[Pw_forum] PAW-pseudo potential in splittet phonon calculation

[Thomas Gruber]

Dear all,

I splittet a phonon calculation with start_q, last_q and start_irr & 
last_irr into each representation. Everything worked perfectly fine with 
QE 4.3.1 for my system when I use *.pz.vbc-pseudopotential. But I wanted 
to use PAW-pseudopotentials (which I created with atomPAW) and when I 
use the same splitting I get differences in the frequencies. I followed 
the GRID_example for the AlAs system and got the following differences 
in the frequencies:
without splitting:
      omega( 1-3) =      -0.420632 [THz] =     -14.030759 [cm-1]
      omega( 4-6) =      22.373134 [THz] =     746.287424 [cm-1]
with splitting:
      omega( 1-3) =       0.225621 [THz] =       7.525921 [cm-1]
      omega( 4-6) =      22.384860 [THz] =     746.678569 [cm-1]

Did I copied the wrong files for each job or is the a problem with the 
code when using PAW-pseudopotentials?

Regards,

Thomas