Spacebits 6 took place on 26 of June at Serra da Estrela, the highest mountain in Portugal. The spot was chosen after a presentation on TEDxCovilhã. It has been almost 9 months after our last flight and there was a nostalgia feeling in the air …
We arrived Friday at the Serra da Estrela chalets and started to unpack and prepare the Spacebits payload. In the past, we had done five flights, and recovered all of them, but we never had a perfect flight. Between software bugs (sorry!!!), cameras that stopped working and many other small problems, we collected a list of DOs & DON‘Ts. With two days there was enough time to solve them and to improve the design. We worked hard to cover and improve every failure scenario that we thought but it seems it is not enough … Oh, boy, you can’t get paranoid enough!!!
With so much effort and value put in the project, we had designed Spacebits keeping in mind that if anything happens to the main board we would be releasing a few thousand euros into the wild. So worked has started to design a failsafe module. It has been proven invaluable and again, Fernado Afonso and Filipe Varela started to check the module and test it thoroughly.
Taking pictures of the earth was the major goal, but we wanted also to track the flight on real-time. Celso Martinho, our fearless leader, developed a web application based on SAPO Maps that could show the position and also some sensor information on the web. We had test it already with success but we also add a few tweaks.
There was also applets for the iPhone and iPad developed by Filipe Varela. This is handy but you need to have 3G coverage, which is not always the case. Nevertheless, it is an invaluable add-on.
There where also some work to do on the main board, where data is collected, recorded and sent to our base station. The external temperature sensor had to be repacked and all sensors were carefully tested. With the rest of the team happy to work inside, Filipe Valpereiro headed into the sun, testing the sensors and checking the GPS signal. After a few minutes in the sun his monitor reported a 48º measured with the external sensor. It was really hot …
Celso invited everyone to join in our mailing list to improve the project. There where some requests in the mailing list for an Android version and Miguel Duarte coded one just in time for Spacebits #6.
It was also a great opportunity to improve our ground station and Delfim Machado worked on a app to receive data from the HAB and stream it to our server. In the past we had a laptop with a 3G connection to stream data into the web. But there was not always 3G signal while driving into the country.
This time we had the perfect solution. Serra da Estrela is the highest place in Portugal and we could just leave an antenna there, pointing into the direction where we expected the ballon to fall. In a few hours, Delfim and Celso managed to connect a XBee radio to an Android Sapo A5. All pieces fitted perfectly in the plastic enclosure that Celso bought the day before. Late at night everything was ready, just in time to another TO-DO item: get some good sleep. We learned that the hard way!!!
Sunday started shinny and there was another TO-DO item: ask the caffe manager to allow us to place our secret ground station and YAGI antenna in the terrace. We had line of sight from the terrace into the spot where the ballon was predicted to fall, near Castro Daire. We turned the main board and immediately everyone received some SMS with the ballon coordinates.
Filipe Valpereiro configured the radios to work in broadcast and Delfim Machado assembled another Android Sapo A5 ground station to go in the car. Just in case someone might try to mess with our YAGI antenna and left us in the dark for the next few hours … At this moment the expectations where why, and we started to gather around the cars to unpack the material and prepare the balllon.
We had rented a 7 person van, filled with hand bags, computers, some food and 9000L of compressed Helium. Not the kind of luggage that you want to carry around in a car. Living dangerously
Celso instructed all the attendees to took some care before handling the ballon. There was a lot of wind that day, as expected and we need a lot of hands to help us secure the 3 meters wide ballon until it’s full. It’s not easy to handle in mild winds, and everyone took extra precautions not to wear any sharp object, wedding rings included!!!
After some ground work we started to pump Helium into the ballon, very slowly. It takes about an hour to fill the ballon. If you go to fast, there’s a chance to release it at cold temperatures and the latex rubber can froze. There was some wind, but we got plenty of time, and patience, more or less
We have enough hands to help.
Meanwhile, Fernando and Filipe inspected the payload and added the Reward-if-Found sign. As silly as it sounds, it helped us to recover our payload in the second flight, found in a beach near Cádiz, one week after a our failure. Name, phone, email and skype should be enough
It was time to turn on the GoPro camera. Very light, easy to operate, water prof enclosure available, and can shot 1080p. This is a must piece, i only wish they could have a model with swappable lens!!! We had no time to fully dehydrate the camera and the enclosure. This turned to be a problem latter on. During the flight, the camera crosses colder layers of the atmosphere and the lens get some internal condensation. Not critical, but some of footage got blurred.
With the ballon filled, it was time to measure the lift force. Fernando crafted a weight mechanism with a dynamometer attached to a wood board and some solid concrete weights. With the ballon secured to a nylon rope it was easy to measure the lift force.
It’s a simple and elegant solution, release the attached ballon and with a simple dynamometer you can measure the lift force. We got a few measurements but we pumped a few more liters of Helium, to account for the wind force. Better safe then sorry
Countdown: 3, 2, 1 … GO
In a few seconds the ballon reached hundreds of meters above our head. The hunt had begin, we packed everything, and rushed to the car, to get some data from our Android based ground station.
Unfortunately we had no data. No radio signal, no GPS, nothing. The ballon climbed silently, and we pointed the antenna hopping to get some data, but it was hopeless.
We suspected that the main board switch had been turned off, maybe by a near cable when closing the box. In the rush, no one remembered to test the radio communication. A simple procedure, that could save us from failure. The flight predictor pointed Castro Daire as the most probable landing spot, and the team moved on.
At this moment we hoped to get the ballon landing in a GSM covered area. The failsafe module is programmed to send SMS with the last know coordinates. It helped us to recover some of the balloons, and we hopped the module would work again, but things weren’t prospective, the payload could land in an area with no GSM coverage.
After a few hours driving and burning the car tiers, we went to Castro Daire. Our flight simulations showed this to be the most likely spot. We stopped for some quick snack and a brief meeting waiting for some message from the failsafe module. Lunch was abruptly interrupted when our phones started to ring, Spacebits #6 had landed.
In less then one hour we recovered the ballon. We discovered latter that hour that the Google Maps API rounded the position to the nearest route. While useful driving a car, it’s not something you wanted to happen in the wild, and this sort of “bug” lefts us wandering around, looking for the ballon. Luckily, Rafael and his wife had a Garmin GPS, and they were able to locate it easily.
After opening the box we found the problem. There was a Killer-Switch and one of the cable add turned it off. So much for failure scenarios and hard work. Shortly after being turned on, the radio started to transmit, and our ground station at Serra da Estrela (80Km distance) started to receive some data. In a few seconds we where able to see data in the Spacebits website.
Time for a group photo and prepare to return home.
Based on the ascent velocity and the amount of video recorded we estimate that the ballon reached 38Km high. At this high, you can start to see the round shape of the Earth, which appears more dramatic with the GoPro fish eye lens.
You can also see part of the parachute in the left corner, moments before opening. It takes approximately 30 minutes for the payload to return into the Earth.
It was a great weekend, full of surprises, work and fun. Above all, it is a great way to share knowledge, experience real life problems and meet new people. Thank’s to everyone who show up in Serra da Estrela.
And if you want to see more, please enjoy the video bellow.
Categories : Arduino, Collaborations, Hacks, Physical Computing