Noise characterization for oscillator (both ring and LC)

[amitbhaiji]

Hi all,
Rather than posting as a query here i am posting this message to invoke some discussion on the Jitter and Phase noise of the both ring oscialltor and LC tank oscillator where each one of us can share there views regarding the same.
So Can we discuss
1) How jitter or phase noise comes in picture. what all are reason behind it.
2) Why the phase noise of LC tanks oscillator is less in comparison to
the ring oscillator.
3) How to reduce the phase noise and supply sensitivity of both ring
osciialor and LC tank oscillator.
4) What is the effect of No of stage in ringoscillator on total jitter.
5) How both short term jitter and long term jitter is characterised.
6)Whats the relation between jitter and phase noise.
and lot more to discuss.
My sole idea is to share our view regarding the same.

Amit

 

[v_naren]

1) it is due to the LTV nature of the phase to amplitude transfer function of a periodic wave form.

2) it is because the output waveform is a pure sinusoid....any other oscillator that is not a pure sinusoid will have lot of phase noise.

3) I read about this here **broken link removed**

4) .....

5) ....

6) jitter has no meaning for a pure sinusoidal osc like an LC type....so for a square wave osc...more the phase noise...rather the narrower the slope of the phase noise the lesser the jitter in the square wave

 

[Mazz]

2) I don't think the reason is in the waveform...If so XO with output as pure sine or square wave should have different phase noise!
I'm not a VCO expert, but I think that the different Q and number of noise sources is the right answer.
Mazz

 

[v_naren]

2) in Ali Hajimir's thesis he has proved that LC oscillators(sine) have the lowest phse noise...because the o/p wave is a pure sine. this is true given all other things are constant when compared to a square wave oscillator.