[Wien] Spin orbit coupling problem

[Aaron]

Dear wien2k developers and users:

   There is an error in my spin-orbit coupling calculation and I can't fix
it even after I read the wien2k FAQ. So I post this mail in forum and hope
someone give me some advices.
   I followed these steps in my calculation:
1, usual scf with hybrid (-eece)
2, initso in w2web with settings listed below:
******************************
WFFIL
4  1  0                      llmax,ipr,kpot
-10.0000   1.50000           emin,emax (output energy window)
   0.  0.  1.                 direction of magnetization (lattice vectors)
******************************
3, run wien2k with command:
runsp_lapw -so -eece -it0 -p -i 100 -ec 0.000001 -NI
    Then errors occur after lapw0 finished, with message:
******************************************
**  Error in Parallel LAPW1
**  LAPW1 STOPPED at Sun Apr 12 18:54:28 CST 2009
**  check ERROR FILES!
'SELECT' - no energy limits found for L= 1
'SELECT' - E-bottom   -1.03000   E-top -200.00000
......
******************************************
    I have checked with all uplapw1_*.error files. They all include the
lines:
'SELECT' - no energy limits found for L= 1
'SELECT' - E-bottom   -1.03000   E-top -200.00000
    I found two explanations in FAQ. One says "Usually in such cases the
mixing factor in MIXER was too large and too large charge fluctuations
occured (You may see this using 'grep :DIS case.scf' or 'grep :NTOxx
case.scf')". I checked with my DIS in case.scf and found these:
......
DIS  :  CHARGE DISTANCE       ( 0.0000009 for atom    2 spin 2)
 0.0000011
......
DIS  :  CHARGE DISTANCE       ( 2.7457405 for atom    2 spin 2)
 2.7081582
......
    Is this the phenomenon of too large charge fluctuations?
    Another explanation contribute the problem to that EF got quite large. I
found the highest EF result of "grep :FER case.scf" in my calculation is
1.09. But I don't know if it is too large for the default E-parameters (0.30
Ry).
    I have tried changing E-parameters in case.in1 to 0.50 Ry and the mixing
factor in case.inm to 0.1, and do the calculation again. But the error kept
showing up.
    Could anyone give me some suggestions? Thanks in advance.

Best regards,
Aaron
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20090415/c0cb475f/attachment.html