HFSS Resonant cavity

View attachment Project1.rarHi
I am going to simulate a cavity in HFSS. I want to find resonant frequencies and S11 parameter. All time I try do that a program shows stupid solutions. Is there anybody that can tell me how to do that?
File in attachments.

Jacob

kubajed said:

View attachment 134469Hi
I am going to simulate a cavity in HFSS. I want to find resonant frequencies and S11 parameter. All time I try do that a program shows stupid solutions. Is there anybody that can tell me how to do that?
File in attachments.

Jacob

Click to expand...

Hi Jacob,

First use the Eigenmode solver to find the resonant frequencies of the cavity. This requires no input port and will tell you where your cavity resonates. Then adjust your dimensions to move the resonant frequency if desired. I suggest using the optimetrics tool to sweep a dimension, and then use results>create modal solution data report>rectangular plot>swept dimension and the desired mode.

Next use the Driven Modal solver. Place an antenna or waveguide and specify the input to be a Wave Port Excitation. Then after you have specified the solution frequency you can right click Setup1> Add frequency sweep> specify range and increment. After it has completed go to results>create modal solution data report>rectangular plot> freq sweep with S11 or VSWR on the other axis.

NOTE: If you make the frequency increment too small (more than 100 points) the solver will take a long time to solve all unknowns at each frequency. If you make the increment too big (~10 points) the plot will not show any useful information. BUT if you solve using interpolated or fast with ~10 points, you can go back AFTER the simulation has finished and specify a smaller increment which it will update to show a useful S11 in a shorter amount of time. There may be a better way of doing this, but this is how I have gotten around the issue.

Hope this helps,

Cheers,

Kurt

Thank you for answer.
What is solution frequency in driven modal?
I ran both solutions and I think something is wrong. Why there are a lot of resonant frequencies? Why S11 is so "ragged"?

The solution frequency is the only frequency that is used for the adaptive analysis. I've never seen different results for a sweep when using a different solution frequency but I usually choose it to be near the middle of the sweep.

It looks like you did not decrease the frequency step size after the solution was found. To do this right click your frequency sweep >properties > and change the step size until you have a few thousand points. You cannot do more than 25,000 and it will warn you if you go above 100, just click OK and update the plot.


Side note: If you do frequency sweeps for parameterized results, for example one sweep for 100 different geometries, decreasing the step size will take a long time. I try to narrow the frequency band as much as I can so that I can get the same resolution for less computation time.

Hi
Thank you. I think I still do something wrong, because there shouldn't be as much eigenfrequencies. How to find only few, most important eigenfrequencies? The most important frequency is 2,45 GHz.

kubajed said:

Hi
Thank you. I think I still do something wrong, because there shouldn't be as much eigenfrequencies. How to find only few, most important eigenfrequencies? The most important frequency is 2,45 GHz.

Click to expand...

Do you see many dips in the reflection curve around 2.45 GHz? If so, simulate the structure using the Eigenmode solution type first. That will easily show what modes you need to suppress assuming they are too close to the desired mode at 2.45 GHz. If you are desiging something to be powered using a microwave oven magnetron, you'll need to suppress anything within +-60 MHz to ensure that you don't have any degeneracies. If you're using a high power commercial magnetron it may have a tighter bandwidth ~+-10 MHz.

I'd recommend you experiment with different perturbations to the geometry to suppress the surrounding modes using the Eigenmode solution type. Then go back to the driven modal once you have your resonance where you want it to make sure your transmission line is matched to the structure.

Hi
I want to decrease sensitiveness of eigen mode. It is possible? There is too much resonant frequencies.

kubajed said:

Hi
I want to decrease sensitiveness of eigen mode. It is possible? There is too much resonant frequencies.

Click to expand...

No. Even if you specified the solver to somehow 'pass over' unwanted resonant frequencies, the natural frequencies will still be there. You need to make structural changes to suppress unwanted modes.

If you downsize the structure such that you are operating closer to the fundamental (first) mode, then they will be spaced further apart.

Which mode is which (TExxx, Tmxxx)? How to obtain that?

Plot the fields and you should be able to see which is which.

If I try plot fields in eigen mode I see only one plot. For what mode is that plot? No more for other modes.

Right click on field plots and go to edit source. Then type 1 in the eigenmode you want to see and 0 in all the others.

Hi
I try to change cylinder length and I use optimetrics. I attach file and images. Could you tell me how to check fields for each length?
View attachment Project1.rar