Coupled nonlinear oscillators, e.g., Kuramoto models, are commonly used to analyze electrical power systems. The cage model from statistical mechanics has also been used to describe the dynamics of synchronously connected generation stations. Whereas the Kuramoto model is good for describing high inertia grid systems, the cage one allows both high and low inertia grids to be modelled. This is illustrated by comparing both the synchronization time and relaxation towards synchronization of each model by treating their equations of motion in a common framework rooted in the dynamics of many coupled phase oscillators. A solution of these equations via matrix continued fractions is implemented rendering the characteristic relaxation times of a grid-generator system over a wide range of inertia and damping. Following an abrupt change in the dynamical system, the power output and both generator and grid frequencies all exhibit damped oscillations now depending on the (finite) grid inertia. In practical applications, it appears that for a small inertia system the cage model is preferable.Index Terms-Power system stability, Power system transients, Power system protection, Rate of change of frequency or ROCOF, Renewable energy sources, Synchronous generators.