In my oscillator design, the output amplitude looks as expected, until I add the AD8061 to the simulation. Granted, the AD8061 model I obtained is in Beta, but the behavior I'm seeing is not just signal at its output, but how in interferes with the behavior of the circuit at its input, even as it is isolated from this circuitry via high valued resistors. If I use an ideal op-amp model in its place, it works as expected.
the expected output at "out" is shown below when the circuit is simulated with an ideal op-amp (AOL=100k, GBW=200MHz):
However, if replace U3 with the [beta] model for the AD8061, the signal at "OUT" looks like this:
What could the model (even if buggy) possibly be doing to the signal at 'OUT" which is isolated from the device by >50kohms? Am I safe in assuming this is a simulation artifact and not real?