Compact Remote High Altitude Balloon Launcher

1
COMPACT REMOTE HIGH
ALTITUDE BALLOON LAUNCHER
GROUP DESIGN PROJECT FINAL
REPORT
Supervisor:
Dr András Sóbester
Team Members:
Christian Balcer
Sabin Kuncheria Purackal
Reetam Singh
Rahul Kharbanda
Sullivan Pal
Binay Limbu
6th
April 2014

2
Academic Integrity Statement
We the undersigned confirm that the material presented in this final project is all our
own work. References to quotations from, and the discussion of work of any other
person have been correctly acknowledged or cited within the report in accordance with
the University guidelines on academic integrity.
Signed: ………………………………………………………………………………………
Names: ……………………………………………………………………………………...
Date: ………………………………………………………………………………………

3
1 ABSTRACT
Written by Christian Balcer
The mission of the project was to create a system that could be placed in a location and
subsequently remotely launch a high altitude balloon for atmospheric research upon receiving a
GSM signal. A lightweight, portable, and remotely activated high altitude helium balloon
launcher is designed and constructed. A radiosonde that can relay atmospheric data via radio is
also assembled. This report details background research, design, engineering analysis and testing
of the final system. Engineering analysis including a Failure Mode Effect Analysis,
Computational Fluid Dynamics study, Wind tunnel test, and Finite Element Analysis is
conducted in order to aid the design process and reduce the risks of failure on testing. An
outreach program in partnership with the European Association of Aerospace Students took
place in order to inform interested parties about the project. The final launch test as of this date is
documented and demonstrates a successful launch bar some minor failures that can be easily
rectified.

4
2 ACKNOWLEDGEMENTS
We would like to thank Phil Oxborrow for providing nylon piping and assistance in sourcing
parts for the gas system. All members of the Engineering and Design Manufacturing Centre
(EDMC) for design advice and manufacture of the gas release mechanism. Phil Crump for
helping track the payload during launch and unabated enthusiasm. Dave Cardwell for assistance
and running of the wind tunnel. The ISVR department for general electrical advice and provision
of solder. Matthew Brejza and UKHAS members for providing technical advice regarding the
payload. European Association of Aerospace Students (Euroavia) for providing sponsorship and
a platform for the outreach activity. Dr Andras Sobester for continued advice, support, and
provision of equipment.

5
3 TABLE OF CONTENTS
1 ABSTRACT 3
2 ACKNOWLEDGEMENTS 4
4 LIST OF FIGURES 8
5 LIST OF TABLES 12
6 NOMENCLATURE 13
7 STATEMENT OF OBJECTIVES 14
8 INTRODUCTION 15
9 BACKGROUND RESEARCH 16
9.1 HIGH ALTITUDE BALLOONS 16
9.2 CUSTOMER REQUIREMENTS 18
9.3 LEGAL ISSUES 22
10 HELIUM GAS SUPPLY 23
10.1 INTRODUCTION 23
10.2 CALCULATIONS 23
10.2.1 CALCULATION OF LIFT AND GAS VOLUME REQUIRED AT GROUND 24
10.3 PLOTS 24
10.4 GAS CANISTER CHOICE 26
10.5 GAS CANISTER TEST 28
11 GAS DELIVERY AND BALLOON RELEASE SYSTEM 29
11.2 GAS RELEASE DEVICE 29
11.2.1 REQUIREMENTS 29
11.2.2 CONCEPTS 29
11.2.3 DESIGN AND MANUFACTURE OF THE GAS RELEASE MECHANISM 31
11.3 BALLOON INTERFACE 34
11.4 GAS PIPING 35
11.5 BALLOON RELEASE MECHANISM 35
12 LAUNCH BOX DESIGN 40
12.1 DESIGN REQUIREMENTS 40
12.2 CONCEPT DESIGNS 40
12.3 MATERIAL SELECTION 43
12.4 OPENING MECHANISM 44
12.4.1 SERVO AND TORQUE CALCULATIONS 45
12.5 FINAL BOX DESIGN 48
12.6 CONSTRUCTION 50
13 BOX ELECTRONICS 53
13.1 PRESENTATION 53

6
13.2 STRATEGY 55
13.3 CONCEPT 57
13.4 CALIBRATION 59
13.5 CODING 64
14 PAYLOAD ELECTRONICS 65
14.2 OBJECTIVES 65
14.3 PAYLOAD BINARY MATRIX 66
14.4 PAYLOAD OPTIONS AND SELECTION 66
14.5 HARDWARE SELECTION 68
14.5.1 MICROCONTROLLER 68
14.5.2 FINAL MICROCONTROLLER SELECTION 70
14.5.3 RADIO TRANSMITTER 70
14.5.4 GLOBAL POSITIONING SYSTEM 71
14.5.5 SENSORS 72
14.6 DATA TRANSMISSION AND DECODING 72
14.7 DECODING 78
14.8 PAYLOAD CONSTRUCTION 79
14.9 FINAL PAYLOAD AND ACTIVATION AT LAUNCH 80
14.10 THERMAL 83
14.11 PARACHUTE 85
14.12 CONCLUSION 86
14.13 FUTURE WORK 86
15 ENGINEERING ANALYSIS: 87
15.1 CURVE LEVER DEFLECTION CALCULATION 87
15.2 OPENING MECHANISM DESIGN MODIFICATION 90
15.2.1 TEST-1 90
15.2.2 CONSIDERED DESIGN SOLUTIONS 90
15.2.3 FINAL DESIGN SOLUTION 92
15.2.4 CONCLUSION 94
15.2.5 FUTURE WORK 95
15.3 COMPUTATIONAL FLUID DYNAMICS ANALYSIS 96
15.3.1 INTRODUCTION 96
15.3.2 DESCRIPTION OF APPROACH 96
15.3.3 COMPUTATIONAL PROCEDURE 100
15.3.4 RESULTS AND ANALYSIS 101
15.3.5 CONCLUSIONS AND DISCUSSION 105
15.4 WIND TUNNEL TESTS 107
15.4.1 EXPERIMENTAL SETUP 107
15.4.2 RESULTS 109
15.4.3 FUTURE WORK 118
15.5 FAILURE MODE EFFECTS ANALYSIS (FMEA) 119
15.5.1 FMEA TABLE 119
16 OUTREACH 131
17 LAUNCH 134
17.1 PLANNING 134

7
17.2 EXECUTION OF LAUNCH 135
17.3 TRACKING 137
17.4 CONCLUSION OF SECOND LAUNCH 140
18 POTENTIAL MARKET 141
19 PROJECT CONCLUSION 143
20 REFERENCES 144
21 TEAM TASK ANALYSIS 146
22 FINANCIAL REPORT 149
23 APPENDICES 151
23.1 APPENDIX A: TABLE OF NICHROME-WIRE GAUGES AND TEMPERATURES REACHED FOR
DIFFERENT CURRENT (IN AMPS) [36] 151
23.2 APPENDIX B: INTERNATIONAL STANDARD ATMOSPHERE CONDITIONS [3] 152
23.3 APPENDIX C: CODE FOR BOX WITHOUT SENSORS USED IN THE FLIGHT TEST 153
23.4 APPENDIX D: TRIGGER CODE 156
23.5 APPENDIX E: CODE FOR BOX WITH SENSORS (NOT IMPLEMENTED) 157
23.6 APPENDIX E: DATA USED TO CALIBRATE THE WIND SENSOR 160
23.7 APPENDIX F: PAYLOAD PROGRAM 161
23.8 APPENDIX G: WIND TUNNEL TESTS RAW DATA 165

8
4 LIST OF FIGURES
Figure 9-1: A typical weather balloon.....................................................................................16
Figure 9-2, Tree diagram displaying the key customer requirements for the launcher design......19
Figure 9-3: Binary weighting matrix of customer requirements ...............................................20
Figure 9-4: Normalised importance of the customer requirements............................................21
Figure 10-1, Plot of how the diameter of the balloon varies with altitude for various masses of
helium...........................................................................................................................25
Figure 10-2, Plot showing how ascent rate at ground varies with payload weight for various
masses of helium............................................................................................................25
Figure 10-3, Maxxiline 2.2L gas canister [4] ...........................................................................27
Figure 11-1, Sprocket to be attached to release mechanism [6].................................................31
Figure 11-2, Engineering drawings of gas release mechanism supplied to the EDMC
accompanying manufacture instructions..........................................................................32
Figure 6-3, Solidworks model of release mechanism (not showing threads) ..............................32
Figure 11-4, Release mechanism connected to ........................................................................32
Figure 11-5, release mechanism screwed to gas cylinder with rubber o-ring..............................33
Figure 11-6, one-way valve ....................................................................................................34
Figure 11-7, Left: valve glued to adaptor, Right: valve glued to adaptor with balloon taped on..35
Figure 11-8, Diagram of concept hotwire method for releasing balloon when inflated...............36
Figure 11-9, Diagram of concept cutting method for releasing balloon when inflated................36
Figure 11-10, Diagram of concept fuse-wire method for releasing balloon when inflated...........37
Figure 11-11, Left: Photo of final hotwire design attached to tape around piping. Visible is the
outline of the one-way valve above, and below the outline of the o-ring, Right: Photo of
hotwire design only........................................................................................................38
Figure 11-12, Final design schematic showing gas system and balloon release mechanism........39
Figure 11-13, Photo of the gas system excluding the canister and release mechanism................39
Figure 12-1, Box concept design 1..........................................................................................40
Figure 12-2, box concept design 2 ..........................................................................................41
Figure 12-3, Box concept design 3..........................................................................................42
Figure 12-4, concept design for the opening mechanism (side view).........................................44
Figure 12-5, 3D diagram of the box opening mechanism.........................................................45

9
Figure 12-6, Curved lever used for the box opening mechanism (dimensions in mm).Concept
design 1. ........................................................................................................................45
Figure 12-7. Concept design 2 for the box opening mechanism................................................45
Figure 12-8, free body diagram of the box lid with the forces ...................................................46
Figure 12-9, image of biplex fluted polypropylene sheet [1] .....................................................48
Figure 12-10, final box design ................................................................................................50
Figure 12-11, image of gusseted plastic angle bracket [ [14]] [15]] ............................................51
Figure 12-12, constructed curved lever ...................................................................................51
Figure 12-13, construction box for the wind tunnel experiment................................................52
Figure 12-14, images of the constructed launch box ................................................................52
Figure 13-1: launch process....................................................................................................53
Figure 13-2: A photo of the bare electronics used in the test.....................................................54
Figure 13-3: Time taken for GSM to initialize.........................................................................56
Figure 13-4: Initial Flow Diagram for the Program .................................................................56
Figure 13-5: Illustration of the "on" and "off" time ..................................................................57
Figure 13-6: System Wiring ...................................................................................................58
Figure 13-7: Subsystem Wiring ..............................................................................................59
Figure 13-8: Time for the hotwire to melt the one-way valve connection..................................60
Figure 13-9: Hotwire Anemometer Calibration at 8.3°C avg....................................................61
Figure 13-10: Hotwire Calibration at 9.8°C average ................................................................61
Figure 13-11: Overview of the value read by the sensor to the wind and temperature................62
Figure 13-12: Alternative angle to show temperature effects ....................................................63
Figure 13-13: Hotwire Curve fit .............................................................................................64
Figure 14-1 Binary Matrix for the payload..............................................................................66
Figure 14-2: From left Raspberry pi, Arudino Mega 2560,Arduino Uno, Arduino Pro Mini. ....70
Figure 14-3 Radiometrix NTX2B radio transmitter.................................................................71
Figure 14-4 Adafruit Ultimate GPS Breakout. ........................................................................71
Figure 14-5 DHT22...............................................................................................................72
Figure 14-6 BMP085 Pressure Sensor.....................................................................................72
Figure 14-7 Examples of voltage divider schematics................................................................73

