Previous Article Next Article Table of Contents
Engineering and Technical Services: The Aladdin...(1Mar94) 1024 x 1024 InSb Array: A Project Update (from the Director's Office, NOAO Newsletter No. 37, 1 March 1994) In the December 1993 NOAO Newsletter we introduced the Aladdin project. This a progress update on that effort. The critical design review on the readout design was held in December and the final design is now complete. The readout design tapes have been submitted to Orbit, which is doing the mask fabrication and wafer processing. There have been no changes in the specifications from those shown in the December Newsletter. The detector focal plane assembly (FPA) will mount in a Yamaichi socket (IC51-1244-410), and the hybrid package is a custom 124-pin LCC made specifically for this project. The array is organized into four independent 512 x 512 quadrants. Each quadrant will be read from its corner towards the center so as to minimize any interface effects at the quadrant borders. If the various bias and clock lines for each quadrant are combined, 11 bias voltages and 10 clocks are required for operation. The array has 8 outputs per quadrant and the design goal is to read out the whole array in 50 ms. The design has both NMOS and PMOS output drivers, and the actual selection is made at the time of bonding. The NMOS drivers may have LED problems but provide for lower voltage operation while PMOS drivers are reputed to have lower noise and less tendency towards LED effects. During the readout evaluation we will carefully investigate these effects and make a determination as to which is the best approach. The addressing structure is a CMOS shift register. The slow shift register is a three phase design while the fast register is a two phase design. The clocks are buffered on chip so the clock driver does not have to drive the on-chip capacitance. Resetting is done a row at a time so faster resets will be possible when using some of the low noise readout approaches. The row enable levels, which are critical to the unit cell performance, are under user control as an external bias. This makes sure the unit cell can be operated in an optimal manner and not be subject to internal variations in the shift register. In selecting various design approaches, the philosophy was to give the user as much direct control as possible. In the past this has proven to be best and it has been fully implemented in this design. It does make for more bias levels and clocks, but the advantages far outweigh the disadvantages. After these modes have been fully explored, a later variation on the Aladdin design may eliminate or combine some features. By the next Newsletter we will be reporting on actual test data taken with the bare readout. More information can be obtained by accessing the NOAO anonymous ftp area. This can be done as follows: ftp ftp.noao.edu (or 140.252.1.24) log in as anonymous use your e-mail address as the password cd aladdin get README get any other files of interest exit with quit The README file will contain any updates and a short description of the contents of other files in the directory. If you have any questions or comments they should be sent directly to me. Al Fowler (afowler@noao.edu)
Previous Article Next Article Table of Contents