oscillator jitter measurement(jc , jcc)

Dear satya,

Thank you for the additional information- it is helpful in trying to answer your questions!

Objective is to measure the jitter of the oscillator. which one to choose, jc or jcc. if jcc what must be the value of k.

The jitter measurement you are interested in is dependent on your application. For example, if the clock serves to access data from a memory device, the cycle to cycle jitter may be most important (jcc). Note that jcc is not dependent on the accuracy of the clock period from any specific reference clock. It is considered a short term jitter measurement. Jc is the k cycle jitter and is a measure of the difference in time between k cycles of the waveform and a single cycle time and is expressed as standard deviation. The value of k specifies the number of cycles over which the measurement is made. It is considered a "long term" jitter measurement. The value of k you choose is a function of what your oscillator is driving. The larger values of k (1000, 10,000,...etc) provide more certainty in your estimate of jcc or jc.

If the clock serves as a reference for a frequency synthesizer, its long term and short term jitter is more important. The latter can be characterized more readily by integrating the phase noise characteristic over the specific frequency range of interest.

I hope this helps your understanding.

freq=50MHz, Oscillator freq=5GHz

from transient analysis i yget the oscillator frquency as 5GHz, but from pss and pnoise it is 4.92GHz.

The difference you observe in frequency may be due to incomplete settling of your transient simulation or an inaccurate oscillator period measurement (for example measuring the frequency over too short a time or including the settling portion of the oscillator in your measurement period).

Shawn

Dear satya,

Thank you for the additional information- it is helpful in trying to answer your questions!

Objective is to measure the jitter of the oscillator. which one to choose, jc or jcc. if jcc what must be the value of k.

The jitter measurement you are interested in is dependent on your application. For example, if the clock serves to access data from a memory device, the cycle to cycle jitter may be most important (jcc). Note that jcc is not dependent on the accuracy of the clock period from any specific reference clock. It is considered a short term jitter measurement. Jc is the k cycle jitter and is a measure of the difference in time between k cycles of the waveform and a single cycle time and is expressed as standard deviation. The value of k specifies the number of cycles over which the measurement is made. It is considered a "long term" jitter measurement. The value of k you choose is a function of what your oscillator is driving. The larger values of k (1000, 10,000,...etc) provide more certainty in your estimate of jcc or jc.

If the clock serves as a reference for a frequency synthesizer, its long term and short term jitter is more important. The latter can be characterized more readily by integrating the phase noise characteristic over the specific frequency range of interest.

I hope this helps your understanding.

freq=50MHz, Oscillator freq=5GHz

from transient analysis i yget the oscillator frquency as 5GHz, but from pss and pnoise it is 4.92GHz.

The difference you observe in frequency may be due to incomplete settling of your transient simulation or an inaccurate oscillator period measurement (for example measuring the frequency over too short a time or including the settling portion of the oscillator in your measurement period).

Shawn