Hi,
when I can I always prefer to plot Loop Gains by the Middlebrook method, while using the Probe FRA script.
Unfortunately, it does not always complete, often quitting without generating any error prompt ...
I was wondering whether the FRA script might find it easier to digest data if fed with already "switching period" averaged signals.
I explain myself better:
- take the example of the voltage at the output of a dc-dc converter, the result of a transient analysis will faithfully reproduce also the switching frequency ripple.
- Now, when running the set of FRA individual simulations, if the sinusoidal modulation injected in series into the closed loop has to be maintained reasonably small in amplitude ==> at the converter output the switching ripple might represents a significant level of noise superimposed onto said injected modulation.
- However, on individual plots, the ripple at the switching frequency can very easily be made to disappear by plotting (in this case the switching period is 10us)
AVG(V(out), 10us) instead of V(out)
Does someone know a way to have the FRA script processing AVG(V(out), 10us) instead of V(out) ?
Thanks,
Luca