SP or PSS+PSP to measure input/output/inter-stage impedance?

I am trying to measure input/output/inter-stage impedance for a Class E Power Amplifier working at 433MHz (switch mode PA) in Cadence. Should I use ZM in SP analysis or ZM in PSP analysis? Which one is correct?

In the driver stage, there are two parallel LC tanks which are tuned at fundamental and the 3rd harmonic. The problem is that the real part of the output impedance of the driver stage is negative 700 ohms at the fundamental by using ZM in SP analysis. If I use PSS+PSP, I got around positive 5 ohms for the real part. So does the negative real part mean that it is because the two LC tanks? Then does it mean I should use PSP instead for measuring impedance for my PA?

If I use PSP, it is weird that the L matching network does not (or barely) change the impedance measured by PSP analysis. Do you know why? What shall I do to match up the impedance measured by PSP? Thanks a lot!

There is no tool which name is Cadence.
Use correct terminologies.

ZM in PSP analysis.
Here you have to consider that operation points of circuit is affected
by not only drive power but also port impedance's value.

Yes.

You have to do both SP and PSP Analyses.
Negative impedance by conventional SP-Analysis shows possibility of unstable in quiescent state.
Again surely see my appends in https://www.edaboard.com/thread190138.html

No one can answer unless you show your circuit information and simulation conditions.

Yes.

About measurement of interstage impedance, see The Designer's Guide Community Forum - time variant impedence in a passive mixer
Here you have to use PSS/PAC not PSS/PSP.

Well, here is how I set up PSS and PSP.
PSS:

PSP:

Did I set it up correctly for input impedance measurement for Class E PA?

Before I added a L matching network, the input impedance of PA is 25.4682+j0.039. So I use this website
(Impedance Matching Network Designer)
to calculate the L and C values for the L matching network. I got 50-ohm port + capacitor of 7.21pF in parallel + inductance of 5.86nH in series + PA circuit in Cadence Spectre schematics. Then after running PSS+PSP again, the measured input impedance doesn"t change much as 28.1109+6.29j. Don't know where went wrong.

The calculation for L matching network from the website I said above works well for impedance matching by using SP analysis in Cadence Spectre. Don't know why the calculation of the website doesn't work for PSP analysis.

Show me netlist regarding analysis statements and signal source definitions.

I think Ls should be 9.17nH for 433MHz.

If you add Cp and Ls in front of gate of MOSFET, nonlinear operation point of MOSFET will be affected.
So an intrinsic input impedance value can be never same as 25.4682+j0.039.
But your case is very suspicious.

If you show me full netlist including device model, I can resimulate instead of you.