change oscillator bias after startup to increase harmonics

[Terminator3]

Hello all! I realized, that it is possible to change bias after oscillator startup to make better harmonics output. Please tell me, is it good way for practical usage? Using that bias directly is impossible, unable to startup. If some related articles available please give a link

 

[BigBoss]

If your oscillator start-up is depended on biasing conditions, meaning there is a design problem.
Why don't you solve this problem first ?? Design an oscillator so that the circuit oscillate under all conditions.( Temperature,component tolerances etc).It's a real design and it will also be robust for your free of mind. In this case you will not have to change this bias.

 

[BradtheRad]

Since you wish to increase harmonics, you may be interested in class C amplifier operation. This can be made to produce a non-sine waveform. The transistor is operated in a non-linear manner.

The link below discusses the Hartley oscillator...
including something called: Automatic ‘Sliding’ Class C bias (scroll down 60 percent of the webpage):

https://www.learnabout-electronics.org/Oscillators/osc21.php

This linked article (pdf) discusses an oscillator which starts up in class A, then shifts to class C.

It also mentions a 'startup resistor', which (as far as I can tell) is only needed to apply sufficient bias until oscillations have become strong enough to create the necessary swings in bias current.

**broken link removed**

 

[tony_lth]

The original post said he wanted a better harmonic, so I guess that he want to suppress the harmonics, not to increase.

- - - Updated - - -

Hi, All, Sorry, i misunderstand BradtheRad.

 

[vfone]

When change dynamically the bias of the oscillator, the frequency stability and phase noise are affected, due to changing the impedances seen by the resonator.
Harmonics also (by definition) would be affected by these variations.
When is about a good oscillator design (including harmonic generators) I would let the circuit operate in a stable bias conditions and do any other implementations to the circuits which follows the oscillator.

 

[biff44]

yes, this is a well know phenomenon. Hewlett and Packard, in their garage, made the world's purest audio oscillator at that time by making an oscillator with a filament lamp in the gain portion. As the oscillations grew, the resistance of the filament changed--lowering the gain. As the gain lowered, the natural output of the oscillator more approximated a pure sine wave.

There are crystal oscillators, and some fet oscillators, that still use this phenomenon to yield better phase noise.