Where to begin. I guess the best place is
aaaaaarrrghhhh
[for those just joining the epsiode. A group of several Sharp GP2D12s on a robot puts a lot of noise spikes on the power lines. Very large capacitors reduces the noise to 35 mv, however, the robot still jumps around erratically and will not move]After struggling for several hours this evening, I determined the erratic motor operation was due to the motor driver getting into a funny state. Re-setting the control BasicX didn't help; the system power had to be recycled. Looked at everything with a scope, didn't see anything but there was a lot of 72 MHz noise from the RC. Finally, getting ready to leave I unplugged the D12s and I still had the problem! Previously I had disconnected the D12s by disconnecting the power connector to the board since that was easier than pulling each D12 connector. Whoa - problems and nothing is plugged in except the regulator. Since this board lies on top of the signal wires to the motor driver, I now suspected some kind of induction, but from a regulator?, and moved the board away. The problem disappeared. Huh ? I moved it closer, no problem, closer, no problem, on top of the wires, no problem. Aaaaaaghhh, ground loop. The D12 board is fastened to the metal deck with metal screws and I had not adequately scrapped away a ground trace under one of the screws. Let me repeat that - AAAAAGGGHHH. I guess the hint should have been that the RC antenna wire is supported by a metal whip that is, yes, attached to the metal base.
Oh somewhere in this favored land, the sun is shining bright ...
John Piccirillo