Start Circuit for LC Oscillator

[mazelk]

Hi,

I have recently worked on an integrated LC oscilator design, and I didn't realize about how it would start (experimentally)

Somebody who has a lot of experience on this field suggested me about adding a STARTING CIRCUIT...

I'd like to know which STARTING CIRCUIT topologies there are, and how does it work... in order to guarantee START for every turn-on.

I thank you for suggestions and advices...

 

[pfd001]

The LC oscillator can start automaticly because of noise.

 

[Miguel Gaspar]

There are two kinds of oscillators:

Autostart oscillators is one with gain more than unity and start automatically due noise, these requires a amplitud limiting circuit to prevent saturation.

Not autostart oscillators are those with gain = 1, and need a signal to start, one harmonic must be the frequency os oscillation.

 

[khouly]

integrated LC oscillator supposed to start oscilation automatically , coz when u design the oscillator u select the startup factor to be 2 or 3 , the startup factor is the ratio between the negative impedance of the active circuit and the resistance of the tank circuit

so in the beginign of the oscillation the negative resistance is 2 time larger than the tank resistance so the oscillator will start oscillation automatically


khouly

 

[mazelk]

The LC oscillator can start automaticly because of noise.

As you said early, the literacy also said that an oscillator start automaticly because of noise or gain more than unity. But, in real and realiable integrated LC oscillator, I guess at a START CIRCUIT must be applyed.

I would like to know your suggestions about these kind of circuits.

 

[elbadry]

I believe that in real LC oscillators, no startup circuits are used...

Startup circuits would be needed for stable circuits incurring +ve feedback ( such as bias circuits, current generators ). No papers on LC oscillators mention startup circuits..

In real world, the spike at supply startup can be a good source to initiate oscillation...

Regards