Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Electromagnetic waves not propagating through perfect electric conductor and problems with ports

Please login with a confirmed email address before reporting spam

Hey,

just a heads up that I'm pretty new to Comsol so the problem I'm experiencing might be quite trivial.

I'm currently trying to model a simple resonator to check how simulated results match up against experimental results and to use this simulation as a basis for other such models.

I'm currently having problems with ports and PEC's. It seems that no matter the changes I make to my setup, waves do not seem to propagate from port 1 to 2 (even though both are connected through a PEC) and the S21 parameter is thus in the -800dB's. This problem seems to appear with both regular ports and lumped ports and when looking at the electric field it seems to not propagate anywhere.

I would also like advice as to the type ports I should be using for models with these types of geometries and what to take note of when using such ports. Currently the main problem I have with using these types of models is that waves do not seem to propagate along conductors to other ports (thus S21 parameters are in the -70dB's for example) and I do not know why.

I would be happy to hear about any solutions that could shed light on this problem.


7 Replies Last Post 2023年7月28日 GMT-4 15:59
Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 2023年7月25日 GMT-4 07:17

Nico,

there are several issues.

  1. Your geometry is kilometers big. You may have chosen the wrong geometry unit.

  2. The port must connect the stripline and ground unless I misunderstand your design.

There is a microstripline example in the application library you can use as a guide for a proper setup.

Cheers Edgar

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Nico, there are several issues. 1. Your geometry is kilometers big. You may have chosen the wrong geometry unit. 2. The port must connect the stripline and ground unless I misunderstand your design. There is a microstripline example in the application library you can use as a guide for a proper setup. Cheers Edgar

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 2023年7月25日 GMT-4 08:13
Updated: 1 year ago 2023年7月28日 GMT-4 17:11
  1. Thank you for pointing out the scaling problem. I unfortunately overlooked it while looking for issues. An issue still remains as now the S21 parameter is in the -30dB which is in the same order of magnitude as with other similar designs and it doesn't seem like propagation is happening along the conductor.

  2. I'm not trying to use microstriplines but cpw lines. I have tried to use the cpw example for modeling my designs but have been unable to find success. Regular ports nor lumped ports seem to work for my specific example as there is no propagation along the conductor.

I'm thinking it might have something to do with how I am assigning ports on interior boundaries but I am having a hard time pinpointing the nature of the problem.

I have an example photo of the electric field where I used rectangular ports. This kind of phenomena of the electric field being strongest at the back of the port seems to be happening with all types of ports, lumped ports included.

Regards, Nico

1. Thank you for pointing out the scaling problem. I unfortunately overlooked it while looking for issues. An issue still remains as now the S21 parameter is in the -30dB which is in the same order of magnitude as with other similar designs and it doesn't seem like propagation is happening along the conductor. 2. I'm not trying to use microstriplines but cpw lines. I have tried to use the cpw example for modeling my designs but have been unable to find success. Regular ports nor lumped ports seem to work for my specific example as there is no propagation along the conductor. I'm thinking it might have something to do with how I am assigning ports on interior boundaries but I am having a hard time pinpointing the nature of the problem. I have an example photo of the electric field where I used rectangular ports. This kind of phenomena of the electric field being strongest at the back of the port seems to be happening with all types of ports, lumped ports included. Regards, Nico

Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 2023年7月25日 GMT-4 10:19

I still think, that the lumped ports need to connect the conductor and ground.

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
I still think, that the lumped ports need to connect the conductor and ground.

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 2023年7月26日 GMT-4 04:19

Thank you, seems to have worked in solving this issue as there is now propagation between ports, although this seems to only happen with uniform lumped ports. The problem still persists with other types of lumped ports and regular ports.

On another note what do you think would be the most accurate port type for modeling these sort of resonators? I mean these types where ports are on an inner surface. And I would like you do add what should be taken into account when modeling using your suggested port type.

Thank you, seems to have worked in solving this issue as there is now propagation between ports, although this seems to only happen with uniform lumped ports. The problem still persists with other types of lumped ports and regular ports. On another note what do you think would be the most accurate port type for modeling these sort of resonators? I mean these types where ports are on an inner surface. And I would like you do add what should be taken into account when modeling using your suggested port type.

Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 2023年7月26日 GMT-4 11:20
Updated: 1 year ago 2023年7月26日 GMT-4 11:23

When working with RF transmission lines of any kind, using a lumped port means making a circuit approximation to the true EM field configuration at the subject port. Often, especially if: (1) you choose a lumped port with close to the correct impedance, and (2) its dimensions are very small compared to the wavelength in the medium in question, this may (i.e., still not always, but often) be a very good and adequate approximation for many/most purposes. However, if you want or need something more accurate, then you need to specify (or solve for) the detailed field distribution at the port(s). A user-defined or numerical port can potentially provide a more accurate representation, but the price is that you will need to put in more work (and more study & thinking) to properly specify or compute (whether analytically or numerically) the fields at those kinds of ports. Comsol Multiphysics provides a variety of RF port specifications and types. I encourage you to familiarize yourself with many of them, along with studying example applications (see examples in the Application Library) that make use of them.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
When working with RF transmission lines of any kind, using a *lumped port* means making a *circuit approximation* to the true EM field configuration at the subject port. Often, especially if: (1) you choose a lumped port with close to the correct impedance, and (2) its dimensions are very small compared to the wavelength in the medium in question, this *may* (i.e., still not always, but often) be a very good and adequate approximation for many/most purposes. However, if you want or need something more accurate, then you need to specify (or solve for) the detailed field distribution at the port(s). A *user-defined* or *numerical* port can potentially provide a more accurate representation, but the price is that you will need to put in more work (and more study & thinking) to properly specify or compute (whether analytically or numerically) the fields at those kinds of ports. Comsol Multiphysics provides a variety of RF port specifications and types. I encourage you to familiarize yourself with many of them, along with studying example applications (see examples in the Application Library) that make use of them.

Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 2023年7月26日 GMT-4 11:20
Updated: 1 year ago 2023年7月26日 GMT-4 11:22

Not sure what your application exactly looks like. Lumped ports are typically used in a situation when some cable connects the structure. General ports may be more appropriate in case the structure is coupled by radiation.

Edit: This time Robert was faster and his response has more detail.

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Not sure what your application exactly looks like. Lumped ports are typically used in a situation when some cable connects the structure. General ports may be more appropriate in case the structure is coupled by radiation. Edit: This time Robert was faster and his response has more detail.

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 2023年7月28日 GMT-4 15:59
Updated: 1 year ago 2023年7月28日 GMT-4 16:00

Thank you, your answers were enlightening and helped me deal with the problems concerning ports and lumped ports, and I've managed to get reasonable results from the simulations.

Regards, Nico

Thank you, your answers were enlightening and helped me deal with the problems concerning ports and lumped ports, and I've managed to get reasonable results from the simulations. Regards, Nico

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.