Well, deriv should be OK. May be a bit noisy, because it's using just the points
which came out of the simulator.
You could always write a verilog-A model to do the derivative using ddt().
Are you measuring the frequency of the VCO with Verilog-A? (the benefit of
doing it with verilog-a is that the crossing points are controlled, and so you
don't get jitter from numerical interpolation). There's a component, freq_meter,
On Tue, 2 Mar 2004 14:11:34 +0100, "Stefan Joeres"