> fsr's (force sensing resistors) are good for this purpose,

Save yourself the trouble.
Been there, done that, still have the emotional scars.  :)

Wanna buy the very expensive "evaluation kit" of Tekscan 
Flexiforce FSRs for which I paid about $100?
No, I won't take your money, I am not a thief or a con artist.

They had the best specs of the bunch, and while I bought
several of these sample kits, the bottom line is that
these things are only good for the most general impression
that "force has been applied" or "more force than before"
or "less force than before.

If you look carefully at their detailed specs, which they
grudgingly provide when you ask pointed questions about 
applications where the sensor acts as anything more than 
a switch, and they spout nonsense like:

"DRIFT:  < 3% / logarithmic time (Constant load - 25 lb)"

Don't feel bad if you do not understand this spec at all.
Neither did I.

The concept of "the log of time" is a highly entertaining 
idea, one we would like to see published in the journals 
for discussion.  :)

What they are covering up here is that these things are
utterly useless when under any load at all for any time
at all.  The resistance of the sensor drops faster than
the price of honey at a diabetic convention, and slowly
reaches some sort of minimal measurement, but never really
stops decreasing, even after months.

If you think about a log curve, you can imagine how 
rapid the degradation of the sensor is under a constant
load, and just how fast the reading "drops off".

There's another spec that is a real knee-slapper, and
would be crucial to the design of Wheatstone bridge 
to slide into the circuit "across from" the Flexiforce 
sensor as a temperature compensating strategy.

The spec says:

"0.36% per degree C"

Which equates to a "3600 TCR PPM/Deg C".  Now compare 
this with known standard values for commonly-used 
resistor materials, and you will find:

TCR PPM/ Deg C    Material
--------------    ------------------------
	3600        Flexiforce Sensor
	3700        Gold
	3800        Aluminum
	4000        Copper
       660        Palladium Silver Ink (Dupont 6134)
      3350        Pure Silver Ink (Dupont 6160)
      
How could the sensor have a "better" TCR than gold?
What might one use in a Wheatstone bridge with this
sensor?

The specs are absolute and utter garbage, just like 
the FSR sensors as an entire product category if you
had any idea of using them for any sort of quantitative 
measurement.

Yes, they can be used to make qualitative comparisons,
but not anything one could dignify with the term
"measurement".

******************************************************
* Full guidelines for BEE-L posting are at:          *
* http://www.honeybeeworld.com/bee-l/guidelines.htm  *
******************************************************