[Pw_forum] Eigenvalues not converging

[Lorenzo Paulatto]

In data 18 agosto 2009 alle ore 11:39:31, Samir Ranjan Meher  
<samirmeher at gmail.com> ha scritto:
> Any suggestions??? My SCF input file is as follows:-

Dear Samir,
I've had look at your input file: it looks like you've put the atoms very  
close together. For example, Nitrogen and Gallium atoms are only 0.6  
Angstrom apart. Furthermore, it looks like a surface slab geometry, which  
is not what you said you're doing.

I suspect you've done some simple mistake in your input file: celldm(1)  
must be in *bohr* units, while celldm(3) is in units of celldm(1)  
(celldm(3)*celldm(1) is in bohr).

If the atoms are overlapping too much the ultrasoft generalized  
orthogonality conditions breaks down, often causing problem during  
diagonalization.

My advice is to recheck and fix your input. You can use xcrysden  
<http://www.xcrysden.org> to visualize it.

Remember reset the diagonalization method to davidson, using  
conjugate-gradient is probably unnecessary once the input is correct.

best regards

-- 
Lorenzo Paulatto
SISSA  &  DEMOCRITOS (Trieste)
phone: +39 040 3787 511
skype: paulatz
www:   http://people.sissa.it/~paulatto/

     *** save italian brains ***
  http://saveitalianbrains.wordpress.com/