Problem with simulation of OpAmp oscillator with LC-circuit.

[Kirill.K]

Hello!



I am trying to simulate some garmonic self-excited oscillators using LC-circuits and OpAmp. From the theoretical point of view, both circuits should work as follows: small oscillations of the noise resistor initially excite the circuit, and then the amplitude increases, due to the nonlinearity of the operational amplifier, the gain becomes equal to unity and the circuit reaches the steady-state oscillation mode. When I use a parallel LC circuit, this is exactly what happens. But when using serial LC circuit, when the amplitude of the oscillations reaches about 2-3 times the power supply voltage of the amplifier, a sharp attenuation of the oscillations occurs. I understand that the increase in the amplitude of the oscillations can be simply limited in one way or another, but I would like to understand why does the oscillator go into stationary mode in one case, but in the other it does not?

I'm attaching some screenshots of both oscillogramms and two models: a parallel LC circuit with no problem and a serial LC circuit with the problem I described.

[VA.KI]

Hello,

I wonder what TMAX you have set?

There is a guide and recommendations on the web.

I hadn't checked the files yet.

[Kirill.K]

I set recomended value of TMAX, but I still have sharp reduction with strange waveform in case of series circuit and steady-state oscillation mode in case of parallel circuit.There are screenshots of oscillograms after setting recomenned TMAX and models.

[VA.KI]

Hello,

What I see now I think connected not with the simulation itself but most probably relates to the schematic issue. I saw that generated sinusoidal signal changes its frequency over time and at some point it cancels itself. This has to be proved yet but I'm sure the issue comes from the schematic, not the simulation.