Re: colpitts Oscillator
**broken link removed**
There are various ways to design such an oscillator. Many people use "rule-of-thumb" and a bit of trial and error. Lately people follow a more scientific approach by using CAD type open loop analysis as a first pass and following up with a non linear analysis and optimization. They look at loop gain, matching, loaded Q and negative resistances. Accurate component models are needed for this, either from the vendor or by doing your own measurements. Finally performance will be compared by measuring the performance of the actual hardware on test equipment.
It is not difficult to get some (lousy) oscillator design working, but it takes quite a bit more effort to optimize such a design for stability, phase noise etc.