Building our own Shark Tank Drone
A perfect excuse for me to head into the shed. The sharks loved the XCraft, XPlus One so much I thought I should have a go. There are going to be plenty of folks that just want to buy off the shelf, and many more that perhaps want to increase the size of the platform and would like to roll their own.
Let’s park the legal challenges facing the sharks now the FAA is getting interested in all classes of model aircraft.
Four years ago I played with VTOLs having been inspired by the Japanese ball. Since then autopilots have become cheaper faster and easier to setup.
There are many many tail sitting projects from home brew to military backed. Only the military ones have really had the cash to solve the tail sitting elephant in the room. The transition from hover to forward flight. Put simply the autopilot is fixed and uses that fixed point as its level reference. As you tilt over it’s not sure what’s what. That is the part really clever folks are required at and was previously above my pay grade.
For the tail sitter and ball I used a copter control board from the OpenPilot project they worked but the transition was something my thumbs handled.
We also started building a bicopter tail sitter (two motors, but life took over. I don’t actually fly anywhere near as much as I used to. That needs to change!
They have come up with Open Source code that will allow the airframe to take off, fly a route and then come home for an autonomous landing.
I should quickly circle back on the why bother with blending a wing and rotors and the answer is quite simply efficiency. A tail sitter like Google X’s first delivery drone prototype can lift more and fly further than a conventional quad or octocopter.
If you have been paying attention though you will have noticed Google has moved away from a tail sitter to a quad/aircraft hybrid. This I think is because the first one presented such an area to the wind that it had trouble keeping a stable or perhaps I should say accurate hover hold in any breeze.
The shark tank drone is going to suffer from this a little as well.
To a small degree Google X seeks to mitigate that problem by dropping the package down on a string. So not actually landing a point where significant drift could cause the airframe to tip over.
Back in 2013/2014 Master’s students Cyriel Notteboom, Menno Hochstenbach and Maarten Verbandt worked on VertiKUL at KU Leuven university in Belguim. For me they have come up with the most elegant four motor wing solution. That old aviation saw, if it looks right it should fly right applies.
I like it for several reasons, it has a purposeful sized wing. By that I mean it’s not just an equally sized X with symmetrical aerofoils. It is made to fly one way up and the longer wings will make it more efficient and easier to see which way it’s going.
A small detail the speed controllers for the motors are exposed in cooling air. They are also individual units so they can suffer a single failure. Other projects I have seen use 4 in 1 speed controllers, these are easy to use because they are a single unit to control four motors. If you have a failure in one though it might take out the others.
By using 4 motors you are able to hover with more weight on board and remove traditional moving flying control surfaces. Quite a saving in the complexity and parts count.
Roll pitch and yaw are handled by the motors. Notteboom, Hochstenbach and Verbandt like Xaircraft use a Pixhawk for their flight controller.
I have a Pixhawk without an airframe to call home and some foam that escaped from insulating a wall. If I wanted a simple platform without autonomous flight I would have used a KK 2 board and OpenAero VTOL
What I need to find is 4 matching or close to matching motors. Best I start turning the house over and find some then get the pencil out and get designing.
If it flies I will share the plans.
