I've wrestled :'( with this problem for a week. Now, I'd like your help.

I designed 2 stripline elliptical LPFs with Fc=1.35GHz/2GHz. I'm interested in seeing the rejection up to 7th harmonic frequencies. I started with ADS linear sim then Momentum RF with infinite ground. Then came trouble. I simulated the same circuit with finite ground stitched with vias to tie the top and bottom metal. Then I added 2 ground reference ports for input and output ports. The result that I got after an agonizing wait didn't quite resemble the result I got from sim with infinite ground plane (which was pretty close to the linear sim). I did try breaking the circuit down in to 2. The finite ground sim still didn't resemble infinite ground sim.

I've checked with an apps engineer who agreed that I set up the sim correctly (he saw my file). I haven't tried this MW mode yet because I'm not patient enough.

Apps engineer told me that most people don't try to do this. He said they just trust sim result with infinite ground plane. I find that hard to believe.

Is that true?

I'd appreciate your advice.

I designed 2 stripline elliptical LPFs with Fc=1.35GHz/2GHz. I'm interested in seeing the rejection up to 7th harmonic frequencies. I started with ADS linear sim then Momentum RF with infinite ground. Then came trouble. I simulated the same circuit with finite ground stitched with vias to tie the top and bottom metal. Then I added 2 ground reference ports for input and output ports. The result that I got after an agonizing wait didn't quite resemble the result I got from sim with infinite ground plane (which was pretty close to the linear sim). I did try breaking the circuit down in to 2. The finite ground sim still didn't resemble infinite ground sim.

I've checked with an apps engineer who agreed that I set up the sim correctly (he saw my file). I haven't tried this MW mode yet because I'm not patient enough.

Apps engineer told me that most people don't try to do this. He said they just trust sim result with infinite ground plane. I find that hard to believe.

Is that true?

I'd appreciate your advice.

You don't have to stitch finite grounds together with vias to get a finite ground - just define a box or waveguide around your filter to simulate package effects.