Check if the system geometry is set up correctly. For some systems you might need SpinPolarized calculations.Check also convergence vs k-grid and energy cutoff, use MP mixing and higher smearing if needed.
You don't have convergence from the very beginning, i.e. dDmax is not
decreasing constantly but jumping. I would go to even smaller mixing values.
- [SIESTA-L] SCF Convergence L. Li
- Re: [SIESTA-L] SCF Convergence Demetra Psiachos
- Re: [SIESTA-L] SCF Convergence Oleksandr Voznyy
- Re: [SIESTA-L] SCF Convergence Riccardo Rurali
- [SIESTA-L] SCF convergence Vasilii Artyukhov
