[Wien] MAE; S-O calculation

[Claudio Lee]

Dear wien-users,
What I am trying to do is that I want to calculate the
difference in energy between two magnetization
directions [001] and [111] of Fe bcc.
First step I just run it as a simple spin polarized 
calculation "runsp_lapw -cc 0.000001 -i 100", when
this step finished I use clean_lapw command to remove
unnecessary files and then second step I continue
running "initso" to prepare for S-O coupling and
finally run it with "runsp_lapw -so -cc 0.000001 -i
100". I try to test with many values of RKMAX:
7,8,9,10; EMAX=10; and number of K-point in whole BZ
=20000. If I chose these parameters were fine and
those step above were correct then the final result
must be well. Unfortunately, the calculation gives me
the difference between total energy along
magnetization directions [001], [111]
delta=E[001]-E[111] too large compare to experiment's,
and the sign of delta changes from "-" to "+" if
exchange correlation potential changes from GGA to
LSDA respectively.
So my questions are:
 1. If those steps I say above to calculate for S-O
coupling are correct?
 2. Parameters RKMAX, EMAX, etc for this system as I
choos above are good?
 3. In the total enegy, if I want to use 8 digits
(instead of 6) in the decimal part then where I can
modify it in the code wien2k?
       
If anyone has the answers please help me.
Many thanks in advance.

Claudio.



__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com