[Pw_forum] starting and expecting charges are differ when using ph.x

Peng Tao ptao10b at imr.ac.cn
Mon May 28 05:11:36 CEST 2012


Dear all,

I'm trying to calculate the phonon dispersion of charged system. I set tot_charge=0.5, and phonon calculation is executed after scf run. And an error occurs as follows:



Largest allocated arrays     est. size (Mb)     dimensions
         Kohn-Sham Wavefunctions         0.04 Mb     (    286,   10)
         NL pseudopotentials             0.10 Mb     (    286,   24)
         Each V/rho on FFT grid          0.61 Mb     (  40000)
         Each G-vector array             0.06 Mb     (   8439)
         G-vector shells                 0.03 Mb     (   3989)
      Largest temporary arrays     est. size (Mb)     dimensions
         Auxiliary wavefunctions         0.17 Mb     (    286,   40)
         Each subspace H/S matrix        0.02 Mb     (  40,  40)
         Each <psi_i|beta_j> matrix      0.00 Mb     (     24,   10)
 
      The potential is recalculated from file :
      ./trat/_ph0/e_q2/e.save/charge-density.dat
  
 
  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
      from potinit : error #         1
      starting and expected charges differ
  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 
      stopping ...


What's wrong with it?
Here are my scf and phonon input files:
&CONTROL
  prefix='e',
  calculation = "scf",
  tstress=.true.
  wf_collect=.true.
  tprnfor=.true.
  forc_conv_thr=1.0D-5,
  pseudo_dir  = "./",
  outdir      = "./trat/",
/
&SYSTEM
  ibrav       = 6,
 celldm(1)=35,
 celldm(3)=0.153,
  nat         = 6,
  ntyp        = 2,
  ecutwfc = 30.80
  ecutrho = 300.0
  occupations = "smearing",
  smearing    = "gaussian",
  degauss     = 0.03D0,
  tot_charge = 0.5
/
&ELECTRONS
  mixing_beta = 0.3D0,
  conv_thr    = 1.D-10,
/
ATOMIC_SPECIES
......
ATOMIC_POSITIONS crystal
......
K_POINTS AUTOMATIC
1 1 80 0 0 0
-------------------------------------------------
-------------------------------------------------

phonon for **
&inputph
tr2_ph=1.0d-14,
prefix='e',
fildvscf='edv',
outdir='./trat/',
fildyn='e.dyn',
trans=.true.,
ldisp=.true.,
nq1=1, nq2=1, nq3=5
/

Thank you very much!

Warmest regards,
Plato Tao







More information about the users mailing list