[DFTB-Plus-User] Question regarding H data in SK files

[Reinaldo Pis Diez]

Dear folks,

I was reading in some detail various papers related to the 
development of the DFTB method, including the one of Foulkes and 
Haydock relating DFT to TB models.

I have now a question concerning the tabular data representing 
different Hamiltonian contributions to the DFTB energy, that is Hss, 
Hsp, etc.

Formally, there is a reference density, n0, to be used to obtain the 
above terms. Then, a small fluctuation (dn0) that relates the true 
density, n, with n0 is used to develop the SCC terms. Please, 
correct me if I'm wrong.

My question is, How are the different Hamiltonian terms obtained? 
Are they self-consistent or not? I understand they do not enter the 
self-consistency within the DFTB model but it is not clear for me if 
tabular data in SK files are obtained from SCF Kohn-Sham like 
calculations of from non-SCF KS ones.

Sorry for the long mail. My best regards,

Reinaldo