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.

Coupling reduction between antennes.

Status
Not open for further replies.

cuongtran

Full Member level 3
Full Member level 3
Joined
Feb 6, 2007
Messages
184
Helped
16
Reputation
32
Reaction score
9
Trophy points
1,298
Location
WhiteHouse
Activity points
2,110
Hi everybody,

I am now simuling a structure array antenne to investigate the application of EBG structure in coupling reduction by hfss.
Has anyone has exemples in this domain? thanks for help me here.
I posted here an article of monopole coupling, but I deal with patch antenne.

THANKS.
 

Nevertheless, if you has the idea of structure set-up in hfss for this simu, please post it here, for example we must feed to all antennes or only one antenne...?
Thanks.
 

wt type of antenna u r using?
 

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
Dear Friend Coungtran,

I tried to download the attched PDF....It was not clear at all....Can you put down your specs so that I may try simulating it on my PC. Any PPT kind of thing if you can draw and attach would be useful....

Sai
 

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
Decoupling two antennas can lead to improve the antenna system performance. But sometimes decoupling is not enough and you also need to decorrelate the antennas: high C11 (autocorrelation) and low C12 (cross-correlation). C11 is somehow the radiation efficiency, and C12 is the energy radiated from one antenna but coming from the other antenna.

There are several ways to decouple and decorrelate two antennas:

-By a matching network (like the previous example). Commonly applied to monopoles.

-By a metamaterial spacer. Applicable to monopoles, patches,...

Good luck,

Pere
 

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
Thanks very much pere,
Your files give alot of methods in this domain,
If you can we could discuss abt this problem,
One strange thing easily to find out that, S11 after the decoupling is always more weakly? why?
 

Hello Cuongtran,

I don't know what do you mean as a weak response of S11 after decoupling. Notice that matching networks are narrowband devices, so when having decoupling between two antennas (low S21), a good matching (low S11) can be also achieved, but as a narrowband response. In this way, when using a Metamaterial AMC Spacer, a wider bandwidth of operation can be achieved for both S11 and S21.

Moreover, to obtain a good performance, decorrelation (high C11 and low C21) should be also achieved, to ensure orthogonal radiation patterns.

If you want to discuss further about this topic, just send me an email pj.ferrer(at)tsc.upc.edu.

Bye,

Pere
 

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
thanks u all, are you have the simulation file for them?
 

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
hi all, for the "decoupling is enough", what do you propose the value of decoupling?

I read an article, this value reach til |-24dB| !? ok??
Thanks all.
 

Hi Cuongtran,

For decoupling (S21), a good value is below -10 dB (mostly <-15 dB). This is ok if S11 is also < -10 dB.

You can have a look to the "Decoupling with a Metamaterial AMC Spacer" document, and you will see that if you put a metallic sheet between the antennas, S21 < -27 dB across the whole frequency band; however, this result is not good because the antennas are unmatched.

This is always a tradeoff between S11 and S21. For this reason, Correlation coefficients should also be used, because they link both S-parameters.

Bye,

Pere
 

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
Hi perej,
Very nice to discuss with you, I found your paper here
**broken link removed**
but a pity I can not download it, would you please send me one version by eeee

Sorry, I ve just found it out in my pc, thanks .

Thanks very much and see you soon.
 
Last edited by a moderator:

You have it in a post aboive this one. Just go up, and you will see a free mirror. If not, try to go this link: **broken link removed**

Bye,

Pere
 
Last edited by a moderator:

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
There is one paper of F. Yang and Samii regarding mutual coupling between EBG surfaces and its applications fro antennas. I suppose you must have read that paper..
Praful
 

I know it, but you have to be very careful with the type of antenna. EBG surfaces are commonly used to decouple patch antennas, or two dipoles over the same surface.

But sometimes you need more surface homogenization, and you have to use Spiral Resonators (or CLLs, SRRs,...). You loose on thickness, but a bidirectional AMC designs work properly for two closely spaced monopoles. In this case, the same application with patch antennas is not easy, since you need resonators with the same height as the dielectric substrate (1-2 mm) which supports the patch antennas.

It's always a tradeoff between antenna type and maximum allowed thickness.

If you have any doubt about this topic, just send me an email to: pj.ferrer(at)tsc.upc.edu.

Bye,

Pere
 

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
Hello perej,
I think that we all know the article of Samii et Yan, but I don't know why Yan did not consider the C coefficient in his paper, or may be since his patchs has the distance of them far enough?
Im agree with you about the problem of bidirectional AMC, it is very interesting if we can find out a material which can eliminate this limit of space for the thickness ...or is it possible to "dip" them into dielectric..?
 

They don't use the C coefficients for two reasons:

1) The patches are far away, so S11 < -10dB and S21 < -15 dB.

2) The formulation with links s11 and s21 was not developed yet.

However, if 1) is fulfilled, you can say that the system is decoupled and decorrelated. So, from this point of view, you don't need to show C coefficients, although they can provide you a good performance estimation.

Bye,

Pere
 

    cuongtran

    Points: 2
    Helpful Answer Positive Rating
Thanks perej,
I thinks that I would like to talk about the distance between the minimum point of s11 and the maxi of s21 at the point of resonance, thanks for your discussion.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top