Learning Forum

[armzn2021]

Hello,

I am trying to calculate the self-resonant frequency of an inductor (simple spiral). However, in Ansys Maxwell, electric or magnetic type solutions are separated. For instance, in Eddy current, the value of AC resistance and inductance can be calculated but the parasitic self-capacitance cannot be calculated. Similarly, the inductance will not be calculated in the electrical studies of Maxwell.

[armzn2021]

Please reply.

[AndyJP]

The equivalent circuit is WRONG. I do not know how long that ignorance from 19th century will continue replicating. The capacitance grows with the coil length, and in the circuit it decreases. The capacitance is a multi-dimentional solution of Maxwell equations and can not be approximated simply by wire layers. At RF it may be approximated in cylindrical coordinates through the wave equation with periodic boundary conditions. But the good solution requires FEM simulation.

[AndyJP]

[armzn2021]

Thank you for your reply. I did not mention "Equivalent circuit" anywhere in my question. I was looking for the FEA-based calculation to see whether I need to use HFSS or COMSOL.

[AndyJP]

I got impression that you consider the image of the equivalent circuit attached seriously. As far as I know, COMSOL may give better results at mm waves, better converging, smoother detailed resonant curves.... however, setting up comsol is another kind of art, and that depends.... I do not use comsol. But "once upon a time" my coworker used and got better data much faster.

[armzn2021]

Thank you for your reply. I used to work with Ansys Maxwell (I'm a power electronics engineer) but it was not clear how to define excitation in HFSS. Would you please help me what is the best way to define excitation in HFSS (for this kind of simulation)? I appreciate it if you send me any simulation files if you have any. Thank you in advance.

[CB]

For your simulation needs I would suggest HFSS or Q3D. Do you have access to these Ansys Tools? For your HFSS port excitation - using a lumped port is a good place to start. For examples you can view the Getting Started guides or some of our wizards may help you with model and result generation.

[AndyJP]

wires are finely simulated with lumped ports. But speaking of pure RF resonance, I was using a manually designed "test waveguide", and Floquet ports for plane TEM wave imitation.