Spurious column select actvity driven from the Atmel

[P33M]

Probing any of the column select lines, you get variations on the theme of following scope trace:

There is a 3.28uS glitch on column select lines occuring at an offset consistent with the switching of other column lines.