Everything about the physical swap — dropping a Propeller 2 Edge Module + breakout board into the spot where the Raspberry Pi normally sits on the Robot Dog connection board. This folder holds the build walk-through, the electrical wiring map, and the adapter-plate CAD:
HARDWARE_SETUP.md— photo walk-through: model + print the adapter plate, assemble the P2 board onto it, mount it on the robot.P2_MIGRATION_WIRING.md— the electrical side of the swap: power architecture, the 3.3 V provenance gotcha, the header→P2 wiring map, and the 3.3 V-vs-5 V level-shifting hazards (the P2 is not 5 V tolerant).- Adapter-plate CAD — the editable plate models (see Files below).
The rest of this README is the CAD story — how the adapter plate was modeled and printed.
The plate is modeled in Shapr3D to match the Pi's mounting footprint, with raised standoff bosses positioned to carry the P2 Edge breakout board:
So I'm a heavy Claude Code user, and I decided to try an experiment. The geometry and pin mounting points for the P2 Edge modules are well-known and publicly documented, as are the Raspberry Pi mounting holes. I asked Perplexity, in this case, to study and gather the mounting information (mounting hole sizes and locations), and then to overlay the two with the same orientation. I described how I wanted the platform to look, and then I had Perplexity give me the constraints for drawing it. Perplexity can't produce files, so it gave me the specification for the model file that I'd want. I copied that specification into Claude Code, and I said, "OK, now write the 2D design file that Shapr3D can read." It did that beautifully, and in fact, it did it in one pass. The first one was correct. I was able to use it.
The next thing I did was import the file into Shapr3D and apply the standard extrusion techniques to create a 3D model. I printed some lightweight tests so I could prove that all the geometry was correct, and pleasantly so, the geometry was perfectly correct. I never had to move a hole placement. The only thing that might have been better is if the inner diameter of the standoffs had been narrower, so the screw could have threaded itself. As a result of not paying attention to that, I had to use nuts to hold the screws in place for the P2 Edge standoffs.
The final thing is, you'll see the very first print I produced was perfectly rectangular. When I laid it down, I realized that the pins on the board underneath, where I would have to connect that header to the P2, were too close to this board. The last thing I had to do then was create an inset so the header would have more clearance around it, allowing me to plug in the cables. That was the final modification, which is why the first one in the following print regression is rectangular, while the second and third have the inset. Anyway, I thought this would be interesting for you to know. -Stephen
Getting the standoff height and hole alignment right took a few iterations. The trick is to print thin, quick test plates first — just enough plate to verify the mounting-hole positions and the board footprint line up — before committing the time and filament to the full-height (taller-standoff) plate. Here are three of those test plates, ending in the V2 design:
| File | Format | What it is |
|---|---|---|
HARDWARE_SETUP.md |
doc | Build walk-through — model/print the plate, assemble the P2 board, mount on the robot. |
P2_MIGRATION_WIRING.md |
doc | Electrical swap — power architecture, header→P2 wiring map, 3.3 V-vs-5 V hazards. |
pi_p2_mount_plate.dxf |
DXF (2D CAD) | Outline/drilling drawing of the mount plate — Pi-footprint mounting holes mapped to the P2 board. Open in any CAD tool or for laser/CNC. |
RoboDogP2PlateV2.3mf |
3MF (3D model) | Printable V2 mount plate for the P2 platform. Slice and 3D-print. |
MIT License - See LICENSE for details.
Part of the Iron Sheep Productions Propeller 2 Projects Collection