CPW ports definition, explicit ground and union of metals

For GCPW (grounded coplanar waveguide), if the side grounds are connected to the bottom ground with a via, it is important to have the via on the port plane so that its effect is included in the 2D cross-section port solve. Otherwise, if there is no via on the port face, the port solve sees four conductors that are disconnected from each other and the mode is different from that of the GCPW.

the metal surface on the bottom is connected to the coplanar waveguide's ground planes through via holes, ensuring that both metal surfaces are at the same potential (i.e., ground). This is essential for maintaining the desired electrical properties, such as a consistent characteristic impedance and minimizing electromagnetic (EM) coupling effects.

Thank you for your reply. Actually, I was thinking that a "connect to lower" statement would be enough to satisfy the condition that you are suggesting. To be redundant, I used also additional via holes on the upper metal sheet, connected to the lower metal boundary, as you can see in the figure.

On the other hand, I am still confused on the opportunity to use a negative "-1" port to define a CPW, independently of the presence of a ground to have a GCPW. Do you find more reasonable results when using negative ports, or ports defined as "none" and not necessarily having a negative definition, like simply "2"?

By the way, to eliminate the problem of "metal union" still maintaining the memory of the previous separated metals, and forcing me in defining additional differential ports, I simply added a metal having the right size to close the metal boundary and now the differential ports is automatically extended over the entire section.

Are you seeing disagreement between the various modeling methods?

Below are shown the results of a quick-and-dirty simulation of a CPWG transmission line, using the AWR MESFET example process.  The dimensions are W=20, S=10, and L=200, and the EM simulator chosen is AXIEM.

The darker trace in each group is the Closed Form (CPW1LINE) model with Acc=10, and the other traces reflect the EM structure modeled with differential ports, as well as sequentially numbered ports, grounded at the schematic level.

The agreement between the two EM simulations is very good.

Are you seeing disagreement with your particular structure?  I guess if you are, I would encourage you to file a support case so that Cadence can troubleshoot it with you.

And just one added point regarding my previous post:

Grounding or floating Ports 3-6 in the schematic with the 6-port EM structure (with sequentially-numbered ports) resulted in a very minimal change in simulation results, since the rectangular ground bars I've added rather robustly connect the topside grounds to the backside ground plane.

Hi, I am still studying the differences between CPW and CPWG with Microwave Office AXIEM. Actually, for CPWG I solved the simulation problems using a number of via holes physically connecting the lateral ground plane on the top of the substrate with the bottom metal, to assure a ground to the studied structure. Nevertheless, I have a few conceptual problems: (1) the presence of via holes solved for me the comparison between a semi-lumped approach simulated circuitally (where the parameter GND=1 emulates the CPWG library elements) and the electromagnetic result; but, is this only a SW problem, or experimentally, I really need the via holes? (2) If I am simulating the simple CPW, is the differential port definition enough to have a "real" result, or do I also need more specific boundary conditions to try a direct comparison with the experiment?

Hi, I am still studying the differences between CPW and CPWG with Microwave Office AXIEM. Actually, for CPWG I solved the simulation problems using a number of via holes physically connecting the lateral ground plane on the top of the substrate with the bottom metal, to assure a ground to the studied structure. Nevertheless, I have a few conceptual problems: (1) the presence of via holes solved for me the comparison between a semi-lumped approach simulated circuitally (where the parameter GND=1 emulates the CPWG library elements) and the electromagnetic result; but, is this only a SW problem, or experimentally, I really need the via holes? (2) If I am simulating the simple CPW, is the differential port definition enough to have a "real" result, or do I also need more specific boundary conditions to try a direct comparison with the experiment?