I'm trying to design a matching network for the MMG3006NT1 and am a bit confused by a possible disagreement between the datasheet's reference design and the S-parameters data file available on NXP's website.
Look at the input matching solution given on the data sheet for 900MHz. If you assume the source is 50Ohm and compute the impedance at the output of the given matching network, it appears the reference design transforms 50Ohm to about 4.0 - j14Ohm.
On the other hand, the S-parameter file I downloaded for the MMG3006 gives a magnitude and phase of S11 at 900MHz of |S11| = 0.95, arg(S11) = 177deg. This is a complex form this reflection coefficient is -0.94 + j0.05, which works out to an impedance of 1.38 + j1.35.
What am I missing here? Are the S-parameters in that file de-embedded? I just cannot see how one gets from the given S-parameters to the given matching solution and until I can see the linkage its very difficult to use the S-parameters to optimize for different frequencies.