[mmaimcal]Re: [Almacal] memo 372 review reply - gurwell

[David Woody]

More comments about gain and sideband gain ratios.

The system temperature is inversely proportional to the gain
and is used as such in the OVRO mm array.  At the digitizer or
any power detector in the system Tsys is used to convert the
output into units of temperature in Kelvin.
The total power is just the system noise.  For non-solar-system
mm sources the source contribution to the total power is negligible
at mm and sub-mm, unfortunately.  If Tsys is determined on
blank sky then a correction to the Tsys must be made when
observing bright solar system objects.
The Kelvin scale is this then converted to Jansky
using the collecting area and aperture efficiency.
This calibration is directly applied to the
output of the correlators.

Thus the Tsys is just the inverse gain including the attenuation
from the atmosphere.

At mm and sub-mm wavelengths the total power into the IF
can change by a factor a few during a track from elevation
dependent sky emission.  Thus we use an ALC circuit to
keep the 2-bit digitizers at their optimal input power setting
while continuously (or nearly continuously) keeping track
of Tsys.  The above calibration procedure works very well
in this situation.  Our new system will use DSB autocorrelator
output to produce DSB Tsys for each spectral channel.

The complication comes when trying to get a priori SSB
system calibration.  The Tsys measured on individual
telescopes is inherently a DSB measurement.  At OVRO
we use astronomical observations of a continuum source
and the interferometer with its excellent sideband separating
capability to determine the gain ratios on all of the antennas
which are then used to convert the Tsys,DSB measurements
into Tsys,USB and Tsys,LSB which are then used to
convert the sideband separated correlator output into
Kelvin and then into Jansky.  The complication of different
atmospheric attenuation in the two sidebands is calculated
from an atmospheric model and included in the real time
calibration.  We don't operate near water lines and hence
only have to use the static oxygen absorption model.
This gain ratio measurement uses astronomical objects
but doesn't require knowing their flux, just that their
spectrum is reasonably flat.

This gives a reasonably good real time system calibration
and the various schemes being discussed are improvements
upon the same general procedure.  But they do involve
measuring the sideband gain ratio and using it in the
a real time calibration.

Larry suggested that you just use an astronomical source
to calibrate the system gain in the desired sideband
interferometrically.  This works great if you know the
flux of the source.  This would not make much, if any, use
of the instrument calibration or sounding that he proposes
we do.  The thermal load antenna based instrument
calibrations and atmospheric soundings can not be
applied to single sideband astronomical observations
without knowing the sideband gain ratios for all
of the antennas and the differential atmospheric
attenuation between the sidebands.

The one exception to this is the photonic calibration
scheme.  This ideally would measure the instrument
gains as a function of sky frequency separately in the
two sidebands.

********************************************
| David Woody
| Assistant Director of Instrumentation
| Owens Valley Radio Observatory
| Caltech
| P.O. Box 968, Big Pine, CA 93513
| phone 760-938-2075ext111
| dwoody at ovro.caltech.edu
********************************************
----- Original Message -----
From: "Larry D'Addario" <ldaddari at tuc.nrao.edu>
To: "Bryan Butler" <bbutler at aoc.nrao.edu>
Cc: <almacal at cv3.cv.nrao.edu>; <awootten at nrao.edu>;
<bbutler at zia.aoc.nrao.edu>; <carter at iram.fr>; <cerni at astro.iem.csic.es>;
<cwalker at zia.aoc.nrao.edu>; <dbock at astron.berkeley.edu>;
<dpw at ovro.caltech.edu>; <efomalon at nrao.edu>; <ewine at strw.leidenuniv.nl>;
<gab at gps.caltech.edu>; <guillote at iram.fr>; <jmangum at tuc.nrao.edu>;
<jsr at mrao.cam.ac.uk>; <ldaddari at tuc.nrao.edu>; <lgm at astro.umd.edu>;
<lucas at iram.fr>; <mgurwell at cfa.harvard.edu>; <mholdawa at tuc.nrao.edu>;
<mmaimcal at cv3.cv.nrao.edu>; <mwright at astron.berkeley.edu>;
<plambeck at astron.berkeley.edu>; <pnapier at zia.aoc.nrao.edu>;
<richard at mrao.cam.ac.uk>; <smyers at zia.aoc.nrao.edu>;
<wilson at physics.mcmaster.ca>
Sent: Wednesday, August 28, 2002 1:49 PM
Subject: Re: [Almacal] memo 372 review reply - gurwell


> Mark Gurwell writes:
>  > ...Only with a
>  > knowledge of the SBR can one accurately use the measured system
>  > temperature (which is inherently a DSB operation) to get the single
>  > sideband system temperature which is the first order correction
>  > to amplitude.
>
> The point is that you don't need the "measured system temperature" at
> all, but rather just the gain.  If you know the gain in the desired
> sideband, then you don't care what the gain is in the undersired
> sideband.  Using interferometric calibration with sideband
> suppression, you obtain directly the gain in the desired sideband.
>
> Your method is traditional for single dish telescopes, where the
> desired sideband gain must be inferred indirectly from separate
> measurements of the DSB gain and the gain ratio.
>
> --Larry
>