+ Post New Thread
Results 1 to 6 of 6
  1. #1
    Full Member level 5
    Points: 3,186, Level: 13
    Achievements:
    7 years registered Created Blog entry

    Join Date
    Jul 2007
    Location
    Trieste
    Posts
    285
    Helped
    42 / 42
    Points
    3,186
    Level
    13
    Blog Entries
    1

    How to calculate external Q using HFSS

    Hi
    My problem is that I need to calculate coupling which is ratio of Q and Q_ext.
    I know exactly how to do it using CST as with Eigen mode solver of CST you can directly calculate Q and Q external.
    Can any body help how to do the same using HFSS.
    I tried to used PML boundary to model perfect matched condition and then run Eigen Mode solver of HFSS which gave me Q but not external Q.
    Please let me know what is the way to find Q ext
    Thanks in advance

    •   Alt19th May 2017, 10:17

      advertising

        
       

  2. #2
    Member level 5
    Points: 1,191, Level: 7

    Join Date
    Jul 2014
    Location
    UCSD
    Posts
    92
    Helped
    25 / 25
    Points
    1,191
    Level
    7

    Re: How to calculate external Q using HFSS

    Q factor of what, exactly? If it's something like a cavity resonator, then you can simulate your cavity with the excitation method and measure Q from the frequency response. That is loaded, or external Q. The eigenmode solution should be unloaded Q, or Q0. This ratio could give you coupling. I believe you can also extract Q from the fields solution using the fields calculator, but I'm not sure you could disentangle loaded and unloaded Q this way.



    •   Alt20th May 2017, 22:56

      advertising

        
       

  3. #3
    Full Member level 5
    Points: 3,186, Level: 13
    Achievements:
    7 years registered Created Blog entry

    Join Date
    Jul 2007
    Location
    Trieste
    Posts
    285
    Helped
    42 / 42
    Points
    3,186
    Level
    13
    Blog Entries
    1

    Re: How to calculate external Q using HFSS

    Thanks that finally I got some answer and you have understood my problem correctly.
    I have a high Q spherical cavity as a part of design of Pulse compressor. I have to couple it with E-Rotator. I have already optimized E-Rotator and Spherical cavity separately now I need to couple the two together so to decouple the big problem I am just trying to couple the output circular port of E-Rotator with high Q spherical cavity.
    I am using driven model to so I am available with S11 and the problem is that any slot in the cavity will shift the resonant frequency from the design value. I tried to run frequency sweep to see what is the new frequency but the issue is that S11 is very bad around -5dB while I was expecting some better value. Why it is so?
    I tried to run the sweep for different aperture size to find some point with critical coupling but the best I got is -5 dB.
    I know coupling and definition but this is first time I am doing it with S parameters and also please suggest how can we calculate Q and loaded Q with Eigen solver.
    I tried to explain my problem and the way I am working if you need some further information let me know.
    Thanks for your patience for reading all



    •   Alt21st May 2017, 10:25

      advertising

        
       

  4. #4
    Member level 5
    Points: 1,191, Level: 7

    Join Date
    Jul 2014
    Location
    UCSD
    Posts
    92
    Helped
    25 / 25
    Points
    1,191
    Level
    7

    Re: How to calculate external Q using HFSS

    In the past I have used Dr. Kajfez's work to help me with Q factor measurements: https://engineering.olemiss.edu/~eed.../rfqmeas2b.pdf

    He also has a book and some handy MATLAB code.

    So it sounds like you have a high Q cavity, but you measure a low loaded Q and shifted resonant frequency because the placement of your E-rotator is not giving you critical coupling? Could you include a picture of your design? It sounds like you've already tried doing what I would suggest, like sweeping aperture size to find the best coupling. Make sure your frequency sweep is very narrow, if your cavity is very high Q (>1000) it can be easy to sweep to coarsely and not find the resonant frequency from S11.



  5. #5
    Full Member level 5
    Points: 3,186, Level: 13
    Achievements:
    7 years registered Created Blog entry

    Join Date
    Jul 2007
    Location
    Trieste
    Posts
    285
    Helped
    42 / 42
    Points
    3,186
    Level
    13
    Blog Entries
    1

    Re: How to calculate external Q using HFSS

    http://obrazki.elektroda.pl/3376681000_1495434512.png
    http://obrazki.elektroda.pl/7465307300_1495434545.png
    http://obrazki.elektroda.pl/8440790000_1495434566.png
    thanks for your continuous interest.
    I have attached all the sweep results and model I am working with.
    In first step I tried to sweep the aperture with a very small seep of 0.1 mm and then used frequency sweep around my working frequency to see what frequency it is shifted and what is s11 you can see in the attached image.
    In the second step I selected the aperture size which gave me the best s11 and swept it for same frequency sweep and mesh settings for cavity radius with a step of 0.2 mm.
    Please have a look at the results and comment about them.
    Issue at hand is that I need to over couple the cavity so I need to find some point of critical coupling and then I need to bring that point to working frequency which is 2.99801 GHz
    I will for sure go through the suggested material but thought to update you about the results so far so that you may have some better picture of what I am talking about.
    thanks for your continued support and interest



    •   Alt22nd May 2017, 07:35

      advertising

        
       

  6. #6
    Member level 5
    Points: 1,191, Level: 7

    Join Date
    Jul 2014
    Location
    UCSD
    Posts
    92
    Helped
    25 / 25
    Points
    1,191
    Level
    7

    Re: How to calculate external Q using HFSS

    Yes that looks like what I would expect, not that I am familiar with the E-rotator cavity scheme you are proposing. You might have tried this, but I would recommend doing a finer frequency sweep around one of those known peaks, for example sweep centered about marker m2 in the third image at 1/100th the bandwidth and 100x the resolution. High Q peaks far apart can be tricky.



--[[ ]]--