radiation of pcb antenna placed between two FR4 layers

[aguijon]

Hi everybody,

I have a question that puzzles my brain for a long time. I am accustomed to see patch or other types of printed antennas placed on top layer of pcbs. But few weeks ago I saw a microstrip koch fractal antenna that is placed in the second layer of the pcb. (between two dielectric layer). I firstly thought that it should be a trick to reduce the dimensions of the antenna more. But I have still no idea how radiates this antenna in this way and what are the drawbacks of using printed antennas when upper and lower layers are FR4 or other dielectric material except for air.
Also, I googled it on the internet but probably I couldn't find the most convenient search term to get satisfactory results. So, It would be very useful if anyone could address a document or website about this phenomena.

Thanks a lot.

 

[davenn]

I don't have any www sites to aim you too

But in the past have been involved in the installation of antennas above 1 GHz that have been in a fibreglass radome
and the principle is not really any different to what you are describing

The dielectric is going to affect the radiation resistance values and the electrical length of the antenna ( Im not into the maths of that )
It is well known that adding a dielectric ( radome etc) increases the antenna to air capacitance and therefore lowering the resonant freq of the antenna

We found that the antenna elements had to be cut a little shorter than what their free air electrical length would be so that once encased they would be on the correct freq

I'm sure there may be one or 2 antenna theory experts on here that could show or point you to the math/formula for working this out

Dave

 

[FvM]

I doubt that the behaviour of the "embedded" antenna structure is completely different from the respective surface placed design. You say it's a microstrip, so the distributed capacitance would primarly depend on the substrate height above the ground plane. The upper substrate height has some effect mainly on smaller traces. You can refer to PCB impedance caculation tools to get an idea.

To determine the exact antenna behaviour, you'll use a 3D field solver.

 

[vfone]

Usually Koch fractal antenna don't need a ground plane underneath, as the normal patch antenna needs.
So, you can place the active element on either side of the substrate, from the moment the antenna design take care about nearby layers and components.
Overall will be a big compromise between antenna performance and PCB size.