Date: Thu, 14 Nov 2002 12:32:23 GMT From: prn@apo.nmsu.edu (Pete Newman) Subject: 2.5m obslog 20021113 (MJD 52592) APACHE POINT OBSERVATORY SDSS 2.5M OBSERVING LOG Wednesday November 13, 2002 (MJD 52592) ---=== OBSERVING TEAM ===--- Swing: Dan Long Night: Steph Snedden, Pete Newman Support: Eric Neilsen (FNAL, by phone) ---=== OBSERVING PLAN ===--- This is the shake-up night for the end of the 2002 November run. The plan is to test items in the shake-up plan during bright moonlight then do science once dark enough. The shake-up plan can be found at: http://hello.apo.nmsu.edu/~sjnk/sdss/obsplans/ ---=== OBSERVING SUMMARY ===--- Most shake-up tests completed, then two plates finished (806, 896). IOP v3_92_0 needs more work (sorry, Eric!), which has led to several sections of this log being incomplete. ---=== OBSERVING LOG ===--- During pre-flight checks, Dan changed the imager instrument block to fix the systematic -0.01deg rotator offset seen in skippy measurements. An imaging run showed the rotator error to now be -0.02deg, so we corrected the rot_inst_ang value the other way. This has not yet been tested on sky. We ran test versions of IOP and TCC under normal spectroscopic observing conditions while the sky was too bright for science. Software, procedures and PRs tested =================================== TCC 2.6.8: Ran all suggested tests except a pointing model, as we did not mount the engineering camera. Everything worked as advertised. We suggest leaving the pointing model test until shake-down, but we are confident it will pass the test. ------ sdssProcedures v1_55, Spectrographic Procedures: Fix to broken link in specObs_summary.html back the main spectrographic procedures has been checked in to CVS, but otherwise not problems and the installed version can be used. ------ sdssProcedures v1_55, Imaging procedures: Ok. ------ PR4191: SoS emails summary at wrong time. -> Closed. Craig discovered a Linux problem with clock on sos.apo.nmsu.edu and fixed it by restarting the crond and atd system processes. ------ PR3863: Spectro guider behaving oddly. -> Needstest. This was already closed, but a recent change to the guide-camera gain made the guider efficiency meter give readings oo large. We know the change in gain was a factor of 3.5, so we have fixed the instrument magnitude zero-point by adding 1.35 to the previously-tested value of 20.8 in the guiderFlatField procedure in guiderMon.tcl and have checked in that change for formal tests in shake-down. ------ PR4214: endStare -abort did not reset spectrograph's timer. -> Feedback. We are not sure if it is behaving as Eric intended. While an exposure was underway, we issued a mechSend s: sp1> mechSend s spMechVersion v2_5_0 ... Requested_exp.time 1200.0 Exp_time_left 1148.1 Last_exp.time 139.5 Exp_state Exposing ... After the endStare -abort, the exposure remaining was 0, as hoped for, but so was the requested exposure time. Is this what you intended, Eric? The Last_exp.time values also don't appear to add up. Do they include chip flushing time, etc.? Here is status: sp1> mechSend s spMechVersion v2_5_0 ... Requested_exp.time 0.0 Exp_time_left 0.0 Last_exp.time 89.4 Exp_state None ... ------ PR4385: IOP murmur mnemonic table incomplete. -> Assigned. After restarting the murmur server with a cleanupMurmur command, we eventually saw more unknown_mnemonic messages, and confirmed they came from the test (V3_92_0) version of IOP. PR updated with the details. ------ PR2561: MJD in sdR headers incorrect. As it requires testing across the fMJD roll over at 10:20MST, we propose this test be done by day crew (e.g. Jon Brinkman). ------ After 06:15Z, we tried science observations. 