Another 2 cents worth... The project currently on my bench will have such an unused input in some configurations, hence I've been musing about this. Here are the 6 (well, 8) choices: 1) input, no resistor: Bad, bad, very bad. Oscillation isn't really the problem here (who cares how an unused pin changes), but the possibility it will drift to Vcc/2 and hence turn on 2 fets from rail to rail is the scary thing. 2) input with pull-up: OK, make the pullup the largest value you dare. Megs would be OK in most enviroments. Dissipation will be practically zero. 3) input with pull-down: same as pull-up. 4) output no resistor: zero component count. Again no disipation for either output state. Only danger is if the board is somehow shorted pin to pin. If this is possible, perhaps one would use the same hi or lo state as the adjacent pins to minimize uncontrolled short current. 5) output with pulldown: Adds a component for no reason. A high value resistor will of cource be chosen... which is in parallel with the low impeadance output stage... in the equivalent circuit the resistor drops out. So why use it? 6) output with pullup: same as #5 I'm of course assuming a pulled-up output will be driven high, and pulled-down driven low (ah, there are numbers 7 & 8). To cross these is to just waste power. Incidentally, pulling up a high output will disipate zero current, as the output only has its .7 drop from the rail IF there is current being pulled out of it. No current, no drop. High outputs source current, not sink it. My choice is #4, output with no resistor. Since my free pin is next to the ground pin, thats where I nailed it... output in the zero state. At 12:24 PM 6/11/99 -0700, you wrote: >Greetings All: >Today's Question >What is the best way to treat unused pins on a PIC for absolutely lowest >possible power consumption? (and the theory behind your choice please). > >Options are: output, drive hi, no termination > output, drive lo, no termination > input with pullup > input with pulldown > input floating (not really an option--pin will oscillate sometimes) > >Some thoughts from my end: >1. Inputs require a pull-up or -down resistor, possibly in the megohm >range, but still some current flow, plus an extra part. >2. Outputs don't require external parts, but with big ground planes and >potential board leakage, driving low might be the best. However, most of >the programmers I have used tend to drive the pins hi. > >So What is the consensus? >kelly >**************************************************************************** >******** >All legitimate attachments to this email will be clearly identified in the >text. >William K. Borsum, P.E. >OEM Dataloggers and Instrumentation Systems > & > >