I think the highest resistance would come from heating the relay. Yesterday I grabbed an automotive cube relay from my tool box and measured the coil resistance: 72 ohms. I hooked it up to 14V for several minutes and checked it again. 96 Ohms, at room temperature. I bet the relay is specified to reliably switch at low battery voltage and in Death Valley temperatures. The ability to reliably switch at the absolute worst conditions results in it dissapating ~2W under nominal conditions.

I am not sure that baking a relay at 100C and then measuring coil resistance is the best way to choose the pullup current, although that would definitely be one way to do it. I was hoping that there was a generally agreed upon range of contact currents that avoid dry switching problems.

What do you think of the idea of polling each switch input? I could easily turn on the pullup resistor for a brief moment, measure the input voltage, then turn off the pullup. This would reduce power consumption to almost nothing. This approach could reduce battery drain when the ignition is on but the engine is not running.

- Lee