10
Figure 14-8 Configuration of Radiometrix NTX2B for data transmission.................................73
Figure 14-9 Circuit diagram for voltage divider calculation......................................................74
Figure 14-10 Circuit diagram when GPIO is high (5V) and GPIO is low (0V). .........................75
Figure 14-11Flowchart of the sketch running on the Arudino. .................................................77
Figure 14-12 Radio receivers connected to computers running dl fldigi for decoding.................78
Figure 14-13 Dl Fldigi decoding the data................................................................................78
Figure 14-14 Balloon plotted on the spacenear.us....................................................................79
Figure 14-15 Payload components being soldered together......................................................80
Figure 14-16 Final Payload with all the components working together.....................................81
Figure 14-17 Final Payload with a net weight of 59gm. In reference to a ruler ..........................81
Figure 14-18: circuit diagram of payload ................................................................................82
Figure 14-19: thermal profile of atmosphere based on us standard atmosphere 1976 .................83
Figure 14-20: Thermal insulation of the payload.....................................................................84
Figure 15-1, Free-body diagram of the force acting on the curved lever ....................................87
Figure 15-2 Left: Curved lever meshed for the FEA analysis. Right: Displacement plot of the
curved lever ...................................................................................................................88
Figure 15-3, Left: Stress plot of the curved lever, Right: Strain plot of the curved lever..............89
Figure 15-4, Free-body diagram of the lever design .................................................................92
Figure 15-5, Image of the new lever design attached to the canister..........................................93
Figure 15-6, closed box model ...............................................................................................97
Figure 15-7, balloon model....................................................................................................97
Figure 15-8, balloon with box with same dimensions as previous models .................................98
Figure 15-9, representation of mesh and domain for box with balloon model ...........................99
Figure 15-10, Contours of static pressure for steady balloon simulation at 4m/s where flow is in
the x-direction..............................................................................................................101
Figure 15-11, Pathlines released from surface showing a 3d view of airflow around the balloon at
1.305 seconds after the simulation is initiated. ...............................................................102
Figure 15-12, Pathlines released from surface showing a view from the rear of airflow around the
balloon at 1.305 seconds after the simulation is initiated.................................................102
Figure 15-13, contours of static pressure, flow in z-direction..................................................103
Figure 15-14, Path lines emitted from surface of the airflow around the box and balloon at 2.93
seconds after the simulation was initiated......................................................................104

11
Figure 15-15, Side view of path lines emitted from surface of the airflow around the box and
balloon at 2.93 seconds after the simulation was initiated...............................................104
Figure 15-16, front view of the experimental setup inside the wind tunnel..............................108
Figure 15-17, angled side view of the experimental setup.......................................................108
Figure 15-18, view from inside the box, displaying the attachment of the nuts and bolts to the
strut.............................................................................................................................109
Figure 15-19, Box lid fixed at an angle of 225° facing away from the wind .............................110
Figure 15-20, Box lid fixed at an angle of 225° facing towards the wind .................................111
Figure 15-21, Graph showing how drag varies with wind speed for two distinct cases.............111
Figure 15-22, graph showing how drag varies with wind speed at different lid opening states ..113
Figure 15-23, Graph showing how drag varies with wind speed when the box lid is shut ........114
Figure 15-24, Graph showing how drag varies with wind speed when the inflated balloon is
attached to the box.......................................................................................................115
Figure 15-25, Front view of experimental setup with inflated balloon secured to the box.........116
Figure 15-26, Graph showing the short term repeatability demonstrated during the experiment
...................................................................................................................................117
Figure 16-1, Christian, Rahul & Reetam at the UKSEDS Annual Conference 2014 at Leicester
...................................................................................................................................131
Figure 11-2, The group explaining the project to Conference Attendees .................................132
Figure 16-3, Reetam giving a presentation on the conference about the project.......................133
Figure 17-1: Photograph of the launch site at New Forest......................................................134
Figure 17-2: Map displaying 50 simulations ran using the ASTRA planner prior to launch .....135
Figure 17-3: Photograph of the inflated balloon ready to be released......................................136
Figure 17-4: Photograph of the balloon and the payload ascending after being released ..........137
Figure 17-5: Photograph of the retrieved balloon and payload, found near Amesbury.............137
Figure 17-6: Eagle eye view of the flight of the balloon..........................................................138
Figure 17-7 Flight of the balloon across different cities and town. ..........................................138
Figure 17-8 Altitude vs. Temperature measurements.............................................................139
Figure 17-9 Altitude vs. humidity measurement....................................................................140
Figure 17-10 Altitude vs. pressure measurement ...................................................................140

12
5 LIST OF TABLES
Table 10-1, Table showing variation of balloon diameter with altitude and important
atmospheric parameters as assumed for isa conditions for when the mass of helium is
0.03767 kg (the mass contained in the maxxiline 2.2L 100Bar canister).............................27
Table 11-1, SUMMARY OF PROS AND CONS FOR DIFFERENT GAS RELEASE
SYSTEMS.....................................................................................................................30
Table 12-1, advantages and disadvantages of each concept design ...........................................42
Table 12-2, Advantages and DISADVANTAGES OF the considered box material [ [10]][ [11]][
[12]] [13]........................................................................................................................43
Table 12-3, calculated values of wind pressure and force at different wind speeds .....................46
Table 12-4, calculated values of the required torque at different box lid opening angles.............47
Table 14-1: pros and cons of different payload concepts. .........................................................68
Table 15-1, boundary conditions for different meshes..............................................................99
Table 15-2, results for the steady closed box simulation for various mesh sizes .......................101
Table 15-3, results for the steady balloon simulation for two mesh sizes.................................101
Table 15-4, table showing results for open box with balloon steady simulation .......................103
Table 15-5, Table showing how the overall drag force exerted on the balloon at certain wind
speeds was obtained .....................................................................................................116

13
6 NOMENCLATURE
Cd Coefficient of drag
ρ Density of air (kg/m3
)
V wind velocity (m/s)
I second moment of inertia ( m4
)
δx Deflection in the horizontal direction (m)
δy Deflection in the vertical direction (m)
F Force (N)
T Torque (Nm)
M Bending moment (Nm)
E Material young’s modulus (MPa)
Pa Pascals
Patm Atmospheric pressure at ground level (Pa)
P Pressure at specified altitude
R Specific gas constant
Mhelium Mass of helium
Mpayload Mass of payload
r Radius of balloon
d diameter of balloon
ISA International standard atmosphere
Lnet Lift of the balloon subtracted by the gravitational force of the balloon and
payload.
Lneck Lift of the balloon subtracted by the gravitational force of the balloon mass.
g Gravitational constant
S Frontal surface area of balloon assuming it is perfectly spherical
N Newtons
GPS Global Positioning System

14
7 STATEMENT OF OBJECTIVES
 To design and construct a ‘shoe-box’ sized, reusable and portable launcher that can be
stacked and transported easily.
 To design and construct robust gas supply system for inflating the balloon.
 To design and construct an automated method of sealing and releasing an inflated
helium balloon.
 To design and construct a lightweight sounding system with live telemetry.
 To design an electronic subsystem inside the box capable of enduring two weeks in the
field.
 To conduct computational and experimental testing of the design to estimate physical
behaviour.
 To conduct an outreach event to raise public awareness about the project.
 To demonstrate a successful automated remote launch.

15
8 INTRODUCTION
Written by Reetam Singh
Over the years meteorologists and scientists have been known to use weather balloons for upper
atmospheric observations to carry payloads to very high altitudes, at a very reasonable cost.
These stratospheric balloons are used by the scientific community on a regular basis to gather
atmospheric data; using a payload which is tethered to the balloon. The sensors on this payload
measure temperature, humidity, wind speed and also atmospheric gas concentration. The data
from these sensors are transmitted through radio channels to a ground station and a continuous
track is maintained on the probe through GPS. These balloons are made of synthetic rubber or
latex which have high elasticity and coefficient of expansion which bursts eventually at high
altitudes (35km-40km) thereafter the probe lands on the ground with the aid of a parachute. For
inflation of these balloons; gases like Hydrogen, Helium or Methane are used which are known
to have lower mass compared to atmospheric air. Hydrogen and Methane due to its explosive
nature are only used in cases when weight of the payload is more than 1 kg. For amateur
purposes, Helium gas is used due to its inert nature and availability.
The Southampton University’s ASTRA group is known to develop new technologies for making
low cost atmospheric research. It came up with the requirement to develop a weather balloon
system which it could deploy anywhere around the world; send a remote signal for its launch
and receive live time measurements and telemetry to its ground station to conduct atmospheric
experiments. The key requirements for the system was that it had to be completely automated
with minimal or no human contact, robust, possess sufficient lifetime, re-usable and had the
ability to perform flawlessly in any given environmental condition. Portability and open software
for the launch system played an important aspect of design as future work on the project were
based on these issues apart from its commercial application.
Through this project, such problems are being answered and a solution to the above-presented
challenge for ASTRA group was provided with a working model.

16
9 BACKGROUND RESEARCH
9.1 HIGH ALTITUDE BALLOONS
Written by Rahul Kharbanda and Sullivan Pal
FIGURE 9-1: A TYPICAL WEATHER BALLOON
In practice, it is required to know the payload and parachute weight (i.e. the flight train), to
determine how much lifting gas is required to lift the payload. As the balloon ascends, it will see
a decrease in pressure, which will result in an increase in balloon volume. Assuming the perfect
gas equation holds for the system,
Where the mass of the lifting gas m is constant. Additionally, T reduces by about 30% compared
to sea level, and P within the balloon is almost equal ambient local pressure. Therefore, a change
in balloon volume can be expressed as:
This shows that the balloon’s volume drastically changes as it ascends. There is a point where the
balloon's material will not be able strain further, and the balloon will burst. The payload will then
descend to earth with the parachute deployed.
The amount of gas put in the balloon depends primarily on the mission requirements. For near
space photography, a high altitude is desirable. In this case, a minimal amount of gas is used to
delay the bursting altitude.
On the other hand, for meteorological sounding, a fast ascent rate is better since the radiosonde
is able to gather data almost instantaneously. The flight time for high ascent rates, low altitude
balloon is shorter. This makes recovery easier as they land nearer to the launch site.

