[Pw_forum] too many bands are not converging

Dimpy Sharma dimpy.sharma at tyndall.ie
Tue Dec 22 21:41:52 CET 2009



Hi there,

I have been trying to run a nscf calculation, after finishing my scsf calculation, howvever I am getting the following error 'too many bands are not converged', what I can get from previous question that 'changing mixing beta or diagonalisation ' will run it.I have tried that but it wonnt work.I have even number of electrons , so I ignored occupation. Also due to memory problem I have decided to take less number of k points, I mean I will break my kpoints in parts and will continue my calculations. Can anybody please help me! 
&CONTROL
     calculation ='nscf'
    restart_mode ='from_scratch'
          outdir = '/'
      pseudo_dir = '/'
          prefix = 'Kpoint4'
         tstress = .true.
         tprnfor = .true.
         forc_conv_thr=2.D-4
         etot_conv_thr=1.D-8
         nstep=600
   
/
 &SYSTEM
        ibrav = 0
    celldm(1) =14.5152
          nat = 321
         ntyp = 3
      ecutwfc = 35
      ecutrho = 120
         nbnd = 742
         nosym=.true.
        smearing='gaussian'
/
 &ELECTRONS
     diagonalization ='cg'
    mixing_mode = 'plain'
    conv_thr = 1.0d-8
     mixing_beta = 0.7
/
 CELL_PARAMETERS cubic 
     1.000000000    0.000000000    0.000000000 
     0.000000000    5.800000000    0.000000000 
     0.000000000    0.000000000    5.800000000
ATOMIC_SPECIES
   Si   28.0855   Si.pz-vbc.UPF
   O    15.9994   O.pz-rrkjus.UPF
   H    1.00800   H.pz-vbc.UPF
 ATOMIC_POSITIONS angstroms
Si           7.4734   12.9443   16.0594
Si           0.9063   12.9082   11.9628
................

K_POINTS crystal
 2
 -0.500000000   -0.500000000   -0.500000000       1.00000000
 -0.460000000   -0.500000000   -0.500000000       1.00000000


Dimpy





































































































-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.democritos.it/pipermail/pw_forum/attachments/20091222/c12c1f68/attachment.htm 


More information about the Pw_forum mailing list