Dear All,

I am doing PSS analysis on a class-e switching amplifier that is a differential single-stage inverter.

Test bench schematic for the analysis consists of input PORT0, input transformer for providing differential signal to the amplifier
inputs, output transformer for converting the amp differential outputs into ground-based single output, and output PORT1.

The sine PORT has a frequency of 900MHz that is one-period of 1.1111ns. After PSS finished, I plotted 4 graphs.(Refer to the attached ADE figure)

1. Input/output voltage Time graphs at 50-ohms input and output
  ports. They are plotted by selecting the analysis result,
  "pss-tran.pss", that has the steady-state time data of one period.

2. Voltage-Gain(Av in dB) spectrum which covers 10 harmonics
3. spectrum of input voltage time graph obtained in 1 by taking DFT.
4. spectrum of output voltage time graph obtained in 1 by taking DFT.


My questions is

Can I get the same PSS gain spectrum as in 2 when I know the spectrums of time graphs of 3 and 4.

At my first look, the odd-harmonic gain seems to be determined by simply substracting the input fund voltage at 900MHz of 3, which is +9.76dB, from output voltages at each odd of 4, which are -4.20dB,
-25.78dB, -59.67dB, -68.15dB, and -69.34dB.

When subtracted as above-mentioned, PSS gains at odd-harmonics end up with -13.96dB, -35.51dB, -69.42dB, -77.88dB, and -79.05dB. At even harmonics it is not true.

It would be much apprecated if anybody can explain to me the relationship between PSS harmonic gains and input/output
spectrums.