Friends, it's time to sum up the results of our project-competition "Server in the Clouds". If someone is not in the know, we started a fun geek project: we made a small server on Raspberry Pi 3, attached a GPS tracker and sensors to it, loaded all this stuff into a balloon and trusted the forces of nature. Where the balloon will land is known only to the gods of the winds and the patrons of aeronautics, so we invited everyone to put points on the map - whose points are closest to the actual landing site, receive "tasty" prizes.
So, our server has already flown to the clouds, and it's time to sum up our competition.
Links to previous publications about the competition
- (a prize for the first place in our competition is participation in a sailing regatta, which will be held from November 3 to November 10 in the Saronic Gulf (Greece) together with the RUVDS team and Habr.
- How did we do"Β» of the project - for lovers of geek porn, with details and analysis of the code.
- with full description.
- , on which it was possible to monitor the movement of the ball and telemetry in real time.
- from the launch site.
And experience, son of difficult mistakes
As you remember, we planned to broadcast data from the server via a GSM modem. It was the main channel for transmitting information. It seemed to us that we had foreseen any surprises with the coverage of the cellular network by inserting two SIM-cards of operators with the best coverage in the Dmitrovsky district into the modem. In addition, the modem had a good omnidirectional antenna. But, as they say, a person proposes, and opsos dispose. When the balloon rose above 500 meters (the height of the Ostankino television tower), the cellular connection disappeared completely.
In hindsight, this seems obvious, but that's what it's like in hindsight. Of course, cellular antennas are designed for coverage on the ground, not in the air. Their radiation patterns βbeatβ along the terrain and do not βshineβ into the clouds. So cellular communication at a height of half a kilometer and above is just an accidental reflection of the lobe of some antenna. So there was no communication with the ball via a cellular channel for half of the route. And on the descent, when we went down below 500 meters, the cellular connection started working again.
How did we get telemetry from the ball? For this, thanks to the duplicating data transmission channel. We installed a kit on the balloperating at 433 MHz.
Its throughput is small, but for our tasks it was quite enough. As for determining the location of the ball using GPS, there were no problems with this, the tracker worked without hesitation.
And in flight, it turned out that the USB cable connecting the telemetry module to the Raspberry Pi 3 turned out to be defective. He worked on the ground, but refused in the sky. Probably afraid of heights. We found out the fault of the cable after landing. Fortunately, we managed to set up data transfer directly from the telemetry module via LoRa.
And about good
Luck smiled at the habrausers @ (first place), @ (second place) and @ (third place)! The most fortunate in November will have a lot of impressions (hopefully pleasant) from, and we will soon present good smartphones to the winners of the second and third places. And of course, all three of them will get a free virtual server rental from RUVDS as a gift.
You can see how the launch went in this short video:
Source: habr.com