[Pw_forum] More pseudopotential troubles

[Miguel Martínez Canales]

Dear people,

I'm now doing things step by step, so that I don't crash into the deep 
abyss of pseudopotentials, and I have found a new (for me) "error":

      ns=   3 l=   1 sum=    0.000034911 r(ikk)     1.899836304
      from gener_pseudo  : info #        -1
      chi too large beyond r_c
plus two very long arrays of values

I'll later put the input used to get this. The thing is: when I run said 
input, I don't get an error but neither do I get a pseudopotential. 
Examining atomic/gener_pseudo.f90 one finds (line 167 onwards, espresso 3.1.1):

     if (sum > 2.e-6_dp) then
         write(6, '(5x,''ns='',i4,'' l='',i4, '' sum='',f15.9, &
              & '' r(ikk) '',f15.9)') ns, lam, sum, r(ikk(ns))
         call infomsg ('gener_pseudo ','chi too large beyond r_c', -1)
         do n=ikk(ns),mesh
            write(6,*) r(n),gi(n,1)
         enddo
         stop
      endif

Seeing that it doesn't generate any pseudopotential, shouldn't it be "call 
errore(...)"? Please excuse my ignorance if this is something obvious. The 
input used to generate this was this:

  &input
    atom='Ca',
    beta = 0.2,
    rel=1,
    config='[Ar] 3d0 4s0 4p0',
    iswitch=3,
    dft='PBE'
  /
&inputp
    pseudotype=3,
    lloc=2,
    nlcc =.true.,
    rho0=0.01,
    file_pseudopw='Ca.pbe-van.UPF',
/
4
3S  1  0  2.00  0.00  1.10  1.10
3P  2  1  6.00  0.00  1.30  1.60
3P  2  1  0.00  0.05  1.30  1.60
3D  3  2  0.00  0.00  1.80  1.80

I'd also be grateful is someone explained to me where the issue is or 
pointed me to a reference, as searching the archives for "chi too large" 
returns nothing useful. And changing rc's or rcutus' changes slightly the 
value of "sum" but it isn't enough. The same with pseudization energy.

Thanks in advance,

Miguel

PS: Removing the second reference energy for the p channel and changing 
rcutus to 1.30 (thus generating a NC PP) does work... but the estimated 
cutoff is 110 rydbergs, which is a tad too high for running a phonon 
calculation.