vivaldi antenna design

Hi... I'm designing a dual-polarized vivaldi antenna which should operate between the frequency range of 2 to 12.6GHz. I'm using the stripline-line vivaldi notch antenna's design. However, i still have many doubts even after reading various journals and articles.

When i'm calculating the wavelength of my stub, cavity and taper, which frequency should i use to determine the wavelength? And how will i know what is my resonance frequency?

And can i use the same concept of the patch antenna to design the microstrip of the vivaldi?

Anyone designing a similar type of antenna? Probably we can discuss more about it.. Btw, i'm using Ansoft HFSS v10 in my design and simulation.

Thanks!

I have designed and fabricated Vivaldi's over multi-octave bandwidths. I wonder where you have doubts about them.

When you ask about the stub length I would note that the antenna is operating over a very wide range. Chose a stub length that produces the best match and gain performance. HFSS will allow you to examine this issue in detail. It is a good tool for this application.

A Vivaldi is a slot like antenna while patches are modeled as cavities and thus have very dirrerent philosophies. Patches are not normally thought of as being particularly broadband. Most are well under 20-30 %

Thank Azulykit..

However, i am still having some problem in achieving a good matching stub. Is there a systematic way in doing this? Any formulas or calculations for the stub design? I dont seem to get a good VSWR and S11 reading. Not sure if i'm doing the simulation properly.

If you dont mind, could i take a look at your design?

The specific details of that design are the property of a former company and I would not want to post them. In reality, I do not have them now for the same reason.

I would start with a stub that is a quarter wavelength (somewhere midband, like the geometric mean) and adjust the length to improve the bandwidth and s11. You might also consider a "pie" shaped stub. There are reports in the literature of people who have had good luck there. The wider stub seems to increase the bandwidth. I have also had reasonable performance with a coax across the slot.

You don't say what you think of as a reasonable match. I can't really comment without knowing your target.

Hi Azulykit,

I'v attached are my design as well as my VSWR and S11 results. I've tried adjusting the stub and cavity size but i still unable to get a good S11 and VSWR reading throughout the bandwidth. My target for S11 is less than -10dB and VSWR of less than 2. Could you please give me some suggestion on how to improve on this? And could you help me check if i have done the simulation correctly? Not sure if there is any problem in my waveport or boundary condition.

Btw, i adjusted my stripline width to match a 50Ω SMA connector which i will connect later on.

Thanks!

If it was my problem to evaluate I wold start by modeling the antenna in parts. First I would create the feed line. When that was working I would put a pair back to back and make sure s21 and s11 were both acceptable. Next I would look at the stub and slotline combination and then examine them back to back. If that works I would try a single (planar) vivaldi to see if it was appropriately sized and radiating as expected. Finally I would look at the dual pair.

Locating a problem is easier if one is sure that the design is on firm ground. Make sure the parts are working first and then put them together.

How did these preliminary steps work out in your effort?

Vivaldi model and results presented in attached article

@freekman
i hope the following paper can help you
Parameter Study and Design of Wide-Band Widescan
Dual-Polarized Tapered Slot Antenna Arrays,IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 48, NO. 6, JUNE 2000

I woud like to know the measured 3dB beamwidth of the vivaldi antenna.

Regards....

Hi All, I am actually trying to design a very basic CPW fed Vivaldi antenna which can operate in the UWB band of 3.1-10.6 GHz. I haven't started yet but I plan to do this as a part of my Mtech dissertation. I have tried designing with IE3D but ram constratints dont allow me to design in IE3D. Please suggest simulator and also any prior designs would be of great help as i don't find exact dimensions in any of the papers.

@avinashprakhya and manjushree
Vivaldi are basically directional antennas, if u design them on CPW then the pattern becomes ominideational. What else specific details of your design are.

My target of the design of the vivaldi is to realize a very wide beamwidth in the E-plane pattern.
If any of the designers can tell me the 3dB beamwidth they could measure of the vivaldi.

Regards

In my case of the vivaldi design, I am not quite bothered about the directionality of the antenna. Due to some other problems in my overall project i would have to stick to using a CPW feed.

If anybody has already done a design of vivaldi using CPW(preferrably) or for that matter any other feeding mechanism in 3.1-10.6 Ghz then it would be of great help to me. Please respond !

Added after 1 minutes:

@ manjushree : mam do you have any preliminary design of any vivaldi antenna ? if yes then you may upload !

@avinashprakhya
Here is a design of vivaldi antenna with CPW. Interestingly the radiation pattern is quite different.

@ snkhan : Hey thanks a lot man for the pdf, but I had that already. What i wanted wasnt a twin structure. I have enough pdfs but nothing mentions the exact dimensions including the one u had jus attached. So i wanted a design in IE3d or cst so that i have a working antenna in place which can then be used for my project.

But thanks once again for the effort.

Hello all,
I'm now designing vivaldi antenna with the frequency range of 0.5GHz to 2.5GHz. I got the good S1,1 result in the frequency bands I want. But the main thing I want to know is how to get the constant Beamwidth 90 degree for all the frequency bands. I test the reflector with different kind of angles. But I didn't get the constant beamwidth 90 degree for all frequencies. If you want to see the design I did, then I'm going to upload it.
If anyone can tell me how to get the constant beamwidth in Vivaldi, please give me advices how to do it.
Thank you so much,
S' Tay Zar Oo

Added after 47 minutes:

Here is the VSWR I got in my design.
If anyone can tell me how to reduce beamwidth for 90 degree.
regards,
S' Tay Zar Oo

Added after 1 minutes:

VSWR has nothing to do with beamwidth

u have to design and optimise flare for particular beamnwidth desired..normally beamwidth decreases with frequency..so design optimally..see some paper in IEEE they have lots of info on this

or leture note o notch antennas by simons

Hello Akeda,
Thank you so much for your helps,
I use reflector in Vivaldi antenna. I want the beamwidth 90 degree in that 3 frequencies ,0.9GHz, 1.8GHz and 2.1GHz almost 90 degree beamwidth. I tested with circle reflectors but the difference of beamwidth is almost 50 degrees in 0.9GHz and 2.1GHz. I want the beamwidth difference about 20 degrees in 0.9 and 2.1. Would you please give me some idea for that?
Would you please tell me how to get the lecture not on patch antenna by simons?
I want to learn from that.
If you don't mind, would you please help me how to get that lecture notes?
Thank you so much for your helps,
stayzaroo

can anybody please tell me how to design a omnidirectional vivaldi antenna ?

Vivaldi antenna is "end-fire" antenna,
IMHO, it is impossible to design Vivaldi antenna with omni-directional radiation pattern .