My latest project has been to build my own Quadcopter.
I spent a while researching in late 2015, got some parts for Christmas, and have been building since. As you can see from the costings graph above, my main expense was the RC transmitter, as I decided to go for one that supports telemetry. For a flight controller, 
the Arducopter project appealed to me as it uses completely open source software and hardware, so I can customise/hack any part of it I like. I chose an S500 frame as it’s suitable for aerial photography but not too large, and it comes with a PDB (Power Distribution Board), which makes cable management a lot easier.
It had its maiden flight today, and now that it’s working in its most basic form I’m looking forward to adding a camera, and having a fiddle with the source code 🙂
Ben James 
This is awesome!
This is super amazing! 😀
I have also been looking to build a quadcopter from scratch, and I’ve been doing research. I want to make a small prototype to understand how they work before I build anything larger, and I have four tiny motors (around 2cm long) from an old RC helicopter, as well as some small LiPo batteries. Based on your experience, how feasible do you think it would be to build a tiny quadcopter controlled over the local network with Raspberry Pi or Arduino without any motor or control board? Do you think it would be challenging to write software from scratch?
Hi Amanda,
That sounds like a fun project, and I know you’re very capable when armed with an Arduino/Pi 🙂
Here are some things that spring to mind which might prove difficult:
1) Weight. The smaller the quad, the less margin for error you have to play with in terms of weight. If you look at the tiny consumer mini quads, the CX-10 by Cheerson has a weight of roughly 12 grams, whilst the Husban X4 (about double the size of the CX-10) weighs in at about 50. I’m guessing that’s the kind of size you’d be looking at building, so it might be worth seeing how much your motors, battery, accelerometer, chosen chip etc weigh.
2) Motor control. I believe you basically have two options: hook them up to a transistor controlled by an Arduino/Pi pin, then PWM it (primitive and not terribly smooth), or buy a control board (weight 😦 ).
3) Software. If you have a lot of time to dedicate to this project, I’m sure you could produce some decent software, especially given the fantastic example of the APM code to refer to (http://dev.ardupilot.com/wiki/learning-the-ardupilot-codebase/), but it would take a lot of coding.
Hope this helps 🙂
Thanks, Ben, for the tips! 🙂 I hadn’t thought too much about the weight, but I definitely need to see if the motors I have are going to be able to generate enough lift to actually get it off the ground. If I can get one, I was thinking the Pi Zero would be great to use because it’ll reduce the weight significantly.