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.

[SOLVED] A small HFSS question: can't plot field on a non-model object

Status
Not open for further replies.

kazbek

Newbie level 5
Joined
Mar 21, 2013
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,358
Dear users,
the question is: how to plot a field cross-section in HFSS without disturbing model solution. If I just draw a non-model rectangle, I can't plot field on it - I got an error: "No plot can be generated because only non-model objects and/or faces have been selected". If I make this rectangle a model object, then it changes solution significantly, even though I don't assign any material. So I can't plot any field cross-secion in the place where I need.
Could anyone please give a hint?
The question was asked in the forums, but not answered, at least I couldn't find it.
Best luck, health and happiness to everone -_-
 

Dear users,
I got an error: "No plot can be generated because only non-model objects and/or faces have been selected"

Not really sure if I understand all your question, but whenever I get that error I just need to copy/paste my model in a new file then add whatever object I need, in that case you won't get an error.
 

In order to plot the field at a particular location, I add a model object of of vacuum or whatever the default background material is. As long as it does not intersect with any other objects, it has never had a significant effect on the results of my simulation.

In any case, a non-model object cannot be used.
 

Many thanks to everybody for your replies!
Unfortunately nothing of what you proposed works for me. I created a rectangle which doesn't intersect anything in the model except for the vacuum box in which the whole model is located. I can plot field on this rectangle, but the resulting s-parameters dependencies on frequency look completely different from what they look like when there is no rectangle.
Assigning vacuum to the rectangle as jeeudr proposed is impossible. Then I created a box instead of rectangle and assigned vacuum to it. The result is the same: s-parameters curves look completely different compared to the case without any additional boxes or rectangles.
So, I would formulate the problem once again: I just want to plot field around a certain place in vacuum. How can I do it without disturbing the simulation (i.e. without disturbing s-parameter dependencies on frequency)?
Again, thanks for help, peace and harmony for everybody!
 

Here is the file. It's a CPW resonator, and I just want to plot field above the surface in YZ plane.
 

Attachments

  • Project5.rar
    8.2 KB · Views: 78

the main problem with your structure is that it does not resonate !!!, are you sure about design?
ive done some modifications check it again please
 

Attachments

  • Project5r.rar
    42.6 KB · Views: 76
  • Like
Reactions: kazbek

    kazbek

    Points: 2
    Helpful Answer Positive Rating
Thanks for your help, redcrystalline!
However what do you consider as an improvement in your model over mine? In my model mag(s(1,1)) dependence on frequency had a clear resonant character with resonances at full wavelength and half wavelength and with Q-factor (defined as peak frequency over the peak width at half the intensity) of about 10. mag(s(1,1)) drops from 1 to 0.55 at resonance. As far as I have noticed you just reduced the waveport's height and assigned an integration line to it. As a result in your model there are hardly any resonant peaks seen at all at mag(s(1,1)) vs frequency plot.
And the main problem is still unresolved: how to plot any field in the free space above my resonator?
 

hi :D
in fact i have modified solver option not structure, by the way i'm not sure about your design solver convergence ,i believe that is the main problem, because when you add a non-assigned rectangular surface, design meshes change and so s-parameters change too! is n't?
to prove my idea i ll upload a example design later
regards
red
 
  • Like
Reactions: kazbek

    kazbek

    Points: 2
    Helpful Answer Positive Rating
Perhaps you are right with your suspicion about the convergence - when I make a solution there is a warning: Adaptive Passes did not converge based on specified criteria. And also my results changed a lot when I changed some parameters just a very little bit. Thanks a lot, Redcrystalline, thing start to clear up a little bit! But what is the cause of this convergence problem?
 

hfss use adaptive solver, this means that software in first make meshes according to "solution frequency" and
changes meshes in every iteretion to converge error to assigned "Maximum Delta-S" maximum allowed iteretions are assigned by "Maximum Number of passes"
now what make you sure about convergence is located in HFSS>RESULTS>SOLUTION DATA>convergence
in this root you could find maximum magnitude of delta-S or relative error
 
  • Like
Reactions: kazbek

    kazbek

    Points: 2
    Helpful Answer Positive Rating
You can with plot the field on non model in HFSS15. it is new.
 

Looks like you just have to place the object you want to plot field on (rectangle for example) not very close to the electrodes. It's a partial solution of the problem, but looks like there is nothing better. Thanks everybody for help!
 

It seems to me now I understand what the problem was. The problem was that the solution didn't tend to converge indeed, that's why any change to the mesh (which happens when you include a rectangle or whatever into your model) made a significant change to the solution (s-parameters in my case). After I solved convergence problem I could include rectangles that I wanted to plot field on without signifcant changes.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top