17
“The ability to collect physical, chemical and biological observations across a wide range of
altitudes is essential for an understanding of atmospheric processes.” These are then used as
boundary conditions in numerical engineering models. Some of the applications where high
altitude atmospheric surveillance is used extensively include: [1]
 Weather forecasting: Large numbers of weather balloon soundings are integrated into
numerical weather prediction models.
 Pollution and aerosol monitoring: In order to sample ash clouds, density measurements
are used to formulate maps with contour lines around safe zones. In addition, fallout
monitoring after nuclear incidents require accurate atmospheric observations.
 Observation of extreme weather phenomena: A map of wind speeds, temperatures and
humidities across hailstorms, tornadoes and other extreme events can facilitate the
understanding of the physics behind them and lead to improved prediction models.
 Aeronautical engineering research: Investigating icing and contrail formation
phenomena is requires a thorough analysis of the associated atmospheric variables.

18
9.2 CUSTOMER REQUIREMENTS
Written by Rahul Kharbanda
The major stakeholders to consider when designing the remote weather balloon launcher are
educational institutions such as schools, universities, etc., hobbyists and weather balloon
enthusiasts and meteorological researchers. The following is a comprehensive list of the major
customer requirements and a brief description.
Low Weight- The box must be quite light in order to ensure that it is portable and can be carried
to remote places. The payload must also have a low weight in order for the balloon to increase
the lift and successfully ascend at a good rate
Compact size- The box must be small, easy to carry and fix at any location. The payload should
be contained in a minimalistic design
Ability to launch in windy conditions- The box should be able to withstand high surface winds and
the servo should be powerful enough to open the lid at these high speeds
Communication via GSM network- The box should be remotely launched from a location that can
transmit and receive GSM signals
Reusable- The box must be designed to be reusable, with easy access to the gas canister and
electrical components in case they require replacement
Waterproof- The box itself, the electrical components inside and the launch mechanism must be
made waterproof to avoid electrocution and component failure
Impact resistant- The box must be made of a material that is impact and corrosion resistant and is
capable of withstanding large stresses
Low cost- The box and the internal components should be cost-efficient and should minimise the
need for high maintenance costs to enable mass production
Standard gas supply connection- An international standard gas supply connection should be used
to allow for easy replacement of parts
Low power requirement- Due to the limited operation time for the batteries used to control the
payload, the power should be used efficiently and incorporate a backup source
Payload live telemetry- There should be a constant uplink/downlink of data signals emitted from
the radio transmitter to identify the coordinates of the payload after it is released at all times
Balloon size within launch regulations- Since the balloon expands at higher altitudes, it must be
ensured that even at the maximum burst diameter, it is legally permitted to be in the air without
requiring any authorised approval from a governing body.

19
Some of these customer requirements can be considered as sub-groups of one major design
feature. As such these requirements can be split into primary and secondary levels. This can be
then displayed diagrammatically in a tree diagram as follows.
FIGURE 9-2, TREE DIAGRAM DISPLAYING THE KEY CUSTOMER REQUIREMENTS FOR THE
LAUNCHER DESIGN
With so many requirements from different stakeholders for the launcher’s design a compromise
must be made between them to optimise the final design selection process. As a result the binary
weighting matrix is used to compare and rank the requirements against each other.

20
FIGURE 9-3: BINARY WEIGHTING MATRIX OF CUSTOMER REQUIREMENTS

21
FIGURE 9-4: NORMALISED IMPORTANCE OF THE CUSTOMER REQUIREMENTS
From the matrix and the graph above it can be seen that the key customer requirements are that
the balloon size should be within the permitted legal requirements at its maximum diameter, the
box must be compact, the box and payload must be light and most importantly, there must be
communication between the box and the mobile phone used to send a text message via GSM
network. However it must not be misconstrued that if a customer requirement is not as highly
rated as the other it becomes irrelevant when designing the launcher. The weighted importance
will be used to focus the priorities in the next stage of the design and help in making
compromises on conflicting design options such as weight and strong material properties.
0 1 2 3 4 5 6 7 8
Communication via GSM network
Low weight
Compact size
Balloon size within free-launch regulations
Waterproof
Can be stacked easily
Payload live telemetry
Ability to launch in rain
Withstands standing water <5mm
Can detect wind speed
High launching burst altitude
No hazardous systems (eg heat wire/blades)
Payload data backup
Low pressure system
High reliability
Impact resistant
Long endurance
Low power requirement
Easy to manufacture
Standard gas supply connection
Minimal setup
Copes with extreme hot/cold temperatures
Low cost
High speed of launch
Ability to launch in windy conditions
Reusable
Usable on varying surface gradient
Aesthetics
Normalised Importance (%)
Customerrequirements

22
9.3 LEGAL ISSUES
Written by Rahul Kharbanda
In the UK, the Civil Aviation Authority (CAA) grants weather balloon flight permission. In
order to get permission for the first launch, an application form was filled and sent to the CAA at
least well in advance (has to be at least 28 days prior to launch). The CAA then issues a
NOTAM to air traffic and granted us permission to launch the balloon. The regulations required
the use of a standard meteorological balloon and for the payload to descend by parachute.
According to Air Navigation Order (CAP393), the maximum weight of the payload must not
exceed 4kg and permission is required by the CAA of the balloon is being flown at more than 60
metres above ground level. [2]
However, section 253(a) of CAP393 lists the exceptions from application of the Order, which
applies to any ‘small balloon’ that is defined to be a balloon that is of no more than two metres in
any linear dimension.
As a result, to eliminate the launch-location requirements (except near airports) for the target
stakeholders, the balloon was carefully selected to have a maximum burst diameter of 1.6m and
the payload and the attached string combined were designed to be less than 0.2m, thereby
classifying it as a small balloon.
The possibility of obtaining insurance against injury or damage caused by the payload landing
was investigated so as to reduce any risk of liability. However, the insurance was very costly and
upon advice from the members of the UK High Altitude Society (UKHAS), adequate flight
planning using the ASTRA flight predictor and the UKHAS predictor was undertaken to ensure
a landing would not take place near main roads, airports or built-up areas.

23
10 HELIUM GAS SUPPLY
This chapter details how the gas supply was chosen. The requirements for the gas cylinder are
explained and the amount of helium required for a successful high altitude launch is calculated.
A suitable in-box helium cylinder is chosen taking into account the size of the box and the
amount of helium required for a successful high altitude launch.
10.1 INTRODUCTION
The burst altitude of the balloon, and mass allowable for the payload is determined by the
amount of helium gas used to inflate the balloon at ground level, the size of the balloon as
determined by the burst diameter, and the weight of the balloon.
The Pawan 100g balloon is chosen due to its burst diameter of 1.6m. The key reason for this
decision is that the balloon will never exceed a 2m length in any direction and allow 40cm for
the payload and parachute combined. As shown in the legal section above this allows us to
conduct unregulated launches thereby drastically improving the flexibility of use for the device.
The calculations for the volume of gas required at ground level can be based on the specifications
of this balloon. Sufficient helium must be supplied so that the balloon can generate enough lift to
take a payload of reasonable mass up to a high altitude greater than 10km using the Pawan 100g
balloon.
Furthermore the gas canister must be adequately small to that the original design requirement for
the box to be of a ‘shoebox’ size and light enough for a human to carry. It should also be easy to
source and not too expensive.
10.2 CALCULATIONS
The volume of helium required at ground level is calculated, thereby determining the maximum
allowable payload mass, ascent rate, and burst altitude.
Assumptions:
All assumed atmospheric parameter values are taken from engineeringtoolbox.com using the
international standard atmosphere. These values are shown in appendix B. All other constants
are taken from the same website [3]. It is assumed that the balloon is launched from sea level. Air
density is calculated using the ideal gas law. The balloon is assumed to be rigid and perfectly
spherical and to have a Cd of 0.25 (the value stated on the website of the balloon supplier).

24
10.2.1 CALCULATION OF LIFT AND GAS VOLUME REQUIRED AT GROUND
To calculate volume of helium at given temperature and pressure:
(Equation 1)
To calculate diameter of balloon from given volume of helium:
( ( ) )
(Equation 2)
To calculate balloon neck lift from the local air density, volume of helium and mass of balloon of
0.1kg:
(Equation 3)
To determine the maximum allowable payload weight for a given ascent rate at the ground
assuming steady state conditions the drag force and lift force are equated:
10.3 PLOTS
Using equations 1 and 2, and ISA values for how temperature and pressure vary with altitude the
plot shown in Figure 10-1 can be obtained. The plot shown in Figure 10-2 is obtained using
equation 3.

25
FIGURE 10-1, PLOT OF HOW THE DIAMETER OF THE BALLOON VARIES WITH ALTITUDE
FOR VARIOUS MASSES OF HELIUM
FIGURE 10-2, PLOT SHOWING HOW ASCENT RATE AT GROUND VARIES WITH PAYLOAD
WEIGHT FOR VARIOUS MASSES OF HELIUM.

26
Figure 10-1 shows that increasing the mass of helium used for inflation reduces the burst altitude.
Halving the mass of helium from 0.04kg to 0.02kg roughly increases the burst altitude by 2km.
Figure 10-2 shows that either increasing the mass of helium, or accepting a reduced ascent rate at
ground level can increase the allowable payload weight.
10.4 GAS CANISTER CHOICE
The gas canister required is determined by the mass of helium required. As shown above, this is
a function of ascent rate, payload mass and burst altitude. These values are determined as
follows:
 Steve Randall, member of the UKHAS, recommends that the ascent rate should be at
least 4m/s.
 A preliminary plan of the payload design demonstrated that it is possible for the mass to
be below 67g.
 The design objective was for the burst altitude to be at least 10km.
A 2.2L helium canister manufactured by Maxxiline, as shown in Figure 10-3, was selected for
use since it fit our constraints. The helium is pressurised to 100 bar, therefore the mass of helium
it supplies is 0.03676 kg. With this amount of helium equation 3 shows that an ascent rate of
2.91m/s is achieved at the ground with a payload of 0.067kg. Using equations 1 and 2, and
considering table 1, this mass of helium causes the balloon to expand to the burst diameter of
1.6m at approximately 17.5km.

