Hello.

This is a fundamental yet important question.

Where to begin?

I have been working on a stacked patch antenna. In the substrate definition model, the groundplane is listed as an open boundary with infinite dimensions.

When I simulate my antenna with an infinite groundplane, I notice very smooth curves and well-defined vallies.

However, I know this is not the case for finite groundplanes. On ADS's help section, they list methods of implementing finite groundplanes.

This is important because a finite groundplane will introduce odd and even harmonics.

One method is to include driving your metallization layer with an internal port. The finite groundplane is modeled as an additional cond layer. You add port(s) to the finite ground sheet and label them as ground reference ports with respect to the driving port on the patch antenna.

So I did this and ran an adaptive sweep from 0-10 GHz. It simulates but shows that for certain frequencies, there exists an "unphysical bahvior" and notes other warnings such as coarse mesh size, grouped slot and strip ports, etc.

Now the notes says that, this condition exists where a metallization layer is absent in the presence of groundplane.

So If I sweep from 0-4 GHz. The problem disappears.

So I have a problem: I get three different plots, one with infinite groundplane, one with finite(0-10 GHZ with unphysical behaviors for some frequencies) and the other with no warnings.

On my finite ground sheet plot, my resonant frequency is off and my impedance BW is well above VSWR=2:1.

On my two finite ground planes, I get 200 MHz of impedance BW and an S11 between 19-22 dB down. But the sweep from 0-10 GHZ has some anomalies at higher frequencies.

So, which plot do I trust? Has anyone run into this problem? what is the best and correct method for implementing finite groundplanes and properly grounding them?

Any help will be greatly appreciated.

This is a fundamental yet important question.

Where to begin?

I have been working on a stacked patch antenna. In the substrate definition model, the groundplane is listed as an open boundary with infinite dimensions.

When I simulate my antenna with an infinite groundplane, I notice very smooth curves and well-defined vallies.

However, I know this is not the case for finite groundplanes. On ADS's help section, they list methods of implementing finite groundplanes.

This is important because a finite groundplane will introduce odd and even harmonics.

One method is to include driving your metallization layer with an internal port. The finite groundplane is modeled as an additional cond layer. You add port(s) to the finite ground sheet and label them as ground reference ports with respect to the driving port on the patch antenna.

So I did this and ran an adaptive sweep from 0-10 GHz. It simulates but shows that for certain frequencies, there exists an "unphysical bahvior" and notes other warnings such as coarse mesh size, grouped slot and strip ports, etc.

Now the notes says that, this condition exists where a metallization layer is absent in the presence of groundplane.

So If I sweep from 0-4 GHz. The problem disappears.

So I have a problem: I get three different plots, one with infinite groundplane, one with finite(0-10 GHZ with unphysical behaviors for some frequencies) and the other with no warnings.

On my finite ground sheet plot, my resonant frequency is off and my impedance BW is well above VSWR=2:1.

On my two finite ground planes, I get 200 MHz of impedance BW and an S11 between 19-22 dB down. But the sweep from 0-10 GHZ has some anomalies at higher frequencies.

So, which plot do I trust? Has anyone run into this problem? what is the best and correct method for implementing finite groundplanes and properly grounding them?

Any help will be greatly appreciated.

"Demystifying Ports and Grounds in Momentum":

https://edasupportweb.soco.agilent.com/ ... ref=search

The document number is 315140.