The paper studies robustness with respect to time-sampling of the energy regulation for one-dimensional sine–Gordon system. Such a problem is a new to control of invariants for hyperbolic partial differential equations (PDEs). In the absence of analytic results, the problem is studied numerically. The properties of four sampled-data algorithms are computationally studied with respect to three performance criteria. The four speed gradient algorithms are the ‘‘proportional’’, ‘‘relay’’, ‘‘adaptive-relay’’ and combined ones, by using state feedback with in-domain actuators. The three performance criteria are limit error, transient time, and threshold of stability for the sampling interval. An unexpected result is that the best performance for all three criteria was exhibited by the simplest, speed-gradient-proportional, algorithm. Simulation results are also presented for other energy tracking controllers to add insight into the parameter choice for improving the closed-loop robustness in the PDE setting over sampled-data algorithms.