An On-the-Fly observing mode (OTF) for heterodyne observations works
on an experimental basis at the 30m telescope. In this mode, the
telescope beam drifts continuously across the source, while data are
dumped at a rate of up to once every 1 or 2 seconds, depending on the
backends and number of frequency channels. Any scanning direction can
be chosen and successive scans can be concatenated to make a map.
Individual scans can have any length; they are preceded and/or
followed by OFF-source reference scans of any duration. This mode
ensures a better homogeneity of the data and a much more efficient use
of observing time. It is well suited for the observations of large
(larger than about 1' x 1'), fully-sampled maps of relatively strong
(typically 
1 K) emission lines. Up to 4 receivers can be used
simultaneously. The backends can be the filter banks and/or the
autocorrelators.
By now OTF observations have been carried out successfully on several programs. Our experience shows that in its current form OTF is not appropriate when weak or broad (> about 30 km/s) lines are observed and/or the best possible spectral baselines are required.
All data processing, from raw backend-counts to maps, is done off-line on an HP workstation. Because of the very large amount of data, a special setup of software and disk space is needed, and there is no automatic computation of calibrated spectra (".30m files") and no real-time display of the data during observations. We are working on improvements of the observing mode, and we are developing new versions of CAL and CLASS with specific enhancements for OTF data, but at present the software remains experimental.
Because of the special requirements of this observing mode and its present shortcomings, projects using OTF observing need to involve an astronomer of the IRAM-Granada team, who will be responsible for the technical aspects and for the data reduction.
If you want to use OTF, please note it on the proposal form ``special requirement'' line, and contact H.Ungerechts (ungerechts@iram.es) or W.Wild (wild@iram.es) at IRAM-Granada well before the proposal deadline for more information.
Frequency switching is possible with up to 3 receivers switched simultaneously. The baseline quality depends on the receiver, the tuned frequency, frequency throw, phase time, and atmosphere, and is usually better at 3 mm than at 1 mm. Certain limitations exist (maximum frequency throw of 45 km/s, backends, phase times etc.; for details see [8], or contact IRAM-Granada). Baselines may be better when using one single receiver. Observers should be prepared to use another observing mode if the baseline quality is not sufficient for their project.
An instantaneous IF bandwidth of 1 GHz is available for the 230G2 receiver. The two 1 MHz filter banks (512 MHz each) can be combined to provide 1 GHz bandwidth. The use of the 1 GHz wide filter bank excludes the simultaneous use of any other backend with the 230G2 receiver (the other receivers are not affected).
Many proposals underestimate the observing time needed to carry out the programme, even under excellent weather conditions. We ask you to pay special attention to this matter as time underestimation is now a major criterion for proposal rejection.
A handbook collecting most of the information necessary to plan 30-m telescope observations is available [10].
Finally, to help us keeping up a computerized source list, we ask you to fill in your `list of objects' as explained below.