I'm trying to use this device to measure the resistance of human skin (in the range of 50k to 10 M ohm). As a test, I made a resistive divider with 3 resistors: 10M, 680k (a proxy for skin resistance) then another 10M ohm to ground. With 5 volts across the three resistors, the current flow is 242. uA, which is way higher than the input bias current of the MCP6241 (only 1. pA typical at 25 degrees C), so I assumed this would not overtly affect the operation of the part.
I configured the MCP6251 as a voltage follower and tried to measure the voltage at the higher side of the middle 680k resister which should be about 2.58v but it's off be several hundred mV. Why?
1- Maybe my 6.5 digit DVM can't handle the load impedance either and is itself wrong?
2- I forgot to terminate the second opamp in the package and it's driving the circuit crazy?
3- There's more to input bias current than I thought.