Another possibility is something in the connector, or the test setup.
But 200 ps seems a lot for the connector.
IBM Rochester, MN
I have just finished measuring the intra-pair skew in seven samples of a
6-foot twin-ax cable. This cable consists of 4 twisted pairs (ie. 8
lines) and an individual shield for each pair (ie. 4 shields). The entire
bundle is enclosed by an outer shield.
In all seven samples, the measured intra-pair skew in three of the pairs
was within 50ps. The fourth pair consistently exhibits at least
150ps-200ps of skew. In these samples, it is always the <underline>same
pair</underline> that consistently measures this variation.