Patch antenna at 433 MHz

[FLJdesign]

I have to build a patch antenna at 433 MHz with 8 dB of directivity.
Using PCAAD, I end up with a patch 28 x 30 cm at 5 cm from the ground plane using air. The probe is at 7 cm from the edge and the program calculated 8.5 dB and the BW is 31%. I like the 31% BW because it mean that it will be less sentitive to tolerances. I will build it tomorrow.

Is there another way to make this patch? I read a tread where Higgler said he had made a 433 MHz patch 2x4 in long! I try to simulate this and the resonnance is at 1.4GHz???
Please,I will like more info on the 433 MHz design.
Thank you
F.Methot

 

[FLJdesign]

I buid it. I loaded the S11 file generated by PCAAD into a Eagleware linear simulator and got a perfect match with a 4 pF in series. I install a 4 pF between the probe and the patch. I mesure the S11 and got 25 dB return loss at 433 MHz!

Is there a way to build it smaller a get 8 dB of directivity?

http://www.FrancoisMethotCV.com

 

[Higgler]

The patch I have very small (2"x4"x0.060") and is a quarter wave patch on 60 mil FR4 and it's made for 390 MHz with tuning tabs that can be cut off to change the frequency up to 450 MHz. The bandwidth is small, about 20 Mhz or 4%. It's gain is closer to 2 dBi since the antenna is small. Your thick design sounds friendly, and the only way to get near 8.5 dBi gain is with a larger patch and air dielectric.

Your 31% bandwidth is very large, if that isn't a typo, is the reason for it so that you can use the 390 Mhz European band for similar devices? or did someone want 31 MHz bandwidth and type something in error?

I think you'll find that getting more than 20% bandwidth on that design might be a challenge due to the input probe. I'm speaking from experience. Or did you build it yet?

What kind of probe are you using? If it's just a coax. center conductor it'll have to be a large connector wih a thick center conductor. When you make a thick patch the input feed inductance limits your bandwidth and you need to compensate for it.

kch

 

[FLJdesign]

This is my first patch design and build.
The 31% was calculated from PCAAD 5.0.

Yes I build it. I got 10 dB return loss +/- 60 MHz centered at 433 MHz. The 4 pF in series provide a good 25 dB return loss +/- 10 MHz.

This is to install inside the floor of a museum. They install a TXRX unit under the tile (the floor is 2 feet suspended) and when a visitor approach the TXRX with a mobile and earphone, he or she can listen about the painting just in front. The problem to solve is because there is painting that are close to each other and the system get confuse...
So, I suggested designing a very high directive patch antenna. We will test next week.
But I was wondering what other type of antenna I could use to get directivity!
I was thinking of having a smaller patch and make an array of 4 or 6 and maybe more. I just have to split the signal accordingly.

Will I get better directivity with and array of 4 compare to one large patch?
I can send you some pictures if you want. The probe is a small coaxial cable, at one end I have a BNC femele with flange support. I keept the shield up to the top. Install a leaded 4 pF in series at that point.

Thank you for your time.

http://www.FrancoisMethotCV.com

 

[Higgler]

Sounds like you got very good results on that patch, congrats on that. Patches by themselves aren't very directive, but an array would do alot better.

It might be good to see just how many elements you truly need by building 4, 8, or 16 single elements, purchasing a power combiner with some attenuators and trying them out. People with deflect the energy and possibly corrupt your results.


The tiles you place atop your patch will also change the patch frequency, plus when people stand on them that may change the patch frequency too. Hopefully you check that out when you test antennas.


It seems that mounting the patch antennas on a vertical wall might do better than on the floor. You might want to ask about that. Having the patch on the bottom of a vertical wall might be a solution too.

Sounds like a challenging effort with political challenges too.

kch

 

[FLJdesign]

Thank you for your input.

The program estimated the directivity to 8.5 and this is why I build this one. I will test to see if it is true.

I will definitively look at array...so you say that I can build a 2 x 4 on FR4 60mils...where do I place the probe?

http://www.FrancoisMethotCV.com

 

[FLJdesign]

Regarding the 2 x 4 patch 62 mils with dielectric of 4;
I enter these values in PCAAD. It calculate 0 +J 1.5 of impedance at 433 MHz? I tried different location from the edge.
How did you calculate the matching?

http://www.FrancoisMethotCV.com

 

[Higgler]

The thinner the patch, the closer to the center of the patch (standard half wave patch) you need to feed it. That makes it more sensitive to placement when it's thin.

