Problem with dual-band Sierpinski antenna

[pancho]

sierpinski antenna

Dear people of this forum!!

I'm a student of the technical university of Delft and started in november
2000 with my thesis which has as title " Analysis and Design of a new type
of dual-band patch antenna
for mobile communication". As solution was chosen the Sierpinski patch
antenna, which design is researched deeply at the university of Barcelona
(UPC, C. Puente). IE3D and Microwave Office
are used as simultion package for simulation this type of antenna.

But now having gain insight how to match triangular patch antennas (which is
basic form of the Sierpinsi patch antenna) the problem of my thesis is how
to produce two resonance frequencies near to each other such that the two
UMTS bands could be received with one patch antenna configuration. How has
the slot influence two the two bands.

After three months now simulating using the software package IE3D and
MWOffice, I'm happy to say to you that two triangular patch antennas are now
finished using IE3D (very user friendly!!). But still one problem persists:
the third antenna of my thesis has to be of the form of a one-iteration
Sierpinski antenna. The two frequencies at which it must work are: 1.95 GHz
and 2.14 GHz each with a bandwidth of 60 MHz and as substrate I have chosen
TMM4 (epsilon=4.5, loss tangent=0.002, height of substrate 3.2 mm).

After having done a lot of simulations I still can't bring two resonance
frequencies close to each other forming the two UMTS bands. It seems that is
a limit in the Sierpinski patch antenna between the heigth h1 of the lowest
gasket and the heigth h2 of the highest gasket.

My tutor said to me that two possible solutions are:

1) Using instead of one gap, two gaps close to each other
2) Using one gap with an inset matching network.

Because solution 2 ressembles more to the Sierpinski patch antenna, I choose
this solution but using optimization with 9 variables didn't appear to solve
my problem. Therefore I included in this mail the geometry which I'm now
working about and maybe you can give some information or take a look what
possibilities there are for obtaining the two bands (1.92-1.98 GHz and
2.11-2.17 GHz with a S11<-10 dB). Is it possible or do I have to choose for
a totally different type of patch antenna??

I think it must be possible but optimization will be very difficult. Are
there other ways.It would be a pity not continuing to investigate the
Sierpinski patch antenna because of the work I had done during almost two
years now.

I hope someone can take a look quickly to the design......

Regards, Rafael

 

[erickki]

sierpinski antenna in mobile application

Cross-posting is not allowed! The other post is deleted!

erickki

 

[pancho]

sierpinski antenna design parameters

What is cross posting?? I had included two files, that's allowed ?!

I think you say that I posted two posts which were the same, yes the reason was that I accidentally was I posted it first in the wrong thread ok?

Sorry!!


That's cross-posting! You should delete the other one.

erickki

 

[JNovak]

fractal antenna thesis

we made measurements with these kind of antennas, and they showed that MWO does not perform well with this kind of geometries

 

[loucy]

triangular antenna in ie3d

MWO is inherently not suitable for antenna simulation because the simulated structure is always enclosed with metallic sidewalls. You might get "good" results for some antennas, but you shouldn't bet on it with new antenna design. IE3D is the right choice for the antenna shown here.

A few months ago, there were some discussions in this forum on fractal antennas, and links to several commercial fractal antenna vendors were posted. Maybe you can get some idea from those sites.

If I were you, I would start looking for alternatives. The two band are just too close to each other. More importantly, the concept of fractal antenna is questionable. ----My Opinion

 

[JNovak]

antenna sierpinski

well, like I said, we did experiments in anechoic chamber far field and near field with originial scheme of power measurement, that is much more valuable than all kins of MW simulations.

The concept is not questionable it is working, indeed not in the cell phones or like. The ones we did work well past X-band, very broadband. Indeed they do not provide much gain. But we didn't expect that to be the case, since we were doing that for the different purposes. No violations of the Maxwell equations though )

 

[loucy]

sierpinski antenna design

Can you give more detail on the performance of your working model at X-band? What was the ratio of the two center frequencies?

Even if it works, it has little to do with the idea of fractal geometry. Look at the figure above--who would think of fractal?

 

[pancho]

umts antenna cross problem

Yes indeed the geometry doesn't like a fractal antenna but wasn't it the fractal form that induced two or more bands??

I think that this antenna (like the figure given above) indeed seems like a slotted triangular patch antenna...

But I think it must be possible also for such a structure to generate two frequency bands which are close to each other??

Each antenne behaves namely at a pair of resonance frequencies which generally are always far away from each other (I refer to the first and second resonance), therefore I think it must possible with using the gap in the structure above for changing the distribution of the current such that the distance between the two resonance frequencies will be closer to each other.....

The three sections of microstrip lines are only for matching at the given two frequencies which were:

-UMTS1 1.95 GHz (1.92-1.98 GHz)
-UMTS2 2.14 GHz (2.11-2.17 GHz)

The ratio is thus something like 1.097.....

My problem was determining the right heigth, position and width of the slot, and to find the right dimensions of the different microstrip sections..

Can you give some advice on how to continue further.........

Thanks for your suggestions already, Rafael

 

[loucy]

sierpinksky antenna

I personally don't believe fractal is responsible for USEFUL dual band performance.

Your optimization code should have told you whether it is possible to get the desired 1.1 ratio. Don't start with the assumption that the code could find the correct configuration because the question of existance has not been answered. Rather, you should look at the output of intermediate steps (those "bad" geometries checked by the optimization code) and study the trend. The resonant frequencies should change smoothly with any of your parameters mentioned above.

If you opt to try antennas other than the fractal one, a two layer antenna, each layer with a patch resonant at one of the two frequencies, is a possible way out. The two patches can even be triangular. Depending on one's definition, that is also fractal .