What is it?
Gravity Fall is a gravitational energy storage system — a mechanical battery. A 15.65 kg mass is lifted to 1.80 m and released. As it descends through a 1:9 chain drive, it spins a multi-pole hub motor. A magnetic braking system controls the descent, an MPPT charge controller regulates the output, and the energy is stored in a 12V LiFePO4 battery. A DC-DC buck converter provides 5V USB output.
Why did we build it?
Solar and wind depend on weather. Lithium batteries degrade, require rare minerals, and produce toxic waste. Gravity is free, constant, and available everywhere. We wanted to test whether gravitational potential energy could be reliably converted to electricity using accessible components — and measure exactly how efficient the conversion is.
Key results
- 13W peak electrical output at 1.02 N·m torque
- 58% system efficiency (gravitational PE → electrical energy)
- Angular velocity stabilized at maximum load, varying only 0.01 rad/s — the magnetic brake works
- Stable 9.5-second descent time across all configurations, validated by MPPT algorithm
- Non-linear power scaling: 56% mass increase produced 155% power increase
Honest limitations
This is a proof of concept. Charging an iPhone 16 requires 394 descent cycles. The prototype cost $15,000–$20,000 MXN (~$860–$1,150 USD) due to professional fabrication, though we estimate a DIY replication at ~$7,000 MXN (~$400 USD).
What's next?
- Reducing the 41 cm dead space in the 1.80 m structure (30% more gravitational potential)
- Replacing chains with timing belts to cut the 10% frictional loss
- Automating the mass return with a winch system
- Testing with heavier masses to map the efficiency curve further
Open source
All FreeCAD design files, experimental data (xlsx), photos, and a complete build guide are available on GitHub. MIT License.
Step by step and entire list of materials/components also on Instructables: https://www.instructables.com/Gravity-Fall-a-Mechanical-Battery-That-Converts-Gr/.
valeriamayara22
This is our first open-source project and we'd love to hear from this community. If you have questions, suggestions, or ideas on how to improve the system — especially on reducing friction losses or automating the mass return — we're all ears. Any feedback is appreciated.