sfc-convergence-error remains-even after lowering convergence criterion-iteratio

[Sabyasachi Sen]

Please find the attached .py file and related .out file for which is showing error in scf convergence as follows:
+------------------------------------------------------------------------------+
|  99 E = -695.443 dE =  6.250511e-04 dH =  1.643567e-03                       |
+------------------------------------------------------------------------------+
################################################################################
#                                                                              #
# Warning: The calculation did not converge to the requested tolerance!        #
#                                                                              #
####################################


Even after using higher damping, higher tolerance, higher history cycle, the error remains (see the attached .py and .out file). The error remains even with larger no. of iterations (100 to 120). I doubt the impact of increasing cycles or reducing other criterion does not become effective in the final output. Kindly see the attached .py file and suggest me the remedy,
(i) how do I understand that their effect is actually influencing the final output.
(ii) Kindly suggest me a way out to get rid of the said convergence problem.

Sabyasachi Sen

[Anders Blom]

Results converged to such low tolerance can probably not be fully trusted. But it looks like you are on your way to convergence, so one simple way is to increase the max number of steps to 200 or 300 and keep going. Since I have no idea what your system is, I can't say anything about how to make it converge better.