[Pw_forum] problem with scf and nscf calculation!!

[Federico Iori]

Hi pw users!

I'm doing some calculation on metallic systems but I have several 
problems with scf and than nscf (or bands) calculation for the band 
structure calcualtion:

what happen is that the eigenvalues calculated with a SCF calculation 
and with a NSCF (PWversion =< 3.0) or BANDS (for PWversion >= 3.1)
on the same kpoints differ: the band strcture is completely different in 
the NSCF (I should say wrong,because the character of the electronic 
structure completely change) with respect the SCF calculation.
I use this smearing:

        occupations='smearing',
        smearing='gaussian',
        degauss=0.005,



Could someone suggest me something, please? ( a part to change job of 
course....)

cheers,
federico iori.