Did you make a quarter wave shorted patch with that 2x4 inch dimensions or a half wave standard patch, mine was a quarter wave shorted patch?

The standard patch would be approx. 7 inches long in FR4 (Er= 3.8 to 4.0 ish) and the feedpoint is one inch off the center, or 2.5 inches from the edge.

kch

 

[FLJdesign]

Here I need more info (limited knowledge with patches):

What is a quarter wave shorted patch or half wave standard patch in relation with the program that calculate the impedance?

If I enter 2 x 4 x 62 mils rectangular patch in the software and if I select Probe fed, the software calculate the impedance at the location of the probe, so why quarter or half wave?

 

[Higgler]

I don't know PCAAD, sorry. There's likely no option for quarter wave patch, only a full halfwave patch.

If you make a half wave patch, and cut it exactly in half and short the wall, you have the same frequency, though less bandwidth and less directivity.

i.e. if you make a 2"x7" FR4 433 Mhz patch, then cut the 7" length to 3.5" and connect the top to bottom of the patch at the cut length, you've converted a 1/2 wave patch to a quarter wave patch, still at 433 Mhz. Slight asymmetry in the patterns, but if you need a smaller antenna, it's a good way to go. 1/4 wave patch is about the smallest convenient flushmount antenna to make.

kch

 

[FLJdesign]

OK, it's like a short stub transmission line.
So the 2 inches is the width and 7 in is the lenght of the half wave or the lenght of 3.5 but shorted at the cut will be quarter wave...and this is feasible only when the thichness is small. OK, I learn something new today. I will Google some of that.

Thank you

http://www.FrancoisMethotCV.com

 

[Higgler]

Sounds like you understand it now.
I think it works for any reasonable thickness. The location of the feed changes from the edge for thick patches to the center for very thin patches.
This quarter wave cutting technique is just to make the smallest size antenna. It seems that on a full wave patch, the voltage inside the middle of the patch is zero which lends itself to being short circuited. It works pretty well.

The shorted end is often just done by a set of short circuited via's (1-6 via's). The quantity and size of those shorts do change the frequency of the patch a little.

kch

 

[FLJdesign]

I simulated the 2 x 7 and I found the probe location with little iteration for 50 ohms. Good!

Now, I did lots of search to find a program that calculate the dimensions, and no success. It is easy to find a simulator but not easy to find a program that calculates different types of patches. Do you know a program that can do that? It seems that we need to go find a book for the equations for every type of patches!

http://www.FrancoisMethotCV.com

 

[Higgler]

Three programs are good.
IE3D (Zeland),
HFSS (Ansoft) and
CST Microwave Studio.

these programs all require you to provide the dimensions. They have optimizers to improve the performance of VSWR or Antenna Patterns and can require alot of time to run.

Each is $50K. They are accurate, costly and take some time to learn.

kch

 

[FLJdesign]

...but do you know a program that calculates the dimensions?

I am a consultant. I can't afford big software for specific application. PCAAD is affordable and I can do a lot with Eagleware (now Agilent) linear and the EM 2D + 3D simulator which work very well for patches.

One day I might put all the equations that I can find togethere and write a Visual Basic program!

http://www.FrancoisMethotCV.com

 

[Higgler]

No, there aren't any dimensional programs you'll find cheap and accurate. You will find approximate formulas around though.

Absolute accuracy all depends on too many factors to just have a simple program for dimensional calculations. Most narrow band patches have less bandwidth than the tolerances of the dielectric you etch them on. Hence you have to do 5 different designs sometimes, measure the patches and adjust your artwork to meet the material dielectric to be on frequency.

kch

kch

 

[DaveAA1A]

I don't know where you are presently at with this project, but have you thought of a helical ( circular polarized ) antenna since you have 2' of space under the flooring??

Regards, Daveaa1a

 

[FLJdesign]

Hi Dave,
I build a 28 x 30 cm and 5cm off the ground plane. I measure 60 MHz BW for 20 dB return loss. We tested with a transmitter-receiver and got very good directivity. We are building 5 others and we will tested at the museum.

I said that the I had 2 feet deep but in fact I have only one foot.

I have good experience with helical antenna, I tested lots of 800 MHz for the mining and shaft. But here, I have less than 12 in.

The raised 28 x 30 cm patch provide good match even if we are standing close or over it.
Thank you for your suggestion.

http://www.FrancoisMethotCV.com