Pegasus High Altitude Balloon Project

Pegasus V

Mission Successful (sort of)

Launched but underfilled the balloon so climbed very slowly, reached 10km altitude and then cutdown fired and the payload parachuted down, landing in a field (here). The KML file of the flight can be found here. As the payload was undamaged it after some software updates (to get the compass actually logging and also perhaps time the camera to take pictures facing particular directions) it'll be launched again - perhaps a sunrise launch...

Will update this site with details of the construction soon.

Aim

Originally the plan was to investigate the post burst, apogee chaos with accelerometers and an upward facing camera videoing the burst and subsequent parachute deployment. Instead the mission aim was changed to test transmitting the payload position by morse by toggling a GPIO connected to the "lego brick" 434Mhz 10mW radio transmitter. Also onboard would be a digital compass to measure the rate of spin that the payload experiences during the flight. Using this data it should be possible to time the triggering of the camera to take panoramics or taking pictures facing a specific direction. Finally all the components will be connect via a expansion board to allow them to be easily changed and reused - there will also be 2 switches exposed on the payload surface to allow the payload to be pre-set up and then just turned on at the launch site.

Equipment

• Gumstix 200xm connex
• GPSstix
• Nokia 5110
• 434mhz 10mW Radio Transmitter
• Olympus mju 300 (3.2 MP) Camera
• i2c Digital Compass
• 4 AA lithium cells

Theory of Operation

• GPS/Phone system stays the same - every 2 minutes takes gps coords and sends by sms - also logs its cell ID, battery levels and reception.
• For the radio - perl scripts parses the gps log - converts the gps coords into a sequences of on/off which is used to toggle a gpio to the radio. Encoding will probably be alternating Morse code and turning the radio on and off equal to the number (e.g. if its 5 then on/off 5 times). To speed everything up a C program will control the on/offs and I'm going to delve into PXAregs to directly access the gpio.
• Digital Compass will be attached by i2c - probably record spin alternating minutes. Note - requires a 5v power supply
• Camera - will face directly up - to monitor balloon burst and parachute. Have directly traced the shutter and power switch so will attach gpios to control. With a 512mb memory stick it'll do 22minutes of 640x480 and at 160x140 over 3 hours. Might upgrade to a bigger card - if not will turn the camera on at a particular altitude in an attempt to record the balloon burst. Change of plan! - instead camera will face sideways taking pictures every minute

Planning

8/02/07

In the planning stages - have all the equipment so there will be a features freeze - I'm not going to add any more stuff (apart from a cutdown device - though I'd really like to go to burst altitude - might be just in case it drifts to far away). At present trying to work out a payload case which is partially reusable. Also waiting to get some connectors - am going to construct a connector/expansion board to save direct soldering to the gumstix and also allow in the future to add more stuff - will include an power on switch for the phone and gumstix so at launch everything is put together. Just a matter of turning it on and letting it go.)

Construction

• GPS is onboard the GPSstix and is attached to ttyS2, read by gpsparser3.pl script)
• Nokia 5110 is connect on ttyS1, gnokii is used to communicate between the gumstix and the phone.
• Digital compass is attached to the i2c (SDA and SCL) using the compass binary compiled from the manufactuers example source code here
• The radio is powered from the VCC pin off the gumstix and is toggled high and low by a GPIO using a C program that itself uses pxaregs to directly access the GPIOs (source can be found here). The gpsparser3.pl while logging to /gps.log also logs the lat,long and alt to /lonlatalt.log which is read by another perl script which translates this to a string that is feed to the C morse program and is then broadcast. The radio broadcasts 10 seconds of continous tone and then the long, lat and altitude encoded as morse code at 11wpm.
• Cutdown is controlled by a single GPIO to the cutdown charger circuit described on the UKHAS wiki. The software has both a square of coordinates that if exceeded triggers the cutdown or the cutdown can be triggered after 2 hours from the powering up of the flight computer.
• The camera was damaged after being dropped of a cliff and so had to be slightly adapted to allow it to be turned on, the hack permenantly leaves it on and so on/off is controlled by removing the battery. Triggering the camera is done via infared instead of the direct shutter button hack. Here the remote trigger is permentantly turned on but shorting out the push button, the 3v cell is removed and instead the remote is powered via a gpio off the gumstix - turned on when the picture needs to be taken.