This week I set out to tackle the two parts of my project that I believed would be the most challenging: 3D printing a frame for my light art project, and wiring up multiple LEDs with a shift register. Before I start, huge thanks to Will Schwartz for printing the two models for me!

I began the week by ordering the parts I needed: 50 pre-wired LEDs, six shift registers, and a pack of solderless breadboards. The first thing I did in the week was design two possible sockets in which to mount the LEDs.

This view shows what would be the inside of the sphere. In the left socket, the LED is inserted from the inside, and held in place by a ring at its base. In the second design, the LED is inserted from the outside and held in place by its wires. In both of these designs, there is a hole on the outside for the LED to shine through.

After 3D printing this test part, both of the sockets worked reasonably well. However, the LEDs I am using are very bright but with a very narrow cone. Therefore, in order to diffuse their light more, I decided to cover up the outside of the holes in the final design.

After seeing the print, and testing it with my LEDs, I was ready to finalize and print the bottom half of the sphere. This is what that ended up looking like in Fusion 360:

As you can see, there are no holes for the light to shine through. However, since the sphere is printed clear plastic, the LEDs can still shine through. Once this was printed, another issue came up with the LEDs I bought. The wires they came with were too narrow to stick into a solderless breadboard (they would just fall out). I overcame this obstacle by soldering small leads of normal wire onto the ends. With that, I was able to insert LEDs into the sockets of the printed half-sphere and see what it looks like with them all on (just connected to a battery):

I’m quite happy with how this turned out, and while the printed sphere has some imperfections, I think it will work fine and does not require a reprint. The last thing I wanted to do this week was get the wiring working with a single shift register. This doesn’t enable all LEDs in the bottom half to function (there are 14, and a shift register has an 8-bit output), but it shows that everything is working:

Hopefully I will be able to demo this in class, but the wiring is rather finicky and might come apart. Overall, I didn’t experience any major failures this week, just smaller obstacles that I was able to overcome. I think that I am in a good place for the project, and my next step will be to create a more permanent solution for the wiring than a solderless breadboard.