[Pw_forum] Using multiple s shells

[Stephen Daunheimer]

Dear Users,
I am trying to make an ultrasoft pseudopotential for neutral K and I 
would like to include the 3s, 3p and 4s shells in the valence.  I saw 
that the US pseudopotential for K the pwscf website was made this way.  
I keep getting an error though that says "too many beta functions" and I 
am assuming that this has to do with having two s shells.  I have tried 
setting lloc to -1 thinking that it might help but it did not. This is 
what my input file looks like:

&input
       title='K',
       zed=19.0,
       rel=1,
       beta=0.5,
       xmin=-7,
       lsd=0,
       rlderiv=2.5,
       eminld=-20.0,
       emaxld=20.0,
       isic=0,
       config='[Ar] 4s1',
       nld=2,
       iswitch=3,
       file_logderae='K.PZ.AELOG',
       dft='PZ',
/
&test
       pseudotype=2,
       nconf=1,
       file_wavefunctionsps='K.PZ.PP',
       configts(1)='3s2 3p6 4s1',
/
&inputp
       lloc=-1,
       rho0=1d-2,
       file_pseudopw='K.US.PZ.RRKJ3',
       file_logderps='K.PZ.PPLOG',
       zval=9.d0,
/
6
3S  1  0  2.00  0.00  1.0800  1.2000
3S  1  0  0.00  2.30  1.0800  1.2000
3P  2  1  6.00  0.00  1.2800  1.3500
3P  2  1  0.00  1.50  1.2800  1.3500
4S  3  0  1.00  0.00  3.9800  4.2000
4S  3  0  0.00  0.50  3.9800  4.2000

Thanks for any help.

Best,
Stephen Daunheimer