<DIV>Dear user,</DIV> <DIV> I posted this question earlier but <BR> made a mistake (I quoted the </DIV> <DIV> mixing parameter as 0.0 instead </DIV> <DIV> of 0.05) so ignore the </DIV> <DIV> previous posting. Basically </DIV> <DIV> my problem is the following:</DIV> <DIV><BR> I have observed that at the beginning <BR> of a new geometry step in mini, <BR> the charge distance (grep :DIS *.scf) is <BR> quite large. Eventually DIS is small <BR> but it takes a long time for convergence. <BR> I used a mixing factor of 0.05, PW scaling <BR> of 0.1 and min -j 'runsp_lapw -i 120 -fc 1 -I' . <BR> How do I fix the fluctuation? <BR> Also how is the friction term ETA for <BR> NEWT chosen? I'm using NEWT <BR> because it works way better than <BR> PORT and it is easier to adjust <BR> delta(i).<BR> <BR> Thanks<BR></DIV>