Electronic – Switches for current source

I may be way off base, but I believe this is a case where you don't want to over-think the problem, KISS.

When SW2 is closed the input pin is tied to ground and you should read 0. This is just like the case where an unused pin is grounded rather than left floating.

When SW1 is closed the input is tied to 5V (either Vcc or Aref), and you should read the max value (that is all the binary bits will be 1)

For both of the above cases, you could remove the resistors from the above circuit as they have no affect on the level of the input pin (that is one or the other of the switches is closed).

The only remaining case to consider is when both switches are open. If you did not have any resistors then the input pin would be floating. When the input to the ADC is left floating one would expect to read out semi-random values from the port, certainly not an unchanging value of 0 or MAX (all bits on). While I don't recommend omitting the resistors, you should be able to make this work without them.

While not explicitly part of your question, the more interesting case is if you wanted to detect one of several switches being closed with a single analog input pin. For a reasonable number of pins, say 10, this can easily be done with a resistor ladder ( http://en.wikipedia.org/wiki/Resistor_ladder ). You can also Google R2-R ladder.

For a walk-through of how this is implemented see: http://embedded-lab.com/blog/?p=4040

As an aside, if you only wanted to sense a single switch closure one would most likely use a digital I/O pin rather than an analog one, if available. The most common reason for using an analog rather than a digital pin for detecting switch closures is when the number of digital pins available is less than the number of switches.

I think your assessment that there are fewer N.C. type switches is somewhat short sighted. It is true that most TACT type switches are indeed N.O. type but many other switch types offer Form C type contact arrangements that offer both N.C. and N.O. on the same part.

If the input voltage on your regulator is always present you can leverage that to your advantage with some simple additional circuitry to provide the inversion logic that you require for use of a N.O. type switch. Here is how to connect it up.

Use an NPN transistor with its emitter connected to GND. Tie the collector to the inhibit pin on the regulator chip. Then connect a resistor from the base of the NPN to the input voltage supply rail. A 10K resistor may be a suitable value. Finally connect the N.O. tact switch between the base and GND. A small signal type transistor such as a 2N3904 is probably the best type to use.

Pressing the switch turns off the NPN which in turn frees the GND connection at the inhibit input.