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.

How to apply an ac current

Please login with a confirmed email address before reporting spam

Hi
I have the following problem. In my device I have a stripline ( see attached picture) and want to apply an ac current on both sides (the two lines at the bottom of the picture). The currents should have a phase shift of pi, so that I get a current antinode at the top. But I have no idea how to apply an ac current.
Thanks for answering

Kind regards
Thorsten


19 Replies Last Post 12 mar 2015, 11:33 GMT-4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27 mag 2012, 17:07 GMT-4
Hello Thorsten,

Electrical circuit interface provides some nice functionality on this.

In the attached file, I've attached two ac current sources to two sides of a triangle, in a time-dependent solution. Is this what you wanted?

Best Regards,
Magnus
Hello Thorsten, Electrical circuit interface provides some nice functionality on this. In the attached file, I've attached two ac current sources to two sides of a triangle, in a time-dependent solution. Is this what you wanted? Best Regards, Magnus


Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 28 mag 2012, 04:12 GMT-4
Hi Magnus

nice simple model, but I find it slightly confusing with all these different GND and extra terminals

Terminal 2 in EC can be replaced by a GND BC conditions (is equivalent to connect it to the default GND of CIR node "0"). Then one can disable CIR terminal 2 and the GND on "3". Furthermore the "gnd" on CIR node 4 is not required, one can replace the node "4" by node 0 which is GND in the CIR Terminal BC definition .

I see nothing "physically" wrong with your approach, I just find it excessively complex to analyse

--
Good luck
Ivar
Hi Magnus nice simple model, but I find it slightly confusing with all these different GND and extra terminals Terminal 2 in EC can be replaced by a GND BC conditions (is equivalent to connect it to the default GND of CIR node "0"). Then one can disable CIR terminal 2 and the GND on "3". Furthermore the "gnd" on CIR node 4 is not required, one can replace the node "4" by node 0 which is GND in the CIR Terminal BC definition . I see nothing "physically" wrong with your approach, I just find it excessively complex to analyse -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 28 mag 2012, 06:00 GMT-4
Thanks Magnus and Ivar,

That is exacly what I meant. I've tried it, but I get no solution. There must be something wrong in my model and I have to admit that I don't understand the cir interface quite well. I've attached my model and hope you can correct my mistakes :)

kind regards
Thorsten

Thanks Magnus and Ivar, That is exacly what I meant. I've tried it, but I get no solution. There must be something wrong in my model and I have to admit that I don't understand the cir interface quite well. I've attached my model and hope you can correct my mistakes :) kind regards Thorsten


Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 28 mag 2012, 06:49 GMT-4
Hi

first of all you do not really need the CIR physics, you can as well use the raw Terminal in MEF, use the "phasor" notation for complex excitation I0*exp(i*phase_in_radians)

Then I do not believe your time dependent solver in 10 steps over 1 sec is OK to resolve 1GHz signals ?

Why not use a frequency domain (harmonic) solver at 1GHz, you are normally working is steady state sinus excitation case, I assume or are you looking for transients ? anyhow then it's in the nsec range and not seconds you should choose for the time stepping

Read the doc about the harmonic solvers, and check a few model library examples its worth the reading ;)

Then at your frequencies you do not really need to simulate the "bulk" material thickness, current will flow in the surfaces anyhow, and with your nesh of only 1 element in the tickness you resolve nothing inthere. You will also gain the mesh size and solver time.

Finally check the mesh in your antenna region, I would have used a higher density at the antenna junction


--
Good luck
Ivar
Hi first of all you do not really need the CIR physics, you can as well use the raw Terminal in MEF, use the "phasor" notation for complex excitation I0*exp(i*phase_in_radians) Then I do not believe your time dependent solver in 10 steps over 1 sec is OK to resolve 1GHz signals ? Why not use a frequency domain (harmonic) solver at 1GHz, you are normally working is steady state sinus excitation case, I assume or are you looking for transients ? anyhow then it's in the nsec range and not seconds you should choose for the time stepping Read the doc about the harmonic solvers, and check a few model library examples its worth the reading ;) Then at your frequencies you do not really need to simulate the "bulk" material thickness, current will flow in the surfaces anyhow, and with your nesh of only 1 element in the tickness you resolve nothing inthere. You will also gain the mesh size and solver time. Finally check the mesh in your antenna region, I would have used a higher density at the antenna junction -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 28 mag 2012, 08:29 GMT-4
Thanks Ivar
Now it works ;)
Thanks Ivar Now it works ;)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 30 mag 2012, 09:23 GMT-4
Hey, I've read your entries about the problem in the topic. I'm starting my diploma thesis and i've to simulate a transmission network with DC and also AC currents. I tried to implement the AC Current with the information given above, but i have not been succesful at the moment. It would be nice if I could get more detailed information how to simulate the AC current?! (I could not open your example files)
Hey, I've read your entries about the problem in the topic. I'm starting my diploma thesis and i've to simulate a transmission network with DC and also AC currents. I tried to implement the AC Current with the information given above, but i have not been succesful at the moment. It would be nice if I could get more detailed information how to simulate the AC current?! (I could not open your example files)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 30 mag 2012, 09:28 GMT-4
Hi,

