SIPI

Calculators / 03

BER, Q-Factor and Total Jitter Calculator

BER → Q → total jitter analytical
Gaussian tail: error probability against distance from the edge

Random jitter has no hard edge — its Gaussian tail reaches every distance with some probability. A target BER picks the distance: how many standard deviations out the tail has thinned to that probability. That multiplier, Q, is why a 1 ps RMS jitter costs about 14 ps of eye at 10−12.

Reading the numbers

Every value is in the address bar as you change it, so a link reproduces the calculation exactly. Type values with SI prefixes — 100n, 2.2p, 16G — or drag the sliders; a bare number keeps the prefix already shown.

The formulas

BER = ρ · Q(QBER),    Q(x) = ½ erfc( x / √2 ) Q(x) is the Gaussian tail beyond x standard deviations. At ρ = 1 and 10⁻¹², Q_BER = 7.03.
TJ = DJ(δδ) + 2 · QBER · RJrms Dual-Dirac: deterministic jitter enters once, the random tail twice.
Go deeper — assumptions, how it is checked, and sources

Assumptions. RJ is Gaussian and DJ is well described by its dual-Dirac value DJ(δδ), which is a model-fitted quantity and is smaller than the peak-to-peak DJ a scope reports. Extrapolating a Gaussian to 10−15 assumes nothing else lives out there.

How it is checked. Q(0) = ½; Q is antisymmetric about 0; the two numerical branches of the tail agree where they meet at x = 3; the inverse round-trips; and at 10−12 the multiplier matches the published 14.069.

Source. INCITS TR-35-2004, Fibre Channel — Methodologies for Jitter and Signal Quality Specification (MJSQ).

Related