How to predict groundplane effects on Patch Antenna? 1

[Hilux]

Can anyone point me to a source of information regarding the effects of groundplane size and shape when used under a ceramic patch for GPS?

I am finding a considerable shift in the antenna's resonant frequency with different groundplane dimensions and I am concerned that the performance of our equipment (a GPS dependant device) may be compromised by its surroundings. For instance if it is tuned to work correctly while 'hand held' using the pcb structure as a groundplane then what might happen if it used mounted on a larger metallic surface or over other good grounds?

The equipment groundplane is approximately 50 x 100 mm and cannot easily be varied. The antenna feedpoint will be about 20mm from one end and close to the longtitudinal centreline.

I cannot find much of use on the internet so far so I'm open to suggestions.....

Thanks in advance,

John

 

[Higgler]

That's a common problem and one great reason to have a wider bandwidth antenna (if possible) so that in-situ shifting won't lose you so much gain.

The groundplane size will change your gain and I have articles that show that change, a simple curve with zero dBi gain for a 1 lambda ground plane, and +5 dBi Gain for a 4 lambda ground plane. You typically need a half wave ground plane to minimize frequency shift.

But adding a human hand is a lot of dielectric loading. Fat= Er6, Muscle = Er 30, Blood = Er 50, water = Er 81, that's alot of loading. The FCC has a website with more details.

You won't find much frequency change vs. ground plane size info since it depends alot on the antenna and it's bandwidth. You will see antenna pattern changes in a dipole mounted different distances from the ground plane edge.

Using a thin ceramic to shrink your antenna is the worst possible combo for sensitivity as you have found. I can see why you'd do that though. It would be ok without the hand placement retuning it. Cell phones designers often avoid patches and use PIFA's for this reason. PIFA's are similar size but wider bandwidth and can handle environment changes better. It's essentially having a microstrip 50 ohm line go past the edge of the ground plane and widen out, then adding a short circuit to help tune it. If there's enough room, you can get huge bandwidth by avoiding the short circuit, i.e. 10:1 bandwidth with 2:1 VSWR.

The cell phone antenna designers also know that if you let alot of current get on your ground plane and tune the antenna for that situation, then changing the electronics board requires a new antenna design. Hence the better designs try to choke off the currents so that the circuit board and peoples hands won't have a large effect. That limit's size of the antenna too and is more space challenging.

I'd suggest two things;
1) empirically take your patch with the smallest ground plane, then add metal tape to enlarge your ground plane to evaluate effects. Probably will have so-so results in the real world though.
2) Switch to a wider bandwidth antenna, PIFA or printed monopole and set the lower frequency at L1 (assumes you don't need L2) without a hand touching it, then note the frequency shift with a hand in standard positioning and see how it does.
3) More complicated is a self adjusting varactor tuned antenna, PIFA probably. I've been testing a design whereby your gain changes when your hand touches it, but the frequency is not shifted and is set via a capacitor and short circuit. Pretty cool result and minimizes the overall loss due to the hand. This is a 925 Mhz antenna. The GPS equivalent sized one for L1 would be 0.58" wide by 0.2" tall addition to the end of your circuit board. So if you have a circuit board, you'd just lengthen it by 0.2 inches and print the antenna on the end.

khiggins

http://www.toyon.com

 

[Hilux]

Thank you Higgler, that's a very comprehensive response.

The GPS requirement is only single frequency so we don't need massive bandwidth other than to avoid the possible 'environmentally induced' variations but I will certainly have a look at the PIFA possibilities. We hadn't even considered any form of linearly polarised antenna because of the extra loss compared to the correct circular polarisation - and then went on to try a reduced size ceramic patch..... Hmmm, funny thing the human mind.

Thanks for the suggestions. We've already spent some time with the copper tape and it's difficult to draw too many firm conclusions from the results!

I like the idea of varactor tuning to compensate for changes but I think it is beyond the scope of this particular project.

So it looks like I will spend some time playing with PIFAs over the next few days to see just what can be done!

Many thanks for the suggestions - I will post any meaningful results on the forum for everyone's benefit.

John

 

[Higgler]

Hilux,
the wider bandwidth is only for the purpose of still being efficient when your hand is on the antenna and dielectrically shifts the frequency of efficient operation, not for having wider bandwidth per se.

Most people think you need circular polarization. Not so, linear still gives plenty of S/N and is much easier to make.

Typically you can take a good Circular antenna and add a 10 or 15 dB pad and it'll still work with a good GPS receiver. It just slows acquisition times with a weaker signal.


Good luck,

kevin.