[molpro-user] huge gradient norm error

[Lukáš Félix Pašteka]

Hi,
im trying to optimize geometry trough simple HF-SCF, but for some starting geometries (all of which are close except one bond length) it fails with error complaining about large gradient norm:

 Norm of gradient contribution is huge!
 Probably due to wrong coordinates.

 ERROR IN QEXIT: ACTIVE=^T  CALLED WITH

 GA ERROR fehler on processor   0

Does anybody know, what exactly does this mean and what could be wrong?
Thanks in advance

Lukas Pasteka

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20080531/096a015e/attachment.html>