27
TABLE 10-1, TABLE SHOWING VARIATION OF BALLOON DIAMETER WITH ALTITUDE AND
IMPORTANT ATMOSPHERIC PARAMETERS AS ASSUMED FOR ISA CONDITIONS FOR WHEN
THE MASS OF HELIUM IS 0.03767 KG (THE MASS CONTAINED IN THE MAXXILINE 2.2L
100BAR CANISTER)
Altitude
(km)
Temperature
(K)
Pressure
(Pa)
Volume of
Helium
Balloon
diameter (m)
0 288.15 101325 0.217123119 0.745708595
1 281.5 89874.56043 0.239136394 0.77010336
2 275 79495.19744 0.264116698 0.796035456
3 268.5 70108.52041 0.292400119 0.823492486
4 262 61640.20644 0.324519848 0.852604356
5 255.5 54019.88001 0.361111522 0.883515639
6 249 47180.99341 0.402936234 0.916387663
7 242.5 41060.70795 0.450909481 0.951400952
8 236 35599.77593 0.506137826 0.988758108
9 229.5 30742.42336 0.569965604 1.028687214
10 223 26436.23351 0.644034809 1.071445881
11 223 22632.03131 0.752290166 1.128397861
12 223 19283.76849 0.88291117 1.190253488
13 223 16348.40969 1.041437969 1.257606949
14 223 13785.81936 1.235026671 1.331146654
15 223 11558.64956 1.472996867 1.411673949
16 223 9632.22873 1.76759243 1.500126341
17 223 7974.451311 2.135050291 1.597606537
18 223 6555.668394 2.597119556 1.705419038
FIGURE 10-3, MAXXILINE 2.2L GAS CANISTER [4]
Whilst this is below the recommended value, the balloon will still ascend at a reasonable speed.
In order to obtain an ideal ascent rate of 4 m/s, a possible configuration would be to use a

28
payload of mass 30g as demonstrated by Matt Brezja [5], and a helium gas canister of same size
pressurised to 133 Bar.
In a scenario where the gas canister does not contain as much helium as stated due to a
manufacturing defect, or a failure occurs that leads to a reduction in the amount of gas available,
the tolerance of the setup allows the balloon to lift even with a 25% reduction in the mass of
helium available albeit at a slow ascent rate of 1.47 m/s.
The Maxxiline cylinder is of a suitable size (Diameter: 102 cm, Length: 325 cm), and weight
(280g). Furthermore the price is reasonable and the canisters are disposable. The choice of
canister informs the design requirements for the size of the box. It will be built so that the
canister fits snugly in order to keep the size as low as possible.
10.5 GAS CANISTER TEST
An indoor test was undertaken to verify the theory that inflating the balloon with the Maxxiline
cylinder would provide adequate lift. It was impossible to accurately measure the ascent rate
however the balloon behaved as expected by easily lifting a 67g weight. The time required for the
balloon to inflate was determined from this test to be 4.5 minutes. This would inform the design
of the on-board box Arduino code. Furthermore the method of sealing the balloon was
demonstrated to be adequate (at least at ground level) as the balloon remained inflated for over
24 hours.

29
11 GAS DELIVERY AND BALLOON RELEASE SYSTEM
11.1 INTRODUCTION
This chapter details the design and manufacture of the gas delivery system and balloon release
mechanism. These are key components in the overall system as they allow the balloon to inflate
with gas and then be released. Conceptual designs for the gas release mechanism are discussed
and the final design is explained. The plan for the brass release mechanism that was constructed
at the EDMC is also presented. The type of piping used is described and a schematic of the entire
system shows how it works. Conceptual designs of the balloon release mechanism are briefly
described and the chosen method involves the use of a hotwire to melt the tape connection. The
use of a one-way valve on the balloon as a method of avoiding the difficulty of remotely sealing
the balloon is shown. Electronic calculations demonstrate that the onboard power supply should
be able to supply enough current to adequately heat the hotwire.
11.2 GAS RELEASE DEVICE
11.2.1 REQUIREMENTS
The gas release device must have the ability to release gas upon receiving a signal from the on-
board Arduino. The gas supply must be engaged, but does not need to be disengaged since the
helium canister must be effectively emptied into the balloon.
The device must have the ability to use an on-board power supply that would be of a reasonable
portable voltage, i.e less than 18V.
Furthermore the system must be able to cope with a high-pressure source since the gas canister is
rated at 100 bar.
11.2.2 CONCEPTS
A feasibility study narrowed down the possible options for this system to two choices:
1. A solenoid valve (a valve actuated by electrical signals) could be used to control the gas
supply. The gas cylinder could be left permanently open and the solenoid valve could
control either the unregulated gas supply, or the regulated gas supply if a regulator was
attached to the cylinder.
2. A method of opening the gas cylinder directly could be developed. Gas is released from
the Maxxiline cylinders by twisting a supplied female threaded part (hereby referred to as
the release mechanism) onto the top that contains a pin. Eventually this pin pushes the

30
valve releasing the helium. The flow rate of the helium can be controlled to a degree by
the amount of twist applied. A servo could be connected to the release mechanism in
order to twist it and release the helium into the piping. A custom-built brass release
mechanism with a British Standard Pressure (BSP) threaded output could be
manufactured for easy and secure connection to the gas piping.
The positives and negatives of these options are summarized in the table below:
TABLE 11-1, SUMMARY OF PROS AND CONS FOR DIFFERENT GAS RELEASE SYSTEMS
ADVANTAGES DISADVANTAGES
1. SOLENOID
VALVE
Concept proven to work in
many other applications
Easy to plan and implement
Self-contained unit will not
interfere with other systems.
Connections would be secure
as most solenoid valves are
supplied with standard BSP
inputs and outputs.
Operating pressures must be
adhered to so use on unregulated
gas supply is difficult.
Regulator is expensive and difficult
to source for the Maxxiline canister
that stores the helium. Furthermore
the regulator would add weight and
size.
High-pressure solenoid valves are
expensive and typically need >12V
DC supply.
Without a regulator it would be
dangerous to leave the cylinder
open and have high pressures (100
bar) existing in gas piping.
2. GAS
CYLINDER
DIRECT
RELEASE
MECHANISM
A servo only requires 6V.
Weight and size is kept low
Cheaper
No issues with high pressure
being stored in piping.
Possibility to link to same
servo that opens lid, reducing
possible failure modes.
Manufacturing part provides
design flexibility in BSP output
Difficult to design and implement
Custom-built brass release
mechanism could be difficult to
manufacture to such a high
tolerance that there is no leakage.
Any modifications or re-builds
would be very costly.
Servo required to rotate the release
mechanism, the torque required is
difficult to determine.

31
to allow for simple pipe
construction.
The gas cylinder direct release mechanism was chosen due to the key advantage of the possibility
of sharing a servo with the lid opening mechanism since a servo would be required in any case
for that purpose. Sharing the servo reduces the possible failure modes and the in-box power
supply required. A simple rotation of the servo can open the lid and engage the gas
simultaneously. Another key advantage was that the device could be manufactured so that it
would be possible to attach via screws an off-the-shelf sprocket that would allow a chain link to
the servo.
As shown in Figure 11-1 the sprocket also has holes allowing arms to be screwed on as an
alternative method of opening the canister if the torque required was so great that a larger
moment arm was required. The design is kept flexible in order that it can be adapted to the
torque requirements of the gas cylinder and release mechanism when manufactured.
FIGURE 11-1, SPROCKET TO BE ATTACHED TO RELEASE MECHANISM [6]
11.2.3 DESIGN AND MANUFACTURE OF THE GAS RELEASE MECHANISM
The release mechanism is manufactured from brass on advice of personnel at the University of
Southampton Engineering and Design Manufacture Centre (EDMC) who were contracted to
manufacture the part. Brass provides the required strength whilst being suitable to work with for
this type of part.
The following instructions and design plan (Figure 11-2) were provided to the EDMC:
Manufacture of brass inflator to BSP standards. The inlet is 1/8" BSP parallel female, and the outlet is
1/4" BSP parallel male. The inlet includes a pin inside that will push open a valve inside a disposable gas
canister when rotated. A plastic equivalent (without the correct outlet) will be provided to aid the
construction process. The BSP standard threads are not modelled on the plans. A holed plate on the inflator
will allow a sprocket to be screwed in place, as the inflator will eventually be turned via a chain. 8 holes are
on the plate, but 4 would be sufficient if there is a great reduction in cost. In order for the sprocket to fit, the

32
outer diameter of the inlet must be no greater than 12mm (sprocket hole diameter is 1/2").
FIGURE 11-2, ENGINEERING DRAWINGS OF GAS RELEASE MECHANISM SUPPLIED TO THE
EDMC ACCOMPANYING MANUFACTURE INSTRUCTIONS.
FIGURE 11-4, RELEASE MECHANISM CONNECTED TO
PIPING WITH SPROCKET SCREWED ON.
FIGURE 11-3,
SOLIDWORKS MODEL OF
RELEASE MECHANISM
(NOT SHOWING THREADS)

33
The release mechanism was extensively tested with the gas canisters to ensure that gas could be
released without leakage by screwing the part onto the cylinder. In order to reduce costs a flat-
ended screw was used as the pin. The thread was made slightly deeper to allow a more smooth
rotation, and a small rubber o-ring was placed around the bottom of the canister male thread in
order that an airtight seal would be created before the part was turned tight enough such that the
pin impinged on the cylinder valve and allowed the gas to flow (see Figure 11-5).
It was determined during testing that the mechanism would need to be rotated by approximately
40 degrees from a point where there was zero gas flow rate to where the gas mass flow rate was
reasonable (approximately 0.000136kg/s since it took 270 seconds for the canister to empty and
fully inflate the balloon).
FIGURE 11-5, RELEASE MECHANISM SCREWED TO GAS CYLINDER WITH RUBBER O-RING

34
11.3 BALLOON INTERFACE
A small and light one-way valve (shown in Figure 11-6) that is designed to be placed in the necks
of standard helium party balloons for easy filling is used to avoid the difficulty of remotely tying
the balloon. The valve has an input of 10mm in diameter allowing a 10mm nylon air hose to fit
snugly inside. Using duct tape an airtight seal is created between the valve and the pipe allowing
for inflation to take place.
The head of the one-way valve is 2.5cm in diameter, and the neck of the balloon has a 5 cm
diameter. In order to create an airtight seal between the neck and the valve an ‘adaptor’ is
manufactured by sawing the bottom half and the top off a popular branded fruit juice bottle. The
valve is glued in place at the neck of the adaptor, and the balloon neck can stretch over the wider
end of the adaptor and taped in place. Figure 11-7 shows these constructions. The suitability of
these connections was tested by filling helium balloons up to launch diameter. The tests were
successful and the balloons remained inflated for over 24 hours.
FIGURE 11-6, ONE-WAY VALVE

35
FIGURE 11-7, LEFT: VALVE GLUED TO ADAPTOR, RIGHT: VALVE GLUED TO ADAPTOR
WITH BALLOON TAPED ON.
11.4 GAS PIPING
Nylon air hose is used for the helium transport. It is available in a variety of outside diameters,
operates from -40°C to +80°C, and has a maximum operating pressure of 22 bar [7]. It is
compatible with push-in fittings that allow the design of the system to be flexible. Right angle
elbow fittings are used to direct the flow, and a rotary elbow fitting allows the release mechanism
to be twisted whilst maintaining an airtight seal. A fitting that takes ¼ inch BSP male thread and
outputs a female push-in interface for a 8mm nylon air hose is placed between the release
mechanism and the rotary fitting. For connection onto the one-way valve shown above, an
adaptor is used that takes 8mm nylon air hose and provides an interface for a 10mm nylon air
hose. To reduce the risk of leakage from the push-in fittings superglue is placed in the gap
between the pipes and edges of fittings. See Figure 11-12 for a schematic of the full system and
Figure 11-13 for a photograph of the resultant system.
11.5 BALLOON RELEASE MECHANISM
The use of a one-way valve attached to the balloon allows the release to occur by simply cutting
through gas piping system. Figure 11-8, Figure 11-9, and Figure 11-10 show sketches of concepts
for this system.

