<p>Dear all:<br>   I encountered the error message "wrong total_weight" in my phdos calculation with serial matdyn.x. The QE-4.0.2 was compiled with ifort-10.1.015_intal64, lam-7.1.4(and also openmpi-1.2.7) and MKL-10.0.3.020 under RedHat Enterprise Linux Server 5.2 environment. The CPU is Quad-Core AMD Opteron(tm) Processor 2356. The detail information of my calculation are as follows:<br>
   I want to obtain the phonon-dos of silicon vacancy (ground state configuration) with Si63 supercell at gamma point.  The step-by-step inputfiles are:<br>for scf calculation<br>-------------------<br> &CONTROL<br>    calculation  = 'scf'<br>
    verbosity    = 'high'<br>    restart_mode = 'from_scratch'<br>    pseudo_dir   = '/home/pub/pplib/'<br>    outdir = '/tmp/'<br>    prefix = 'Si-V1'<br>    tprnfor = .true.<br>    nstep   = 200<br>
 /<br>&SYSTEM<br>    ibrav =  1 <br>    celldm(1) = 20.70<br>    nat  = 63 <br>    ntyp = 1<br>    occupations = 'smearing'<br>    degauss     = 0.002<br>    smearing    = 'mv'<br>    ecutwfc = 34<br> /<br>
&ELECTRONS<br>    conv_thr = 1.0D-8<br>    diagonalization='david'<br>    mixing_mode  = 'plain'<br>    mixing_beta = 0.3<br> /<br>ATOMIC_SPECIES<br> Si 28.0855 Si.pbe-rrkj.UPF<br>ATOMIC_POSITIONS (crystal)<br>
.<br>Si coordinates<br>.<br>etc<br>K_POINTS <br>1<br>0.0 0.0 0.0 1.0<br>-----------------<br>for phonon calcuation:<br>----------------------<br>&INPUTPH<br> amass(1) = 28.0855<br> outdir = '/tmp/'<br> prefix = 'Si-V1'<br>
 tr2_ph = 1.0d-12<br> alpha_mix(1) = 0.1<br> iverbosity   = 0<br> reduce_io = .true.<br>! recover  = .true.<br> fildyn = 'Si-V1.dynG'<br>! epsil  = .true.<br> trans  = .true.<br>/<br>0.0  0.0  0.0<br>--------------</p>

<p>for the IFC calculation:<br>-----------------------<br>&input<br>  fildyn = 'Si-V1.dynG'<br>  flfrc  = 'Si-V1.fc'<br>  zasr   = 'crystal'<br>/<br>1 1 1<br>1<br>Si-V1.dynG<br>-----------------<br>
All the above calcualtions finished well. However, the forenamed error message appeared when I calculated the phonon dos with dense q-points.<br>for the phonon dos calculation:<br>-------------------------------<br>&input<br>
  flfrc = 'Si-V1.fc'<br>  asr = 'crystal'<br>  dos = .true.<br>  nk1=10, nk2=10, nk3=10<br>  deltaE = 0.1<br>  fldos = 'Si-V1.phdos-10-0.1'<br>  flfrq = 'Si-V1.freq-10-0.1'<br>  flvec = 'Si-V1.modes-10-0.1'<br>
  ntyp  = 1<br>  amass(1) = 28.0855<br>/<br>--------------------</p>
<p>The matdyn.x stopped and complained that<br>--------the matdyn.x outfile-----<br>     warning: symmetry operation # 34 not allowed.   fractional translation:<br>      -0.0000387  0.0000000  0.0000387  in crystal coordinates<br>
Norm of the difference between old and new effective charges:    0.00000000000000000000<br>Norm of the difference between old and new force-constants:   0.01586035815826266834<br>  0.833333333333333     <br> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%<br>
     from frc_blk : error #         1<br>     wrong total_weight<br> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%<br>     stopping ...<br>---------------------<br>So, what's wrong with it? code bug or miscompiled?Is something wrong with the input-parameters in my inputfile? I searched the mailist and found nothing could solve the problem. BTW, with the similar calculation, I got optic frequency of 514.4 cm^-1 (expt. 518 cm^-1) at gamma point using the perfect Si64 suprecell. If you would like to test the problem, the Si-V1.dynG can be found at <a href="http://www.divshare.com/download/5725901-c73">http://www.divshare.com/download/5725901-c73</a> due to the froum file limits. Please be patient for 15 seconds adversting when you download it :).<br>
Any ideas are appreciated! Thanks a lot!<br>Best regards!<br> <br>shany<br>Graduate School of Chinese Academy of Sciences </p>