What effects are you studying?

Perhaps matlab simulink simpowersystems is more appropriate for system level studies?

/ Magnus
Hi, What effects are you studying? Perhaps matlab simulink simpowersystems is more appropriate for system level studies? / Magnus

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 30 mag 2012, 09:46 GMT-4
I have to check the electric fields on the surface of an silicon insulator which is assembled between the power grid line and the grid pole. The static case with DC Current on the grid line is no problem. Now I need more detailed information how to simulate an AC Current (what do I have to fill in as the electrical potential, where do I have to generate the time depending sinusoidal Function for the current etc....)
I have to check the electric fields on the surface of an silicon insulator which is assembled between the power grid line and the grid pole. The static case with DC Current on the grid line is no problem. Now I need more detailed information how to simulate an AC Current (what do I have to fill in as the electrical potential, where do I have to generate the time depending sinusoidal Function for the current etc....)

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 30 mag 2012, 10:40 GMT-4
Hi

either you use a time dependent solver, set the temperature range therein, select the solver subnode and set the time stepping tab to "intermediate" and NOT FREE, then add your voltage as V0*sin(2*pi*t[1/s]) and off you go

But if you apply a steady state sinus voltage, the frequency domain solver is much more efficient, you define the frequency (or range of frequencies) in the Frequency Domain Solver range settings, you define the voltage amplitudes V0 in the terminal and off you go.

Be aware that not all physics accept frequency domain solvers, as not all physics are today set up for harmonic developments of the equations (check the equations for a frequency domain solver and a time dependent solver, and if possible a stationary solver to better catch the differences

--
Good luck
Ivar
Hi either you use a time dependent solver, set the temperature range therein, select the solver subnode and set the time stepping tab to "intermediate" and NOT FREE, then add your voltage as V0*sin(2*pi*t[1/s]) and off you go But if you apply a steady state sinus voltage, the frequency domain solver is much more efficient, you define the frequency (or range of frequencies) in the Frequency Domain Solver range settings, you define the voltage amplitudes V0 in the terminal and off you go. Be aware that not all physics accept frequency domain solvers, as not all physics are today set up for harmonic developments of the equations (check the equations for a frequency domain solver and a time dependent solver, and if possible a stationary solver to better catch the differences -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 6 giu 2012, 08:12 GMT-4
Dear all
with my device above I get a good result. Now I want to check if comsol calculate the standing wave correctly.
So I've created a simple device (coplanar strip) where to ac currents converge, so that there should arise a standing wave.
I created the ac currents like described above. But I get no meaningful result. I should see the standing wave in the E-field, but the E-field is only at the ends. So I need again your help :).
I attached the file in this link:

daten-transport.de/?id=a9Z3dwPNLrb8

The normal upload does not work right now.


Thanks in advance
Thorsten
Dear all with my device above I get a good result. Now I want to check if comsol calculate the standing wave correctly. So I've created a simple device (coplanar strip) where to ac currents converge, so that there should arise a standing wave. I created the ac currents like described above. But I get no meaningful result. I should see the standing wave in the E-field, but the E-field is only at the ends. So I need again your help :). I attached the file in this link: http://daten-transport.de/?id=a9Z3dwPNLrb8 The normal upload does not work right now. Thanks in advance Thorsten

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 giu 2012, 02:37 GMT-4
Hi

clear your solver and your mesh, and reset the file history, then save (with a new name) and upload that file, its much smaller ;)

--
Good luck
Ivar
Hi clear your solver and your mesh, and reset the file history, then save (with a new name) and upload that file, its much smaller ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 giu 2012, 04:32 GMT-4
Ok,
Here is the new and smaller version :)

regards Thorsten
Ok, Here is the new and smaller version :) regards Thorsten


Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 giu 2012, 05:01 GMT-4
Hi

first of all at 1.5 GHz I'm not sure you can use correctly ACDC for 1 meter long structures, your wavelength in vacuum is only 0.2 m

then if you impose a current of +1 on one side and -1 on the other side I do not understand the flow, are you sure you ddi not mix up your GND and Terminal 1 and 2 ?

