Hello all, I have done a phonon calculation for a fully relaxed system with QE 4.3.2 and everything went fine. Then I increased the volume (~3%) and did a relax calculation:
&control calculation = 'relax' restart_mode='from_scratch', tstress = .true. tprnfor = .true. pseudo_dir = '/home/gruber/software/Espresso/espresso-4.3/pseudo/', outdir = './' forc_conv_thr = 1e-4 etot_conv_thr = 1e-5 wf_collect = .TRUE. / &system ibrav= 0, A = 1, nat= 104, ntyp= 2, ecutwfc =70, occupations='smearing', smearing='methfessel-paxton', degauss=0.001 / &electrons conv_thr = 1.0d-8 / K_POINTS automatic 4 4 4 0 0 0 Then I split my phonon calculation into several calculations: &inputph tr2_ph=1.0d-15, recover=.true. ldisp=.true., nq1=4,nq2=4,nq3=4, start_q=8, last_q=8, start_irr=27, last_irr=28, amass(1)=28.086, amass(2)=6.914, alpha_mix(1)=0.1, / For the first q-point it works, but for the last 2 q-points (maybe more) I get the following error message: Band Structure Calculation Davidson diagonalization with overlap %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from cdiaghg : error # 868 diagonalization (ZHEGV*) failed Can someone tell me why and what I should do to get my representations? Sincerely, Thomas Gruber -- Thomas Gruber Institut f?r Theoretische Physik TU Bergakademie Freiberg Leipziger Str. 23, 09599 Freiberg, Germany Tel: +49 (0)3731 392006 Email: thomas.gruber at physik.tu-freiberg.de