When I do a generic search for "Colpitts oscillator circuit" I get lots of results with circuit examples that use an NPN transistor in a common emitter or common collector configuration as the means for providing the amplification/feedback necessary to sustain oscillations in an LC tank circuit. I have also seen a number of circuit examples that use a JFET instead of a BJT. However, I see almost no circuit examples that use a MOSFET.
Is there any reason MOSFETs in Colpitts oscillators seem so rare? Are JFETs and BJTs in fact used in Colpitts oscillators more frequently in practice? If so, why? What are the pros and cons of using each type of transistor in a Colpitts oscillator?
Best Answer
The most basic problem with a MOSFET in a colpitts oscillator is that the gate-source junction is much more capacitive than either a JFET or a BJT (like a hundred to a thousand times more). So, that gate-source capacitance throws a big wrench in the works.
I mean, they can work (sub 1 MHz for instance) but due to the unknown value of the gate-source capacitance, the frequency they produce is a little unpredictable and, with varying gate bias conditions, the capacitance can also vary and add further woe to the performance and stability. After all, you want an oscillator to be stable and predictable.