VCO without tail current source

[xinzhizhu]

My vco has no tail current source. As we all know, the amplitude can be known and limited, Atank=Ibias*Rp. But in my case, how can I control the swing voltage without the Ibias? Is the Ibias equal to Idrain?

And if the swing voltage is large, the transistors may enter the triode region. But decreasing the swing voltage may lower the performance. So what kind of trade off should I make?

Thank you!

 

[BradtheRad]

This may be a case where bias and collector current combine to sustain tank oscillations. The tank only needs a tiny boost of energy during each oscillation to replace lost energy.

The transistor bias turns on in a range from .3 to .7 V. So as you say that may set the upper limit you can get for voltage swing. Any greater V and it overcomes the zener threshold in the transistor. (I guess that's what you mean by triode mode.)

Are you using one transistor to sustain a tank circuit? I have not seen schematics for 1-transistor tank oscillators. I've seen Clapp, Colpitts, Hartley, twin T, phase shift, etc.

Sometimes you must let it be enough for one device to perform one purpose. Therefore it may be necessary to add a second stage (second transistor) to get voltage amplification.

 

[xinzhizhu]

I use the cross-coupled structure in my circuit.(Sorry, I don't know how to upload my schematic.)

It is hard to obtain 0.6V swing voltage.The smallest swing is about 0.9V and the best phase noise achieved at 1MHz offset frequency is -97dBc/Hz for 43GHz oscillating frequency. I found that small width of transistors can decrease the current and the amplitude, but it may generate more flicker noise.

And if I want to compare 4 different VCO performances, is it correct to just compare the best FOM for each VCO without considering different oscillation frequencies?

Thank you for your answer!

 

[BradtheRad]

I believe I found a webpage showing a cross-coupled LC vco. Is this the same one you say you use?

vco-design.blogspot.com

Sorry I misunderstood the number of transistors.

Here are some facts I've observed about tank circuits. They are not immediately obvious from the usual mathematical formulas.

>>>> Large coil and small capacitor produces greater voltage swing than large cap and small coil.

>>>> Large cap and small coil use greater current than large coil and small cap.

>>>> To maximize Q, reduce ohmic resistance in tank loop. This allows more purity of interaction of inductance and capacitance. Oscillations fade less quickly.

>>>> The oscillating loop needs a certain degree of isolation. It is easily 'swamped' if there is low resistance connection to other parts of the circuit.

And if I want to compare 4 different VCO performances, is it correct to just compare the best FOM for each VCO without considering different oscillation frequencies?

Edaboard has discussions about figure of merit. A search for 'FOM' will find them.