On Mon, 2013-07-01 at 14:35 +0200, Vincenzo Verdolino wrote: > thank you so much. I was sure that the problem was something > really "stupid"
not necessarily so. In principle, everything should work even - if the cutoff for the charge density is larger than needed - if the plane-wave basis set is too small to yield a good description of the system - if starting from random wave functions (files for most HGH PP do not contain atomic wavefunctions) - if there are unconverged eigenvalues - ... For some reason, however, wavefunctions after the first diagonalization are so bad that the charge density is not correct (notice the "negative charge": this is something that should never happen for norm-conserving PPs). It is a rather peculiar case. You should first of all try a simpler case, e.g. bcc iron or something similar OP. -- Paolo Giannozzi, Dept. Chemistry&Physics&Environment, Univ. Udine, via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222
