Hi Juan,

using a LM318 is not a good idea, because the LM318 needs higher supply voltages, +/-5V at least. If you power your circuit from a 9V battery then the supply voltage can easily drop to 7V.

Also, your design is very high ohmic and by this very susceptible to hum and EMI.

I wouldn't use a Greinacher circuit either, but a simple rectifier. Take care, in the original circuit from Murata there are some bugs! E.g., the discharging resistor is missing.

For the comparator a CMOS type should be used, because the bipolar version shows too much input bias current.

I would give the following scheme a try:



The threshold voltage generation with the diode at TLC393 is tolerant against supply voltage changes. Use 1%-resistors for the two 10kOhm voltage dividers. The diodes can be 1N4148.

All four Opamps of the LM324 need to be cascaded, because the LM324 is not a fast one.

Shield the whole circuit and the sensor and bond the shield to signal ground. A grounded metallic wire mesh can be used to extend the Faraday cage over the sensor.

Kai Klaas