[CST MWS] Inaccurace in "broadband" farfield monit

Hello all.
I noticed that when calculating far fields in CST Microwave Studio single fequency and broadband farfield monitors give different results even at the limits of frequency interval (Fmin and Fmax).
Does anybody have an idea about such behaviour?

hi there
can you explain clearly what difference you are talking about?
return loss , gain or pattern?
regards

Antenna patterns, obtained by these two monitors are different (and the difference is increasing with decreasing of frequency), and therefore gain is different. I didn't measured return loss.
Distance from the structure to open boundary is enough (>lambda/8), even for the smallest frequency.