Our aim is to develop an intelligent payload, the skyPod v1.0 (to be attached to a weather balloon) which is capable of reaching altitudes up to the stratosphere, while recording environmental data to be logged on board and to be transmitted back to a station on the ground. The payload will record:
Some of the intelligent features include:
The intention is for the balloon to carry the payload up to a height at which atmospheric pressure causes the balloon to expand and burst. The payload is brought back down to earth safely with aid of a parachute and the payloads dual-tracking system increases chance of recovery, allowing data (which has not been transmitted back) to be extracted and processed for examination and analysis.
Space based equipment:
For all the SUNSET launches, there are two ground stations. One is a mobile unit, one is a student run satellite tracking station based at the University of Strathclyde, called STAC.
== STAC ==
An array of sensors was constructed to measure the following:
This was formatted (nearly) to the UKHAS standard and sent back to anyone listening on earth
A meteorological balloon was sourced to carry the payload to high altitude. Such balloons were only available from a limited number of manufacturers. An established one being the Japanese based company Totex Corporation (1). They produce a popular range of balloons with detailed documentation freely available online. This data was used as the basis of the calculations in the following sections.
The following parameters were important in balloon selection:
These are inter-related and conflicting parameters, which are determined by the gas volume used and burst diameter of the balloon selected. For any one balloon a greater volume of gas results in a greater ascent rate but the balloon will burst at a lower altitude. These parameters, and others, are summarised can be seen online at http://ukhas.org.uk/guides:balloon_data
Some of the terms from this online table require further explanation:
From analysis of the data from the table, the KCI 1200 was selected as the balloon of choice. This balloon offered the best compromise of ascent rate, burst altitude and cost.
For a maximum payload mass of 1.65 kg and a burst altitude of 29 km, the data shown in Table 1 was calculated using a pre-existing spreadsheet which can be found online at http://ukhas.org.uk/_media/guides:burst3.xls.
|Burst Altitude||29000 m|
|Launch Diameter||2.27 m|
|Balloon Area||4.047 m2|
|Launch Volume||6.124 m2|
|Burst Volume||336.535 m3|
|Free Lift||33.704 N|
|Ascent Rate||7.436 ms-1|
|Time To Burst||65 minutes|
|Payload Mass||1.65 kg|
The balloon was sourced from a UK supplier of Totex sounding balloons, Random Engineering, at a cost of £70.25.
A 9.01m3 canister of Helium gas was sourced from BOC for £91.10. BOC were able to offer a significant discount through the University and thus no alternative suppliers were considered. Welding gas was selected as it was cheaper than party balloon gas and was of a higher purity (99.9%).
To increase the chances of recovering the payload a ‘cut-down’ mechanism was built in to the system. The purpose of the cut-down mechanism was to detach the payload from the balloon by cutting the balloon cord if the payload were to drift into an area which may cause it to land in an inaccessible location (e.g. in a body of water, the middle of a busy city or a mountainous area).
As the payloads live location was being continuously monitored, the cut-down mechanism could have been deployed upon receiving an actuating signal from the ground. However, as it could not be guaranteed that transmission would be successful it was decided to make the payload fully autonomous. Thus, an acceptable perimeter was pre-set in the flight software prior to launch and the mechanism would actuate if the payload crossed the pre-set boundary.
There were two practical solutions to detaching the balloon from the payload:
|Mechanical Actuator||Hot Wire Cutdown|
|Weight||Motor/Solenoid is Heavy||Almost 0 added weight|
|Space||Requires space in package||Very small internal footprint|
|Reliability||Moving parts, freeze/seize risk||Wire - freeze risk|
|Power||Power requirement is greater||Low power requirement|
|Cost||Greater costs associated||Very cheap to implement|
Once the advantages and disadvantages were identified for each, it became clear that the hot wire system would provide a more effective and robust solution as no advantages were identified for the mechanical actuator over the hot wire system.
The location of the cut-down mechanism required some consideration, since its location would affect its performance. Figure 1 shows different arrangements considered.
Cutting the balloon cord inside the payload with the parachute attached to the balloon cord would result in the parachute being detached from the payload, meaning it would be left to fall back to Earth with no parachute.
Attaching the parachute directly to the payload instead of the balloon cord would have a number of disadvantages: · The parachute could become tangled in the payload and fail to open when the balloon is detached. · The hanging parachute could obstruct the camera ports. · While ascending the parachute could open and act against the lift of the balloon.
To prevent the parachute flapping around the payload and obstructing the cameras it can be held further up the balloon cord by feeding the balloon cord through a hole in the parachute canopy and tying a knot to prevent it falling down, as illustrated in Figure 1. However, the lower knot would prevent the balloon chord slipping through the spacer and the hole in the parachute canopy and so the whole system would still be carried upwards by the balloon.
