RFID Patch antenna 1

[ElCap]

Hey guys,
im designing and creating a rfid system as a project with a few friends.
i have the task of designing the reader antenna.
were operating at 915MHz and i was gonna design a patch antenna on a pcb as its simpler.

What software is there to design this kind of antenna?
(freeware would be best!!!)

and any design help or tutorial would be appreciated!

 

[VEBill]

Why is a patch 'simpler' if you need software and you have to make a PCB?

A monopole or a dipole would be simpler. All the 'software' you need is L(ft)=k/F(MHz) where k is 234 for a monopole or 468 for a dipole. The only tool you need is a pair of wire cutters. If you need more gain, then fold some cardboard & tin foil into a corner reflector.

Even if you really really want a patch antenna for the final system, then I'd recommend that you also make a dipole for initial testing and trouble-shooting.

 

[waveboy]

Hi,
At 915MHz, your monopole would measure about 8cm. -However, if you make a planar antenna on a PCB, you could make it a spiral antenna, which would be generally smaller and more compact. -A spiral antenna is a little like taking a dipole/monopole and then "twisting" it into a spiral. The compactness could be an advantage in a typical "RFID" application. The PCB could also offer advantages in that it would give support to the spiral element. Also, it would be easy to make a spiral with a ground plane behind it. Remember that if you decide on a "balanced" configuration of antenna, you may also need to design a suitable balun. -the design criteria for a balun for a "near-field antenna" (such as an RFID antenna) is presumably the same as for a 'normal' far-field antenna(?).

You can make your own PCB antennas (quickly) by buying (or mixing) some etchant and etching a spiral on a sheet of FR4 coated in copper. -You have to use any math/drawing program to draw a suitable spiral/patch, then use some kind of 'photo' method to expose the FR4 with this shape on top before etching it away. (I have abbreviated this technique for brevity). Am i right in saying that you may be advised to make a prototype, test it, then modify it -rather then suffer for long periods with antenna design software which is rather "difficult" with spiral antennas?

 

[MikeHalloran]

A CNC mill can make spiral antennae faster and more precisely than etching, especially if you want to use thick copper.





Mike Halloran
Pembroke Pines, FL, USA

 

[waveboy]

I think the nature of typical RFID applications dictates that the RFID 'tag' is small and unobtrusive. Due to this, I suspect it would be unlikely to use CNC for RFID patch or spiral antenna. Usually they are cut out of copper sheet with lasers or etched out of copper-coated FR4.

In fact an RFID antenna is usually electrically small. It is not usually the most optimal size for the frequency that it "responds" to. -Thus everyday metal objects which are nearer to wavelength size can activate the RFID receiver.

 

[Higgler]

I think it's a "reader antenna", which is a full size efficient antenna. I would think from a company cost saving perspective that you'd just buy one for $200. Being so common nowadays, there has to be alot available.

But if that's not the case, a patch antenna is so simple, you don't need any computer program to design and build one. I think you should make the antenna yourself. You probably want circular polarization.

Rogers 6002 material is the best temp stability and was made specifically for patch antennas, but for lower material cost you can use low cost FR4 (McMaster Carr) and mount it on spacers and solder a cable/connector to it.

Do you have a network analyzer and antenna test facilities?

kch

 

[ElCap]

yeah its for a reader antenna.
and yes we do have access to a network analyzer and antenna test equipment.

been doing a few calculatuions for it and it seems that for a patch antenna at 915mhz that it would be about 30cm X 30cm, which is a bit big for our application.

ive heard about a short-circuit patch antenna design which would reduce the size to about 1/2 to a 1/3 of the size of a normal patch antenna. is this correct? is the design of this difficult?

 

[VEBill]

A dipole for 915 MHz would be about 15 cm end-to-end and as thin as you like. With well-known tricks such as end-loading , it can be made significantly smaller than that while maintaining high efficiency.

PS: I suspect a patch can be made smaller than your calculations indicate.

 

[Higgler]

El cap, you have a dimension inaccuracy.

wavelength is 11.803 inches/frequency(Ghz), so one wavelength is just over a foot.
Standard patch in air is one half wavelength or just over 6 inches. Using FR4 (dielectric constant of about 4) makes your antenna only about 3 inches in size for a "half wavelength patch antenna".

I recently made a 4 inch long quarter wave patch at 433 Mhz, it took about 2 hours to prototype and test. Very little effort is required. About 5,000 of them have been made so far with. These are the shorted patches you wrote about too.

Hence you're antenna can be as short as 2 inches or 5 cm if needed.

kch

If you want more info, look up Toyon one the internet and call me.
Kevin Higgins.

 

[ElCap]

yeh i was including the ground plane as well in my calculations but yeh i was exaggerating a bit.

how would i go about making a shorted patch?
what changes would have to be made to the design with respect to the basic 1/4 wave microstrip patch design?

any specific websites or material you can point me in the direction of?

thanx

 

[Higgler]

ElCap,
If you know how to make a standard half wave patch, make one, then cut it in half and short out the part you cut. That simple. Bandwidth and gain shrink since the antenna is smaller. It'll takes some cut and try of course to perfect the VSWR. Gain will be about +3 dBiL typically and the antenna pattern in the E plane is slightly assymetrical, more gain on the open side than the shorted side.

kch

 

[ElCap]

are u talking bout a 1/4 wave microstrip patch?
cos i know how to make those (sort of).
i was sorta asking bout a 1/8 wave one.
is that possible?

 

[Higgler]

1/8 wave is easy, use FR4 for dielectric and your total length is 1/4 wave/sqrt(4). Your antenna should end up approx. 8 inches long x 4 inches wide x 0.062" thick.

A former coworker (may he rest in piece), when asked "George, I realize you need to analyze antennas then build them, but when is the best time to build one?", Georges answer was always "immediately". As an antenna Engineer, I agree with George on many simple designs. You can get prototype dimensions immediately from books or the web. Building and testing a prototype, then matching computer software to it, then optimizing the computer for a final design has been standard in my company.


kch