thermoelectric device, the Peltier junction, when the system is cooling and -13.8V when

the system is heating. It also sends 0V when the pulse width modulated signal is low.

This system also has a manual override using a 12 button key pad. In case the user

wants to break out of the automatic setting and choose the heating or cooling intensity a

keypad was added to the system. The keypad again provides the user with six different

modes of heating or cooling. It also allows the user to revert back to automatic mode if

desired.

To make the user interface more convenient and LCD screen was added. The

LCD screen displays the current ambient temperature and what mode the system is in.

The LCD and keypad will be extended from the system, on the back of the wheelchair, to

the user. Thus, allowing the user to have constant monitor over the ambient temperature

and constant control over how hot or cold the system gets.

Our system offers a more independent life for quadriplegics with some mobility.

It offers a safe way of giving them the ability to support extreme weather. The next step

is to universalize this design. Some quadriplegics do not have any sort of mobility, but

still experience problems in body temperature regulation. However, controlling this

system would be impossible. A way offer them an ability to use this system would be to

add a voice activated command that would respond to the users control requests with no

39 | P a g e

need to be mobile. Another improvement on this design could be done on the casing. A

better insulated system could lead to better performance in extreme weather. The casing

could also be altered to function with no need to be covered, which can also lead to better

results.

The system designed satisfies all design requirements, making this a safe product

to be used for body regulation with quadriplegic. Refer to the figure 22 below for the

schematic of the system.

Figure 22 - schematic of the system

40 | P a g e

Appendix 1 – Technical roles, responsibilities, and work

accomplished.

The technical roles were divided up at the beginning of the semester and can be

seen in Table 1.

Table 1 Initial Technical Roles

Name Technical Roles

Jarred Davis Power

Christa Pline Circuit Assembly

Rachelle Bouserhal User Interface

Jawad Zaheer Temperature Sensors

Merav Nahoom Software/Programming

This was adapted many times during the semester and collaborative efforts were required

for most areas of the project. Following is a personal description from each member of

the technical work they did throughout the semester.

41 | P a g e

Merav Nahoom

As the only Computer Engineer, I was assigned to the role of developing software

for the microcontroller and implementing a user interface. After the hardware

components were chosen, I had to focus my research on the data sheet of the PIC. As

a result of this, I learned how to access different data, read ports, and understand the

memory management of the PIC. In addition, I learned how to effectively use the

MPLAB, which is the C language editor and compiler.

I had to design a workflow chart for myself and break the work into sections. I

started by listing all the components that needed to be implemented in order to

develop a successful product. After listing all the components, I prioritized them. The

priority of each component was my guideline for the rest of my work. I started by

implementing the outlined goals of the automated part of the system. My team and I

decided that we will detect ambient temperature. The ambient temperature will be the

deciding factor of whether the system will be cooling or heating. I started by reading

the temperature from the thermistor, by utilizing the ACD library that the PIC

supports. Next, I used a simple “if” and “else” statements and divided the different

temperature ranges to perform different functions. A big portion of the development

42 | P a g e

of the automated system was based on the function I wrote to implement Pulse Width

Modulation that will satisfy our design specifications.

I started by developing a test case for Pulse Width Modulation, which is the most

important aspect of the automated system of our project. First, I produced high and

low pulses. Next, I had to implement it by using a clock. A clock, in this case, is very

important in controlling the period of each cycle. Another major test case was the

user interface, which includes the keypad and the LCD. I approached this by

separating the LCD to a separate test code from the keypad. This way I could test

them separately. After perfecting the code for each one of those components I

combined them together to the main code.

In my work chart I broke down the components of our design to different

functions. I also followed this pattern in developing my code. Before adding any

component to the main code, I created a separate file that implements only a specific

component. Only after it was working on its own, according to the standards that I

specified in my design stage, I combined it into the main code. By doing this, I

managed to debug the code more easily and eliminate potential problems in the early

stages of the development.

Extensive testing should take place in every new product developed with an

emphasis on safety and reliability. Ideally, the programming stage would be the last

thing in developing this kind of a system. However, because of time constraints, I had

to develop the software as the hardware was being developed. Therefore, I developed

a small test board that simulated the system and tested my code on that board. I

believe this gave us some more time for testing and fixing bugs.

43 | P a g e

Christa Pline

The first major technical aspect that I worked on was changing the thermoelectric

device. The original test of the system showed that this element was drawing quite a

bit of current. A new model and new component of the same model were both

ordered. After changing the Peltier Junction I ran the same kinds of tests that had

been run on the old system, for current and temperature. There was less current draw

but there was also less temperature response. The Peltier was changed again with the

new component of the same model number had the same temperature response of the

first component but less current draw.

The next section that needed to be worked on was making the relay system. To

change the polarity of the current we decided on a relay system. Two relays were

required with transistor stages to boost the current and it took the whole group effort

to get the desired output of the second relay.

At the beginning of the project I wrote some pseudo code for how the automatic

and manual setting to function. The programming was not an area that I worked

extensively on but there were some issues with the automatic program that I helped to

work out. Because of the large number of changes in the code some of them are used

in the final version and some of them have been built upon and some done away with.

44 | P a g e

Assembling and testing the system was the final part of the project and I did a lot

of work in this area. Once the programming was done for the day and uploaded on the

microcontroller I would assemble the system and test for functionality. This was

greatly a group effort as well. Each of us has done a fair share of testing of the

components separately and as a whole system throughout the design process.

Jarred Davis

My technical portion of the project involved implementing power boosting

techniques and integrating the software of the project with the hardware. Merav

Nahoom wrote the code required to run our entire project. Once Merav finished

compiling a portion of the code I had the opportunity to test the code using the

hardware of our device. Using the debugging mode of the MPLAB software I was

able to trouble shoot the programming, if there were any problems in the functionality

of the project I could locate the problem in the code and make corrections.

