In this period I had the opportunity to gain different experiences in the design and construction of antennas. In particular, the latest experiences have focused on the study and design of patch antennas and antenna arrays.
The last opportunity, after the realization of the project for the Wireless Electromagnetics Technologies course, allowed me to get in direct contact with the university laboratories and to personally create an integrated antenna after having designed it.
CAD design
The first step to use a numerical control cutter is to create a 3D design that allows correct processing, both for the CNC and for the assembly. In the considered design, the weak point is the shorting pin which, considering the available instrumentation, cannot be created inside the substrate. So this prototyping began with a redesign that allowed its correct assembly.

Starting from the original design, the patch was moved to the edge so as to have the side of the short circuit at the edge of the substrate and to be able to make it through adhesive copper. The other dimensions have not changed.

The gerber file was then imported into the control software for the milling machine and cutting started.

A 4MILL300 from MIPEC and the accompanying software, TrackMaker, are used.

Assembling the patch
After cutting the patch antenna you need to assemble it. In other words, we proceed to add the SMA connector by soldering it. The shorting side is then added using an adhesive copper foil and some solder points.
Performance
The last step was to test its performance. Using a Vector Network Analyzer from pico Technologies it was possible to measure the reflection coefficient. After a SOL (short-open-load) calibration it is possible to connect the antenna to the cable calibrated at 50 Ω.

Then the reflection coefficient is obtained and its performance can be analyzed.

Final remarks
There are some points that have led to differences between the design specifications and the actually measured characteristics.
The redesign phase leads to a first shift of the resonant frequency from what was measured in the project.

The project was carried out considering an FR4 with an electrical permittivity of 4.8, slightly different from what is available in the laboratory. Furthermore, during the cutting phase there was a machine error which led to producing a slightly less thick substrate in some places by introducing an air gap.