need help self excited class c oscillator simulation and ltspice problem

A system with positive feedback and a loop gain greater than 1 at a particular frequency will sustain an oscillation at that frequency, so that once it is oscillating it will continue to oscillate. Something has to start the oscillation though. In a real circuit there will always be some noise to get it going, but in SPICE that is not necessarily the case.

You could try telling SPICE to solve for the initial operating point with the independent sources off, and then to turn the sources on at the beginning of the simulation. Whatever transient that causes will hopefully be enough to start the oscillation. You can do that with the "startup" directive (which I see that they used in some of the examples).

Other options include: replace all your DC sources with PULSE(...) sources that are off for the first few microseconds and then turn on and stay on; specify an initial condition on a capacitor or inductor in the loop; specify a smaller max timestep, one much smaller than your expected period of oscillation.

Jonathan

It seems to me that class C is problematic because the natural quiescent state is off, no current flow. Unless there is a large turn on transient or noise spike, the circuit will never transisition through its active region where gain occurs. Therefore regeneration and oscillation may never start. I think spice sees that, the initial state is off and it stays off. There is nothing to "push" the circuit into its active region. Aren't most oscillators of the types you mentioned class A or maybe class B where their initial state is somewhere in the active region of amplification? Maybe you should design your circuit where it starts class A then transitions to Class C as oscillation builds up. This can be done by using the oscillation, rectified to produce the class C bias required. Bob