What information does an eye diagram depict?

An eye diagram is a powerful visualization tool in digital communications and high-speed electronics to assess signal quality and channel performance. This FAQ will briefly explain an eye diagram and the overall and specific metrics engineers should know to understand an eye diagram.

When capturing signals from a digital bitstream such as Ethernet or USB, an oscilloscope places many symbols over each other. The result is a display showing multiple transitions, which creates a diagram in the shape of an eye. This diagram includes logic levels, such as the high and low voltage levels represented as digital “1” and digital “0,” and signal crossings where the signals transition between logic levels.

Different modulation schemes, such as the Non-Return-to-Zero (NRZ) and Pulse Amplitude Modulation (PAM) formats, can form an eye diagram. NRZ refers to two different levels with transitions between them, while the PAM refers to multiple distinct levels, with the number of eyes corresponding to the modulation order. An NRZ signal, with its two levels representing a digital 0 or 1, transmits one bit per symbol. PAM4, with four levels, transmits two bits per symbol where each of the four levels represents a different combination of bits. NRZ is PAM2 modulation, though NRZ is the more common nomenclature.

Figure 1 illustrates an NRZ and a PAM format, with a PAM4 of “4” representing the modulation order. PAM has the advantages of NRZ: higher bandwidth efficiency, increased data rates, reduced channel loss, and lower system cost. The smaller eye height of PAM4 signals makes them more susceptible to noise and jitter.

Figure 1. An illustration of a PAM4 and an NRZ eye diagram shows how PAM4 uses four levels to transmit two bits per symbol as where NRZ transmits one bit per symbol. (Image: QSFPTEK)

What are the overall quality metrics that can help analyze an eye diagram?

Two quality metrics give us first-hand information about an eye diagram: the eye opening and the eye mask. The eye opening is a clear area in the center of the diagram, with a larger opening indicating better signal quality and maintaining signal integrity. The eye mask is a defined region within the eye opening that the signal should not violate and is used for compliance testing. Figure 2 shows an opening and what an eye mask looks like. Many oscilloscopes come with preloaded eye masks based on industry standards. You can also create your own eye masks.

Figure 2. The eye-opening is the area between the high and low signals, and the eye mask is within the eye-opening. (Image: ANSATA)

The metrics used to measure the quality of an eye diagram can be classified into timing and amplitude types. The timing metrics include jitter and duty cycle distortion. Jitter refers to variations in the timing of signal transitions, and duty cycle distortion refers to the asymmetry in the time spent at the high and low levels of the signals with respect to the crossing points. Figure 3 shows the metrics jitter and duty cycle distortion. Oscilloscopes with jitter analysis also let you see jitter in the form of a histogram, which lets you see the distribution of the jitter. That can help you identify the source of the jitter. The time between the centers of the crossing points represents a unit interval.

Figure 3. Jitters can be observed as variations in the cross-over timings and duty cycle distortions as asymmetric high and low signal levels. (Image: Cadence Design Systems)

Amplitude metrics include noise and intersymbol interference (ISI). Noise refers to random variations in signal amplitude, visible as the vertical thickness of the signal traces. ISI is caused by previous bits leading to eye closure. Noise and jitter might look similar, but an FAQ focuses on their differences.

How do specific metrics aid in measuring the signal quality using an eye diagram?

Physical impairments originating from imperfect devices can commonly be found as a clipping effect, pattern effect, overshoot, eye-skew effect, insufficient extinction ratio (ER), and mismatching power, depicted in Figure 4. Let’s briefly discuss these phenomena, why they occur, and the consequences.

Clipping effect

Clipping manifests as narrowing the vertical and horizontal openings in an eye diagram. The eye becomes less open, indicating less margin for noise and jitter, caused primarily by a bias in voltage. The voltage bias, in turn, prevents the device from operating in a linear region, leading to distorted waveforms.

Pattern effect

A pattern effect occurs when the tail end of a previous bit overlaps with the current bit, causing interference. This effect is critical in high-speed digital communication systems, where it is more likely to occur and degrade overall system performance.

Overshoot

Overshoot occurs when the signal amplitude temporarily exceeds its intended final value during transitions. In an eye diagram, this phenomenon is observed at the rising or falling edges of the signal waveform due to high-speed transitions and impedance mismatch.

Eye-skew effect

Skew is the timing offset between a received signal and its reference clock caused by differences in the length of signal paths in propagation delay among multiple channels.

Insufficient ER occurs when the difference between the optical power levels of the on-state (logical 1) and the off-state (logical 0) is too small. This leads to increased bit error rates and degraded system performance. The primary source of this phenomenon is non-ideal signal shapes due to bandwidth and component limitations.

Mismatched power

The effect of mismatched power is seen as noise and closely resembles how insufficient ER leads to signal distortion. This phenomenon is attributed to transmitter aging, as reaching the rated output power is challenging, especially in laser applications.

These six specific metrics provide extensive information about the system’s performance and identify areas that deserve attention.

Summary

By analyzing the various metrics of an eye diagram, engineers can quickly assess signal quality and identify potential issues. This helps them make informed decisions about system design and optimization. The PAM format is gaining popularity because of its higher bandwidth efficiency, but the key metrics for studying the quality of signals on eye diagrams remain the same.