RFIC design： problem with FEM ports definition

What is the correct way to define TML calibration ports, when using the same ground metal for multiple transmition lines?
I am designing a distributed amplifier, I am using the lowest metal (Metal1 in this example) for ground and the top metal for the transmission lines (I attached a picture to better discribe the setup).
I am using the differential ports to set Metal1 as ground - for each port I set two pins - one targetting the Metal1 and second targetting the TopMetal and then I set the pin targetting the Metal1 to be negative potential for the pin targetting the TopMetal. I do this for all the 4 ports. With this setup I chose the resulting ports to be TML zero length (because I want no extra inductances added).
If chose Momentum simulation, I get an error saying:
"A calibrated pin must not be positioned on an edge on which other pins are positioned. (Using a single calibrated pin simultaneously in several ports is also invalid in this respect.) The pins involved are "P#" and "P#""
If chose the FEM simulation, I get “triangle ports” and warning saying: “Calibrated port P2 and port P4 are in almost identical reference planes. Still the reference planes are sufficiently misaligned, prohibiting ports to be co-calibrated or to become part of the same waveguide. Adjust the ports to exactly share single reference plane if co-calibration or waveguide behavior is desired”.
Further on if I make the pins to be edge/area pins and make the ports out of them, that solves the “triangle ports” problem, but for each port I get a warning saying: “Calibration will not be used for port P1 (pins P1 and P5 are not on the edge between a conductive and not conductive region)”

Falcon on Apr 14, 2015 5:49 AM

附图/附件

TWA.PNG

申明：网友回复良莠不齐，仅供参考。如需专业帮助，请学习易迪拓培训专家讲授的ADS视频培训课程。