The sensitive, broadband seismometer described in this paper is based on frequency modulation and a two-dimensional (2D) Fourier transform method to obtain time vs frequency (and seismic amplitude) recordings. The intent was to develop an instrument that could compare seismic recordings with variables (as signals) that, for rational reasons, might be active before a seismic rupture and, thus, possibly serve as quake warning indicators. Five such variables were eventually incorporated into the instrument and tested as amplitudes: electric field changes, the electromagnetic spectrum between 3 and 30Hz (extremely low frequency as Schumann-like emissions), the electromagnetic spectrum between 0.3 and 3Hz (possibly ionospheric Alfvén waves acting through the geomagnetic field), and short period seismic resonances at 10.1 and 15.9Hz. All five of the sensors actually detected possible precursors of the seismic events that followed them. However, the 3-30Hz electromagnetic signals showed up in less than 10% of the recorded earthquakes. The signals of the 0.3-3Hz band were also not in all quake recordings. Broadband strobing was recorded before several earthquakes and was especially prominent for hours before a quake near Wyatt MO and Bardwell KY on 2017-03-19. The strobe issue appears to be worthy of further scrutiny. The lack of a good electric conduction mechanism between earthquake hypocenters and the earth-ionosphere cavity may be the reason why the Schumann-like emissions of the 3-30Hz band were not observed for most earthquakes. A magnetometer is a preferred next step because it could increase the percentage of quakes with observable effects. In addition, a short period seismic resonator for 20Hz is another desirable addition.