[Smeagol-discuss] Convergence

[Ivan Rungger]

Dear Christian,

    one reason could be incorrect leads files for left and right leads. Do
you construct the left leads spin up, and the right ones spin down? The
corresponding left leads files have to be different from the right lead
files for an antiparallel calculation.

Cheers,

 Ivan

2008/6/6 Christian Santiago Helman <helman at tandar.cnea.gov.ar>:

> Dear Ivan,
> First at all, thanks for your suggestions, let's  clarify
> > Just as a clarification: did the SIESTA calculation converge for both
> > ferromagnetic and antiferromagnetic?
> Both converge with SIESTA and also the Ferromagnetic calculation with
> SMEAGOL.
>
> I change MixSCF1 to TRUE and no present appreciable loose of charge.
> However still without converge. I paste the tail of buffer.out and maybe can
> suggest something.
> Thank you!
>
> siesta:  742   -24158.9917   -24156.9110   -24156.9110  4.65948  1.67055
> 0.00000
> siesta:  743   -24163.0123   -24163.1561   -24163.1561  3.78361  0.95474
> 0.00000
> siesta:  744   -24167.4128   -24161.8277   -24161.8277  3.66278  3.15323
> 0.00000
> siesta:  745   -24160.4453   -24161.5182   -24161.5182  3.92038  1.91812
> 0.00000
> siesta:  746   -24164.9195   -24162.7977   -24162.7977  3.81111  3.01683
> 0.00000
> siesta:  747   -24160.0780   -24159.7864   -24159.7864  4.87890  0.45763
> 0.00000
> siesta:  748   -24160.6820   -24160.1509   -24160.1509  3.83874  0.20810
> 0.00000
>
>
> 2008/6/2 Ivan Rungger <runggeri at tcd.ie>:
>
> Dear Christian,
>> > I change some siesta parameters like DM.MixingWeight( from 0.005 to
>> > 0.02) , DM.NumberPulay (from 0 to 8), DM.NumberKick (from 0 to 5)
>> > ,DM.KickMixingWeight (0.02) but nothing happened, i'm far away from
>> > the converge criteria.
>> When running smeagol always also add the "MixSCF1 True" option.
>> >
>> > Also the system loose charge, from 228 to 221, i change NEnergReal,
>> > NEnergImCircle, NEnergImLine,NPoles but the amount of charge
>> > calculated is independent of these parameters, always around 221.
>> This clearly indicates that there is something wrong, probably in the
>> way the leads are joined to the scattering region. For a 0-bias
>> calculation NEnergReal is not used, the others you can set about 32 for
>> an electronic temperature equal to room temperature.
>>
>> Just as a clarification: did the SIESTA calculation converge for both
>> ferromagnetic and antiferromagnetic?  And did smeagol converge for the
>> ferromagnetic, but not for the antiferromagnetic configuration?
>> Depending on which calculations converged and which not there might be a
>> different problem.
>>
>> Cheers,
>>
>>  Ivan
>>
>> >
>> > Any suggestions?
>> > thanks !!
>> >
>> >
>> > --
>> > Lic. Christian Helman
>> > Unidad de Actividad Física
>> > Centro Atómico Constituyentes-CNEA
>> > Buenos Aires - Argentina
>> > Tel:++54-11-6772-7102
>> > ------------------------------------------------------------------------
>> >
>> > _______________________________________________
>> > Smeagol-discuss mailing list
>> > Smeagol-discuss at lists.tchpc.tcd.ie
>> > http://lists.tchpc.tcd.ie/listinfo/smeagol-discuss
>>
>>
>> --
>> =================================================
>> Ivan Rungger,
>>
>> School of Physics and CRANN,
>> Trinity College Dublin,
>> Dublin 2,  IRELAND
>> Phone: +353-1-8968454
>> Email: runggeri at tcd.ie
>> =================================================
>>
>>
>>
>
>
> --
> Lic. Christian Helman
> Unidad de Actividad Física
> Centro Atómico Constituyentes-CNEA
> Tel:++54-11-6772-7102
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.tchpc.tcd.ie/pipermail/smeagol-discuss/attachments/20080610/2da17b9f/attachment.html