What this personal reconstruction project provides.
fanhao375’s Microduck Replica ↗ connects public simulation geometry and runtime source to reconstructed drawings, CAD-importable assemblies and electronics research. Its English and Chinese documentation makes it useful for investigating how a MicroDuck-style build might fit together.
Our review uses revision cf347f2, checked September 8, 2026. Read the original project for later changes. This page attributes the author’s findings; we have not assembled its hardware.
Separate build evidence from design work.
| Material | Available evidence | What it does not establish |
|---|---|---|
| Physical build log ↗ | September 1–2 printing and trial-fit entries, with photos of early parts and screws fitted to leg pieces. | A complete powered robot, calibrated sensors or a reproduced walking result. Several print and measurement fields remain TBD. |
| Reconstructed geometry ↗ | Assembly drawings and files derived from the public robot model. | Manufacturing tolerances, every fastener fit or cable clearance on a finished robot. |
| Official RPI Robot HAT ↗ | A separate Pollen Robotics hardware project with a KiCad design. | A complete MicroDuck mechanical BOM or a whole-robot assembly procedure. |
| Community imu_to_dxl board ↗ | A third-party schematic and PCB design with review notes. | A fabricated, electrically tested replacement. The author explicitly states that hardware validation is still missing. |
There is an official hardware source worth knowing.
It would be inaccurate to say that no Pollen hardware files are public. The RPI Robot HAT repository ↗ describes a board with IMU, motor communication and audio, designed in KiCad 9. It also notes that Feetech use requires cable adaptation. Inspect that board’s own schematic and revision when planning an interface.
A published board is one part of a robot. It does not resolve the rest of a replica’s mechanical, power, firmware and calibration questions.
Use the files as a sequence of experiments.
- Start with the physical log. Record the last demonstrated stage. Compare photos and measurements with the particular parts you plan to use.
- Keep one source revision. Save the chosen drawings and notes together. An updated connector drawing should trigger a new wiring review.
- Trial-fit one assembly. Measure holes, fasteners, clearances and cable paths before committing to a complete set of printed parts.
- Treat the adapter as a prototype. Review its exact connector model and pin order. The board notes contain a September 8 connector correction, so a remembered Feetech pinout is not sufficient.
- Record each test before integration. Keep mechanical observations, electrical checks, firmware responses and motion trials as distinct results in your build log.
The adapter’s design-review page ↗ is the place to inspect its current assumptions. Passing a PCB design-rule check does not measure current handling, signal quality or robot behavior.
What would make a complete replica tutorial?
We would look for a dated whole-robot BOM, tested fabrication files, a matched wiring and firmware version, calibration instructions and a repeatable first-motion record. The build should identify substitutions and unresolved faults alongside successful results.
For now, this project is a good research companion for experienced makers. If your goal is to follow an established hardware build sequence, use the separately documented Open Duck Mini v2 route, or start with the Chinese tutorial collection.
What would you like to understand next?
Keep the source, your environment and your observations together.