As one of the most important oscilloscope specification, jitter is a digital signal’s unwanted phase modulation, and is the combination of the oscilloscope’s noise floor and sample clock jitter.
However, oscilloscope vendors specify jitter measurement floor in various ways. To verify how an oscilloscope’s jitter is specified, users can use these simple steps.
The first step is to find a very low-jitter sine-wave source with much greater bandwidth than the oscilloscope.
The next step requires users to connect the sine wave generator to the oscilloscope’s input.
Beginning with one GHz on the sine wave generator, users should then input the sine wave into the oscilloscope.
Finally, after doing the three steps, users should then activate the time-interval error measurement of the oscilloscope and record the measurement result. It has to be noted that the clock recovery can be set by users to the constant clock recovery setting of the oscilloscope.
The above-mentioned steps correspond to the jitter measurement floor curve’s first measurement. Users should repeat Steps 2, 3 and 4 — in either 500 MHz or one GHz more than the bandwidth of the oscilloscope — to complete the curve.
Although these easy to perform steps are quite straightforward, users should bear in mind several additional issues.
First, users should note that increasing the sine wave will cause a decline in bandwidth. This is attributed to faster rise time, which results to reduced noise in the jitter measurement floor.
The noise contributes more to the oscilloscope’s jitter measurement floor rather than to its actual sample clock rate when measuring rise times that are slower than 30 ps.
Users should also consider that as the offset and amplitude of the sine wave alters, the results generated by the oscilloscope also change. Thus, an oscilloscope of a particular vendor may be very sensitive to offset changes, significantly worsening jitters when adding offset. Meanwhile, an oscilloscope from another vendor can be enhanced with 75 percent of the scale input. The jitter measurement floor can be increased by raising the scale input above 90 percent.
If there is time, users can also alter key measurement variables such as offset, amplitude and percentage of screen of input, while measuring similar jitter measurement floor curve. This exercise brings to mind another consideration – a decline in frequency causes rise time to increase, resulting to an increase in noise contribution to the oscilloscope’s jitter measurement.
Finally, users should also be aware of the measurement’s jitter being close to the jitter of the oscilloscope. When this happens, more jitter will be contributed by the oscilloscope to any measurement being made. Although users can view the oscilloscope’s lowest theoretical jitter, doing so requires the device to have a significantly lower jitter. Taking for instance an oscilloscope with a lowest jitter measurement floor of 150 fs and a device featuring 150 fs of jitter, an error of around 30 percent and 40 percent will be added to the jitter measurement. This leads the real-time oscilloscope to realize a jitter measurement of 200 fs, at best.
Leave a Reply
You must be logged in to post a comment.