Finallyyour mesh is a too coarse for a 20 cm wavelength, you must sample correctly your fields

--
Good luck
Ivar
Hi first of all at 1.5 GHz I'm not sure you can use correctly ACDC for 1 meter long structures, your wavelength in vacuum is only 0.2 m then if you impose a current of +1 on one side and -1 on the other side I do not understand the flow, are you sure you ddi not mix up your GND and Terminal 1 and 2 ? Finallyyour mesh is a too coarse for a 20 cm wavelength, you must sample correctly your fields -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 giu 2012, 06:11 GMT-4
Again thanks Ivar

First of all, I deleted my mesh completly :) . Normally it is very fine. But I have the feeling that the AC current is not correct. If I only apply one current and replace one terminal by a GND, I get no flow.

And I tried with many different frequencies, so that can't be the mistake.
And I don't mixed up the terminals and GND. If I switch a terminal with a GND I would not get a standing wave?!

Regards Thorsten
Again thanks Ivar First of all, I deleted my mesh completly :) . Normally it is very fine. But I have the feeling that the AC current is not correct. If I only apply one current and replace one terminal by a GND, I get no flow. And I tried with many different frequencies, so that can't be the mistake. And I don't mixed up the terminals and GND. If I switch a terminal with a GND I would not get a standing wave?! Regards Thorsten

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 giu 2012, 07:17 GMT-4
Hi

but what I see is one stripe with GND on both ends and one stripe with current flowing out at each end, where does it enter ?
But I might have missed something, quickly done ...

--
Good luck
Ivar
Hi but what I see is one stripe with GND on both ends and one stripe with current flowing out at each end, where does it enter ? But I might have missed something, quickly done ... -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 giu 2012, 07:33 GMT-4
okay, the currents should flow in and form a standing wave. It is the same idea as above in my stripline. I just wanna check if comsol does calculate my strilpine correct. But there is no standing wave in this model so I think there is no standing wave in my stripline as well. I made an even simpler model. There is only one AC current and one GND, but the result is in my eyes nonsense. So I believe my current is wrong.

Do you believe that I applied the currents correct?

Sorry that I ask so much, but I am really desperate right now :)
okay, the currents should flow in and form a standing wave. It is the same idea as above in my stripline. I just wanna check if comsol does calculate my strilpine correct. But there is no standing wave in this model so I think there is no standing wave in my stripline as well. I made an even simpler model. There is only one AC current and one GND, but the result is in my eyes nonsense. So I believe my current is wrong. Do you believe that I applied the currents correct? Sorry that I ask so much, but I am really desperate right now :)


Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 giu 2012, 08:06 GMT-4
Hi

first, for me your mesh is too coarse, you are hardly at the Nyquist criteria for air, minimum "Finer" if not "Extremely fine"

second, MEF and ACDC is for models that are "small" compared to your EM wavelength, but in your case your model is 5x larger than a wavelength in air. This is more for the RF module hypothesis that solves the Maxwell equation in more complete form (wavelengths are "short w.r.t. model size").
Now, it's true your are still somewhat in the middle "grey" zone, where neither or both might be still more or less OK to solve your issue

--
Good luck
Ivar
Hi first, for me your mesh is too coarse, you are hardly at the Nyquist criteria for air, minimum "Finer" if not "Extremely fine" second, MEF and ACDC is for models that are "small" compared to your EM wavelength, but in your case your model is 5x larger than a wavelength in air. This is more for the RF module hypothesis that solves the Maxwell equation in more complete form (wavelengths are "short w.r.t. model size"). Now, it's true your are still somewhat in the middle "grey" zone, where neither or both might be still more or less OK to solve your issue -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 7 giu 2012, 08:13 GMT-4
Okay. thanks
I will try the RF module.

Best regards
Okay. thanks I will try the RF module. Best regards

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 12 mar 2015, 11:33 GMT-4
Hello,

I'm really new to comsol and am trying to simulate a multielectrode array and have current run through it. I have built the MEA on comsol but I am not sure how set up a circuit on comsol to apply a current? Are there any papers you know of that could guide me through this?

Thank you!
Hello, I'm really new to comsol and am trying to simulate a multielectrode array and have current run through it. I have built the MEA on comsol but I am not sure how set up a circuit on comsol to apply a current? Are there any papers you know of that could guide me through this? Thank you!

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.