Tetris Table

So originally the thought was to have a 20x10 LED array, but then this would be strictly purpose built for Tetris, whereas there is so much time and effort going into the design and hardware (as well as expense). So changing to a 20x20x4 LED table will be used instead. This drastically increases the potential power consumption (2x of original estimate), but build isn't entirely harder. Created a 3D mockup of the 20x20 LED table.

So after a long-overdue timespan, I have picked up the project again! Recently I tested the boards and noticed that I had an atrocious time soldering all of these pads with the LED's, and I was burning out too many of them to viably make something this large. So I did a new revision of the LED boards. Now they'll be back to the standard 2.54mm headers on both sides of the board for I/O and power. Also I spaced the LED's to have a little bit larger room to solder the LED's together, instead of the 1mm space I had before it's closer to 3mm. The boards should be in before the end of September, and once again, went through OSH Park due to the awesome history of customer service I had with them. Software is on hold until some hardware works.

In addition, I realized using something like solder paste might be a fantastic idea for this project, so in conjunction with this project I am attempting to build a mini-fridge to store the paste!

Yesterday the second revision of my boards came in. These boards were even smaller, coming in at 15mmx11mm, which through OSH park came out to be $2.60 for 6 boards. As of yet I have yet to test the boards to ensure validity, as I need to ensure there are no shorts as I was not the most confident in the board layout. Once I get my first LED's soldered, and the Headers actually arrive from USPS, I hope to have some picture up showing the design and how it is coming along.

After working with mbed.org for a bit, I became dissatisfied with the results I was seeing. For compiling a simple code that just flips an LED it was consuming 14K memory. Even though the 411 has 512K it also means to me there might be some performance implications associated with it, so I am shying away from it.

For this I will be approaching using gcc on an ARM-M4 compiler via a Linux VB to see if I can get it to work. If this does work I'll post a link/my instructions to get it to work.

I suppose this is a bit late starting to log; but better late than never.

The progress that has been made so far is developing the PCB for the WS2812's that will be used for the discrete pixels. This has been developed over the course of learning how to use KiCad. I sent these off to OSHPark ~ 4 weeks ago, and after USPS screwed up the first order; I finally got the boards. Things I noticed with the design (see the github link) is that I did not leave enough space to solder the LED's effectively, so the first attempt at soldering was.. botched at best. The header drill holes also are not large enough for the headers I purchased; but this made me think to replace these headers with a SMT variant; such that I can decrease the size further.