Mixing tag

[simoneca]

Hi all,

I am trying to build a potential energy surface, but i found problems with electronic convergence. Among previous posts I found the suggestion to set the mixing tags as:

AMIX = 0.2
BMIX = 1E-4
AMIX_MAG = 0.8
BMIX_MAG = 1E-4

This solved my problems. Moreover it is not necessary a further run from the produced WAVECAR without such tags to get a more accurate result.

Here's the question:
I am using gaussian smearing (ISMEAR=0) with SIGMA=0.20 but using these mixing options all the three values (Free energy, energy without S, and energy s-> 0) are the same. Why?
I know this is correct in the case of no smearing (tetrahedrons), but does it have any sense in my calculations? Can I trust my energies??

Thanks in advance.
Simone

[admin]

this result can of course be correct, it depends on the electronic structure of the system: If the system has a large gap, or if you do a molecule with a discrete line spectrum and a Gamma-only k-mesh, the three energies have to be the same