Consider the source

Fiber-optic telecommunication systems are moving data worldwide at 10 Gb/s, and future systems presently in development will be operating at 40 Gb/s. Even though the information is digital in nature, the actual signals are analog. A true digital pulse signal only possesses two states, either "zero" or "one." An analog-digital pulse signal possesses many other characteristics, including amplitude, rise/falltime, over/undershoot, ringing, long-term droop, etc. To design, characterize, and troubleshoot gigabit-per-second systems, engineers and technicians eventually need to observe the actual system pulse waveforms. To make this measurement, engineers generally use a photodetector and an oscilloscope.

The most common time domain measurement for a transmission system is the eye diagram (see figure). The eye diagram is a plot of data points repetitively sampled from a pseudo-random bit sequence and displayed by an oscilloscope. The time window of observation is two data periods wide. For example, for 10 Gb/s, the period is 100 ps, and the time window is set to 200 ps. The oscilloscope sweep is triggered by every data clock pulse. An eye diagram allows the user to observe system performance on a single plot.

Eye diagrams of the same system as recorded by three different oscilloscopes show the variations in test data that can be introduced by differing instrumentation. (Picosecond Pulse Labs)

To observe every possible data combination, the oscilloscope must operate like a multiple-exposure camera. The digital oscilloscope's display persistence is set to infinite. With each clock trigger, a new waveform is measured and overlaid upon all previous measured waveforms. To enhance the interpretation of the composite image, digital oscilloscopes can assign different colors to convey information on the number of occurrences of the waveforms that occupy the same pixel on the display, a process known as color-grading. Modern digital sampling oscilloscopes include the ability to make a large number of automated measurements to fully characterize the various eye parameters.

Unfortunately, eye-diagram measurement problems exist in the marketplace with disagreements between buyers and sellers. Simply stated, the problem comes back to what the truth in the measurement is. Any eye-diagram measurement should always specify which oscilloscope was used. No oscilloscope is perfect. Consider a 10-Gb/s eye diagram measured by three different oscilloscopes (see figure). There are definitely observable differences. The pulse responses of oscilloscopes from different manufacturers and different models are all different. Minor variations even occur between different serial numbers of the same models.

All engineers are familiar with the risetime slowing effect of limited bandwidth (BW). This is summarized by the approximate relationship

T_r(10-90%) * BW(-3dB) = 0.35

where T_r is risetime and BW is bandwidth. However, even using oscilloscopes that have the same bandwidth, different measurement results will be obtained. The current state-of-the-art in oscilloscopes is 50 GHz. 50 GHz is adequate for 10-Gb/s systems, but it is woefully inadequate for measurements on 40-Gb/s systems and components. The 40-Gb/s industry needs oscilloscopes with bandwidths of at least 100 GHz.

To really know the performance of one's oscilloscope, the total pulse response, with risetime, overshoot, perturbations, and settling time—not just the bandwidth—should be calibrated and traceable to an international standard. These calibrations can be obtained from the National Physical Lab (NPL) in England.¹ With an oscilloscope calibrated by NPL, an engineer can correct for some of the oscilloscope's bandwidth limitations. This requires using a computer to process measured data and deconvolve out the oscilloscope's impulse response. This deconvolution process works extremely well for measurements on single pulses, but it is not suitable for eye-diagram waveforms, which consist of random samples of multiple waveforms. Unfortunately the oscilloscope manufacturers have not included this deconvolution capability into their instruments. oe

References

1. See www.npl.co.uk/npl/cem.