In the summer of 2014, I decided I wanted to build a drone. I chose for a quadcopter frame layout, which means the craft would have four motor/propeller pairs. I didn’t know much more about quadcopters at that point. After I had done enough research, I settled on a list of components I would need to put together the quadcopter. For a beginner, these were good components and the whole set fit my price range. All parts ended up costing me €400, including the RC remote.
After a few days, a package containing the majority of the components arrived. Within were mostly smaller boxes, wiring and electronics.
The battery has 3300 mAh capacity (14,8 V), which should be enough to keep the craft in the air for 15 to 20 minutes. It had a 4mm bullet connector soldered to it when I arrived, but I changed it for an XT-60 connector because that was my connector of choice for my main power wiring.
The selected motors are NTM 28-26’s with a maximum power of 280 W. This should be more than enough for my quadcopter and flight purposes. I will install 10″ propellers on these motors.
Instead of beginning with a simple flight controller, I chose to use one with GPS support. This is beneficial because it allows autonomous flying and landing, in case I lose control during flight.
The first thing to do: put the frame together. It consists of four aluminium tubes with several fibreglass parts cut into different shapes. Additionally, it comes with several bags of nuts and bolts to hold everything together.
The flight controller is mounted on the top centerplate using nylon spacers. This way, the board does not touch the plate itself, which reduces sensor vibration interference.
Because I’m building the drone in X-configuration (as opposed to a + configuration), I’m using two black and two silver-coloured arms. By mounting the silver arms at the front, I’ll be able to see which side is the front while the quadcopter is in the air.
The legs assemblies consist of two fibreglass plates with spacers inbetween. They are made in such a way that they can rotate relative to the frame tubes. This helps reduce frame stressing during rough landings.
A leg is then mounted at each tube end.
To let the legs rotate, one end is secured to the arm using a bolt, while the other end is held on by a spring. At landing, the spring will be able to stretch, thus absorbing the kinetic energy while landing.
After finishing the arm assemblies, they were ready to be attached to the centerplate. Each arm is secured to these middles plates using two bolts. The frame is now finished.
Next up are the motors. Each motor is mounted with the wires facing to the side, because frame bolts prevent the wires from being oriented towards the center. This does not cause any further problems.
The four speed controllers (ESCs, one for each motor) did not have connectors soldered to them. I soldered on 3mm bullet connectors and sealed the connection using heatshrink tubing. This will prevent short circuits.
Instead of mounting the speed controllers on the arms of the frame, I chose to put them in the frame. In my opinion, this looks better, but space is limited between these centerplates. This makes things a bit more difficult. The wires coming out of the top plate will later be connected to the flight controller.
The speed controller output wires weren’t long enough to reach the motors. To solve this, I had to make 12 extension wires to bridge the gap. Above, a photo is shown where heat shrink is not yet applied to these extension wires.
After installing the wires, I secured them to the frame using tie wraps to prevent them from interfering with the propellers and other hardware during flight.
Back to the flight controller. On the left, the power supply is plugged in to the board. The supply itself is mounted to the right of the bottom-left speed controller. On the right side of the flight controller, the four speed controllers are wires are plugged in.
Additional electronics were then added, including the GPS and receiver, respectively below and above the flight controller. After installing these components, I plugged in a USB cable to set up the board and install the MultiWii software.
After checking the motor rotation direction and calibrating the speed controllers, I mounted the propellers onto the motors.
The battery is placed under the frame and is secured using a simple velcro strap. There’s a battery alarm on the battery which starts making noise when the voltage drops below a certain treshold.
The quadcopter is finished – for now. On the right, the charger is hooked up to charge up the battery for the quadcopter’s first flight.
Things weren’t that easy. During the first flight, I quickly found out that my gyroscope sensor was broken. Additionally, one of the motors made a weird sound and got very hot. I had to replace both the flight controller and the affected motor. This time, I chose for an APM 2.6 flight controller. The company where I initially ordered my parts was nice enough to refund me the costs of the faulty parts.
While waiting for my replacement parts, I also ordered a new LiPo charger since my old one was quite lacking. This one is much better and does exactly what it should do.
Some time passed, and the new flight controller arrived. This one also has its own GPS module, so the old one had to come off. Because this controller has no mounting points that are compatible with my frame, I cut my own mounting plate from balsa wood. I stuck the controller to this plate using double-sided tape.
After another test flight, everything seemed fine, which concludes the build log for now.
After a few months and lots of flight attempts, not much was left of the original aluminium/fibreglass frame, in the sense that arms were bent and some parts were broken. This might have been a result of my flying skills. I bought a new frame, an F450 with plastic arms that should be a bit more durable. This turned out to be a better frame in the end, as it has more room to accommodate electronics. It also seems to be easier to fly, stability-wise.