To support the pull-down resistor argument
on my 16c73a the Vol is 0.6V max wheras Voh is Vdd-0.7V
i think this would be because of differences in the complementary fets in
the totem pole outputs.

so a pull down resistor with 0.6v would consume 15% less power than the
same resistor used as a pullup



On Fri, 11 Jun 1999, William K. Borsum wrote:

> This brings up another question:
> For inputs that occasionally have external inputs plugged in, but are
> otherwise normally disconnected from external voltages, should the default
> be pull-UP or pull-DOWN resistors on the input pins?  Again, the object is
> to absolutely minimize power draw. Based on the arguments below, a
> pull-down might seem best--assuming the external voltage sources and the
> PIC code don't care.
>
> Any more thoughts?
>
>
>
> >I talked to microchip tech support at one time about this, there really is
> >no concensus, because if your device is held in reset all the pins are in
> >tristate so i suppose it could draw lots of power.
> >
> >The most economical, and safe from a software perspective seems to be just
> >configuring them as outputs.  Next to that probably putting a 100k
> >resistor to ground and configuring as outputs with latches low.  The 100k
> >resistors take care of having a tristate input during reset/input during
> >startup.
>
> >> 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.
>
> ****************************************************************************
> ********
> All legitimate attachments to this email will be clearly identified in the
> text.
> William K. Borsum, P.E.
> OEM Dataloggers and Instrumentation Systems
> <borsum@dascor.com> & <www.dascor.com>
>