CCL:G: BSSE problem

 Sent to CCL by: angelo vargas [vargas*_*chem.ethz.ch]
 Dear cclers
I have a problem when I try to calculate bsse for the interaction between trifluoroacetone and trimethilammonium ion. The calculation for the trimethylammonium ion works but when I try to calculate the same for trifluoroacetone it doesn't. 
 the input is
 $RunGauss
# MP2/Aug-cc-pVDZ Opt=Z-matrix Guess=Indo SCF=tight massage gfinput gfprint IOP(6/7=3) pop=full 
 Calcolo
 !
 !
 !
 !interaction between  trimethylammonium ion and trifluoro acetone
 !
 0 1
  c
  c   1 cc2
  c   1 cc3        2 ccc3
  f   2 fc4        1 fcc4         3 dih4
  f   2 fc5        1 fcc5         3 dih5
  f   2 fc6        1 fcc6         3 dih6
  o   1 oc7        2 occ7         5 dih7
  h   3 hc8        1 hcc8         7 dih8
  h   3 hc9        1 hcc9         7 dih9
  h   3 hc10       1 hcc10        7 dih10
  n   7  2.82266   1  132.681     3 -179.82
  c  11  1.49578   7  113.497     1  -63.89
  c  11  1.49577   7  113.175     1  64.357
  c  11  1.49649   7   95.056     1 -179.82
  h  11  1.04005   7   11.518     1   -0.61
  h  12  1.09746  11  108.708    14  57.964
  h  12  1.09826  11  108.484    14  -61.84
  h  12  1.09644  11  108.752    14 178.067
  h  13  1.09645  11  108.751    12  56.946
  h  13  1.09826  11  108.485    12  -63.15
  h  13  1.09746  11  108.706    12 177.052
  h  14  1.09826  11  108.571    12  62.516
  h  14  1.09720  11  108.685    12  -57.60
  h  14  1.09720  11  108.683    12 -177.37
 cc2         1.550892
 cc3         1.493784
 ccc3        116.263
 fc4         1.362949
 fcc4        110.618
 dih4        179.733
 fc5         1.346107
 fcc5        110.533
 dih5         60.178
 fc6         1.346139
 fcc6        110.499
 dih6        -60.739
 oc7         1.230561
 occ7        118.122
 dih7       -119.819
 hc8         1.095608
 hcc8        110.000
 dih8         -0.288
 hc9         1.102232
 hcc9        109.199
 dih9        121.322
 hc10        1.102172
 hcc10       109.241
 dih10      -121.963
 11  Nuc 0.0
 12  Nuc 0.0
 13  Nuc 0.0
 14  Nuc 0.0
 15  Nuc 0.0
 16  Nuc 0.0
 17  Nuc 0.0
 18  Nuc 0.0
 19  Nuc 0.0
 20  Nuc 0.0
 21  Nuc 0.0
 22  Nuc 0.0
 23  Nuc 0.0
 24  Nuc 0.0
 The output ends this way:
 Requested convergence on RMS density matrix=1.00D-08 within 128 cycles.
  Requested convergence on MAX density matrix=1.00D-06.
  Requested convergence on             energy=1.00D-06.
  No special actions if energy rises.
  SCF Done:  E(RHF) =  -488.590793281     A.U. after   13 cycles
              Convg  =    0.9757D-08             -V/T =  2.0018
              S**2   =   0.0000
ExpMin= 2.97D-02 ExpMax= 1.47D+04 ExpMxC= 5.03D+02 IAcc=3 IRadAn= 5 AccDes= 0.00D+00 
  HarFok:  IExCor= 205 AccDes= 0.00D+00 IRadAn=         5 IDoV=1
  ScaDFX=  1.000000  1.000000  1.000000  1.000000
Warning: off-atom basis functions, so minimal bfn integration tests in XC quadrature. 
  Range of M.O.s used for correlation:     8   370
  NBasis=   370 NAE=    28 NBE=    28 NFC=     7 NFV=     0
  NROrb=    363 NOA=    21 NOB=    21 NVA=   342 NVB=   342
  **** Warning!!: The largest alpha MO coefficient is  0.70613767D+02
  Disk-based method using ON**2 memory for 21 occupieds at a time.
  Estimated scratch disk usage=  2073594272 words.
  Actual    scratch disk usage=  2072878496 words.
  JobTyp=1 Pass  1:  I=   8 to  28 NPSUse=  1 ParTrn=F ParDer=F DoDerP=T.
  Logic failure, MDV=     4335045 NBuf=          -3 but FulOut=F.
Error termination via Lnk1e in /usr/local/api/g03/l906.exe at Sun Nov 6 17:15:35 2005. 
  Job cpu time:  0 days  1 hours 16 minutes 57.1 seconds.
File lengths (MBytes): RWF= 15900 Int= 0 D2E= 0 Chk= 4 Scr= 1 
 Segmentation fault (core dumped)
Note that I am using the massage keyword with G03. I am relaxing the trifluoroacetone in presence of the basis of the trimethyl ammonium ion.
The same calculation with B3LYP 6-31Gdp ends all right. Other similar calcualtions have converged when using guess=indo, but this one does not. 
 Any help is appreciated
 Angelo