rfLib tenplexer - power calculation

Hello all,

1. I find a strange result when using the tenplexer (or triplexer) component in combination with a PSS analysis (50 harmonics, fundamental frequency  = 15GHz): the output power at the fundamental frequency is higher than the DC power dissipated in the transistor, so the efficiency is higher than 1 ?

The voltage amplitude at the port for the fundamental termination ( v /I58/out1; pss (V), the yellow trace below)

seems too high. This is also the case with the triplexer.

The mtline component uses simulated s-parameters of a quarter-wavelength TL (in a s2p-file), the 3.5  Ohm resistor is parasitic;

the 1 uOhm resistor is added to suppress the warning that the two bottom ports of the mtline are shorted.

2. In a HB simulation, lots of warnings of the form 

Warning from spectre at time = 93.8861 ps during periodic steady state analysis `hb'.
    WARNING (CMI-2682): M0: The bulk-drain junction forward bias voltage (916.975 mV) exceeds 'VjdmFwd' ...')

occur. These warnings do not appear when the PSS-simulation (option HB) is used. The magnitudes and phases

for the tenplexer were extracted from a passive network.

With this real passive network at the drain, these warnings do not occur in HB or PSS-HB and all power calculations are normal (18.3 dBm output power, 99 mW dissipated DC power, identical drain voltage and current)

3. And less importantly: hhe output log mentions:

    WARNING (CMI-2495): Risky extrapolation of data given in RLGC file `/home/design/jjolambr/Desktop/setup_working_dir/wk/pa_thesis/klasseB_v2/realistische_kwartgolflengte_15G_500G_v5.s2p'.

But the s2p file contains the "S-parameters" in DC. However, this seems not really an issue since Ids and Vds seem ok.

Thank you for your time,

Greetings,

Joris