Step 6 : Details on the PXLC10


I've had a few emails from people who ordered the board and couldn't figure out how to assemble it. My subcontrollers are SIX channels each, not 8. That means my Cat5 cable has six signals, one ground, and one wasted wire.

This gets a little confusing, because the first 74LS374 gets 6 channels from the first Cat5 cable, and two from the second. Then the next 74LS374 gets 4 from the second Cat5 cable and 4 from the third. The next 74LS374 gets 2 from the third Cat5 cable and 6 from the fourth, etc.

Also watch out because the 4 connectors on the left side of the chip are ascending in bits, but on the right side, they are descending. You have to think twice when wiring them up. Sorry for making it so confusing, but it made the PCB layout easier.
Note that for the subcontroller on Cat5 cable labelled B, 6 channels come from the last chip on the right (brn+wht, grn, blu+wht, blu, grn+wht, ora). The last to outputs of that same chip go to subcontroller A (brn+wht, grn). I use the brown wire for ground in all cables, these go to the J61, J62, .. J67 points on the board

Step 7 : Channel order on PCB


Here is another pic showing the pattern. My subcontrollers are A, B, C, D, E ... and I denote the 6 channels on A as A1, A2, A2, A4, A5, and A6, etc. The pattern repeats every 4 subcontrollers. 4 subcontrollers are served by 3 74LS374's, i.e. 3 x 8 (bits per chip) = 24, 4 x 6 (bits per subcontroller) = 24.

You can see a bigger size of the above photo at

http://home.comcast.net/~donlaw/lights/pcbnum.jpg

Step 8 : More information

I did two presentations on the project to our local linux users group, FLUX. The slides are available online at flux.donlaw.com
if you want some more info. The
May presentation talks more about the hardware described here.