3x3 patch antenna trasmission line designing problems

[Enio]

Hello,

I'm currently engaged on a project where I need to design a 3x3 rectangular patch antenna array operating 1.5~5.8GHz on FR4 h=1.6mm substrate.

I've studied the single element design and some TL matching techniques like the quarter wavelength but I can't put together a 3x3 elements antenna or the most optimal ways to feed it. Is there anyone or any material available to guide me through it?

I have ADS available for simulations.

 

[EM_Guy]

Check out Balanis' book: Antenna Engineering: Analysis and Design. Chapter 6.

Start with playing around with linear arrays. Once you are comfortable with that, play around with planar arrays.

What are the requirements of your design? This is very interesting to me. Is it for school or for industry? Or is it just a hobbyist thing? I'd love to see the actual problem, the requirements of the design, and whatever solution / white paper that comes out of it. I'm assuming you are trying to optimize the gain. Note that you can focus the beam in a given direction by altering the phases of the individual elements! I'm a little envious!

 

[Jeff Rosenbury]

Hello,

I'm currently engaged on a project where I need to design a 3x3 rectangular patch antenna array operating 1.5~5.8GHz on FR4 h=1.6mm substrate.

I've studied the single element design and some TL matching techniques like the quarter wavelength but I can't put together a 3x3 elements antenna or the most optimal ways to feed it. Is there anyone or any material available to guide me through it?

I have ADS available for simulations.

That's a tough job.

I hope I'm not understanding it right, because I don't think it can be done. That's an array with a nearly 2 octave signal. When the elements add at one frequency (for gain) they will subtract at another.

Hopefully you can skimp on at least some of the requirements. Do you need the whole band for some sort of UWB signal or can you do something like a log periodic design? Are you set on a 3x3, or can you use different size/shape elements for different frequency ranges? What sort of radiation pattern are you looking for? How big can your VSWR be across the band?

I've had some limited success with (sometimes hollowish) triangular (actually trapezoidal) patches for UWB. But those were less than one octave. I fed them with coax through the groundplane into the middleish area of the patch. You can also consider making each element a slightly different size (for broadbanding).

Do you have access to a vector network analyzer? You might want to fabricate and test any design you make.

 

[Enio]

The university does have a VNA but that's only after I'm done simulating and drawing the design. I think my professor wants a 3x3 array with horizontal and vertical subarrays of 1x3. I'm not sure what he wants with that frequency range but anything near that is already fine.

I've read a lot of papers from IEEE but none of those use simple rectangular elements or a regular ground plane or microstrip fed antenna. None of those papers achieved such a BW either. Well I guess I'm screwed

 

[EM_Guy]

Not sure if this will help or not. But it may be worth checking out.

Obviously, the big challenge here is optimizing the bandwidth. I'm familiar with Tschebyscheff impedance matching. But I'm only familiar with that when performing impedance matching. But it looks like they are using Chebyshev tapering in this paper. Check out Collin's book: "Foundations for Microwave Engineering" - chapter 5 - the section on Chebyshev transformer. Also note that with a Chebyshev transformer, you not only optimize the bandwidth around a given center frequency, but you can also design a multiband array, in which you have more or less good performance over a very wide band, but periodically (in frequency) you get bad performance. (See Figure 11 in the attached).

I would love to see what you come up with.

 

[EM_Guy]

See the attached: 16-element linear array with good impedance matching over a 4.4:1 FBW for beam scan angles within +/- 45 degrees from broadside.

VSWR < 2 from around 2.6 GHz to 8.5 GHz.

Hope this helps.

 

[Enio]

Oh thanks, I'll give those a read.

By the way, is it possible to design a multiple frequency antenna using a simple miscrostrip feed line and different sized elements?

 

[Jeff Rosenbury]

Well I guess I'm screwed

Perhaps not. Go back to your professor and find out what she really wants. It is likely she just wanted an antenna that works for some application.
There is a good chance she doesn't need everything I think she does. With a little more freedom, this is likely doable.

 

[Jeff Rosenbury]

Oh thanks, I'll give those a read.

