Strange for G92 and spartan in optimization of dimethyleacetylene

[Peng <xiaopeng-0at0-astro.ocis.temple.edu>]

 Dear netters,
 	I encounter a stupid mistake..It must because I did something
 "stupid" wrong. could anybody kindly point out?
 	I try to optimize the structure of din=methylacetylene, however,
 no matter the 2 methyle group is anti conformation or eclipse
 conformation..strange) I do
 it always turrns out the optimized structure is a saddle point(at
 rhf/6-31g level) one of the input file for Gis attached below and another is
 just cheange D4=0.0
 	If I use Spartan4.0 to optimize it, it also turns out a saddle
 point. but we know this must be a local minium
 	Please help!
 	Thanks in advance!
 %chk=butyne
 #RHF/sto-3g opt optcyc=100 test
 run the optimization of 2-butyne
    0   1
  C
  C 1 Rz
  X 1  1.0  2  A1
  C 1  R1   3  A1  2   180.0
  X 2  1.0  1  A2  3   180.0
  C 2  R2   5  A2  1   D1
  H 4  RH1  1  AH1 3   D2
  H 4  RH2  1  AH2 7   D3
  H 4  RH3  1  AH3 7  -D3
  H 6  RH4  2  AH4 7   D4
  H 6  RH5  2  AH5 10  D5
  H 6  RH6  2  AH6 10 -D5
       variables:
  Rz=1.18
  A1=90.0
  A2=90.1
  R1=1.501
  R2=1.5001
  RH1=1.094
  AH1=109.4
  RH2=1.0941
  AH2=109.41
  RH3=1.942
  AH3=109.42
  RH4=1.09411
  AH4=109.411
  RH5=1.09412
  AH5=109.412
  RH6=1.09413
  AH6=109.412
  D1=180.0001
  D2=180.0011
  D3=120.0
  D4=180.001243
  D5=120.01
 --link1--
 %chk=butyne
 #RHF/sto-3g freq geom=checkpoint guess=read  test
 freq calculation to confirm the local minium
 0 1
 --link1--
 %chk=butyne
 #RHF/6-31g(d) fopt geom=checkpoint guess=read  test
 further optimize at 6-31g level
 0 1
 --link1--
 %chk=butyne
 #RHF/6-31g(d) freq geom=checkpoint guess=read  test
 freq calculation to confirm the local minium
 0 1
 xiaopeng
 =======================================================
 			Chemistry Graduate student
 			215-2047149(office)
 			215-5357188(home)
 			xiaopeng-0at0-astro.ocis.temple.edu
 ======================================================