How to simulate single/coupled ring resonators in HFSS?

[godsound]

At present, I have been trying to simulate single/coupled open-loop resonators in HFSS. But I encounter several problems on how to set ports on the coupled open-loop resonators and input/output tapped single open-loop resonator in HFSS for EM simulation.

For coupled resonators, when I try to find a relationship between spacing "s" and coupling coefficient "k" by simulating them in HFSS, should I add extra tapped input/output line onto the side that is opposite to the coupled side and then set a de-embedding length which is equal to its length so as to get rid of the parasitic effect it adds? or is there some other better way to do it? If the former way is ok, where should I connect input/output port onto the resonator? Is the input/output location related with the offset "d" between two coupled resonators when they are not positioned on same horizontal line?

Besides, for the tapped single input/output resonator, how long should I set for the tapped input/output line? Is there any requirement or guidance on that?

I am really in trouble with those EM simulation sets for open-loop resonatorsin HFSS. Would anyone please do me this favor and help me out? Thanks.

 

[logbook]

ANSOFT support should help out. If you pay the same fees as we do annually then you might as well get value for money from the support/apps people. I was at one of their training days recently (good value for money) and the trainer certainly seemed to know how to make the tool work.

As for your posting, it is not clear if you are going in via transmission line or waveguide (just as a for-example). I doubt that your question is simple to resolve without a higher level of detail.

 

[godsound]

Thanks logbook!

In fact I sent my question to Ansoft support but didn't get the right answer. Probably as you said I should give more detail.

We know coupled open-loop structures are nowadays used in band pass filter design. when we simulate two coupled open-loop structures in HFSS, we should set the ports to get the S-parameter. But for open-loop structure, I have no idea where I can put the port onto the open-loop ring. So my question is about the locations of the ports on this two coupled open-loop structure. Anyone can help me? Thanks.

 

[yechi]

I'm investigating the same topic. As much as I know ,for the coupling coefficient (k) to be extracted you should couple each of the two ports weakly to each of the two resonators.

 

[godsound]

Thanks yechi!

As what you said, it would be fine if I put the two ports as far as possible.

But I still get one question: since a tapped line should be used with one end connecting to port while the other end connecting to the resonator, is the position of the line also affect the resonators performance? I mean in which way it is like a structure comprised of two tapped open-loop resonators. And from one tapped open-loop, we know the positition of the tapped line really does affect the performance.

I think using de-embedding selection in HFSS may remove this effect. And I will do some simulations. But if anyone clearly know this or has done before, would you please share your experience with me? Thanks.

 

[yechi]

Dear godsound,
You should distinct between two operations :
(1) Extracting the external Q, Where you should use tapped lines if you're going to use them in your design.
(2) Extracting the k (coupling coefficient
) ,where you should put the ports as to not load the resonators.

 

[godsound]

Thanks yechi!

The second opeation you said is the rigth one which makes me stuck a lot - extracting the k, how or where do you put the ports in the simulation not to load or weakly couple the resonators? I really no idea about this.

Thanks!

 

[yechi]

Hi,
I found one article that explains his procedure for determining the k. It states that "Two peaks specified by the two resonators were tuned to be equal height and the valley (between them) was set to be at least 10 dB lower"
The article is : "Performance of high-temperature superconducting band-pass filters with high selectivity for base tranceiver applications of digital cellular communication systems" .In general the sniffing lines should be at the same impedance as that of the resonators and so are the ports.

 

[godsound]

Thanks buddy!