Power Amplifier simulation in Spectre

[zhaofeng1983]

I have a question:
IS spectre PSS simulation can accurately predict power amplifier performance?
Then how about large power amplifiers?

 

[khouly]

check this Appnote , it is very interesting
www.ecse.rpi.edu/Courses/F05/ECSE6967/optional PA HW/SpectreRF_PA533AN.pdf
and also check the HB simulation in ADS it is great in strong nonlinear circuits

 

[zhaofeng1983]

Thanks! I have this file but I doubt whether spectre will be accurate in large power amplifier design.

Then does anybody know how to use .dspf document generated by assura for spectre?

 

[totowo]

In simulation on power amplifier, HB is more powerful than PSS method......The disadvantage of pss is the low speed and "fragile" convergence...

The PSS is also believable if your simulation steps is correct enough.(e.g. the order of your highest harmonics)..........of course, The precondition is, there is no divergence.......

The difference between HB & PSS is trivial, since the error of transistor model is more notable, especially in large-signal region......

 

[khouly]

yeah , the HB is more robust and converge faster
the HB engine in ADS id very powerfull

khouly

 

[BigBoss]

The best simulation results are obtained by doing "an accurate transient simulation" and then FFT. The others are valid for "mild nonlinearities",hard nonlinear circuits can create convergence problems...

 

[pipida]

When I tried to simulate Class-E PA with large DC feed inductance, transient simulation cannot converge. The current through it cannot reach stable. HB simulation can provide quick and "accurate" results. PSS is not powerful enough compared to HB, especially when you try to feed signal of a lot of harmonics.