2/11/26
So the final build is finally finished. I had to tweak some of the design a bit:
Tweaked areas causing thin walls during printing
Fixed antenna bump body separation
Resized bottom cap a bit larger for tighter fit.
Adjust plate sizes and mounting
Thanks to a reddit user, I was informed that the RAK solar chip is able to charge a PO4 battery correctly, and that I would need to add the Adafruit board back in. I decided to just leave in the LIPO cell in for now. I will likely branch the project, and redesign a PO4 model.
I also need to get a better antenna....
I have added the print files to Printables. I will continue to update as needed if any changes or improvements are needed. The assembly is fairly easy, but if there is further interest I can put together a comprehensive assembly manual.
2/5/26
Milestone: UFO Rev.01
The design went through many iterations, but I have finally arrived at a final design. After a couple attempts at a simple rectangular case, I decided to go a different direction with a square/round hybrid. I've started referring to it as the UFO.
This design allows for the bottom to be unscrewed to access internals, eliminating the need for case screws. An electronics mounting plate houses the PCB's and the battery, it also allows for cable routing under the plate making for an extremely clean layout with predefined cutouts for cable passthrough. The solar panel will be 'permanently' affixed in place using silicone/RTV, to keep everything nicely water proof.
A 1/4-20 heat insert in the bottom cap allows the device to be mounted on basically any standard camera mount. With how the bottom screw cap is designed, no further gasket or O-ring is technically needed, even if the device is mounted at an angle. This also in theory allows for easier integration of other mounting platforms as well, like magnetic mounts and etc. It is completely weather tight while being completely accessible for tool-free maintenance. Although I may end up adding in a small grub/security screw just to prevent any movement, we'll see how testing goes. The gasketed panel mounted SMA connector allows for either external or direct mounted LoRa antennas.
I embarrassingly didn't realize that the RAK19003 companion board has built in solar/battery circuits, so I ditched the BQ25185 controller completely. I also ditched the 18650 battery holder in favor of just a 18650 battery with leads. This simplified the internals significantly For my test build I'm using a standard 2200mAH LiPo, but will be switching to a LiFePO4.
The original design was quite large, but over 3-4 prototype runs I was able to reduce the final size by a little over 60%. Admittedly the circular design is technically less efficient size-wise compared to a rectangular housing, but I thought it was a fair trade off for the improved mounting capabilities and streamlined internal access. Plus I wanted to do something a little different. It took several iterations to dial in the electronics plate, mainly due to the tolerances being so small. Total print time for all the components is about 4hrs. The final dimensions are around 113mm dia x 40mm thick (~4.5in x 1.5in)
I have completed a full PLA prototype and I'm working on the final revision using PETG. I will be using a lighter gray color to reduce heat build up in direct sunlight, and an accent color for the bottom cap for aesthetics; just waiting for the filament to arrive. I'm looking into possibly using ADA or even fusion jet (nylon).
Once I finish the final PETG version, I will be posting a final photo shoot of the device. In addition I will be releasing the print finals, assembly manual, and a complete BoM. Following that, I will be mounting the device and monitoring over the next few months and make any updates as needed.
Deployment
At the moment I have 4 confirmed deployment locations (private property) pending testing period and funding to build devices. Two of these will be deploying with high gain pole mounted fiberglass antennas. I also have some irons in the fire for additional deployment locations, but I'm holding off until I can get few more in the wild before aggressively perusing those leads. I want to ensure the platform/design is robust enough. I also want to ensure these are deployed strategically enough to prevent too much coverage overlap.
1/2/26
I've been interested in mesh networks for a while now. I've been getting a little burnt out on my fold writer deck project, so I decided to take a little break and work on a simpler project for a bit. This will be utilizing the Meshtastic system.
My area doesn't really have any nodes, so I thought it would be fun to see if I could develop and deploy independent nodes throughout the area. Building a decentralized network that is insulated against disruptions in public infrastructure outages (both cell and power). I have a few people already interested in deployment on their property. If I'm successful in building this, I hope to start deploying these through my area to build the network.
I'm not really doing anything novel with this build. Mesh networking already has a pretty robust community behind it. There are dozens of clever enclosures and builds out there. But I thought it would be fun to design my own. The components are all off-the-shelf and well tested.
Components list:
RAKwireless Mini Meshtastic Starter Kit US915 RAK19003 + 4631
Adafruit BQ25185 solar controller
18650 LiFePO4 battery (1800mah)
5V 0.6W solar panel (Adafruit 5856)
The enclosure is fairly simple. The main concerns being weatherproofing and UV resistance. The idea is that once these are deployed, they shouldn't need maintenance for long periods of time. So it will be important that the design is well sealed and robust. The biggest concern I have at the moment is UV exposure. I don't have the capability to print in ASA or ABS, and PETG still has noticeable UV degradation vs those. Depending on cost, I might get them made from nylon. But for now, I'm focused on getting the design finished. Once I have a prototype, I'll start digging into alternative materials if PETG doesn't pan out.
I have ordered the various components and finished the first revision of the enclosure. Once I get everything, I'll be printing out a prototype and seeing how things go.