filter design using ring resonators

Dear all:

Recently I read something talking about ring resonator filter design. One typical design flow is outlined as follows:
1. Characterize single unit ring resonator to get its Q
2. Extract coupling coefficients for different coupling schems(electric coupling, magnetic coupling, mixed coupling) to generate the design chart.
3. Based on the design chart of Q and coupling coefficients and make use of the traditional LPF design topologies to synthesize the physical parameters of the filter (layout design).

However, I really have no idea on how the Q and coupling coefficients of the ring resonators(especially the split-ring resonators) can be obtained using EM solvers. If you have the design experience or example project in any of the popular EM solvers such as ADS, HFSS, Designer... Kindly share with me your design knowledge.

Thanks in advance!

Best regards,

Hi mate
see tbis book " microstrip filters for rf/microwave applications"
it has all what you need

you can downloaded here
https://www.edaboard.com/viewtopic.p...ters+microwave

alos see

https://www.edaboard.com/viewtopic.p...rowave+filters

Hi, rf100:

Thanks a lot for your reply.

Actually I've referred to the first book: Microstrip filters for microwave/RF circuits and I'm tring to re-produce some of the results presented on chapter 8 (page 264-271). However, I just cannot get the expected results using HFSS.

For example: I tried the open-loop ring resonator with the following dimension:
cell dimention a=7mm, trace-width=1mm, gap=0.4mm Spacing=2mm, for the dielectric, DK=10.8 and h=1.27mm (exactly as presented in the book). The results are too..... far from what presented in the book. I believe there is nothing wrong with the solver setup, however, there could be some tricks regarding modeling methodology. Thus, I hope someone here who has the relevant experience can share with me his/her knowledge for the Q/coupling coefficient extraction using EDA tools.

Regards,

Hi,

I'm trying too to simulate those same result but on the CST software, but unsuccessfully. I too not been able to reproduce those two resonant frequencies of the electric and magnetic coupling. I think they (Lancester at al) require a special mode setup of the simulation for the S-Param simulation. By simulating EigenValues of the open ring model, I do get resonant frequencies quite close to the mentioned frequencies, and visualising the resonant do match electric and magnetic coupling. However, doing a transient simulation, I can't get those resonant peaks. I'm also stucked.

Eyal

Hi
did you try with momentum
I think it is easy and reliable
good luck

Hi, Eyalc:

I'm still trying to reproduce the design chart. Please do let each other know if one of us successfully get the results.

I tried Ansoft HFSS and Ansoft Designer, they gives similar results for different kind of model set-ups. Till now, I can only get the resonant frequencies of the unit resonator at different modes using two-port set-up either in HFSS or Designer. However,I still cannot get the similar results for the coupled structure. :(

Hope I can make some progress and update with you before give up.


Hi, rf100:

I believe it doesn't matter on what kind of solvers you are using. The most important thing is how to set-up the model physically in any of the solvers and extract the related information accordingly. For example, the model setup in HFSS designer and in Agilent ADS should be very similar using Edge Port, I believe the results you're going to get is also similar provided that the physical layout is the same. Do you think so?

Regards,