dear Adam Narbudowicz . thats right. i am sumulating siw (substrate integrated waveguide) phase shifter. my structure is Ultra wide band. i reach to different answers in each of the mentioned software.!!!

I work with SIW , in all my designs experience HFSS result agree well with measurements. of course you can simulate the possible fabrication error and tolerance with HFSS (for example the permittivity error -+5%). which type of calibration you use?

i reach to my goal by cst but not by hfss. i want to fabricate my phase shifter, it costs so much for me. if the measurements not be matched with cst results so i wasted my money!!! i change the meshing but see the same results.

you should verify the convergence

sir Tarek Djerafi , what kind of solver you use for your siw,s. time domain, frequency solver???my frequency range is 26-40 GHZ. what is the best settings for my meshing??

sir Tarek Djerafi , i work on some IEEE papers which all of theme are written by senior members. they use tapered transition to feed the siw. but the width of their microstrip line isnt matched with normalized impedance(50 ohm)??

for example for duroid/5880 , thickness 0.508mm , Er=2.2 the width must be 1.8mm but the papers do not apply this rule!!!

i am confused