The Automated Planet Finder (APF) is a facility purpose-built for the discovery and characterization of extrasolar planets through high-cadence Doppler velocimetry of the reflex barycentric accelerations of their host stars. Located atop Mt. Hamilton, the APF facility consists of a 2.4-m telescope and its Levy spectrometer, an optical echelle spectrometer optimized for precision Doppler velocimetry. APF features a fixed format spectral range from 374 nm - 970 nm, and delivers a "Throughput" (resolution * slit width product) of 114,000 arc-seconds, with spectral resolutions up to 150,000. Overall system efficiency (fraction of photons incident on the primary mirror that are detected by the science CCD) on blaze at 560 nm in planet-hunting mode is 15%. First-light tests on the RV standard stars HD 185144 and HD 9407 demonstrate sub-meter per second precision (RMS per observation) held over a 3-month period. This paper reviews the basic features of the telescope, dome, and spectrometer, and gives a brief summary of first-light performance.

We monitored 13 moderate luminosity active galactic nuclei at z=0.36 to measure flux variability, explore feasibility of reverberation mapping, and determine uncertainties on estimating black hole mass from single-epoch data. Spectra and images were obtained with approximately weekly cadence for up to 4 months, using the KAST spectrograph on the 3-m Shane Telescope. In broad band we detect peak-to-peak variations of 9-37% and rms variations of 2-10%. The observed flux variability in the g' band (rest-frame 2800-4000\AA) is consistent with that in the r' band (rest-frame 4000-5200\AA), but with larger amplitude. However, after correcting for stellar light dilution, using Hubble Space Telescope images, we find nuclear variability of 3-24% (rms variation) with similar amplitudes in the g' and r' bands within the errors. Intrinsic flux variability of the H$\beta$ line is also detected at the 3-13% level, after accounting for systematic errors on the spectrophotometry. This demonstrates that a reverberation mapping campaign beyond the local universe can be carried out with a 3-m class telescope, provided that sufficiently long light curves are obtained. Finally, we compare the H$\beta$ FWHM measured from mean spectra with that measured from single-epoch data, and find no bias but an rms scatter of 14%, mostly accounted for by the uncertainty on FWHM measurements. The propagated uncertainty on black hole mass estimates, due to the FWHM measurement errors using low S/N (10--15 per pixel) single-epoch spectra, is 30%.