By the way, is it possible to design a multiple frequency antenna using a simple miscrostrip feed line and different sized elements?

Are you aware of the log periodic antenna? That can be http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4642887&url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel5%2F2213%2F4642878%2F04642887.pdf%3Farnumber%3D4642887, at least for some value of simple.

A log periodic antenna basically uses many different antennas for broadbanding. Unfortunately this creates phase reversals between each element. This makes them unacceptable for a single UWB signal. But they work fine for multiple narrow band signals. They are very popular for such applications. You might even be able to buy one off the internet if that's what you need. (Make sure to test it though. There's a lot of crap out there.)

 

[Enio]

Are you aware of the log periodic antenna? That can be http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4642887&url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel5%2F2213%2F4642878%2F04642887.pdf%3Farnumber%3D4642887, at least for some value of simple.

A log periodic antenna basically uses many different antennas for broadbanding. Unfortunately this creates phase reversals between each element. This makes them unacceptable for a single UWB signal. But they work fine for multiple narrow band signals. They are very popular for such applications. You might even be able to buy one off the internet if that's what you need. (Make sure to test it though. There's a lot of crap out there.)

I think that's exactly what she wants. She wants the single antenna operating in versatile narrow bands for GPS jamming, ISM and some cm wavelengths

I need to design one using rectangular microstrip antennas though. That's the thing with this project.

 

[EM_Guy]

Oh thanks, I'll give those a read.

By the way, is it possible to design a multiple frequency antenna using a simple miscrostrip feed line and different sized elements?

I think so. Again, check out Figure 11 of Taper Optimization Microstrip Patch Array design. It only shows the S11 from 5 to 7 GHz, but it looks like there is some kind of periodicity going on in the frequency domain.

I know in my experience I have seen lots of multi-band antennas. If the requirement is to be multiband - rather than broadband - then it should be easier to achieve.

Also, you need to know what your requirements are. For instance, do you need to keep the VSWR below 3, below 2, below 1.5, or something else? If at a given frequency you have a great VSWR (say 1.2 or something) when your requirement is VSWR < 2, but you have a narrow band array, then you should be able to make design adjustments that would make the antenna more broadband at the cost of slightly compromising the performance of the array at that resonant frequency. On the Smith chart, the goal is to get a tight circle across the bandwidth around the center of the Smith chart. If you have ADS, you should be able to optimize your design using ADS optimizing tools.

 

[Jeff Rosenbury]

I think that's exactly what she wants. She wants the single antenna operating in versatile narrow bands for GPS jamming, ISM and some cm wavelengths

If you are intending to use a lot of power, you will want a very good VSWR. I doubt a "one size fits all" (like a log periodic) antenna is the best choice for such applications. Also, I would question how much power you can put through a patch antenna. That leaves aside the legality of jamming other people's signals.

May I ask where you go to school?

 

[Enio]

Since it is a microstrip array it is for low power applications. The desired performance for the different frequencies are a S11 of -10dB (that translates to some VSWR maximum but I'm on bed right now :(. So let's say the S11 graph would have multiple spots where S11 peaks down at -10dB (okay now english is getting very hard for me when it comes to the technical terms hahaha).

Answering Jeff: I'm currently at a brazilian university but I'll be moving to UVM (Vermont) this Fall.

I don't know if there's such a thing, but designing an array of different sized elements (different operation frequencies, therefore different input impedances etc) would solve the project. Then the only issue left would be matching the transmission line impedances (not sure if another design problem would come up, I'm still very fresh into antennas and all I've learned so far was studying on my own).

 

[EM_Guy]

I don't know if there's such a thing, but designing an array of different sized elements (different operation frequencies, therefore different input impedances etc) would solve the project. Then the only issue left would be matching the transmission line impedances (not sure if another design problem would come up, I'm still very fresh into antennas and all I've learned so far was studying on my own).

Usually, this approach would not be an array design, but a diversity design. When you are designing an array, the most significant factor is usually the array factor, which is a function of the geometry of the array, the number of elements of the array, the spacing of the elements, and the progressive phase difference between the elements.