I am having a bit of difficulty generating the power contours at a Pin where Pout is 17dBm. I am doing a PSS analysis in Cadence Virtuoso. I am sweeping the phase along with a parametric sweep of the magnitude to generate the power contours. Initially I did a small signal match to verify all parameters it was found that 17 dBm was achieved at around Pin = 9 dBm. When I try the large signal setup, the power contours seem to give a completely unexpected output as seen in the attached image. But when I try for a different power at an arbitrarily negative value, say -30 it seems to be alright. I have attached screenshots of the power contours at Pin= -30dBm, Pin=9 dBm, and my small signal results along with the schematic. If anyone knows how to size the components for large signal with the load-pull set up in order to achieve the optimal output impedance.
best test the "horrible" looking points directly, i.e. not in a swept simulation. Best inspect the voltages in time domain, e.g. with a transient analysis. May your PA is not stable under that conditions? Also vary the input amplitude to see if the small signal gain fits to an AC or S-parameter analysis.