Hello
The energy and Hamiltonian differences dE and dH should decrease steadily towards the specified tolerance. If it's not doing the same what could be the possible error in the job. Should Poisson solver needs to be changed or the geometry should be changed?