For our system to function properly we need to amplify the output signal of our

microprocessor. I implemented a line driver component in an attempt to boost the

current level. This implementation proved to be useless since the line driver caused

the current to increase to the appropriate level but unfortunately the voltage dropped

to a level too low to run our system. As a result my teammate, Rachel Bouserhal,

45 | P a g e

implemented a current booster using transistors which proved to be successful in

running our project. After Rachel completed the assembly of the transistor stage I

tested the system with Christa Pline, and we discovered that the system was

functioning properly.

The keypad was wired in two different ways. I was responsible for the first

method of wiring. I used three different voltage dividers, which allowed each row of

the keypad to have a different voltage value. I used this method because the

architecture of the keypad does not allow the microprocessor to distinguish between

rows; it could only distinguish columns. Thus I used the voltage dividers to make

each row distinguishable. For this method to work properly three microprocessor pins

are needed. This method however, was not implemented into our current design.

Jawad Zaheer developed the current wiring method using some of the concepts from

my design. This design is an optimized version of the previous implementation, and

only requires the use of one microprocessor pin.

There were multiple issues with in system operation when the hardware was

transferred from a proto board to a bread board. It was my responsibility to evaluate

these problems, and correct them. This involved testing the microprocessor input

voltages and making changes to the programming based on these values. Hardware

connections also had to be checked and corrected if necessary.

46 | P a g e

Rachel Bouserhal

I am a very analog and hardware oriented person. My technical role involved the

manipulation of switching circuitry and analog circuits. Our project needs both a

+13.8V supply and a -13.8V supply. Since power is supplied from the wheel chair its

unsafe to convert the supply voltage. I created a switching circuit using relays that

will supply the Peltier junction both +13.8V and -13.8V when it is cooling and

heating respectively. Also, using circuit manipulation I created an “OFF” state where

0V is being sent to the Peltier junction when it is low in a pulse width modulation

cycle. This involved a lot of trial and error. Working on it until I achieved my desired

results, helped me better understand relays.

Since the inputs of the relays are outputs of the microprocessor, each input needed

a current booster to trigger the relay. I also created each current boosting stage from

the outputs of the microcontroller into the relay. This involved some analog circuitry.

I had to look back to my analog classes to design a transistor stage that I can drive

into saturation given the specifications of our system. This involved using the correct

transistors and the correct base resistances. This is a useful circuit to understand. It

can be used in many applications.

47 | P a g e

After everything was implemented, testing was a very important part of our

project. Understanding the effects of different duty cycles of pulse width modulation

on the temperature of the vest could only be done through testing. After both the

manual settings and automatic had been programmed I tested the system. To ensure

that the results we expect are received. Our project was testing heavy and everyone

did a portion of it. Testing helped us better understand our system and make any

calibrations to it we need to have it perform the way we want it.

I have focused my education on circuits and analog theory. I am glad that I was

able u se my knowledge in those areas and contribute to my team. Everyone on the

team had significant roles that were necessary for the completion of the project.

Jawad Zaheer

During the entire design process, I was given several responsibilities. In

the beginning, I was assigned to develop a temperature sensor that can detect the

ambient temperature. It is one of the integral parts of the project as the heating and

cooling of the system primarily depends on it. I researched several elements that

measure temperature such as thermistors, thermocouples. Thermistors are nonlinear

to any temperature scale and needs calibration while thermocouples have low output

levels and require cold junction compensation. Then I came across National

48 | P a g e

Semiconductor part, LM 34, which fulfilled our project needs. The output voltage of

LM34 is linearly proportional to the Fahrenheit temperature and does not require any

external calibration.

I was also involved in developing the user interface. I setup the LCD to the

PIC microcontroller so software can be implemented on it. I also created the current

keypad design implementation with multiple resistors going into single pin of the

PIC. I also worked with the Sponsor, Stephen Blosser, for the case for the user

interface. The user interface is extended from the system to the case using the printer

port cable to make it easier for wearer to use.

I was also responsible for integrating different components together as

well as troubleshooting hardware and software issues. After everyone finished their

task, the whole system was first implemented on the breadboard. After that was

successful, the whole system was transferred to project board and tested again. There

was extensive testing required to understand the Pulse Width Modulation required

optimizing the system. Once all issues in the design were accounted for, I finally

made the PCB layouts of the project using the program Eagle. From designing to

troubleshooting, it gave me an opportunity to apply my knowledge and gave me a

chance to learn about various aspects of the design.

49 | P a g e

Appendix 2 – Literature and website references

2N222. Digital image. Kmitl. Web.

http://www.kmitl.ac.th/~kswichit/npn/2n2222.gif

ECE 302: Electronic Circuits. Wierzba, Gregory M. Print.\

Embedded C, Pont, Michael J. Print.

Expiremental Hypothermia in Traumatic Quadriplegia. Totel, G.L., R.E. Johnson,

F.A. Fay, J.A. Goldstein, and J. Schick. International Journal of Biometeorolgy 15.2

(1971): 346-55.

High-Performance Modules. http://www.tetech.com/Peltier-Thermoelectric-Cooler-Modules/High-Performance.html

MPLAB C18 Libraries

http://ww1.microchip.com/downloads/en/devicedoc/MPLAB_C18_Libraries_51297f.pdf

MPLAB C18 User’s Guide

http://ww1.microchip.com/downloads/en/devicedoc/MPLAB_C18_Users_Guide_51288j.

pdf

PIC18F4520 Data Sheet

http://ww1.microchip.com/downloads/en/DeviceDoc/39631B.pdf

Transistor Circuits. Welcome to the Electronics Club. 12 Nov. 2009.

http://www.kpsec.freeuk.com/trancirc.htm#pnp