The ABINIT Users Mailing List ( CLOSED )

["Anglade Pierre-Matthieu" <anglade@gmail.com>]

This error message happens only with conjugate gradient based mixing (I think) so you can simply change "iscf" to 7 for instance.
I've experienced getting this message also when having too few bands for my systems. May be specifying "nband"  could help if you like to keep the CG based mixing scheme.

On 5/15/06, yshtogun@cas.usf.edu <yshtogun@cas.usf.edu> wrote:

Dear Abinit users!
I try to calculate Etotal for different lattice constant.
For some lattices constant I got ERROR in my convergence:

scfcge : ERROR -
  Potential-based CG line minimization not converged after 13 restarts.
  Action : read the eventual warnings about lack of convergence.
  Some might be relevant. Otherwise, raise nband.
-P-0000
-P-0000  leave_new : decision taken to exit ...

This is my input files:
#Definition of the unit cell
acell 33.0665018318 33.0665018318 4.2840000000 angstrom   # The size of unit cell in angstrom


#Definition of the atom types
ntypat 1          # There is only one type of atom
znucl 6           # The keyword "znucl" refers to the atomic number of the
                  # possible type(s) of atom. The pseudopotential(s)
                  # mentioned in the "files" file must correspond
                  # to the type(s) of atom. Here, the only type is Silicon.


#Definition of the atoms
natom 28          # There are 28 atoms
typat 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
                  # They all are of type 1, that is, Carbon.

xangst            # This keyword indicate that the location of the atoms
                  # will follow, one triplet of number for each atom
                  # Triplet giving the cartesian coordinates of all atoms, in Angstrom
19.2887927159  16.5332509159  0.0000000000
19.0159083159  17.7288357159  0.7140000000
18.2513031159  18.6876202159  0.0000000000
17.1464166159  19.2197055159  0.7140000000
15.9200852159  19.2197055159  0.0000000000
14.8151987159  18.6876202159  0.7140000000
14.0505935159  17.7288357159  0.0000000000
13.7777091159  16.5332509159  0.7140000000
14.0505935159  15.3376661159  0.0000000000
14.8151987159  14.3788816159  0.7140000000
15.9200852159  13.8467963159  0.0000000000
17.1464166159  13.8467963159  0.7140000000
18.2513031159  14.3788816159  0.0000000000
19.0159083159  15.3376661159  0.7140000000
19.0159083159  17.7288357159  2.1420000000
18.2513031159  18.6876202159  2.8560000000
17.1464166159  19.2197055159  2.1420000000
15.9200852159  19.2197055159  2.8560000000
14.8151987159  18.6876202159  2.1420000000
14.0505935159  17.7288357159  2.8560000000
13.7777091159  16.5332509159  2.1420000000
14.0505935159  15.3376661159  2.8560000000
14.8151987159  14.3788816159  2.1420000000
15.9200852159  13.8467963159  2.8560000000
17.1464166159  13.8467963159  2.1420000000
18.2513031159  14.3788816159  2.8560000000
19.0159083159  15.3376661159  2.1420000000
19.2887927159  16.5332509159  2.8560000000


#Definition of the planewave basis set
ecut  20.0         # Maximal kinetic energy cut-off, in Hartree

#Exchange-correlation functional
ixc 1             # LDA Teter Pade parametrization

#Definition of the k-point grid
kptopt 1          # Option for the automatic generation of k points, taking
                  # into account the symmetry
ngkpt 1 1 4       # This is a 1x1x4 grid based on the primitive vectors
nshiftk 1
shiftk  0 0 0.5


#Definition of the SCF procedure
nstep 100         # Maximal number of SCF cycles
toldfe 1.0d-4     # Will stop when, twice in a row, the difference
                  # between two consecutive evaluations of total energy
                  # differ by less than toldfe (in Hartree)
diemac 12.0       # Although this is not mandatory, it is worth to
                  # precondition the SCF cycle. The model dielectric
                  # function used as the standard preconditioner
                  # is described in the "dielng" input variable section.
                  # Here, we follow the prescription for carbon.

  Does somebody have idea why for some lattices it works and for some doesn't?
  I think than my input is corect.
  Thank you in advance!

-- 
Pierre-Matthieu Anglade