[evlatests] 25 MHz correlator offset poblem

[Gustaaf van Moorsel]

A week or so ago I reported that turning off selftest effectively
removes the correlator offset observed in 25MHz BW spectral line
observations.  As a result selftest is now automatically disabled
when observing in 25MHz spectral line.

I ran another three hour test this weekend to make sure that this
indeed was the case.  I ran a 25MHz spectral line observation of
an empty field, without giving any directives about turning off
selftest.  The result was successful: no offset was detected,
strongly suggesting that switching off selftest automatically
works as expected.

As an aside, just like last time, an effect reminiscent of the
correlator offset problem was seen, but disappeared after flagging
antenna 23, stokes 'RR', IF 1.  Note that the 'real' offset problem
was not antenna specific.

Gustaaf