Probem with Ports/boundaries

[cnhoff]

Probem with Ports/boundaries in CST MWS

Hello,

i am trying to simulate a suspended stripline filter actually, but as i am new to CST i have some basic problems.

I tried to simulate Microstrip line (Ro4350 lossy 50 Ohm line made of Gold). I dont have bottomside metallization, as i am using electric walls on all sides. Now i first only set one port at the beginning of the line (substrate edge), the waveguide end was obviously shorted by the electric boundary when i examined the smith chart. Phase-shift was believable also.

When i set a second port at the other end of the line, i have non-believable results. S11 is lower than -60 dB and S21 neglectably low and Smith Chart is only one single matched dot. When i considerably change MS line width, results basically stay the same, although port impedance changes correctly...

I really don't know whats happening there i just do not get resonably results with 2 ports. Btw i am using the planar filter preset.



Can anyone help please?!

Chris

 

[iyami]

Re: Probem with Ports/boundaries in CST MWS

what material do you have below the microstrip line? A mistaken setting there could explain your results.

 

[cnhoff]

I have used lossy and loss-free RO4350 material in the correct thickness for this line. Metallization thickness is 35um.

 

[iyami]

without more details it is hard to say anything. But no transmission and no reflection suggests wrong parameters somewhere.

 

[volker_muehlhaus]

Re: Probem with Ports/boundaries in CST MWS

I dont have bottomside metallization, as i am using electric walls on all sides

If you use electric walls, you should increase the distance to the side walls and to the top cover.

When i considerably change MS line width, results basically stay the same, although port impedance changes correctly...

From what I remember, depending on the simulator settings, you will get S-parameters that are normalized to a fixed impedance (such as 50 Ohm) or S-parameters relative to the line impedance at the ports. I assume you have the second case. Check if there is some option to normalize the results to 50 Ohms.