\n
![\"A120_cropped\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/A120_cropped-1024x692.png\")
<\/a><\/p>\nFor my fourth Raspberry PI weather balloon flight I wanted to pack as much into the flight as I could, making full use of what the Pi has to offer:<\/p>\n
- \n
- 3G link for video and backup telemetry<\/li>\n
- Send telemetry over 3G from the launch site<\/li>\n
- A live video stream from payload itself, with telemetry overlay, before and for the first few seconds of launch site, over 3G<\/li>\n
- Radio telemetry and SSDV images as usual<\/li>\n
- Transmit larger photos than before, using twice the transmission speed<\/li>\n
- Switch to taking and sending even larger photos when above 30km<\/li>\n
- Take full-size images for storage on the SD card<\/li>\n
- 3G to automatically reconnect on landing<\/li>\n
- Send the landing position over 3G<\/li>\n
- Live video stream from the landing site (hopefully in the middle of an empty field)<\/li>\n
- Upload all images to a web server by ftp after landing<\/li>\n<\/ul>\n
That’s quite a lot more than before, and with all the overall control software written the day before launch I wouldn’t have been at all surprised to see things go wrong! In the event, the only part that failed was the ftp upload (the very last thing I added).<\/p>\n
With more electronics than usual to pack inside the payload, what better choice than a TARDIS? Sadly I must have missed my university class on transdimensional engineering, so I had to build a fairly large TARDIS to fit everything inside. Actually the real reason for choosing the TARDIS over say an Apollo Command Module (that’s for next time) was that it’s easy to make from flat sheets of extruded polystyrene! Unsurprisingly, the internet is full of plans and diagrams so it didn’t take long to find what I needed, including the vital fact that the correct colour paint is Prussian Blue :-). A flashing white LED completed the construction.<\/p>\n
![\"tardis\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/tardis-211x300.jpg\")
<\/a><\/p>\nFor the inside I built mounted the electronics either side of corrugated plastic sheet. One side carried the Pi itself (a model B – I needed both USB ports), plus switched-mode regulators for the 3.3V and 5V lines, and a USB 3G dongle for the live video stream and backup telemetry:<\/p>\n
![\"1\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/1-714x1024.jpg\")
<\/a><\/p>\nOn the other side went the 6 Lithium AA cells plus a custom board carrying the NTX2 radio transmitter, UBlox GPS receiver, BMP085 pressure\/temperature sensor and an A\/D to monitor the battery voltage. A second temperature sensor was added outside the base.<\/p>\n
![\"2\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/2-699x1024.jpg\")
<\/a><\/p>\nLast but not least I needed a camera to photograph the TARDIS, and again I used a Logitech C270 webcam this time mounted in a fluorescent pink ball (easy to spot at distance after it lands!). You can see that mounted precariously at the back of this image:<\/p>\n
![\"3\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/3-682x1024.jpg\")
<\/a><\/p>\nThe launch day was overcast and cold, but my main concern was the flight path. Too far west and the landing spot would be near Bath; too far east and it would be near Salisbury Plain (where men play with their toys). So I planned the ascent rate\/\/descent rate and burst altitude to land between those extremes. Anthony Stirk<\/a> (who helped out on my first Pi launch) came along to help, and to launch his own balloon too which he successfully got to float and was last seen flying over Germany. Also Alex Eames of http:\/\/raspi.tv\/<\/a> came to record the launch on video. We launched Anthony’s flight first and then the TARDIS:<\/p>\n
![\"4\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/4-748x1024.jpg\")
<\/a><\/p>\nThe live video streaming from the payload worked well for the launch, though it soon hit the low cloud and then the 3G dropped out (as expected). At 4km the flight software stopped trying to connect through 3G, closed the streaming program, and started to take still images which were then transmitted over the radio link. I’d tested all this stuff with a pretend flight but it was great to see it actually work for real! I’d chosen to transmit images and telemetry at 600 baud, which may sound slow but it’s twice the rate normally used for images and 12 times the normal telemetry speed, and I was concerned that we’d lose lots of image data. However we lost very little – more than on a 300 baud flight, but not enormously so especially considering it was a midweek launch with fewer listeners than normal. here’s one of the images it sent:<\/p>\n
![\"snap84\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/snap84.jpg\")
<\/a><\/p>\nWith the launch kit packed back in the car, we stopped by at home so I could tune my receiver there in to help with the downloading, then we set off in the general direction of where the flight was supposed to land. We took my 4×4 which has a car PC installed with a touchscreen on the dash (top of the pic below), plus a Google Nexus 7 mounted below for the map. Anthony had his netbook plus mine to use.<\/p>\n
![\"IMG_1483\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/IMG_1483-1024x768.jpg\")
<\/a><\/p>\nWe took the wrong road out of Marlborough which did slow us a little but the flight path meant we had enough time. Once the balloon burst we could see that the landing point was going to be near Melsham, so we headed there, and then on to Whitley as the landing prediction homed in. After a short stop for fuel, and a shorter stop for another purpose (the garage didn’t have a toilet!) we parked up by the main road expecting the payload to land about 300 metres away. In the event it was about 1.3km away, so we didn’t see it. The radio signal dropped out on landing, and the last position we had was for it 200 metres above the ground. We jumped back in the car and headed there, and very soon the map updated with the actual landing position, sent over the GSM link. About that time the live video link burst back into life too! Here’s a still from that:<\/p>\n
![\"landed\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/landed.png\")
<\/a><\/p>\nWe parked up about 200 metres from the payload. I entered the co-ords into my Android phone and we followed the directions from that, but pretty soon Anthony spotted the flight lying in the field:<\/p>\n
![\"IMG_1488\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/IMG_1488-1024x768.jpg\")
<\/a><\/p>\nHere’s the flight path to the landing spot (top-right downwards) and the path we took carrying the payload back to the car. Note the HV electricity lines! By my calculation the TARDIS was 37 metres above the ground when it crossed them ….<\/p>\n
![\"3\"](\"http:\/\/www.daveakerman.com\/wp-content\/uploads\/2013\/03\/3-1024x756.png\")
<\/a><\/p>\nAnd here’s the video that the Pi streamed live as Anthony and I recovered it!<\/p>\n