X-mode FDTD codes, in use, suffer from long-run growth and, under relatively high levels of turbulence, are unstable. The state-of-the art kernel in use for reflectometry simulations, the Xu-Yuan kernel (XYK) [1] is not free from these setbacks. An innovative kernel proposed by Després and Campos Pinto (DPK) addressing these issues and build from scratch to be conservative is being implemented in REFMULX (IPFNs X-mode code). The results obtained with DPK show an improved long-run stability, superior to XYK. Some of the ideas developed in DPK, namely the addressing of fields within each position cell, were applied to the XYK resulting in a modified XYK (MXYK) with a much better performance than the original XYK. This new numerical advance enlarges the possibilities of X-mode simulations in the ITER cases. Although the implementation presented in this work is two-dimensional, these are generic three-dimensional kernels by construction, which makes them of the utmost importance to the 3D reflectometry codes in development.