06:25Z Plate 806, cartridge 1. Done in one exposure when combined with previous exposures. We continued to take further exposures as we did not have another plate available. After getting the required exposure, we tried to determine the nature or source of the "dirt" seen on guider flat-field images on cartridge 1. We first air-dusted both ends of all fibers (plate, slit-heads, guide fiber termination block) immediately before mounting the cartridge. After observing, we returned to instChange, took a gFlat (with the slitheads unlatched), air-dusted the guider termination block, took two more gFlats with the heads unlatched and latched, then air-dusted the plug plate and took another gFlat with the heads latched. The files are stored in sdsshost:/home/prn/52592/ as: guider_flat_cartridge_1_a.fits Initial gotoField guider_flat_cartridge_1_zenith.fits Pre-second-dusting guider_flat_cartridge_1_clean_term_block_unlatched.fits Dust term. block guider_flat_cartridge_1_clean_term_block_latched.fits ", latched guider_flat_cartridge_1_clean_plate_latched.fits Dust plate, latched We discern no difference between the images! We also compared the first image with one taken yesterday (in /home/prn/52591/) and also saw no difference. We surmise: 1) Air-dusting is insufficient to remove dust which is "glued" to the ends of the fibers, OR 2) The air-dusting we did was not as efficient as it could have been. It was difficult to find the guide fibers on the plate end in the dark with a flashlight, so broad sweeps with the air jet were made, OR 3) This is fixed-pattern noise in the guide fibers due to damage to individual fibers in the coherent bundles, OR 4) Some combination of the above. We wish we had a better archive of these images so that we could make more detailed comparisons, but we think some careful experimentation with guider flat-field exposures either side of alcohol-swab cleaning should tell us definitively if the "dirt" is fixed or removable. 08:19Z Switched to imaging but orographic clouds formed almost immediately! We switched back to spectroscopy. We both got error messages in logTool and logViewer apparently from the setCartridgeForLog command followijng the instrument change. See problem section #1 for details. Many items in this log are incomplete as a result. 09:57Z Plate 896, cartridge 9. Done in 2 exposures when combined with previous data. 11:14Z Plate 1059, cartridge 5. Not done. We determined from the first exposure that we stood no chance of finishing this plate tonight, so being the end of the run, we stopped at 11:55Z (about 10min before 18deg twilight). endNight finished normally. ---=== IMAGING RUN SUMMARY ===--- Run Time Stripe Lambda Last Flavor Comments Start End Begin End Frame ------------------------------------------------------------------------- 3483 08:44Z 08:59Z 100 O -106.36 -102.61 39 bias On telescope, unlatched 3484 09:05Z 09:28Z 29 N -64.22 -61.50 52 ignore Clouds Incomplete... ---=== IMAGING RUN DETAILS ===--- Run 2484 - Strip 29N - Frame Lambda Notes ----- ------ ---------------------------------------------- -help expected integer but got "cartridge %s" ftclParseArg: Invalid integer value USAGE: setCartridgeForLog -help while executing "shTclParseArg $args $opts setCartridgeForLog" (procedure "setCartridgeForLog" line 7) invoked from within "setCartridgeForLog $cartridge $plateName" invoked from within "if {[regexp loadingCartridge $matchLine]} { set plateNameIndex [expr [llength $newLine] - 1] set cartridgeIndex [expr $plateNameIndex - 1] set cart ..." (procedure "readMurmurLogLine" line 73) invoked from within "readMurmurLogLine 2" (script bound to file event - binding deleted) We then found that the focus button on logTool produced a focus line in its originating IOP window, but not in the logTool window. Issuing an explicit sc4l command in the IOP window did not help. The offsets button fails similarly. PR4416 opened. We later discovered other problems with that set of commands. s4l (spectroForLog?) fails: im> s4l UT Exp Time flavor comment =============================================== ----- sequence 16993, plate -9999 ------- 22:36Z 16993 0.0 bias {{}} ----- sequence 16994, plate 1143 ------- 00:18Z 16994 0.0 bias {{}} ----- sequence 16995, plate -9999 ------- 00:48Z 16995 0.0 bias {{}} ----- sequence 16996, plate 1143 ------- 01:07Z 16996 10.0 flat pre-flat 01:10Z 16997 2.0 arc pre-arc 01:37Z 16998 1200.1 target science Error: expected integer but got "1200.1" im> iopVersion v3_92_0 and logTool is not picking up spectro exposures. Or mirror MIG values...etc.