Analysis of startup, shutdown and middle sequences, Campaign 31 Abstract: The behavior of the detector arrays in Campaign 31 follows previously established trends, with SL and LL showing further evidence of slow but steady deterioration in noise performance compared to previous campaigns. SL exceeded its SUR read-noise limit 11 of 14 times; LL exceeded its limit 9 times, and LH 14 times. SH passed all tests. The behavior of LH is somewhat troubling, as it clearly grew noisier as the campaign proceeded. Analysts: Lei Hao & Greg Sloan Procedure version: 2.7 Campaign 31 , Startup, shutdown and middle anneal YEAR 2006 DOY 130 CE1 used for all data collection AORKEY = 17227008 for startup AORKEY = 17225216 for the first mid-campaign anneal sequence AORKEY = 17225728 for the second mid-campaign anneal sequence AORKEY = 17225984 for the third mid-campaign anneal sequence AORKEY = 17226240 for the forth mid-campaign anneal sequence AORKEY = 17226496 for the fifth mid-campaign anneal sequence AORKEY = 17226752 for shutdown Number of files = 64 for each No known errors. Steps 2 through 5 run OK. Unruly.pro results for startup (IRS_001): Noise characteristics of the modules (in DN) SUR data SRS data noise pre post noise pre post Module limit anneal anneal limit anneal anneal SL 8.8 10.3 9.1 15.0 11.9 12.0 SH 8.8 7.2 7.1 15.0 12.4 12.6 LL 14.5 18.2 14.4 15.0 12.4 12.8 LH 12.0 11.0 11.3 15.0 12.5 12.7 Unruly.pro results for the_first_middle (IRS_002a): Noise characteristics of the modules (in DN) SUR data SRS data noise pre post noise pre post Module limit anneal anneal limit anneal anneal SL 8.8 8.6 11.1 15.0 12.3 12.3 SH 8.8 7.9 8.5 15.0 12.9 12.8 LL 14.5 15.8 15.1 15.0 13.0 13.0 LH 12.0 12.0 11.6 15.0 13.0 12.8 Unruly.pro results for the_second_middle (IRS_002b): Noise characteristics of the modules (in DN) SUR data SRS data noise pre post noise pre post Module limit anneal anneal limit anneal anneal SL 8.8 8.2 9.1 15.0 12.2 12.1 SH 8.8 7.4 7.5 15.0 12.9 12.9 LL 14.5 15.7 14.1 15.0 13.2 13.0 LH 12.0 12.6 11.8 15.0 13.0 12.8 Unruly.pro results for the_third_middle (IRS_002c): Noise characteristics of the modules (in DN) SUR data SRS data noise pre post noise pre post Module limit anneal anneal limit anneal anneal SL 8.8 9.1 9.4 15.0 12.6 12.2 SH 8.8 7.3 6.9 15.0 12.9 12.8 LL 14.5 16.0 13.9 15.0 13.0 12.9 LH 12.0 12.8 11.7 15.0 12.8 12.8 Unruly.pro results for the_forth_middle (IRS_002d): Noise characteristics of the modules (in DN) SUR data SRS data noise pre post noise pre post Module limit anneal anneal limit anneal anneal SL 8.8 8.7 9.3 15.0 12.2 12.2 SH 8.8 7.6 7.2 15.0 12.9 12.8 LL 14.5 15.4 14.3 15.0 12.9 12.9 LH 12.0 12.3 12.2 15.0 12.9 12.9 Unruly.pro results for the_fifth_middle (IRS_002e): Noise characteristics of the modules (in DN) SUR data SRS data noise pre post noise pre post Module limit anneal anneal limit anneal anneal SL 8.8 9.1 9.3 15.0 12.3 12.3 SH 8.8 7.4 8.2 15.0 12.8 12.9 LL 14.5 17.2 15.7 15.0 13.2 13.1 LH 12.0 12.9 12.5 15.0 13.0 13.0 Unruly.pro results for shutdown (IRS_003): Noise characteristics of the modules (in DN) SUR data SRS data noise pre post noise pre post Module limit anneal anneal limit anneal anneal SL 8.8 11.2 12.5 15.0 12.4 12.4 SH 8.8 7.8 7.9 15.0 12.9 13.0 LL 14.5 16.4 14.0 15.0 13.1 13.1 LH 12.0 14.0 14.0 15.0 12.9 12.9 The analysis applied the PIRAT procedure unruly.pro to FITS files generated by MIPL. For SUR data, the quoted values are the standard deviation of the read noise map. For SRS data, these are the median of the standard deviation map. Comments: The SRS results are as usual and are all within the limits. The stability is also good throughout the campaign (see IRS_001/rn.ps, second page). The SUR results are slightly worse than in previous campaigns. Once again, SL and LL exceed their limits more often that not, with 11 and 9 failures, respectively, out of 14 tests. SL failed every pre-anneal test, and LL failed every post-anneal test. SH is within the limit in all SUR tests, but LH failed 8 of 14 times. More troubling, the LH SUR read noise appears to be increasing through the campaign (see IRS_0001/rn.ps for an illustration of this effect). Changes in read noise between anneals and from startup to shutdown: SUR data SRS data 002c-pre/ 003-pre/ 003-pre/ 002c-pre/ 003-pre/ 003-pre/ Module 001-post 002c-post 001-post 001-post 002c-post 001-post SL 0.99 1.20 1.23 1.05 1.01 1.03 SH 1.02 1.13 1.09 1.02 1.01 1.02 LL 1.11 1.18 1.14 1.01 1.02 1.03 LH 1.14 1.19 1.24 1.01 1.01 1.02 Comments: The modules showed more variability during this campaign than usual, perhaps as a result of its length. The variation in the SRS tests is up to 5%. Typical variations would normally be 2% or less. Again, this might be a result of a longer campaign. The SUR comparisons have two cases over the yellow limit of 20%, SL and LH, in both cases the 003/001 comparison, with variations of 23% and 24%, respectively. The variation in LH gets worse as the campaign proceeds. Module 002c-pre to 001-post 003-pre to 002c-post 003-pre to 001-post yellow red yellow red yellow red SL 1 0 17 2 18 2 SH 0 0 0 0 0 0 LL 5018 8 5079 5 5821 14 LH 5 2 7 2 6 3 Comments: The SH results are still good. LH is also good, except for the usual bad pixels. The orders can be seen in all of the LH comparisons. Latent images appear in the SL blue PU field, especially in the 003/002c and 003/001 comparisons, forcing a few pixels over the limit. Pixels over the red limit: (coordinates) [times] SL: (108,28) [2], (99,37) [2] LL: (120,1) [2], (123,3) [1], (52,4) [2], (3,94) [2], (47,105) [1], (54,38) [1], (76,64) [1], (85,76)[1], (87,99) [1], (23,119) [1] (The 14 red pixels in 003/001 are not recorded.) LH: (8,84) [3], (15,99) [1], (11,72) [1], (84,114) [2]