** dual-polarized p4tch antenna with 2% bandwidth

dear all antenna fans,
I am working on a project involving design of a two port patch antenna with prob feed, good isolation between ports, with a bandwidth of just 2% and a wide pattern (-3db pattern: 150 degrees and 90 degrees in H and E or E and H planes). the simpler the design, the better ofcourse. please welcome me by your suggestions and ideas,
looking forward your immidiate respons,
Mat

dear all,
unfortunatly I do not see any response, please give me any advise if you can,
Mat

Hi Mamali,
I was just looking t some papers and I came across this one. THis might help you.

77 GHz Dual-Polarized Microstrip Antennas on Thin Dielectric Membranes
Gildas P. Gauthier, Nihad Dib, Linda P. Katehi and Gabriel M. Rebeiz

May be you could modify this to meet your specifications. Please let me know if you want the paper.

-svarun

svarun said:

Hi Mamali,
I was just looking t some papers and I came across this one. THis might help you.

77 GHz Dual-Polarized Microstrip Antennas on Thin Dielectric Membranes
Gildas P. Gauthier, Nihad Dib, Linda P. Katehi and Gabriel M. Rebeiz

May be you could modify this to meet your specifications. Please let me know if you want the paper.

-svarun

Click to expand...

dear svarun, in that frequency, i think its so easier to have 2% bandwidth, in 1.2GHz, its really hard; bytheway, I'll be grateful if you upload the paper.
anyway, tnx,
Mat

I have read some articles that used rectangular microstrip p4tches loaded with peripheral slits (on opposite or on all four sides)... The slits reduce the size of the p4tch with bandwidth reduction as a side effect...

Maybe this side-effect could be of use to you

Try a smaller p4tch with peripheral slits in order to get your resonance at 1.2GHz but with reduced bandwidth...

mogwai.

mogwai said:

I have read some articles that used rectangular microstrip p4tches loaded with peripheral slits (on opposite or on all four sides)... The slits reduce the size of the p4tch with bandwidth reduction as a side effect...

Maybe this side-effect could be of use to you

Try a smaller p4tch with peripheral slits in order to get your resonance at 1.2GHz but with reduced bandwidth...

mogwai.

Click to expand...

Dear mogwai,
as you know, the rect p4tches on thin substrates hardly have 1% bandwidth, and I need 2% atleast, so i can not yous size reduction technics such as loading. bytheway, the clear specification is this:
a dual polarization antenna in 1240~1260MHz (S11<-15dB) with port isolation better than -20dB. the sub parameters is (h=.5mm or 1mm or 1.5mm or 2.5 mm and er=4.4).
the size limitation is a rectangle of 12cmx12cm.
teh pattern should be stable and have at least 90degrees in one plane and 120 degrees in the other.

maybe hard, but please help if you can, any ideas is welcome.
Mat

Hello mamali,

The substrate, which is FR4 if I'm not wrong, is not the best one for antennas, since the epsilon is high, and it affects strongly the input bandwidth. I think as well that it is too thin, and it will hardly allow you to reach your required bandwidth.

On the other hand it has some advantages, since being so high the epsilon and so thin the material, the crosspolarization and port to port coupling will be very good.

Usually, to get wide bandwidth the substrates are chosen with lower epsilon and larger thickness than the values you mention, although this leads to narrower beamwidths and higher (poorer) port isolation.

In summary, with the substrate you mention my recommendation is to stack several layers to build a thicker material (e.g. combine 2 of the 2.5mm layers to increase the thickness to 5mm). This will lower the Q factor of the antenna and will give you more bandwidth (I haven't checked if that is enough to get 2%).

As said above, with the FR4 of these thicknesses and the frequency of operation you dont have to worry on the isolation btw ports.

For the beamwidth: I have made a calculation assuming a circular patch (0.15lambda radius for the FR4) and the numbers I get are 160deg for the E plane beamwidth and 84deg for the H plane beamwidth, maybe they are enough to your application (?).

Hi Mamali,
May be you can use Silicon and if you can do some etching under the patch to remove say 60-80% of the material using cheap chemicals like potassium hydroxide thereby creating a composite substrate-of air and silicon with a low synthesized dielectric constant, you could actually get rid of the disadvnatages of high permittivity of silicon and obtain reasonable bandwidth. Well you need to do some full wave simulations to finalize on the design. You should refer this paper. It is very nice.

" Papapolymerou, I.; Franklin Drayton, R.; Katehi, L.P.B., Micromachined patch antenna, IEEE Transactions on Antennas and Propagation , Vol 46, Feb. 1998, Issue: 2
Pages: 275 – 283 "

-Arun

Hi Mamali,

what about using thick, low permittivity substrates? This would met the size constraint of 12 cm x 12 cm, with no problem for bandwidth.
Look at this design with aperture feeding for 1240 MHz.
Regards

Z

dear dowjones,svarun and zorro
tnx for comments and calculations; I know well about the problem with FR4, but its unfortunaltely a must! bytheway, i designed a two port circular patch, wich met the specifications, the problem is that the BW is exactly 1.24~1.26GHz (i satcked layers to get a 3.5 mm thick sub, as dowjones mentioned too); I am still working on it, and i will comment if any new idea comes out. tnx all, and please share more ideas, specially for two port CP p4tches, I mean with two inputs, not two port with one input (which uses dividers to get better axial ratio).
sincerely,
Mat
+its really cool that i have such a great community of friends