36
FIGURE 11-8, DIAGRAM OF CONCEPT HOTWIRE METHOD FOR RELEASING BALLOON
WHEN INFLATED
FIGURE 11-9, DIAGRAM OF CONCEPT CUTTING METHOD FOR RELEASING BALLOON
WHEN INFLATED

37
FIGURE 11-10, DIAGRAM OF CONCEPT FUSE-WIRE METHOD FOR RELEASING BALLOON
WHEN INFLATED
During testing it was determined that the use of a latex tube or skirt is not suitable since it
behaves like a small balloon and inflates to a large extent before the activation pressure of the
one-way valve is reached. The cutting method was deemed too dangerous, and doubts were cast
over the possibility of obtaining 5 amps from the power supply. The chosen option was to use the
hotwire method, and use standard packing tape instead of latex (see Figure 11-12 for a diagram
of the final design).
32 SWG (Standard Wire Gauge) nichrome wire has a resistance of 17 Ohms/metre [8].
Therefore a 10cm long strip has a total resistance of 1.7 Ohms. With the on-board 6V power
supply, assuming a negligible battery resistance, the current across the length of the wire is will
theoretically be 3.53 amps (from Ohm’s law). According to the table shown in appendix A, this
current would heat the wire up-to approximately 1000 degrees centigrade when held straight in
clear air. This is far in excess of the melting point of the polypropylene tape which is
approximately 160 degrees centigrade [9].
Extensive testing was conducted on a hotwire release mechanism and many different methods of
attaching the hotwire were attempted. The final design shown in Figure 11-11 is very secure,
easy to manufacture and has never failed after 10 tests. An important addition is the use of an o-
ring (cut from another one-way valve) placed under the hotwire, whilst the edge of the one-way

38
valve is above it. This allows the hotwire to be set in place very tightly thereby allowing uniform
contact with the tape. Without this addition the hotwire was very unreliable as non-uniform
heating of the tape often occurred.
FIGURE 11-11, LEFT: PHOTO OF FINAL HOTWIRE DESIGN ATTACHED TO TAPE AROUND
PIPING. VISIBLE IS THE OUTLINE OF THE ONE-WAY VALVE ABOVE, AND BELOW THE
OUTLINE OF THE O-RING, RIGHT: PHOTO OF HOTWIRE DESIGN ONLY.

39
FIGURE 11-12, FINAL DESIGN SCHEMATIC SHOWING GAS SYSTEM AND BALLOON
RELEASE MECHANISM.
FIGURE 11-13, PHOTO OF THE GAS SYSTEM EXCLUDING THE CANISTER AND RELEASE
MECHANISM.

40
12 LAUNCH BOX DESIGN
Written by Sabin Kuncheria Purackal
In this chapter, the detailed design procedures for the launch box and opening mechanism are
discussed. This includes initial considered box concept designs, positives and negatives of each
design, opening mechanism designs, required torque calculations and the materials considered
for the box. The detailed methodology for the launch box construction is also discussed in this
chapter.
12.1 DESIGN REQUIREMENTS
The following were the key customer design requirements for the launch box:
 Impact resistant
 Weatherproof
 Portable (<5 kg)
 Waterproof
 Reusable
 Ease of manufacture
 Cost efficient
12.2 CONCEPT DESIGNS
FIGURE 12-1, BOX CONCEPT DESIGN 1

41
Three concept designs for the launch box were developed in order to match the customer
requirements. The advantages and disadvantages of the designs are tabulated in Table
12-1.Figure 12-1 shows the first box concept design. The gas canister and the electrical systems
are kept beneath the box. This will allow these systems to be effectively sealed for water
proofing. The payload, the balloon, and the balloon release mechanism are placed on the top
shelf of the box. In this design, the box lid is opened as the balloon inflates. The gas and
electrical systems underneath the box can be accessed via side door. The advantage of this design
is that it is portable, easy to manufacture, cost efficient for mass production and has no electrical
systems that can fail. The primary disadvantage of this design is the failure associated with the
box lid opening. As mentioned above, the box lid is opened as the balloon is inflated. This
inflating balloon can successfully open the box at steady and low (<2 m/s) wind speed. But at
wind speed higher than 2 m/s, the balloon will be blown out of the box as it inflates and the
payload, attached to the balloon neck, will get jammed inside the box between the box and box
lid preventing the balloon from launching. This was confirmed during the wind tunnel tests
conducted.
Figure 12-2 shows box design concept-2. In this design, the box is divided into two halves, first
half for the Arduino and electrical system and other half for the balloon, release mechanism and
pipes. The box lid is opened using a servo, which is controlled and operated by the Arduino. The
primary advantage of this design is that box lid is opened by a powerful servo. This enables the
lid to be opened high wind speeds and is able to move any heavy object (e.g., a bird, a branch) on
top of the box lid during launch time. Powerful servos are expensive, so this design will be cost
inefficient. This design is also susceptible to electrical, servo and Arduino failure.

42
Figure 12-3 shows the third concept design. The interior design of the box is similar to concept
design 1, but different in the opening mechanism. The box lid is a sliding door, which slides in a
gear rail with the help of gear that is connected to a servo. The main advantage of this design is
that the servo requires little torque, compared to concept design 2, to open at high wind speeds.
This design is also susceptible to servo, electrical and Arduino failure. In addition to above
failures, foreign objects can also obstruct the gear rail preventing the lid to open during launch.
TABLE 12-1, ADVANTAGES AND DISADVANTAGES OF EACH CONCEPT DESIGN
Box design Opening
mechanism type
Advantages Disadvantages
Concept design 1 Box lid opens as
balloon inflates
 Portable
 Excellent
manufacturability
 Waterproof
 Cost efficient
 Fails at wind speed
greater than 2 m/s
Concept design 2 Box lid opened by
a servo
 Portable
 Water proof
 Lid can open at
high wind speeds
 The Servo can
open heavy object
on top of box
 High torque servo
required
 Susceptible to failure
due to electrical
systems, Arduino,
servo
 Poor
manufacturability
Concept design 3 Sliding box lid
operated by servo
 Portable
 Water proof
 Low torque servo
required
 Can open at high
wind speeds
 Susceptible to failure
due foreign object
obstructing the gear
rail
 Also vulnerable to
electrical and
Arduino failure
 Poor
manufacturability

43
12.3 MATERIAL SELECTION
According the customer requirement, the box should be ideally impact resistance and have good
weatherability properties. Four materials were initially considered for the box material. The
advantages and disadvantages of the considered materials are tabulated in Table 12-2.
TABLE 12-2, ADVANTAGES AND DISADVANTAGES OF THE CONSIDERED BOX MATERIAL [
[10]][ [11]][ [12]] [13]
Material Advantages Disadvantages
Acrylic  Good weatherability
 Transparent
 Good insulation
properties
 UV resistant
 Abrasion resistance
 Low water absorption
 Poor impact resistant
 Expensive to
manufacture
Polyethylene foam sheet  Cheap/Easy to
manufacture
 Good insulation
properties
 Light weight
 Poor impact resistance
 High water absorption
Metal sheet  Cheap/Easy to
manufacture
 Durable
 Impact resistant
 Poor weatherability
 Poor resistance to
corrosion
 Electrical insulation
needed
Biplex fluted polypropylene
sheet
 Crack resistant
 Flexible
 Easy/cheap to
manufacture
 Environmentally
sound
 Good corrosion
resistance
 Flammable
 Difficult to bond
 Low temperature impact
poor
 Poor oxidative
resistance in presence of
metals like copper

44
12.4 OPENING MECHANISM
FIGURE 12-4, CONCEPT DESIGN FOR THE OPENING MECHANISM (SIDE VIEW)
Figure 12-4 shows the initial concept design that was considered as the opening mechanism. This
design consist of three sprockets, connected to each other by a plastic chain. Sprocket-1 is
attached to the box lid, sprocket-2 attached to the canister and sprocket-3 attached to the servo.
As the servo is activated, the sprocket 3 which is attached to the servo rotates sprocket 1 and 2.
This will simultaneously open the canister and the box lid. The canister will be only activated
when the box lid is fully opened.The major disadvantage of this design is that sprocket-1 or chain
can easily get caught to an external object, whilst being displaced, result in damaging the box
and the internal systems.
The above design was slightly modified by introducing a curved lever with a sliding mechanism
instead of sprocket-1. The primary advantage of the curved beam is that, it ensures complete
opening of the box lid without any parts of the chain or servo sticking out of the box. Also with
just 90° servo rotation, the curved beam allows 220° rotation of the box lid. The lever is attached
to the servo which is operated by an Arduino. The tip of the lever is attached to a sliding
mechanism, which is attached to the lid of the box Figure 12-5. Two concept design for the
curved lever were considered (Figure 12-6 ). The box wasn’t able to fully open when concept
design-1 was used. So concept design-2 was used which ensure the box lid to open 220°.

45
FIGURE 12-5, 3D DIAGRAM OF THE BOX OPENING MECHANISM
FIGURE 12-6, CURVED LEVER USED FOR THE BOX OPENING MECHANISM (DIMENSIONS IN
MM).CONCEPT DESIGN 1.
FIGURE 12-7. CONCEPT DESIGN 2 FOR THE BOX OPENING MECHANISM
12.4.1 SERVO AND TORQUE CALCULATIONS
The torque for the required servo was calculated using simple moment and force analysis. The
wind force acting on the box lid was calculated using the following equations.
where ρ is the density of air, V is the wind velocity and Cd is the coefficient of the drag of the box
lid. The box lid is assumed to be a flat plate. Therefore, the Cd is assumed to be 1.17 (flat plate

46
perpendicular to the flow direction) and the wind pressure shape factor as 1 [13]. The calculated
wind pressure and force experienced by the box at different wind speeds are tabulated in Table
12-3.
TABLE 12-3, CALCULATED VALUES OF WIND PRESSURE AND FORCE AT DIFFERENT WIND
SPEEDS
Velocity
(m/s)
Wind pressure (Pa) Force (N)
1 0.6125 0.0860
2 2.4500 0.3440
3 5.5125 0.7740
4 9.8000 1.3759
5 15.3125 2.1499
6 22.0500 3.0958
7 30.0125 4.2138
8 39.2000 5.5037
9 49.6125 6.9656
10 61.2500 8.5995
Figure 12-8 shows the free body diagram of the forces acting on the box lid, where d is the
moment arm and is the angle made the box lid with horizontal when opened. As the box lid
opens, both and the weight of the box lid (W) change as a function of . That is,
and . The calculated values of the required torque are tabulated in Table
12-4.
FIGURE 12-8, FREE BODY DIAGRAM OF THE BOX LID WITH THE FORCES

47
TABLE 12-4, CALCULATED VALUES OF THE REQUIRED TORQUE AT DIFFERENT BOX LID
OPENING ANGLES
Angle
(degree)
Weight of box lid (N) Moment
due to lid
(Nm)
Max
torque
required
(Nm)
Torque in
kg.cm
0 2.9430 0.6769 1.5368 15.6671
10 2.8983 0.6565 1.5164 15.4582
20 2.7657 0.5978 1.4577 14.8597
30 2.5491 0.5078 1.3678 13.9426
40 2.2551 0.3975 1.2574 12.8175
50 1.8927 0.2800 1.1399 11.6200
60 1.4729 0.1695 1.0295 10.4942
70 1.0083 0.0795 0.9394 9.5760
80 0.5131 0.0206 0.8805 8.9758
90 0.0016 0.0000 0.8600 8.7661
So the maximum torque required to open the box lid is 16 kg.cm. The servo is also connected to
a 50 mm diameter sprocket, which is attached to the canister, via a chain. The force required to
turn this sprocket should also be added to the above torque. The measured torque required to
rotate the canister sprocket is approximately 1-2 kg.cm, but as approaching the gas release point
more torque is required. An approximate of 18-20 kg.cm servo torque will be required.

