[evlatests] EVLA test meeting of September 24

[Barry Clark]

EVLA Test Meeting of September 24, 2009

1.  Overview of new bands.

E. Momjian reports that we now have a new Ku band receiver on
antenna 24.  This can easily be be evaluated with the interferometers
using the last three VLA antennas with Ku band receivers.  He
finds Tsys/efficiency of about 54 K, about as expected.  (The old
VLA systems run 170 and up.)

There are now four antennas with S band receivers.  In the best
IF (C) Tsys/efficiency runs from 45 to 60 depending on frequency.
In other IFs, some antennas behave quite differently from others.
Possible aliasing in the conversion scheme was discussed.

The new version of the L band OMT has been installed in antenna
24.  This works well down to 1 GHz as expected, with Tsys/efficiency
about 110 K, dropping to about 80 K in the rest of the band.

2.  T304 IF converter.

K. Morris reports on the response of the T304 final IF converter.
There is a pronounced rolloff to 12 GHz, of about 8 dB, mostly in
the last GHz.  This in particular affects X band, as this is fed
directly into the T304 from the frontend.  (For other bands we may
be able to use a preceding LO to avoid that part of the input.)
In the 8bit sampler output, the full 8 dB slope is seen.  This is
ugly, though it should be well within the dynamic range intrinsic
to the 8bit sampler.  In the case of the 3bit sampler outputs, the
bandpass compensation devices already in that path can handle the
slope, but this results in about 8db less output power in the high
band than in the other band.  This is probably not a problem, but
still slightly worrisome, in that the T304 input attenuator may have
to be adjusted to provide a little more power to the output stage,
slightly decreasing the dynamic range.

3.  Q band stability.

C. Chandler reports that selecting the best half dozen antennas
(based on evidence of behavior within individual short observing
runs) and looking at the gains reported by calib, she finds that
these repeat to about 5% over a six week interval.  This is a
little worse than seen at lower frequencies, but is still very
good.

4.  Observing in the presence of interference.

R. Perley reports that he tried observing in the 1612 radio astronomy
band, using a 3 MHz wide channel, nestled among the kilojansky
interference from Inmarsat, Gps, Glonas, and Iridium.  The source
observed was 3C 286 (13 Jy).  Flagging only moderate amounts of
data, he made a map with 1 mJy rms near the source, 0.25 at the
edge of the field.  There were no obvious interference related
artifacts.

5.  Mapping.

There are mildly discrepant results between maps of the calibrator
made in AIPS (V. Dhawan) and in CASA (S. Myers).  Not yet clear what
the discrepancy is due to.  Particularly in the CASA case, application
of a BLCAL corrrection results in large improvement; not yet understood
why.

When S. Myers tried to make a map of a strong, off center, source
using only the external calibration from another calibrator, the map
had a dynamic range of only a few hundred to one.  Not yet clear what
this is due to, or even if this is unexpected.