I recently have come across a project for the Comma3X that utilizes pogo pins for the quick disconnect of a comma3x. This is my attempt to create an open source alternative to that mount that is DIY and 3D printable. I was inspired after seeing the mount that @cloudj0816 made in the OpenPilot community discord. All credit goes to him for this initial idea! If you'd like to purchase one of his connectors (full CNC metal), you can find the wait list in the #hw-unofficial channel.
- Inexpensive
- Open Source
- 3D Printed
Angle adjustment (for imperfect mounting positions)- This is more of a future goal. I'm first going to focus on getting the static mount correct.
Preliminary BOM:
| Colloqiual Name | Part No. | Quantity |
|---|---|---|
| Pogo Pin | ||
| USB-C Connector | 2 | |
| ASA Filament | n/a | <100g |
| K&J Magnets | D83-N52 | 4 |
| 3M Adhesive | 3M™ VHB™ Tape 5952 | ~100mm2 (~1.5in2) |
- Pogo Pins
- USB-C Connector
- Magnets
- Adhesive
- General layout
- Trace routing
- Trace power calculations - Data analysis pending (2/15/24)
- Initial refinement (MKI)
- Finalization
- Ordered
- Testing - Failed
- Troubleshooting/diagnosis begins for MKI
- Results finalized, head into next revision.
- Secondary Refinement (MKII)
- Ordered
- Testing - Failed
- Troubleshooting/diagnosis begins for MKII
- Project frozen.
- Results finalized, head into next (final) revision.
- Tertiary Refinement (MKIII)
- Ordered
- Testing
- Troubleshooting/diagnosis begins for MKIII
- Results finalized.
- PCB imported and modeled.
- Male mount modeled
- Female mount modeled
- Mock-up mount printed.
- Mount refinement
- Test-bed mounts printed
- Test-beds tested
- 'Gravimatrix' mock-up printed
- Project frozen.
- 'Gravimatrix' refinement - Minor adjustments being made
- Preliminary testing
- Production unit trials
- Endurance testing (50 hours usage)
- Vibration testing
- Heat testing (7 day constant connectivity)
- Comma3X Weight: 311g (11oz)
- Power Usage: See the summary spreadsheet for this data.
I've gotten the first hand-built boards put together and making a solid connection! The most likely candidate for the mounting system will be the Gravimatrix design, as that is the most solid. But as of 3/10/24 I have a working prototype! Soldering USB-C connectors by hand is extremely difficult, who'da thunk it?
These connectors will likely be soldered on from the manufacturer going forward, this will ensure consistency and quality at the cost of a few cents per board. A compromise I am more than willing to make.
PCB no worky! I believe I have some crosstalk/interference throughout the coupler. I'll be re-assessing and redesigning the PCB to hopefully counteract this and get a working connection.
This weekend I'll be using an oscilloscope to check signal integrity. Hopefully its simple as that, and adding a ground plane(s) will solve it. I had no idea what a ground-plane was until I asked for ideas/advice on the electrical engineering stack exchange. Heres the thread where that was suggested!.
Oscilloscope readings showed little to no interference. Messages were coming and going normally, with solid definition. The second PCB design has run into the same error, in the same way. I will be doing some more testing including soldering wires directly from pad-to-pad in an attempt to isolate this to one component. If that doesnt work, I'll be doing one final re-design before putting this project on hold for a while.
I finally got around to wiring the pads directly to one another for testing! Bad news is that it still doesnt work. I'm unsure, at this point, why this board doesnt work correctly. I feel as if I'm getting some sort of cross-talk between the connections. I'll be tring one more configuration, flipped, to see if that yields any other results. If this doesnt work then I have a backup plan which involves some consumer-hardware and a 3D printed case to house it. I'm hoping this custom solution works, as I put a lot of time into it, but if not then it is what it is. I should've gotten an EE degree, physics is useless here! Actually it was useful for the magnet force calculations.
Unfortunately it seems that I've hit the limit of my knowledge/ability with this project. I cannot get the communication to work correctly across the interlink. I have already spend about $400 in material on this project, putting together another PCB for order would only add to that and would likely not work. Not to mention the hours spent testing and designing the mounts/PCB revisions. This was still an excellent journey to go on! Maybe one day I'll revisit it. This repo does contain some useful info, though, such as power statistics and the weight of the unit, hopefully that will be useful to someone should they stumble across this.