Photo by flberger - Source
These flying machines are becoming more and more popular. They are being built by a huge range of people, from hobbyists to professionals and even military contractors. For my final year project at Uni, I was going to build a remote control car, giving it extras such as 'traction control' and possibly going to the extent of giving it the ability to follow a list of way-points - I ended up doing a much better final year project and sadly didn't get anything done with regard to the RC car. After finding out about Quad Copters, and giving it some thought, I decided to change my target.
My new home project is to build as much of a quad copter as possible. By this I mean designing, building and developing the speed controllers (ESCs) and the flight control board as well as the software that runs on both of these systems (and possibly more! - flight frame etc...). It's going to be a huge learning exercise, getting the brush-less motor controller operational is step one, however that is significantly easier than the maths that will be involved in the flight control systems.
Multi Rotor Theory
Let's start with single rotor aircraft - Helicopters. They are everywhere, very well understood, and are easily recognised with their single, large rotor, on the top, and their smaller rotor on the tail. As the top rotor turns, it pushes against the air it is passing through. When on the ground this isn't an issue as the friction between the ground and the landing gear keeps the airframe pointing straight ahead. Once airborne however, the airframe has nothing to grip on to, or to hold it facing forwards. As a result, the airframe wants to start to rotate - in the opposite direction to the rotor blades (this is why in films, when the tail rotor gets damaged, the helicopter starts to rotate uncontrollably). This effect is called torque reaction. To prevent this, the pilot has foot controls to adjust the amount of lateral force that is given by the tail rotor. This is a very brief explanation of helicopter operation, and most of the more complex principals and techniques will not be used on my quad copter - at least not for now! (hint) If you would like a more in-depth explanation of helicopter operation, this is a good website.
Now that we understand the reason that a helicopter has a tail rotor, we can hopefully understand why a tandem rotor aircraft like the Boeing CH-47 Chinook doesn't have one. It's because the rotors spin in opposite directions. This causes the rotational forces induced by each rotor to cancel each other out - resulting in zero rotational force. If we then consider a quad-, hex-, octo-copter (the list goes on) we can infer that the forces generated by clockwise rotating rotors, must be matched by the forces generated by the counter-clockwise rotating rotors. This even holds true for tri-copters, though we'll leave those nasty brain-aches there for now.
A quad copter in plus '+' configuration - Source
So, we now have a quad-rotor aircraft that can hover without spinning - fantastic!.. Or at least, that's a start... Surely we want to be able to be able to control the speed of rotation (the yaw) so that we can fly around? Well this isn't actually a tricky problem to overcome. If you wish to rotate clockwise, then you slow the rotors that are spinning clockwise (clockwise rotors create a counter-clockwise force). And the same is true for the inverse, to rotate counter-clockwise, you slow the counter-clockwise rotors.
Magic! Almost... What happens if you want to travel forward, or sideways? Infact, what happens if you start to fall to one direction, and need to level the aircraft out? In this situation you just control the upthrust provided by the relevant rotors. For example, if you want to travel forward, you slow the front rotor, and speed up the rear rotor (assuming a plus '+' configuration for simplicity). This has the effect of tilting the aircraft forward, which then causes the upthrust to be vectored forward as well. From here, it is easy to understand how to control the pitch and roll of the aircraft.
So there you have it, multi rotor theory in a 4 paragraph nutshell.
I always find it interesting to ask the question "Has anyone else done this before?" Of course, in this case the answer is very definitely yes! But putting that to one side, when do you thing the first quad copter was built? Now remember people like da Vinci, I bet your date has changed... Unfortunately I've not found evidence of multi rotor aircraft quite that far back, but a cursory look brought me to a man named Etienne Oemichen. He worked for Peugeot, and apparently began working (or playing) with rotating wing aircraft in around 1920. I'm not going to go on about it (I hated history lessons as a child), but for those interested, this page makes a good read.
I am approaching this project as a massive learning exercise. I'm hoping to build as much of the aircraft as possible (in an electrical sense).
Electronic Speed Controllers
Initially, I plan to build the Electronic Speed Controllers (ESCs) and along the way I will gain an in depth understanding of Brushless DC Motors, PCB design, soldering surface mount components, high power electronics, and tune up my skills with hard real-time systems. This component also has potential to introduce an amount of analogue electronics with regard to sensing the back Electro-Motive Force (EMF) which then allows correct commutation.
Read more about the ESC here.
The next step will be to develop a control board which will allow me to practice my new-found PCB design skills in a more complex environment. This will initially have sensors such as 3-axis accelerometers, and 3-axis gyros, along with a simple CPU and communications to each of the speed controllers. The major challenge for me in this part will be the maths involved. The values read from the sensors need to be pushed through some pure magic in order to get speed values for the motors. I will initially work to get a see-saw style arm balancing, using a motor and propeller at each end. This will hopefully then progress to achieving 4-motor stabilisation, and once that is complete, I plan to implement some form of radio control link, so that the aircraft can be controlled.
Once I have a basic quad copter flying and operating correctly, I plan to take things to the next level. This will quite possibly involve some form of automation, for example, following a set of way-points. This section is far away, and because of that I need to forget about it - for now.
I plan to keep this section as a sort of diary, noting down important events, observations, lessons... that sort of thing. Because of this, the content for this section will be on a separate page.