I built a drone with only one motor [video]

**TL;DR: A drone driven by a single motor that uses hinged blades for pitch and yaw control — mechanically extremely simple, yet it required custom firmware, an infrared lighthouse, and a homemade flash-imaging setup, along with a tremendous amount of software and testing work, before it could fly successfully.** ## A Drone That "Shouldn't Exist" A quadcopter needs four motors. A helicopter needs two motors plus multiple servos. Even Tom Stanton's swashplateless helicopter needs two motors. This drone, on the other hand, hovers and steers with just one motor and some electronics. The author's motivation was straightforward: fewer actuators means smaller, cheaper, easier-to-manufacture drones. But mechanical simplicity comes at a price — the software needed to make this thing fly is extraordinarily complex. To that end, he had to write custom firmware for the flight and motor controllers, build an infrared lighthouse so the drone could know its orientation, and even construct a simple high-speed camera from scratch. ## How a Single Motor Controls Both Pitch and Yaw ### Hinged Propeller Hub The secret lies in the hinged propeller hub. In early versions, these hinges were tilted at a **45° angle**. When the motor accelerated, the blades lagged behind the hub due to inertia — causing one blade to pitch up while the other pitched down, generating unequal lift on each side. By timing the accelerations correctly within each revolution, the drone could tilt in any direction — at least in theory. ### Replacing the Drone ESC with a Servo Controller The initial attempts failed. Drone ESCs estimate motor position using back-EMF, and acceleration pulses occurring more than 50 times per second caused the ESC to lose sync, because the violent torque changes made the back-EMF signal too noisy. Other hinged-blade drones work around this by using massive propellers — large blades spin much more slowly, so the problem doesn't manifest. But the author wanted a small, cheap drone, and a propeller as big as his forearm clearly wasn't going to work. The solution was to replace the drone ESC with a **Motius R4 servo controller**. Servo drives measure position directly via magnetic encoders, so rapid torque changes don't cause loss of sync. The result: a 6-inch propeller spinning at nearly **20,000 RPM**, with the motor accelerating and decelerating **300 times per second**. To the author's knowledge, this is the fastest and smallest hinged-blade propeller ever demonstrated on a drone — with over six times less blade area than the Pulsar drone. ### 2 Milliseconds of Latency Equals 20° at 20,000 RPM The high blade speed made timing a critical problem. In the original design, one MCU read the rotary encoder, combined the data with flight controller commands, and sent back control instructions — the entire loop taking about **2 milliseconds**. That sounds incredibly fast, especially in the mechanical world — but at 20,000 RPM, 2 milliseconds corresponds to the motor rotating a full **20°**. So the author rewrote the motor controller firmware to communicate directly with the flight controller and compute motor commands on board. Response became near-instantaneous, while also eliminating an extra microcontroller — saving both weight and cost. ## Hub Redesign ### Single Pivot and PPACF Printing After shrinking the propeller, the hub was still very prominent, so the author redesigned it to use only a single pivot, cutting the volume by more than half. He drilled and tapped threads into the motor, allowing the hub to screw directly into it, further reducing the profile height. This small part had to withstand considerable forces: a single flanged bolt held the blade in place, and it had to transmit all of the motor's torque and the propeller's thrust simultaneously. So the author printed the hub in **PPACF** — an engineering-grade filament with astonishing strength and rigidity — on the **H2C** sent by Bamboo Lab. He joked: "We've all seen PPACF bells, but have you ever seen a PPACF prop hub?" He doesn't believe this propeller could survive in any other material. The hub also packed in two ball bearings and a miniature thrust bearing to ensure the blade could spin freely under heavy loads. ### Restoring the Centrifugal Centering Force with 90° Counterweights The single-pivot design works on the same principle as the dual-pivot version — motor acceleration pitches one blade up and the other down. But there was a critical difference: **moving the hinge to the center eliminated the centrifugal force that normally pulls the blades back to neutral**. Without it, the blades would skew off-center at high speeds due to aerodynamic drag, causing severe vibration. The solution was to place mass at **90°** to the blades — the screws protruding from the hub serve exactly this purpose — passively generating a restoring force that keeps the blades neutral at high speeds. ## Becoming a Flyable Drone ### Anti-Torque Blades and Airframe The author printed the airframe, complete with a built-in battery compartment, and included anti-torque blades to counteract the propeller's rotational force. Typically, a tail rotor or servo-driven flaps would be used, but those would add extra actuators, weight, and complexity. The trade-off was that the airframe would rotate slowly and continuously — a problem to be dealt with later. ### 160 Hz Vibration Drowning Out the Accelerometer Slowly advancing the throttle while testing over a bed, the drone crashed almost immediately, and it sounded terrible. The accelerometer was overwhelmed by violent **160 Hz vibration**, making control impossible. The problem: high blade speeds meant tip velocities exceeding **500 km/h** — approaching half the speed of sound. The blades became almost a blur, and the author couldn't see what the propeller was doing at all. A high-speed camera could capture it, but that would cost thousands of dollars and require extensive lighting. ### DIY "Ultra-High-Speed" Imaging: Freezing Blades with a Flash The trick was that you don't need to record thousands of frames per second — you just need to **repeatedly capture one perfectly timed frame**: 1. Modify the motor controller to emit a pulse when the propeller reaches a known angle; 2. A microcontroller combines that pulse with the camera's shutter signal to trigger a flash at the precise moment; 3. The flash acts as an ultra-high-speed shutter — the pulse lasting only a few microseconds, freezing the 500 km/h blades into a sharp image. By varying the motor's roll command during each capture, the author was able to see the blades pitching back and forth for the first time. But the images also revealed an unexpected phenomenon: **at high speeds, the blades flapped and flexed under the enormous loads — and this was the root cause of the 160 Hz vibration**. Swapping in much stiffer blades made the vibration disappear. ## A Long Cycle of Test Flights and Crashes Flight testing was far from smooth: - "This thing is a death machine." - He checked the failsafe parameters for the first time. - "This thing isn't crash-proof. It just shattered." Over the course of several weeks, each test yielded a tiny improvement, followed by yet another crash. He broke propellers, motor controllers, receivers, and even cracked several TPU airframes — "come on, how does flexible TPU crack?" He even felt like the drone was homing in on his face like a heat-seeking missile. Once he attempted a hand launch — and, predictably, it went poorly. **Crashing every 3 minutes was simply not sustainable.** The solution was to build a test rig, weight it down with a delivery suit bag, and seriously tune the **P controller**. The drone finally began showing signs of stability. ## Solving Heading Drift with an Infrared Lighthouse The anti-torque blades could only slow the airframe's rotation — not stop it. Over time, the drone would lose track of which direction was forward and become uncontrollable. A compass is typically used to track heading, but with a large motor mere centimeters away, the magnetometer was completely unreliable. So the author built a lighthouse — or more accurately, a "reverse lighthouse," since it was the drone rotating rather than the beacon: - Salvaged an infrared emitter from his **motion tracker project** and aimed it at a corner of the room, creating a fixed infrared beacon; - Added an infrared sensor to the drone; - Wrote custom flight controller firmware that fed each detection into an **attitude estimator**, correcting the orientation estimate every time the lighthouse came into view and providing an absolute reference. Finally, the drone could take off, stay level, and respond to controls for the first time. The entire setup: one motor, one 6-inch propeller, and some electronics. ## Conclusions and Next Steps The author readily acknowledges that the mechanical simplicity concealed an enormous software effort to get this thing flying. To his knowledge, this is the **smallest and highest-RPM hinged-blade propeller** ever demonstrated on a flyable drone. He's also glad he stuck with the small scale — "If this thing had a 15-inch propeller, I probably wouldn't be alive right now." The project is far from over: the drone still depends on the infrared lighthouse, meaning it can't fly outside his bedroom. The next video plans to install a **line-scan camera sensor** that sweeps the surroundings while rotating, using that information to maintain heading and enable autonomous navigation. The code and 3D files are all freely open-source, with a repository link below the video. Source: I built a drone with only one motor — hakonjdjohnsen (https://www.youtube.com/watch?v=Ndftcft6XhU)