Page 1 of 1

ERROR FEXCF:supplied exchange-correlation table is too small

PostPosted: Sat Nov 04, 2017 10:35 am
by deepashri
Dear All,

I am getting following error while optimizing a monolayer having vacuum of 12Angstrom,

###################################################################################################
WARNING: Sub-Space-Matrix is not hermitian in DAV 10
-0.249457526494363
DAV: 5 -0.146118915545E+02 0.41352E-02 0.35853E+02 1984 0.308E+01BRMIX: very serious problems
the old and the new charge density differ
old charge density: 22.00000 new 6058.80162
0.845E+03
ERROR FEXCF: supplied exchange-correlation table
is too small, maximal index : 11653
#####################################################################################################

I am using VASP-5.4 with latest pseudopotentials including kinetic energy density, as I wish to use METAGGA for band calculations later. Following is my INCAR file,

ISTART = 0
INIWAV = 1
NWRITE = 2
LWAVE = .TRUE.
PREC = medium
EDIFF = 1.0E-6
EDIFFG = -0.001

ISPIN = 2
ISMEAR = 0
SIGMA = 0.001

ICHARG = 2
LCHARG = .TRUE.

ISYM = 1

NELMIN = 4
NELM = 80
NELMDL = 0
NBANDS = 16

IBRION = 1
ISIF = 3
NSW = 500
POTIM = 0.4

ALGO = F
LDIAG = .TRUE.

LPETIM = .FALSE.
LREAL = .FALSE.
LORBIT = 1
NPAR = 2
NSIM = 28
LPLANE = .TRUE.

LELF = .FALSE.

I have tried changing various parameters, but some other errors appear, such as,

ERROR in subspace rotation PSSYEVX: not enough eigenvalues found

Could anybody post the possible source of error?

Thanks.
Deepashri.