48
12.5 FINAL BOX DESIGN
A box concept similar to concept design-3 with a servo for the opening mechanism was decided
as the final box design. In this design, all the electrical system, piping system and the canister are
placed in the bottom shelf and the balloon, balloon release mechanism and the payload are
placed at the top shelf. The top shelf is constructed in a way that it can be easily removed. This
allows better accessibility to the piping system, electrical system and the canister for purposes like
canister and battery replacement. The payload is also placed in a separate compartment in the
top shelf in order to prevent any payload movement inside the box when the box is displaced.
It was decided to use biplex fluted polypropylene sheet as the box material (Figure 12-9). The
materials exhibits excellent impact, corrosion and crack resistance. It is environmentally sound
with good weatherability properties. The primary advantage of this material is that it is
lightweight, can be easily cut, can be attached to the required box size and cost efficient to
manufacture without using the EDMC.
FIGURE 12-9, IMAGE OF BIPLEX FLUTED POLYPROPYLENE SHEET [1]
The calculated torque required to open the box lid and to activate the canister is approximately
20 kg.cm. So we decided to use a servo with 24 kg.cm torque as it meets our calculated
requirements. The servo is attached to the side of the box, and to the servo a 100 mm diameter
sprocket with a chain attached. The chain is linked to a small 50 mm diameter sprocket, attached
to the canister, and as the servo rotates the chains activate the canister. To the 100 mm diameter
sprocket, the curved lever (lever concept design-2) is attached which opens the box lid. The tip
of the leaver is attached to a sliding mechanism which enables efficient opening and closing of
the box lid. Initially, several materails were considered for the manufacture of the curved lever
and sliding mechanism, but due to budget constrains, recycled polypropelene was used for both
of them. One of the main advantages of using biplex fluted polypropylene sheet is that can be
easily cut and can be attached into the required box size. Initially, we considered using the
EDMC for the box manufacture, but due the budget constraint and high manufacture cost
charged by the EDMC, we decided to construct the box manually as it was the most cost

49
efficient method to serve the value of money. Cutting method like laser cutting was also
considered. Precision cutting method like laser cutting can be employed to cut/drill complex
shape or patterns. Due to high operating cost of the laser cutter and due to simple required box
design shape, the means of using laser cutter was neglected. The jigsaw machine was used to cut
the biplex fluted polypropylene sheets.
Two boxes were constructed. One box was used to understand the behaviour and simulate the
effects of the force acting on the box. This was achieved by the wind tunnel test (Figure 12-13,
construction box for the wind tunnel experiment). The second box was used as the final launch
box. The edges of the box were firmly attached using gusseted plastic angle brackets (Figure
12-11, image of gusseted plastic angle bracket). The bracket permits additional rigidity allowing
the box to resist external impact and deformations. The brackets were firmly attached to the box
using nuts and bolts.
Due the torque of the rotating sprocket attached to the canister, the canister can rotate about its
centre of axis. To prevent the induced rotation due to the torque, the canister was firmly kept in
place using a pair of Velcro straps. The Velcro straps are attached to the plastic corner brackets.
The servo is also firmly attached to side of the box using nuts and bolts. The nut and bolt allows
firm attachment of the servo to the sidewall and helps to prevent any induced rotation due to the
torque. The combination of canister attachment using Velcro strap and the firm attachment of
servo to the sidewall will not only prevent any induced rotation due to torque but also prevent
any resulting chain slipping. The images of the constructed box are given in Figure 12-14, images
of the constructed launch box.
Figure 12-11 shows the 3-D model of the final box design. Box dimensions are 460x250x220
mm.

50
FIGURE 12-10, FINAL BOX DESIGN
12.6 CONSTRUCTION
One of the main advantages of using biplex fluted polypropylene sheet is that can be easily cut
and can be attached into the required box size. Initially, we considered using the EDMC for the
box manufacture, but due the budget constraint and high manufacture cost charged by the
EDMC, we decided to construct the box manually as it was the most cost efficient method to
serve the value of money. Cutting method like laser cutting was also considered. Precision
cutting method like laser cutting can be employed to cut/drill complex shape or patterns. Due to
high operating cost of the laser cutter and due to simple required box design shape, the means of
using laser cutter was neglected. The jigsaw machine was used to cut the biplex fluted
polypropylene sheets.
Two boxes were constructed. One box was used to understand the behaviour and simulate the
effects of the force acting on the box. This was achieved by the wind tunnel test (Figure 12-13,
construction box for the wind tunnel experiment). The second box was used as the final launch
box. The edges of the box were firmly attached using gusseted plastic angle brackets (Figure
12-11, image of gusseted plastic angle bracket). The bracket permits additional rigidity allowing
the box to resist external impact and deformations. The brackets were firmly attached to the box
using nuts and bolts.

51
Due the torque of the rotating sprocket attached to the canister, the canister can rotate about its
centre of axis. To prevent the induced rotation due to the torque, the canister was firmly kept in
place using a pair of Velcro straps. The Velcro straps are attached to the plastic corner brackets.
The servo is also firmly attached to side of the box using nuts and bolts. The nut and bolt allows
firm attachment of the servo to the sidewall and helps to prevent any induced rotation due to the
torque. The combination of canister attachment using Velcro strap and the firm attachment of
servo to the sidewall will not only prevent any induced rotation due to torque but also prevent
any resulting chain slipping. The images of the constructed box are given in Figure 12-14, images
of the constructed launch box.
FIGURE 12-11, IMAGE OF GUSSETED PLASTIC ANGLE BRACKET [ [14]] [15]]
The jigsaw machine was also employed to manufacture the curved lever. The jigsaw machine
was used to cut the recycled polypropylene (PP) sheet and then a double cut flat file was used to
file the work piece to achieve the curved lever shape (figure 11). Precision cutter like the laser
cutters can be also used to cut the curved lever, but initially the obtained recycled material for
lever was thought to be polyvinylchloride (PVC). Cutting PVC using a laser cutter produces
chloride fumes, which are a potential hazard when inhaled. So the method of using laser cutter
was opted out. Also, it was not cost efficient to use laser cutter due to the budget constraints.
Later, the recycled material was confirmed as PP by the University’s nCATs department.
FIGURE 12-12, CONSTRUCTED CURVED LEVER

52
FIGURE 12-13, CONSTRUCTION BOX FOR THE WIND TUNNEL EXPERIMENT
FIGURE 12-14, IMAGES OF THE CONSTRUCTED LAUNCH BOX

53
13 BOX ELECTRONICS
Written by Sullivan Pal
The box electronics is a core subsystem to the box that controls all aspects of the launch, and it is
therefore critical to come up with a robust design. This chapter deals with the concept generated
for power management, justifies the uses of some subcomponents, and mentions the methods of
calibrating the subcomponents such that all work in synchrony.
13.1 PRESENTATION
The box electronic system needs to be carefully designed as the failure of such a part can lead to
the failure of the entire balloon launcher. It controls the entire aspect of the launch phase.
Its mission is to check for any messages prompting for the launch of the balloon, judging the
current wind speed, to then decide whether to open the lid which shelters the components from
the weather, fill and release the balloon with the aid of a Nichrome hotwire.
FIGURE 13-1: LAUNCH PROCESS
Care must be taken to design the electronics subsystems such that it is:

54
-Robust: tolerance to programming bugs, resilience to physical shocks and environmental effects.
-Economical: Power wise, this would enable the remote balloon launcher to stay on standby for
a much longer time, or alternatively, require less batteries to operate, in turn reducing the overall
weight of the payload which is an extra benefit.
-Smart: Must be able to estimate surface winds and determine if it is possible to launch a balloon
within those conditions.
The most challenging aspect of the electronic subsystem is to save the most power possible.
FIGURE 13-2: A PHOTO OF THE BARE ELECTRONICS USED IN THE TEST
The core hardware chosen for the electronics subsystem are the Arduino Uno, and the Arduino
GSM shield, manufactured by Telefonica. These can be found for less than £100 combined. Any
pay as you go SIM card (purchased separately) is required to operate the GSM shield.
Two sensors are used in conjunction to estimate the absolute surface wind speed: the hot wire
anemometer and a temperature sensor for compensation. Wind tunnel experiments have been
conducted to calibrate the wind sensor.

55
A simple breadboard has been used for this project, but ideally, a printed circuit would offer
much more resistance to wires coming off out of the breadboard (it happened a few times while
setting up the hotwire). The entire circuit fits on a single breadboard, and therefore it was
possible to fit most of the critical electronics in Tupperware. Simple holes would need to be
drilled in to let the wires pass through, and some sealant/superglue at the Tupperware’s interface
would stop the wires from coming off the breadboard so easily.
13.2 STRATEGY
In order to save power, either putting the Arduino in a lower power state or shutting the Arduino
off completely is required for endurance. At regular intervals, the Arduino should wake itself up
and check for messages. For the first option, a NE555 timer circuit would send pulses at regular
intervals to the Arduino to make it wake up. The advantage of such method would be the cost,
and the low power consumption of such circuit.
In the configuration used (Arduino + GSM shield), is not possible to use a sleep interrupt code,
as the GSM shield already uses the pins necessary for it.
A much simpler solution to implement, albeit more costly would be to use a low power Arduino
Pro-Mini (3.3V, 8MHz) as a trigger to power up the main Arduino board, responsible for the
launch process. It has been reported that an Arduino Pro Mini could just draw about 7mA when
idle and 15mA when transmitting signals through digital pins. Because of the variable time taken
for the GSM shield to connect to a network, a certain feedback from the main Arduino can to be
sent to the Trigger Arduino to let it know if it is ready to be shut down once it has checked for
messages, making this option much more flexible and robust.
To get an idea of what intervals we need between “on” and “off” times for the main Arduino, it
is necessary to know the average current draw when the main Arduino is running, the average
“C” battery capacity, and the average time it takes for the Arduino to connect to a network to
check messages that could prompt a launch. Experimentally, it takes on average (over 8 runs) 18
seconds for the Arduino to connect to the network. If an average current draw is assumed to be
100mA on start-up (peaks at over 122mA but stayed within 70-80mA most of the time), and
referring to the diagram in Figure 13-4, the “on” time should be within a minute on average. It
has been observed that the servo “twitches” at the end of each loop, meaning as soon as the first
twitch is observed, the Arduino has successfully connected to the network and checked for any
available messages once.

