[Pw_forum] problem with scf and nscf calculation!!

[Paolo Giannozzi]

On Mar 12, 2007, at 17:39 , Federico Iori wrote:

> Infact another "strange thing" happen: when my system is semiconductor
> (all the valence states are doubly occupied)  I can get the right  
> band structure
> both with the PW-3.0 and PW-3.2 calculation. BUT when my system  
> become metallic (inserting a substitutional impurity) and I have a  
> semi-occupied level,
> the calculation for the band dont' work anymore........!!!

I'll believe this when I'll see this. Please provide an example - ONE
THAT CAN BE RUN - showing the above phenomenon
---
Paolo Giannozzi, Democritos and University of Udine, Italy