I m going to design a circuit to measure a resistor,R n display to LCD. this circuit will using a constant current. by using voltage drop trough the R, I will programming the PIC to measure the resistance of the R using formula:

R = Va-Vb
I(constant)

however the circuit is using voltage source. And my problem is how to make the current flow trough the R always constant although the change of value of R. Because I have simulate a circuit that I get from internet, the current still changing when I change the value of R...WHY?

 

There are lots of op amp constant current circuits. Often these are voltage controlled. It would be nice to have one with one terminal grounded, but many simple circuits are not. To use a non grounded circuit you may want to measure the voltage across the resistor with another op amp ( or dedicated ic ) circuit: the differential amplifier. For different ranges of resistance consider using different constant currents. For high current constant currents you can use a low voltage voltage regulator, good for currents in the 100 ma range. ( an op amp can be boosted by a transistor to give similiar currents ).

 

The wikipedia page has good info on constant current sources:
http://en.wikipedia.org/wiki/Current_source

 

What range of resistance?
What accuracy?

IIRC National Semiconductor and Linear Technology make three terminal current
source ICs. You would put them in series with the resistor you are testing.
A programmable current sink is another option.

For a lower accuracy you could use a current regulator diode or a JFET.

(* jcl *)

 

This is half experimentation, half intelligable guesswork, and a touch of awesome, but looking at finger-sized contacts composed of a shined up metal, i'd say you're working with a range of 500k/finger to 200/finger. 500k being a perfectly dry finger with little pressure to 200 being someone who just dipped their fingers in a saline solution of maximum concentration. That's why they use specialized plastic-like pads or a contact gel for medical purposes, it keeps the resistance at a fairly regular level.

 

For an easy to apply IC you could consider the LM334 which may be also compensated for temperature. It seems good for what you are trying to do.

 

resistance is in few megaohms range
error can be +/- 5%