Phase Noise Simulation in VCO

[ashishjain_1000]

Hello,

When we do the phase noise simulation of a VCO using spectre, which node we should probe for noise?The VCO output is any way oscillating.

If we are doing VCO of freq range. like 6GHz, then do we need spectre RF or spectre is ok?

Are there separate noise models for doing phase noise simulations using spectre?

What decides the required tuning range and hence the control voltage of the VCO?

Regards,
Ashish.

 

[pfd001]

you can use spectre RF for phase noise simulation.
If the input frequency of the pll is fixed, the tuning range is dependant on the process variations. The tuning range of the VCO should cover the process variations.

 

[analog_prodigy]

Phase noise in frequency domain is same as the jitter in the time domain. So, can we use the transient analysist too

 

[amarnath]

transient does not include noise, so it does not give correct info about jitter.

 

[ashishjain_1000]

I do not have access to spectre RF.What to do?Is there any freeware software for doing this?

Added after 3 hours 15 minutes:

why noise is not there in transient simulation?WOuld it not have noise at the frequency of the transient?Noise at that freq. will corespond to the jitter in time domain, if I am correct.

 

[ashishjain_1000]

Hi PFD001,

If we have a VCO stage (not a current controlled one) where we have delay control by a voltage (saturated gain VCO Stage) (Alper Eken paper IEEE JSSC Vol 39 No 1 Jan 2004) then we do not have any control of the VCO fereq. with the process variation.Now in this case if the proces variation is larger than the tuning range, how do we make sure that the loop will work?Do we need to put some trimming options in the divider??

Regards,
Ashish.

 

[Hunter_BJUT]

u'd better use PSS simulation of SpectreRF.
Remeber to highlight "Oscillator" item ,then it prompt to select nodes. then you choose output node and leave reference node empty (default /gnd!), it is ok.

Added after 1 minutes:

correction:
PSS and Pnose combinations for phase noise.

 

[neoflash]

any one familiar simulating divider's phase noise directly?