EldoRF analysis on divider

eldorf

I'm using EldoRF on frequency divider or large signal amplifier.

The purpose is to estimate the amplifier or divider's contribution to periodic clock phase noise.

Simulation results are easily generated. However, i am wondering its accuracy. I never give the trigger point voltage to EldoRF, in which way it calculate period?

thanks,
Neoflash

thanks

Hi neoflash,

In EldoRF (or in any HB based simulators) the circuit signals are described
as a truncated Fourier serie (a finite sum of harmonically related phasors).
The phase noise represents the spectral density of the phase fluctuations
(of one of these phasors). Then the period is known (as long as you define
the harmonic at which the phase noise is computed), and there is no notion
of triggering. The phase noise is the same for all the time points as it concerns
the full period.

Concerning accuracy of the results, this will only depends on the number of harmonics
that are specified for the steady-state analysis.

Bozzo.

that's very true.

However, if send the signal to a receiver with different trigger point, what will the output jitter be?

Will that be the same for different trigger points?

Hi neoflash,

There is a big difference between phase noise and time jitter.
Both are two interpretations of a common phenomenon, they have the same origin.

I agree that jitter is related to a time event whereas phase noise is not.

The big question is the following: is there a direct relation between phase noise
and jitter ?

For autonomous circuits there exist in EldoRF fonctions that compute period_jitter
(a scalar term) and long_term_jitter (a waveform vs time).

For forced circuits the answer is not so obvious.
There exists in SpectreRF a fonction that provides jitter and that is equivalent
to the EldoRF long_term_jitter but I think
it is not adapted because they are not related to a triggered time.

There exist in litterature numerous papers dealing with the subject
and I have experienced the different formulas
(with Matlab) and none of them lead to correct results
(getting the jitter info from a phase noise spectrum).

Then it is still an open subject...

Finally there exist in Eldo an analysis called NOISETRAN (which is a transient noise
analysis) that could provide you some interesting results (it is not going through
phase noise and is fully related to time events).

I hope it helps.

Bozzo.

Why you think autonomous circuits are unique?

Doesn't it require information of trigger point?

Hi,
I am new designer in RFIC and I would like to learn using Mentor graphics tools for designing an Rf Mixer. I have used ADS simulator for such example and I would like now to use Design Architect, Eldo RF.
Can any body help me to find tutorials about Da_ic and Eldo RF.
Thinks