CA-forum minutes - 6 Sep 2001
agenda:
WVR (R.Sault)
Pointing (D.Rayner)
S-Band OMT update (G.Moorey)
3mm issues - held over till next forum. But see (Tony Wong's presentation)
Water Vapor Radiometer (R. Sault)
See Bob's presentation .
Two WVRs are operational - antennas CA03 and CA04. A number of problems have been identified - some
apparent non-linearity in one of the A-to-Ds; poor match between a skydip and the theoretical modelling.
However, the results are encouraging - the phase uncertainty is reduced to ~400 mircors phase.
The operation of the WVRs is being revised, and some further work is required.
Previously, a notebook communicated directly (through a PCMCIA data acquisition card) with the WVR. The notebook used the
antenna ethernet connection back to the central site. The notebook produced significant interference at L-band (It is
bolted to the L-Band receiver frame).
Suzy Jackson is finalising upgrade #1. An interface card has been built, allowing a dataset collection of the data. This
dataset will use the dataset bus on the mm receiver package. During the september shutdown interface cards will be
installed in CA03 and CA04. A fibre from the mm receiver to the pedestal room will mean that the notebook could be
placed in the pedestal room. A changeover from the PCMCIA to dataset protocol will be required. This should be completed
during the shutdown. It should be noted that this mode of operation means that the ACC cannot monitor the mm receiver
while the notebook is logging data - the fibre cannot be shared. The ACC and the notebook do share the ethernet
connection back to the central site.
Once this hadware upgrade is completed, the PCMCIA mode will not be a simple option - a reversion would require ~1 day's
work.
If upgrade #1 turns out to be unsatisfactory, then the PCMCIA mode would likely be restored.
Upgrade #2. The ACC should have control of the mm receiver dataset bus - so the logging of the WVR data will have to be
reworked. Suzy will undsertake this, but the time scale is unknown at this stage.
The main issue for the WVR is now the low-frequency fluctuations, thought to be 1/f noise in the amplifiers.
The next steps:
- The (1/f) issue needs to be addressed - GaAs instead of InP ?
- Have Astrowave build the remaining WVRs (4 ?); Possibly without the LNA; the AT will look to these.
See Dave's presentation
1. The encoder upgrade (balancing the COS and SIN channels) undertaken by Shaun James and Ravi works very well.
Ravi recommends that we continue to monitor the encoder performance - the upgrade should be stable, but we should check.
2. Dave 's work shows that there are problems in transferring pointing models between receivers (different turret stations),
different polarisations and frequencies within a given receiver band. All are a worry.
Reference pointing will take care of some of these issues.
The C/X receiver on antenna CA04 was repositioned in order to remove a substantial offset. The plan is to use
the plumb-bob whenever a receiver is installed - this ensures the correct positioning, but possibly at the expense of a
straining of the receiver assembly --- G. Moorey will investigate.
The offset between polarisations (particularly at 3cm, CA03) is possibly of the same family as the notorious OMT problem
with the S-Band feed. (see Derek McKay's investigations, caforum #9; See also C.Granet's report - RPP 3794, 1995,
B.Thomas' report, AT/39.3/053 (1995), and M. Sinclair's note AT/39.3/054).
The problem with the S-Band OMTs was investigated by B. Thomas (AT/39.3/053) and C. Granet (RPP/3794).
Their conclusion was that a redesign of the OMT fins would help. One prototype set of fins has been built and tested.
Graham recommends :
- Install the fins on a narrabri antenna during the October shutdown.
- Test. If satisfactory, build the remaining 5 sets of fins.
- Upgrade the L-Band LNAs. The natural time for this is alongside the LNAs for the Arecibo project.
- Upgrade the S-Band LNAs - this involves replacing the first transistor of the amplifier. G. Moorey will do this.
Action item on the astronomers - the S-band needs to be defined in order that the S-band system be appropriately optimised.