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.

Multiple current sources for interference

Please login with a confirmed email address before reporting spam

I want to simulate interference effect in COMSOL: two current sources injecting at different frequencies 10 kHz and 10,01 kHz. I have the following questions:

  • Should I create two Terminals and two Grounds under the same Electric currents node like that

or create separate ec node?

I assume I need to run time-dependent study for this task. If I have two ec nodes each with its own Terminal and Ground, then Time Dependent step results only in one of the sources "working". I tried having also two Steps in the study, but not sure if this is a correct way to go.

Lastly, as my freq1 and freq2 differ by 10 Hz only, to see the interference, my study duration has to be quite long - at least 100 or 200 ms, but because modulating freq are in kHz, the simulation takes very long time even with simple geometries and basic mesh. What am I doing wrong?

Thanks for help


3 Replies Last Post Jul 6, 2021, 12:15 p.m. EDT
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 3 years ago Jul 5, 2021, 5:15 p.m. EDT
Updated: 3 years ago Jul 5, 2021, 5:15 p.m. EDT

Question: Is all the physics in your model linear, such that superposition holds? if so, I suggest you instead run two frequency-domain models, one with the first source at 10 kHz (and the second source off) and the second model with the second source at 10.01 kHz (and the first source off), and then sum the fields (or currents) from the two separate runs. This should work far more easily than attempting to maintain sufficient accuracy in a long run time-domain model.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
Question: Is all the physics in your model *linear*, such that superposition holds? *if so,* I suggest you instead run two frequency-domain models, one with the first source at 10 kHz (and the second source off) and the second model with the second source at 10.01 kHz (and the first source off), and then sum the fields (or currents) from the two separate runs. This should work far more easily than attempting to maintain sufficient accuracy in a long run time-domain model.

Please login with a confirmed email address before reporting spam

Posted: 3 years ago Jul 6, 2021, 3:26 a.m. EDT

Question: Is all the physics in your model linear, such that superposition holds? if so, I suggest you instead run two frequency-domain models, one with the first source at 10 kHz (and the second source off) and the second model with the second source at 10.01 kHz (and the first source off), and then sum the fields (or currents) from the two separate runs. This should work far more easily than attempting to maintain sufficient accuracy in a long run time-domain model.

Thanks for the suggestion, Robert. How do you mean to sum them up: when plotting already?

Also, I think in this simple case superposition holds. Do you have an example when that wouldn't be the case? I guess it doesn't matter if I simply increase number of sources.

>Question: Is all the physics in your model *linear*, such that superposition holds? *if so,* I suggest you instead run two frequency-domain models, one with the first source at 10 kHz (and the second source off) and the second model with the second source at 10.01 kHz (and the first source off), and then sum the fields (or currents) from the two separate runs. This should work far more easily than attempting to maintain sufficient accuracy in a long run time-domain model. Thanks for the suggestion, Robert. How do you mean to sum them up: when plotting already? Also, I think in this simple case superposition holds. Do you have an example when that wouldn't be the case? I guess it doesn't matter if I simply increase number of sources.

Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 3 years ago Jul 6, 2021, 12:15 p.m. EDT

In a linear medium, the response is directly (linearly) proportional to the stimulus. For example, a simple ideal resistor is a linear component, because if you double the voltage across it, the current through it doubles. Ideal inductors and ideal capacitors are also linear. In contrast, semiconductor junctions, vacuum tube circuits, magnetizable materials (such as iron) etc., provide examples of non-linear behavior.

In regard to summing the results, there are various approaches. If you can set up and execute two solutions (or solutions corresponding to two different runs of a model), you can usually find a way to display/plot the sums of the outputs from them. You just have to figure out/understand how Comsol names them. But if you look around in the various pull-down menus for plottable quantities (when looking around in the Results), you'll probably figure it out.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
In a linear medium, the *response* is directly (linearly) proportional to the *stimulus*. For example, a simple ideal resistor is a linear component, because if you double the voltage across it, the current through it doubles. Ideal inductors and ideal capacitors are also linear. In contrast, semiconductor junctions, vacuum tube circuits, magnetizable materials (such as iron) etc., provide examples of non-linear behavior. In regard to *summing* the results, there are various approaches. If you can set up and execute two solutions (or solutions corresponding to two different runs of a model), you can usually find a way to display/plot the sums of the outputs from them. You just have to figure out/understand how Comsol names them. But if you look around in the various pull-down menus for plottable quantities (when looking around in the Results), you'll probably figure it out.

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.