<div dir="ltr"><div dir="ltr"><div dir="ltr"><div>Dear Iurii, </div><div><br></div><div>Thanks for getting back to me so quickly. I briefly "debugged" and two manifolds don't seem to be a problem when only one atom species is present, e.g. </div><div><br></div><div><br></div><div>nat=1</div><div>ntyp = 1</div><div>...</div><div>ATOMIC_SPECIES<br>C 12.011 C.pbe-n-kjpaw_psl.1.0.0.UPF<br>Dy  1.008 Dy.GGA-PBE-paw-v1.0.UPF<br></div><div>...</div><div>ATOMIC_POSITIONS (angstrom)<br>Dy            0       0       3<br></div><div>...</div><div><br></div><div>HUBBARD ortho-atomic<br>U Dy-5d 5<br>U Dy-6s 1<br></div><div><br></div><div>No problem with projwfc.x:</div><div><br></div><div>Lowdin Charges: <br><br>     Atom #   1: total charge =   9.9929, s =  3.9930, p =  5.9999, d =  0.0000, <br>                 spin up      =   4.9965, s =  1.9965, <br>...<br>     Spilling Parameter:   0.0007<br> <br>     PROJWFC      :      3.25s CPU      3.71s WALL<br><br> <br>   This run was terminated on:  17:39: 5  18Jun2023  </div><div><br></div><div><br></div><div><br></div><div>Then when using</div><div>nat=2, ntyp=2</div><div><br></div><div> ATOMIC_POSITIONS (angstrom)<br></div><div>Dy            0       0       3<br></div><div>C              0      0        0 </div><div>...</div><div><br></div><div>HUBBARD ortho-atomic<br>U Dy-5d 5<br>U Dy-6s 1</div><div><br></div><div>projwfc.x also works!</div><div><br></div><div>Lowdin Charges: <br><br>     Atom #   1: total charge =   9.5105, s =  3.3338, p =  5.9973, d =  0.1794, <br>                 spin up      =   5.0089, s =  1.8779, <br>                 spin up      =   5.0089, p =  2.9980, pz=  0.9982, px=  0.9999, py=  0.9999, <br>                 spin up      =   5.0089, d =  0.1330, dz2=  0.0339, dxz=  0.0495, dyz=  0.0495, dx2-y2=  0.0000, dxy=  0.0000, <br>                 spin down    =   4.5016, s =  1.4559, <br>                 spin down    =   4.5016, p =  2.9994, pz=  0.9997, px=  0.9998, py=  0.9998, <br>                 spin down    =   4.5016, d =  0.0463, dz2=  0.0463, dxz=  0.0000, dyz=  0.0000, dx2-y2=  0.0000, dxy=  0.0000, <br>...<br>     Spilling Parameter:   0.0041<br> <br>     PROJWFC      :      3.57s CPU      4.05s WALL<br><br> <br>   This run was terminated on:  17:43:40  18Jun2023  <br></div><div><br></div><div>BUT (and this might be a bug), if C appears before Dy in the ATOMIC_POSITIONS:</div><div><br></div><div>ATOMIC_POSITIONS (angstrom)<br>C             0       0       0<br>Dy            0       0       3<br></div><div><br></div><div>pw.x has no problem but projwfc.x does seem to think that I requested the 6s on the C and not the Dy.</div><div><br></div><div> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%<br>     Error in routine offset_atom_wfc (1):<br>     Mismatch between the requested and available manifolds<br> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%<br><br></div><div><br></div><div>(tested now also on the last QE from gitlab).</div><div><br></div><div>so the "hot fix" seems to be to keep the atom with the two manifolds in the front of the atoms in ATOMIC_POSITIONS. Or, what also seems to work (oddly) swap the order of species in the ATOMIC_SPECIES card (so that they match the appearance of the atom in the position card).</div><div><br></div><div>in and outputs are available here: <a href="https://drive.google.com/drive/folders/1yzwd5XiGcci_8ZCv9qWVDbuHv7Sopgr9?usp=sharing">https://drive.google.com/drive/folders/1yzwd5XiGcci_8ZCv9qWVDbuHv7Sopgr9?usp=sharing</a></div><div><br></div><div>HTH!</div><div>Chris </div><div><br></div><div><br></div><div><br></div></div></div></div>