Peak Distortion Analysis

Peak distortion analysis (PDA) calculates the worst case bit pattern on the step responses of the channel, for both VerifEye and Quick Eye analyses. Peak distortion analysis is calculated for each combination of Eye Source and Eye Probe.

For Quick Eye analyses, the worst case bit patterns can be pushed to all Eye Sources and repeated by setting the PD_COUNT parameter in the Quick Eye analysis setup. When the PDA data is enabled, any bit patterns set in the Eye Sources are ignored. Figure 1 shows an example of the eye diagram from Quick Eye resulting from one iteration of the worst-case bits (no repeats).

Figure 1. QuickEye Result using Worst-Case Bit Pattern from Peak Distortion Analysis

For both QuickEye and VerifEye analyses, the Peak Distortion domain data (bit list, error contribution) can be reported as a standard 2D plot or in a data table. Figures 2 and 3 show the 2D plot and data table for the bit pattern above. The data table shows just a portion of the data.

Figure 2. 2D Rectangular Plot of Worst-Case Bit Pattern from Peak Distortion Analysis

Figure 3. Data Table for Worst-Case Bit Pattern from Peak Distortion Analysis

NOTE: Peak Distortion Analysis (PDA) considers any 8b10b or 64b66b encoding when generating the worst-case bit pattern.

The PROBE parameter in the PDA setup determines the target for worst-case analysis. Here is a reference diagram:

When PD_COUNT=-1, PDA calculates the worst-case bit pattern for EYESOURCE1 to EYEPROBE1 and from EYESOURCE2 to EYEPROBE2, but no bits are sent.

When PD_COUNT=0 or greater, the target PROBE must be specified. If PROBE=EYEPROBE1 in the example above, PDA calculates the worst case bit patterns from EYESOURCE1 and EYESOURCE2 to EYEPROBE1, and those bit patterns are sent to the analysis.