CCL:G: Constraints?

 Sent to CCL by: "David  Close" [closed(~)]
   I'm having a problem imposing constraints on a series of stacked molecules.
 I'm trying to look at the influence of neighbors in a problem involving a
 crystal structure.  When I get up to three molecules in a stack the middle
 molecule tends to move away from the stack.  I need to impose constraints, but I
 keep getting a message "Curvilinear step not converged.  Error in
 constraints".  Thuis is in GAussian.  On another system the program seems
 to fuss about symmetry?
   My questions is, what calculations are being performed to decided whether or
 not a constraint is suitable?  If one knows what calculations were being
 perfomred, is it possible to add suitable constraints? It seems as if a few bond
 constraints work, but when I add constrained angles, the curvilinear message
 appears after 100 "tries".