Byron A Jeff wrote:
> The toughest part is going to be the sensors. Door loops are subject to quite
> a bit of noise. In my prototype alarm system (which used a button and beeper
> interface, i.e. three beeps to activate, two to disarm, 4 for status) the
> door loop falsely triggered until I debounced with a 200ms debounce routine.
>
> 2) Attached the 120V interface wires to an optoisolator and a reversed biased
> led through a 10Kohm 2W resistor.

  Just a note on the door loops, it's best to use a say 4.7k
resistor at the sensor end, and a resistor at the alarm
end.   Then use a 339 set up for a window comparator around
2.5V.  That way, even if they short or open the wire, it
still goes off.  Any halfway competent thief can bypass a
cheap system without this by shorting the wires.  And a 339
and a few resistors is cheap anyway.
  Also, I know you're talking simple here, but never trust
an all wireless system.  Amazing how many expensive alarms
can be walked through by setting a signal gen to ~318 MHz
and cranking up the power from outside the house.  Good ones
will still go off after a few minutes of no reports, but do
this a few days running and the police stop coming..
  Still seems just as easy to hack back before the triac,
but at least you did #2!!  Amazing how many commercial
devices violate the PRV spec, even though most magically
keep working anyway..