What is "Trap Ratio"? How does this relate to etch factor and angle?

The "trap ratio" that we ask for in synthesis mode merely refers to the ratio between the width at the top of the trace conductor and the width at the bottom of the trace conductor. This is a much simpler and more direct number than the etch factor and angle. The conversion between our "trap ratio" and the etch factor and angle are as follows:

My insertion losses never matched with field solvers, so I always thought the dielectric loss was higher than my vendor's data would suggest. Gauss 2D actually matches my testing, but seems to show high conductor loss, not dielectric loss. Why is this?

What your testing is indicating is, in fact, a higher conductor loss than you would have anticipated. Quite simply, the insertion loss is the sum of the conductor loss and dielectric loss. If the insertion loss is not correct, that must mean either your conductor loss or your dielectric loss is incorrect. Your earlier conclusion that it was the dielectric loss that was incorrect was based on the assumption that the conductor loss is modeled correctly by the other field solvers you were using and so the error must lie in the dielectric loss, which is dependent on the data you input in your solver.

However, the conductor loss was modeled incorrectly in those other solvers. This is because they were predominantly developed at a time when ignoring ground plane losses would not have shown significant error, because the frequencies at which the industry was operating were quite low compared to today. Ground plane losses become significant at frequencies above ~1 GHz, because of the "proximity effect", which results in an a narrower return path through the ground plane, leading to additional significant losses in this return path. This is amplified by decreasing trace thicknesses in modern day applications. Gauss 2D includes the ground plane losses in the conductor loss, rendering both the conductor loss and insertion loss accurate.

The Average Power Handling Capability that I obtain from Gauss 2D is lower than I obtained from post-processing the outputs from my old field solver? Why is this the case?

Again, this is the result of Gauss 2D accurately modeling conductor loss. Other solvers underestimate the conductor loss, so if you were to take that data and post-process it to try to obtain an Average Power Handling Capability (APHC) for your trace, you would have arrived at an overestimate, as compared to the accurate value Gauss 2D provides in its APHC calculator. If you had been provided with accurate conductor loss in the past, your post-processing calculations would have been in line with Gauss 2D.

The danger of an overestimated APHC is that it can lead to deployment of boards into applications that require higher power handling capability than the lines can actually handle within your specified operating conditions. With Gauss 2D, you can be much more confident in the power handling capability of your traces.

Do you provide S-Parameter Outputs? What about Mixed-Mode S-Parameters for differential traces?

Yes. Gauss 2D provides 2-port s-parameters for single ended traces and both 4-port s-parameters and mixed-mode s-parameters for differential traces.

Can I use your RLGC outputs for SPICE modeling?

Yes, you can. In the session window, we provide the RLGC, formatted as needed for SPICE, as in the example below:

Does Gauss 2D calculate frequency-dependent properties?

Yes, Gauss 2D calculates the quasi-TEM mode frequency dependent behavior for Microstrips and allows you to project frequency dependent losses, dielectric properties, impedance, and RLGC for Striplines, such as in the example below: