The work done at NSLS bearalines is a segment of a complete PX project that often is minor in time, but crucial in its importance. Typical experimenters come to BNL only for brief periods, say 3-4 days, and rarely more than 2-3 times each year. During visits, they are expected to keep the apparatus running around the clock, and they try to use their precious beam-time allocation efficiently. There are several consequences of these visits being short, intense, and widely spaced, and of these measurements taking a short time in the PX process. One is that the experimenters rarely become completely familiar with the apparatus and the software that drives it. Secondly, they often are working with little sleep under stressful conditions. Finally, a team of three or four experimenters often will include one or two who are not really experienced crystallographers. They simply are not able to make all the decisions that might need to be made during data collection. A major goal of those of us who operate these facilities is continually to improve the interface between the user and the instruments. This interface includes both the physical presentation of the apparatus and the appearance of the software that controls it. There are several aspects to this work. Firstly, one must meet continuing advances in the quality of the experimental hardware. Of course computers must be continually upgraded, but the measurement apparatus itself seems to have an obsolescence lifetime of about three years. When apparatus is replaced, the control mechanisms and software also must be upgraded or replaced, along with the documentation and training program. We take very seriously the role that good instrument-control and data-reduction software can play in making use of the facilities reliable and as easy as possible. To that end, we employ a full-time programmer (John Skinner). This allows us easily to respond to new experimental situations (new equipment or software) and also to reply to users' suggestions or complaints. The beamline software we present to our users now has several aspects of the experiment integrated into one package, is fairly easy for the users to learn to use, and is filled with checks to prevent mis-steps and to guide the user in making measurements.
Showing the most recent 10 out of 167 publications