Cutting the balloon cord above the parachute and below the balloon would allow the balloon to be completely detached and the payload to descend with the aid of the parachute. It would, however, require copper wires to run up the balloon cord from the payload to a section of nichrome coiled around the balloon cord.
The next image shows the consequences of mounting the cut-down at each location. As a result of this analysis, location 3 was selected.
The hot wire cut-down mechanism operated by passing a current through a section of nichrome wire coiled around the nylon balloon cord. The nichrome wire was heated to a very high temperature, allowing it to melt the cord. The current was delivered to the small section of nichrome from the payload using normal copper wires which ran up the balloon cord.
Extra care had to be taken when coiling the nichrome around the balloon cord as the nichrome coils needed to be close enough to each other to successfully cut one section of the nylon cord, but not too close that they touched and created a short circuit preventing the nichrome from heating up and cutting the cord.
The nichrome wire used had a 0.1 mm diameter (38 AWG) and had a resistance of 1 Ω/cm.
The system was tested using a 5 cm section of nichrome with a 5 V supply and then with two 1.5 V batteries.
The 3 V solution offered a sufficiently short cut-down time, when compared to the 5 V solution. It was decided to power the cut-down mechanism with the 3 V ancillary power supply as described in power source section. It was decided to keep this separate from the main power supply due to its high current draw (0.55A).
An integral part of the project was the requirement to produce high quality images from within the stratosphere, which meant that the selection and implementation of imaging equipment had to be decided upon in the early stages.
An initial idea was to use a Canon camera with high resolution as it is possible to implement code to automatically control the camera. This is known as Canon Hack Development Kit (CHDK) and is well documented online. The first step was to download the correct CHDK auto-build for the camera, which was identified using the Automatic Camera Identifier and Downloader (ACID). From this the firmware version was identified allowing the relevant full build to be found. The final preparatory step was to set up the SD card which was done using software called Cardtricks.
The process for controlling the camera was as follows:
The script then ran automatically upon powering the camera. The settings could be written into the script as defaults or adjusted while using the camera.
For testing purposes a Canon IXUS 55 was used, the relevant software was downloaded and a memory card was prepared. The script required for taking photos at set intervals is known as an intervalometer script and there are many open-source scripts available online. A particularly good script was found which had numerous configurable parameters including:
This script did not work immediately and so was adapted by using sections of another script which did work but had fewer features. This resulted in the final code which was tested and performed as required.
The Sony Ericsson X10 Mini mobile phone which was purchased for tracking had a built-in five Megapixel camera with high quality lens. Various free applications were available for the phone from the Android Store to take photographs at regular intervals. The selected application was called ‘Timerlapse’. Once set up it performed efficiently and effectively in the background, taking high quality photographs at ten second intervals throughout the flight. The application was also programmed to stop taking photographs after four hours, to preserve battery life and phone memory.
Thus two methods were developed for taking still photographs. Both methods were tested and worked very effectively. However, in order to save on weight and cost only the phone was used to take high quality photographs instead of the CHDK program.
It was stated in the design specification that High Definition (HD) video footage must be recorded during the flight and so it was required to choose a video camera which could record at a minimum of 720p (1280×720 resolution), but ideally at full HD which is 1080p (1920×1080 resolution).
The Kodak Playsport was selected because of its waterproofing, high quality images, smaller size and power saving due to smaller screen and dimming.
The minimum thermal power output by the circuit was calculated as follows:
This was based on the fact that the camera and phone would not always be charging and only the circuit could be assumed to be always drawing current.
It was decided intuitively that a heating circuit should generate twice this thermal power to provide sufficient heating for the circuit components. The heating circuit was comprised of many small resistors rather than one central heating resistor. This created a bigger surface area for faster heat dissipation and allowed heat to be supplied directly to each component rather than just the air in the payload. The heating was to be supplied with two AA Energiser Ultimate Lithium batteries supplying a total of 3 V. The required power output was set at 2.5 W, allowing the following calculations to be performed:
This current was achievable, with the Energiser Lithium batteries being capable of 2 A continuous discharge.
The 3.6 Ω required was created from a network of ten resistors connected in five parallel branches of two series resistors. The individual resistor value required was calculated as follows:
The closest commercially available resistor value was 9.1 Ω leading to the heating resistor network shown in Figure 1.
Figure 1: Heating Resistor Network
The completed network had the following electrical characteristics:
(I've noticed the maths looks funny. It'll be sorted soon)
Thus the heating circuit was designed to dissipate 2.473 W of thermal power, with each individual resistor being required to dissipate 0.25 W from 165 mA of current.
The resistor chosen was RS Product Number: 707-8069 as it had adequate power capacity and a small minimum order quantity as only ten were required.
The lovely graphs will be here soon, people.