[Smeagol-discuss] Convergence problem at high bias

[Ivan Rungger]

Hello Nalumon,
> Do you have any suggestion? Any parameter I should change to reach the 
> convergence?
There can be different reasons. If your zero bias calculation is correct 
and converges, then in principle also as bias is applied the calculation 
should converge. However for example if you have some sharp states (i.e. 
sharp peaks in the transmission or in the DOS) that enter the bias 
window then there might be convergence problems, and also you have to 
make sure to use enough points on the real energy axis. It would be 
useful to see your fdf file to be more specific.
> To solve this problem, if I will increase ElectronicTemperature, does 
> it much affect my electrical conductance?
I would say usually I would use a temperature of 300 K, which broadens 
out the Fermi distribution by some fraction of an eV (this is also 
important if you want to use few poles, e.g. it is enough to use about 
30 poles for 300 K, or even only 16 poles, for lower temperatures you 
would need more). This usually makes convergence easier (but not 
always). The conductance is usually pretty much constant when changing 
the temperature.

Cheers,

 Ivan
>  
> Thank you in advance
>  
> Sincerely yours,
> Nalumon
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> 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
=================================================