Illustration of the issue with setting the zero-point of the logarithmic potential to Φ₀ = 0 in Equation (4), as is assumed in the original code. As an example, we use the best-fit mass model for NGC 1453 in Liepold et al. (2020) with a logarithmic dark matter halo of R_c = 15 kpc and V_c = 633km s⁻¹. The ratio of the potential energy to the maximum kinetic energy is plotted for this halo (dotted), halo plus stars (dotted–dashed), and all three mass components (dashed). When this ratio is much larger than one, as is shown for a large range of radii, even large errors in the kinetic energy would have little effect on the total energy. Energy conservation is therefore effectively not enforced in the original code for a logarithmic potential. The solid line shows the same ratio with all three mass components included, but with the halo zero-point set according to Equation (6).