Ted,
     I built a pendulum that is similar to the one that has been
simulated, except that it was configured to optimize sensitivity at a
period of about 4 s (to study microseisms).  It is described at the
following site by the paper titled "Compound pendulum to monitor
hurricanes and tropical storms"::
 http://physics.mercer.edu/hpage/compound/compound.html
To convert the instrument shown in Fig. 1, the lower (lead) mass would
be discarded, and the upper (lead) mass would be lowered nearly all the
way possible (just above the axis, after which the upper tube would be
sawed off just above the mass).