[Pw_forum] ld1 & search of ghost states

[Paolo Giannozzi]

Latévi Max LAWSON DAKU wrote:

> At this point, Paolo: the distinction you make between
> NC and NC pseudopotentials

did I?

> add a bit more to my confusion.

In norm-conserving PPs, the inner pseudo-state is nodeless,
so the outer pseudo-state must have one node in order to be
orthogonal to the previous. With ultrasoft PP a generalized
orthogonality constraint - with an overlap matrix - is enforced.
In principle, this does not require that the soft parts of
the pseudo-states have a different number of nodes. This is
why I pointed out that my statement was valid only for
norm-conserving PPs

Paolo
-- 
Paolo Giannozzi, Democritos and University of Udine, Italy