Josh Koffman wrote:

> If I had a pull down resistor, and a switch connected to Vdd, would
> that be enough to trigger it? Or for that matter would the opposite
> work?

  You're really barking up the wrong tree here.  A Schmitt input is an
input, forget the rest - just drive it with CMOS logic levels - at least
3V (at 5V Vcc) for "1", and less than 1V for "0".  It's extremely high
impedance, so either a pull-up or a pull-down will work excellently, as
long as the "pull" value is much lower than any other leakage
resistance in the circuit.

  In fact, the Schmitt action relates to hysteresis - the voltage needed
to turn on is significantly higher than the voltage needed to turn off,
so it doesn't have any indeterminate state or sensitivity to trivial
noise at a threshold level.

> It doesn't have to be quad bilateral...just something normally closed
> The only thing I can think of are relays, and I was wondering if there
> is a solid state alternative.

  Yes, it's called a depletion mode FET, generally JFETs (junction FETs)
do this - they conduct fully at 0V on the gate.  They are only available
as discretes (AFAIK) and the gate functions as a diode to the source, so
you have to figure out your control arrangements accordingly.

> Will I run into problems having them all on one port? Will the
> microcontroller be able to drive them sufficiently all at once?

  YOu wouldn't dream of using other than CMOS logic for this (i.e., if
you do use an 8255, use an 82C55), the driving ability of the PIC
(fan-out) to CMOS is immense, unless you want very fast response.
Consider the fan-out as being 10 (ten) as a rule of thumb.

  THE simplest output chip for your function is a 74HC595, 8-bit serial-
in, parallel out double-latched shift register.  You clock 24 bits into
three of these using a clock and data line, then a third line latches
the data into the output registers.  Three lines to output 24 bits from
3 16-pin chips, or indeed any number of bits at 8 per chip.
--
  Cheers,
        Paul B.