[evlatests] 3-bit sampler survey

[Ken Sowinski]

Vivek and I spent an hour looking at three-bit samplers for obvious
problems.  Herewith the short list.  These account for some of the
problems that have been reported by Vivek, but by no means all.

ea19C2 has a 2.5 bin DC offset when the sample multiplier is set
to the canonical 32.  Setting it to 8 produces a nicer looking
state count histogram, but that should not be necessary  We suggest
that something is wrong with this sampler.

ea22-t304-D output cables are swapped n the way to samplers.
What appears in the correlater as D1 is really D2 and vice versa.

ea25-d302 and d304 default to a sampler multiplier of 64 rather
than 32.  I will lookup Mike's explanation of how to remedy this
and take care of it.