|
One of the principal science instruments to be on board the Space Telescope for its first mission will be the Faint Object Spectrograph (FOS). True to its name, this instrument is designed to produce spectrographic measurements data from extremely faint astronomical objects. The photodetector utilized by FOS is a digicon offering a linear array of 512 photon-counting channels. This detector has been frequently described (ref 1,2), but most relevant here is the fact that a digicon is able to achieve photon-counting sensitivity by the utilization of a very high operating voltage, in this case 22,500 volts. No digicon operating at this voltage has yet been used on any long duration space missions. Previously described space applications have been of only short duration aboard sounding rockets (ref 3). The task of packaging such detectors for long duration space use is not to be taken lightly. Indeed, insulation failures at far lower voltages have cut short numerous missions in the early days of space flight, and even today one occasionally hears of a new example. In the present case, the inherent packaging problem posed by the 22.5 kV potential and the space environment is aggravated by the particular requirements of the FOS program and by the extreme sensitivity of the digicon and its attendent preamplifiers and counting circuits which will respond to charge impulses as small as about 5 x 10-16 coulombs. In attacking this problem, one totally false start was made. The errors were exposed by environmental testing and a second design effort was begun. This time the effort began with the fundamentals. A conceptual design phase attempted to select materials which were compatible with each other and with all of the environmental requirements. In parallel a mechanical/structural design effort developed numerous concepts of the encapsulated detector assembly and how it would interface with the rest of the FOS assembly. This effort was supported by computer-aided electric field analysis and mapping which provided identification of trouble spots in the design. Further design iterations sought to alleviate high stresses. Following this conceptual design phase was the development phase which produced detailed engineering drawings and procedures, and validated same through building and testing various non-flight detectors. While the design proved to be satisfactory in terms of insulation, minute surface discharges in view of the photo-cathode were an unacceptable source of noise on the order of 0.01 counts per second per channel. This was cured by an application to the affected surfaces of Cr203 which served to prevent the build-up of static charges. The resulting dark count was about 0.001 count per second per channel at -100C. This value is to be expected for the FOS trialkali photocathode.
|
