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.
Importing/Specifying your own voltage signals
Posted Sep 12, 2012, 11:47 a.m. EDT Low-Frequency Electromagnetics 6 Replies
Please login with a confirmed email address before reporting spam
Hi there,
I am new to COMSOL, very new, so there is every chance I am barking up the wrong tree with this. However, I have been able to construct a 2D model:
5 x 0.5mm radius HV electrodes arranged in a line, equi-spaced. Around these is a rectangular array of ground electrodes so that the HV and grounds are mostly equal distance apart.
In the gap between the HV and GND's there is a dielectric material. I have all this sat in a big box of air.
Now I used the Electrostatics module, and was able to define the edges of the HV's as voltage potentials and set them to 20kV, like wise I defined the edges of the grounds as grounds. I was also able to define the materials for all the regions and I assume thats all you need to do there (just define the material, then leave it floating in space).
COMSOL can calculate the resulting electric field and to me at least, it looks reasonable. It certainly matches what I have managed with equivalent circuits in a SPICE program.
What I want to try now is applying a voltage signal. I have actual oscilloscope data I would like to use of a fast rise time HV signal (0-20kV in 100 ns). Failing that, I have approximate forms generated by single wiebe functions (these at least present nice smooth curves).
How do I actually get such signals into COMSOL? Can I do that with the electrostatics module? Or do I need use a different module - one of the RF modules? - ( I have full access here). Any help is much appreciated.
I am new to COMSOL, very new, so there is every chance I am barking up the wrong tree with this. However, I have been able to construct a 2D model:
5 x 0.5mm radius HV electrodes arranged in a line, equi-spaced. Around these is a rectangular array of ground electrodes so that the HV and grounds are mostly equal distance apart.
In the gap between the HV and GND's there is a dielectric material. I have all this sat in a big box of air.
Now I used the Electrostatics module, and was able to define the edges of the HV's as voltage potentials and set them to 20kV, like wise I defined the edges of the grounds as grounds. I was also able to define the materials for all the regions and I assume thats all you need to do there (just define the material, then leave it floating in space).
COMSOL can calculate the resulting electric field and to me at least, it looks reasonable. It certainly matches what I have managed with equivalent circuits in a SPICE program.
What I want to try now is applying a voltage signal. I have actual oscilloscope data I would like to use of a fast rise time HV signal (0-20kV in 100 ns). Failing that, I have approximate forms generated by single wiebe functions (these at least present nice smooth curves).
How do I actually get such signals into COMSOL? Can I do that with the electrostatics module? Or do I need use a different module - one of the RF modules? - ( I have full access here). Any help is much appreciated.
6 Replies Last Post Sep 12, 2012, 3:22 p.m. EDT
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
I have done something similar (although with acoustics module, not electical)
Basically, start by importing your signal into a Global Definitions > Interpolation node. You import the signal data right into that table and your arguments should be t (global time variable for Comsol)
Then, in your physics interface where you are applying the signal, specify that the signal is int1(t) where int1 is whatever name you give the interpolation function (should appear after name in parentheses).
Then you need to add a time-dependent study. Right-click root node > Add Study. And configure with whatever time range and step is reasonable given your input data.
Will
I have done something similar (although with acoustics module, not electical)
Basically, start by importing your signal into a Global Definitions > Interpolation node. You import the signal data right into that table and your arguments should be t (global time variable for Comsol)
Then, in your physics interface where you are applying the signal, specify that the signal is int1(t) where int1 is whatever name you give the interpolation function (should appear after name in parentheses).
Then you need to add a time-dependent study. Right-click root node > Add Study. And configure with whatever time range and step is reasonable given your input data.
Will
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
I have done something similar (although with acoustics module, not electical)
Basically, start by importing your signal into a Global Definitions > Interpolation node. You import the signal data right into that table and your arguments should be t (global time variable for Comsol)
Then, in your physics interface where you are applying the signal, specify that the signal is int1(t) where int1 is whatever name you give the interpolation function (should appear after name in parentheses).
Then you need to add a time-dependent study. Right-click root node > Add Study. And configure with whatever time range and step is reasonable given your input data.
Will
Ok tried this and COMSOL at least ran. But my result is wierd, it just shows the same surface plot for each time step. Not sure what I have done wrong. My data has a time column and a voltage column. t goes from 0 to 5e-7 in steps of 1.e-9 (so I put this in the range setting in COMSOL).
Gonna play with some settings see if I can get it to work.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Once I have defined my function how do I actually implement it?
I have put this in the Box for electrical potential V0:
int1(0<t<<5e-7)
and in the study i ask it to do the range 0 to 5e-7 with 1e-10 timesteps and i get singularity errors. Bit stuck!
I have put this in the Box for electrical potential V0:
int1(0<t<<5e-7)
and in the study i ask it to do the range 0 to 5e-7 with 1e-10 timesteps and i get singularity errors. Bit stuck!
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Have you tried int1(t) without the extras?
The solver should iterate t for you automatically and solve your interpolation function for the correct potential at each time step.
Sorry, at this point is where my lack of experience with the electrostatic package is showing :)
The solver should iterate t for you automatically and solve your interpolation function for the correct potential at each time step.
Sorry, at this point is where my lack of experience with the electrostatic package is showing :)
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
doesnt seem to like me putting in int2(t)
its orange and the help bubble says unexpected unit of output. Not sure what that means. Do I have to specify the Units in the interpolation pane? Im sure when I do this it mucks my wave form up! If i leave them blank and press plot, I see my proper sharp rising curve, if I put units in the units box, plot prints something entirely different!
EDIT: wait, im putting t in as a unit of time.... doh! should be putting s
its orange and the help bubble says unexpected unit of output. Not sure what that means. Do I have to specify the Units in the interpolation pane? Im sure when I do this it mucks my wave form up! If i leave them blank and press plot, I see my proper sharp rising curve, if I put units in the units box, plot prints something entirely different!
EDIT: wait, im putting t in as a unit of time.... doh! should be putting s
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
yes! this works now! thanks for the help!