Ciao Lorenzo I am not sure but one problem could be that you're using cosab 0.5 (60 degrees) instead of cosab=-0.5 (120 degrees) another one (again just a guess) is your cut-off seems to be really low even for an ultrasoft pseudopotential.
anyway, you should try with this electron setting mixing_mode='local-TF', mixing_beta=0.6, I've not problem in converging it (using different pseudo though) Cheers Davide P.S. the last test you can do is to use a bit of smearing, for instance: occupations='smearing', smearing='marzari-vanderbilt' degauss=0.05 ________________________________________ ------------------------------ Message: 2 Date: Sun, 29 May 2016 21:11:49 +0200 From: Lorenzo Don? <[email protected]> Subject: [Pw_forum] CPO-27-Zn scf not converge To: PWSCF Forum <[email protected]> Message-ID: <[email protected]> Content-Type: text/plain; charset="us-ascii" Dear all can you help me or give me some tips to make scf convergence on this calculation: &control pseudo_dir='/home/utente/Scaricati/espresso-5.4.0/pseudo/' outdir = './', wf_collect = .true., verbosity= high, / &system ibrav= 5, a= 25.226, cosab= 0.5, space_group = 148, nat= 9, ntyp= 4, ecutwfc =15.0, occupations='fixed', smearing='gauss', input_dft= PBE, / &electrons mixing_mode='plain' diagonalization='david' mixing_beta = 0.01 conv_thr= 1.0e-4 electron_maxstep=500, / ATOMIC_SPECIES Zn 65.409 Zn.pbe-van.UPF C 12.010 C.pbe-van_ak.UPF O 16.00 O.pbe-van_ak.UPF H 1.00 H.pbe-van_ak.UPF ATOMIC_POSITIONS crystal_sg C 7.127430447398E-02 -2.455696438145E-01 7.991093030472E-02 C 1.197468901477E-01 -2.066445045066E-01 2.152449225957E-01 C 2.117353032357E-01 -1.117780631842E-01 2.553368090727E-01 C 1.646071563067E-01 -1.528403540257E-01 1.440875844425E-01 H 1.629458066146E-01 -1.419696182443E-01 -4.760745519579E-03 Zn -1.943259345214E-02 2.834354147594E-01 3.060064447988E-01 O 7.229455751443E-02 -2.290060932141E-01 -9.085450716092E-02 O 3.018234203718E-02 -2.947952345384E-01 1.422041978763E-01 O 2.519030126218E-01 -6.097111538485E-02 1.721974260943E-01 K_POINTS gamma Thanks a lot to help me dearly lorenzo _______________________________________________ _______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
