[evlatests] K-Band Tuning Span Restriction

[Ken Sowinski]

On Tue, 8 May 2007, Rick Perley wrote:

>    Ken explains that with the current LO setups for transition, we
> cannot in fact tune the AC and BD IFs apart by more than 2.0 GHz
> (whereas I had thought it was 4.0 GHz).  This restriction only applies
> to K and Q bands.  I leave it to him to give a more complete
> explanation.  He reports that with some coding change, the separation
> can be increased to 4 GHz -- I have encouraged him to consider this.  In
> the meantime, the 'EVLA Returns' page should have this existing
> restriction stated.

Some background.  There is a single first LO at K and Q
(and Ka) bands which has to be set to move both AC and
BD sky frequencies to somewhere in X band to present to
the downconverters.  The LO tuning scheme now in use
sets that LO to place the AC sky frequency as close to
10 GHz (the middle of X band as it can manage, and BD
follows along.  This has the consequence that BD must
lie within 2 GHz of AC.

The LO tuning scheme can be modified to move (f_AC + f_BD)/2
to 10 GHz, and then the two IF pairs can span close to 4 GHz.

We can also invoke the T303 conversion path to increase
the frequency spanned by the two IF pairs to the full BW
supplied by the receiver.  The LoIfSetup object already
knows how to do this, so making it work is an exercise in
obs2script.  However antennas ea14 and ea16 (and ea13 for
a few more days) do not have the M301 module necessary for
the conversion path to be selected.