Do these verification measurements need to be made at the AOS?
What fraction of the primary operating conditions need to be sampled for these verification tests?
Pros and Cons
Production OPT Measurements:
Same limitations as prototype OPT?
Single-Antenna Beam-Switched Radiometry:
Can FS and offset point using limb of a planet (like Jupiter).
Requires working nutator.
Can probably get about 3 mK RMS in one second with a bandwidth of 1 GHz and assuming Tsys ~ 100 K.
Single-Antenna Total-Power Radiometry:
Testing method used by AEG to check prototype antenna FS performance (used continuum mode).
For example, can do FS verification by switching to the limb of Jupiter (or the Moon) using total power.
Can be done either in spectral line (i.e. SiO masers) or continuum.
Requires good sky and receiver total power stability.
Requires more "stuff" to work.
Signal-to-Noise Considerations for Pointing Measurements:
Acceptable sky coverage possibly a limitation as quasars at least as bright as 3 Jy are required (see following bullets).
Can use dimmer sources (like quasars) than the single-antenna radiometric techniques. For example, should be possible to get less than 1 arcsec measurement errors for quasars with fluxes as low as 3 Jy. (Richard Hills)
Interferometric SNR is ~ 15 on a 10 Jy quasar with 16 ms integrations (the fastest correlator dump time for cross-correlation) in Band 3 at the OSF using 1 baseband and 1 polzn. Using all 4 basebands and both polzns would be sqrt(8) better but I would be less confident of all that working during early interferometry. Is this SNR good enough and how does it compare with switched power on a planet? (Peter Napier)
There might be an advantage in doing the planet or Moon limb tracking offset pointing, tracking, and fast switching measurements interferometrically so that atmospheric emission and Rx noise fluctuations (but not gain fluctuations) are removed. (Richard Hills)
Remember that we can in principle read out the cross-correlations at ~60Hz which is plenty fast enough.
I reckon that signal to noise is marginal on quasars so this would probably have to wait until we have short baselines available and can use planets.
Inevitably this measurement includes the atmospheric direction of arrival fluctuations as well as the antenna, although in principle we might be able to use the derivative of the WVR reading plus knowledge of the wind speed and direction to apply a correction when the geometry is right.
Longer term tracking errors can be made pretty easily and effectively with interferometric 5-points.
For offset pointing and fast switching measurements, can simply move between a bright source and a dummy position by prescribed amounts and position angles. Make interferometric pointing measurements toward the bright source position.