problems with magnetic LSDA+U calculation

[loschen]

Hi,

in order to do magnetic DFT+U calculation I did try to reproduce the calculations on the antiferromagnetic (AF) NiO as given in one of the hands-on sessions available via the net. No problem for the simple LDA calculation (4_2_NiO), giving a mu_b exaclty as in the example calculation, but as soon as I try to do the AF LSDA+U calculation with the Dudarev approach (4_3_NiO), I get different results, even with the unchanged inputfiles downloaded from the vasp homepage.
For the magnetization I get a much to low/wrong value...:

# of ion s p d tot
----------------------------------------
1 -0.003 -0.002 0.234 0.228
2 0.003 0.001 -0.181 -0.177
3 0.001 -0.026 0.000 -0.025
4 0.001 -0.026 0.000 -0.025
------------------------------------------------
tot 0.002 -0.053 0.053 0.001

Does anyone has a clue what is going wrong?
(using vasp 4.6.19 on a an Intel Xeon system and also on a IBM PowerPC970)

Thanks in advance,

Christoph






<span class='smallblacktext'>[ Edited ]</span>