Clapp oscillator AC analysis

I have this clapp oscillator which I have been working on for quite some time now (target operating frequency is 2.4GHz):

**broken link removed**


For the oscillation frequency, I've been using the approximated formula w0=1/sqrt(LC), where C is the 3 capacitors in series --- I'm aware the Clapp actually uses a varactor, but I'm using a static capacitor for simplicity.

My problem is in trying to prove that the Barkausen conditions are met. I know practically everything about my NMOS; what confuses me is the distinction between the Tank circuit and the Negative Resistance generator part.
I'd like to separate these into separate schematics and test their impedances separately.

The farthest I've gotten is taking (what I consider to be) my "feedback network" and testing impedance at the ports to test for negative resistance behavior. Port 1 would be my output voltage (of the whole circuit) and port 2 the feedback voltage.

**broken link removed**

Any suggestions on perhaps a more insightful way to approach this?

What's the matter...wrong forum?

/bump

dalarev said:

What's the matter...wrong forum?
/bump

Click to expand...

OK, I´ll try to give you some hints.
In general, you have two alternatives to analyze the CLAPP oscillator:
a) The circuit is considered as a two pole oscillator based on the negative resistance principle. In this case, there is no "Barkhausen" criterion because this condition is based on the principle mentioned in b). The oscillation condition requires that at a certain frequency a grounded positive resistance is in parallel with a negative resistance with a value which is equal or somewhat larger.

b) For my opinion, it is easier to see the CLAPP oscillator as a four-pole circuit which has to fulfill the Barkhausen criterion (loop gain equal or a bit larger than unity).
For this purpose, you have to redraw the circuit with the aim to see an active part (transistor) and a feedback circuitry.
Advice: Apply the principle of virtual ground, which in fact is nothing else than to select another reference point. To do this, you 1.) connect the source pin to ground (now you have a common source stage) and 2.) you disconnect all grounded elements from ground potential but leave them connected with each other. Now you have an amplifier and a frequency selective feedback network which can be analyzed as usual (by hand calculation or by simulation).

Further questions?

Good luck
LvW

Added after 3 hours 8 minutes:

In addition to my comments, here is a link for a corresponding article:

**broken link removed**

Hadn't seen the link at the end of your post. I will study that first and then report back, thanks.