Menu

Discussion Forum

New Discussion
Filtern

Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

frequency domain study with external circuit

Posted 19.12.2012, 03:49 GMT-5 Low-Frequency Electromagnetics Version 4.0a 6 Replies

David Segal
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam


Hi

The comsol model gallery includes a frequency domain model of a capacitor. I added an external circuit coupled to

the capacitor and a time dependent study runs OK (as shown in a separate gallery model).

Now say I have a voltage sine source in the circuit, how do I run it in a frequency domain study?

Is it possible?

I keep getting errors. Must I define it as an AC source and run it in a time dependent study ?

thanks

David
6 Replies Last Post 20.12.2012, 01:36 GMT-5
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 19.12.2012, 07:33 GMT-5
Hi

have you read carefully the help of COMSOL ? at least in v4.3a it states:

"
Note:
For the AC source, the frequency is a global input set by the solver so do not use the Sine source unless the model is time dependent.
"

so it should be set to AC for a frequency domain and SINUS for a time domain study, at least that is how I read it, but so far I have never used CIR as active source, only as passive circuitry, so I haven't checked this ;)

--
Good luck
Ivar
Hi have you read carefully the help of COMSOL ? at least in v4.3a it states: " Note: For the AC source, the frequency is a global input set by the solver so do not use the Sine source unless the model is time dependent. " so it should be set to AC for a frequency domain and SINUS for a time domain study, at least that is how I read it, but so far I have never used CIR as active source, only as passive circuitry, so I haven't checked this ;) -- Good luck Ivar
David Segal
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 19.12.2012, 08:26 GMT-5


Hi

Thanks Ivar, I tried defining an AC voltage source in the circuit

and still getting errors when trying to run in frequency domain.

David
Hi Thanks Ivar, I tried defining an AC voltage source in the circuit and still getting errors when trying to run in frequency domain. David
David Segal
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 19.12.2012, 11:28 GMT-5


Hi

The easiest way to check this problem is to open model 12695 (transient modelling of a capacitor in a circuit)

change the study to frequency domain and try to set the voltage source so that the model will work.
Hi The easiest way to check this problem is to open model 12695 (transient modelling of a capacitor in a circuit) change the study to frequency domain and try to set the voltage source so that the model will work.
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 19.12.2012, 15:13 GMT-5
Hi

indeed if you add a frequency domain study, set the frequency to a range such as {50 100 200 400} Hz and select CIR voltage source to AC it does not converge, you get two flat lines insted of a nice converging display plot.

But then while lloking at the solver sequence I see its in segregated iterative moe, so I added a fully coupled node, and selected the "Direct" solver node and said Enable, then solved the model , and that worked nicely, you get complex I and V terminal values, no I havent checked the results, but they do not look too bad (worth an analyitcal check) so probably the ssue was the default solver settings not adequate for this model

--
Good luck
Ivar
Hi indeed if you add a frequency domain study, set the frequency to a range such as {50 100 200 400} Hz and select CIR voltage source to AC it does not converge, you get two flat lines insted of a nice converging display plot. But then while lloking at the solver sequence I see its in segregated iterative moe, so I added a fully coupled node, and selected the "Direct" solver node and said Enable, then solved the model , and that worked nicely, you get complex I and V terminal values, no I havent checked the results, but they do not look too bad (worth an analyitcal check) so probably the ssue was the default solver settings not adequate for this model -- Good luck Ivar
David Segal
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 19.12.2012, 16:15 GMT-5



Hi

Regarding model 12695 I found that It's runs OK in frequency domain if I change the conductivity of air

from zero to one

David
Hi Regarding model 12695 I found that It's runs OK in frequency domain if I change the conductivity of air from zero to one David
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 20.12.2012, 01:36 GMT-5
Hi

indeed that is another issue, "0" conductivity is too little for current flow, and you get numerical underflows, anyhow "air" has a very small conductivity.

It would be interesting to compare the 1[S/m] conductivity results with the direct solving results

Indeed the results are quite different, as one should expect, from my understanding, but can you explain the differences ?

By the way it's sufficient to set the solver to "Fully coupled" (Direct is not required)

Try changing the conductivity of air 1, 1e-3 1e-6 [S/m] compared to the 1E-14[S/m] of the glass and see the differences of the current and voltage on the terminal, plot them for a frequency scan both real and imaginary (phase part)

--
Good luck
Ivar
Hi indeed that is another issue, "0" conductivity is too little for current flow, and you get numerical underflows, anyhow "air" has a very small conductivity. It would be interesting to compare the 1[S/m] conductivity results with the direct solving results Indeed the results are quite different, as one should expect, from my understanding, but can you explain the differences ? By the way it's sufficient to set the solver to "Fully coupled" (Direct is not required) Try changing the conductivity of air 1, 1e-3 1e-6 [S/m] compared to the 1E-14[S/m] of the glass and see the differences of the current and voltage on the terminal, plot them for a frequency scan both real and imaginary (phase part) -- Good luck Ivar

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.