I am working on extracting the coupling factor of microstrip open-loop resonator with a center frequency of 2.5 GHz.
Somehow I cannot find too peaks in S21 with which I can figure out the coupling coefficient,only in the ELECTRIC coupling case,the peaks appear around 5 Ghz instead.Is it common in some occasions?
I will later attach the HFSS project files (version 15) and a photograph of the layout if possible.Do I miss something important?
However,with the help of the HFSS Eignemode,too resonant frequencies can be found around 2.5 GHz.Is it accurate to extract coupling coefficients this way?
Hi, Wen.
This filter will not work with this configuration, if you want magnetic coupling.
There are two or three problems with magnetic coupling, the first and important problem in your configuration is a small cross section that must be found between transmission line and resonators to make enough magnetic coupling.
To solve this problem you must increase the cross section by making the transmission line as T ship.
If the problem stays, please tell me to solve another problem, but I think other problems are not influenced on the general form of response.
With my best regards.
Dear Wen,
The answer to your last question is yes, it is accurate to extract the coupling coefficient from the two lowest frequencies given by the HFSS eigenmode solver, provided that the structure is symmetrical as in your case. However corresponding peaks at S21 could not be obseverd because there is no coupling between the resonators and the input / output probes, as Mr. Ahmed points out.
Best regards,
Juan S Galaz
After a few days of trying，I want to answer my question according to my current knowledge.
HFSS eigenmode solver is a powerful tool!But the coupling coefficients I get from it don't agree well with the example displayed in the paper. Especially when it comes to small coupling coefficients.
The paper doesn't provide much detail on how to feed a planar square resonator while extracting.
Feeding the resonator with offset of zero(weak coupling right in the middle of the magnetic side)somehow arouse a 5GHz resonant frequency.Obviously you can find too peaks around 5GHz. Although it doesn't seem to matter according to the formula proposed by the paper(5GHz=2.5GHz*2), I am not sure it's right.
I try to feed it on the mixed coupling side.It turns out to be OK.And I use the results to built a four poles filter,simulation result is fine.
I still need more time study on the whole thing.Hope it can help some of the beginners just like me.
Article Couplings of microstrip square loop resonators for cross-cou...
I would say the easiest way to understand this is to start with a simple ring or square filter at 2.5GHz. What you will observe is that there will be multiple resonances at integer multiple of the designed resonance frequency. Now if you introduce a split in the ring/square (such as at 0, 90, 180 and 270 degrees), the frequencies will change but there will be a pattern to it. Some cases you will observe even frequencies vanishing and sometimes odd. Also, you will see multiple sub-resonant frequencies appearing along the actual resonant ones. This all depends upon the symmetry and location of the split. Also, sometimes, you need to modify your feed mechanism as well.
In your case, you are having a peak shifted to 5GHz and your actual peak at 2.5 GHz isn't appearing (if I have precised it correctly ?). I would suggest use a stripline based feed and keep the design same. Observe the frequencies from 1 to 10 GHz. Start changing the position of the split and see where you get your desired results. Play with it and you will observed interesting facts for your filter design.
- Badges
- Science topic