I am using HyperLynx-LineSim v8.1.1 and running FastEye analysis on the Xilinx Virtex-6 GTX eldo simulation kit.

The problem I am having is how to interpret and correlate the Statistical contours and Bathtub curves generated. I attach both plots below.

In the Statistical contours, at voltage of -100m (-0.1V), the eye is widely open from 0.4UI to 0.65UI with BER down to 10^-30.

In the Bathtub curves, the BER curve (dark blue line) corresponding -0.1V shows the best BER is worse than 10^-5 even with Zero-width Eye (the bottom point of dark blue curve) .

I thought these two plot represent the same information just in different ways. Statistical contours show the a series of horizontal cross-section views of 3-D Eye , while Bathtub curves show a series of vertical cross-section views of the same 3-D Eye. So why such huge difference? Which one to trust? Am I interpret the Bathtub curve wrongly?

Any thoughts/explanation are welcomed.

Weiming

This problem with the factor of 2 in the voltage will be fixed in the next release of HyperLynx.