Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Strange Points in Dispersion Diagram Simulation. Please Help!

Status
Not open for further replies.

jianke

Member level 3
Member level 3
Joined
Oct 22, 2010
Messages
55
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Visit site
Activity points
1,764
**broken link removed****broken link removed**Hi all,

I got a simulation result of the dispersion digram for an unite cell. But some stange points are found in the dispersion digram. Anyone can help me explain it and remove them. The simulation setup of HFSS (version 13.0) is,

Solution Setup
Minimum Frequency: 1GHz;
Number of Modes: 3;
Maximum Nuber of passes: 15
Maximum Delta Frequency Per Pass: 1%
Maximum Refinement Per Pass: 20%
Minimum Coverged Passes: 2

Setup Sweep Analysis
px: linear step from 1deg to 180deg, step=15deg

Everything is fine, when the sweep starts from "15deg" to 180deg, step=15deg. However, when I change the start "15deg" to "1deg" or "0deg" or "10deg", some strange points appear. Please find them in the picture (in the red circles), discreted and continued versions are provided.

3modes_Discreted.JPG 3modes_Continued.JPG
 
Last edited:

Hi jianke,

Nothing is wrong with your dispersion diagram. The culprit is in fact your minimum frequency.

The problem arises when you are looking at very small angles, and the lowest mode, which extends to DC on the slow slide of the light line, is below your minimum frequency. Therefore, it shifts all the modes upwards when it's solving, until the lowest mode is found and you have the sudden discontinuity.

Having to correct manually for things like this in simulated dispersion diagrams is fairly common. If you're going to be working with dispersion much I'd recommend observing the fields to tell which mode is which - this is an invaluable tool for complicated dispersion solutions.
 
  • Like
Reactions: jianke

    jianke

    Points: 2
    Helpful Answer Positive Rating
Hi PlanarMetamaterials, thanks so much for your reply. According to your suggestion, I decreased the Minimum Frequency from "1GHz" to "0.1GHz" (other parameters keep the same). However, the simulation result is not correct, and it shows "Adaptive Passes did not converge based on specified criteria."

I want to dertermine the bandgap (or called stopband), so the dispersion curve better starts from "0deg" or small angle, just as some published papers show. Do you know how to modify my setup to let it start from very small angle. Thank you!
published work.JPG
 

Hi jianke,

At 0 degrees, you would have to have a minimum frequency of 0 Hz to capture the lowest mode. Obviously, you can't do that.

That data that you previously obtained in your first post appears to be correct; I'm not sure why you think there is a problem.

When you change the minimum frequency, the solution does no reset - so if you start it again, it will start doing adaptive passes wherever it previously left off. If this was fairly close to 15, then that error message is just letting you know that you should increase the maximum number of passes, say for example, 30. I would also reset the solution so that it will start from scratch again.

Good Luck
 
  • Like
Reactions: jianke

    jianke

    Points: 2
    Helpful Answer Positive Rating
Hi PlanarMetamaterials, thanks again for those good suggestions. If you saw the attached picture of my "second" post, you can see that the dispersion curve almost starts from 0 (of course, it cannot be 0). I also want to get the similar curve, so that I can precisely determine the stopband.

I change the minimum frequency from 1GHz to 0.1GHz, or 10MHz, or 6Mz. And change the start sweep to 1deg, 5deg, or 10deg (their end points are all 180deg). In all those simulation, the Maximum Number of Adaptive Passes is increased to 30. However none of them is correct. HFSS error is "hfsscomengine, process hf3d : Unexpected internal software error code 0 with module id 0."

Now it seems that only one combination of (minimum frequency 1GHz) and (sweep:15deg-180deg) is correct. If I want my dispersion curve to start from small value (like the published work shows), what should I do? The simulation file is attached. Please help me! Thank you so much!
 

Attachments

  • Unitcell.rar
    31.6 KB · Views: 120

Hi PlanarMetamaterials, how to "get each mode individually"? Do you mean to simulate 1 mode at a time?
If you set the Number of Modes is 1, it simulates the first mode. However, if it is set 3, how can you simulate 3 modes seperately?

Could you please explain the procedure a little bit detail? I completely have no idea. Thanks again!!!
By the way, do you know the meaning of HFSS error - "hfsscomengine, process hf3d : Unexpected internal software error code 0 with module id 0."?
 
Last edited:

Yes, your guess was correct. However, due to the collusion of the upper two modes, you probably want to get those two as a pair.

- Set the number of modes to 1
- Set the minimum frequency to a low value.
- Simulate from 1 degree to 180 to get the first mode
Then,
- Set the number of modes to 2
- Set the minimum frequency to 20 GHz
- Simulate from 0 degrees to 180 to get the upper two modes

Good Luck
 
  • Like
Reactions: jianke

    jianke

    Points: 2
    Helpful Answer Positive Rating
Last question. After get all those curves for each mode, how do we combine them togehter to show just in one figure.
We can do it in HFSS? Or, we have to export the data, and use the external program to plot them together (such as Matlab)?

Please accept my sincere thanks for your kind help!
 

No problem, glad to help.

You can put them all on the same plot if you create two different setups - one for each of the setups in my last post (i.e. one having one mode and one having two). That being said, I prefer my own plots, so I always just export the CSV file and take it from there :)
 
  • Like
Reactions: jianke

    jianke

    Points: 2
    Helpful Answer Positive Rating
Hi PlanarMetamaterials, sorry to disturb you again for help. According your suggestions, I successfully finished the simulation for an UniteCell by seperating modes (from 0 to 180). But when I extend the unite cell to a more complicate stucture (see the attached file) which has 9 modes, the simulation failed.

I tried more than 20 combinations of solution ans sweep setup, none is correct. Those failed simulations setup included increased/decreased Minimum Frequency, different Number Modes, increased Maximum Pass, different Sweep Start, separated simulation for each or two modes. If possible, could you help me check the simulation file. Thanks so much.
 

Attachments

  • ZJ32.rar
    22.1 KB · Views: 117
Last edited:

Hi PlanarMetamaterials, thanks so much for your reply again. I solved the problem. I think the error was from insufficient memory due to the tempory HFSS files in computer. After deleting those files and running again, everything is fine. But, please allow say thank you again for your hlep!!!
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top