Issues with strange resonance in S11
I get a strange resonance in S11, when the strip to groundplane (zmin boundary) distance is increased.
For 60 um it is ok. For 180 um the resonance comes up.
Any ideas.
You should indicate the complete circuit. In other words, indicate a voltage source and where you connect both the + and - terminals. It is possible that your problem is that what looks like ground planes to you (the two plates with holes) are really patch resonators. Whether or not this happens depends on how you connect your voltage source to the circuit.
Once you have shown us how you connect the voltage source, you should indicate a complete path for your current flow. The complete path will be from the + terminal of your voltage source, through the circuit, and then back to the - terminal on your voltage source. The path back to the - terminal is called the "ground return path". It is just as important as the signal path. If you do not have a complete path, current will not flow. You might also want to indicate how you terminate all other ports, short circuit, open circuit, 50 Ohms, etc., in order to allow a complete path for current flow.
I find that for problems like this, once you go through the above steps, the cause of unexpected problems often becomes clear.
@rautio
It is a via hole connected with two microstrips; each strip has its ground plane.
There are waveguide ports (not visible in the pic) at each strip, which excite the quasi TEM mode.
If the distance of the whole configuration to the bottom end of the calculation room is too far, I get this strange resonance.
The bottom end of the calculation room is also metal (not visible in the pic).
Thanks for reply.
Hi elektr0
Try to connect the two ground plane like this just to see if it solves the problem.
If your "ground" plates are acting as resonant patches, plasma's suggestion should help. Metal in the locations where he added the red lines will give the ground current a complete path. Without metal there, the ground return has an open circuit in the middle of it. Having an open circuit in your ground return has the same effect as an open circuit in your signal line. Keep in mind that even with the extra metal, the plates could still give a problem.
Good idea to view the current distribution at resonance. If the plates are the resonators, they will have lots of current on them at resonance.
If the outside edges of the plates are connected to perfectly conducting sidewalls, and your ports are also at those sidewalls, the plates probably won't be the problem. If your circuit is in a conducting box, it might be a box resonance. To get a box resonance, you need at least 1/2 wavelength in two dimensions. Free SonnetLite (www.sonnetsoftware.com, I work for Sonnet) includes a box resonance calculator if you want to know the resonant frequencies of your box (with dielectric, without circuitry). Your circuit has too many layers for the free version of Sonnet, but if you can get the full version of Sonnet, you can check your results.
It might happens what he is seeing is not resonance
is just a transition. The signal comes from a transmision
line then find a transition that is not matched and
then another matched TL. And it might happens that
is coupling to the other TL. I think you should separate
both vias, make them look as close to a coaxial as
possible.