PA Loadpull simulation and ac simulation

Loadpull simulation is based on the harmonic simulation, which is a large signal simulation.

I match input , output and interstage impedance of PA from the loadup simulation, and then run the HB1TonePAE simulation( Harmonic simulation) , which gives me the input/output return loss at certain given input signal, for example, 0dBm, 10dBm.

Then I run PA for ac simulation, which is a small signal simulation, why the input and output return loss( S11, S22) is quite different from that of HB1TonePAE simulation( Harmonic simulation)? Why?

Thanks,
Shanw

You compare two absolutely different things,Load Pull simulation is NOT to find Input/Output Return Losses, it's a large signal simulation and is used to find Optimum Load Impedances while the Input is-generally-conjugate matched.

AC simulation is a frequency domain task and it gives you small signal transfer function ( gain and phase) of a circuit.There is no relationship between these two because no loading/sourcing condition is defined in AC simulation.

Yes, you are right.

From the SP simulation, or ac simulation, I get a good S11 (-20dB) and S22 (-15dB) .

But I got worse Input Return Loss(-5dB) and Output Return Loss( -2dB) when I did HB1TonePAE simulation. Why the Input/Output Return Loss from here is not equal to S11/S22 from SP simulation?

Thanks.

Because s-parameters are defined as small signal scattering parameters and working under large signal conditions are absolutely different.That's why Load Pull technique has been developed in order to find the optimum impedance that gives max. Output Power or Highest Efficiency for a Power Amplifier.Otherwise s-parameters were sufficient but they aren't.

Note that s-parameters bias dependent Network parameters and they are not valid under large OP swings because of their natural definition.