What limits the offsetting accuracy of the 4-m Mayall telescope? Tests during the past several years have shown that even guided offsets were sometimes off by 2-3" or more over 10' , and that guided offsets were often LESS accurate than unguided offsets. This has compromised spectroscopic observations of faint point sources, and it has long been apparent that the culprit was the limited precision and accuracy of the encoding of the rotator angle. Although nominally precise to 0.1 degrees, "reality checks" on the sky have shown that the actual accuracy of the encoder was 0.2-0.3 degrees. (It is easy to align the rotator to 90 degrees using the telescope and guider motion as a fiducial; the reproducibility of the zero-point has been no better than this.) This is wholly consistent with the offsetting problem.
Our engineers Scott Bulau, Tony Abraham, David Stultz, and Shelby Gott have long had a design for an upgrade to this sytem, but limited resources combined with only once-a-year access to the rotator have prevented its implementation until this summer, despite the on-going impact to science at the 4-m. We are glad to announce, though, that the upgrade has been built and bench-tested, and will be installed during the present summer shutdown. On-sky tests will begin as we bring the telescope out of shutdown at the beginning of September. The design goal is an offset accuracy of 0.3" over 10', consistent with the demands of observing with the R-C Spectrograph or CryoCam. This requires an encoder accuracy of 0.03 degrees; the new system should exceed this by roughly an order of magnitude. We expect that our closed-loop offsetting will be limited only by the mechanical accuracy of the X-Y stage of the guide probe; our tests will reveal the actual offsetting accuracy of the new system.
Phil Massey, 4-m Telescope Scientist