Thanks Deyu Lu - it looks like
1) the converter fhi2ups might have been changed in the last 5 years, but now things work properly 2) we should remove O.pbe-mt.UPF from the pseudo table, and replace it with a O.pbe-mt-new.UPF, that is your O.cpi.UPF. Paolo, what do you think? As a side comment - wonderful work. This is really precious. nicola On 6/28/11 3:48 PM, Deyu Lu wrote: > Thank Nicola and Paolo for the reply. I did some further tests. > The oxygen psp (both fhi form and input file) was downloaded from > http://www.abinit.org/downloads/psp-links/gga_fhi. In the previous > test, I converted fhi to UPF directly. Following Nicola's suggestion, > I regenerated the cpi file with 08-O.GGA.ini using fhi98pp. The output > (O.cpi) is converted to UPF using fhi2upf. The mesh of the new psp > (O.cpi.UPF) starts from 7.81250000000E-04, the same as 08-O.PBE.fhi.UPF, but > is slightly different from 7.81252980657E-04 in > O.pbe-mt.UPF. The resulting lattice constant is 4.633 A, consistent with what > I found before. > > In the attachment, I include the input/output using O.pbe-mt.UPF (2.d) and > O.cpi.UPF (10.d). > > Best > Deyu -- ---------------------------------------------------------------------- Prof Nicola Marzari Department of Materials University of Oxford Chair of Materials Modelling Director, Materials Modelling Laboratory nicola.marzari at materials.ox.ac.uk http://mml.materials.ox.ac.uk/NM