56
FIGURE 13-3: TIME TAKEN FOR GSM TO INITIALIZE
FIGURE 13-4: INITIAL FLOW DIAGRAM FOR THE PROGRAM
So based on the average time it takes for the Arduino to connect, and given a battery capacity of
7800mAh for C [16], this equates to a 78 hour maximum endurance. However, a significant
charge must remain such that the hotwire can be turned on for 5 minutes. The hotwire has a
resistance of 3Ω and therefore draws 2A when activated on 4 “C” batteries, meaning that a total
of at least 1667mAh battery capacity must remain after two weeks. An allowance of 5800mAh
was therefore given for the main Arduino to run checks for a period of 2 weeks. As such, the
main Arduino can be on for a total of 2320 minutes (taking into account the additional 50mA
current draw from the relay that enables the powering of the main Arduino board) or 2320
cycles. The delay between each “on” state of the Arduino was calculated as:

57
Where 1,209,600 is the amount of seconds in two weeks, 2320 is the number of check cycles, and
60 is the estimated “on” time of the Arduino during which it will check for text messages. That
means there is a compromise between outdoor endurance and response time between the text
sending and balloon launch.
FIGURE 13-5: ILLUSTRATION OF THE "ON" AND "OFF" TIME
The trigger will always keep the main Arduino on, until ready, but the delay between the “on”
times will always be 461 seconds.
13.3 CONCEPT
There were many initial designs, but as the project progressed, the concept design evolved and
finalized into the following key features:
-The system is controlled by a trigger (an 8MHz, 3.3V Arduino Pro Mini that completely powers
on or off the main Arduino, via a transistor-relay system). The transistor relay system is used to
reduce the current draw on the lithium cell. The coil resistance of the relay is about 100 Ohms so
if the main board was required to stay on, the trigger board will need to supply a constant 50mA
during that duration, which would result in a lower overall endurance. A transistor was therefore
added to control the relay with the main power source rather than from the trigger power source.
For the actual test, the Arduino was directly connected to a USB power source to avoid potential
harder to find problems and have a quicker launch.

58
FIGURE 13-6: SYSTEM WIRING
As such, only the Arduino Pro mini is constantly powered, and draws only 9mA when idle
compared to over 76mA registered on the main Arduino, while idle. A single 18650 Lithium-Ion
battery is able to power such an Arduino for about one week and a half on a full charge. Due to
the relatively large size of the box with respect to the electronics, it is possible to add an
additional battery set in parallel for additional endurance. For the Trigger Arduino to know
when to stop powering the main Arduino, a high state can be sent from a digital pin from the
latter Arduino to the trigger Arduino while it is busy. A voltage divider is used since one
Arduino operates on 5V and the other one on 3.3V.
-The whole system is powered by 4 “C” Batteries, with the exception of the trigger, which is
powered by a lithium battery. The reason is that a voltage regulator is more efficient if the input
voltage is closer to the desired output voltage.
-The Futaba servo which controls the opening of the lid and the gas canister is hooked directly
on the raw power supply (6V), as it is rated for 7.4V anyway, and in addition, the main Arduino
cannot provide enough power for both the subsystems (sensors, hot wire…) and the servo
without compromising the regulated power stability. Were the servo connected to the 5V
regulated power supply by the Arduino, the hot wire anemometer may have its results affected
[17].
A schematic of the subsystems of the main Arduino is shown below.

59
FIGURE 13-7: SUBSYSTEM WIRING
Note that due to budget and time restrictions, a trigger Arduino was not used in the practical
experiment, and the Arduino was directly powered using a USB charger. An Arduino Mega was
used as a trigger to test the above circuit with success.
13.4 CALIBRATION
Calibration is required to ensure that the whole system will work in most of the conditions it will
go through. However, to run calibrations, tests must be run and they tend to be costly in terms of
Helium spent. The subsystems that were required to calibrate were:
 The lid open time: the lid has to be open for as long as the helium from the canister is
transferring gas into the balloon. An under full balloon will have a poor ascent rate or
may never lift off from the box. A simple timing was done to estimate roughly the time
taken. It usually takes up to four minutes for the balloon to fill, but there is always a
possibility of external factors contributing to increasing the filling time (e.g. outside
pressure, temperature, wind, or the lid not fully opening, opening the valve only
halfway…). An additional minute has been added in case external factors come into
play.
 The hotwire: the hotwire needs to be turned on until the balloon is released. It was
initially considered to put magnets on the payload and a reed switch as a feedback
system to let the Arduino know when the balloon has been released. However, due to
payload weight considerations as well as to keep the electronics simple, the idea was

60
scrapped. On the first day of testing, the hotwire failed to release the balloon, and
external temperatures were likely to blame. A time temperature relationship has been
made, but further experimentation is needed to get more data points. In addition, the
design of the hotwire evolved throughout the project to be more reliable. As such, more
testing should be carried on to adjust the heating time.
FIGURE 13-8: TIME FOR THE HOTWIRE TO MELT THE ONE-WAY VALVE CONNECTION
The anemometer: the anemometer is used to measure the wind speed and determine whether or
not the launch of a balloon is possible. A too high wind speed can result in catastrophic balloon
bursts due to the balloon hitting the ground. This was observed in one of the first wind tunnel
experiments. A lower wind speed also reduces the risk of the balloon’s fabric getting damaged if
it hits the ground.
A hotwire anemometer works on the principle of heat transfer, and will require a larger current
drain as the wind speed increases. In the wind sensor used for this project, a Modern Device wind
sensor, the internal circuit will output a bijective relation between output voltage and wind
speed. The Arduino code maps the function to integer (the analogue pins are 10 bits in
resolution). The following plots are from the wind tunnel experiment. The spreadsheet
containing the data can be found in the Annex.
0
10
20
30
40
50
60
0 5 10 15 20 25
Time(s)
Temperature (Degrees Celcius)
Melting time

61
FIGURE 13-9: HOTWIRE ANEMOMETER CALIBRATION AT 8.3°C AVG.
FIGURE 13-10: HOTWIRE CALIBRATION AT 9.8°C AVERAGE
Tests have been conducted in ascending and descending order to test for hysteresis properties.
From Avionics 2 slides [18], “a transducer might give a different reading depending on whether it has
been reached by an increasing or decreasing change in value”. It can be seen that there is still a very
small difference, but it is insignificant in comparison to the temperature’s effect. By personal
experimentation, going outside and inside a house can result in a large change in readings (tested
at 20°C and around 0°C during the winter) so an error in wind speed by up to 5 meters per
second if temperature is not considered. Unfortunately, due to the limited amount of days that
was available in the wind tunnel for the group, not a wide range of temperatures were available
to calibrate the wind sensor, so the data is sparse.
0
50
100
150
200
250
-2 0 2 4 6 8 10 12
AnalogValue
Wind Speed (m/s)
Hotwire calibration at 8.3C
Ascending
Descending
0
50
100
150
200
250
0 5 10 15
AnalogValue
Wind Speed (m/s)
Hotwire Calibration at 9.8C
Ascending
Descending

62
FIGURE 13-11: OVERVIEW OF THE VALUE READ BY THE SENSOR TO THE WIND AND
TEMPERATURE

63
FIGURE 13-12: ALTERNATIVE ANGLE TO SHOW TEMPERATURE EFFECTS
The plots match our expectations: the lower the temperature, the higher the heat transfer,
therefore the larger the read value. As the temperature increases, the heat transfer from the
resistor to the ambient air is smaller and therefore the value decreases. Different orientations
were tested, and the impact of it seems negligible. As such, an expression of wind can be
expressed as:
Where U is the value returned by the wind sensor normalized into integer, x is the actual wind
speed that needs to be estimated, T is the temperature in Celsius provided by the DHT11
Temperature sensor, and a, b and c are constants which are determined from the calibration.
Given the lack of data, it was considered too hasty to include temperature effects, so a two part
single variable linear function was used.
⁄

64
⁄
FIGURE 13-13: HOTWIRE CURVE FIT
13.5 CODING
Three sets of codes have been written. The first one in Appendix C shows the code in its simplest
form: connect, receive and open the lid if any message is received, activate the hot wire and then
close the lid. It does not take into consideration any wind/temperature. This code was used for
the launch test, so the launch conditions were already predetermined.
The second one is the trigger Arduino code, which is again very short. It can be found in
Appendix D
The third one is a more comprehensive version of the first one code, which involves sensor
readings, and a “busy” feedback signal for the trigger Arduino (which is basically just a few more
lines, and an infinite loop at the end of the code). It can be found in Appendix E.

65
14 PAYLOAD ELECTRONICS
Written by Binay Limbu, assisted by Reetam Singh
This chapter describes the objectives of the payload, identifying the customers and the customer
requirement. Different options available for the payload electronics are looked into and the
selection of particular electronic component justified. A detailed explanation of how the radio
transmitter operates and how the tracking works is also explained. The flowchart for the program
that goes onto the main payload is also given. Finally the whole payload is reviewed as to
whether it could achieve its objective with some recommendation given for future work.
14.1 INTRODUCTION
The payload plays an important role in the project. The whole concept behind the project being
that the payload must be able to measure the atmospheric parameters such as temperature,
humidity and pressure at the corresponding altitude and be able to transmit these data in real
time so that the acquired data can be examined for research purposes. Therefore the aim of the
payload is that it must be able to function as a radiosonde. A radiosonde is a small weather
station coupled with a radio transmitter attached to weather balloon. Weather personnel launch
these across the world to collect atmospheric data for weather prediction. During the
radiosonde’s ascent, it transmits data on temperature, pressure and humidity to a sea-, air-, or
land based receiving station. Often, the position of the radiosonde is tracked through GPS, radar
or other means. Thus the radiosonde flight produces a vertical profile of weather parameters in
the area above which it was launched [19].
14.2 OBJECTIVES
The objectives of the payload can therefore be defined as:
 Measure the atmospheric parameters such as temperature, pressure and humidity.
 Measure the altitude and give the location of the weather balloon using GPS.
 Should be extremely light to conform to the low lift generated by the small weather
balloon used in the box.
 Should use the unlicensed frequency band for telemetry as this reduces the legal hassles
involved.
 Keep records of the data collected by the payload for future reference.

66
14.3 PAYLOAD BINARY MATRIX
The main customers of the payload are the students who have undertaken the project and the
potential customers such as high altitude balloon enthusiasts and academics who will be using
the device for research purposes.
FIGURE 14-1 BINARY MATRIX FOR THE PAYLOAD.
14.4 PAYLOAD OPTIONS AND SELECTION
An extensive research was done online to check if there were off-the shelf products that would
slide in perfectly into our design satisfying the requirements. A number of options became
available. Therefore in order to cut down and make a decision all the pros and cons of the
concept generated were listed [20].
1. EAGLE FLIGHT COMPUTER [20]

67
2. VAISALA RS92 SGP RADIOSONDE
3. BUILDING A CUSTOMISED PAYLOAD

68
TABLE 14-1: PROS AND CONS OF DIFFERENT PAYLOAD CONCEPTS.
EAGLE FLIGHT
COMPUTER
VAISALA RS92 SGP
RADIOSONDE
BUILDING A
CUSTOMISED
PAYLOAD
PROS 1. Flight computer
2. GPS Enabled
3. Measures
pressure,
Temperature and
Humidity.
4. Light (36 g)
1. Measures the
atmospheric
parameters
accurately.
2. Considerably cheap
1. Custom built to
satisfy the
project’s need
such as
transmission in
the unlicensed 434
MHz frequency
band.
2. Can include any
number of sensors
and devices.
CONS 1. Really expensive.
2. No telemetry
1. Requires ground
check set.
2. Water activated
battery.
3. Requires additional
software for telemetry
decoding.
4. Transmits data on the
licensed 403 MHz
frequency band.
1. Time limit for the
completion of the
project.
In order for the project to be successful the payload needed to satisfy the important customer
requirements listed in the binary matrix. Since this was not possible to achieve with the help of
off the shelf products, a decision was undertaken to build a customised payload to suit the
project.
14.5 HARDWARE SELECTION
14.5.1 MICROCONTROLLER
One of the main components in the payload is a computer that would be able to coordinate all
the activities of the sensor and pass it on to the radio transmitter to transmit the data. The on-
board computer needed to be programmable so as to be able to control the sensors. The
following boards were considered for the operation.
1. RASPBERRY PI
The Raspberry Pi is a credit-card sized computer that plugs into the TV and a keyboard.
It is a capable little computer with a fully functional operating system. Some of the
common distribution used with the raspberry pi is the NOOB installer. The
recommended distribution is the Raspbian, which is specially designed for the Raspberry
Pi and which is always optimised [21].

