Elerion:
And it works with 100 kohm base resistor, 24 ohm emitter resistor and 12 ohm feedback resistor.
A little distorted on the bottom, but pretty decent.
And obviously (?) it works if feeding a larger resistor load.
Of course, with larger resistor load (without external load current) it works just fine, because it calculated for R_l=1kOhm, and normalized resistance of dissipation in oscillator has similar value.
Elerion:
Where did you take all those rules from? From experience?
These rules are result of multiple oscillator simulations data analyzing.
Elerion:
This circuit is not exactly the typical colpitts oscillator seen in manuals.
It should be so, because I was trying to resolve several problems together:
1. Define optimal inductive (capacitive) reactance of tank components.
2. Define optimal values ratio of capacitors in tank.
3. Increase efficiency, by value of emitter resistor decreasing.
4. Create distinct path for distortion control.
Elerion:
Also, this circuit it doesn't use a voltage divider at the transistor base, just a single "pull-up"
resistor.
In this circuit may be used voltage divider (see attachement).
Elerion:
Did you explore these alternatives?
When I asked "did you explore these alternatives?", I meant common collector and common emitter versions of
colpitts oscillator.
Explore for what? What are problems, need to be solved?

 

I agree.

I tried your method with several frequencies and it woks very well.
But of your last two circuits, one of them doesn't work for me (I tried three or four different transistors, but I don't have the exact model you use), and the other one is a little bit distorted (left, see the attachment).
I managed a better output, but using a single base resistor, not voltage divider (right, see the attachement).
The top traces show the inductor and tank capacitor currents. When there's distortion, currents stop being nice sine waves.

Do you think it is better to use voltage divider?

 

I agree.

I tried your method with several frequencies and it woks very well.
But of your last two circuits, one of them doesn't work for me (I tried three or four different transistors, but I don't have the exact model you use), and the other one is a little bit distorted (left, see the attachment).
I managed a better output, but using a single base resistor, not voltage divider (right, see the attachement).
The top traces show the inductor and tank capacitor currents. When there's distortion, currents stop being nice sine waves.

Do you think it is better to use voltage divider?

Both schematic are working, only change 510k to 390k and 150k to 100k for your model. Try asc. files in attachment.
Sweep temperature from -40C to +70C for circuits with voltage divider and with one base resistor and you will see what is better.

 

Both schematic are working, only change 510k to 390k and 150k to 100k for your model. Try asc. files in attachment.
Sweep temperature from -40C to +70C for circuits with voltage divider and with one base resistor and you will see what is better.

Thank you!

I did a sweep (0C to 80C, which is what matters for me) for you last voltage divider circuits and my single base resistor circuit, and I didn't see any difference. Common sense tells me that voltage divider is better than single resistor from supply voltage, because bias is more stable. This is basic electronics.
Another thing would be a collector-base feedback resistor (shunt-shunt feedback).
Here we only have a load for AC (the choke), but from DC bias point of view (there's only a 1 ohm resistor), the base resistor is not a feedback path for bias. Am I right?

 

I did a sweep (0C to 80C, which is what matters for me) for you last voltage divider circuits and my single base resistor circuit, and I didn't see any difference. Common sense tells me that voltage divider is better than single resistor from supply voltage, because bias is more stable. This is basic electronics.