Discussion Forum

You appear to be attempting to use a purely RF model to represent an RLC circuit, while expecting purely circuit-like behavior. I perceive this as an unusual thing to do. Is it, perhaps some kind of academic exercise and/or an attempted demonstration that it is possible? After all, you could instead more cleanly represent an RLC circuit with the "electric circuit" formalism in the AC/DC module (specifically, the "ec" capability in the AC/DC module.) Instead, your approach is (deliberately or not) including more couplings (parasitics) than an explicit circuit-connected set of discrete R, L, and C components, since the RF fields extend all over. And to do that right (if that is your intention) you will need to think carefully about specifying your computational boundary conditions, the size of your computational space, and the details of your mesh. The EC formalism avoids all that additional complexity. Of course, if all you want to do is model a pure circuit, you should probably use one of the many free or low-cost SPICE codes (see https://en.wikipedia.org/wiki/SPICE) available, rather than Comsol Multiphysics. I think it is fair to say that the EC capability within Comsol Multiphysics is essentially an abbreviated version of a SPICE code, set up for convenient interfacing with FE computations. If you don't have any FE computations to join to it, it would be a bit like using a cement mixer to drive yourself, alone, back and forth to work each day; there are more appropriate tools for such a simple task.

Remember that a lumped circuit assumes frequencies low enough that there is no variation/wave across each component. Without opening your file, I would suggest you make the comparison at reasonable wavelength relative to the size of your components.

Thank you for the replies,

I am trying to simulate a more complex system of a helical resonator using the RF interface, and I am trying to add a capacitance to my system through a lumped element in order to see how the resonance frequency and Q factor change with the added capacitance. Both should decrease like 1/sqrt(C), and the resonance frequency was behaving like this, but the Q factor was increasing as I increased my C.

I was thinking I might be using the lumped elements incorrectly, so I wanted to create this model as a very simple case to confirm whether or not I was doing this correctly.