Atlas

• Test out:
  • New flight computer
  • Crystal oven to stabilise the frequency drift
  • 30% duty cycle on radio Tx
  • Long range receiving + coordination
  • New code
  • Launch line launch technique

• New PCB design

• ATmega328 flashed with arduino bootloader
• NTX2 434.075Mhz 10mW radio transmitter
• LD1117 3.3v Voltage regulator
• Crystal Oven (DS18b20 temp sensor, P2N222A transistor + 2x 10ohm resitors)
• FSA03 Ublox GPS module + breakout board
• Internal DS18b20 temp sensor + External DS18b20 temp sensor
• External Photodiode

• Arduino Libraries:
  • PID_Beta6
  • TinyGPS (ublox adapted)
  • OneWire
• Code

Good outcomes:

• Crystal oven works - manged to keep the NTX2 radio crystal at 0 degC through out the flight, this resulted in less drift on the radio - had a number of reports that it wasn't necessary to retune the radio as often - certainly something that will be included in future flights
• The launchline technique works really well - such an easy launch especially for flight trains that have low tensile lines.
• It is possible to recruit lots of listeners - had excellent coverage throughout most of the flight - struggled at the end and so would be advantageous to recruit more listeners in the Netherlands.
• Using a preformed box was great - saved a lot of cutting etc and was well sealed - ideally need a smaller box for future flights as had a lot of spare space.

Problems

• GPS - gps was very unreliable, firstly it struggled to get an initial lock - hence the late launch. Secondly it was stuck in portable mode so stopped working at 12km altitude. The fsa03 has had a number of problems in the past including a wobbly antenna.
  • Use a different GPS - there is always the lassen IQ (though it is old) or another type of ublox gps - such as the GPSbee from seeedstudio.
  • If using a ublox gps at regular intervals reset the mode to Airborne < 1g so that it can work above 12km.
• Did not float - we launched too late (15:20) and probably overfilled a little with an ascent rate of ~3.5m/s
  • Be more prepared
  • Use more accurate scales to calculate lift.
• Radio duty cycle - people had difficulty with tracking as the duty cycle was low, instead of using 1 string repeated twice per minute instead go back to continous transmission at least for the first part of the flight and then perhaps switch low power mode.

Thoughts

• Backup beacon
  • Add GPS, power up every 5 minutes of so
  • Different design of antenna to improve reception.
  • Different payload box, perhaps on the vent itself.