How to predict the bandwidth of a patch antenna?

Some literatures(by Chu etc.) have discussed the foundamental limits of electrically small antenna.
Are there any similar conclusion on modern antennas(patch antennas for example)?
Thank you!

Δf/f_res ~ 1.2 (d/D) (in free space)

where d is the height of the patch above the ground plane and W is the width (typically a half-wavelength)

Hi,

Jackson and Alexopolus has developed an empirical formula to predict the BW of a rectangular and square patch antenna

For VSWR which is < 2 ;

BW = 3.77(( εr - 1)/εr2)(W/L)(t/λo)

where;
εr = substrate dialectric constant
L = length = 0.49λo/√εr ( for square or rectangular)
W = width = L (for square patch ) or
W = width = 2L ( for rectangular patch )
t = substrate thickness
λo = wavelength in free space

note that increasing substrate t will increase the BW but will also reduce directivity.



Goodluck...

Most patches seem to have bandwidths on the order of 5 to 15 percent. When slots and cuts are added to the mix as well as increased thickness it seems that 25 to 35 percent is achievable. These seem to be well into what is reported as broadband. I have seen only a few references to patch antennas with around 50% bandwidth.

I am interested in patch like antennas that operate close or above 50%. bandwidth (in the microwave band) .

If anyone has found references in the literature for high bandwidth patch like antennas I would appreciate the citations so I can run down additional information.

hi vfone,
many thanks for ur reply!
it's indeed quite convenient to make a prediction for traditional patch.
Actually, what i'm focusing is if a bandwidth limit exists, for a given specific patch size, no matter what bandwidth enhancement technique we apply?

Added after 26 minutes:

hi livemath,
many thanks for ur reply!
As far as the last conclusion u mentioned is concerned, an increased thickness of the substrate corresponds to a reduced electric energy stored under the patch, i.e in the cavity-like substrate, accordingly a reduced Q and enhanced bandwidth. This is my understanding and i'm not sure if it is resonable.

Added after 18 minutes:

hi Azulykit,
many thanks for ur reply and explanation!
I'm busy with designing a compact patch, about 1/4 wavelength size, very low profile and fairly large bandwidth being required. The problem seems that the miniaturization and large bandwidth is hard to be achieved simutaneously.
The largest bandwidth(VSWR<2)) i ever noticed is about 40%+, when the U-slot on the patch is applied, when it comes to traditional "patch antenna". i wonder if those planar monopoles for UWB systems could help fix ur problem, if such antennas could still be considered as "patch-like".
Thank you!

In general terms, I think of a patch as some shape over some kind of groundplane. In that context wide bandwidth and thin substrate pull in opposite directions. To make things worse patch radiators are usually on the order of half wavelength by half wavelength. Quarter wavelength sizing and thin is probably not in the cards unless -5 to -10 dBi is an acceptable gain level.

There is another rule for antennas: The antenna is always too dang big.

There are a raft of monopoles that have wide bandwidth but one needs to consider the size of the groundplane as well when looking at the real size impact. It is practical to design for bandwidths well beyond 50%. Consider LPD's and spirals. I have designed and tested many forms of both operating from L through Ku band and occasionally pushed the band edges both lower and higher.