[Pw_forum] troubleshooting cdiaghg-have checked faq

[mohnish pandey]

Dear Vijaya,

Your 'mixing_beta' is to high. Try reducing it. Generally the default value
which is 0.7 is good. I see no reason to use 0.9 as you are using. 'cg' is
quite robust but its slower than 'david' so you can use 'david' as
diagonalization scheme for faster calculations. But try the first
suggestion for mixing parameter (May be 0.5 will be good)

On Tue, Jan 8, 2013 at 1:57 AM, vijaya subramanian <vijaya65 at hotmail.com>wrote:

> Hi
> I am trying to run a simple scf calculation on a polymer crystal.
> The system has already been studied in a paper in PRL.  I am using the
> input parameters (not all input parameters are specified in the paper)
> as closely as possible as well as atomic coordinates from the paper.
> I have looked at the structure and its periodic images using Xcrysden and
> dont see any problems with my pwscf input file.  I am using pbe-Vanderbilt
> ultrasoft pseudopotential provided in the QE website for C, F and H.
>
> I am using QE version 5.0.1 on cygwin and inbuilt libraries, ie. I am
> using the algebra packages provided by
> QE.  I get the cdiaghg error and have eliminated the following causes for
> the error:
> 1) Atomic coordinates are correct-I dont see any  contacts that are too
> close.
> 2) Vanderbilt pseudo potential is used in the paper (though they dont
> specify where they got it) -I am assuming
> the versions provided on the QE website are equally reliable.
> 3) I have used diagonalization=cg and still have the problem.
>
>
> So is there any other reason I am getting this error?
>
> My input parameters are as follows:
>   calculation  = "scf",
>   prefix       = "pvf2",
>   pseudo_dir   = "/home/vijaya/espresso-5.0.1/pseudo",
>   outdir       = "/home/vijaya/tmp",
> /
> &SYSTEM
>   ibrav     = 0,
>   nat       =24,
>   ntyp      = 3,
>   ecutwfc   = 35.D0,
>   ecutrho   = 340.D0,
> /
> &ELECTRONS
>    diagonalization = 'cg'
>   conv_thr    = 1.D-5,
>   mixing_beta = 0.9D0,
> /
> &IONS
> /
> CELL_PARAMETERS bohr
> 16.2138 0. 0.
> 0. 9.2785 0.
> 0.  0. 4.8376
> ATOMIC_SPECIES
> C 12.011 C.pbe-van_bm.UPF
> F 18.998 F.pbe-n-van.UPF
> H 1.00794  H.pbe-van_bm.UPF
> ATOMIC_POSITIONS {angstrom}
> C        0.000000000   0.000000000   0.000000000
> C        0.000000000   0.854340000   1.280000000
> C        0.000000000   0.000000000   2.559999997
> C        0.000000000   0.854340000   3.839999997
> F        1.081080000   1.743050000   1.280000000
> F       -1.081080000   1.743050000   1.280000000
> F        1.081080000   1.743050000   3.839999997
> F       -1.081080000   1.743050000   3.839999997
> H        0.900900000  -0.608840000   0.000000000
> H       -0.900900000  -0.608840000   0.000000000
> H        0.900900000  -0.608840000   2.559999997
> H       -0.900900000  -0.608840000   2.559999997
> C        4.290000000   2.455000000   0.000000000
> C        4.290000000   3.309340000   1.280000000
> C        4.290000000   2.455000000   2.559999997
> C        4.290000000   3.309340000   3.839999997
> F        5.371080000   4.198050000   1.280000000
> F        3.208920000   4.198050000   1.280000000
> F        5.371080000   4.198050000   3.839999997
> F        3.208920000   4.198050000   3.839999997
> H        5.190900000   1.846160000   0.000000000
> H        3.389100000   1.846160000   0.000000000
> H        5.190900000   1.846160000   2.559999997
> H        3.389100000   1.846160000   2.559999997
>
>
> K_POINTS {automatic}
> 1 2 2 0 0 0
>
> Thanks in Advance
> Vijaya
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>



-- 
Regards,
MOHNISH,
-----------------------------------------------------------------
Mohnish Pandey,
PhD Student,
Center for Atomic Scale Materials Design,
Department of Physics,
Technical University of Denmark
-----------------------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20130108/94cd9590/attachment.html>