SCF convergence problem when using COSMO + Symmetry

[Dr_Nick]

Dear all,

I am having trouble converging serial Turbomole 6.4 BP86/def2-SV(P) ridft calculations on a series of C3 symmetric molecules, when using COSMO at the same time:

Code: [Select]

ITERATION  ENERGY          1e-ENERGY        2e-ENERGY     NORM[dD(SAO)]  TOL
 129  -1032.8300891532    -4187.4823349     1753.6876868    0.915D-01 0.504D-10
                            Exc = -126.028852111     Coul =  1879.71653890
                              N = 137.99588874
                          Ediel =-0.152851784008E-01
                            current damping = 0.300

          Norm of current diis error: 0.10683
          max. resid. norm for Fia-block=  9.914D-04 for orbital     13e
          max. resid. fock norm         =  1.135D-03 for orbital     13e
          irrep e   : virtual orbitals shifted by    0.07967
 WARNING
 Symm. of e-repr. not preserved in        4943 cases
 Max deviation in the symm. blocks:  1.665733858134285E-003
 WARNING
 Antisymm. of e-repr. not preserved in        4946 cases
 Max deviation in the antis. blocks:  1.136402495277015E-003
 WARNING
 Symm. of e-repr. not preserved in        4950 cases
 Max deviation in the symm. blocks:   979.917033488670
 WARNING
 Antisymm. of e-repr. not preserved in        4950 cases
 Max deviation in the antis. blocks:   613.185325593434
COSMO: using sym. in pot. calc.

This reminds me of a similar and probably related problem reported in this forum:
http://www.turbo-forum.com/index.php?topic=377.0

Without symmetry the calculation converges just fine in less than 15 steps. I already tried to play with damping factors and starting from a calculation without solvation, albeit without success. Any suggestions?

Thanks in advance