Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

[SOLVED] How to improve efficiency of patch antenna

Status
Not open for further replies.
G

Guest

Guest
how to improve antenna gain

I have met a puzzled problem. I use pin to reduce the size of patch antenna, but efficiency reduce sharply, meanwhile S11 reaches to -30 dB, the max gain is about -5 dBi.
What is the matter?
Help !

thanks in advance.
 

how to improve gain of the antenna

please, explain a bit further. what do you mean with pin?. is it a shorting post close to the coaxial input?.

bear in mind that the gain of an antenna is associated to its area. if the method you use to reduce the size reduces it too much, the gain of the antenna will also be reduced!
 

tm02 mode circular patch

Hello e.m. fans,

First of all,Football, are your data measured or calculated? Probably calculated.
What frequency are you using?
How much is the Gain of the unperturbated patch?
Anyway, considering your problem I can say that inserting a pin, a slot, a lumped element..., you can have or a broadbanding or a size reduction effect....however you have to consider that ohmic losses (working frequency plays an important role) are associated to the pin you insert. That's the first aspect.
It's quite typical for such type of structures.
the second one is that you are modifying the surface current density forcing a point of the patch to be at V=0V, so you are not using all the patch surface for radiation.. that's another reason for Gain decrease.
You have the size reduction because patch surface current "path" is deeply modified respect to a standard patch antenna. It's associated to a lower resonant frequency (lower than frequency of an unperturbated patch).

DowJones: in your equation you have to consider the different working frequencies of the 2 antennas (standard probe-fed patch, shorting pin loaded patch antennas).....The area you are considering is not linked to the size reduction effect.

Hope it can be useful

Regards
Lupin
 

dowjones said:
please, explain a bit further. what do you mean with pin?. is it a shorting post close to the coaxial input?.
bear in mind that the gain of an antenna is associated to its area. if the method you use to reduce the size reduces it too much, the gain of the antenna will also be reduced!
My design is a two-layer circular patch antenna.
Shorting pin is used in my design. the shorting pin connects patch and ground. Size is reduced to 30% of its original size. But gain decrease sharply. I think pin as a inductance that will absorb too much energy, so though s11 is near -40 dB(say that input power is high efficiency), radiated power is therefore very low.

Is that right?
 

Lupin said:
Hello e.m. fans,

First of all,Football, are your data measured or calculated? Probably calculated.
What frequency are you using?
How much is the Gain of the unperturbated patch?
Anyway, considering your problem I can say that inserting a pin, a slot, a lumped element..., you can have or a broadbanding or a size reduction effect....however you have to consider that ohmic losses (working frequency plays an important role) are associated to the pin you insert. That's the first aspect.
It's quite typical for such type of structures.
the second one is that you are modifying the surface current density forcing a point of the patch to be at V=0V, so you are not using all the patch surface for radiation.. that's another reason for Gain decrease.
You have the size reduction because patch surface current "path" is deeply modified respect to a standard patch antenna. It's associated to a lower resonant frequency (lower than frequency of an unperturbated patch).
DowJones: in your equation you have to consider the different working frequencies of the 2 antennas (standard probe-fed patch, shorting pin loaded patch antennas).....The area you are considering is not linked to the size reduction effect.
Hope it can be useful
Regards
Lupin

These data are simulated by ie3d. Operation frequency is in a band centered @2.44GHz. The gain is calculated at horizon. Unperburbed gain is about 0dBi, but using shorting pin gain will reduced to below -6 dBi.
I have read of two methods applied to enhance gain,say integrated antenna and very high permittivity superstrate. but the two is not practical.

And my design is for microstrip line feed not probe feed.
thx, it is very grateful of you.
 

hello again football.

again I insist on the argument I gave on the reduction of the area. Based on your data you say that a given antenna size (without pin) gives 0dBi in a given direction. When you put the pin the area is reduced by 30% and so is the gain of the antenna (0.3 is -5.2dB) which exactly coincides with the number you report, and this does not depends on the match of the antenna, provided it has a reasonable value.
As I said before the gain of the antenna is related to its area, and a reduction in the area is associated to an equivalent reduction of its gain.
To lupin: I only consider the area of the antenna, which other could I?

regards
 

Hi e.m. friends,

Football: you have a circular patch fed by a microstrip line with a shorting pin don't you?
What materials are you using? What about their thickness?
Theese parameters can affect Antenna Gain... I'ld like to know them just because I think the unperturbated Antenna Gain ( o dBi) is too low...6-8 dBi are typical values.
Anyway, you have a Gain decrease because surface
currents distribution are modified as I said you in my previous post...
To answer your question I should know shorting pin thickness...
The frequency is relative low to consider skin effect losses..
Anyway I suggest you to use CST, HFSS, Ensemble codes.. I used very seldom IE 3D so I can't help you in problems about mesh or other software aspects...
Check patch antenna pattern, you could have an unbalanced pattern..
It depends on the point where you put the shorting post..
To increase Antenna Gain, you can change dielecdtric constant and thickness or, if you have no problems about accomodation or about antenna pattern, you can use a patch antenna array...
What's your requirement about Gain?

