Double PCB antenna one beaneath the other
I want to design a new board that is going to have two pcb antennas in diversity but the problem is that I don't have enough space to my board. The problem is that a cell coin battery is going to be included also on the board and this creates a massive ground plane underneath of one of the two antennas. In other words a have free space (no battery underneath) for only one antenna. The two antennas are connected to the transceiver through a switch so one antenna is going to be used at a time. I am thinking of using this one http://www.ti.com/lit/an/swra117d/swra117d.pdf . So I was also thinking to put the two antennas one beneath the other. This is going of course to affect the behavior of the antenna but how much and how? I mean is they are still going to be functional? Any other suggestions?
Also what is the smallest solution for a pcb antenna at standard ISM band 2.4 Ghz?
I won't expect a reasonable diversity effect if the two antennas aren't at least λ/4 (about 30 mm) apart.
There are designs where even-though antennas are close to each other diversity still works like this http://www.sparkroberts.info/?q=node/10. So what do you think regarding my first question? As for the smallest antenna I think the best choice is an inverted -F antenna right?
Selecting antenna for diversity is depending on what type of diversity. Common types: space, time, directive or polarization diversity.
A small PCB antenna: www.antune.net/demo/bluetooth/
Do same antenna in nearest other PCB corner. It gives some isolation between antennas for diversity.
Another possibility is polarization diversity. It can be same antenna structure with dual feed or separate antennas. Simplest type is crossed dual dipoles. It is often complicated to design as embedded antenna/s as it needs VNA and skilled measurements.
For space and time diversity can most types of antennas be used, if antenna distance is big enough.
A quarter-wave antenna use ground as part of the antenna so actual PCB size and antenna placement relative the other antenna, can then be critical for best isolation/diversity.
The guys from the posted PCB link (Spark&Roberts), they can place antennas at any distance lower than λ/4, but doesn't mean the antennas would work properly in a diversity system.
When I design a diversity system usually I am looking for about 20dB isolation between antennas, and if both antennas are resonant, this barely happen for a distance shorter than λ/2.