- README: the HMI is generated turned round (nothing to set in the Editor); PC radio link: bridge mode done, no USB links any more, what's left (Unity driving the real robot, three nodes at once). - remote/README checklist: sticks and IMU checked, display uploaded, radio checked against the sim through the dongle; replace the IMU board. - CLAUDE.md: display runs the new HMI; next steps. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6.1 KiB
Hexapod
A six-legged walking robot (Teensy 3.6) and its handheld remote (Teensy 4.0 + Nextion touchscreen), linked over nRF24L01.
The remote above is the new design (hardware/openscad/remote_v2.scad, not
printed yet). The original printed remote is on the
archive page.
robot/ robot firmware: IK, gait, body pose, IMU stabilization, state machine
remote/ remote firmware: 2 x 3-axis sticks, MPU6050 tilt, Nextion UI, RF24 TX
shared/ HexapodProtocol: the CmdPacket wire format + radio settings used by both
sim/ the unchanged robot firmware compiled into a DLL for the simulation
Hexapod_Simulation_Unity/ Unity project that runs that DLL on the 3D model
dongle/ USB nRF24 stick firmware (STM32F030): the PC-side radio, see dongle/README.md
hardware/openscad/ OpenSCAD model of the robot (rebuilt from the Fusion designs) and the new remote
hardware/openscad/stl/ print-ready STL exports
hardware/fusion360/ Fusion 360 designs: body, legs, servo, servo horn, remote case
scripts/ PlatformIO helper (per-environment sketch folder)
This is the 2026-current/ folder of the hexapod-robot repo; the earlier
generations (2009 onwards) are in the folders next to it, see
../README.md.
Build (PlatformIO)
pio run # build both
pio run -e robot -t upload
pio run -e remote -t upload
pio device monitor -e robot
Simulation
From this 2026-current folder in PowerShell (the dev PC is Windows):
.\sim\build.ps1 # build hexapod_sim.dll + run the headless firmware checks
Then open Hexapod_Simulation_Unity and press Play. See
sim/README.md.
Printing the remote (v2)
STLs in hardware/openscad/stl/remote_v2/,
already in print orientation; everything prints in ASA without supports.
Regenerate them from remote_v2.scad with -D 'PART="..."'.
| File | Part | Notes |
|---|---|---|
remote_v2_top_inlays.3mf |
top shell with both inlays | One object with three parts on filaments 1/2/3: case black, honeycomb red, ON/OFF white. The inlays are the first 3 layers (0.6 mm) of the face-down shell, so the colour swaps only happen there. Made by hardware/openscad/export_3mf.py from the three STLs below |
remote_v2_top.stl + _inlay.stl + _text.stl |
the same, as STLs | Load the three at once and keep them as one multi-part object |
remote_v2_top_multicolor.3mf |
OrcaSlicer project (owner) | all parts on two plates; after re-exporting STLs, update its parts with Orca's "Reload from disk" |
remote_v2_bottom.stl |
bottom shell | flat underside on the bed |
remote_v2_kickstand.stl |
table stand | |
remote_v2_eyelet.stl |
neck-strap eyelet | on the top face (slots to balance it) or on the front edge |
remote_v2_bail.stl |
body brace | swings down and back to rest against your body |
Hardware (besides the electronics of the old remote):
- 14 x M3 heat-set inserts (4.0 mm hole, ~5.7 mm long): 4 shell screws, 2 + 2 eyelet (top and front), 2 brace pivots, 4 main-board clamps
- M3 x 12 x 4 (shells), M3 x 8 button head x 2 (eyelet), M3 x 16 button head + washer x 2 (brace), M3 x 8 button head + 7 mm washer x 4 (main board clamps, no holes in the board), M3 x 40 + nut (kickstand pin)
- M2.5 x 14 x 8 (joysticks, into their own boxes), M3 x 5 machine screws x 4 (display, tap the posts M3), M3 self-tapping x 2 (UPS board), M2 self-tapping x 2 (switch)
- The display is mounted turned round so its connector clears the Teensy's USB
plug.
remote.HMIis generated for that (display direction 270 in the Editor's terms, see remote/NEXTION.md), nothing to set.
Design notes, checks and open points: hardware/openscad/PLANNING.md.
Planned
- PC radio link: a USB nRF24L01 stick on the PC as the PC-side radio, on
the same air settings as the robot and the remote. It's reverse-engineered
and runs our own firmware (STM32F030 + Si24R1 + AT2401C, updated wire-free
over USB), see dongle/README.md. Done (2026-11-17):
its bridge mode lets the real remote drive the Unity simulation over the
air, with the sim's telemetry going back to the remote's display as the ack
payload (sim/README.md, "Using it in Unity"). There's no USB
link to the remote or the robot any more; everything goes over the radio.
Still to do:
- Unity driving the real robot (the dongle transmitting instead of listening) and mirroring it;
- three nodes at once: today the dongle answers the remote instead of the robot. Needs an address per node (or per pipe), a packet type byte, and a rule for who has priority. Owner's decisions: the remote wins, with a toggle on the remote that hands the robot to Unity and lets the remote's sticks control Unity instead; the PC is the hub while it's online, the robot is the hub when it isn't; goals: mirror the real robot in Unity, drive it from Unity, remote data to the PC, and PC data to the remote.
- Remote battery level: see remote/README.md, "Planned".
- Robot power: a stable 6 V supply for the 18 servos, see hardware/openscad/PLANNING.md, "Notes for future designs".
Where to go next
- robot/README.md: placeholder geometry to measure, the bench-check checklist, and the serial test harness for driving the robot without a remote.
- remote/README.md: wiring, control scheme, stick calibration, and the remote's serial console.
- remote/NEXTION.md: how to build the touchscreen HMI.
- shared/HexapodProtocol/HexapodProtocol.h: the protocol. Change it only by reflashing both boards.