Dowjones: Antenna Gain depends on equivalent Antenna Gain and on working frequency...the areas of the unperturbated and perturbated patch antennas of Football configuration are just the same...the working frequency decreases and it corresponds Gain decrease as well...For that reason you can speak about "size reduction" but the area you use is just the same....

Regards
Lupin
 

hi again folks.
well, it seems this is a fruitful discussion on effective antenna area.

Lupin, I agree that the physical dimension of the antenna maybe unperturbed, but if the working frequency of the antenna is reduced, the electrical size of the antenna is reduced, and that justifies the gain reduction (although as you point out, care must be taken to check that the pattern shape has not been distorted. I assume this in my reasoning).

Football: if you are interested in the horizon direction and if the horizon is the plane of the antenna, why dont you change the selection of the antenna type?. One patch in the dominant mode radiates in boresight (vertical from the antenna plane), which is not your intended direction (if I understand correctly).
 

Football,
what do you mean by horizon direction? In a spheric reference system, Teta=90°? In that case, it's quite typical to have a very low Gain.. Do you have a groundplane don't you? As Dowjones says, patch antenna is not valid to have gain in teta= 90° but in Broadside (Teta=0°) direction.
Think about it, maybe you should change antenna typology...

Regards
Lupin
 

Lupin said:
Hi e.m. friends,
Football: you have a circular patch fed by a microstrip line with a shorting pin don't you? What materials are you using? What about their thickness?
These parameters can affect Antenna Gain... I'ld like to know them just because I think the unperturbated Antenna Gain ( o dBi) is too low...6-8 dBi are typical values.
Anyway, you have a Gain decrease because surface
currents distribution are modified as I said you in my previous post...
To answer your question I should know shorting pin thickness...
The frequency is relative low to consider skin effect losses..
Anyway I suggest you to use CST, HFSS, Ensemble codes.. I used very seldom IE3D so I can't help you in problems about mesh or other software aspects...
Check patch antenna pattern, you could have an unbalanced pattern..
It depends on the point where you put the shorting post..
To increase Antenna Gain, you can change dielecdtric constant and thickness or, if you have no problems about accomodation or about antenna pattern, you can use a patch antenna array...
What's your requirement about Gain?
Dowjones: Antenna Gain depends on equivalent Antenna Gain and on working frequency...the areas of the unperturbated and perturbated patch antennas of Football configuration are just the same...the working frequency decreases and it corresponds Gain decrease as well...For that reason you can speak about "size reduction" but the area you use is just the same....
Regards
Lupin

yeah, the antenna is fed by microstrip antenna with shorting-pin for size reduction. FR4(dielec=4.3, loss=0.02) is used as substrate, height=1.6mm.
If radius of circular patch is selected equated to 10mm, TM01 resonate @4GHz, while TM02 at 8.5GHz. TM02 will generates omnidirctional radiation at horizon(theta=80-90 degree) with ((1)--(-7)) dBi gain. Smaller the radius of the shorting-pin, lower the resonant frequency. For resonance at2.4GHz, the radius of the circular patch is about 6GHz if one shorting-pin placed near the feedline with radius 0.15mm. Yes you are right that the pattern is affected varied the point place of shorting-pin.
The key problem lies in that with or without shorting pin result in efficiency of below 6% or about 80%(of couse at diffirent resonant frequency). May be there is a trade-off between gain and size.
 

Lupin said:
Football,
what do you mean by horizon direction? In a spheric reference system, Teta=90°? In that case, it's quite typical to have a very low Gain.. Do you have a groundplane don't you? As Dowjones says, patch antenna is not valid to have gain in teta= 90° but in Broadside (Teta=0°) direction.
Think about it, maybe you should change antenna typology...
Regards
Lupin
The ground plane is finite and slightly larger than the patch, and this change resonant frequency slightly. HFSS can simulat the plane of theta=90, but simulation of IE3D is not valid at theta=90.
 

The goal to is produce an omni-directional radiation pattern at horizon. In order to generate omni-radiation at horizon, TM02 mode of circular patch is used. But can there some type antenna generate omni-radiation with keeping size and pattern suitable?
 

football, why dont you check a dipole/monopole?. a dipole gives 2.15dB, and although it is not flat it is small and radiates in the horizontal direction (although linear polarization)
 

I agree with Dowjones, you can use a vertical dipole...

