The problem is only in the input stage of the [OP AMP][1].

The eddy-current is generated from two coils by ferrous movement passing through them.  The coils are 10 Ohm each with 0.254mH.  The eddy-current coming out of them is a sine wave that goes positive and negative. The circuit works 95% of the time as it is, but at times it seems that the difference between the two OP AMP inputs (2 and 3) is not enough to result in an output.  

On the bench, if I simulate things and feed a positive 3V pulse into **J1** going to Pin 2 of the Op Amp **AND** also ground **J2** (going to pin 3) then it works 100% of the time.  Of course, in real life the eddy-current is not only positive.

I suspect that when the positive and negative eddy-current is generated in the form of a sine wave, at times the input via J2 going into pin 3 of the OP AMP overrides J1 going into Pin 2 of the OP AMP...thus the OP AMP does not put out any amplified signals.  Should I change R2 to a lower value like 2-3K? Or should I add also a capacitor in series between J2 and pin 3 of the OP AMP?

 [![enter image description here][2]][2]


[![-20mV to +60mV][3]][3]


  [1]: https://ww1.microchip.com/downloads/aemDocuments/documents/MSLD/ProductDocuments/DataSheets/MCP6L91-1R-2-4-Data-Sheet-DS20002141C.pdf
  [2]: https://i.sstatic.net/IHWeg.jpg
  [3]: https://i.sstatic.net/npZR2.jpg