FHALP-2

28/02/08 - FHALP-2 has been found in a field in Cambridgeshire

Pictures from the camera

A few good pictures:

• Picture of parachute at apogee
• Cool horizon picture
• Flares from the Sun picture

Flight Log:

• Flight
• Total Log
• Kml file of flight

The fhalp-2 design was an updated version of the fhalp-1 design with the hope of running it off a single 3.7 2000mAh Lipo battery. The GM862-GPS module from telit formed the main flight computer which has an onboard SirfIII GPS, GSM module and a python running microprocessor. The theory of operation was:

• acquire GPS coords by AT commands (as the GPS library wasn't available in the firmware I was using) which was already formatted out of the NMEA.
• Check to see whether there was a GPS fix and compare to flight parameters (e.g. cutdown alt)
• Attempt to send a SMS
• Use the morselib to convert the coords into morse code and then transmit them down the radio by toggling a GPIO line attached to the datain line of the radio.
• Log the coords to the main memory of the module (while this isn't advised due to the limited read/writes as the flight was relatively short I thought it would be okay.
• Power save for 30 seconds and then repeat.

Therefore the wires that were plugged into the GM862-GPS breakout board were:

• On/Off switch connected to On/Off and Ground - turning on involved pulling the line to ground
• DTR to Vcc which booted the module straight into the python script engine
• An LED between Vcc and Status (an open collector) to show the state of the module
• GPIO 7 to reset line on the 555 IC to stop it using the radio (pulled low to allow the module to transmit morse)
• GPIO 3 to Radio Data In
• GPIO 6 to cutdown circuit
• 3.7 2000mAh Lipo battery to Vcc and Gnd

The cutdown device and radio also required a power supply so the was a 'power strip' of pins drawn off Vcc and Gnd to accommodate this.

• The radio was a circuit design 10mW module that transmitted on 434.075Mhz, I added a CMOS 555 IC which produced a 1hz pulse when the GM862-GPS module wasn't transmitting morse, when the module wanted to transmit the morse it pulled the reset line to GND allowing the module to toggle the data in line.
• The cutdown device was based upon the design on the ukhas wiki here

As a final addition a seperate camera payload was added using an Olympus Mju 300 digital camera which I had hacked up at an earlier date (howto on the ukhas wiki). I added a second CMOS 555 timer which ran off the camera's battery and again toggled the shutter lines. This fitted snuggly into a cardboard box I found and was hung at a slight angle below the main payload.

• Test GM862 module as a flight computer
• Test radio module (especially the use of Morse)
• Compare Morse decoding software (Cocoamodem and CWget)
• Measure GSM reception

• GM862-GPS - got
• Circuit Design Radio Module - (Morse Beacon) - got
• Cutdown Circuit (based on a Mosfet) - got and built circuit
• 3.7v Lipo Battery - sparkfun UK Supplier - got
• GPS antenna - got
• GSM antenna - still to get Active Robots
• uALFAT Data Logger - got

So the plan is to make a morse beacon that pulses on and off at 1Hz and then every minute the gm862-gps takes over and pulses out the time,lat,long and alt. To achieve this without using a complicated PIC I've got hold of some CMOS 555 ICs which will be set up to pulse the radio at 1Hz and the GM862 will pull the reset to ground and toggle the pin itself.

Its not the best design in the world however testing seems to work, when the gm862 is off or not transmitting it emits a 1hz pulse and when the gm862 wants to transmit it resets the 555 and transmits its morse via a GPIO.

Picture of GM862, lipo 2000mAh battery, radio module and data logger (not yet wired up) - lots of the wires are for development and connection to my laptop!

The day before the launch during testing I discovered that the GM862-GPS wasn't detecting the presence of the SIM card and therefore wasn't sending and SMS messages. After a lot of testing I came to the conclusion that it was something internal. However I decided to continue the launch as in theory the module was spare as I would still need to order a new one to get a fully functioning payload.

The launch was actually went really well, we all got there at 10am - the balloon was already inflated as Steve had been using it for tethered flights the day before, it was quite a large balloon with definitely enough lift. While Fergus and I set up the payload Steve prepared the parachute and Ed packaged up the camera payload. Once everything was connected up we checked that the radio was working and released it at 11.12am. Using Doug's binoculars we were able to track the balloon for over an hour while still listening to the radio transmissions (however due to drift it was difficult to fully decode full gps strings). However after about 1 hour the radio transmissions ceased and a few minutes later Doug reported that he could see something seperating from the balloon (quite possibly the cutdown activating). After this we didn't receive any further radio signals and as the sim card wasn't working didn't get a text message on landing. In the hope that the radio might have started working again we attempted to predict the landing site and visited a few areas, making sweeps with the yagi however we didn't pick up any signal and the payload remains lost just East of Cambridge. My name and number is on the side of the camera payload - if you find it please get in touch!

• Doug put some launch pictures up: here
• Radio log here
  • The data comes in the format 'FIREFLY latitude longitude altitude' e.g. FIREFLY5215.11Ø9 ØØØØ5.5293 5200
• Flight KML here
• Flight Forecast (Uni of Wyoming Trajectory Forecaster) here
• Flight Forecast 24hr in advance here

• Molex connectors aren't the greatest in the world and end up taking quite a lot of space up, in hindsight it would be good to have 2 breakout boards, one that has headers for development while the other has the wires directly soldered to it and will be the payload main board. Its then a matter of shifting the module between boards.
• Seperate radio beacon for searching when everything goes wrong!
• Morse isnt good - either go for RTTY or perhaps just pulse out the number, that everyone can take part and all they have to do is listen and write down the numbers.
• Its quite important to have a working GSM module!
• The SIRF III chipset is irritating with its altitude limit - it might be better to use the cheaper GM862 (without GPS) and then add a GPS on the 2nd serial port which is able to cope with the higher altitude.

• The GM862 carried on logging throughout the flight which means that there wasn't a power failure. It actually logged until 1:00am so there is no need to worry about it running out of battery power when using a 2000mAh LiPo.
• On opening up the payload there didn't seem to be much damage and all the connectors appeared to be in place which suggests that there wasn't failure there.
• While the ground plane wires are still attached the downward facing antenna isn't which could be the cause of the loss of radio signal
• The cutdown has fired however the log shows that the payload didn't reach the 23000m cutdown limit possible reasons for it firing include:
  • that the GM862 reset which meant that when turning on it brought a GPIO high therefore triggering the cutdown.
  • Or there is an error in the cutdown code (more likely knowing me!)
• The camera + 555 timer worked well, once turned on it worked through out the flight. However the pictures are quite blurry suggesting a slow shutter speed. It mightbe due to the payload case covering the light sensor making it appear darker which is interpreted requiring a slower shutter speed - will require some tests.

• uALFAT-TF Data Logger - UART, I2c or SPI
• Cheap MicroSD cards