New RxH3 looking good!

Until September 2001 the holography system at the JCMT (RxH2) operated at 94 GHz. It consisted of a source fixed to the UKIRT building and a receiver in the JCMT Cassegrain cabin. It did not measure phase directly with a reference signal (as a conventional holography system does) but "recovered" the phase from the recorded amplitude maps made at two or more different focus settings. This technique is known as "Fresnel-region phase-retrieval holography". Although put together as a quick solution to allow dish measurements at he JCMT, it has worked remarkably well for many years. However the limitations of the old system became a significant factor to achieve a more accurate surface.

A new holography system for the JCMT (RxH3) to replace RxH2 has been constructed at MRAO/RAL/JAC and became operational in December 2001. Main features of the new system are:

• Map-to-map repeatabitily of 5 micron rms
• Much improved handling of systematic errors
• Ability to make a measurement in 10-30 minutes (not reached yet)

The new system was a significant improvement over the existing holography system which yielded about 12 micron RMS repeatability, was affected by systematic errors, and which required 2 hours for a dish measurement. In order to eliminate problems caused by spurious paths and standing wave patterns (the membrane in front of the dish has been identified as a significant source of problems), the frequency can be modulated over up to 10 frequencies separated by 5 MHz at a rate of 1 to 10 msec. The new holography setup is thus capable of measurering the dish at many frequencies simultaneously. Operating at the higher frequency of 160 GHz will also result in much improved resolution. Thermal properties of the telescope will also be easier to characterize with the much faster data-taking. With the expected improvement in accuracy, surface measurement errors should be reduced to a minor component of the overall surface error budget.

The new system incorporates the following specific features:

• A 2-channel source located inside the UKIRT dome and a 4-channel receiver in the JCMT receiver cabin.
• Two frequency operation at 80.3 and 160.6 GHz.
• Full-phase measuring system (Main and reference signals at 2 frequencies).
• Frequency stepping to eliminate the effects from spurious signal paths.
• Backend with simultaneous high/low gain ranges and cross-correlators.
• Real-time VME based data acquisition system.
• Time-stamped data samples for better accuracy.

The new system was scheduled to be fully commissioned and operational at the telescope by the end of 1999, but problems with the data acquisition software prevented the system from being fully commissioned. First maps have been obtained in August 2001, in cooperation with Richard Hills and John Richer (MRAO, Cambridge). A phase-map of the dish is shown here. The reduction software is being developed as new data are coming in. For some time there were problems with the stability of the Gunns and they were out of lock when the temperature in the receiver reached too high values, preventing new observations. This problem has been solved in December 2001 and we now routinely obtain maps at 80 and 160 GHz (at 6, 24, or more different frequencies) in order to study the effect of standing waves on the maps, and find the best combination of frequencies which cancel them in an average map. Some results are shown in the phase maps at 80 GHz and 160 GHz. The 160 GHz map shows a large amount of details. The rings in some of the panels are caused by refraction around the secondary, which effect was not yet fully corrected. The software now derives adjuster moves files from the maps, and on 4 November the surface was adjusted for the first time using RxH3 and its reduction software. The rms improved from about 35 micron to about 25 micron.

Note that the new holography system should have sufficient sensitivity and resolution to actually measure the scalloping of individual panels and will hopefully provide us with enough data to start work on this problem.

Poster paper on RxH3 at URSI General Assembly (August 2002).
Newsletter article about RxH3 (March 2002).
picture of RxH3 and micro before the JCMT shutdown in July 2002.