AnsweredAssumed Answered

momentum SMD port calibration

Question asked by rinc on May 20, 2014
Latest reply on May 28, 2014 by rinc
Hello everyone, here i have a question about simulate matching network using real lumped elements.
Here are some information i use: ADS 2011.10, substrate: Roger 4003c with er= 3.65, thickness= 0.504, conductor thickness = 17 um., Frequencies from 2 to 3 GHz
From these information, i calculated optimum width of 50 transmission line is: w=1.12 mm. (using momentum)
Here is my schematic of matching network for my Power Detector: 
Fig1: simpleMacth1_CouplSche: the first 2 parallel S2P files are identical inductors with values of 1nH. The third S29 data item is shunt resistor of 56 Ohm. And the rest is my detector. The matching network is all 3 S2P data items. All these files i got from manufacturers, so it should include parasitic effect. Thats why i use SMD calibration instead of Delta Gap calibration (still keep default values of 50 Ohm for every ports)
Now i realized this in layout, and here is the simple one:
Fig2: Layout1: Port 1 is RF port, set as "TML" in EM setup, "Port" 2, 3, 4 are relevant to matching inductors, matching shunt resistor, and first diode respectively. These 3 ports are set to SMD port, with very small additional area which has same widths of the lumped package (0603, width = 0.8mm, i add small length: 0.001 mm).
Then i create EM model, and applied it to schematic to compare with my "ideal" schematic simulation.
Fig3: Sche1: EM model and schematic setup
Fig4: Sche2: ideal schematic setup
Here is the result to compare Sche1 and Sche2 (Fig5: Comparison) 
I agree that there will be some slightly changes between 2 simulations, but according to the smith chart, this change is dramatically, or other words, destroy my circuit. 
I dont know where did i make some mistakes? Could you please try to help me.

btw: Fig1 and Fig4 are the same, just replication