Today I have mainly spent the day learning how not to print. I gave myself the challenge of creating a device from the ground up so that I could think about the process we are taking our students through. It is a combination of the software / hardware side of Arduino plus the sketching and prototyping of a physical interactive to be used by staff and students. The device is quite simple - a rotary encoder with push button and a small OLED to display the device number of one of fifty LED rings we are controlling via MQTT in the lab.
Once the device was technically working and the components locked in, I started to sketch out some ideas for the physical presence of the device. I started with a simple box with button, dial and display - and then got stuck for about 20 minutes. I couldn’t think of other forms for the device. But I set myself the challenge of filling up the pages with sketches and after an hour had several plausible designs to explore. This stuff takes time.
After the sketchbook I started to sketch in Fusion 360. Initially I focused on the wrong things - the external surface appearance - I wanted to have a twisty effect so spent way to much time exploring splines on curved surfaces….
I then realised I wasn’t spending enough time thinking about where things will go in the enclosure. I had however learned really quick ways of creating egg shaped structures so I started with a fresh sheet of screen and started with Lumina 2. For this one I started to think about screen position.
The model and render looked great but the print? Not so much. See the header image of this post for the failed and gnarly mess that came out from trying to print on thin air (ie the top of the void for the screen). I also realised that getting the electronics in and out of enclosure was going to be a pain. So another clean sheet of screen and lumina 3 was born.
For this one I started to explore a screw on base. The thread function works really nicely but I was having loads of problems getting prints to work with this design. The top section was fine but the middle and bottom were badly designed with wall thickness too large which I think was causing the tangled mess. I might try this design on a Prusa 4 at work to see if better cooling helps.
Lumina 4 is where I have now arrived. I replaced the screw fit base with one connected by magnets. Reducing the wall thickness to 2mm has halved the printing time and 10x better print quality on the Prusa 3S. I have also taken a component approach to this one so will end up printing 5 parts and assembing them. Took a few goes to get the sizing right for a snug fit for some of the hardware so being able to print smaller sections helped.
Final assemblies will be added to Flickr with the tag 3dprint:
After thought: I spent a long time at the start trying to find a model of the Arduino MKR1010 to import into Fusion 360 to get me started - that was a “red herring” - a rough size is all you need to get going.