- Joined
- Jan 22, 2008
- Messages
- 52,513
- Helped
- 14,758
- Reputation
- 29,798
- Reaction score
- 14,128
- Trophy points
- 1,393
- Location
- Bochum, Germany
- Activity points
- 298,483
There is of course a danger of models being inaccurate. For the J110, I'm rather sure from datasheet values, that it should work. The distortions in the example circuit clearly indicate considerable surplus gain.
Oscillator simulation problems may also come from a too perfect simulator. In a real circuit, noise would assure start of oscillations, but Spice doesn't consider noise in transient analysis. Numerical limitations may have a similar effect sometimes, but not necessarily. So it's general advisable to have an additional oscillation start support. That can be simply an initial condition for a node voltage sufficient different from bias point. Or even better, a pulse current source, that would cause a damped oscillation in case of insufficient gain.
Oscillator simulation problems may also come from a too perfect simulator. In a real circuit, noise would assure start of oscillations, but Spice doesn't consider noise in transient analysis. Numerical limitations may have a similar effect sometimes, but not necessarily. So it's general advisable to have an additional oscillation start support. That can be simply an initial condition for a node voltage sufficient different from bias point. Or even better, a pulse current source, that would cause a damped oscillation in case of insufficient gain.