CCL:G: Unable to achieve self-consistence

 Sent to CCL by: "Tommaso  Marcelli" [marcelli _ theochem.kth.se]
 I am having trouble in calculating frequencies with a 2-layers Oniom for a 236
 atoms system in Gaussian 03: 200 atoms AM1 and 36 atoms B3LYP/6-31G(d,p).
 Besides frequencies the system seems to work fine, I have no problems in
 geometry optimization and also transition state search with local Hessian works
 very nice using the same level of theory. This is the end of my output:
 ALL CONVERGERS ARE NOW FORCED ON
  SHIFT=1000, PULAY ON, CAMP-KING ON
  AND ITERATION COUNTER RESET
 
 """""""""""""UNABLE
 TO ACHIEVE SELF-CONSISTENCE
           DELTAE=   0.4558E+05     DELTAP=   0.1788E+01
  Error termination via Lnk1e in
 /pdc/vol/gaussian/G03RevD.01/amd64_fc3/g03/l402.exe at Thu Jul  3 00:22:30 2008.
  Job cpu time:  1 days  2 hours 14 minutes 42.3 seconds.
  File lengths (MBytes):  RWF=     48 Int=      0 D2E=      0 Chk=     13 Scr=
 1
 Command exited with non-zero status 1
 61748.03user 32734.44system 13:07:41elapsed 199%CPU (0avgtext+0avgdata
 0maxresident)k
 0inputs+0outputs (0major+940278minor)pagefaults 0swaps
 I already included maxcycle=1200 and vshift=500 in my input but the problem
 stays exactly the same. Any suggestions on how to fix this?
 Thanks in advance