HFSS and Coax Feed Question
I am fairly new to EM and Antenna modelling and simulation, therefore, I beg a pardon on some of the following non-sense questions that I might inquire.
Recently I have been playing around with HFSS in order to simulate a patch antenna. The patch antenna is fed by a coax cable. To perform this simulation I followed the Help on the HFSS manual examples and set everything accordingly. Regarding the antenna simulation and modelling I understood its working principle. My main question is on the coax feed; according to the HFSS setup, the coax feed is simply set by using a inner conductor (PEC) with 0.7 mm radius and an vacuum outer cylinder with 1.6 mm radius, both considering 0.5 cm length. After this, we set up a circle with the same radius as the outer cylinder at the end and assign a wave port excitation. My question is concerning the behaviour of such structure, since in real life we will have both inner and outer conductors as well as a different dielectric. Since we do not set up the external shield, how can the simulator know the excitation mode/way we employ?
Other doubt I have concerns the usage of such feeding method when exciting either unbalanced or small antennas. Since this situation usually leads to higher currents on the outer shield of the coax cable, I am predicted that this kind of simulation won't be precise enough to predict the real world behaviour. Any thoughts on that? Also, why do antennas with small ground planes or electrically small usually force large currents on the outer shield of the coax?
I would sincerely appreciate any help in order to better understand this conditions.
Thank you for your time.
Kind regards.
you are right to be suspicious of an HFSS simulation.
In real life, RF currents run on the OUTSIDE of the coaxial line (generated at the coax shield - to - antenna ground plane connection), and act like a second antenna element (which totally screws up the radiation pattern). Unless you model a length of that cable also in HFSS, you would not see any of that real world effect.
After a quick calculation I saw that the conditions that HFSS example indicates (Inner radius = 0.7 mm, Outer radius = 1.6 mm and er = 1 ) lead to a 50 Ohm coax cable :). However, I have still not found answers for the remaining questions.. Again, thank you for your time.
Using such internal ports is a common way to model antenna excitation. Pro: it is efficient to model it this way. Con: as you already pointed out, this "ideal" model that doesn't account for common mode signals that might be exist between a coax shield and an external ground. However, to get those effects, you would need to model the antenna with coax of reasonable length and realistic ground environment. That's usually overkill. The typical approach is to use internal grounds in simulation, and avoid unwanted modes in the hardware by using proper baluns or ferrites.
Thank you very much. In order not to create a new thread, I will ask here another question hoping you guys can clarify me :)
During the simulation of a patch antenna I am viewing both the Vector Electric Field and surface J. The surface J current models the charges movement and density around the area (e.g. ground or square patch), which I understood, I guess. However, my question is now regarding the Electric Field Vector. When I plot both on ground and patch, I get almost nonexistent field on the ground, while the patch demonstrates to have the expected fields. My question is: How are those electric field vectors calculated? Are they the resulting electric field of the net charge on each unit area, or are the electric fields resulting from the interaction of both the patch element and the ground, i.e. the overall antenna structure? Because the latter would make more sense to me, which reasonably explains why the ground plane has no vector electric field.
Any help on this?
Again, thank you for your time.
Best regards.