Anyway, if you want to keep microstrip technology I woul act as it follows:

- circular patch witch a circular gorund plane; ground plane size, dielectric thickness, patch size play an important role for the shape of the radiation pattern; chek it in litterature.. you have a lot of parameters to taper your pattern as you like....
- excitation by probe, not by microstrip.. you must keep the simmetria for the pattern you want; edge input would degrade radiation pattern....
- for size reduction I would use rectangular slots allocated in a simmetrical way respect to the center of the circular patch ("radial" simmetria);rectangular slots on the radiating patch can taper the pattern
-I prefer exciting the dominant mode just because I'm sure about the size reduction effects of the slots for this frequency.. anyway the choice of the second order mode with radiating slots migh be a challenging aspect...

So if you like my idea, we can go on with our discussion in my topic "Slotted patch antennas".. it coul be good also for me because I'm writing a code about slotted patch antennas and the above structure could be interesting...

Regards
Lupin
 

Lupin said:
I agree with Dowjones, you can use a vertical dipole...

Anyway, if you want to keep microstrip technology I woul act as it follows:

- circular patch witch a circular gorund plane; ground plane size, dielectric thickness, patch size play an important role for the shape of the radiation pattern; chek it in litterature.. you have a lot of parameters to taper your pattern as you like....
yeah, you are right that slot antenna may be my next point. So there must some question to be going to bother you.

What I need is a planar antenna which could produce omni-directional pattern in horizontal plane, so monopole or dipole may not be preferred.
 

dowjones said:
football, why dont you check a dipole/monopole?. a dipole gives 2.15dB, and although it is not flat it is small and radiates in the horizontal direction (although linear polarization)

thanks for your reply.

Do you know how much dBi(average value) a planar antenna can produce at horizon(theta=90)
 

hi football,
there is not a simple answer to your question. it depends on many factors. assumed it is an antenna working in the dominant mode, the main factor is the material choice. If you choose a low dk, the antenna is directive (think of 9dBi at boresight) and has low gain on the horizon, you can think of -12dBi or so. In this case the antenna is broadband-or better, not so narrowband-.

If you choose higher dk (about 2.5) the beamwidth is larger and it has lower peak directivity (6dbi at boresight). Probably, at the horizon the gain may be around 0dBi (i.e., the pattern falls 6 or 7db at 90deg). The antenna has narrower bandwidth.

Well, this is more or less the theory. now, playing with the ground plane size you may change the pattern and modify these values (this was what you were simulating, isn't?)

If you want a planar antenna looking at the horizon, a good alternative was the suggestion of lupin (slot antennas), or even printed dipole antennas, although this depends on the pattern specification you have

regards
 

Hi Football,
if you want, you can write down your RF and mechanical requirements and with my microstrip antennas knowledge I can try to help you.
Your design sounds challenging...

Dowjones: do you study slotted patch antennas? Another idea for Football's design could be a circular patch with slotted circular ground plane...a "sort" of PBG structure to taper the pattern....

Regards
Lupin
 

Lupin said:
Hi Football,
if you want, you can write down your RF and mechanical requirements and with my microstrip antennas knowledge I can try to help you.
Your design sounds challenging...
Dowjones: do you study slotted patch antennas? Another idea for Football's design could be a circular patch with slotted circular ground plane...a "sort" of PBG structure to taper the pattern....
Regards
Lupin

The size of my antenna is limited to below 8 mm*mm, operating @2.5GHz, with omni-gain above -1dBi at horizontal plane(Bandwidth>3.5%). And it must be microstrip edge-fed. So it is so difficult to realize. The difficult lies in the contradiction between gain and size and so on. Microstrip edge-fed can significantly damage azimuth pattern.
 

This happens. say a corner shorted rectangular microstrip antenna..
There the resonant frequency is determined by approximatly corner to corner distance and which is (\Lambda/4)[yes and not even \Lambda/2) hence this a shorted antenna resonates at a very low frequaency almost 1/4 th of that corresponds to a rectangular microstrip of same size.


Now this is a compact antenna wih almost 1/4 area of a nomal rectangular microstrip antenna.

We know D = 4*pi*A_eff/(Lamda^2)
A typical antenna is having 4 to 8 dB directivity (Say 5 dB for this case). So the new antenna will have 1/16th(~12.5 dB) directivity which is 5-12.5=-7.5 dB.

Another thing is efficiency. A typical good corner short antenna gives 50% efficiency and in some case they may go even upto 10%.


So ultimately for the bast possible case you will endup with a -10.5 dB gain for this case.:((

:!: :idea: :?:
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top