Hi,
1- I have the same problem when analysing microstrip p@tches. the reason of my problem was that because of radiation boundary condition the resaults converge very late (after 20 passes), but i had sett maximum of passes less than ten.
so I propose
change maximum of passes higher that 20, max delta per passe about 0.03 and refinement per passes about 10%.
2- in this condition if again solution don't convege, please send us result of "delta per pass" for all passes.
hope to sacceed.