Time-lapse of our 70mins long flight
This was our final year project undertaken in a group of three i.e. Srijal Poojari, Adithya Ramanujam and myself. We worked in collaboration with Drishti Works a startup based in Mumbai, developing robots for waste management and agricultural uses.
The drawback of traditional drone systems is the limited flight time due to the battery capacity of the drone. Our aim is to develop a tethered multirotor with a very long flight time
The major issue with building a tethered drone is that the drone requires very high current (200A-300A). Any cable which is capable of transmitting this current without a significant voltage drop will be very thick and hence very heavy. This will limit the maximum flight altitude of the drone.
To solve this we designed a high voltage DC (HVDC) transmission system which will step-up the transmission voltage and hence reduce the current carried by the tether; in turn, reducing the weight of the tether which will allow the drone to go up to a higher altitude.
As a part of this system, we developed a step-down module which will be on board the drone. This will step-down the voltage from 130V DC to 32V DC @20 Amps. Unfortunately our first iteration of the board did not work and blew up as soon as it was turned on. This was caused due to improper layout of PCB. As it was a high frequency, high voltage switching supply, special care has to be taken during the layout of the PCB. We took the learning from this design and designed a second version of the board. The board worked this time around. Although we were not able to test the board to the full 32A output current as the mosfets were improperly selected which restricted the cooling we could supply to the mosfets and hence it restricted the output power. We should have selected mosfets in a package which would have been more suitable for mounting a heat sink, say a to-220 package. It would be a less power efficient design than the current design but it will enable us to supply the board with the necessary cooling it requires.
We performed a successful 70 min continuous and semi-autonomous flight at the end of semester VII.
