Broadband Microstrip antenna

[fractalguy]

Hello

I need to design a Microstrip rectangular/square patch antenna with following specifications:Operating band:1.75Ghz to 2.5Ghz and center frequency:2.0Ghz
It needs to have a bandwidth of 750 Mhz.The substrate has dielectric = 2.5 and thickness=3mm.No air gap can be used.A capacitive feed is to be employed.
Thanks

 

[zirtapoz]

45.36x46.69mm patch size ll be suitable. How about loss tangent?? and the other parameters(gain axial ratio etc. etc.)

 

[vfone]

The bandwidth of a patch antenna is only a few percent (~5%) of the resonant frequency.
Your requirements are 30% bandwidth, which is almost impossible to get in a standard patch antenna.

 

[mauloftin]

Hi,
You can use book: Antenna.Theory-Balanis ,for example

https://www.amazon.com/Antenna-Theory-Analysis-Design-2nd/dp/0471592684

page 722


ML

 

[noor_awad]

Hi,
You can use book: Antenna.Theory-Balanis ,for example

https://www.amazon.com/Antenna-Theory-Analysis-Design-2nd/dp/0471592684

page 722


ML

i use this site
https://www.emtalk.com/mpacalc.php?...m&Wa=46.69&W_units_list=Wmm&Rin=223.730988773

to calculate the dimension of the patch
but the size is different than what appear on the site
which correct???
please reply

 

[thylacine1975]

There are a heap of different analytical/empirical models (and corresponding simplifying assumptions) for [patch] antenna designs, and it's been my experience that most web calculators don't a) tell you much about the model they use, and b) give slightly different answers from web calculator B.

Given the patch antenna's inherent sensitivity to substrate parameter variation, if you're not 100% sure of your substrate (such as if you're using FR4) I've found a workable approach to be:
* Choose a material,
* Use web calculator X,
* Design your patch,
* Make your patch, plus _2_ others - with lengths +/- a few% of your calculated value.
* Measure the resonant frequency of all three patches, and from a line of fit of your f vs L data, determine the optimum length for your desired frequency.
* Go make *yet another* patch from the same batch of substrate, and you should be pretty close.

Regarding the original bandwidth requirement, I second vfone's comment: 30% is unattainable in a plain rectangular patch. You will only achieve anything approaching this value by using some (unconventional) "loading" techniques, such as adding slots, shorts, resistive loss elements, fractal networks, parasitic elements, high permittivity superstrates etc. ...and of course trading off gain/polarisation/pattern purity etc in the process. Resistive loading (connecting a low value R to the ground plane at some location across the patch) might get you part way there, at the expense of a large reduction in gain (ref: Chen, H.-T., "Compact circular Microstrip Antenna with Embedded Chip Resistor and Capacitor", IEEE AP-S Int. Symp. Digest, 1998, pp 1356-1359).

 

[noor_awad]

thanx realy for ur reply
i will do as u said
but if i have a square patch wxw
just only use the equation of the width only
is that correct??

 

[thylacine1975]

Aye, that's fine.

The resonant frequency is only a sensitive function of the "length", which can be arbitrarily defined as the axis containing the feed point (for linearly polarised patches). The resonant frequency is *much* less sensitive to changes in the other dimension, which predominantly affects the feed impedance. I seem to recall there are some other reasons why L = W (i.e. square) is desirable... E/H pattern symmetry? ..simplified circular polarisation operation? Whatever It'll work!

 

[noor_awad]

thanks too much