Multicopters: Overview

Around 2014, “drones,” quadcopters in particular, were becoming popular.

Commercially-available options were either out of my budget, lacking in features, or had unimpressive specifications. I chose to build my own, and this started a multi-year revision process.

My objectives were to build a drone with first-person live video feedback, an onboard, 2-axis stabilized GoPro for HD video recording, sufficient battery life, and an industrial design philosophy that emphasized ruggedness.

This page details the various revisions I made over time, to achieve those goals.

While sturdy, this revision weighed too much. Flight times were on the order of 12 minutes.

The drone in the above video is revision 2. It uses 0.070″ thick aluminum top and bottom plates that I machined myself, along with 0.500″ carbon fiber tubes and fiberglass landing rails. Most other parts are off-the-shelf.

Minor changes occurred over time, before I eventually settled on a printed ABS design to save on weight (to the benefit of flight endurance).

Rendering in Fusion360. Both sides are held together with nylon 6-6 machine screws to save on weight, compressing the carbon fiber tubes in place.
CAD model used to find a propeller spacing that wouldn’t enter the GoPro’s FOV, without negatively affecting MOI and COG

I solvent-welded a battery compartment to the bottom, and it flew well:

Revision 3/4: FPV feedback, GPS, with on-screen heads-up-display showing key telemetry data. Slower brushless motors with larger propellors.
Flight of the Clamshell Design:

Not long after the last video, I lost the drone during a crash. The clamshell assembly was too weak. The aerodynamic performance was better than expected; the low MOI (thanks to most of the mass being highly concentrated about the COG) meant that it was very responsive to inputs.

While I still have confidence in the design, I chose to pursue fixed-wing UAV’s after this crash, gutting this drone and utilizing the parts on the next build.