How to draw a curve of vivaldi antenna ?

Hi to all,
I'd like to design a Vivaldi antenna in the 9 - 11 GHz frequency range on FR4 substrate (Er=4.4) with CST. So I can not draw a curve in CST. I drew a curve on AutoCAD but it could not import to CST. How can I do?

Thanks in advance
Paow

You can draw curves in CST .

Select Curves> new curves>
use arc or circle to create an arc. make a line at its start for the width you need.
now select sweep curve option and click on the arc and line , you will get a solid 2D arc , now select face and extrude for thickness.

Regards
GV

I also use a text file of precalculated table of points with my own made software, then I import this list to CST to select the curve points. Bezier is useful.

In CST may be you can make a simple VB script for doing this, please, read the help.

B.R

As byteptr suggested, you can draw a line/curve using equation in Matlab and save that two column matrix as a text file. Then in CST, Curves > New curve and select Spline. Press Esc to have a window where you can import the .txt file. Select the .txt file and press OK. That's the curve you are looking for.

Regards,
Shameem

Thanks a lot. I will try to do.

Just one remark: I don't think it's a good idea to implement an antenna at 9-11 GHz on FR4. Tolerances on epsilonr and thickness are just too high to get good results, and moreover you'll get a lot of losses in the dielectric. Use some high frequency laminate which is designed for this.

check this, might be useful to you,

http://www.cst.com/Content/Applicati...cle.aspx?id=15

Regards
GV

Dear raiohead,
What's the substrate that I should be use? Or how can I do, if I use the FR4. I know that FR4 is not suitable high frequency.

Hi paow,

I guess Rogers RO5880 (ε_r = 2.2) would be a better substrate for your vivaldi. While designing an antenna, most of time try to use lower permittivity material since radiation from these kind of substrates are easier.

Regards,
Shameem

Thanks shameemkabir,

I will design my antennas on RO5880. But the results are simulated from CST not from measured.