John,
I would suggest that the likely mode of coupling is what is called
common mode coupling.  Shielding the coil probably isn=92t going to help =
=96
you could put a short there and the problem would probably still exist.
Both the wires to the high and low sides of the coil are probably acting
as a single antenna.  If so, twisted wire won=92t help either.  I will
guess that your input circuit is reasonably high impedance in relation
to the sense coil resistance.  If so, proceed as follows.  Put a
resistor of about 10% of the value of the resistance of the sense coil
in series with each line to the input of the op-amp and physically close
to the op-amp.  Then at each node formed by the new resistor and the
op-amp, bypass the node to the op-amp circuit ground with a cap that
forms a lo-pass of 1KHz or so with the resistor ( C =3D 1/(2 * pi * f *
R).  This will give a practical 20 to 30 dB of attenuation at the AM
station frequencies and will probably suffice for your purposes.  You
can get more attenuation by using two sections in series instead of
one.  Depending on your op-amp input, you may need to add another series
resistor of the same value after the cap if the op-amp input goes
unstable and oscillates.
Charles R. Patton


__________________________________________________________

Public Seismic Network Mailing List (PSN-L)