Page 1 of 1

Error EDDDAV: Call to ZHEGV failed. Returncode = 25 2 2

Posted: Wed Mar 11, 2009 1:59 pm
by wahyu
Dear all,
I have been troubled by this error.
structure: CsCe2Cl7 (WICKOFF Pna2_{1} #33)
80 atoms per unit cell.
PAWPBE with the standard Ce.POTCAR (12 valence)
I am running a SCF to get DOS.

INCAR:
PREC = Normal
ISMEAR = 0
SIGMA = 0.05
ALGO=Normal
LREAL=.FALSE.

#AMIX=0.4
#BMIX=0.001

ISIF = 3
IBRION = 2
NSW = 0
NELMDL = 12
ISPIN = 2

LMAXMIX=6

#IALGO=48
LPLANE=.TRUE.
NPAR=4
LSCALU=.FALSE.
NSIM=4
ISYM = 2
LORBIT = 10
LCHARG = .TRUE.
LWAVE = .FALSE.
#NBANDS = 24

OUTPUT:
entering main loop
N E dE d eps ncg rms rms(c)
DAV: 1 0.468655775615E+04 0.46866E+04 -0.20928E+05 13992 0.877E+02
DAV: 2 0.438411341675E+03 -0.42481E+04 -0.37086E+04 16508 0.239E+02
DAV: 3 -0.337189612854E+03 -0.77560E+03 -0.74655E+03 14728 0.104E+02
DAV: 4 -0.371778341399E+03 -0.34589E+02 -0.34407E+02 17952 0.231E+01
DAV: 5 -0.372802553578E+03 -0.10242E+01 -0.10184E+01 18464 0.372E+00
DAV: 6 -0.372829127090E+03 -0.26574E-01 -0.26476E-01 19256 0.590E-01
DAV: 7 -0.372830329967E+03 -0.12029E-02 -0.12022E-02 18768 0.107E-01
DAV: 8 -0.372830380870E+03 -0.50903E-04 -0.50879E-04 12688 0.187E-02
DAV: 9 -0.372830384838E+03 -0.39678E-05 -0.39657E-05 7472 0.707E-03
DAV: 10 -0.372830386216E+03 -0.13785E-05 -0.13776E-05 7184 0.430E-03
DAV: 11 -0.372830386928E+03 -0.71150E-06 -0.71034E-06 7168 0.295E-03
DAV: 12 -0.372830387313E+03 -0.38508E-06 -0.38525E-06 7168 0.212E-03 0.386E+01
Error EDDDAV: Call to ZHEGV failed. Returncode = 25 2 2

I have used DAV, NELMDL=12, and LMAXMIX=6. I have checked and relaxed the structure which is correct and able to converge by using Ce_3.POTCAR (ZVAL=11). Note that the Ce_3.POTCAR runs well, but I get the above error when I used the standard Ce.POTCAR. Unfortunately I have to use this Ce.POTCAR to get the f-orbital in the band structure. Any help would be greatly appreciated since I will need to deal with Ce alotinthefuture. Thx.

Error EDDDAV: Call to ZHEGV failed. Returncode = 25 2 2

Posted: Wed Mar 11, 2009 2:34 pm
by admin
the fact that the Ce_3 PPs do not give this error indicates that the f-electron of Ce actually is localized (or is there any experimental hint that it should be itinerant?). You can however try to decrease the mixing parameter (note that the error occurs at the first step of charge density update)