We report the results of simultaneous Chandra and RXTE observations of the Seyfert 1 galaxy Mkn 509. We deconvolve the broad and narrow Fe-K emission-line components for which we measure rest-frame equivalent widths of 119/-18 eV and 57/-13 eV respectively. The broad line has a FWHM of 57,600((sup 14,400)(sub -21,000)) km/s and the narrow line is unresolved, with an upper limit on the FWHM of 4,940 km/s. Both components must originate in cool matter since we measure rest-frame center energies of 6.36((sup 0.13)(sub -0.12)) keV and 6.42/-0.01 keV for the broad and narrow line respectively. This rules out He-like and H-like Fe for the origin of both the broad and narrow lines. If, as is widely accepted, the broad Fe-K line originates in Thomson-thick matter (such as an accretion disk), then one expects to observe spectral curvature above approximately 10 keV, (commensurate with the observed broad line), characteristic of the Compton-reflection continuum. However our data sets very stringent limits on deviations of the observed continuum from a power law. Light travel-time delays cannot be invoked to explain anomalies in the relative strengths of the broad Ferry line and Compton-reflection continuum since they are supposed to originate in the same physical location. We are forced to conclude that both the broad and narrow Fe-K lines had to originate in Thomson-thin matter during our observation. This result, for a single observation of just one source, means that our understanding of Fe K line emission and Compton reflection from accreting X-ray sources in general needs to be re-examined. For example, if an irradiated accretion disk existed in Mkn 509 at the time of the observations, the lack of spectral curvature above approximately 10 keV suggests two possibilities. Either the disk was Thomson-thick and highly ionized, having negligible Fe-K line emission and photoelectric absorption or the disk was Thomson-thin producing some or all of the broad Fe-K line emission. In the former case, the broad Fe-K line had to have produced in a Thomson-thin region elsewhere. In both cases the predicted spectral curvature above approximately 10 keV is negligible. An additional implication of our results is that any putative obscuring torus in the system, required by unification models of active galaxies, must also be Thomson-thin. The same applies to the optical broad line region (BLR) if it has a substantial covering factor.