David VanHorn wrote:
>
> Easy:
>
> Use a pair of diodes.
> One charges the cap through a known 1% resistor.
> The other discharges the cap through the unknown.
>
> The cap value is common to both measurements, and cancels out.
>
> You need a schmidt input for this, something with high and low thresholds
> that are not both at VCC/2. An HC14 works well.
>
> You are measuring the time it takes to charge/discharge C from VthLow, to
> VthHigh and back down again.
>
> Measurement is a little complicated, but not bad.
> Start with the cap charged (reading a "1")
> Discharge till you get a zero, then IMMEDIATELY change the output to charge
> again, and measure the time taken to get there.
> When you get there, IMMEDIATELY set to discharge, and measure the time it
> takes to get to "0" again.
> You can average up the measurements.
>
> If you don't immediately reverse on hitting the threshold, then you risk
> charging to some unknown voltage, (not the threshold)
>
> If this isn't practical, then arrainge the HC14 as an RC oscillator through
> the diodes and resistors, and simply measure the high and low times.

Why do it the hard way (diodes)?
You just need ONE 74HC14, and your unknown
R is the feedback resistor and the C hangs off the input pin.
You get an oscillator period which is proportional
to R if the thresholds don't move with temperature.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.