Building Gearbox Dynamometer

Summary

I built this custom gearbox dynamometer (aka dyno) to test marine transmissions I design and build from the ground up. Typically, my transmissions serve as a lower-unit gearbox for hydrofoiling electric boats. As a consulting mechanical engineer in the electric marine industry, I needed a way to run up to 1000 hours of durability testing on the transmissions I develop. But with dyno labs charging $100+ per hour, outsourcing was not an option for me. The solution? Build my own AC Dynamometer. What started as a necessity quickly turned into a fully integrated, data-driven test rig, complete with custom torque measurement, water cooling, CAN bus integration, and a Python-based dashboard. This project became a thesis-level deep dive into mechanical testing. From designing the 8020 aluminum frame, tuning motor controllers, to calibrating strain gauges, I picked up what felt like a degree in control systems and test engineering. Many things I learned the hard way. This post is all about the dyno itself: how I built it, the challenges I ran into, and the technical details behind getting it to work.Dyno OverviewLeft to right: Input Motor (aka Drive) with a vertical shaft going into the gearbox. Out from the gearbox is a dynamic torque sensor, then a belt reduction (increasing speed). After the belt is the Output Motor (aka Dyno, or Regen). In essence, this dyno measures the shaft power into the gearbox, and shaft power out. Shaft power = torque * speed. With the right parameters, I can simulate anything from light cruising loads to full-throttle acceleration, while monitoring how the gearbox is performing. The dyno system works by inputting power to the Drive Motor, and then regenerating power output from the Dyno Motor. The system has a variety of losses, mechanical and electrical, so the regenerated power is a fraction of the input power (about 70% overall efficiency at optimal conditions). The approximately 30% losses end up coming from the 220VAC split-phase input to th...