Stefano Baroni wrote: >> >>> conv_thr = 1.0d-2 >> >> too small for meaningful results > > you mean "too large", Paolo, don't you? definitely! Paolo -- Paolo Giannozzi, Democritos and University of Udine, Italy