Clock jitter refers to the deviations in a clock’s output transitions from their ideal positions. These deviations can either lead or lag the ideal position, as measured at a specific voltage level. Jitter is often expressed as an absolute value in picoseconds.
In a crystal-controlled oscillator, jitter is linked to random noise and has a uniform distribution. This means that any period width deviation within the specified range is equally likely. In other words, jitter follows a Gaussian distribution.
To determine the rated values, measurements are taken on an oscilloscope, and a histogram of this Gaussian distribution’s width is recorded. This is known as peak-to-peak jitter. The histogram should be based on at least 1,000 waveform cycles.
Standards like Gigabit-Ethernet, Fibre-channel, and SONET/SDH specify total system jitter in unit interval (UI) or RMS units. To obtain the RMS jitter value, divide the peak-to-peak value by 6.
It’s important to remember that the measured jitter value includes both the oscillator’s jitter and the jitter inherent in the measurement equipment. For a Digital Storage Oscilloscope (DSO), this contribution comes from sources like trigger jitter, time base stability, and delay jitter. When taking measurements, minimize the oscilloscope’s influence by setting zero delay, using a trigger signal derived from the oscillator itself (for the highest signal-to-noise ratio), and maximizing the measured waveform’s dynamic range.
Since the oscilloscope’s jitter sources are usually uncorrelated with those of the device under test, they can often be subtracted from the measured data using quadrature subtraction:
tDUT = ( tmeas² – tinstr² )½
Where:
- tDUT = jitter of the device under test
- tmeas = total measured jitter
- tinstr = jitter due to instruments
In the values provided in Suntsu Frequency Control, Inc. datasheets, this subtraction has not been performed. Many users are unaware of the test equipment’s impact, so we leave the instrument jitter in the numbers, as their measurements will likely also include it.
Cycle-to-Cycle and Period Jitter:
- Cycle-to-Cycle Jitter: The deviation between the periods of two consecutive cycles. This perspective helps detect large-displacement failures. The measurement requires capturing adjacent cycles, so more than two successive edges are needed.
- Period Jitter: A measurement of the maximum difference between non-adjacent periods within a sample. The sample size depends on the specific sample rate and capabilities of the equipment (oscilloscope) used. This can lead to variability if measurement conditions are not known or consistent. Period jitter is crucial for detecting short-cycle failures, which can be a challenging failure mechanism to identify in certain applications.
Need more information?
You can download a PDF version of these notes here, reach out to us by phone at (949) 783-7300 or through our online contact form.