Lilypad LED Matrix: part 4.b, adjusted layout
2012-Oct-01, Monday 12:42 amLast we looked in on this project I'd just added in the adapter for the keypad and a battery holder. Since then a couple of things happened.
One; in the two-and-a-half years since starting this project, the silver on the conductive thread has started to corrode. This manifested first as the programmed time delays no longer seeming to work right, and then with an un-even light disruption across the LED grid (this happens anyway, but wasn't noticeable before this). So I rather carefully went over every single trace of conductive thread with a thin paintbrush and a bottle of fray-stop glue. Hopefully it will slow down further corrosion.
Two; I got my hands on a IR-detector unit, and wanted to include it in the top (as talked about in the Dublin Maker Faire post). To make this possible the keypad-adapter would have to double up as an ir-adapter. But this required a change of circuit layout.
The keypad has five wires; 3 output, 1 led, 1 ground. The ir-detector has four wires; 1 output, 1 led, 1 ground, 1 power. The adapter has five slots. But two of them are pins 0 and 1 (Rx & Tx) and I had trouble getting them to work consistently. So to fit everything in I decided to move the power and ground lines out to their own adapter. But where was I to find a two-pin breakout board? They come in large, small (used for the existing adapter), and that's it. *brainwave* I remembered a set of empty LED boards I hadn't gotten around to using.

And sure enough, two separate connections, just add headers! I pulled out the existing stitching and stitched the new layout in. The power lines are actually run along the back in channels of bias-binding for insulation because the whole thing is so crowded at this point.

Turned it on, fixed up the code (new pin numbers), adjusted the timing on the patterns, and tried my hand at interrupt coding. However, before I could finish it up, my top needed to go off with
Jeffrey_Roe and
PartFusion to NY MakerFaire to represent tog, so development of new features was promptly abandoned, to be continued once it returns. (Although, I hear it was a bit wet, so I expect my top probably stayed in somewhere nice and dry.) *edit* Spot the LED top. But now I have a definite list of things that need doing, and then it will actually be finished!
Part 1 | Part 2 | Part 3 | Part 4.a
One; in the two-and-a-half years since starting this project, the silver on the conductive thread has started to corrode. This manifested first as the programmed time delays no longer seeming to work right, and then with an un-even light disruption across the LED grid (this happens anyway, but wasn't noticeable before this). So I rather carefully went over every single trace of conductive thread with a thin paintbrush and a bottle of fray-stop glue. Hopefully it will slow down further corrosion.
Two; I got my hands on a IR-detector unit, and wanted to include it in the top (as talked about in the Dublin Maker Faire post). To make this possible the keypad-adapter would have to double up as an ir-adapter. But this required a change of circuit layout.
The keypad has five wires; 3 output, 1 led, 1 ground. The ir-detector has four wires; 1 output, 1 led, 1 ground, 1 power. The adapter has five slots. But two of them are pins 0 and 1 (Rx & Tx) and I had trouble getting them to work consistently. So to fit everything in I decided to move the power and ground lines out to their own adapter. But where was I to find a two-pin breakout board? They come in large, small (used for the existing adapter), and that's it. *brainwave* I remembered a set of empty LED boards I hadn't gotten around to using.


And sure enough, two separate connections, just add headers! I pulled out the existing stitching and stitched the new layout in. The power lines are actually run along the back in channels of bias-binding for insulation because the whole thing is so crowded at this point.

Turned it on, fixed up the code (new pin numbers), adjusted the timing on the patterns, and tried my hand at interrupt coding. However, before I could finish it up, my top needed to go off with