If the problem is with floating point precision / overflow then there's no modding this, you need to recompile the engine with double precision.