Discussion Forum

Posted: 1 decade ago 18 mai 2011, 11:27 UTC−4

Could anyone help to answer this question? Thanks.

Posted: 1 decade ago 18 mai 2011, 11:57 UTC−4

Hi, I think a frequency domain analysis is always based on harmonic excitation and you can't apply a square wave. The frequency response analysis can show you what way your antenna responds to the spectral components of the square wave. You probably could run a parametric study using a sufficient number of harmonics of your square wave and add up the results with appropriate scaling factors. However, I don't know how to add results from a parametric study. If you want to apply a real square wave you will need to perform a transient analysis and define the port input as a function of time. Cheers Edgar

Posted: 1 decade ago 18 mai 2011, 12:08 UTC−4

Hi Edgar, Thanks a lot for your reply. Your point makes sense. However, if I do not add a square-wave AC source, do you know how to add a sinusoidal AC source? My issue is I do not know how to use a kind of AC source in lumped port. Best, Tianying

Posted: 1 decade ago 19 mai 2011, 04:04 UTC−4

Hi Tianying, it is easy. Switch on wave excitation in the lumped port node, specify the terminal type and respective input and choose your frequencies in the Frequency domain study node. Cheers Edgar

Posted: 1 decade ago 19 mai 2011, 19:39 UTC−4

Hi Edgar, Thanks a lot. I will try it. Best, Tianying

Posted: 1 decade ago 20 mai 2011, 03:20 UTC−4

Hi Tianying, I recently had some problems with the lumped port. Actually the coaxial and uniform type didn't work and the solver terminated with an error message that is not very instructive, as most error messages from COMSOL :-(. Finally I could set up a user defined lumped port that requires defining width, height and direction of the excitation gap. A support case with COMSOL support is running. If you make it to define a uniform or coaxial port I would be interested to see how you did it. Cheers Edgar

Posted: 1 decade ago 23 mai 2011, 16:31 UTC−4

Hi Edgar, An error massage often occurs when I use the lumped port. More, I have not used the user defined lumped port. I am sorry for not helping you. I am trying my best to solve this issue. I will tell you once I make it. Best regards, Tianying

Posted: 1 decade ago 29 août 2011, 06:55 UTC−4

[QUOTE] Hi Tianying, I recently had some problems with the lumped port. Actually the coaxial and uniform type didn't work and the solver terminated with an error message that is not very instructive, as most error messages from COMSOL :-(. Finally I could set up a user defined lumped port that requires defining width, height and direction of the excitation gap. A support case with COMSOL support is running. If you make it to define a uniform or coaxial port I would be interested to see how you did it. Cheers Edgar [/QUOTE] Hi Edgar, I'm trying to excite the spiral antenna using a 1mm rod in the center . It worked in CST. I'm trying to simulate it using a single terminal . Is it possible to excite it using lumped port ? I tried with the user defined lumped port , it did not give me an error but the results are wrong. Lumped port in all other forms gave me an error. I also used the port boundary condition I got an error again Theoretically you need 2 terminals for port . But since the excitation rod is off very small magnitude it sees the ground of the coaxial cable it self. It would be great if you could help me in this. Thanks a lot for being active on this discussion forum, your comments have proved very instrumental. Regards Glenston

Posted: 1 decade ago 29 août 2011, 07:30 UTC−4

Glenston, I have some difficulties to get the picture. If you could provide your model, it might be easier to have a look. Cheers Edgar

Posted: 1 decade ago 29 août 2011, 08:14 UTC−4

[QUOTE] Glenston, I have some difficulties to get the picture. If you could provide your model, it might be easier to have a look. Cheers Edgar [/QUOTE] I'm communicating with you through 2 threads. Sorry for that. Anyways i'm attaching the model again . Thanks a lot for your help. Regards Glenston

Attachments:

• spiral antenna.mph

Posted: 1 decade ago 29 août 2011, 08:40 UTC−4

This attachment is with perfect electric boundary conditions Regards Glenston

Attachments:

• spiral antenna.mph

Posted: 1 decade ago 15 sept. 2011, 05:36 UTC−4

Hello sir I also need to apply AC source to my model. i given as Function name : AC Type : sine Angular frequency : 2*pi*1e8 Phase : 0 Amplitude:2 and then through electric potential i called function name as V0 : AC but it is saying unknown variable AC . please help i need to apply AC signal ,how can i apply?

Posted: 1 decade ago 9 févr. 2012, 02:00 UTC−5

I have the same problem.... I use AC/DC module, quasi-static electric current. I set the boundary Vo=5, or Vo=5*sin(2*pi*nu_emqvw*t) But when I solve it, both show "No subdomain scalar data on screen." for the electric potential value Anyone know how to do a AC simulations? ============================================ I searched some previous threads and found something useful... http://www.comsol.com/community/forums/general/thread/23481/ This might help. I will try it now

Posted: 1 decade ago 9 févr. 2012, 03:58 UTC−5

Hi it depends what you want: transient or steady state AC fixed frequency or frequency domain scan ? You must change the solver accordingly and either use amplitude phase, or with the sin(2*pi*f*t) explicitely. Note "t" is only defined once you launch a time / transient solver ANd do notforget that EC is "quasi standard hecen loe frequency) again check the doc ACDC versus RF -- Good luck Ivar

Posted: 1 decade ago 16 juil. 2012, 06:40 UTC−4

[QUOTE] Hi it depends what you want: transient or steady state AC fixed frequency or frequency domain scan ? You must change the solver accordingly and either use amplitude phase, or with the sin(2*pi*f*t) explicitely. Note "t" is only defined once you launch a time / transient solver ANd do notforget that EC is "quasi standard hecen loe frequency) again check the doc ACDC versus RF -- Good luck Ivar [/QUOTE] Hi Ivar Thanks for the reply. I wanted to use AC voltage as an input into piezoelectric devices physics. According to your post above I should use TIme Dependent study and then put sin(2*pi*f*t) as input to piezo - Electrical Potential...right? but it does not take the input. sin(2*pi*f*t) is red indicating it cannot take the input. am i making a mistake? should pi be defined? Thanks in advance Megha

Posted: 1 decade ago 16 juil. 2012, 14:25 UTC−4

Hi normally it should work, but could it be that "f" is already used, or that it does not have the units of "Hz" ? If you have a steady state sinus excitation I would recommend that you consider the frequency domain sweep using a harmonic development of the PDE equations, check the doc -- Good luck Ivar

Posted: 1 decade ago 18 janv. 2013, 03:41 UTC−5

Hello! Do you know how to add AC source now? Will you please describe how to do it in COMSOL 4.2? Thanks!

Posted: 1 decade ago 18 janv. 2013, 04:51 UTC−5

Hi check the doc for Frequency solver or harmonic development (you define amplitudes and the frequency, and your physics is described in omega (frequency domain) (unfortunately I have only latest 4.3a up running ;) -- Good luck Ivar