Mark Willis wrote: [snip] > If you have one master block, you'll need all power to be fed through it > (Could just use a track that the "=" signs plug into, and perhaps slide > on (like track lighting tracks?), to feed power to all the rest), power > coming from just one place with 25 resistive connections to the end of > the circuit gives you power distribution and ground lead lifting > problems, potentially. Of course, the only "major" current used here > will be at an equals sign, all other currents are pretty small, really. > OTP 12C508A's for most blocks, I imagine? The CMOS parallel/serial SR's consume less than 600 uA per module, the total equation (lets say 20 modules) will consume less than 12mA from the master's power supply. Imagine a very poor contact (5 Ohms) at the power pins connector (more than one pin can be used), it will be a maximum V drop of 60mV from the master to the module #1, 57mV between #1 and #2, 54mV between #2 and #3, and so on until the last module will receive master's VCC - 1.14V. Suppose Master feeds 6V to the modules, still plenty of voltage (4.86V) for those low consume cmos shift registers, and the module #1 even supplied wikh 5.94V will understand perfectly the PIC logic signals from 0-5V Even very cheap connectors (if contacts are protected against the skinny boy chocolate fingers) can offer very low contact resistances, below 0.5 Ohms. Problems with those cheap connectors are basically current, if it is high it heats the contact metal and starts oxydation, and that is what increases the contact resistance. I am talking about currents in magnitude of more than one amp per square millimeter of a really cheap contact. This configuration expects a maximum of 20mA at the master's VCC output pin contact with 10sqmm. Remember the christmass series lamps, with 50 lamps in a single loop? Can someone tell me a worse contact than the ones used at those contacts? some of them still working after 3 or 4 years, even under rain during a full month (here in Florida), then cooking its oxydation during another 11 months... :) So many other solutions are possible, with a rechargeable battery per module (or even a 3.6V coin cell), microcontroller per module, even a communication protocol to master blinks a red LED at the wrongly positioned operator, but all of this increase prices. Do we have $50 per module? oh man, I can make the module talk... sing... and dance... strip? extra $50! :) Wagner