Jean MASSON (Mr.) wrote:
"But the exact value of capacity is at best 5%. So accuracy is 5% !"

Please read my posting complete !
I suggest also converting resistance to a time relation. Then the exact capacitor value was without influence.

Also its not too hard to calibrate the measuring routine with an accurate referenz resistor and store the correction value inside the data EEPROM of the micro.
Yes, the initial accuracy of capacitors are about 5%, but the deviation over time and temperature was extremely better. And only this parameter was meaningful in this case.


Jean MASSON (Mr.) wrote:
"A few bucks digital multimeter gives you a better precision without pain.
There is sense to build a measuring device only to obtain a better
precision."

Mostly the precision does NOT determine the solving of a measuring task.
Its more probably, that Donald want other things than only view the result. E.g. the resistance represent another measuring dimension and the relation was nonlinear.
Or he want log the resistance over time or many others ...


Peter