69
2. ARDUINO
Arduino is an open-source physical computing platform based on a simple
microcontroller board, and a development environment for writing software for the
board. Arduino can be used to develop interactive objects, taking inputs from a variety of
switches or sensors, and controlling a variety of lights, motors, and other physical
outputs. For the payload the three versions of Arduino, the Arduino Mega 2560, the
Arduino Uno and the Arduino pro mini were given serious consideration [22].
ARDUINO MEGA 2560: [22]
 54 pins
 Operating Voltage: 5V
 Supply voltage:7V-12V
 Flash Memory: 256 KB
 Easily programmable and source code available for sensors compared to
other micro-computers
 Power using 9V battery under sleep mode configuration.
 Thermal protection and casing using 5mm EPP Foam.
ARDUINO UNO:
 20 pins (14 digital, 6 Analog I/O)
 Operation Voltage: 5V
 Supply Voltage:7V-12V
 Flash Memory:32 KB
 The Arduino Uno can be powered via the USB connection or with an
external power supply. The power source is selected automatically.
ARDUINO PRO MINI:
 22 Pins (14 digital, 8 analog)
 Operating voltage: 5V (16 MHz model)
 Supply voltage: 5 – 12 V
 Flash Memory: 16 KB
 The Arduino pro Mini can be powered with an FTDI cable or breakout
board connected to its six pin header or with a regulate 3.3V or 5V supply
(depending on the model) on the Vcc pin.

70
FIGURE 14-2: FROM LEFT RASPBERRY PI, ARUDINO MEGA 2560,ARDUINO UNO, ARDUINO
PRO MINI.
14.5.2 FINAL MICROCONTROLLER SELECTION
The advantages and disadvantages of both the Raspberry pi and the Arduino system were
considered. The Raspberry pi is a small computer with a RAM and CPU. Therefore integrating
it into the payload would be complex as the system would need to be customised to handle the
pins input/output and handle the sensor reading. While not impossible to integrate the raspberry
pi into the payload, the Arduino system presented a simpler solution to handle sensor reading
and programming solution. Another advantage that the Arduino provided over the raspberry pi
was it was inexpensive with the Arduino pro mini costing only £8. The Arduino software could
also run on cross platform including windows, Macintosh OSX, and Linux operating systems.
The main advantage of Arduino over the raspberry pi however was the simple, clear and easy-to-
use programming environment which made editing, uploading and debugging program into the
avr much easier. The Arduino software is also published as open source tools, available for
extension by experienced programmers. Therefore a large number of libraries are available for
the Arduino to be able to work with. In the end the Arduino pro mini was selected because of its
extremely light weight (2g) and its versatility.
14.5.3 RADIO TRANSMITTER
RADIOMETRIX NTX2B
The inspiration for using the radiometrix ntx2b in the payload was drawn from the UK high
altitude society. The radiometrix NTX2B is the perfect unit for licence-exempt operation in the
433MHz (EU) and 458Mhz (UK) bands and has been extensively used by the members of the
society. Special features of the ntx2b are the fully integrated sigma-delta PLL synthesizer based
design, high stability TCXO reference and an operating voltage of 2.9V-15V at 18mA. The data
bit rate is 10 kbps max with a transmit power of +10 dBm. Hence the low power consumption
along with an operating range of over 500 m with transmission in the licence-exempt frequency
band made the ntx2b an ideal selection. Another reason for using the radiometrix ntx2b was its

71
proven reliability of working over a large temperature range and at an extremely high altitude
[23].
FIGURE 14-3 RADIOMETRIX NTX2B RADIO TRANSMITTER.
14.5.4 GLOBAL POSITIONING SYSTEM
ADAFRUIT ULTIMATE GPS
The GPS plays an important role in the payload as it helps tracks the balloon giving important
information such as the latitude, longitude and altitude of the balloon. For the payload the
adafruit ultimate gps breakout provided the ideal solution. It is a high quality GPS module that
can track up to 22 satellites on 66 channels, has an excellent high-sensitivity receiver (-165 dB
backing), and a built in antenna. It can do up to 10 location updates a second for high speed,
high sensitivity logging or tracking. Power usage is incredibly low, only 20 mA during
navigation. Also it exceptionally light weight (8.5g) along with small size (25.5mm x 35 mm x
6.5 mm) and its functionality up to 27 km made the decision to use this particular GPS much
easier [24].
FIGURE 14-4 ADAFRUIT ULTIMATE GPS BREAKOUT.

72
14.5.5 SENSORS
DHT22 TEMPERATURE-HUMIDITY SENSOR
The DHT22 is a basic, low cost digital temperature and humidity sensor. It uses a capacitive
humidity sensor and a thermistor to measure the surrounding air, and spits out a digital signal on
the data pin. It is fairly simple to use with an operating voltage of 3-5V and 2.5mA max current
use during conversion (while requesting data). The sensor is good for 0-100% humidity readings
with 2-5% accuracy and for -40 to 80o
C temperature readings with ± 0.5o
C accuracy [25].
FIGURE 14-5 DHT22
BMP085 PRESSURE SENSOR
The BMP085 is a high-precision, low-power barometric pressure sensor with digital two wire
interface. The sensor offers a measuring range of 300 to 1100 hPa with accuracy down to 0.03
hPa in ultra-high resolution mode. Other advantages of using this sensor are its ultra-low power
consumption, with low noise measurements and factory calibrated settings [26].
FIGURE 14-6 BMP085 PRESSURE SENSOR.
14.6 DATA TRANSMISSION AND DECODING
The key idea behind transmitting data using the radiometrix ntx2b is to use the arduino to get the
ntx2b to transmit between two frequencies slightly giving two tones. This is achieved by
adjusting the voltage on the NTX2’s TXD pin which changes its transmission frequency slightly.
The difference in this frequency is called the shift. By doing this in a controlled fashion 1 and 0
can be transmitted and therefore data. In the payload the voltage on the TXD pin of the NTX2
is controlled using a voltage divider [27].

73
A voltage divider is a simple circuit which turns a large voltage into a smaller one. Using just two
series resistors and an input voltage, an output voltage which is a fraction of the input voltage
can be created [28]. Examples of the voltage divider schematics is shown in the picture below:
FIGURE 14-7 EXAMPLES OF VOLTAGE DIVIDER SCHEMATICS.
Assuming that the three values of the above circuit i.e. the input voltage (Vin), and both resistor
values (R1 and R2) are known the Vout can be calculated using the following equation. Thus
the given formula is used to calculate the shift generated by the voltage divider in the radio
transmitter.
The NTX2 is a FM (Frequency Modulation) module intended to have a voltage applied to the
TXD pin of between 0 and 3 volts. This voltage range changes the output frequency of the
module by up to 6 KHz. Therefore for 1 Hz change in frequency the voltage needs to be changed
by (3/6000) 0.0005v. So to get a shift of 500hz the voltage applied to the TXD pin needs to be
toggled by 500x0.0005=0.25v. It must be specified that getting the shift totally accurate isn’t
entirely essential so long as it is in the 300-600Hz region. The circuit diagram for the radio
transmitter and the voltage divider is shown below:
FIGURE 14-8 CONFIGURATION OF RADIOMETRIX NTX2B FOR DATA TRANSMISSION.
Vcc = 5V
GPIO low = 0 V

74
GPIO high = Vcc
R4 = R5 = 4.7K
R3 = 47K
NTX2 deviation ~ 2000Hz/v
The two 4k7 resistors, R4 and R5, set the bias (centre) point of the two levels to be half the
supply voltage. The GPIO pin then deviates the voltage from the centre point by an amount
controlled by R3. The calculation for the two cases when the GPIO pin is high and low is given
below:
FIGURE 14-9 CIRCUIT DIAGRAM FOR VOLTAGE DIVIDER CALCULATION.
Therefore when the resistors are in parallel, where ǁ signifies resistors in parallel the value is
given by
R1ǁ R3 = R2ǁ R3 =
( )
= 4272.72 ohms
When the output is high (5v) Vout in the TXD pin is given by:

75
When the output is low (0v) Vout in the TXD pin is given by:
FIGURE 14-10 CIRCUIT DIAGRAM WHEN GPIO IS HIGH (5V) AND GPIO IS LOW (0V).
Thus voltage shift = 2.619 – 2.380 = 0.2380v
Therefore the frequency shift is = 0.238 x 2000 = 476 Hz. This is the frequency shift generated by
the payload.
Thus by controlling the GPIO pin of the arduino the voltage of the NTX2 TXD pin of the radio
transmitter can be controlled and using 1’s and 0’s data can be transmitted via RTTY.
RTTY or Radioteletype is a telecommunications system consisting originally of two or more
electrochemical teleprinters in different locations. These have however been replaced recently as
the work can be carried out using radio transceiver with a computer and soundcard. RTTY
works by shifting the transmitted frequency. They are known as fsk (frequency shift keying) and
afsk(audio frequency shift keying) [29].
In the payload a program is uploaded onto the arduino pro mini. The program reads the data
from the GPS and the sensors and drives the NTX2’s pins connected to the GPIO. The sensors
and the GPS data are saved in a string format in the program called the DATASTRING. This
DATASTRING is passed to a procedure called rtty_txtstring in the program which takes care of
transmitting the data by breaking it down into characters, which then transmits the individual

76
bits of those characters. With the circuit set up as shown above the program in the payload
transmits the DATASTRING at 50 baud, 7 bits ASCII 2 stop bits. Checksum is also added at
the end of the DATASTRING for the purpose of detecting errors that may have been introduced
during its transmission or storage [27]. The flow chart for the program is given below. The code
running on the Arduino pro mini is provided in the appendix of the report.

Compact Remote High Altitude Balloon Launcher

Compact Remote High Altitude Balloon Launcher

Recommended

Recommended

More Related Content

Similar to Compact Remote High Altitude Balloon Launcher

Similar to Compact Remote High Altitude Balloon Launcher (20)

More from Reetam Singh

More from Reetam Singh (12)

Recently uploaded

Recently uploaded (20)

Compact Remote High Altitude Balloon Launcher