Finishing the hardware

Post date: Jan 09, 2015 12:57:56 PM

I finished the hardware in time for Christmas, meeting my goal. I connected the cathode boards together, wired the outputs to the cube, made the anode control board, and connected everything to the microcontroller.

When the pcb standoff spacers came in the mail, I was ready to connect the cathode control boards together. I used a belt sander go shave off a bit of the sides of the perf boards so that they don't hit the components when the boards overlap so the spacers can fit in the holes. Then I wired all the cathode control boards together of the same color, daisy chained together with common ground, voltage source, and serial connections, as well as the thick ground to the power supply. The other end of these wires merge together with the other colors' boards and go to the microcontroller. The green data wire was soldered between each shift register in the input and out the output until it reached all of them. Then I soldered on the ribbon cable, 64 pins per color. I screwed in the spacers and mounted the red, green, and blue levels of boards on top of each other.

To solder the ribbon cable to the cube, all the wires were first stripped and tinned, and then I fastened the boards above the cube so that the ribbon cable hung down and just touched the wires sticking up from the inverted cube. My first attempt involved me starting with one color cable, but by the time I got halfway through, I realized that there was no way for me to go back and do the other colors. So I unsoldered everything and started over. I soldered one row at a time, alternating between the red, green, and blue ribbon cables, being careful and going in order. I kept it neat and organized, although it looks like a random mess. I compared the process to trying to braid someone's tangled hair when their hair is dyed three different colors and the braid has to be perfect, not one hair out of place, or else it might have to be done again, and on top of that, you are color blind and can't tell which color is which.

When designing the anode circuit, this website helped me to figure out how to use the mosfets. I also got some help from a friend. Once I had the working circuit on a solderless breadboard, it took some time to figure out how to squish it to get 8 of them fit on one board. Next I placed the components on a solderable breadboard, soldered them on, cut the leads, and connected the right solder tabs with solder trails to complete the board. I then wired and mounted it to the other boards and connected the outputs to the cube, finishing the hardware.

To test it, I plugged in the power supply, which made a loud plasma zapping noise and died, releasing a small bit of magic smoke. It turns out that a snipped component lead had fallen into the power supply and shorted it out to the grounded metal case. Whoops. I tried replacing the fuse, but that wasn't the only problem. A bunch of components were fried, so I ordered another one. I hope it comes in time for my presentation, otherwise I might have to use a variable power supply which can only output 2 amps max, which might not be enough current.

Now it is time to program. I already have one spatial dimension done and working on my proof of concept, so I just need to add two more spatial dimensions to the base code and then I can finally start programming animations. One is added by extending and folding the existing spatial dimension, and the other is added by cycling through the anodes. A library of functions will be built as I create animations to make programming future animations easier.

-Jeremy