How to use Greens function to implement ground planes and dielectric substrate for stripline filter using EDITFEKO?
I'm trying to simulate a 6th order parallel-coupled bandpass filter on stripline using EDITFEKO but I am not getting the correct response. The filter example was taken from the attached slides (pg. 26 - 29) and all measurements were converted to meters with two edge ports and excitations at the input and output lines [touching the top and bottom conductors] (EDITFEKO and POSTFEKO files attached). Since EDITFEKO requires that every connected edge have the same length, even when all the associated sections are 'unioned' (i.e. same labels; 'LA' card), correctly implementing practical ground conductors is quite tedious and so I'm first trying to make use of two ideal infinite ground plane models for the top and bottom conductors of the structure, as I've done successfully before for microstrip implementations [but only using one ground plane] (EDITFEKO and POSTFEKO files also attached). Is my stripline implementation incorrect or is it not possible to implement stripline technologies using FEKO's Greens function approximation or do I need to manually implement a finite substrate and ground conductor (enclosure/box)?
Answers
-
It is interesting that you prefer to use EDITFEKO for model construction.
In general your meshing is very coarse. A rule of thumb is that the triangle edge lengths should be similar to the gap widths.
In the below filter, here using a finite substrate, the meshing had to be really really fine, but the requirement will differ from model to model.
Try rerunning the model with 50% more mesh elements. If the results are stable then the mesh is sufficient in your specific case.
Also have a look at the following two models in in the Example Guide:
E1: Microstrip filter
E2: Microstrip coupler
0