Coupler Position Simulations

The full geometry of the position cavity was used in a lossy eigenmode computation to simulate the effects of moving the coupler. The full experimental results are at ClicCouplerPosition.

In simulation, the coupling is more sensitive to the feedthrough-waveguide spacing but the frequency is less sensitive. This may be due to the rounded ends of the feedthrough not being modelled accurately in simulation.

Coupled Resonators

The waveguide-coaxial transitions create resonances that vary strongly with feedthrough-waveguide separation in both frequency and quality factor. The external quality factors of these modes are lowest when their frequencies coincide with the cavity dipole resonant frequency. it is here that the coupling between the resonant modes is strongest. This is why the strongest coax-cavity coupling is measured at ~0.1mm feedthrough-waveguide separation. It must therefore, be determined, how great an influence these modes have at other separations.

This effect was also investigated at finer spacing. This confirms that the lowest external quality factor for the waveguide-coax modes is seen where the resonant frequencies of the waveguide-coax and cavity dipole modes. This is also where the lowest measured loaded quality factor is.

Low Conductivity and Lyapin Feedthrough

The simulations were repeated, once with a lower surface conductivity that brought the loaded Q down to the measured value and once with a bead at the end of the feedthrough.

Further work to understand these effects can be found here.

Mesh Dependence

The split in mode frequencies is due to the accuracy of the meshing. The dependence of the mode splitting and resonant frequency on mesh density is shown below. The accuracy is not sufficient for temperature simulations and so for these, the mesh will be made symmetric and dense and the solution may have to be third order.