CCL:G: [CCL] RE: G: increasing NStep for partial optimization

 Sent to CCL by: "Robinson, James" [James.Robinson : evotec.com]
 -----Original Message-----
 > From: Robinson, James
 Sent: 07 May 2006 00:43
 To: 'CCL Subscribers'; 'owner-chemistry{=}ccl.net'
 Subject: RE: G: increasing NStep for partial optimization
  I have experienced this problem too. The work around I have used is to
 specify maxcycle with the OPT keyword and then when the run stop because
 an initial maximum number of cycles have been performed, I then simply
 insert the RESTART keyword.
 #p opt=(z-matrix,maxcycle=999) b3lyp/cc-pVDZ nosymm ..
 Job then dies after a maximum number of steps, related to degrees of
 freedom.
 Then change to #p opt=(z-matrix,restart,maxcycle=999) b3lyp/cc-pVDZ
 nosymm .. And off you go again. Hope this helps.
 James
 -----Original Message-----
 > From: owner-chemistry{=}ccl.net [mailto:owner-chemistry{=}ccl.net]
 Sent: 04 May 2006 02:54
 To: Robinson, James
 Subject: CCL:G: increasing NStep for partial optimization
 Sent to CCL by: Sengen Sun [sengensun..yahoo.com]
 In my experinece, there are often converge problems when one or two
 constraints are applied between two molecules. Your case appears to me
 to be such a case, as 100 steps are quite a big number if your system is
 not very big. A possible indication for this converge problem is
 oscillating energies in your output files.
 That is, the energy jumps back and forth without a general trend.
 I was suggested to do something to deal with this problem. But it is too
 complicated to me as I had too many systems to be examined. As I spent
 more and more time in this, I found that this problem depends on
 different software programs, and sometimes on the levels of computation
 and basis sets. Some products are simply bad doing this kind of things.
 A good product for this purpose is Spartan. I am quite amazed at its
 speed and stability, and never have had this annoying converge problem
 again since I bought it. The only relationship I have with
 Wavefunctions, Inc. is my purchase of their product - Spartan. I don't
 have any other relationships with them, neither am I an advertising
 agent for them. I just want to clarify that.
 Sengen
 -----------------------------
 Christopher.DeDobbelaere||student.ua.ac.be wrote:
  Dear ccl members,
  I'm doing some popt jobs in g03 where the constraint during the
 optimization is  the distance between two molecules. The optimization
 takes some time and stops  after 100 steps. I already increased
 maxcycles but that just increases the  cycles per scf step.
  Is there a way to increase NStep to a larger value?
  This is the error i get:
   Optimization stopped.
      -- Number of steps exceeded,  NStep= 100
      -- Flag reset to prevent archiving.
  This are the options i'm using right now:
  p b1b95/6-311+G* popt(MaxCyc=250) freq iop(1/8=10)  Thanks in advance!
  Greetings,
  Christopher
  Undergrad. Student
  University of Antwerp
 __________________________________________________
 Do You Yahoo!?
 Tired of spam?  Yahoo! Mail has the best spam protection around
 http://mail.yahoo.comhttp-:-//www.ccl.net/cgi-bin/ccl/send_ccl_messagehttp-:-//www.ccl.net/chemistry/sub_unsub.shtmlhttp-:-//www.ccl.net/spammers.txt